Cabeza, Roberto - Nyberg, Lars - Park, Denise C-Cognitive Neuroscience of Aging Linking Cognitive and Cerebral Aging-Oxford University Press (2017)

i
COGNITIVE NEUROSCIENCE
OF AGING
ii
iii
COGNITIVE NEUROSCIENCE
OF AGING
Linking Cognitive and Cerebral Aging

SECOND EDITION
Edited by
Roberto Cabeza
Lars Nyberg
Denise C. Park
1
iv
1
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.
Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.
© Oxford University Press 2017
First Edition published in 2005

Second Edition published in 2017
All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.
You must not circulate this work in any other form

and you must impose this same condition on any acquirer.
Library of Congress Cataloging-in-Publication Data

Names: Cabeza, Roberto, editor. | Nyberg, Lars, 1966– editor. | Park, Denise C., editor.
Title: Cognitive neuroscience of aging : linking cognitive and cerebral aging /
edited by Roberto Cabeza, Lars Nyberg, Denise C. Park.
Description: Second edition. | New York, NY : Oxford University Press, 2017. | Includes index.
Identifiers: LCCN 2016030377 (print) | LCCN 2016031903 (ebook) |
ISBN 9780199372935 (hardcover : alk. paper) | ISBN 9780190660222 (UPDF) |
ISBN 9780190660239 (EPUB)
Subjects: LCSH: Brain—Aging. | Cognitive neuroscience.
Classification: LCC QP356.25 .C64 2017 (print) | LCC QP356.25 (ebook) | DDC 612.8—dc23
LC record available at https://lccn.loc.gov/2016030377
9 8 7 6 5 4 3 2 1
Printed by Sheridan Books, Inc., United States of America
â•‡ v
Contents
Contributors â•… vii
Introduction â•… 1
I Methods andÂ€Issues
1 MRI Measures of Aging: Methodological Issues â•… 9

Hanzhang Lu and Peiying Liu
2 Molecular Imaging of Aging and Neurodegenerative Disease â•… 35

Anna Rieckmann, Randy L. Buckner, and Trey Hedden
3 Age Differences in Structural Connectivity: Diffusion Tensor Imaging

and White Matter Hyperintensities â•… 71
David J. Madden and Emily L. Parks
4 Age Differences in Functional Connectivity at Rest and During Cognitive Tasks â•… 105
Cheryl L. Grady
5 Multiâ•‰modal Imaging of the Aging Brain â•… 131

Anders M. Fjell and Kristine B. Walhovd
6 Structural and Functional Imaging of Aging: Longitudinal Studies â•… 155

Lars Nyberg, Sara Pudas, and Anders Lundquist
7 Interpreting Age-â•‰Related Differences in Memory-â•‰Related Neural Activity â•… 183

Michael D. Rugg
vi
vi Contents
II Cognitive Processes
8 Selective Attention and Inhibitory Control in the Aging Brain 207

Theodore P. Zanto and Adam Gazzaley
9 Working Memory and Executive Functions In the Aging Brain 235

Patricia A. Reuter-Lorenz and Cindy Lustig
10 Neural Correlates of Age-Related Slowing 259

Timothy A. Salthouse
11 The Aging Hippocampus: Linking Animal and Human Research 273

Shauna M. Stark and Craig E. L. Stark
12 Episodic Memory Encoding and Retrieval in the Aging Brain 301

Wei-Chun Wang and Roberto Cabeza
13 Emotion and Emotional Memory 337

Elizabeth A. Kensinger and Jaclyn H. Ford
III Health and Disease
14 The Middle-Aged Brain: A Cognitive Neuroscience Perspective 363

Denise C. Park and Sara B. Festini
15 The Modifying Role of Hypertension in Cognitive and Brain Aging 389

Karen M. Rodrigue and Gerard N. Bischof
16 Genetics and Cognitive Neuroscience of Aging 415

Goren Papenberg, Ulman Lindenberger, and Lars Bäckman
17 Effects of Exercise on Cognition, Brain Structure, and Brain Function

in Older Adults 439
Kirk I. Erickson and Lauren E. Oberlin
18 The Link of Intellectual Engagement to Cognitive and Brain Aging 461

Martin Lövdén, Lars Bäckman, and Ulman Lindenberger
19 The Challenges of Disambiguating Preclinical Alzheimer’s Disease

from Cognitive Aging 485
Reisa Sperling
20 Late-Life Depression: Translating Neurobiological Hypotheses

into Novel Treatments 507
George S. Alexopoulos and Robert E. Kelly
Index 529
vii
Contributors
George S. Alexopoulos Roberto Cabeza

Weill-Cornell Institute of Geriatric Center for Cognitive Neuroscience
Psychiatry Duke University
White Plains, NY, USA Durham, NC, USA
Lars Bäckman Kirk I. Erickson

Aging Research Center Department of Psychology
Karolinska Institutet and Stockholm University of Pittsburgh
University Pittsburgh, PA, USA
Stockholm, Sweden
Sara B. Festini
Gerard N. Bischof School of Behavioral and Brain Sciences
School of Behavioral and Brain Sciences The University of Texas at Dallas
The University of Texas at Dallas Dallas, TX, USA
Dallas, TX, USA
Anders M. Fjell
Randy L. Buckner Department of Psychology
Department of Psychology and Center University of Oslo
for Brain Sciences Blindern, Oslo, Norway
Harvard University
Cambridge, MA, USA
vii
viii
viii Contributors
Jaclyn H. Ford Anders Lundquist

Department of Psychology Department of Statistics
Boston College Umeå University
Boston, MA, USA Umeå, Sweden
Adam Gazzaley Cindy Lustig

Center for Integrative Neuroscience Department of Psychology
University of California, San Francisco University of Michigan
San Francisco, CA, USA Ann Arbor, MI, USA
Cheryl L. Grady David J. Madden

Rotman Research Institute at Baycrest Brain Imaging and Analysis Center
Toronto, Ontario, Canada Duke University Medical Center
Durham, NC, USA
Trey Hedden
Department of Radiology Lars Nyberg
Harvard Medical School Departments of Radiation Science
Cambridge, MA, USA Umeå University
Umeå, Sweden
Robert E. Kelly
Weill-Cornell Medical College Lauren E. Oberlin
White Plains, NY, USA Department of Psychology
University of Pittsburgh
Elizabeth A. Kensinger Pittsburgh, PA, USA
Department of Psychology
Boston College Goren Papenberg
Boston, MA, USA Aging Research Center
Karolinska Institutet
Ulman Lindenberger Gävlegatan, Stockholm, Sweden
Center for Lifespan Psychology
Max Planck Institute for Human Denise C. Park
Development Center for Vital Longevity
Berlin, Germany The University of Texas at Dallas
Dallas, TX, USA
Peiying Liu
Advanced Imaging Research Center Emily L. Parks
UT Southwestern Medical Center Brain Imaging and Analysis Center
Dallas, TX, USA Duke University Medical Center
Durham, NC, USA
Martin Lövdén
Aging Research Center Sara Pudas
Karolinska Institutet Integrative Medical Biology
Gävlegatan, Stockholm, Sweden Umeå University
Umeå, Sweden
Hanzhang Lu
Department of Radiology Patricia A. Reuter-Lorenz
Johns Hopkins University School of Department of Psychology
Medicine University of Michigan
Baltimore, MD, USA Ann Arbor, MI, USA
ix
Contributors ix
Anna Rieckmann Craig E. L. Stark

Department of Radiology Department of Neurobiology & Behavior
Massachusetts Geneneral Hospital University of California, Irvine
Charlestown, MA, USA Irvine, CA, USA
Karen M. Rodrigue Shauna M. Stark

School of Behavioral and Brain Sciences Department of Neurobiology & Behavior
The University of Texas at Dallas University of California, Irvine
Dallas, TX, USA Irvine, CA, USA
Michael D. Rugg Kristine B. Walhovd

Center for Vital Longevity Department of Psychology
University of Texas at Dallas University of Oslo
Dallas, TX, USA Blindern, Oslo, Norway
Timothy A. Salthouse Wei-Chun Wang

Department of Psychology Center for Cognitive Neuroscience
University of Virginia Duke University
Charlottesville, VA, USA Durham, NC, USA
Reisa Sperling Theodore P. Zanto

Center for Alzheimer Research Center for Integrative Neuroscience
and Treatment University of California,
Harvard Medical School San Francisco
Cambridge, MA, USA San Francisco, CA, USA
x
1
(A)
Neurovascular
coupling
Metabolites
MRI
• Blood volume ↑ scan
Stimulus • Blood flow ↑
• Blood oxygenation ↑
Neuronal Vessel size
activity fMRI
Astrocytes BOLD
signal
Neurotransmitters
(B) (C) 3
BOLD signal (%)

3s 2
4–17s
+ 3s 1
4–17s
+ 3s 0
+ –1
–2
0 100 200 300
time Time (s)
Figure 1.4 Illustration of BOLD fMRI. (A) Illustration of the neurovascular coupling pathway that leads to the observa-
tion of fMRI signal change due to a stimulus. (B) Example of task-fMRI stimulus paradigm. (C) Corresponding BOLD
signal in visual cortex from an fMRI scan using the paradigm in (b). Red bars indicate stimulus periods.
Control Time delay Acquisition

Control
Control-Label
Label Time delay Acquisition

Label
Figure 1.6 The principle of ASL MRI. In the control scan, the magnetization of the blood is unchanged; whereas in the
label scan, the magnetization of the blood is inverted when passing through the labeling plane. Note that the RF pulse train
is still played out during the control scan, but it is essentially a zero-degree RF pulse. The purpose of this is to equate the
magnetization transfer effects between the label and control images. After a time delay which allows the labeled blood to
arrive at the imaging slice, a control image and a labeled image are acquired. The subtraction of control and labeled images
can cancel the static tissue signal and the resulting difference image provides an estimation of CBF to the brain.
2
(A) Glucose metabolism (18F-FDG) (B) Amyloid burden (11C-PIB) (C) Tau burden (18F-T807)
Normal metabolism Regional hypometabolism Low amyloid High amyloid Low tau High tau
SUVR DVR DVR
0 1.2 0 1.5 0 2.0
Figure 2.1 A) Example images for FDG-PET standardized uptake values (SUVR) for two clinically healthy older adults (age > 70). The person on the left shows typical cortical FDG
signal, the person on the right has evidence of temporo-parietal hypometabolism. B and C) Example images for clinically healthy older adults showing evidence for low and high amyloid
burden as measured with 11C-PIB (B) and tau burden as measured with 18F-T807 (C). PET signal for all tracers is standardized with respect to the cerebellum.
D1 receptor (11C-SCH22390) D2 receptor (11C-Raclopride) Dopamine transporter (11C-Altropane)
Young Old Young Old Young Old

DVR DVR DVR
0.5 2.5 0.5 5.0 0.5 5.0
Figure 2.4 Example PET images for three common PET ligands of the dopamine system. An example image for a young person (20-30 years) and a clinically normal older adult (>
65 years) is shown for each ligand. Loss of striatal signal for the old person can be seen for all ligands. Images are voxelwise Distribution Volume Ratio images (DVR, Logan et al.
1990) with reference region cerebellum.
3
(A) (B) (C) λ1
λ2
λ3
(D) (E)
FA = 0.01
FA = 0.8
(F) (G)
A B C
White Matter Voxels
Diffusion Tensors
Figure 3.1 Contributions to tract anisotropy. (A) Water diffuses more easily along the axis of a fiber bundle than it does
across the axis of the bundle, due to the presence of barriers such as membranes and myelin. (B) Typically, multiple dif-
ferent diffusion-weighted images are acquired, with each one sensitized to diffusion along a different direction in space.
(C) One can fit a mathematical model to the measurements in order to estimate certain model parameters that describe
diffusion behavior within each voxel. The most commonly used model, the diffusion tensor model, fits the measurements
to a tensor, or ellipsoid, which is fully characterized by its three orthogonal eigenvectors and their associated lengths, or
eigenvalues (k1, k2, k3). (D) In cerebral spinal fluid (CSF), water diffuses freely in all directions and so FA is close to zero;
in white matter, diffusion is directionally dependent and so FA is closer to one. (E) The long axis of the diffusion tensor
corresponds to the principal diffusion direction. Within a coherent fiber bundle this aligns with the fiber direction. (F) By
following these voxel-wise estimates of principal diffusion directions it is possible to perform diffusion tractography, and
reconstruct estimates of fiber pathways. (G) Variations in diffusion parameters along tracts during normative development
are likely a combination of tract-specific (e.g. myelin content, axonal characteristics) and local environment contributions.
Voxel 1 contains a tract of interest (yellow) as well as a crossing tract (gray), resulting in low anisotropy measurements at
this point. Voxel 2 contains only the tract of interest and exhibits high anisotropy. Within voxel 3 axons from nearby gray
matter join the tract and some axons break off heading toward gray matter targets. The result would be a drop in anisotropy
measurements at this point in the tract. The figure is reproduced with permission from (Walhovd et al., 2014).
(A) (B) (C) (D)
Figure 3.2 Diffusion tensor imaging (DTI) tractography. (A) Seed and target regions. (B) Estimated fibers in the
genu of the corpus callosum. (C) Estimated stream tube. (D) Estimated stream tube with age group differences in
fractional anisotropy (FA) color coded; warm colors representing age-related decline. Modified and reproduced
with permission from Davis et al. (2009).
4
Fractional anisotropy (FA) Axial (principal) diffusion (AD) Radial diffusion (RD) Mean diffusion (MD)
4 4 –4 –4
2 2 –2 –2
z-score
0 0 0 0
–2 2 2 –2
–4 4 4 –4
–6 6 6 –6
5 25 45 65 85 5 25 45 65 85 5 25 45 65 85 5 25 45 65 85
Years Years Years Years
Age-effects on (FA) White matter volume Cortical myelin content Cortical volume
4 4 4
2 2 2
0 0 0
–2 –2 –2
–4 –4 –4
–6 –6 –6
Quadratic age-effect 5 25 45 65 85 5 25 45 65 85 5 25 45 65 85
No quadratic age-effect Years Years Years
Figure 3.3 Multimodal imaging of white matter through the lifespan. Results are based on 430 well-screened healthy participants between 8 and 85 years (mean 41.6 years). Values in the scatter-
plots are expressed in z-scores (standard deviations) to ease comparison between metrics. Values represent for FA, axial, radial and mean diffusion the mean of all voxels that were included in the
left superior longitudinal fasciculus. The tract-based spatial statistics skeleton represents the middle of the tract for all participants (red and green voxels in the lower left brain image). White matter
volume represents the total volume of all cerebral white matter, and cortical volume represents the volume of all cortical gray matter, in both cases corrected for total intracranial volume. Cortical
myelin content is based on the ratio between T1-and T2-weighted MR images in an overlapping sample (n = 339, age 8–83 years), sampled 0.2 mm from the white matter/gray matter boundary into
the gray matter in the superior frontal cortex. Data modified from (Grydeland et al., 2013) and reproduced with permission from (Walhovd et al., 2014).
5
A
R L
P
FA 0.6
0.5
FA
0.4
25 50 75 100 125
Milliseconds
AD 1350
1300
1250
AD
1200
1150
1100
25 50 75 100 125
Milliseconds
RD
750
700
650
RD
600
550
500
450
25 50 75 100 125
Milliseconds
MD 900
850
MD
800
750
700
25 50 75 100 125
Milliseconds
Young (< 52 years) Elderly (≥ 52 years)
Figure 3.5 Age-related differences in the relation between reaction time standard deviation (sdRT) and measures of
microstructural integrity. Voxels showing a significant age x sdRT interaction on diffusion characteristics are displayed
(age, sex, mRT, and sdRT were used as covariates). The effects are corrected for multiple comparisons across space by
threshold-free cluster enhancement at p < 0.05. The results are smoothed to ease visualization of effects, and displayed
on top of the WM skeleton (red on green skeleton, FA; copper on green skeleton, AD; blue on green skeleton, RD; green
on red skeleton, MD). Right, Scatterplots illustrating the relationship between sdRT in milliseconds (x-axis) and diffusion
characteristics (y-axis) across all voxels showing a significant relationship between sdRT and diffusion. As can be seen,
weak relationships exist in the younger half of the sample (age < 52 years), while stronger relationships are seen in the
older half (age ≥ 52 years). Reaction time data drawn from the less-demanding (congruent) condition of the Eriksen flanker
task. The more attentionally demanding (incongruent) condition yielded a similar, though spatially less extensive, pattern.
Figure reproduced with permission from Fjell et al. (2011).
6
(A)
Younger Middle Older
L L L
(B)
L L L
1 19
Figure 3.6 White matter lesion (WML) volume. Panel A: White matter lesions for individual adults 20, 48, and 65 years
of age, in T2-weighted FLAIR images. Participants were healthy, community-dwelling individuals without any sign
of cognitive impairment on neuropsychological testing or history of cardiovascular disease (other than hypertension).
Lesions, as identified from a semi-automated program separating lesions from normal white matter appear in red. Panel
B: Voxelwise lesion maps for 23 younger adults (19-39 years of age), 19 middle-aged adults (40-59 years of age), and 16
older adults (60-79 years of age). Color scale represents the number of individuals within each group exhibiting a lesion,
per voxel, overlaid on a T1-weighted template. Authors’ data.
vPCC
Resting functional connectivity

MTL 0.5 * *
0.4
Correlation
0.3
0.2 Young
Older
0.1
0
vPCC MTL dPCC
dPCC
Figure 4.1 Age differences in FC within DMN subsystems are shown. The images on the left show the patterns of FC
with the ventral PCC (vPCC subsystem), the left parahippocampal gyrus (MTL subsystem) and the dorsal PCC (dPCC
subsystem) that were identified in both age groups. The graph at the right shows that the strength of seed correlation with
these FC patterns differs with age. The mean correlation (across a resting state run) between seed activity and activity in
the relevant brain regions for younger and older adults is plotted. Age differences are indicated by asterisks. Data are from
Campbell et al (2013).
7
(A)
(B) 1.00
0.80
0.60
0.40
Correlations
0.20
Older
–0.00
Young
–0.20
–0.40
–0.60
–0.80
–1.00
RFG LFG medOFC RT ACC
Figure 4.3 (A) The colored brain regions represent a set of areas active in young and older adults during a face-matching task. (B) The graph shows the correlations between activity in the
regions in (a) and activity in right fusiform, left fusiform, medial orbitofrontal cortex (medOFC). Correlations between activity in the regions seen in (a) and behaviour also are shown (reaction
time, RT; accuracy, acc). Young adults show correlations between right and left fusiform and the regions seen in (a), but not with medOFC or behaviour. In contrast, older adults show reliable
positive correlations between the pattern of activity, right fusiform activity, and accuracy on the task, indicating age differences in FC as well as how the FC pattern relates to performance. Error
bars denote 95% confidence intervals for the correlations. Data are from Burianova et al (2013).
8
Left hippocampus Right hippocampus

0.20 0.25
0.15 0.20
0.10 0.15
0.05 0.10
0 0.05
–0.05 0
–0.10 –0.05
20 30 40 50 60 70 80 20 30 40 50 60 70 80
Age Age
Figure 6.4 Cross-sectional estimates of age-related decline in hippocampal activity during episodic memory encoding.
The sample comprised 292 individuals aged 20-80 years, divided into 8 age groups. The blue clusters show the main effect
of encoding across the sample, whereas the age-effect is shown in red. Reproduced with data from Salami et al., 2012,
Journal of Neuroscience.
(A)
(B) (C)
0.6
0.2
Beta change 2003–2008
0 0.4
–0.2
Beta values
0.2
–0.4
0
–0.6
–0.2
–0.8
–0.4
45 50 55 60 65 70 75 80 85 45 50 55 60 65 70 75 80 85
Age Age
Figure 6.6 Discrepancy between longitudinal (panel A) and cross-sectional (panel B) age-effects on activity in the right dorso-
lateral frontal cortex. Reproduced with permission from Nyberg et al., 2010, Proceedings of the National Academy of Sciences
of the United States of America.
9
z=8 z = 24 z = 32
Age decrease Age increase
(A) (B)
30 20
20 10
Brain score
Brain score
10 0
0
–10
–10
–20
–20
–30
10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90
Age (years) Age (years)
Encoding tasks Recognition tasks
Figure 6.7 Scatter plots indicate significant cross-sectional age-related increases in a brain network comprising medial
frontal gyrus, precuneus, posterior cingulate cortex, cuneus and middle occipital gyrus (yellow/orange clusters), as indi-
cated by higher brain scores. Conversely, younger adults had higher activity in a network comprising middle and inferior
frontal gyrus, caudate nucleus, putamen and fusiform gyrus (blue brain clusters), as indicated by lower brain scores. The
changes are common to various task conditions during memory encoding and retrieval. Reprinted by permission of MIT
Press Journals, from Grady et al., (2006), Journal of Cognitive Neuroscience.
10
(A)
(B)
P < 0.05 P < 0.001
Figure 8.3 Gray-matter volume. Age-related declines in gray-matter volume are associated with (A) increased distracti-
bility and, (B) deficient neural suppression of sensory cortex activity in older participants. Previously published (Chadick
et al., 2014).
(A) Anterior (B)
Genu of corpus
callosum
Superior longitudinal
fasciculus
Uncinate
Right fasciculus Left
Splenium of corpus
callosum
Posterior
z = 20 z = 40
P = 0.05 P = 0.01
Figure 8.4 White-matter integrity. Age-related declines in white-matter integrity (fractional anisotropy) are associated
with (A) increased distractibility and, (B) deficient neural suppression of sensory cortex activity in older participants.
Previously published (Chadick et al., 2014).
11
Figure 12.3 Increase in nSMEs as a function of age in several default mode network regions. Reprinted with permission
from Park H, Kennedy KM, Rodrigue KM, Hebrank A, Park DC (2013) An fMRI study of episodic encoding across the
lifespan: Changes in subsequent memory effects are evident by middle-age. Neuropsychologia 51: 448-56.
(A) L DLPFC R DLPFC (B) (C) L DLPFC
L IPL RSC HPC
MTL conn. HPC conn. Amygdala conn.
YAs > OAs OAs > YAs
Figure 12.4 PASA-like effect in functional connectivity. (A) OAs exhibited reduced MTL connectivity with parietal and ret-
rosplenial cortices, but increased connectivity with dorsolateral PFC. Adapted with permission from Daselaar SM, Fleck MS,
Dobbins IG, Madden DJ, Cabeza R (2006) Effects of healthy aging on hippocampal and rhinal memory functions: An event-
related fMRI study. Cereb Cortex 16: 1771-1782. (B) OAs displayed reduced HPC connectivity with OTP regions, but increased
HPC connectivity with PFC regions. Adapted with permission from Dennis NA, Hayes SM, Prince SE, Madden DJ, Huettel SA,
Cabeza R (2008) Effects of aging on the neural correlates of successful item and source memory encoding. J Exp Psychol Learn
34: 791-808. (C) OAs showed reduced amygdala connectivity with HPC, but increased connectivity with dorsolateral PFC.
Adapted with permission from St. Jacques PL, Dolcos F, Cabeza R (2009) Effects of aging on functional connectivity of the
amygdala during subsequent memory for negative pictures: A network analysis of fMRI data. Psychol Sci 20: 74-84.
12
Medial PFC Amygdala Parahippocampus
Lateral PFC Hippocampus Ventral visual processing areas
Figure 13.2 Neuroanatomy of Emotional Memory. The key regions implicated in the formation and retrieval of emotional memories.
13
(B) A IncLg > ConLg B 2.3 3.2
R Z = 22 mm X = –52 mm
Figure 15.2 B) Increased activation to incongruent trials vs. congruent trials in the left middle temporal, left inferior
parietal, and left supramarginal gyri was seen as vascular risk score increased.
(A) (B) 0.8

70
0.6
Cognitive performance
Hippocampal activity
60 0.4
50 0.2
0
40
–0.2
30
Val/Val –0.4 Val/Val
20 Any Met Any Met
–0.6
70 75 80 85 90 95 100 105 20 40 60 80
Age (in years) Age (in years)
(C) (D)
0.90
6.2
Fractional anisotropy
5.8 0.80
5.4
5.0 0.70
4.6
Val/Val
0.60
4.2 Any Met
3.8 0.00
< 65 ≥ 65 < 65 ≥ 65
0 40 50 60 70 80 90
Val/Val Any Met Age (in years)
Age (in years)
Figure 16.3 Effects of BDNF on (a) longitudinal decline in perceptual speed across 13 years, with steeper decline for
BDNF Met carriers. Perceptual speed is measured using the digit-letter task, which required participants to name letters
associated with a digit, according to a template. The y-axis indicates total number of correct responses after 3 min. Adapted
from Ghisletta et al. (2014). Interaction between age and BDNF, reflecting (b) lower hippocampal activity during retrieval
of episodic memories, (c) smaller hippocampal volumes, and (d) lower white-matter integrity in the splenium for older
BDNF Met carriers. Hippocampal activity in (b) indicates parameter estimates of the BOLD response measured in arbi-
trary units in left hippocampus, which is greater during retrieval relative to a baseline condition. White-matter integrity is
indicated by fractional anisotropy. Adapted from Sambataro et al. (2010), Sanchez et al. (2011), and Kennedy et al. (2009),
respectively.
14
(A) (B)
(C) (D)
Figure 16.4 KIBRA genotype groups show different correlations between increasing age and performance on (A) imme-
diate and (B) 30-minute delayed recall of a story, as measured with the Wechsler Memory Scale. (c, d) KIBRA genotype
group differences in the correlation between age and brain activation during an episodic memory task. (C) The KIBRA CC
group (red) exhibits a negative correlation between age and activity in left hippocampus during encoding, which is not
observed for T allele carriers (blue). (D) The KIBRA CC group (red) exhibits a negative correlation between age and activ-
ity in right hippocampus during retrieval, which is not observed for T allele carriers (blue). Hippocampal activity indicates
parameter estimates of the BOLD response measured in arbitrary units, which is greater during encoding and retrieval
relative to a baseline condition. Adapted from Muse et al. (2014).
15
LEFT HIPPOCAMPUS RIGHT HIPPOCAMPUS
(A) Hippocampus 5.2 5.2
5.1 5.1
5 5
4.9 4.9
4.8 4.8
4.7 4.7
4.6 4.6
Baseline 6 months 1 year Baseline 6 months 1 year
(B) Caudate nucleus RIGHT CAUDATE NUCLEUS

LEFT CAUDATE NUCLEUS
4.9 5.3
4.8 5.2
4.7 5.1
4.6 5
4.5 4.9
4.4 4.8
(C) Thalamus THALAMUS
15.00 Exercise
Stretching
14.50
14.00
13.50
13.00
Baseline 6 months 1 year
Figure 17.1 Results from a 12 month exercise intervention in which 120 older adults were randomized to either a walking exercise condition or to a stretching control condition. The results demonstrated
that there were no significant changes in the size of either the thalamus or caudate nucleus with the intervention, but there were significant increases in the size of the hippocampus for the walking exercise
group. Adapted from Erickson et al. (2011).
16
Figure 19.2 The neuropathology of Alzheimer’s disease. A microscopic section from an AD patient autopsy demonstrat-
ing extracellular aggregates of amyloid-beta in a neuritic plaque and intraneuronal aggregates of tau in neurofibrillary
tangles.
Amyloid and Tau PET imaging

PiB Aβ T807 Tau
Clinically
Normal
Aβ–neg
Clinically
Normal
Aβ–pos
AD
Dementia
Aβ–pos
Figure 19.3 PET Amyloid and Tau imaging. Coronal PET images superimposed on structural Magnetic Resonance) of
PiB Aβ (left) and T807 Tau (right) acquired on 3 participants in the Harvard Aging Brain Study. The top row is a clinically
normal older individual with low PiB retention and minimal T807 binding in the medial temporal lobe (MTL). The middle
row shows a clinically normal older individual with elevated PiB retention and T807 binding extending beyond the MTL
into inferior temporal neocortex. The bottom row shows images from an AD dementia patient with extensive PiB and T807
binding in the neocortex. Images courtesy of Dr. Keith Johnson.
1
Introduction
A s we noted in the first edition of this book ten years ago, although cog-
nitive neuroscience of aging has a long past, only lately has it achieved
the critical mass to be considered an autonomous discipline. In the past, the cognitive
and neural mechanisms of age-related changes in cognition were typically studied
independently of each other. Studies in the domain of cognitive psychology of aging
investigated the effects of aging on behavioral measures of cognition and character-
ized a variety of age-related deficits in memory, attention, etc. In parallel, studies in
the domain of neuroscience of aging investigated the effects of aging on the anat-
omy and physiology of the brain. Even if most scientists agree that cognitive aging
is caused by cerebral aging, the relationships between these two phenomena were
largely unknown. Fortunately, the situation changed with the growing popularity of
studies focusing on the relationships between the effects of aging on cognition and
on the brain. This group of studies originated the new discipline we called cognitive
neuroscience of aging, which we attempted to characterize in the first edition of this
book published in 2004.
Since 2004, cognitive neuroscience of aging has continued to expand at a very
rapid pace. In addition to the multipication of aging studies using fMRI, which has
reached thousands of articles, new methods that were less common ten years ago,
such as diffusion tensor imaging (DTI), have markedly increased in popularity. To
illustrate the growth of the discipline, Figure Introduction.1 shows the number of
articles related to fMRI of aging and cognition since 1998 (the last two years are
not included because databases are still incomplete for these years). Although the
database search used for this figure is inexact, it is fair to say that the number of
articles in cognitive neuroscience of aging published per year has tripled druing the
last ten years. The growth is even more dramatic when looking at citations, which
1
2
2 Cognitive Neuroscience of Aging
Number of papers Number of citations for

per year these papers
16,000
350
14,000
300
12,000
250 10,000
200 8,000
150 6,000
100 4,000
50 2,000
0 0
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Figure Introduction. 1. Publications related to fMRI of aging. The search was neither com-
prehensive or selective. The Web of Science search was (topic: “functional magnetic resonance
imaging” or DTI) AND (topic: aging or elderly or older) NOT (topic: infant or child or adoles-
cent) AND (perception or attention or memory or executive).
show an exponential function. Thus, our discipline is not only alive and well but it
is even more vibrant that it was ten years ago.
The second edition has a similar oganization to the first edition, with a first
section focused on methods, a second section aimed at cognitive processes, and
a third section covering clinical and health-related issues. Many of the chapter
authors in this second edition were also contributors to the first, but there are
many new authors including several young researchers. We hope this second edi-
tion will be as successful as the first, and will also be used by both experts and
by students of the discipline. Below, we briefly describe the chapters in the three
sections of the book.
Section on Methods and Issues
In our previous volume, we had a section on imaging measures, with focuses on struc-
tural MRI, dopamine imaging with PET, electrophysiological and optical measures in
studies of cognitive aging, and BOLD fMRI and its relation to cognitive changes in
aging. Methods development continues to be a driving force behind new insights into
the cognitive neuroscience of aging. We see several new trends.
One such trend might be labeled expansion of the scope of imaging methods. For
example, today it is common in PET studies of normal aging to examine amyloid and
tau burden, something that a few years ago was more common practice in studies of
dementia. In MRI research, more and more studies focus on issues such as perfusion
and iron accumulation, which clearly is of relevance for the aging brain. Another
3
Cognitive Neuroscience of Aging 3
trend is “multimodal imaging.” The combination of different MRI measures, or MRI

and PET markers, is today quite common. It is a powerful approach, where the sum
can be greater than the parts. A third trend is a shift away from cross-sectional com-
parisons and toward longitudinal, within-person studies. All of these trends are dis-
cussed in some detail in the chapters in the new volume.
Lu and Liu discuss methodological issues related to MRI measures of aging. They
examine the ever-relevant topic of the basic principles of the MRI signal, such as its
origin and how studies could best be designed to ensure a good quality of the images.
They also present commonly used MRI techniques in studies of cerebral aging, as
well as emerging techniques such as perfusion and iron imaging.
Rieckmann, Buckner, and Hedden contribute a chapter on molecular imaging of
aging and neurodegenerative disease, with emphasis on the most common molecular
PET targets such as glucose metabolism, amyloid burden, tau, and dopamine. They
stress the importance of integrating PET with other neuroscience markers to obtain
multiple markers of the brain basis of cognitive aging, and point to future develop-
ments related to hybrid PET-MR studies.
The chapter by Madden and Parks is focused on the microstructural integrity of
cerebral white matter, in particular as measured by diffusion tensor imaging and
white-matter hyperintensities. They consider the fundamental issue of the relation-
ship between the MRI measures of white-matter integrity and behavioral measures
as well as functional connectivity. They also point to the value of imaging of iron
accumulation.
Grady’s chapter is concerned with functional connectivity at rest and during active
task processing. While resting-state connectivity has been a topic of study for some
20 years, we have witnessed an increase in the number of resting-state connectivity
studies the last few years. This is true in aging studies as well, and commonly studied
networks are discussed in the chapter along with different methodologies to capture
connectivity. Here, too, the issue of the relation between the imaging measures of con-
nectivity and cognitive measures is vital.
Fjell and Walhovd explicitly focused their chapter on multimodal imaging of the
aging brain, and they address the central question of how to best make use of this
approach. A key topic here concerns the possibility that multimodal imaging will con-
tribute to an explanation of increased amounts of aging-related variance in measures
of interest. They also discuss different ways of analyzing multimodal imaging data in
brain-aging studies.
Nyberg, Pudas, and Lundquist contribute a chapter on longitudinal structural
and functional brain imaging. Reasons for (and against) a longitudinal approach
are discussed, and findings from cross-sectional versus longitudinal studies are
contrasted. A main focus is on methodological issues in longitudinal imaging,
including attrition, changes in hardware, and choice of appropriate statistical
models.
Finally, the chapter by Rugg also compares cross-sectional and longitudinal stud-
ies, using functional neuroimaging of episodic memory as a case study. The chapter
considers the various limitations of cross-sectional designs, but also identifies several
important points one can draw from cross-sectional evidence.
4
Taken together, the chapters in the Methods and Issues section of the new volume
contribute knowledge of the opportunities and limitations of current methods, as well
as highlighting important avenues for future methods development.
Section on Cognitive Processes
Having reviewed the various methods used in the cognitive neuroscience of aging in the
first section, the second section of the book turns toward the evidence generated using these
methods in different cognitive domains. In particular, the chapters cover five fundamental
domains: processing speed, attention, working memory, episodic memory, and emotion.
A reduction in processing speed is the most reliable finding in the cognitive aging litera-
ture and it affects all cognitive abilities. Surprisingly, only recently has age-related slow-
ing become the focus of neuroimaging studies. Age-related deficits in attention, working
memory, and episodic memory are also consistently found across cognitive aging studies.
Whether these deficits reflect an underlying insufficiency in shared processing resources
or whether they reflect more specialized mechanisms is still a topic of debate, and func-
tional neuroimaging has contributed valuable evidence regarding this question. The effects
of aging on emotional abilities are subtle, but they modulate many of the effects of aging
on memory abilities.
The chapter on the neural mechanisms of age-related slowing contributed by
Salthouse reviews studies linking speed measures to different neuroimaging mea-
sures, including white-matter integrity and grey-matter volume. The chapter also
considers fundamental questions regarding the interpretation of age effects on
neuroanatomical measures and linking them to age-related changes in cognitive
measures.
Zanto and Gazzaley’s chapter tackles the important issue of how age-related brain
changes affect the neural mechanisms of attention and inhibitory control. Their chapter
reviews selective attention studies in spatial, feature, object, and internal attention domains,
and covers both fMRI and EEG studies. These methods have complementary strength,
with fMRI providing better spatial resolution and EEG better temporal resolution.
Reuter-Lorenz and Lustig contribute a chapter on the neural mechanisms of work-
ing memory and executive function in older adults. Their chapter emphasizes three
issues that became central in this literature since the publication of our previous edi-
tion, ten years ago: differences between maintenance and manipulation, the interac-
tion between attention and interference control deficits, and differential age effects on
verbal vs. visuo-spatial memory. They also discuss the interpretation of age-related
activity differences and the concept of compensation.
The chapter by Stark and Stark and the chapter by Wang and Cabeza focus on
the age-related changes on the neural bases of episodic memory. These two chapters
complement each other: the former goes deeper into age effects on the hippocampus,
covering both human and animal studies, and the latter covers the whole episodic
memory network in human studies. Stark and Stark describe a cross-species model of
the aging hippocampus, and review the effects of aging on hippocampal anatomy and
physiology in animal models and humans.
5
Cognitive Neuroscience of Aging 5
Wang and Cabeza distinguish between two episodic memory processes, asso-
ciative memory (including recollection) and item memory (including familiarity),
and review fMRI studies on the effects of aging on these processes. They focus
on event-related fMRI studies that isolated success-related activity during encod-
ing and during retrieval, and discuss their findings in relation to the concept of
compensation.
Finally, Kensinger and Ford tackle the relatively small but complex functional neu-
roimaging literature on the effects of aging on emotion and emotional memory. Their
chapter covers the effects of arousal and valence, considering not only age effects on
regions associated with emotion generation, such as the amygdala, but also on regions
linked to emotion regulation, such as prefrontal areas.
Taken together, the six chapters in the Cognitive Processes section provide a thor-
ough overview of the cognitive neuroscience studies, particularly neuroimaging stud-
ies, on how aging affects the neural mechanisms of cognition.
Section on Health and Disease
The final section of this book, “Health and Disease,” addresses a broad range
of topics and is considerably expanded relative to the earlier volume, reflecting
the rich and rapid growth of knowledge in this domain. The three chapters on
this topic from the original book are impressive and foreshadow rather clearly
future issues in the cognitive neuroscience of aging. In his 2004 chapter, Randy
Buckner considers the meaning of enhanced prefrontal activation in older adults
and what constitutes a healthy aged brain. He suggests that “on average, healthy
individuals will show the least additional recruitment and the best performance,”
but also notes that that lower performing older adults who show enhanced recruit-
ment would likely perform worse without this additional compensatory activa-
tion. Currently, relating patterns of brain activity to cognitive performance is
a front and center issue in the emerging study of individual differences that is
developing as large studies come online. In the 2004 volume, Cheryl Grady sug-
gests that a new technique—measuring connectivity among brain regions—is an
important advance for understanding differences between normal cognitive aging
and adults who are developing dementia. Grady now has a chapter in the Methods
section of the present volume on connectivity that draws on the large corpus of
work that ensued on this topic since 2004. The last chapter in the 2004 volume
was written by Lars Nyberg on cognitive training and primarily makes the case
that the older brain retains plasticity, and that the focus for future research should
be on developing optimal interventions for older adults rather than focusing on
differential effects of training in the old and young. There was actually little data
to review in 2004. In this new volume, the chapter on cognitive training (Lovden,
Backman, and Lindenberger) had so many studies to choose from that the authors
note that they had to be selective in choosing only those studies that highlighted
particular issues of concern. In sum, a reading of these three original chapters
helps one see clearly how much has truly been accomplished in understanding
6
the cognitive neuroscience of health and disease, and how well these early pio-
neers understood what needed to be done and provided readers with a blueprint
for the future. The seven new chapters in this section are likely to serve a similar
function for readers.
Park and Festini have written a chapter on the middle-aged brain, suggesting
that this is a critically important area where new research is badly needed. They
review the limited extant literature on this topic and propose an agenda for future
research.
The impact of hypertension on the aging brain and cognitive function is the topic of
a chapter by Rodrigue and Bischof. They provide a thorough and insightful review of
the impact of this silent disease on brain structure, function, and cognition.
Papenberg, Lindenberger, and Backman provide a unique summary of the role of
genetics in determining the trajectory of cognitive aging. This is a great account of
what genes matter for cognition and what this burgeoning field can and cannot tell us
about the aging mind.
Erickson and Oberlin provide a terrific overview of the many studies that have
been completed to date on the beneficial effects of exercise on the aging brain. They
consider the impact on brain volume, white matter, and both task-based functional
activity and connectivity when the brain is at rest.
A new chapter on cognitive training by Lovden, Backman, and Lindenberger pres-
ents cognitive training issues in the context of a compelling meta-analysis as well as
a review of the implications of findings in the cognitive training literature for three
major theoretical models of cognitive aging.
Sperling presents an overview of the newly developed diagnosis of “preclinical
Alzheimer’s disease.” This term has arisen from results associated with new PET
imaging techniques that visualize deposits of amyloid and tau proteins in vivo. Until
the recent past, such measures were only available at autopsy.
Finally, Alexopoulos and Kelly provide an overview of how their use of state-of-
the-art imaging techniques has led to a new brain-based model of depression.
Overall, this last section of the book provides gratifying evidence of how imaging
tools are leading to a new understanding of brain–behavior relationships and provid-
ing new conceptual models of neurologic and psychiatric diseases.
7
METHODS AND ISSUES
8
9
MRI Measures of Aging

Methodological Issues
Hanzhang Lu
Peiying Liu
M agnetic resonance imaging (MRI) is one of the most widely used imag-
ing techniques in studies of cognitive neuroscience, including in brain
aging. Compared to other medical imaging modalities, MRI has three important fea-
tures that have contributed to this broad acceptance. First, MRI does not involve ion-
izing radiation and, as a matter of fact, most MRI scans do not require any exogenous
contrast agents. Therefore, it is particularly suitable for studies on healthy participants
in whom injection of exogenous tracer or agent is not desirable. For similar reasons,
one can repeat the MRI scan as frequently as needed in a follow-up or longitudinal
setting. Second, MRI provides an excellent spatial resolution, allowing the visualiza-
tion of the brain with exceptional details. It also shows clear image contrast between
soft tissues, unlike several other imaging technologies such as X-ray and computer-
tomography (CT), thus can easily distinguish the gray and white matter based on
their characteristic MR properties. Finally, MRI is also a versatile imaging modality.
Within the same imaging session, one can perform several MRI pulse sequences and
therefore obtain multiple domains of structural, functional, and physiological infor-
mation from the participant. This advantage reduces subject burden and allows the
integration of multi-parametric datasets in understanding cerebral aging. This chapter
will provide a review of the basic principles of MRI, describe several major MRI
techniques that are commonly used in cerebral aging, and introduce new, emerging
techniques that are on the horizon.
9
10
10â•… Methods andÂ€Issues
Basic Principles of MRI
The Origin of the MRI Signal

The source of MRI signal is from the endogenous nuclei (e.g., 1H, 23Na, 31P, 17O, and
19
F) that form the human body. The vast majority of the MRI studies used in cerebral
aging are based on the detection of the protons in 1H nucleus (more specifically water
protons); thus we will focus our discussion on this nuclear species.
Each proton can be viewed as a small magnet. When these small magnets are
placed in an environment without any significant external magnetic field (e.g., outside
the MRI scanner), they will be oriented randomly (Figure 1.1A) and, as a result, their
net strength, referred to as magnetization, is 0. On the other hand, when these magnets
are placed inside a strong external magnetic field (e.g., inside the MRI scanner), their
orientation is no longer random. Slightly more than half of them will orient them-
selves to be parallel to the external field, while slightly less than half will be anti-â•‰par-
allel to the external field (Figure 1.1B). Thus, the net strength of the proton magnets
is no longer 0 but is parallel to the external field (see M vector in Figure 1.1B), since
slightly more magnets are pointing along that direction.
The magnitude of the net strength, which ultimately determines the intensity of our
MRI signal, is given by:
1 C
M= ρ B0 [1]â•„
4 T
(A) Without External Magnetic Field (B) With External Magnetic Field
B0
M
FigureÂ€1.1â•‡ Illustration of the formation of MRI signal. (A) When the water protons are placed
in an environment without a significant magnetic field, they are oriented randomly, thus the
net signal (referred to as magnetization, M) is zero. (B) When the water protons are placed in
a strong magnetic field, they are aligned along the axis of the external field, B0, with a slightly
larger fraction parallel to the field (relative to anti-â•‰parallel). Thus, the net magnetization is along
the B0 direction. This net magnetization is what an MRI system measures when performing
a scan.
â•‡ 11
MRI Measures ofÂ€Agingâ•… 11
where ρ is the proton density (PD) of the issue, B0 is the strength of the external field,
T is the temperature in Kelvin, and C is a parameter related to several physics con-
stants. While mathematically simple, several observations can be made from Equation
[1]â•„. First, one can appreciate that the MRI signal is greater at a higher field strength,
B0. Second, proton density, i.e., the amount of water in the tissue, plays a major role
in MRI signal intensity. In the brain, CSF contains the highest water density (about 1g
water/â•‰ml) while the gray (0.89g water/â•‰ml) and white matter (0.73g water/â•‰ml) contain
less (Herscovitch and Raichle, 1985). Therefore, in a proton-â•‰density MRI, CSF is
expected to be brighter than the gray matter, which is in turn brighter than the white
matter. Third, the lower the temperature, the greater the MRI signal. Unfortunately,
this is not something that one can easily exploit under in vivo conditions.
While the application of the external B0 field can induce a net magnetic field in
the tissue, this field along the B0 direction cannot be detected. Therefore, a second
external field, referred to as radiofrequency (RF) magnetic field (also known as the B1
field) is needed to “excite” the tissue spins. With these steps, one can detect a signal
in the receiving coil. However, no spatial information is present yet. The spatial infor-
mation is encoded by applying a magnetic field gradient such that water protons in
different spatial locations experience different magnetic fields and therefore have dif-
ferent oscillation frequency. Once these signals are received and recorded by the coil,
an image reconstruction algorithm can then be used to obtain the final MRI images.
The signal one obtains from the final MR image is further modulated by MR prop-
erties of the tissue, in particular T1 and T2. Specifically, the MR signal received is
given by:
1 − e − TR / T1
S= M⋅ sin ⋅ (FA) ⋅ e − TE / T 2 , [2]â•„
1 − cos ( FA ) Ce − TR / T1
where FA is flip angle, TR is repetition time, and TE is echo time. In Equation [2]â•„,
the MR properties T1 and T2 should be differentiated from imaging parameters, e.g.,
repetition time (TR) and echo time (TE). The difference is that imaging parameters
can be chosen by the researcher, whereas T1 and T2 are intrinsic properties of the
tissue that the researcher has no control over. However, by properly choosing TR and
TE (and other imaging parameters for more advanced sequences), one can make the
overall image intensity weighted in a pre-â•‰defined fashion, resulting in different con-
trasts such as T1 weighted or T2 weighted MRI.
Considerations of MRI Image Quality

When evaluating the quality of a set of MRI data, three criteria can be considered.
One is to identify whether the image contains any artifact, which is represented by
the appearance of signals in unexpected areas. Figure 1.2 illustrates a brain image in
which the fat signal was insufficiently suppressed and appeared inside the brain due
to chemical shift between fat and water protons. Many reasons can cause image arti-
facts, and it is preferable to have the images reviewed by an experienced imaging sci-
entist, especially at the beginning of a project, before proceeding with a large sample
12
Image containing Image with fat

fat artifact artifact suppressed
FigureÂ€1.2â•‡ An example of brain images with and without artifacts induced by undesired fat
signal (arrows).
scanning. Another criterion is the spatial signal-â•‰to-â•‰noise ratio (SNR) of the image,
which is a useful index in evaluating the quality of structural MR images. The signal
in the spatial SNR calculation can be obtained by drawing a region-â•‰of-â•‰interest (ROI)
in a representative brain region and calculating the mean of all voxels, while the
noise can be estimated by defining an ROI outside the brain and calculating the mean
or standard deviation of the voxels. For MRI data that have multiple time points, such
as functional MRI (fMRI) or arterial-â•‰spin-â•‰labeling (ASL) MRI, a third criterion, tem-
poral SNR, is often used to assess data quality. In the calculation of temporal SNR,
the signal term is defined similar to that for spatial SNR. For the noise term, however,
it is defined as the standard deviation across the time points. Temporal SNR can be
defined on a voxel-â•‰by-â•‰voxel basis or on an ROI. The estimation of temporal SNR
is preferred in fMRI data assessment because it considers contributions from both
thermal and physiological noise, whereas spatial SNR only considers thermal noise.
Since physiological noise is known to be a major, if not the predominant, component
of the noise source in fMRI, the examination of temporal SNR is more appropriate
for determining the reliability of a functional dataset.
Magnetic Field Strength Considerations

As mentioned above in Equation [1]â•„, higher magnetic field strength usually cor-
responds to a greater sensitivity. 3T is therefore preferred over 1.5T for cognitive
aging studies. This advantage has been experimentally demonstrated for virtu-
ally all brain MRI pulse sequences (although for cardiac and body MRI, 1.5T is
sometimes preferred). As far as the magnitude of the enhancement effect is con-
cerned, it depends on specific pulse sequence and spatial resolution. Going from
1.5T to 3T, a typical SNR increase of 50%–â•‰100% is often reported (Willinek and
Kuhl, 2006; Bradley, 2008). High-â•‰resolution scans tend to manifest a greater
gain because in those scans thermal noise is usually the dominant source of noise
relative to physiological noise, which scales with signal. Some sequences such
as BOLD fMRI and ASL perfusion benefit from additional factors related to
â•‡ 13
enhanced magnetic susceptibility effects and longer T1 at 3T. Other sequences

such as FLAIR benefit less because slower signal recovery at higher field offsets
some of the advantages.
7T MRI is also becoming increasingly available in some research institutions.
There are about 50 to 60 human 7T systems around the world. However, the use of 7T
in cognitive aging studies is still in an early stage. Despite the promise of an increased
sensitivity, 7T MRI still suffers from several technical limitations at present. These
limitations are due to magnetic field inhomogeneity (thereby resulting in image inho-
mogeneity), higher power deposition (thus tissue heating becomes an issue, which is
usually not a problem when going from 1.5T to 3T), and physiological side â•‰effects
(i.e., participant may feel dizzy when entering the scanner), aside from its high costs.
However, these limitations may be resolved with technical efforts and advances, as
shown by highly promising results from the Human Connectome Project.
Practical Considerations When Designing a Cerebral Aging MRI Study

It is recommended that each scan session be less than 60 minutes, as excessive motion
is often observed when the subject has been inside the scanner for a long period of
time. When that happens, the data collected are of low quality and usually need to
be excluded. This is especially likely for elderly participants. For studies that require
more than 60 minutes, one should consider allowing the subject to come out of the
scanner to take a break before entering again for the remainder of the scans.
It is also useful to prioritize the scans such that the pulse sequences that are most
important for the study hypothesis are performed first. It is not uncommon that a sub-
ject would decide to abort the scan session or show excessive motion after staying in
the scanner for a while. Thus, arranging the scan order based on priority can ensure
the successful data collection of the most relevant sequences. Some investigators also
found it helpful to place the functional and physiological scans at the beginning of the
session when the subject is most alert and attentive. The structural scans can usually
be performed even when the subject is asleep.
When calculating the total scan session duration, it should be remembered that the
scan duration displayed on the scanner console often underestimates the real-â•‰life scan
time, as the preparation time necessary at the beginning of every scan is usually not
included in the displayed time. The preparation time is inherent to every MRI pulse
sequence and usually includes B0 shimming, RF center frequency determination,
RF power optimization, and dummy scans to allow the magnetization to approach a
steady state. Collectively, the actual data collection time for a 60-â•‰minute session may
be around 45 minutes. It is useful to keep this in mind during planning. A small num-
ber of pilot scans should be performed before finalizing the protocol and obtaining a
more accurate estimation of the session duration.
Common MRI Techniques Used in Cerebral Aging
Table 1.1 provides a list of MRI techniques commonly used in cerebral aging studies.
They measure various aspects of the brain structure and function. Later chapters will
14
Table 1.1 List of MRI techniques and associated imaging parameters at 3T.
MRI technique Usage Scan duration Typical imaging parameters
T1-MPRAGE Provide brain volumetric 4–7 min 3D Acquisition,

information parallel imaging acceleration factor = 2,
Voxel size = 1×1×1mm3,
field-of-view (FOV) = 256×204mm2,
TR = 8.2(along y-phase-encoding
direction) and 2100ms(along z-phase-
encoding direction),
TI = 1100ms,
TE/flip angle = 3.7ms/12º,
160 sagittal slices with whole brain
coverage
Functional Measurement of 5–10 min Multislice acquisition,

MRI neural activity through voxel size = 3.4×3.4×3.5 mm3,
hemodynamic responses
FOV = 220 × 220 mm2,
TR/TE/flip angle = 2000/30 ms/80°,
39 axial slices with whole brain coverage
DTI Probe microstructural 4–5 min Multislice acquisition,

tissue integrity and voxel size = 2×2×2.2 mm3,
white matter
FOV = 220 × 220 mm2,
connectivity
b value = 1000 s/mm2,
30 directions
Proton density Help detect tissue 3–5 min Multislice acquisition,

MRI lesions in white matter Turbo-spin-echo for fast imaging,
voxel size = 1×1×3 mm3,
FOV = 240 × 210 mm2,
T2-weighted Detect tissue lesions 2–4 min Multislice acquisition,

MRI in white matter Turbo-spin-echo for fast imaging,
voxel size = 1×1×3 mm3,
FOV = 240 × 210 mm2,
T2-FLAIR Detect tissue lesions 3–5 min 2D version:

in white matter voxel size = 1×1×2 mm3,
FOV = 256 × 256 mm2,
TR/TE/TI = 11,000/100 ms/2800 ms,
138 axial slices with whole brain
coverage
3D version:
voxel size = 1.1×1.1×1.1 mm3,
FOV = 240 × 240 mm2,
TR/TE/TI = 4800/278 ms/1650 ms,
150 sagittal slices with whole brain
coverage
â•‡ 15
cover the analyses and interpretation of some of these techniques. The present chapter
will primarily focus on the image acquisition strategies.
Brain Volumetrics
Cerebral aging is associated with pronounced brain volumetric changes (Pfefferbaum
et al., 1994; Good et al., 2001; Ge et al., 2002; Sowell et al., 2003; Raz et al., 2005).
Thus, the ability to assess brain volume and cortical thickness is important in cerebral
aging studies. Brain volumetric parameters are usually assessed with a high-â•‰resolu-
tion anatomical MRI scan that provides sufficient contrast between gray matter, white
matter, and CSF. In principle, this can be achieved with any one of the three main
image contrasts in MRI, i.e., T1-â•‰weighted, T2-â•‰weighted, and proton-â•‰density weighted.
In practice, however, the vast majority of such studies have used the T1-â•‰weighted
pulse sequence (Pfefferbaum et al., 1994; Good et al., 2001; Ge et al., 2002; Sowell et
al., 2003; Raz et al., 2005). This is due to scan time considerations. A proton-â•‰density
weighted pulse sequence usually utilizes a long TR, thus is less time-â•‰efficient. A T2-â•‰
weighted pulse sequence usually uses a long TR, again lengthening the scan duration.
Therefore, a T1-â•‰weighted pulse sequence, which uses a short TR and a short TE, is
considered an optimal approach for rapid, strong-â•‰contrast, and high-â•‰resolution acqui-
sition of brain structure information.
T1 of CSF is greater than gray matter T1, which is in turn greater than white mat-
ter T1. At 3 Tesla, these values are 4,163 ms (Lin et al., 2001), 1,135 ms (Lu et al.,
2005), and 732 ms (Lu et al., 2005), respectively. Thus, in a T1-â•‰weighted image,
white matter is expected to have a higher signal intensity than gray matter, and CSF
will have the least signal intensity. Figure 1.3A shows an example of T1-â•‰weighted
image at a resolution of 1x1x1 mm3. Using image segmentation functions that are
readily available in standard software packages, probability maps of gray matter
(A) (B) (C) (D)
T1-MPRAGE Gray matter White matter CSF

probability probability probability
FigureÂ€1.3â•‡ Example of T1-â•‰weighted image and the corresponding probability maps of gray
matter, white matter and CSF. (A) T1-â•‰MPRAGE image obtained from a MRI scan. (B) Gray
matter probability map. (C) White matter probability map. (D) CSF probability map. (b-â•‰d) are
obtained by segmenting T1-â•‰MPRAGE image shown in (a).
16
16 Methods and Issues
(Figure 1.3B), white matter (Figure 1.3C), and CSF (Figure 1.3D) can be obtained

from the T1-weighted image.
In the early years, T1- weighted high- resolution images were acquired using
a plain short- TR, short- TE gradient- echo sequence. Different vendors have dif-
ferent names for this sequence. On MRI scanners made by General Electric, it is
called Spoiled Gradient Echo (SPGR). On Siemens systems, it is called Fast Low
Angle Shot (FLASH). On Philips scanner, it is called Fast Field Echo (FFE). More
recently, an improved version of the T1-weighted pulse sequence has gained popular-
ity. This sequence is called Magnetization Prepared Rapid Acquisition of Gradient
Echo (MPRAGE) (Mugler and Brookeman, 1991), which uses a preparation pulse
to enhance the tissue contrast between gray matter, white matter, and CSF, thereby
allowing better tissue segmentation and image registration. The MRPAGE sequence
is the most widely used technique for brain volumetric studies at present.
Brain volumetric images are usually collected at a high resolution. The typical
voxel size of a brain volumetric image is around 1x1x1 mm3, although a slightly non-
isotropic voxel size (e.g., 1x1x1.2 mm3) is sometimes also used. This level of spatial
resolution is needed in order to provide sufficient number of voxels across the thick-
ness of cortical ribbon, which is around 2–4 mm. A resolution higher than 1 mm is of
course desirable, but often at a cost of scan duration and/or SNR. Thus, sub-millimeter
resolution (e.g., 0.7 mm) is only used when one is specifically interested in a small
structure of the brain, for example identifying subfields of hippocampus. Under these
circumstances, usually the image will focus on the specific structure but not cover the
entire brain, i.e., only partial brain coverage.
Brain volumetric images are acquired in 3D. In MRI acquisition, there is a clear
distinction between “3D” and “multislice” acquisitions. In multislice acquisition,
although the stack of 2D slices can cover the same brain volume as a 3D scan, the
through-plane resolution is never as good as that of a 3D acquisition. That is, even if
one prescribes the slice thickness in a multislice acquisition to be 1 mm, the actual
resolution cannot reach 1 mm due to imperfection in slice selection profile. This is
because the RF pulse used to excite the spins has a frequency-selection band that
is not perfect. As a result, the spins adjacent to the intended slice are also partially
excited, and this effect decays with distance. Fortunately, brain volumetric images are
always collected in a 3D mode, thus the images will look equally sharp when viewed
in axial, sagittal, or coronal planes. A corollary of this notion is that, from the image
quality point of view, it does not matter if the original scan is performed in axial, sag-
ittal, or coronal orientation, because the data can be reformatted into any orientation
without loss of spatial resolution. From the acquisition efficiency (thereby scan dura-
tion) point-of-view, sagittal orientation is often used. The main reason for this choice
is that the human brain is usually the shortest along the left-right direction, typically
about 160–180 mm. Thus, it requires the least number of slices to cover the brain from
the left to right side. A second reason is that there is no other source of MRI signal to
the left or right of the brain (unless the cushion or headset stabilizing the head contains
fluid). Thus, even if the slice selection profile is not perfect, no spurious signal will be
excited or detected outside the intended excitation volume. Such signals would have
manifested themselves as foldover artifacts in which a spin to the left of the intended
volume will overlap with the right end of the image, and vice versa.
â•‡ 17
Without any acquisition acceleration schemes (e.g., parallel imaging), typical scan
duration for an MRPAGE sequence is between 8–â•‰10 minutes. With a parallel imaging
acceleration factor of 2, this becomes 4–â•‰5 minutes. When describing the MRPAGE
sequence in a report, aside from the typical parameters such as voxel size, field-â•‰of-â•‰
view, and echo time (TE), it is useful to specify an imaging parameter referred to as
the inversion time (TI). In the MPRAGE sequence, the TI is crucial in determining
the contrast of the image. In general, it is useful to describe the imaging parameters as
thoroughly as possible when reading or writing an article.
Functional MRI
Functional MRI (fMRI) provides an approach to probe the brain function nonâ•‰invasively
(Bandettini et al., 1992; Kwong et al., 1992; Ogawa et al., 1992). The most commonly
measured fMRI signal is based on an indirect effect of neural activity on blood flow
and blood oxygenation. This signal is referred to as blood-oxygenation-â•‰level-â•‰depen-
dent (BOLD) signal (Ogawa et al., 1990), since it is sensitive to blood oxygenation in
the venous vessels. There are two forms of fMRI, task-â•‰related fMRI and resting-â•‰state
fMRI. In task-â•‰related fMRI, one aims to measure BOLD signal changes in response
to a time-â•‰controlled task. Thus, the researcher usually presents a preâ•‰defined stimulus
time series to the participant while the MRI scanner is continuously acquiring images
of the brain. Then, in data analysis, mathematical algorithms such as linear regression
are used to search for the expected temporal signal pattern throughout the brain. If any
voxel manifests the expected pattern, one usually calls this voxel “activated” and labels
it with a pseudocolor. In resting-â•‰state fMRI, no explicit task is applied during the experi-
ment while images are continuously collected for a period of a few minutes. Then, in
analysis, one either selects a seed voxel and looks throughout the brain for voxels that
have a signal pattern similar to that of the seed, or one uses more advanced algorithms
such as independent component analysis to identify clusters of voxels that have similar
temporal fluctuation, presumably due to direct or indirect synaptic connections. Thus,
resting-â•‰state fMRI is also referred to as functional connectivity MRI (fcMRI).
Both task-â•‰related and resting-â•‰state fMRIs use a T2*-â•‰weighted echo-â•‰planar-â•‰
imaging (EPI) technique. The fMRI pulse sequence uses a multislice acquisition
scheme to cover the whole brain, in which the images are taken on a slice-â•‰by-â•‰
slice basis. For each slice, a fast acquisition technique, EPI, is used to collect the
data. For a typical brain volume containing 40 slices, it takes up to 3,000 ms to
complete the data collection. Since neural activity and the associated hemody-
namic responses are transient, it is desirable to expedite the rate of data collection.
This can be achieved using a recently developed multiband acquisition technique
in which two or more slices are acquired simultaneously (Feinberg et al., 2010;
Mueller et al., 2010; Setsompop et al., 2012). As such, the time it takes to complete
the data collection of one volume can be shortened substantially. At present, the TR
in an fMRI scan can be as short as 600 ms or less. These recent technologies have
been utilized in major brain imaging initiatives such as the Human Connectome
Project (Van Essen et al., 2012) and are gradually becoming the standard procedure
for fMRI studies.
18
The typical spatial resolution of fMRI is between 3–3.5 mm, but there is a trend
toward a higher spatial resolution of 2–2.5 mm with recent advances in fast imaging
technologies, including multiband, parallel imaging, and high-field MRI.
FMRI signal value at any particular time point, often written in arbitrary units with
a typical range of several hundred, has no physiological meaning. It is the temporal
fluctuation or change of the signal that contains information about neural activity.
Taking task-related fMRI, for example, a task s timulus may cause the fMRI signal to
increase by a fraction of a percent. A signal change of 1% is considered a large signal in
fMRI data, in particular for cognitive tasks. Does a 1% signal change mean that neural
activity has increased by 1%? Not really. FMRI signal is a complex function of several
physiological parameters including cerebral blood flow (CBF), cerebral blood volume
(CBV), and cerebral metabolic rate of oxygen (CMRO2) (Davis et al., 1998; Hoge et al.,
1999; Arthurs and Boniface, 2002). Figure 1.4A illustrates the pathway that leads to the
observation of fMRI signal change due to a stimulus. Neural activation induced by the
stimulus results in an elevation in metabolism and release of neurotransmitters. These
factors could either directly dilate blood vessels or do so via activation of glial cells such
as astrocytes, which have end feet attached to vessels. These effects cause an increase in
CBV and CBF, which are referred to as the hemodynamic responses of the brain. It is
important to note that the increase in oxygen supply (i.e., CBF) during brain activation
is more prominent than the increase in oxygen demand (i.e., CMRO2), thus the oxy-
genation status of the venous blood is increased. Therefore, given a task s timulus such
as shown in Figure 1.4B, BOLD fMRI signal change could be detected (Figure 1.4C).
(A)
Neurovascular
coupling
Metabolites
MRI
• Blood volume ↑ scan
Stimulus • Blood flow ↑
• Blood oxygenation ↑
Neuronal Vessel size
activity fMRI
Astrocytes BOLD
signal
Neurotransmitters
(B) (C) 3
BOLD signal (%)
3s 2
4–17s
+ 3s 1
4–17s
+ 3s 0
+ –1
–2
0 100 200 300
time Time (s)
Figure 1.4 Illustration of BOLD fMRI. (A) Illustration of the neurovascular coupling pathway

that leads to the observation of fMRI signal change due to a stimulus. (B) Example of task-
fMRI stimulus paradigm. (C) Corresponding BOLD signal in visual cortex from an fMRI scan
using the paradigm in (b). Red bars indicate stimulus periods. (See color plate also)
â•‡ 19
How is venous blood oxygenation level related to fMRI signal? It is fortuitous that
hemoglobin in the blood has a differential magnetic property between oxygenated and
deoxygenated states, thus can serve as an endogenous contrast agent for MRI. When
oxygenated, the hemoglobin is in a so-â•‰called diamagnetic state, which does not distin-
guish itself from the surrounding water in terms of magnetic property. In the deoxy-
genated state, on the other hand, it is in a paramagnetic state, which generates a small
magnetic field in its surrounding, causing a disturbance to the homogeneity of the
magnetic field. Since T2* of a voxel is closely associated with the homogeneity of the
field, this MR parameter is dependent on the abundance of deoxyhemoglobin in the
voxel and thereby the oxygenation status of the blood.
When integrating these cascades in a quantitative manner, the fMRI signal change
due to task stimulation can be written as (Davis et al., 1998; Hoge et al., 1999):
∆BOLD   ∆CMRO2  β  ∆CBF 

α −β

= M ⋅ 1 −  1 +  ⋅  1 +  , [3]â•„
BOLD   CMRO2  CBF  
in which ΔCMRO2/â•‰CMRO2 indicates task-â•‰induced change in brain metabolic rate;

ΔCBF/â•‰CBF is the change in blood flow; M, α, and β are variables related to scanner
field strength and imaging parameters, blood flow-â•‰volume relationship, and vessel
geometry, respectively. To provide the readers with a realistic example of these param-
eters in the primary visual cortex, Hutchison et al. (2013) found that flashing grating
caused a ΔCMRO2/â•‰CMRO2 of 15% and a ΔCBF/â•‰CBF of 35%, which resulted in a
fMRI signal increase by 0.7%.
More discussions on fMRI experimental design and analyses in aging studies can
be found in chapters 4 and 6.
Diffusion Tensor Imaging (DTI)

DTI provides a powerful tool to evaluate microstructural tissue integrity and white-
matter connectivity, beyond those of conventional structural imaging (Basser et al.,
1994). Diffusion MRI can reflect structural properties of the brain because tissue
structures such as cell membrane act as barriers that prevent water molecules from
free, random motion. Furthermore, some structures such as axonal membrane and
myelin sheath disrupt free diffusion in an orientation-â•‰dependent manner. That is,
along the white-matter fiber’s principal direction, the water diffusion is minimally
affected, whereas across the fiber direction the diffusion is heavily influenced. As
such, DTI can also provide an anisotropy measure of water diffusion, which can be
used to determine white-matter fiber orientation (Mori et al., 1999; Le Bihan et al.,
2001).
DTI measures the diffusion pattern of water molecules by applying two magnetic
field gradients in the pulse sequence (Figure 1.5A), which can encode the distance of
water diffusion during this period. The strength of the magnetic field gradients can be
represented by an imaging parameter called “b value,” which is associated with both
magnitude and duration of the gradients as well as their time gap:
20
(A)
Acquisition
RF
G
δ δ
∆
(B)
b = 0 s/mm2 b = 1000 s/mm2 ADC map
(C)
Dxx Dxy Dxz
D = Dyx Dyy Dyz
Dzx Dzy Dzz
(D) Dxx Dxy Dxz x

ADC = Dyx Dyy Dyz . y
Dzx Dzy Dzz z
Figure 1.5 Illustration of diffusion tensor imaging (DTI). (A) A simplified DTI pulse sequence.
(B) Diffusion images with two different b values and the calculated ADC map. (C) Components
of the diffusion tensor matrix. (D) ADC can be determined by the tensor matrix.
b = γ 2 G 2 δ 2 ( ∆ − δ / 3) , [4]‌
where δ is the duration of the gradient, G is the magnitude of the gradient, and Δ is
their time gap. The DTI signal is expected to follow a mono-exponential function of
the b value:
S = S0 e − bD,[5]‌
in which S is the DTI MRI signal, D is called the apparent diffusion coefficient
(ADC). Therefore, by applying a minimum of two b values while measuring their
corresponding MRI signal intensities, one can determine the ADC value (Figure
1.5B).
It is important to note that magnetic field gradient has directionality, and it can
be applied in any direction of the 3D space. Accordingly, one can measure an ADC
along any direction in space. Therefore, an infinite number of ADC values exist
for a given tissue, which makes it difficult to measure and to interpret in practice.
21
MRI Measures of Aging 21
Fortunately, with reasonable assumptions (e.g., single tissue compartment), one can
show that these ADCs cannot take arbitrary values. Instead, they follow the con-
straints of a tensor matrix (Figure 1.5C). That is, the water diffusion pattern in a
tissue has six degrees of freedom, corresponding to three directional variables and
three magnitude variables, and once the tensor matrix is obtained, ADC value along
any direction in space can be determined (Figure 1.5D). Other indices can also be
derived from the diffusion tensor matrix:
( λ1 − λ 2 )2 + (λ1 − λ 3 )2 + ( λ 2 − λ 3 )2
Fractional anisotropy ( FA ) = ,[6]‌
(
2 λ12 + λ 2 2 + λ 32 )
Mean diffusivity ( MD ) = ( λ1 + λ 2 + λ 3 ) / 3, [7]‌
Radial diffusivity ( RD ) = ( λ 2 + λ 3 ) / 2, [8]‌
Axial diffusivity ( AxD ) = λ1 ,[9]‌
where λ1, λ2 and λ3 are the eigenvalues of the diffusion tensor.

Conceptually, the three-dimensional description of water diffusion is equivalent
to dropping a small amount of ink in tissue and watching how it spreads with time.
When there are barriers in the tissue (e.g., myelin sheath), the spatial distribution of
the ink will not be spherical and instead will appear like an ellipsoid. This results in
an anisotropy in diffusion.
One needs to apply the magnetic field gradients along a minimum of six non-col-
linear directions. Then, together with a b0 image acquired under the absence of the
gradient, the diffusion tensor can be determined. In practice, more gradient directions
are generally used in order to improve the reliability of the data. At present, most
studies use 30 to 128 gradient directions. In theory, these directions can be arbitrarily
chosen as long as they are not collinear. However, in order to obtain the best sampling
of the 3D space, it is recommended to spread the gradient directions to be evenly
spaced on the sphere. Optimal gradient tables for various numbers of directions have
been reported in the literature (Jones et al., 1999; Skare et al., 2000).
One technical point that an investigator should be cognizant of is whether the gra-
dient directions are defined in magnet coordinates or in imaging slice coordinates. The
magnet coordinates refer to the space defined by the MRI scanner’s physical axes, and
are independent of the patient’s position or the slice orientation that the MRI operator
places. The imaging slice coordinates refer to the space defined by the orientations of
the acquired slices, and they change with the MRI operator’s tilting of the angulation
of the slices. Therefore, depending on which coordinates the gradient directions are
based upon, the data could appear quite differently. In essence, the ellipsoid depicted
by the diffusion tensor could be tilted from the true angle if the gradient directions
are not interpreted correctly. This does not change parametric maps such as mean dif-
fusivity (MD) or fractional anisotropy (FA) of the data. However, when one aims to
use the DTI data for fiber tractography or for determining the white matter principal
22
direction (e.g., color maps), the results will likely be incorrect. Different MRI vendors
and pulse sequence programmers may define the gradient directions differently. The
results are also dependent on the processing pipelines in terms of the orders of regis-
tration and tensor fitting. Therefore, it is useful to verify the specific acquisition and
processing settings before conducting fiber tractography and displaying color maps
in DTI analysis.
DTI data are usually collected in multislice mode, in which a series of 2D axial
slices are acquired to form a whole-brain volume. The acquisition voxel size of DTI is
usually around 2x2x2 mm3, and is often interpolated to a reconstructed voxel size of
1x1x2 mm3. In the recent protocols of the Human Connectome Project, the acquisition
resolution of DTI has been further improved to 1.25x1.25x1.25 mm3 (Sotiropoulos
et al., 2013). Multiband technology (Feinberg et al., 2010) is increasingly used in DTI
to reduce the scan duration.
Typical acquisitions of DTI use a single non-zero b value, between 700–1000 s/
mm2. This range of b values can provide a sizeable signal decay (thereby allowing
a reliable fitting of the decay curve) while maintaining sufficient signal intensity.
Recently, there is a trend to collect more b values (referred to as shells), which could
allow the assessment of diffusion properties beyond a single tensor (Tuch et al., 2003;
Jensen et al., 2005; Aganj et al., 2010; Descoteaux et al., 2011). These advanced mea-
surements may permit the detection of crossing fibers in a voxel, the determination of
axonal diameter, and the quantification of non-Gaussian diffusion metrics.
Anisotropy indices such as FA can be further exploited to conduct analysis of fiber
connections, termed tractography. Readers are referred to Chapter 3 for connectivity
analysis of DTI and application of DTI in cerebral aging.
Tissue Lesion Detection with Proton-Density, T2-weighted,

and FLAIR MRI
Cerebral aging is often associated with tissue lesions, in particular in the white mat-
ter (Gunning-Dixon and Raz, 2000; de Leeuw et al., 2001). MRI techniques that are
capable of detecting tissue lesions include proton-density, T2-weighted, and FLAIR
MRI, because lesion areas usually have high water density and longer T2 relaxation
time. The lesion regions therefore appear bright in these images.
As mentioned earlier in the chapter, these images usually take longer time to
acquire because the TR needs to be relatively long (several seconds) in order to
allow the magnetization to recover sufficiently. Therefore, their spatial resolutions
are often lower than that of T1-weighted image. Typical acquisition voxel size is
approximately 1x1x2 mm3. The images are usually acquired in multislice mode,
although 3D acquisitions are increasingly used. To accelerate the data collection
efficiency, a fast imaging approach, turbo-spin-echo, is used, which allows the
collection of 10–20 lines in the k-space following each excitation. Without this
acceleration, it would have been impractical to collect these images within a clini-
cally relevant time frame.
In terms of the analysis of these lesion-detection images, there is not a standard
analysis protocol. Three categories of analysis approaches have been used. One is
â•‡ 23
through manual, visual inspection of the images by a trained individual, e.g., a neu-
roradiologist. The rater will provide a quantitative rating of the lesion volume based
on his/â•‰her experience, e.g., negligible lesion, modest amount of lesion, severe lesion,
etc. (de Leeuw et al., 2001; Wahlund et al., 2001). Reliability of the rating can be
checked by having a second rater evaluate the images and the consistency between the
raters can be examined. A second category is a semi-â•‰automatic approach, in which the
image is first analyzed with an algorithm, usually based on signal intensity threshold-
ing together with cluster size requirement (Zijdenbos et al., 1994; Hulsey et al., 2012).
The resulting mask is then manually edited by a rater to remove spurious voxels due
to motion, heterogeneity in coil sensitivity, and so on. This analysis approach yields
a quantification of total lesion volume as well as lesion volume by spatial location,
e.g., periventricular lesion volume. A third category of approaches are fully automated
algorithms to delineate lesion voxels based on sophisticated computational methods
(Lao et al., 2008; de Boer et al., 2009). Usually these approaches utilize multiple
image contrasts, e.g., proton-â•‰density, T2-â•‰weighted, FLAIR, and T1-â•‰weighted images,
to improve the specificity of the voxel delineation. The advantage of this type of
approach is, obviously, its reliability and potential sensitivity to image quality; how-
ever, it needs to be further evaluated. Additionally, the requirement of multiple image
contrasts usually lengthens scan duration.
In recent years, some new advances in the application of T1-â•‰weighted and T2-â•‰
weighted MRI are being developed. For example, the ratio between T1-â•‰weighted
and T2-â•‰weighted MRI signal intensity has shown to be sensitive in detecting age-â•‰
related change of intracortical myelin content during normal development and
aging (Grydeland et al., 2013). However, validation studies are needed for such new
techniques.
Emerging MRI Techniques Relevant for Cerebral Aging
In addition to the techniques described above, recent advances in the MRI field have
enabled several other methods that are ready for application in cerebral aging studies.
This section will provide a brief introduction of these promising MRI techniques.
Cerebral Perfusion with Arterial-â•‰Spin-â•‰Labeling (ASL) MRI

Cerebral perfusion, denoted by cerebral blood flow (CBF), is an important index for
brain function and viability. Diminished CBF is a known cause of cognitive decline
and can further cause dementia (Gorelick et al., 2011; Zlokovic, 2011). Many struc-
tural alterations in the brain such as white-matter lesions are also thought to be associ-
ated with CBF dysfunction (Schuff et al., 2009). Therefore, the ability to determine
CBF will provide an important piece of information on cerebral aging.
CBF is traditionally measured with positron emission tomography (PET) (Mintun
et al., 1984). However, safety concerns associated with radioactive tracers as well as
the complexity of the procedure have limited its application in cerebral aging studies.
CBF techniques based on MRI have previously been devised and have been routinely
24
Control Time delay Acquisition

Control
Control-Label
Label Time delay Acquisition

Label
Figure 1.6 The principle of ASL MRI. In the control scan, the magnetization of the blood is
unchanged, whereas in the label scan, the magnetization of the blood is inverted when passing
through the labeling plane. Note that the RF pulse train is still played out during the control
scan, but it is essentially a zero-degree RF pulse. The purpose of this is to equate the magnetiza-
tion transfer effects between the label and control images. After a time delay which allows the
labeled blood to arrive at the imaging slice, a control image and a labeled image are acquired.
The subtraction of control and labeled images can cancel the static tissue signal and the result-
ing difference image provides an estimation of CBF to the brain. (See color plate also)
used in acute stroke (Ostergaard et al., 1996b; Ostergaard et al., 1996a; Sorensen et
al., 1999). However, these earlier CBF techniques require the injection of MRI con-
trast agent. Therefore, for healthy participants who are not clinically implicated to
receive contrast agent, this again presents a practical obstacle.
More recently, an MRI technique called arterial-spin-labeling (ASL) has emerged
as a noninvasive (i.e., does not require the injection of exogenous agent) method to
measure CBF in humans (Williams et al., 1992; Kim, 1995; Wong et al., 1998; Yang
et al., 1998; Golay et al., 1999; Wang et al., 2005; Dai et al., 2008). This method uses
blood water as an endogenous tracer and, by noninvasively labeling it using radiofre-
quency pulses, CBF information can be obtained. Figure 1.6 illustrates the principle
of ASL MRI. One first performs a labeled MRI scan (bottom panel, Figure 1.6), in
which incoming arterial blood is labeled at its entry point to the brain, usually at the
level of the neck or lower part of the brain. A time delay is then applied to allow the
labeled water to reach its destination, the brain parenchyma. An image of the brain
slice is then acquired, which contains signal of the newly arrived water as well as
that of static tissue water (white arrows, Figure 1.6) that had been there all along.
The signal from the newly arrived water molecules is the target signal and reflects
CBF, but the data from the static signal is a nuisance. Therefore, a control MRI scan
(top panel, Figure 1.6) is performed, during which the incoming water is not labeled.
Following a delay time identical to that of the labeled scan, a brain image is acquired.
â•‡ 25
In the control image, the static tissue signal is the same as that in the labeled scan.
Therefore, the subtraction of labeled and control images can cancel the static tissue
signal, and the resulting difference image provides an estimation of CBF to the brain.
With a perfusion kinetic model (Chalela et al., 2000; Alsop et al., 2014), one can also
convert the MRI signal intensity to quantitative CBF values in the units of ml blood
per 100g brain per minute (ml/â•‰100g/â•‰min).
The difference signal in ASL is only 1% or less of the original MRI signal intensity
(Alsop and Detre, 1996; Liu et al., 2011), which limits its sensitivity and is considered
a weakness of this technique. With recent advances in ASL technique, however, this
drawback is mitigated to a large extent, and at present one can obtain a whole-â•‰brain CBF
map within a reasonable scan duration of 4–â•‰6 minutes. There is now also a consensus in
ASL implementation in the field (Alsop et al., 2014), and all major MRI vendors have
included ASL in their product software. Age-â•‰specific post-â•‰labeling delay times have
also been proposed (Alsop et al., 2014). It is generally recommended that a longer delay
time should be used in older participants. Therefore, access to ASL MRI is becoming a
routine procedure in research and clinical and settings. One caveat is that white-matter
CBF is still difficult to measure reliably with the current ASL methods. Thus, at present,
most investigations of cerebral perfusion is focused on the gray matter.
ASL MRI has been applied in cerebral aging studies. CBF was found to decrease
with age, which is most pronounced in the prefrontal cortex (Lu et al., 2011). This CBF
reduction is independent of gray matter thickness (Chen et al., 2011) and is associated
with white matter integrity as assessed by DTI (Chen et al., 2013). CBF was affected
by genetic factors such as APOE e4 status, wherein cognitively normal e4 carriers dis-
played greater CBF than non-â•‰carriers (Wierenga et al., 2013). CBF was also affected
by lifestyle factors such as physical exercise, in which individuals who maintain long-â•‰
term aerobic exercise showed greater CBF in posterior cingulate cortex (Thomas et al.,
2013). Cognitive training was also shown to be able to enhance CBF in frontal execu-
tive network (Chapman et al., 2015). On the other hand, hyperperfusion in the medial
temporal lobe appears to be a sign of dysfunction, as elevated CBF has been observed
in patients with Alzheimer’s disease (Alsop et al., 2008), and hippocampal CBF was
found to be inversely correlated with verbal memory (Rane et al., 2013).
Accounting for Vascular Aging Effect in Functional MRI

Much of our understanding of brain functional changes with age is based on fMRI
findings. However, because fMRI signal is a vascular response (Arthurs and Boniface,
2002) and the brain vasculature has known degradation with age (Lu et al., 2011), it
is not straightforward to interpret age-â•‰related fMRI changes. For example, when an
age-â•‰related signal decline is observed, it could be due to diminished neural activity or
dampened vasodilatory response (or both).
To accurately examine neural activity changes with age, the vascular changes
need to be separately measured and factored out (D’Esposito et al., 1999; Jezzard
and Buxton, 2006; Liu et al., 2013a). One way to estimate vasodilatory capacity of
the blood vessels is to induce a transient hypercapnic condition while monitoring
MRI signal changes (Davis et al., 1998; Hoge et al., 1999; Kastrup et al., 1999;
26
Yezhuvath et al., 2009; Kannurpatti et al., 2010). CO2 is a potent vasodilatory stimu-
lus and is known to result in increased blood flow and oxygenation in the brain.
Hypercapnia can usually be induced via two approaches. One is to ask the sub-
ject to hold breath briefly (for 15–30 seconds) based on instructions displayed on
the screen (Kastrup et al., 1999; Kannurpatti et al., 2010). The advantage of this
approach is that no additional apparatus or equipment is needed, thus it is relatively
convenient to deliver the hypercapnic stimulus. A limitation of this approach is that
the success of this maneuver is highly dependent on the cooperation of the partici-
pant, and furthermore, the actual quantity of the stimulus strength is not determined
because end-tidal CO2, which reflects the CO2 concentration in the lung and arterial
blood, cannot be measured during breath-hold. A second approach to induce hyper-
capnia to a subject is via inhalation of a gas mixture containing a small amount of
CO2 (Davis et al., 1998; Hoge et al., 1999; Yezhuvath et al., 2009). Usually, the
CO2 content in the gas mixture is about 5%, which does not induce discomfort
when breathing briefly (e.g., 1 minute at a time). The advantages of this approach
are that physiological state can be well controlled, end-tidal CO2 can be continu-
ously measured, and it does not strongly depend on the cooperation of the subject.
Its disadvantage is that additional apparatus is required to allow the delivery of the
gas mixture to the subject while he or she is inside the MRI scanner. However, with
an increasing interest in using hypercapnia in research and clinical imaging, many
implementations of the gas delivery apparatus are now available (Slessarev et al.,
2007; Wise et al., 2007; Yezhuvath et al., 2009).
MRI response to hypercapnia is often referred to as cerebrovascular reactivity
(CVR), which is an index representing the ability of the blood vessels to dilate when
stimulated. CVR can then be used to correct fMRI signal (Davis et al., 1998; Hoge
et al., 1999; Sowell et al., 2003; Ances et al., 2009; Gauthier et al., 2013; Hutchison
et al., 2013; Liu et al., 2013b).
Correction of task- related fMRI signal with CVR can generally take two
approaches. One approach uses BOLD MRI pulse sequence during the hypercapnia
scan and obtains a CVR index in the units of %BOLD per mmHg CO2. The fMRI
signal can then be corrected by (Liu et al., 2013b):
S fMRI,corr = S fMRI,uncorr / CVR,[10]
in which S fMRI,uncorr and S fMRI,corr are the uncorrected and corrected fMRI signal, respec-
tively. This approach does not require any change to the fMRI scanning protocol
(thus the investigator can still have a complete dataset of original, uncorrected fMRI).
A second correction method, termed calibrated fMRI, uses a more model-based
approach (Davis et al., 1998; Hoge et al., 1999; Ances et al., 2009; Gauthier et al.,
2013; Hutchison et al., 2013). In this approach, vascular response due to hypercap-
nia is assessed by both BOLD and CBF changes, using an advanced pulse sequence
combining T2* and ASL MRI. Therefore, more information is collected. When fitting
these experimental data to a comprehensive model, task-related changes in cerebral
metabolic rate of oxygen (CMRO2) can be estimated, which are thought to reflect
aggregated neural activity in the brain. The disadvantage of this approach is that the
â•‡ 27
fMRI data acquisition is also required to use the combined T2* and ASL sequence,
which is less efficient due to the longer TR necessary and generally has a lower SNR.
As a result, most of the calibrated fMRI studies have been limited to sensory or motor
tasks (Davis et al., 1998; Hoge et al., 1999; Ances et al., 2009; Hutchison et al., 2013).
There have been few studies that used cognitive tasks or examined regions outside the
visual or motor cortices (Gauthier et al., 2013).
Age-â•‰related fMRI data could reveal very different results with and without CVR
correction. For example, Liu et al. showed that, before CVR correction, fMRI response
in the visual cortex manifests an age-â•‰related decrease (Liu et al., 2013b). However,
after correction it appears that the true neural response actually increases with age,
similar to the age â•‰pattern in the prefrontal cortex. That is, the uncorrected BOLD
signal reduction in the visual cortex may be purely a vascular artifact. A similar find-
ing of age-â•‰related increase in the visual cortex is observed in Hutchison et al. (2013),
using the calibrated fMRI approach.
Aside from the utility of correcting fMRI signal, CVR itself is a useful index of
vascular health of the brain. With age, CVR has been shown to decrease throughout
the brain, and the rate of the decline is faster than baseline perfusion (Lu et al., 2011).
This is consistent with the brain’s autoregulation function, which aims to maintain a
relatively constant blood flow by reducing the vascular tone.
Susceptibility Weighted Imaging (SWI)

The aging brain is usually associated with accumulation of iron (Zecca et al., 2004),
increased incidence of microbleeds (Poels et al., 2010), and decreased blood oxy-
genation (Lu et al., 2011; Peng et al., 2014). SWI is a useful tool to examine these
parameters (Haacke et al., 2004). Unlike most MRI sequences where only magnitude
information of the MRI signal is collected, SWI also acquires the phase information.
The phase of MRI signal is highly sensitive to iron, calcium, and blood oxygenation,
and provides information complementary to magnitude image.
A standard SWI scan employs a 3D gradient-â•‰echo sequence using a relatively high
resolution with in-â•‰plane voxel size of 0.5 to 1 mm and through-â•‰plane thickness of 1 to
2 mm. A whole-â•‰brain scan usually takes 5 to 10 minutes. Depending on the focus of
the study, there exist several implementations of the SWI technique.
A major utility of SWI is to visualize brain venous vessels and detect microbleeds
(Kidwell et al., 2002; Sehgal et al., 2005). In this application, a single gradient-â•‰echo
sequence is performed, and magnitude and phase images are generated. Two post-â•‰
processing schemes are then used to enhance the image contrast so that veins and
bleeds are conspicuous (Haacke et al., 2004; Haacke et al., 2009). One scheme is to
conduct multiplication of the magnitude images using the phase image. This is usually
repeated three to four times to achieve an optimal result. With this scheme, the venous
vessel regions, which already have relatively low signal intensity in the magnitude
image due to short T2*, are further darkened because phase of a vein is character-
istically different from that of an artery or tissue (Figure 1.7). A second scheme to
enhance the image contrast is to conduct an image-processing step termed minimum
intensity projection (mIP). In mIP, the signal intensities along several consecutive
28
Phase
unwrapping
Phase image Phase image
Final mIP image
Magnitude in image
Figure 1.7 An example of images obtained at 7T using Susceptibility Weighted Imaging

(SWI). After phase unwrapping, the phase image is multiplied four times onto the magnitude
image so that the venous vessel regions are darkened in the processed image. The final image
shown here is the minimum intensity projection (mIP) image of 8 slices centered on the dis-
played magnitude and phase images, in which venous voxels are further darkened (Courtesy of
Dr. Yulin Ge, New York University).
slices are compared, and the lowest signal value is assigned to the new mIP image.
Therefore, in the mIP image, venous voxels are further darkened (Figure 1.7).
The above approach is useful for qualitative examination of brain image, as the
image contrast is visually conspicuous. However, it lacks the ability to quantify a
physical or physiological parameter because both phase multiplication and minimum
intensity projection are nonlinear operations to the original image, and the resulting
new image has no direct relationship to the underlying tissue property. For quan-
titative assessment of brain tissue properties, for example iron concentration, other
approaches are needed. One approach is to use the phase value as an approximation
of the magnetic susceptibility, which is thought to be directly related to iron content
(Haacke et al., 2010). Another approach is to use a multi-echo acquisition to obtain the
magnitude signal as a function of echo time (TE), from which an exponential model
to obtain the decay constant, T2*. T2* is also closely associated with iron content
(Rodrigue et al., 2011)
Recently, a more advanced SWI implementation was developed. This method is
termed quantitative susceptibility mapping (QSM), which aims to provide a direct,
quantitative estimation of the magnetic field source inside a voxel, commonly referred
to as magnetic susceptibility (Li and Leigh, 2004; Bilgic et al., 2012). In QSM, the
image acquisition schemes are similar to a conventional SWI. A single-echo or multi-
echo 3D gradient-echo sequence is used, and phase images are obtained. The primary
â•‡ 29
technical advance of QSM lies in the post-â•‰processing algorithms. The main motiva-
tion behind QSM is that a routine phase image does not provide a direct assessment of
quantity of magnetic susceptibility (e.g., amount of iron, amount of calcium, amount
of deoxyhemoglobin, or amount of hemosiderin) in the region of interest. Instead, it
represents a result of convolution between the magnetic susceptibility and an impulse
response. QSM therefore aims to solve the so-â•‰called “inverse problem,” to estimate
magnetic susceptibility from the phase image. This is an ill-â•‰conditioned problem.
It is an active area of research in the MRI field, and various algorithms have been
developed by using different constraints (de Rochefort et al., 2008; Liu et al., 2009;
de Rochefort et al., 2010; Liu et al., 2012).
Conclusion
MRI is a powerful imaging modality that has enormous potential in studies of cerebral
aging. Over the past decade, MRI technologies have undergone rapid progress in both
image quality and acquisition speed. Due to its nonâ•‰invasive nature and the absence of
radiation, the translation of these new methodologies to clinical and cognitive applica-
tions has also been amazingly fast. Many new imaging techniques that were viewed
as emerging methods only a few years ago have now become routine. The MRI ven-
dors have also enthusiastically adopted these new technologies and made them more
widely available to researchers and clinicians. Therefore, MRI has become one of the
most important tools in cognitive neuroscience and cerebral aging.
References
Aganj I, Lenglet C, Sapiro G, Yacoub E, Ugurbil K, Harel N (2010) Reconstruction of the ori-
entation distribution function in single-â•‰and multiple-â•‰shell q-â•‰ball imaging within constant
solid angle. Magn Reson Med 64:554–â•‰566.
Alsop DC, Detre JA (1996) Reduced transit-â•‰time sensitivity in noninvasive magnetic reso-
nance imaging of human cerebral blood flow. Journal of cerebral blood flow and metabo-
lism: official journal of the International Society of Cerebral Blood Flow and Metabolism
16:1236–â•‰1249.
Alsop DC, Casement M, de Bazelaire C, Fong T, Press DZ (2008) Hippocampal hyperperfusion
in Alzheimer’s disease. NeuroImage 42:1267–â•‰1274.
Alsop DC, Detre JA, Golay X, Gunther M, Hendrikse J, Hernandez-â•‰Garcia L, Lu H, Macintosh
BJ, Parkes LM, Smits M, van Osch MJ, Wang DJ, Wong EC, Zaharchuk G (2014)
Recommended implementation of arterial spin-â•‰labeled perfusion MRI for clinical applica-
tions: A consensus of the ISMRM perfusion study group and the European consortium for
ASL in dementia. Magn Reson Med 73:102–116.
Ances BM, Liang CL, Leontiev O, Perthen JE, Fleisher AS, Lansing AE, Buxton RB (2009)
Effects of aging on cerebral blood flow, oxygen metabolism, and blood oxygenation level
dependent responses to visual stimulation. Human Brain Mapping 30:1120–â•‰1132.
Arthurs OJ, Boniface S (2002) How well do we understand the neural origins of the fMRI
BOLD signal? Trends in Neurosciences 25:27–â•‰31.
Bandettini PA, Wong EC, Hinks RS, Tikofsky RS, Hyde JS (1992) Time course EPI of human
brain function during task activation. Magn Reson Med 25:390–â•‰397.
30
Basser PJ, Mattiello J, LeBihan D (1994) MR diffusion tensor spectroscopy and imaging.
Biophysical Journal 66:259–267.
Bilgic B, Pfefferbaum A, Rohlfing T, Sullivan EV, Adalsteinsson E (2012) MRI estimates
of brain iron concentration in normal aging using quantitative susceptibility mapping.
NeuroImage 59:2625–2635.
Bradley WG, Jr. (2008) Pros and cons of 3 tesla MRI. Journal of the American College of
Radiology: JACR 5:871–878.
Chalela JA, Alsop DC, Gonzalez-Atavales JB, Maldjian JA, Kasner SE, Detre JA (2000)
Magnetic resonance perfusion imaging in acute ischemic stroke using continuous arterial
spin labeling. Stroke 31:680–687.
Chapman SB, Aslan S, Spence JS, Hart JJ, Jr., Bartz EK, Didehbani N, Keebler MW, Gardner
CM, Strain JF, Defina LF, Lu H (2015) Neural Mechanisms of Brain Plasticity with
Complex Cognitive Training in Healthy Seniors. Cerebral Cortex 25:396–405.
Chen JJ, Rosas HD, Salat DH (2011) Age-associated reductions in cerebral blood flow are
independent from regional atrophy. NeuroImage 55:468–478.
Chen JJ, Rosas HD, Salat DH (2013) The relationship between cortical blood flow and sub-
cortical white-matter health across the adult age span. PloS One 8:e56733.
D’Esposito M, Zarahn E, Aguirre GK, Rypma B (1999) The effect of normal aging on the cou-
pling of neural activity to the bold hemodynamic response. NeuroImage 10:6–14.
Dai W, Garcia D, de Bazelaire C, Alsop DC (2008) Continuous flow-driven inversion for arte-
rial spin labeling using pulsed radio frequency and gradient fields. Magn Reson Med
60:1488–1497.
Davis TL, Kwong KK, Weisskoff RM, Rosen BR (1998) Calibrated functional MRI: mapping
the dynamics of oxidative metabolism. Proceedings of the National Academy of Sciences
of the United States of America 95:1834–1839.
de Boer R, Vrooman HA, van der Lijn F, Vernooij MW, Ikram MA, van der Lugt A, Breteler
MM, Niessen WJ (2009) White matter lesion extension to automatic brain tissue segmen-
tation on MRI. NeuroImage 45:1151–1161.
de Leeuw FE, de Groot JC, Achten E, Oudkerk M, Ramos LM, Heijboer R, Hofman A, Jolles
J, van Gijn J, Breteler MM (2001) Prevalence of cerebral white matter lesions in elderly
people: a population based magnetic resonance imaging study. The Rotterdam Scan Study.
Journal of Neurology, Neurosurgery, and Psychiatry 70:9–14.
de Rochefort L, Brown R, Prince MR, Wang Y (2008) Quantitative MR susceptibility mapping
using piece-wise constant regularized inversion of the magnetic field. Magn Reson Med
60:1003–1009.
de Rochefort L, Liu T, Kressler B, Liu J, Spincemaille P, Lebon V, Wu J, Wang Y (2010)
Quantitative susceptibility map reconstruction from MR phase data using bayesian regu-
larization: validation and application to brain imaging. Magn Reson Med 63:194–206.
Descoteaux M, Deriche R, Le Bihan D, Mangin JF, Poupon C (2011) Multiple q-shell diffusion
propagator imaging. Medical Image Analysis 15:603–621.
Feinberg DA, Moeller S, Smith SM, Auerbach E, Ramanna S, Gunther M, Glasser MF, Miller
KL, Ugurbil K, Yacoub E (2010) Multiplexed echo planar imaging for sub-second whole
brain FMRI and fast diffusion imaging. PloS One 5:e15710.
Gauthier CJ, Madjar C, Desjardins-Crepeau L, Bellec P, Bherer L, Hoge RD (2013) Age depen-
dence of hemodynamic response characteristics in human functional magnetic resonance
imaging. Neurobiology of Aging 34:1469–1485.
Ge Y, Grossman RI, Babb JS, Rabin ML, Mannon LJ, Kolson DL (2002) Age-related total gray
matter and white matter changes in normal adult brain. Part I: volumetric MR imaging
analysis. AJNR American Journal of Neuroradiology 23:1327–1333.
31
Golay X, Stuber M, Pruessmann KP, Meier D, Boesiger P (1999) Transfer insensitive label-
ing technique (TILT): application to multislice functional perfusion imaging. Journal of
Magnetic Resonance Imaging JMRI 9:454–461.
Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS (2001) A voxel-
based morphometric study of ageing in 465 normal adult human brains. NeuroImage
14:21–36.
Gorelick PB et al. (2011) Vascular contributions to cognitive impairment and dementia: a state-
ment for healthcare professionals from the american heart association/american stroke
association. Stroke; a Journal of Cerebral Circulation 42:2672–2713.
Grydeland H, Walhovd KB, Tamnes CK, Westlye LT, Fjell AM (2013) Intracortical myelin
links with performance variability across the human lifespan: results from T1-and T2-
weighted MRI myelin mapping and diffusion tensor imaging. The Journal of neurosci-
ence: the official journal of the Society for Neuroscience 33:18618–18630.
Gunning-Dixon FM, Raz N (2000) The cognitive correlates of white matter abnormalities in
normal aging: a quantitative review. Neuropsychology 14:224–232.
Haacke EM, Xu Y, Cheng YC, Reichenbach JR (2004) Susceptibility weighted imaging (SWI).
Magn Reson Med 52:612–618.
Haacke EM, Tang J, Neelavalli J, Cheng YC (2010) Susceptibility mapping as a means to
visualize veins and quantify oxygen saturation. Journal of Magnetic Resonance Imaging:
JMRI 32:663–676.
Haacke EM, Mittal S, Wu Z, Neelavalli J, Cheng YC (2009) Susceptibility-weighted imag-
ing: technical aspects and clinical applications, part 1. AJNR American Journal of
Neuroradiology 30:19–30.
Herscovitch P, Raichle ME (1985) What is the correct value for the brain--blood partition coef-
ficient for water? Journal of cerebral blood flow and metabolism: official journal of the
International Society of Cerebral Blood Flow and Metabolism 5:65–69.
Hoge RD, Atkinson J, Gill B, Crelier GR, Marrett S, Pike GB (1999) Investigation of BOLD
signal dependence on cerebral blood flow and oxygen consumption: the deoxyhemoglobin
dilution model. Magn Reson Med 42:849–863.
Hulsey KM, Gupta M, King KS, Peshock RM, Whittemore AR, McColl RW (2012) Automated
quantification of white matter disease extent at 3 T: comparison with volumetric readings.
Journal of Magnetic Resonance Imaging: JMRI 36:305–311.
Hutchison JL, Lu H, Rypma B (2013) Neural mechanisms of age-related slowing: the DeltaCBF/
DeltaCMRO2 ratio mediates age-differences in BOLD signal and human performance.
Cerebral Cortex 23:2337–2346.
Jensen JH, Helpern JA, Ramani A, Lu H, Kaczynski K (2005) Diffusional kurtosis imaging: the
quantification of non-gaussian water diffusion by means of magnetic resonance imaging.
Magnetic resonance in medicine: official journal of the Society of Magnetic Resonance in
Medicine/Society of Magnetic Resonance in Medicine 53:1432–1440.
Jezzard P, Buxton RB (2006) The clinical potential of functional magnetic resonance imaging.
Journal of Magnetic Resonance Imaging: JMRI 23:787–793.
Jones DK, Horsfield MA, Simmons A (1999) Optimal strategies for measuring diffusion in
anisotropic systems by magnetic resonance imaging. Magn Reson Med 42:515–525.
Kannurpatti SS, Motes MA, Rypma B, Biswal BB (2010) Neural and vascular variability and
the fMRI-BOLD response in normal aging. Magnetic Resonance Imaging 28:466–476.
Kastrup A, Li TQ, Glover GH, Moseley ME (1999) Cerebral blood flow-related signal changes
during breath-holding. AJNR American Journal of Neuroradiology 20:1233–1238.
Kidwell CS, Saver JL, Villablanca JP, Duckwiler G, Fredieu A, Gough K, Leary MC, Starkman
S, Gobin YP, Jahan R, Vespa P, Liebeskind DS, Alger JR, Vinuela F (2002) Magnetic
32
resonance imaging detection of microbleeds before thrombolysis: an emerging applica-

tion. Stroke; a Journal of Cerebral Circulation 33:95–98.
Kim SG (1995) Quantification of relative cerebral blood flow change by flow-sensitive alternat-
ing inversion recovery (FAIR) technique: application to functional mapping. Magn Reson
Med 34:293–301.
Kwong KK, Belliveau JW, Chesler DA, Goldberg IE, Weisskoff RM, Poncelet BP, Kennedy
DN, Hoppel BE, Cohen MS, Turner R, et al. (1992) Dynamic magnetic resonance imaging
of human brain activity during primary sensory stimulation. Proceedings of the National
Academy of Sciences of the United States of America 89:5675–5679.
Lao Z, Shen D, Liu D, Jawad AF, Melhem ER, Launer LJ, Bryan RN, Davatzikos C (2008)
Computer-assisted segmentation of white matter lesions in 3D MR images using support
vector machine. Academic Radiology 15:300–313.
Le Bihan D, Mangin JF, Poupon C, Clark CA, Pappata S, Molko N, Chabriat H (2001) Diffusion
tensor imaging: concepts and applications. Journal of Magnetic Resonance Imaging: JMRI
13:534–546.
Li L, Leigh JS (2004) Quantifying arbitrary magnetic susceptibility distributions with MR.
Lin C, Bernstein MA, Huston J, Fain S (2001) Measurements of T1 Relaxation times at 3.0T:
Implications for clinical MRA. In: Proceedings of the 9th Annual Meeting of ISMRM, p.
1391. Glasgow, UK.
Liu J, Liu T, de Rochefort L, Ledoux J, Khalidov I, Chen W, Tsiouris AJ, Wisnieff C,
Spincemaille P, Prince MR, Wang Y (2012) Morphology enabled dipole inversion for
quantitative susceptibility mapping using structural consistency between the magnitude
image and the susceptibility map. NeuroImage 59:2560–2568.
Liu P, Uh J, Lu H (2011) Determination of spin compartment in arterial spin labeling MRI.
Liu P, Hebrank AC, Rodrigue KM, Kennedy KM, Park DC, Lu H (2013a) A comparison of
physiologic modulators of fMRI signals. Human Brain Mapping 34:2078–2088.
Liu P, Hebrank AC, Rodrigue KM, Kennedy KM, Section J, Park DC, Lu H (2013b) Age-
related differences in memory-encoding fMRI responses after accounting for decline in
vascular reactivity. NeuroImage 78:415–425.
Liu T, Spincemaille P, de Rochefort L, Kressler B, Wang Y (2009) Calculation of susceptibility
through multiple orientation sampling (COSMOS): a method for conditioning the inverse
problem from measured magnetic field map to susceptibility source image in MRI. Magn
Reson Med 61:196–204.
Lu H, Nagae-Poetscher LM, Golay X, Lin D, Pomper M, van Zijl PC (2005) Routine clinical
brain MRI sequences for use at 3.0 Tesla. Journal of Magnetic Resonance Imaging: JMRI
22:13–22.
Lu H, Xu F, Rodrigue KM, Kennedy KM, Cheng Y, Flicker B, Hebrank AC, Uh J, Park DC
(2011) Alterations in cerebral metabolic rate and blood supply across the adult lifespan.
Mintun MA, Raichle ME, Martin WR, Herscovitch P (1984) Brain oxygen utilization measured
with O-15 radiotracers and positron emission tomography. J Nucl Med 25:177–187.
Mori S, Crain BJ, Chacko VP, van Zijl PC (1999) Three-dimensional tracking of axonal projec-
tions in the brain by magnetic resonance imaging. Annals of Neurology 45:265–269.
Mueller D, Sauerland S, Neugebauer EA, Immenroth M (2010) Reported effects in randomized
controlled trials were compared with those of nonrandomized trials in cholecystectomy. J
Clin Epidemiol 63:1082–1090.
Mugler JP, 3rd, Brookeman JR (1991) Rapid three-dimensional T1-weighted MR imaging with
the MP-RAGE sequence. Journal of Magnetic Resonance Imaging: JMRI 1:561–567.
33
Ogawa S, Lee TM, Kay AR, Tank DW (1990) Brain magnetic resonance imaging with con-
trast dependent on blood oxygenation. Proceedings of the National Academy of Sciences
87:9868–9872.
Ogawa S, Tank DW, Menon R, Ellermann JM, Kim SG, Merkle H, Ugurbil K (1992) Intrinsic
signal changes accompanying sensory stimulation: functional brain mapping with mag-
netic resonance imaging. Proceedings of the National Academy of Sciences of the United
States of America 89:5951–5955.
Ostergaard L, Weisskoff RM, Chesler DA, Gyldensted C, Rosen BR (1996a) High resolu-
tion measurement of cerebral blood flow using intravascular tracer bolus passages. Part
I: Mathematical approach and statistical analysis. Magn Reson Med 36:715–725.
Ostergaard L, Sorensen AG, Kwong KK, Weisskoff RM, Gyldensted C, Rosen BR (1996b) High
resolution measurement of cerebral blood flow using intravascular tracer bolus passages.
Part II: Experimental comparison and preliminary results. Magn Reson Med 36:726–736.
Peng SL, Dumas JA, Park DC, Liu P, Filbey FM, McAdams CJ, Pinkham AE, Adinoff B,
Zhang R, Lu H (2014) Age-related increase of resting metabolic rate in the human brain.
Pfefferbaum A, Mathalon DH, Sullivan EV, Rawles JM, Zipursky RB, Lim KO (1994) A quan-
titative magnetic resonance imaging study of changes in brain morphology from infancy
to late adulthood. Archives of Neurology 51:874–887.
Poels MM, Vernooij MW, Ikram MA, Hofman A, Krestin GP, van der Lugt A, Breteler MM
(2010) Prevalence and risk factors of cerebral microbleeds: an update of the Rotterdam
scan study. Stroke; a Journal of Cerebral Circulation 41:S103–S106.
Rane S, Ally BA, Hussey E, Wilson T, Thornton-Wells T, Gore JC, Donahue MJ (2013) Inverse
correspondence between hippocampal perfusion and verbal memory performance in older
adults. Hippocampus 23:213–220.
Raz N, Lindenberger U, Rodrigue KM, Kennedy KM, Head D, Williamson A, Dahle C, Gerstorf
D, Acker JD (2005) Regional brain changes in aging healthy adults: general trends, indi-
vidual differences and modifiers. Cerebral Cortex 15:1676–1689.
Rodrigue KM, Haacke EM, Raz N (2011) Differential effects of age and history of hyperten-
sion on regional brain volumes and iron. NeuroImage 54:750–759.
Schuff N, Matsumoto S, Kmiecik J, Studholme C, Du A, Ezekiel F, Miller BL, Kramer JH,
Jagust WJ, Chui HC, Weiner MW (2009) Cerebral blood flow in ischemic vascular demen-
tia and Alzheimer’s disease, measured by arterial spin-labeling magnetic resonance imag-
ing. Alzheimer’s & dementia: the journal of the Alzheimer’s Association 5:454–462.
Sehgal V, Delproposto Z, Haacke EM, Tong KA, Wycliffe N, Kido DK, Xu Y, Neelavalli J,
Haddar D, Reichenbach JR (2005) Clinical applications of neuroimaging with susceptibil-
ity-weighted imaging. Journal of Magnetic Resonance Imaging: JMRI 22:439–450.
Setsompop K, Gagoski BA, Polimeni JR, Witzel T, Wedeen VJ, Wald LL (2012) Blipped-
controlled aliasing in parallel imaging for simultaneous multislice echo planar imaging
with reduced g-factor penalty. Magnetic Resonance in Medicine: Official Journal of the
Society of Magnetic Resonance in Medicine/Society of Magnetic Resonance in Medicine
67:1210–1224.
Skare S, Hedehus M, Moseley ME, Li TQ (2000) Condition number as a measure of noise
performance of diffusion tensor data acquisition schemes with MRI. Journal of Magnetic
Resonance 147:340–352.
Slessarev M, Han J, Mardimae A, Prisman E, Preiss D, Volgyesi G, Ansel C, Duffin J, Fisher
JA (2007) Prospective targeting and control of end-tidal CO2 and O2 concentrations. The
Journal of Physiology 581:1207–1219.
Sorensen AG, Copen WA, Ostergaard L, Buonanno FS, Gonzalez RG, Rordorf G, Rosen BR,
Schwamm LH, Weisskoff RM, Koroshetz WJ (1999) Hyperacute stroke: simultaneous
34
measurement of relative cerebral blood volume, relative cerebral blood flow, and mean
tissue transit time. Radiology 210:519–527.
Sotiropoulos SN, Jbabdi S, Xu J, Andersson JL, Moeller S, Auerbach EJ, Glasser MF,
Hernandez M, Sapiro G, Jenkinson M, Feinberg DA, Yacoub E, Lenglet C, Van
Essen DC, Ugurbil K, Behrens TE, Consortium WU-MH (2013) Advances in diffu-
sion MRI acquisition and processing in the Human Connectome Project. NeuroImage
80:125–143.
Sowell ER, Peterson BS, Thompson PM, Welcome SE, Henkenius AL, Toga AW (2003)
Mapping cortical change across the human life span. Nature Neuroscience 6:309–315.
Thomas BP, Yezhuvath US, Tseng BY, Liu P, Levine BD, Zhang R, Lu H (2013) Life-long
aerobic exercise preserved baseline cerebral blood flow but reduced vascular reactivity to
CO2. Journal of Magnetic Resonance Imaging: JMRI 38:1177–1183.
Tuch DS, Reese TG, Wiegell MR, Wedeen VJ (2003) Diffusion MRI of complex neural archi-
tecture. Neuron 40:885–895.
Van Essen DC et al. (2012) The Human Connectome Project: a data acquisition perspective.
Wahlund LO, Barkhof F, Fazekas F, Bronge L, Augustin M, Sjogren M, Wallin A, Ader H, Leys
D, Pantoni L, Pasquier F, Erkinjuntti T, Scheltens P, European Task Force on Age-Related
White Matter C (2001) A new rating scale for age-related white matter changes applicable
to MRI and CT. Stroke; a Journal of Cerebral Circulation 32:1318–1322.
Wang J, Zhang Y, Wolf RL, Roc AC, Alsop DC, Detre JA (2005) Amplitude-modulated continu-
ous arterial spin-labeling 3.0-T perfusion MR imaging with a single coil: feasibility study.
Radiology 235:218–228.
Wierenga CE, Clark LR, Dev SI, Shin DD, Jurick SM, Rissman RA, Liu TT, Bondi MW
(2013) Interaction of age and APOE genotype on cerebral blood flow at rest. Journal of
Alzheimer’s Disease: JAD 34:921–935.
Williams DS, Detre JA, Leigh JS, Koretsky AP (1992) Magnetic resonance imaging of per-
fusion using spin inversion of arterial water. Proceedings of the National Academy of
Sciences of the United States of America 89:212–216.
Willinek WA, Kuhl CK (2006) 3.0 T neuroimaging: technical considerations and clinical appli-
cations. Neuroimaging Clinics of North America 16:217–228, ix.
Wise RG, Pattinson KT, Bulte DP, Chiarelli PA, Mayhew SD, Balanos GM, O’Connor DF,
Pragnell TR, Robbins PA, Tracey I, Jezzard P (2007) Dynamic forcing of end-tidal carbon
dioxide and oxygen applied to functional magnetic resonance imaging. Journal of Cerebral
Blood Flow and Metabolism: Official Journal of the International Society of Cerebral
Blood Flow and Metabolism 27:1521–1532.
Wong EC, Buxton RB, Frank LR (1998) A theoretical and experimental comparison of continu-
ous and pulsed arterial spin labeling techniques for quantitative perfusion imaging. Magn
Reson Med 40:348–355.
Yang Y, Frank JA, Hou L, Ye FQ, McLaughlin AC, Duyn JH (1998) Multislice imaging of quan-
titative cerebral perfusion with pulsed arterial spin labeling. Magn Reson Med 39:825–832.
Yezhuvath US, Lewis-Amezcua K, Varghese R, Xiao G, Lu H (2009) On the assessment of cere-
brovascular reactivity using hypercapnia BOLD MRI. NMR in Biomedicine 22:779–786.
Zecca L, Youdim MB, Riederer P, Connor JR, Crichton RR (2004) Iron, brain ageing and neu-
rodegenerative disorders. Nature Reviews Neuroscience 5:863–873.
Zijdenbos AP, Dawant BM, Margolin RA, Palmer AC (1994) Morphometric analysis of white
matter lesions in MR images: method and validation. IEEE Transactions on Medical
Imaging 13:716–724.
Zlokovic BV (2011) Neurovascular pathways to neurodegeneration in Alzheimer’s disease and
other disorders. Nature Reviews Neuroscience 12:723–738.
35
Molecular Imaging of Aging

and Neurodegenerative Disease
Anna Rieckmann
Randy L. Buckner
Trey Hedden
T he study of the cognitive neuroscience of aging has evolved to include

focus on molecular and cellular processes in the brain that develop across
the lifespan and affect cognition. While the most common tools to study neural cas-
cades have been noninvasive techniques such as magnetic resonance imaging (MRI)
and electroencephalography (EEG), these techniques measure the consequences of
molecular and cellular processes, including macroscopic atrophy on structural MRI,
white-matter integrity with diffusion-weighted MRI, neuronal activity disruptions
through the indirect window of functional MRI hemodynamics, and glimpses of neu-
ronal activity that can be detected by EEG on the scalp surface. A widely available
in vivo technique that can visualize how the components of age-associated molecu-
lar processes are distributed throughout the brain is positron emission tomography
(PET). PET makes use of a radiopharmaceutical agent that is injected into a person’s
blood-stream, crosses the blood–brain barrier into the brain, and binds to a specific
molecular target of interest. Positron decay emits a signal which is detected by the
PET scanner and computer-reconstructed into an image to reveal the spatial distribu-
tion and concentration of the ligand in the person’s brain. With respect to the study
of brain functions in aging, molecules of primary interest are those related to aspects
of synaptic functions, age-associated pathology, and markers that reflect responses to
pathology. PET studies are necessarily limited by important technical considerations
including the specificity of the agent to the target molecule, uptake and clearance of
35
36
the agent, partial volume effects, co-â•‰registration across modalities in the presence of
atrophy, and the limits of resolution.
In this chapter we concentrate on a few of the most common molecular targets as
illustrative examples of the integral role that PET imaging plays in understanding the
cognitive neuroscience of aging. Our focus will be on more recent studies, with atten-
tion directed toward implications for the study of cognitively healthy aging and likely
future directions in that study.
Two themes will run throughout the chapter. First is an interest in the extent to
which PET imaging can inform the distinction between developmental and pathologi-
cal processes that occur during aging. While the boundaries of aging and dementia
pathology remain difficult to separate (Swerdlow, 2007), we believe it worthwhile
to view aging as characterized by a multifaceted set of progressive processes that
occur at different rates in different people (Buckner, 2004; Hedden & Gabrieli, 2004).
Although some of these processes may be the result of developing pathology, such as
preclinical Alzheimer’s disease (AD) or incipient cerebrovascular disease, others may
have similarities to certain diseases while nonetheless arising from a different etiology
than the pathological pathway linked to disease, such as dopamine loss in Parkinson’s
disease. A recent review by Jagust (2013) discusses the distinction between aging and
disease-â•‰related processes in detail, and we refer interested readers to this work. We
view the distinction as useful in that it sharpens our exploration for distinguishable
facets of age-â•‰related changes in the brain and their complex, interdependent cognitive
sequelae.
A second theme will be the integration of PET imaging with other neuroscience
techniques to build a fuller picture of how individuals vary in their progression along
different pathways as they age. We therefore relate the results from PET studies to
those using other imaging modalities. Although PET may reveal a molecular pathway
that ultimately impacts cognition, it may be that an alteration in regional brain struc-
ture or function acts as an intermediary along this pathway, such that detection of the
intermediary provides more predictive value of the cognitive outcome.
An important future direction that unites these two themes will be the utilization
of multiple markers of brain health to predict which aging individuals are at risk for
progression of different processes so that interventions can be more precisely targeted
toward those who will receive the greatest benefit.
Cerebral Glucose Metabolism
The most widely used PET radioligand is 18F[fluorodeoxyglucose] (FDG), a radio-

actively labeled glucose analog that enters the pathways for natural glucose consump-
tion in the brain (Phelps et al., 1979) and is a marker of a cell’s metabolic rate of
glucose. The brain utilizes a disproportionately high amount of glucose, around 25%
of total body glucose, motivating the use of FDG imaging in the study of brain func-
tion over the last 40 years. FDG studies in aging and AD have been recently covered
in excellent reviews (e.g., Mosconi, 2013; Cohen and Klunk, 2014; Shokouhi et al.,
2014). Here, we highlight a few recent approaches that open up future avenues of
research in the cognitive neuroscience of aging.
â•‡ 37
Molecular Imaging of Aging and Neurodegenerative Diseaseâ•… 37
PET measures of glucose consumption are often equated with a general measure
of synaptic function, including both synapse density (McGeer et al. 1986, 1990) as
well as synapse activity (Schwartz et al., 1979). The primary research focus in aging
and AD over the last 20 years has been on hypometabolism within a distinct set of
regions comprising temporoparietal cortex and posterior cingulate (Figure 2.1A).
These regions display selective hypometabolism in cross-â•‰sectional group comparisons
between AD patients and cognitively healthy controls (Landau et al., 2011; Bohnen
et al., 2012 for review). Longitudinal studies have identified that hypometabolism
in AD-â•‰vulnerable regions predicts cognitive decline and progression to AD and is
consistent with pathological verification of AD diagnosis, which demonstrates the
usefulness of FDG PET as a preclinical biomarker of AD (e.g., Hoffmann et al., 2000;
De Leon et al., 2001; Herholz et al., 2002; Jagust et al., 2006; Mosconi et al., 2008;
2009; 2010; Ewers et al., 2014).
Glucose Hypometabolism and AD Pathology

Use of multiple PET markers within in the same individuals has provided insight about
links between pathological processes and functional disruption. Multi-â•‰tracer stud-
ies suggest that in cognitively healthy individuals with measurable amyloid burden,
regional hypometabolism in AD signature regions is associated with amyloid burden
in a dose-â•‰dependent fashion (Lowe et al., 2014). However, large-â•‰scale studies from
different centers find that AD-â•‰like hypometabolism in patients with mild cognitive
impairment and AD can occur without measurable amyloid burden (Jack et al., 2012).
These individuals have been termed “suspected non-â•‰AD pathology” (SNAP) and can
also be identified in about 20% of clinically healthy older adults (e.g., Knopman et
al., 2013). The neuropathologies underlying hypometabolism without amyloid burden
remain to be identified. One hypothesis may be that hypometabolism in AD-â•‰vulner-
able regions reflects neurodegenerative effects of the APOE4 allele that are indepen-
dent of its well-â•‰established relationship to amyloid deposition (e.g., Kim et al., 2009;
Morris et al., 2010). Supporting evidence suggests that APOE4 genotype is associated
with regional hypometabolism independent of amyloid burden (Jagust et al., 2012;
Ossenkoppele et al., 2013; Knopman et al., 2014). Another emerging hypothesis is
that tau pathology in aging is associated with hypometabolism. FDG PET changes
over time mediate the association between tau markers in cerebrospinal fluid and
cognitive decline (Dowling et al., 2015). Human PET imaging of neurofibrillary tau
tangles remains in its infancy, but a single case study combining FDG imaging with
amyloid and tau imaging found binding to tau, but not to amyloid, spatially over-
lapped with hypometabolism in areas affected by disease (Ossenkoppele et al., 2015).
New developments in tau imaging are reviewed in a later section in this chapter.
FDG Studies in Cognitively Healthy Aging

Outside of the AD-â•‰vulnerable regions, cross-â•‰sectional studies in individuals span-
ning the adult lifespan show widespread decreases in glucose metabolism in frontal
and parietal association cortices (e.g., Loessner et al., 1995; Kalpouzos et al., 2009).
38
(A) Glucose metabolism (18F-FDG) (B) Amyloid burden (11C-PIB) (C) Tau burden (18F-T807)
Normal metabolism Regional hypometabolism Low amyloid High amyloid Low tau High tau
SUVR DVR DVR
0 1.2 0 1.5 0 2.0
Figure 2.1 A) Example images for FDG-PET standardized uptake values (SUVR) for two clinically healthy older adults (age > 70). The person
on the left shows typical cortical FDG signal, the person on the right has evidence of temporo-parietal hypometabolism. B and C) Distribution
Volume Ratio (DVR; Logan et al. 1990) images for clinically healthy older adults showing evidence for low and high amyloid burden as
measured with 11C-PIB (B) and tau burden as measured with 18F-T807 (C). PET signal for all tracers is standardized with respect to the cerebellum.
(See color plate also)
39
Molecular Imaging of Aging and Neurodegenerative Disease 39
It is likely that a portion of this diffuse decrease in FDG signal is due to partial vol-
ume effects (e.g., Ibanez et al., 2004; Yananse et al., 2005); but even when methods
to control partial volume effects are used, subtle hypometabolism in frontal cortex
have been observed (Knopman et al., 2014). A portion of frontal hypometabolism
appears to be unrelated to AD pathology (e.g., amyloid plaques; Lowe et al., 2014),
and the implications of frontal hypometabolism for cognitive aging are not well
established.
Several studies have demonstrated that age-related frontal hypometabolism may be
associated with deteriorating white-matter microstructure as measured with diffusion
tensor imaging (e.g., Inoue et al., 2008; Kochunov et al., 2009; Chételat, 2013). This
is an intriguing finding. White-matter integrity is an important correlate of age-related
decline in executive functions and processing speed in cognitively healthy aging (e.g.,
DeCarli et al., 1995; Madden et al., 2009; Lövdén et al., 2014; Hedden et al., 2016),
and these studies suggest that frontal hypometabolism may play a role in this rela-
tionship. Chételat et al. (2013) investigated whether frontal hypometabolism in older
adults was associated with fMRI functional connectivity of the hypometabolic regions
with the rest of the cerebral cortex but found no evidence for this association. One
reason for the absence of this relationship may be that these two measures operate
on different time scales. Functional connectivity MRI measures synchronized fluc-
tuations of fMRI signal at less than 0.1 Hz and is widely used to probe the function
and integrity of neural networks (Fox and Raichle, 2007; Buckner et al., 2013). In its
current use, FDG offers a snapshot of energy consumption over an extended period of
time with little information on state changes (i.e., fluctuations over seconds or even
minutes). Although, this technical possibility does not explain conceptually why puta-
tive markers of regional and network dysfunction do not converge. Each is believed to
be assaying distinct aspects of network dysfunction.
Two recent methodological advances open new avenues for research into the rela-
tionship between local glucose metabolism as a measure of synaptic integrity and
functional connectivity MRI of large-scale brain networks. Simultaneous PET-MRI
measurements now allow for time-synced measurements of FDG PET signal and
fMRI. Riedl et al. (2014) tested simultaneous FDG PET and resting-state fMRI in
two conditions (eyes open and eyes closed) in a between-subject design. Subjects
tested with eyes open had higher glucose metabolism in visual cortex, anterior insu-
lar cortex, and anterior cingulate, and stronger functional connectivity between these
regions compared to subjects tested with eyes closed. Hence, local glucose metabo-
lism may be a correlate of large-scale neural network activity dynamics. A second
method that could be used in conjunction with simultaneous PET-MRI measurements
allows investigation of changes in glucose metabolism within a scan, on a timescale
comparable to that of an fMRI block design (fPET-FDG, Villien et al., 2014). In this
functional PET method, the brain is supplied with a constant infusion of FDG rather
than the traditional bolus injection, allowing detection of subtle changes in local glu-
cose metabolism in response to task alternations. Combining simultaneous P ET-MRI
and fPET-FDG in future studies of aging and various cognitive tasks will provide
an exciting new opportunity to carefully investigate the metabolic origins of altered
fMRI activity and connectivity of large-scale association networks during aging.
40
Amyloid Burden in Cognitively Healthy Aging
Major strides in understanding the contributions of disease pathology to cognitive

aging have been made over the last decade owing to the development of novel classes
of PET ligands. Since its initial description of rapid memory loss accompanied by
“miliary foci” (later known as neurotic plaques) and neurofibrillary tangles by Alois
Alzheimer in 1906 and 1907 (Alzheimer, 1906; 1907; Maurer et al., 1997), what has
come to be known as Alzheimer’s disease is characterized by classic amnestic behav-
ioral symptoms. But a definitive diagnosis from the hallmark plaques and tangles
could only be confirmed by examination at autopsy until recently. Assays of oligo-
meric amyloid-â•‰beta chains in cerebrospinal fluid can detect lower than usual levels
that signal the likely accumulation into neurotic plaques (Andreasen et al., 1999),
but the presence and location of the plaques must still be confirmed at autopsy. In
2004, researchers at the University of Pittsburgh and Uppsala University led by Bill
Klunk and Chet Mathis reported a PET agent that binds to fibrillar amyloid plaques,
revealing the presence and pattern of these hallmark plaques in the living human brain
(Klunk et al., 2004). This represented an important step toward using brain markers
to definitively diagnose AD in living patients and to evaluating the ability of potential
treatments to reduce or eliminate the plaques.
The compound they developed is a derivative of the amyloid-â•‰beta binding agent
thioflavin-â•‰T labeled with carbon-â•‰11 that has become known as Pittsburgh Compound
B (PIB). PIB has desirable properties, including high affinity, rapid penetration and
clearance of nonâ•‰specific binding, and construct validity in that binding is elevated
in AD (Klunk et al., 2003, 2004; Rabinovici et al., 2007). Several later agents have
been developed, with the most widely adopted of this class being fluorine-â•‰18 labeled
agents, providing a longer half-â•‰life to enable shipping to sites lacking a cyclotron.
The Federal Drug Administration and European Medicines Agency have approved
florbetapir, flutemetamol, and florbetaben for the use of excluding a diagnosis of
AD. None of these agents are yet indicated for predicting the development of AD in
individuals with evidence of amyloid plaques or to monitor a patient’s response to
treatment. New ligands are currently being developed with the aim to increase bind-
ing specificity and detect amyloid deposition at very low levels (e.g., Forsberg et al.,
2013; Ito et al., 2014).
In the absence of a disease-â•‰modifying treatment for AD, the clinical utility of such
diagnostic imaging tools remains limited at present. However, as a research tool, the
ability to visualize the extent and pattern of amyloid burden in the living brains of
individuals opens previously inaccessible research pathways. Clinical trials of anti-â•‰
amyloid therapies are using amyloid PET agents to evaluate the ability of these drugs
to clear amyloid plaques, although to date there is little evidence that such clearance
improves the cognitive or functional symptoms of patients with AD (Mangialasche et
al., 2010; Salloway et al., 2014). Other clinical trials are using amyloid PET agents
to limit enrollment to those with evidence of elevated amyloid burden (Sperling et
al., 2014a). Such selective enrollment ensures that only individuals with the target
pathology are tested, reducing risk for those with little likelihood of being assisted by
the investigative treatment and increasing power by eliminating one source of uncon-
trolled heterogeneity in the population.
â•‡ 41
The initial and subsequent studies revealed the intriguing observation that approxi-
mately 30% of older individuals with no symptoms of AD nonetheless had substantial
amounts of amyloid plaques (e.g., Buckner et al., 2005; Mintun et al., 2006; Johnson
et al., 2007; Pike et al., 2007; Figure 2.1B). This observation did not come as a sur-
prise, as the literature has a long history of pathology-â•‰burdened cases of individuals
who die with no clear symptoms of dementia (Katzman et al., 1988; Morris et al.,
1996). Nonetheless, the ability to identify such individuals while they are living has
inspired research into what the presence of amyloid burden might mean for otherwise
asymptomatic individuals and has led to the development of new guidelines for a pre-
clinical stage of AD (Sperling et al., 2011).
These guidelines imply that the presence of amyloid plaques is sufficient to indi-
cate that an individual is on the pathway to AD or very high risk. This position is con-
cordant with observations in dominantly inherited AD, in which aberrant molecular
processes linked to genetic mutations of the genes for presenilin 1, presenilin 2, or
APP result in the early accumulation of amyloid plaques and to the development of
AD (Bateman et al., 2012). Sporadic AD may not be caused by the exact molecular
cascades as these genetic variants, but the similarity of the pathological and symptom-
atic expressions are difficult to dismiss. Individuals with relatively elevated amyloid
burden are estimated to have 2.6-â•‰fold higher risk for progression from mild cognitive
impairment to AD within two years than are individuals with relatively lower amyloid
burden (Jack et al., 2010).
In contrast, categorizing an individual as having preclinical AD on the basis of the
presence of amyloid burden alone has been criticized as placing too much emphasis
on a single molecular pathway for which insufficient evidence is currently available
(e.g., Fjell & Walhovd, 2012). The observation that some individuals pass away after
an extended life without any signs of dementia, yet possess the hallmark amyloid
plaques, could be taken to demonstrate that amyloid plaques are not sufficient to pro-
duce AD. Alternatively, some individuals may have protective factors, or the very for-
mation of plaques may be a mechanism to sequester toxic amyloid oligomers, thereby
delaying the impact of soluble amyloid on cognition (Shankar et al., 2008). Whether
all individuals with elevated amyloid plaques would eventually develop AD given a
long enough life span is ultimately unknowable. For present purposes, the important
issue is that amyloid represents a disease-â•‰related neuropathology that is associated
with age and may be present in a large proportion of the population under study by the
cognitive neuroscience of aging.
Associations of Amyloid with Other Brain Variables

The presence of amyloid burden in older adults without signs of clinical dementia
has been associated with a number of alterations in brain structure and function.
Associations between amyloid and glucose metabolism measured via FDG were
reviewed in a prior section. Because of its relevance to memory loss associated with
AD, the hippocampus and other structures of the medial temporal lobes (MTL) have
been special targets of interest when examining associations with elevated amyloid.
Because the hippocampal atrophy rate seen in AD is substantially larger than the rate
42
in cognitively healthy older adults (Barnes et al., 2009), it may be reasonable to assume
that the presence of amyloid in a subset of the cognitively healthy older adults may be
responsible for most or all of the association of age with hippocampal atrophy. In cog-
nitively healthy older adults, amyloid has been associated with smaller hippocampal
volumes (Apostolova et al., 2010; Mormino et al., 2009, 2014b; Oh et al., 2014) and
with longitudinal hippocampal atrophy estimated over follow-up periods between one
and four years (Knopman et al., 2013; Mattsson et al., 2014). Individuals with rela-
tively lower hippocampal volume have been estimated as possessing a 2.6-fold higher
risk for progression from mild cognitive impairment to AD within two years compared
to individuals with relatively higher volume, similar to that found for elevated amy-
loid burden (Jack et al., 2010). Specific hippocampal subfields may be preferentially
associated with the accumulation of amyloid (Hsu et al., 2015). However, the impact
of age on hippocampal atrophy appears to occur independently of amyloid (Nosheny
et al., 2015), and significant hippocampal atrophy is observed over one year even in
older individuals without evidence of the amyloid burden or cognitive impairment that
would suggest preclinical AD (Fjell et al., 2013; Insel et al., 2015). These latter results
suggest that presymptomatic AD does not fully explain hippocampal volume loss, but
rather that it is part of the normal aging process that can be augmented by a distinct
pathological cascade.
Similar results have been obtained in studies examining cortical thickness. AD
patients display reduced cortical thickness across many regions relative to cognitively
healthy older adults (Dickerson et al., 2009), and within cognitively healthy older
adults, amyloid is associated with widespread reduced cortical thickness (Becker et
al., 2011; Dickerson et al., 2012; Doré et al., 2013; Villeneuve et al., 2014). However,
the regional patterns are not identical between preclinical individuals and symptom-
atic patients (Whitwell et al., 2013), some studies have failed to find an association
between cortical thickness and amyloid (Wirth et al., 2013), and at least one study
has found increases in thickness associated with amyloid alone but decreased thick-
ness when amyloid is accompanied by tau (Fortea et al., 2014). These results suggest
that there is a somewhat tenuous relationship between amyloid and cortical thickness
among cognitively healthy individuals, which may be indicative of additional fac-
tors that impact the association of amyloid and cortical thickness. Potential factors
could include the co-occurrence with amyloid of the APOE4 allele or of elevated
neurofibrillary tau tangles. Decreases in cortical thickness are also observed in older
individuals without evidence of preclinical AD (Fjell et al., 2013), and the associa-
tion with amyloid in such individuals varies across the cortex and does not appear to
target MTL regions (Fjell et al., 2014). As with the results on hippocampal volume,
the overall pattern from the literature suggests cortical thinning occurs during the
course of aging, but may become more severe in the presence of amyloid burden and
the progression of AD. Thickness in MTL regions may be particularly vulnerable to
AD pathology, especially once symptoms appear at the stage of mild cognitive impair-
ment (Fjell et al., 2014).
In addition to the structural changes observed, aberrant functional activity in the
MTL has been associated with elevated amyloid. In older individuals with elevated
amyloid burden, MTL activity tends to be reduced during associative memory tasks
(Vannini et al., 2012, 2013; Huijbers et al., 2014). Memory- associated cortical
43
regions, many overlapping with what has come to be known as the default network,
also exhibit amyloid-related alterations (Buckner et al., 2005; Sperling et al., 2009,
2010; Mormino et al., 2012), although the direction of the results is not always consis-
tent (Kennedy et al., 2012; Adamczuk et al., 2016, Elman et al., 2014a).
Because task-based activation may vary widely depending upon the task admin-
istered during scanning or the analysis methods used, measures of network func-
tion based on intrinsic properties of network connections have gained prominence.
Elevated amyloid burden has been associated with disruptions of such functional con-
nectivity, particularly of connections between the MTL and cortical regions within the
default network (Hedden et al., 2009; Sheline et al., 2010; Mormino et al., 2011; Jack
et al., 2013; Huijbers et al., 2014). These associations in cognitively healthy individu-
als are mirrored by alterations that occur in individuals carrying dominantly inherited
mutations prior to the onset of symptoms (Chhatwal et al., 2013). Although most
attention has been focused on within-network connectivity, emerging evidence sug-
gests that alterations in between-network connections are also associated with amy-
loid burden (Elman et al., 2014b), as are disruptions of a region’s participation across
multiple networks (Buckner et al., 2009; Drzezga et al., 2011). Although the net-
work connectivity dynamics are highly simplified relative to humans, work in trans-
genic rodent models has found that expression of amyloid plaques is associated with
reduced regional connectivity (Bero et al., 2012), or that impairments in functional
connectivity precede development of plaques (Grandjean et al., 2014). However, it
is important to note that an association between functional connectivity and elevated
amyloid is not always observed (Adriaanse et al., 2014), and there is some evidence
that connectivity may be increased rather than decreased in individuals with elevated
amyloid (Lim et al., 2014; Mormino et al., 2011). These inconsistencies and the small
effect sizes observed in many studies indicate that we still lack a full understanding
of the relationship between amyloid burden and functional connectivity and how they
might inform the separation of clinically normal aging from AD. Using advanced
imaging techniques that enable simultaneous acquisition of multiple slices for faster
sampling, dynamic measures of non-stationarity in brain states may further inform
our understanding of the relationship between amyloid and functional connectivity
(Jones et al., 2012). There may also be interactions between the presence of amyloid
and neurodegeneration that impact functional connectivity (Jack et al., 2013).
Although there is evidence of increased white matter abnormalities in AD patients
(Brickman et al., 2009; Brickman, 2013; Brun & Englund, 1986; Provenzano et al.,
2013) and an association between vascular amyloid deposition and WMH (Gurol et
al., 2013), studies in cognitively healthy older adults indicate that amyloid and white
matter degeneration are independent brain biomarkers. Multiple reports have failed
to find an association between amyloid burden and white matter hyperintensities
(Rutten-Jacobs et al., 2011; Hedden et al., 2012, 2016; Lo et al., 2012; Marchant et
al., 2012) or diffusion-weighted imaging of white matter integrity (Kantarci et al.,
2014) in cognitively healthy older adults. Although one study found an association
of amyloid with impairment in axial diffusivity in some white-matter structures, no
associations were observed in other diffusion measures (Molinuevo et al., 2014). In
contrast, another study found increased integrity on measures of fractional anisot-
ropy and mean diffusivity in multiple structures associated with amyloid burden
44
(Racine et al., 2014). Additional studies have found amyloid burden and white
matter abnormalities to be independently associated with cognitive function in
control or patient groups (Hedden et al., 2012; Barnes et al., 2013; Haight et al.,
2013; Guzman et al., 2013), which is again suggestive of separate pathological
pathways for amyloid and white-matter abnormalities. Evidence from transgenic
rodent models is mixed with regard to an association between amyloid and white-
matter-integrity (Kastyak-â•‰Ibrahim et al., 2013; Qin et al., 2013; Sun et al., 2014;
Grandjean et al., 2014), and it is difficult to lean heavily on rodent models given
they replicate features of pathology but do not recapitulate the full physiology of
the human conditions.
To our estimation, the weight of the current evidence favors amyloid burden and
white-matter abnormalities evolving along separate pathological trajectories that may
nonetheless co-â•‰occur in some individuals and combine to impair cognition. However,
because individuals with both amyloid and white-matter abnormalities may have more
prominent cognitive symptoms and would therefore be less likely to be included in a
sample of cognitively healthy individuals, researchers must be aware of the potential
for sampling bias. Such a bias would account for finding no association in the pre-
clinical stage of AD with an association emerging once symptom onset is observed.
A focus in future research on longitudinal associations between amyloid and white-
matter abnormalities will help answer important questions as to whether these two
neuropathologies have a causal relationship.
Longitudinal studies that examines change in amyloid with change in other neu-
roimaging markers will become an important frontier for future research. Because of
the nonlinear function that characterizes amyloid accumulation (with acceleration of
accumulation early on, eventually reaching a plateau), there will be a dynamic window
in which such associations may be observed, with reduced associations once individuals
reach the accumulation plateau. Cross-â•‰sectional studies commonly categorize individu-
als into those with and without evidence of elevated amyloid burden because of the non-â•‰
normal distribution of this variable. These methods may also require a reexamination
when longitudinal studies require subtle indicators of amyloid accumulation over time.
Amyloid–â•‰Cognition Relationships
Using amyloid imaging PET techniques, individuals who possess a marker of AD
pathology can be identified and examined specifically or culled from studies of cogni-
tive aging. Such individuals have previously been included in samples used to study
aging as a developmental process, representing approximately 30% of the cognitively
healthy individuals in the population. Markers of amyloid burden therefore provide an
opportunity to examine hypotheses with more precision than was previously possible.
Additionally, hypotheses involving differential impacts of AD-â•‰related pathology and
age-â•‰related developmental processes can be examined.
For example, one question that can be tested is whether asymptomatic older indi-
viduals with evidence of amyloid plaques exhibit greater decrements in cognition than
do those without evidence of amyloid plaques. This has been a question of intense
interest, with many studies examining multiple cognitive domains (e.g., Hedden et al.,
45
2012; Lim et al., 2012; Rodrigue et al., 2012). Although both positive and negative
results have been published, the general consensus from studies with large sample
sizes (i.e., those with N>100) is that amyloid burden is associated with subtle cogni-
tive deficits and that certain domains are more strongly impacted by amyloid than
other domains.
In a meta-analysis that included 24 independent cohorts (up to 3,495 subjects)
where amyloid burden was assessed using a variety of methods (including PET imag-
ing, cerebrospinal fluid assays, and pathological examination), episodic memory was
found to be most impacted by the presence of amyloid in asymptomatic individuals
(Hedden et al., 2013). Global cognitive function (which includes episodic memory)
and executive function were also significantly impacted by the presence of amyloid,
while processing speed, working memory, visuospatial cognition, and semantic mem-
ory were not significantly impacted. When examining only studies employing the
PET agent PIB, only episodic memory was significantly impacted by the presence of
amyloid. It is important to keep in mind that, although significant, the effect size of
amyloid on episodic memory observed in the meta-analysis was approximately r = 0.1
(equivalent to Cohen’s d = 0.2, Cohen 1988). This small effect size implies an asso-
ciation that explains approximately 1% of the variation in episodic memory. While
the associations are expected to be relatively subtle, as by definition asymptomatic
individuals do not exhibit grossly abnormal cognitive performance, weak associations
between markers of hallmark AD pathology and cognition pose problems for a clear
separation of clinically normal aging and AD.
Additional studies examining associations of amyloid with longitudinal cogni-
tive change have been published since 2013 (Mormino et al., 2014; Ossenkoppele et
al., 2014; Dowling et al., 2015). Based on these collective results, it appears likely
that the presence of amyloid in asymptomatic older individuals has at least some
specificity for subtle memory deficits, and that assessment of longitudinal change
in cognition may be somewhat more sensitive to the distinction of normal aging and
AD-related processes than cross-sectional assessment. This trend may again reflect
biased sampling in cross-sectional studies because individuals with amyloid and
cognitive symptoms would be less likely to enter a sample of cognitively healthy
individuals.
Markers of amyloid burden also allow cognitive aging to be explored in the absence
of evidence of comorbid neurodegenerative disease. A natural question is whether
the subtle deficits in memory and other cognitive domains associated with the pres-
ence of amyloid burden account for most or all of the cognitive alterations typically
ascribed to the aging process. That is, has the cognitive aging literature mischaracter-
ized normal age-related processes by including individuals with preclinical amyloid
pathology? There are two primary ways to examine this question. One is to remove
all cognitively normal individuals with evidence of elevated amyloid burden from
the sample and examine whether age effects continue to be observed as significant.
Studies taking this approach have observed significantly lower cognitive functioning
with increasing age even in those individuals without evidence of preclinical AD (e.g.,
Nebes et al., 2013). Again, this finding complicates the notion that cognitive decline in
aging is primarily driven by pre-symptomatic disease, at least insofar as the available
markers adequately identify early stages of disease.
46
A different approach is to estimate the impact of both age and amyloid on cogni-
tion in a combined model and examine the effect of age while controlling for amy-
loid. If age continues to have a significant effect, it supports a model in which age
has a significant independent association with cognition irrespective of amyloid bur-
den. The same logic has been applied to the investigation of age-related decreases
in neurotransmitter functions and their impact on cognition, as reviewed later in this
chapter. This approach flexibly accommodates the inclusion of multiple markers of
brain alterations that may be associated with cognition so that the model can be used
to understand whether age continues to account for significant variation in cognition
after controlling for many such markers. This approach also maximizes power by
using the full dataset of aging individuals without making any assumptions about the
future progression of an individual along a particular pathogenic pathway.
If we take this second perspective, the meta-analytic data on the association
between amyloid and cognition can be combined with that from meta-analytic data
on the association of age and cognition to estimate the likely impact of the inclusion
of individuals with amyloid on estimates of the age effects on cognition. Taking
data from a large meta-analysis that looked across multiple domains (Verhaghen and
Salthouse, 1997), we find that the cross-sectional association of aging with cogni-
tion is 2–6 times as large as that of amyloid with cognition (Figure 2.2). Even if
measured amyloid burden shared all of its cognition-related variance with that of the
age-related variance in cognition, amyloid would only account for less than 50% of
the age-related variance in cognition. Because the correlation between amyloid and
age is less than r = 0.4 in most samples of cognitively healthy older adults (e.g., Pike
et al. 2011; Kantarci et al. 2012; Hedden et al. 2016), it is more likely that amyloid
Episodic memory 17%
Executive function 11%
Working memory 12%
Processing speed 6%
Visuospatial function 6%
0 0.1 0.2 0.3 0.4

Effect size
Amyloid Age
Figure 2.2 Relative estimated effect sizes of amyloid burden on cognition and age on cogni-
tion across several cognitive domains. Assuming a correlation of r = 0.4 between amyloid and
age, the percentage (numbers at right) of age-related variance in each cognitive domain that is
shared with amyloid can be computed. Meta-analytic estimates for amyloid are based on Hedden
et al., 2013; meta-analytic estimates for age are based on Verhaeghen & Salthouse, 1997.
47
shares less than 20% of the age-related variance in cognition. Although this is not
insubstantial, it indicates that there is much more to be learned in the cognitive neu-
roscience of aging than discriminating between older individuals with and without
preclinical AD, with the caveat that measures of amyloid burden are imperfect esti-
mates of preclinical AD.
In a recent cross-sectional study that combined multiple brain markers, including
FDG and amyloid PET, to examine age-related differences in cognition, amyloid
was found to share 16% of the age-related variance in memory (Hedden et al., 2016;
Figure 2.3). This accords well with the above meta-analytic estimate. In the same
study, FDG shared 27% of the age-related variance in memory (Figure 2.3), which
could be taken to support a model in which FDG has a more proximal relationship to
memory function than does amyloid (Jack et al., 2013). However, the relationship of
amyloid or FDG to memory in cognitively healthy older adults should be placed in
context with their relationship to each other and to other markers of brain health. For
instance, of the 16% of age-related variance in memory shared with amyloid, 6% was
Amyloid
16%
Cortical
thickness l
na
FDG 25% r hi
to %
27% En 15 pu
s
c am 4%
po 1
hip
Para
WMH
38%
Hippocampus
52%
DTI
41%
Striatum
20%
FPCN DN
6% 7%
Figure 2.3 Percentage of age-related variance in episodic memory performance that is shared
with different brain markers. Highlighted sections indicate molecular markers of glucose
metabolism (FDG) and amyloid (PIB). Entorhinal and parahippocampus indicate thickness
measures. Hippocampus and striatum indicate volume measures. Functional connectivity mea-
sures are estimated from the default network (DN) and frontoparietal control network (FPCN).
White matter integrity measures were estimated using fractional anisotropy from diffusion
tensor imaging (DTI) and white matter hyperintensity (WMH) volume. Pie sections indicate
proportional share of age-related variance in episodic memory estimated for each brain marker
from univariate analyses. Adapted from Hedden et al., 2016.
48
also shared with FDG and 11% was shared with hippocampus volume; only 3% was
found to be uniquely attributable to amyloid (Hedden et al., 2016). Amyloid markers
may indicate likelihood of early progression on the AD spectrum, but because they
are associated with age, they may also have interactive relationships with other brain
measures that alter their relationship to cognition across age.
While cross-â•‰sectional analyses of this type are common and constitute a large por-
tion of the cognitive aging and neuroscience literature, it is important to acknowledge
that they have also been heavily criticized. Using data simulations, Lindenberger et
al. (2011) suggested that a statistically significant cross-â•‰sectional mediation analysis
has little bearing on change in the variables longitudinally. Pertaining to this point,
several studies have shown that cross-â•‰sectional and longitudinal estimates of age-â•‰
related changes in cognitive functions (Rönnlund et al., 2005), brain volume (Raz et
al., 2005), and fMRI activation (Nyberg et al., 2010) diverge. Moreover, associations
between a brain biomarker and cognitive functions in aging may differ depending on
whether the associations are observed cross-â•‰sectionally or longitudinally (Lövdén et
al., 2014; Landau et al., 2012; Mormino et al., 2014b) highlighting again that cross-â•‰
sectional associations and mediation analyses are not able to identify existing causal
links. Rather, such cross-â•‰sectional associations are limited to inferences about the
between-â•‰person variation in relationships among brain biomarkers and cognition in
the context of aging, which may prove useful when trying to classify those individuals
at higher risk for preclinical AD and other age-â•‰related neurodegenerative cascades.
Tau Imaging
Amyloid is only one of the two hallmark pathologies associated with AD. HyperÂ�
phosphorylation and misfolding of the tau protein lead to aggregation visualized as
tangles inside of neurons. This aggregation may be more closely linked to neurode-
generation than are amyloid plaques (Ossenkoppele et al., 2015; Spires-â•‰Jones and
Hyman, 2014) and may occur through a distinct pathway than does amyloid accumu-
lation (Small and Duff, 2008). Relating back to the discussion in a previous section
as to whether amyloid plaques are sufficient to indicate whether an individual is on
the path to AD, recent studies have turned to tau tangles as an equally, or possibly
greater, predictor for the transition to AD. A recent development is the introduction of
novel tau imaging agents (e.g., Shao et al., 2012; Chien et al., 2013; Okamura et al.,
2013; Chien et al., 2014, Fawaz et al., 2014; Figure 2.1C). Early observations sug-
gest that measurements using these agents recapitulate the spread of tangles from the
entorhinal cortex to nearby limbic regions and finally to neocortical regions in a fash-
ion consistent with Braak staging (Braak and Braak, 1997; Villemagne et al., 2015).
However, several of these agents display binding in unexpected locations, and it is
as yet unclear how specific each agent is to various subspecies of tau. Nonetheless,
because the localized expression of neurofibrillary tangles occurs throughout life
from a relatively early age (Braak and Braak, 1997; Nelson et al., 2012), tau may
be a particularly important molecular target for understanding age-â•‰related altera-
tions in cognition (Delacourte et al., 2002). The rapid adoption and integration of tau
imaging into large-â•‰scale studies and clinical trials indicates the level of excitement
â•‡ 49
generated in the field by the development of these novel agents for an important
target (Sperling et al., 2014b; Villegmagne et al., 2015).
Imaging Dopamine
Many PET agents target neurotransmitter systems, providing a window into

another class of synaptic changes in aging. Age-â•‰comparative studies have identi-
fied reduced densities of serotonin receptors (e.g., Wong et al., 1984, Meltzer et al.,
1998, Yamamoto et al., 2002) and acetylcholine receptors (e.g., Dewey et al., 1990).
Cholinergic cell death has been implicated in the cognitive deficits associated with
AD (e.g., Terry and Buccafusco, 2003, for review). More than any other neurotrans-
mitter system, however, the dopamine system has generated substantial interest
because of its relevance to human cognitive functions, psychiatric disease, aging, and
age-â•‰related movement disorders. In 2000, Arvid Carlsson, Paul Greengard and Eric
Kandel shared the Nobel prize for the discovery of dopamine as a neurotransmitter
and the mechanisms by which neurotransmitters like dopamine contribute to syn-
aptic plasticity, the neurochemical basis of learning and memory. Dopamine is syn-
thesized in the midbrain and heavily innervates the striatum and, to a lesser degree,
cortical areas. Striatal cells receive input from widespread areas across the cortex
and project back to cortical areas via the pallidum and thalamus forming “loops” that
broadly differentiate into motor, cognitive, and affective systems. Nigral dopamine
neurons modulate neural excitability in cortico-â•‰striatal circuits by promoting behav-
iors for maximizing reward and minimizing punishment (Schultz et al., 1997; Frank
and O’Reilly, 2006).
In cortical regions, dopamine neurons form structural complexes with other neu-
rons similar to those described for striatum. However, the functional implications are
less understood than in striatum, and it remains to be explored exactly how cortical
and striatal dopamine signals jointly or independently regulate behavior. One hypoth-
esis regarding prefrontal dopamine functions in PFC is that they stabilize neural
representations in working memory and render them robust against interfering dis-
tractors (Servan-â•‰Schreiber et al., 1990; Durstewitz et al., 2000). This hypothesis was
suggested by early primate studies (Brozoski et al., 1979; Sawaguchi and Goldman-â•‰
Rakic, 1991) that demonstrated that dopamine depletion in prefrontal cortex (PFC)
selectively impaired working memory in monkeys.
Dopamine Functions Decrease Across the Adult Lifespan

Post mortem examinations of human brain tissue suggest a linear reduction of around
5% per decade in dopamine concentration and cell density for cognitively healthy
subjects between ages 50 and 90 (Carlsson and Winblad, 1976; Riederer and St
Wuketisch, 1976; Fearnley and Lees, 1991). Linear age-â•‰related reductions in dopa-
mine cell counts are distinct from the pathophysiology in Parkinson’s disease, which
is characterized by exponential and regionally specific neuronal death (Fearnley and
Lees, 1991).
50
The introduction of PET imaging approaches to infer dopamine signaling in the

mid 1980s catalyzed research (Farde et al., 1986). With respect to aging, PET studies
estimating age-related changes across the lifespan align well with postmortem cell
counts, and these data have been extensively reviewed (Reeves et al., 2002; Bäckman
et al., 2006). Striatal markers of the main postsynaptic receptor types D1 and D2,
as well as the presynaptic dopamine transporter (DAT), show reliable differences
between young adults in their 20s and older adults over the age of 65 (Figure 2.4).
The few studies that have assessed binding of more than one radioligand in the
same set of subjects show that presynaptic DAT and postsynaptic D2 receptor binding
are correlated at around 0.6–0.7 (Volkow et al., 1998a; Ishibashi et al., 2009). Further,
lower density of markers of the dopamine system correlates with lower density of
other neurotransmitters (Wang et al., 1995), synaptic function measured by FDG PET
(Volkow et al., 2000) and even white matter integrity (Rieckmann et al., 2016). Lower
neurotransmitter densities in the dopamine system are also paralleled by age-related
reductions of muscarinic receptors (e.g., Dewey et al., 1990) and glutamatergic recep-
tors (Segovia et al., 2001). The available literature therefore suggests evidence for a
global breakdown of synaptic efficiency across the adult lifespan that may be reflected
in decreased terminal density and receptor availability, with the dopamine system
being just one facet reflective of this global cascade.
The broad reductions in markers of the dopamine system in aging are unlike the
patterns seen in disease. For example, in Parkinson’s disease, presynaptic DAT mark-
ers show pronounced decreases compared to age-matched controls, whereas the
postsynaptic D2 receptor functions remain relatively intact or show a subtle increase
(Seeman and Niznik, 1990; Kim et al., 2002). Intact or up-regulated postsynaptic
receptor functions in response to DAT depletion have also been demonstrated for DAT
knock-out mice (Gainetdinov et al., 1999) and explain why dopamine boosters like
L-dopa are effective treatments in Parkinson’s disease. Conversely, in schizophrenia,
presynaptic DAT density is unaltered whereas the postsynaptic D2 receptor response
to dopamine release is increased compared to controls (Seeman and Niznik, 1990;
Laruelle 2000).
There are, however, several observations that complicate the idea of a global neu-
rodegeneration in aging with respect to the dopamine system. First, PET studies of
ligand binding to dopamine-synthesizing enzymes suggest that dopamine synthesis
capacity is unaltered or even increased in clinically healthy aging (e.g., Braskie et
al., 2008). Higher dopamine synthesis in aging may reflect compensatory responses
to the deficits in other parts of the dopamine system, e.g., the pre-and postsynaptic
transporter and receptor functions. Second, in an age-comparative study that utilized
a D1 receptor tracer with affinity for both striatal and cortical D1 receptors, individual
differences in striatal D1 receptors did not correlate with individual differences in
cortical D1 receptors in older adults (Rieckmann et al., 2011a). These findings suggest
that while striatal and cortical dopamine receptor functions show similar trajectories
with chronological age, they appear to be regulated independently. Third, there is
mixed evidence for a progressive linear loss of dopamine markers in old age. While
there are reliable differences in pre-and postsynaptic dopamine markers between
younger and older adults, in some studies focused on older adults only (Reeves et al.,
2005; van Dyck et al., 2008) or using samples with very tight age ranges (Nevalainen
51
D1 receptor (11C-SCH22390) D2 receptor (11C-Raclopride) Dopamine transporter (11C-Altropane)
Young Old Young Old Young Old

DVR DVR DVR
0.5 2.5 0.5 5.0 0.5 5.0
Figure 2.4 Example PET images for three common PET ligands of the dopamine system. An example image for a young person (20–30 years)
and a clinically normal older adult (> 65 years) is shown for each ligand. Loss of striatal signal for the old person can be seen for all ligands.
Images are voxelwise DVR images with reference region cerebellum. (See color plate also)
52
et al., 2015) no association between dopamine markers and age was observed. These
findings may suggest that, across the lifespan, loss of dopamine functions may be
pronounced in middle adulthood and then level off in old adulthood. This pattern
contrasts with that observed for white-matter integrity, hippocampal atrophy (e.g.,
Walhovd et al., 2005), and cognitive functions (Schaie, 1996; Salthouse, 2014), where
negative age associations accelerate in old age. Identifying the true rate of decline of
dopamine functions in aging, as well as comparisons between within-person trajec-
tories for dopamine markers of striatum, cortex, midbrain dopamine synthesis and
atrophy can ultimately only be addressed with longitudinal data. At least two large-
scale longitudinal dopamine PET studies are currently underway, which will likely
yield new insights into age-related dopamine losses over the next decade (Parkinson
Progression Marker Initiative, 2011; Nevalainen et al., 2015).
PET has also taken on an important role for translation from recordings in animal
systems to understanding complex human behavior. PET-based measures of striatal
dopamine functions are associated with learning from feedback, adapting behavior in
response to rewards (Schott et al., 2008; Cools et al., 2009; Jonasson et al., 2014), and
flexible updating of memory representations in prefrontal cortex (inferred by fMRI,
Nyberg et al., 2009). Prefrontal dopamine D1 receptor densities measured with PET
have been linked to working memory task performance in schizophrenic patients
(Abi-Dargham et al., 2002) and in controls (Takahashi et al., 2008). PET studies of
the human dopamine system have shown associations of markers of the dopamine
system with complex cognitive operations such as playing a video game (Koepp et al.,
1998), general knowledge (Karlsson et al., 2011), and human personality traits like
impulsivity (Buckholtz et al., 2010).
Markers of the Dopamine System are Associated with Cognitive

Functions in Aging
The first empirical demonstration that reduced cognitive functions in older adults are
in part explained by individual differences in markers of striatal dopamine functions
was reported in Volkow et al. (1998b). In this study, 30 cognitively healthy volunteers
between the ages of 24 and 86 participated in one PET scan assessing striatal D2 recep-
tor densities and a neuropsychological exam including tests of executive functions,
perceptual speed, and finger tapping, which all showed reliable performance decreases
with increasing age. Partial correlation analyses showed that even after accounting for
the effects of chronological age, dopamine D2 receptor availability and cognitive per-
formance on tests of executive and motor function (see also Wang et al., 1998) were
significantly correlated. Bäckman et al. (2000) extended these initial findings to show
that episodic memory is also associated with lower striatal D2 availability. Moreover,
the amount of variance in tests of perceptual speed and episodic memory related to the
effects of chronological age (between 13% and 52%) was almost fully explained by
the dopamine PET measure. These results are interesting in light of recent multimodal
imaging explorations that link multiple MRI and PET-based measures of brain structure
and function to age-related variance in cognitive functions, explaining around 75% of
the age-related variance (Hedden et al., 2016, discussed in the section on amyloid).
53
Dopamine PET markers were not available in that analysis, and it remains to be explored
in future studies how dopamine PET markers interrelate with other markers of a cog-
nitively healthy aging brain, or how much variance in cognitive performance can be
attributed selectively to individual differences in dopamine functions.
The initial findings reported in Volkow et al. and Bäckman et al. of an interrelation
between age, dopamine, and cognition have been extended to different cognitive tasks
(Reeves et al., 2005) and ligands for DAT (Mozley et al., 2001; Erixon-Lindroth et al.,
2005; van Dyck et al., 2008) and dopamine synthesis capacity (Braskie et al., 2008;
Landau et al., 2009).
Dopamine PET can also be used to assess the change in ligand binding displace-
ment following acute release of endogenous dopamine into the synapic cleft (Laruelle,
2000; Monchi et al., 2006). Few studies have used radioligand displacement to inves-
tigate dopamine release in the context of aging. An exception is a study by Karlsson
et al. (2009), who showed that while an executive task successfully decreased radio-
ligand binding to striatal dopamine receptors in younger adults, no significant change
could be observed in older adults. Although it is unclear whether decreases in D1
binding are indicative of dopamine release or receptor internalization, this study sug-
gests that older adults not only have decreasing numbers of dopamine transporters
and receptors but also an altered response of the dopamine system to a cognitive chal-
lenge. Consistent with an interpretation of reduced dopamine release in older adults,
Floel et al. (2008) demonstrated a change in ligand binding during task performance
in older adults only when dopamine synthesis was boosted by a dose of levodopa prior
to the scan. However, it remains to be seen how these results can be reconciled with
the work by Braskie et al. (2008) suggesting a compensatory increase in dopamine
synthesis capacity in cognitively healthy older adults. Future studies may provide an
answer by focusing on individual differences that link dopamine synthesis capacity,
cognitive performance, and response to pharmacological challenge in older adults.
Multimodal imaging. Longitudinal PET studies of the dopamine system in aging
are not yet available, and it is important to keep in mind that the hypothesis that
dopamine loss and declining cognitive functions are linked in aging has thus far
exclusively been addressed in cross-sectional comparisons. It is not obligated that
the relations among individual differences in change of neurobiological and cognitive
measures follow the same patterns suggested by between-person cross-sectional dif-
ferences (e.g., Nyberg et al., 2010; Lindenberger et al., 2011). Until large longitudinal
data collections are available, cross-sectional multimodal imaging studies combing
PET, fMRI and, where possible, pharmacological interventions can provide potential
mechanisms by which dopamine losses in aging may affect cognitive performance.
In one study in cognitively healthy older adults, Landau et al. (2009) showed that
striatal dopamine synthesis capacity was related to working memory task accuracy
as well as the strength of prefrontal cortex activation during the task as measured by
fMRI. Age-related changes in prefrontal activation during working memory are well
established (e.g., Rajah and D’Esposito, 2005) and the results of Landau et al. provide
initial evidence for an interrelation of prefrontal activation and cognitive performance
with the integrity of the dopamine system. These findings were later extended to dem-
onstrate an association between dopamine synthesis capacity and prefrontal coupling
to striatum during working memory (Klostermann et al., 2012). Using a D1 receptor
54
ligand, Bäckman et al. (2011) and Rieckmann et al. (2011b) showed that the rela-
tion of dopamine functions with fMRI task activation and functional connectivity,
respectively, are found in widespread association networks. These data could suggest
that striatal dopamine functions are a critical modulator of efficient frontal-parietal
recruitment during cognitive task performance. The causal nature of this associa-
tion was assessed by administering a D1 receptor antagonist to the sample of young
healthy participants to examine whether the patterns of activation and connectivity
typical for older adults could be elicited in younger adults by blocking postsynaptic
dopamine receptor functions. The older adults’ patterns of activation (Fischer et al.,
2010) and functional connectivity (Rieckmann et al., 2012) could, at least in part, be
simulated by drug administration.
The combination of multimodal imaging and pharmacological manipulation is an
underutilized experimental design that is of great advantage beyond its usefulness to
provide support for causal relations. Even after accounting for macrostructural partial
volume effects (Meltzer et al., 1990), it is a legitimate concern that age-related dif-
ferences in brain morphology, regional signal-to-noise, or biases in co-registration of
imaging data to group templates introduces factors that may induce a spurious asso-
ciation between PET and MRI signal. Pharmacological challenges in within-person
comparisons of younger adults do not suffer from this problem and can therefore
provide an important complementary examination of neurotransmitter imaging results
in age-comparative studies.
Prefrontal Dopamine Functions. The striatum receives the densest dopamine
innervations from the midbrain, and molecular PET binding targets like DAT and the
D2 receptor are preferentially available in striatum. D1 receptors are also densely con-
centrated in striatum but are the dominant receptor type in cortex, which can be seen
in Figure 2.4. Low availability of binding targets in cortex is not a primary concern
for molecular studies of prefrontal dopamine functions. Rather, almost all dopamine
receptor ligands are non-selective to a certain degree and also bind to serotonin recep-
tors. This is of little effect in the striatum, where there are few serotonin binding sites,
but likely affects the cortical signal of dopamine receptor ligands to some extent and
is an important caveat to consider.
Age differences in dopamine receptor densities are comparable for striatum and
cortical areas (Suhara et al., 1991; Rieckmann et al., 2011a) but may have distinct
implications for cognitive decline. In line with animal research that suggests a
role for prefrontal dopamine functions in maintenance of stable representations, it
has been proposed that prefrontal dopamine functions become particularly appar-
ent in task situations that assess consistency of responses from one trial to the next
(MacDonald et al., 2009, 2012). Inter-individual response variability across many
trials increases with advancing age, and MacDonald et al. showed that extrastria-
tal D1 and D2 dopamine receptor densities mediate this effect. Critically, in these
studies striatal dopamine receptor densities were unrelated to inter-individual
response variability, which suggests that striatal and extrastriatal dopamine recep-
tors are, in part, independent measures in aging. Indeed, Rieckmann et al. (2011),
showed that while striatal and extrastriatal D1 receptor densities are correlated in
younger adults, this is not the case for older adults. This effect was not apparent
in older adults who were comparable to younger adults in terms of performance
â•‡ 55
on an executive functions task. In other words, receptor losses in cortex and stria-
tum do not always go hand in hand, and the level to which striatal and prefron-
tal dopamine functions are “in balance” may also be an important component in
understanding the role of dopamine functions in complex cognitive tasks. Further
exploration of this hypothesis opens up a new avenue in the context of aging
research but is at the core of the dopamine hypothesis for schizophrenia, which
postulates a hyperactivity of striatal dopamine functions, and a hypoactivity of
prefrontal dopamine functions and of understanding the symptoms of Parkinson’s
disease (pronounced loss of striatal dopamine neurons relative to prefrontal dopa-
mine neurons).
Conclusion
This selective overview highlights the importance of quantifying the spatial distribu-
tion and concentration of molecular targets and their relation to other brain markers
for unraveling the association between aging and cognition. Much of aging research
to date is based on correlational analyses with a focus on associations, rather than on
causative mechanisms. To the extent that neurobiological aging can be characterized
by gradual progression or development of cellular and molecular processes, molecular
imaging techniques will provide the most proximal tools to these processes. Their
potentially causative roles must still be verified with longitudinal studies paired with
statistical methods allowing causal inference or, ideally, direct pharmacological inter-
vention studies that target selective molecular pathways.
Molecular targets of pathological proteins, such as amyloid and tau, may help to
reveal how aging processes and preclinical AD pathology have overlapping, interac-
tive, or distinct impacts on cognition. By relating synaptic or neurotransmitter func-
tion as measured with FDG, dopaminergic targets, or other neurotransmitter systems
to macroscopic or functional changes using multimodality imaging, we may be better
positioned to inform how large-â•‰scale brain networks are differentially impacted by
molecular systems during aging, and how these neural alterations underlie cognitive
outcomes. Recent advances that allow simultaneous PET-â•‰MRI acquisition, more sen-
sitive PET cameras, and multiple tracer techniques may enable new approaches to
technical challenges. The field should continue to design a future in which we apply
these developing tools and our emerging understanding to predict those individuals
most likely to be affected by a particular type of pathology or disruption of a specific
molecular pathway. Such studies may provide a foundation for the best application of
interventions as they emerge.
References
Abi-â•‰Dargham A, Mawlawi O, Lombardo I, Gil R, Martinez D, Huang Y, Hwang D-â•‰R, Keilp

J, Kochan L, Van Heertum R, Gorman JM, Laruelle M (2002) Prefrontal dopamine D1
receptors and working memory in schizophrenia. J Neurosci 22:3708–â•‰3719.
Adamczuk K, De Weer AS, Nelissen N, Dupont P, Sunaert S, Bettens K, Sleegers K, Van
Broeckhoven C, Van Laere K, Vandenberghe R (2016) Functional changes in the language
56
network in response to increased amyloid-β deposition in cognitively intact older adults.

Cereb Cortex 26:358–373.
Adriaanse SM, Sanz-Arigita EJ, Binnewijzend MAA, Ossenkoppele R, Tolboom N, van Assema
DME, Wink AM, Boellaard R, Yaqub M, Windhorst AD, van der Flier WM, Scheltens P,
Lammertsma AA, Rombouts SARB, Barkhof F, van Berckel BNM (2014) Amyloid and
its association with default network integrity in Alzheimer’s disease. Hum Brain Mapp
35:779–791.
Alzheimer A (1906) Über einen eigenartigen schweren Erkrankungsprozeß der Hirnrinde.
Neurol Centralbl 23:1129–1136.
Alzheimer A (1907) Über eine eigenartige Erkrankung der Hirnrinde. Allgemeine Z Psychiatrie
Psychisch-Gerichtliche Med 64:146–148.
Andreasen N, Hesse C, Davidsson P, Minthon L, Wallin A, Winblad B, Vanderstichele H,
Vanmechelen E, Blennow K (1999) Cerebrospinal fluid β-Amyloid(1–42) in Alzheimer
disease: differences between early-and late-onset Alzheimer disease and stability during
the course of disease. Arch Neurol 56:673–680.
Apostolova LG, Hwang KS, Andrawis JP, Green AE, Babakchanian S, Morra JH, Cummings
JL, Toga AW, Trojanowski JQ, Shaw LM, Jack CR, Petersen RC, Aisen PS, Jagust WJ,
Koeppe RA, Mathis CA, Weiner MW, Thompson PM, Thompson PM (2010) 3D PIB
and CSF biomarker associations with hippocampal atrophy in ADNI subjects. Neurobiol
Aging 31:1284–1303.
Barnes J, Bartlett JW, van de Pol LA, Loy CT, Scahill RI, Frost C, Thompson P, Fox NC (2009)
A meta-analysis of hippocampal atrophy rates in Alzheimer’s disease. Neurobiol Aging
30:1711–1723.
Barnes J, Carmichael OT, Leung KK, Schwarz C, Ridgway GR, Bartlett JW, Malone IB,
Schott JM, Rossor MN, Biessels GJ, DeCarli C, Fox NC, Fox NC (2013) Vascular and
Alzheimer’s disease markers independently predict brain atrophy rate in Alzheimer’s
Disease Neuroimaging Initiative controls. Neurobiol Aging 34:1996–2002.
Bäckman L, Ginovart N, Dixon RA, Wahlin TB, Wahlin A, Halldin C, Farde L (2000) Age-
related cognitive deficits mediated by changes in the striatal dopamine system. Am J
Psychiatry 157:635–637.
Bäckman L, Karlsson S, Fischer H, Karlsson P, Brehmer Y, Rieckmann A, MacDonald SWS,
Farde L, Nyberg L (2011) Dopamine D1 receptors and age differences in brain activation
during working memory. Neurobiol Aging 32:1849–1856.
Bäckman L, Nyberg L, Lindenberger U, Li S-C, Farde L (2006) The correlative triad among
aging, dopamine, and cognition: current status and future prospects. Neurosci Biobeh Rev
30:791–807.
Bateman et al. (2012) Clinical and biomarker changes in dominantly inherited Alzheimer’s
disease. N Engl J Med 367:795–804.
Becker JA, Hedden T, Carmasin J, Maye J, Rentz DM, Putcha D, Fischl B, Greve DN, Marshall
GA, Salloway S, Marks D, Buckner RL, Sperling RA, Johnson KA (2011) Amyloid-β
associated cortical thinning in clinically normal elderly. Ann Neurol 69:1032–1042.
Bero AW, Bauer AQ, Stewart FR, White BR, Cirrito JR, Raichle ME, Culver JP, Holtzman DM
(2012) Bidirectional relationship between functional connectivity and amyloid-β deposi-
tion in mouse brain. J Neurosci 32:4334–4340.
Bohnen NI, Djang DSW, Herholz K, Anzai Y, Minoshima S (2012) Effectiveness and safety of
18F-FDG PET in the evaluation of dementia: a review of the recent literature. J Nucl Med
53:59–71.
Braak E, Braak H (1997) Alzheimer’s disease: transiently developing dendritic changes in pyra-
midal cells of sector CA1 of the Ammon’s horn. Acta Neuropathol 93:323–325.
57
Braskie MN, Wilcox CE, Landau SM, O’Neil JP, Baker SL, Madison CM, Kluth JT, Jagust
WJ (2008) Relationship of striatal dopamine synthesis capacity to age and cognition. J
Neurosci 28:14320–14328.
Brickman AM, Muraskin J, Zimmerman ME (2009) Structural neuroimaging in Alzheimer’s
disease: do white matter hyperintensities matter? Dialogues Clin Neurosci 11:181–190.
Brickman AM (2013) Contemplating Alzheimer’s disease and the contribution of white matter
hyperintensities. Curr Neurol Neurosci Rep 13:415–415.
Brozoski T, Brown R, Rosvold H, Goldman P (1979) Cognitive deficit caused by regional
depletion of dopamine in prefrontal cortex of rhesus monkey. Science 205:929–932.
Brun A, Englund E (1986) A white matter disorder in dementia of the Alzheimer type: A patho-
anatomical study. Ann Neurol 19:253–262.
Buckholtz JW, Treadway MT, Cowan RL, Woodward ND, Li R, Ansari MS, Baldwin RM,
Schwartzman AN, Shelby ES, Smith CE, Kessler RM, Zald DH (2010) Dopaminergic
network differences in human impulsivity. Science 329:532–532.
Buckner RL (2004) Memory and executive function in aging and AD: multiple factors that
cause decline and reserve factors that compensate. Neuron 44:195–208.
Buckner RL, Sepulcre J, Talukdar T, Krienen FM, Liu H, Hedden T, Andrews-Hanna JR,
Sperling RA, Johnson KA (2009) Cortical hubs revealed by intrinsic functional connec-
tivity: mapping, assessment of stability, and relation to Alzheimer’s disease. J Neurosci
29:1860–1873.
Buckner RL, Snyder AZ, Shannon BJ, LaRossa G, Sachs R, Fotenos AF, Sheline YI, Klunk
WE, Mathis CA, Morris JC, Mintun MA (2005) Molecular, structural, and functional char-
acterization of Alzheimer’s disease: evidence for a relationship between default activity,
amyloid, and memory. J Neurosci 25:7709–7717.
Buckner RL, Krienen FM, Yeo BT (2013) Opportunities and limitations of intrinsic functional
connectivity MRI. Nat Neurosci 16:832–837.
Carlsson A, Winblad B (1976) Influence of age and time interval between death and autopsy
on dopamine and 3-methoxytyramine Levels in human basal ganglia. J Neural Transm
38:271–276.
Chételat G, Landeau B, Salmon E, Yakushev I, Bahri MA, Mézenge F, Perrotin A, Bastin C,
Manrique A, Scheurich A, Scheckenberger M, Desgranges B, Eustache F, Fellgiebel A
(2013) Relationships between brain metabolism decrease in normal aging and changes in
structural and functional connectivity. Neuroimage 76:167–177.
Chhatwal JP et al. (2013) Impaired default network functional connectivity in autosomal domi-
nant Alzheimer disease. Neurology 81:736–744.
Chien DT, Bahri S, Szardenings AK, Walsh JC (2013) Early clinical PET imaging results with
the novel PHF-tau radioligand [F-18]-T807. J Alzheimers Dis 34:457–468.
Chien DT, Szardenings AK, Bahri S, Walsh JC (2014) Early clinical PET imaging results with
the novel PHF-tau radioligand [F18]-T808. J Alzheimers Dis 38:171–184.
Cohen AD, Klunk WE (2014) Early detection of Alzheimer’s disease using PiB and FDG PET.
Neurobiol Dis 72:117–122.
Cohen J (1988) Statistical power analysis for the behavioral sciences, 2nd ed. Hillsdale,
NJ: Lawrence Earlbaum Associates.
Cools R, Frank MJ, Gibbs SE, Miyakawa A, Jagust W, D’Esposito M (2009) Striatal dopa-
mine predicts outcome-specific reversal learning and its sensitivity to dopaminergic drug
administration. J Neurosci 29:1538–1543.
De Leon MJ, Convit A, Wolf OT, Tarshish CY, DeSanti S, Rusinek H, Tsui W, Kandil E, Scherer
AJ, Roche A, Imossi A, Thorn E, Bobinski M, Caraos C, Lesbre P, Schlyer D, Poirier J,
Reisberg B, Fowler J (2001) Prediction of cognitive decline in normal elderly subjects
58
with 2-[18F]fluoro-2-deoxy-D-glucose/positron-emission tomography (FDG/PET). Proc

Natl Acad Sci USA 98:10966–10971.
DeCarli C, Murphy DG, Tranh M, Grady CL, Haxby JV, Gillette JA, Salerno JA, Gonzales-
Aviles A, Horwitz B, Rapoport SI (1995) The effect of white matter hyperintensity vol-
ume on brain structure, cognitive performance, and cerebral metabolism of glucose in 51
healthy adults. Neurology 45:2077–2084.
Delacourte A, Sergeant N, Wattez A, Maurage C-A, Lebert F, Pasquier F, David J-P (2002)
Tau aggregation in the hippocampal formation: an ageing or a pathological process? Exp
Gerontol 37:1291–1296.
Dewey SL, Volkow ND, Logan J, MacGregor RR, Fowler JS, Schlyer DJ, Bendriem B (1990)
Age-related decreases in muscarinic cholinergic receptor binding in the human brain mea-
sured with positron emission tomography (PET). J Neurosci Res 27:569–575.
Dickerson BC, Bakkour A, Salat DH, Feczko E, Pacheco J, Greve DN, Grodstein F, Wright CI,
Blacker D, Rosas HD, Sperling RA, Atri A, Growdon JH, Hyman BT, Morris JC, Fischl
B, Buckner RL (2009) The cortical signature of Alzheimer’s disease: regionally specific
cortical thinning relates to symptom severity in very mild to mild AD sementia and is
detectable in asymptomatic amyloid-positive individuals. Cereb Cortex 19:497–510.
Doré V, Villemagne VL, Bourgeat P, Fripp J, Acosta O, Chételat G, Zhou L, Martins R,
Ellis KA, Masters CL, Ames D, Salvado O, Rowe CC (2013) Cross-sectional and lon-
gitudinal analysis of the relationship between Aβ deposition, cortical thickness, and
memory in cognitively unimpaired individuals and in Alzheimer disease. JAMA Neurol
70:903–911.
Dowling NM, Johnson SC, Gleason CE, Jagust WJ, Alzheimer’s Disease Neuroimaging
Initiative (2015) The mediational effects of FDG hypometabolism on the associa-
tion between cerebrospinal fluid biomarkers and neurocognitive function. Neuroimage
105:357–368.
Drzezga A, Becker JA, Van Dijk KRA, Sreenivasan A, Talukdar T, Sullivan C, Schultz AP,
Sepulcre J, Putcha D, Greve D, Johnson KA, Sperling RA (2011) Neuronal dysfunction
and disconnection of cortical hubs in non-demented subjects with elevated amyloid bur-
den. Brain 134:1635–1646.
Durstewitz D, Seamans JK, Sejnowski TJ (2000) Dopamine-mediated stabilization of delay-
period activity in a network model of prefrontal cortex. J Neurophysiol 83:1733–1750.
Elman JA, Oh H, Madison CM, Baker SL, Vogel JW, Marks SM, Crowley S, O’Neil JP, Jagust
WJ (2014a) Neural compensation in older people with brain amyloid-β deposition. Nat
Neurosci 17:1316–1318.
Elman JA, Madison CM, Baker SL, Vogel JW, Marks SM, Crowley S, O’Neil JP, Jagust
WJ (2014b) Effects of beta-amyloid on resting state functional connectivity within and
between networks reflect known patterns of regional vulnerability. Cereb Cortex. Epub
ahead of print. doi: 10.1093/cercor/bhu259
Erixon-Lindroth N, Farde L, Wahlin T-BR, Sovago J, Halldin C, Bäckman L (2005) The role of
the striatal dopamine transporter in cognitive aging. Psychiatry Res 138:1–12.
Ewers M, Brendel M, Rizk-Jackson A, Rominger A, Bartenstein P, Schuff N, Weiner MW,
Alzheimer’s Disease Neuroimaging Initiative (2014) Reduced FDG-PET brain metabo-
lism and executive function predict clinical progression in elderly healthy subjects.
Neurology Clin 4:45–52.
Farde L, Hall H, Ehrin E, Sedvall G (1986) Quantitative analysis of D2 dopamine receptor bind-
ing in the living human brain by PET. Science 231:258–261.
Fawaz MV, Brooks AF, Rodnick ME, Carpenter GM, Shao X, Desmond TJ, Sherman P,
Quesada CA, Hockley BG, Kilbourn MR, Albin RL, Frey KA, Scott PJH (2014) High
affinity radiopharmaceuticals based upon lansoprazole for PET imaging of aggregated
59
tau in Alzheimer’s disease and progressive supranuclear palsy: synthesis, preclinical eval-

uation, and lead selection. ACS Chem Neurosci 5:718–730.
Fearnley JM, Lees AJ (1991) Ageing and Parkinson’s disease: substantia nigra regional selec-
tivity. Brain 114:2283–2301.
Fischer H, Nyberg L, Karlsson S, Karlsson P, Brehmer Y, Rieckmann A, MacDonald SWS,
Farde L, Bäckman L (2010) Simulating neurocognitive aging: effects of a dopaminergic
antagonist on brain activity during working memory. Biol Psychiatry 67:575–580.
Fjell AM, McEvoy L, Holland D, Dale AM, Walhovd KB (2013) Brain changes in older adults
at very low risk for Alzheimer’s disease. J Neurosci 33:8237–8242.
Fjell AM, Walhovd KB (2012) Neuroimaging results impose new views on Alzheimer’s dis-
ease-the role of amyloid revised. Mol Neurobiol 45:153–172.
Fjell AM, Westlye LT, Grydeland H, Amlien I, Espeseth T, Reinvang I, Raz N, Dale AM,
Walhovd KB, Walhovd KB (2014) Accelerating cortical thinning: unique to dementia or
universal in aging? Cereb Cortex 24:919–934.
Floel A, Garraux G, Xu B, Breitenstein C, Knecht S (2008) Levodopa increases mem-
ory encoding and dopamine release in the striatum in the elderly. Neurobiol Aging
29:267–279.
Forsberg A, Juréus A, Cselényi Z, Eriksdotter M, Freund-Levi Y, Jeppsson F, Swahn B-M,
Sandell J, Julin P, Schou M, Andersson J, Johnström P, Varnäs K, Halldin C, Farde L,
Svensson S (2013) Low background and high contrast PET imaging of amyloid-β with
[11C]AZD2995 and [11C]AZD2184 in Alzheimer’s disease patients. Eur J Nucl Med Mol
Imaging 40:580–593.
Fortea J, Vilaplana E, Alcolea D, Carmona-Iragui M, Sánchez-Saudinos M-B, Sala I, Antón-
Aguirre S, González S, Medrano S, Pegueroles J, Morenas E, Clarimón J, Blesa R, Lleó
A, Lleó A (2014) Cerebrospinal fluid β-amyloid and phospho-tau biomarker interactions
affecting brain structure in preclinical Alzheimer disease. Ann Neurol 76:223–230.
Fox MD, Raichle ME (2007) Spontaneous fluctuations in brain activity observed with func-
tional magnetic resonance imaging. Nat Rev Neurosci 8:700–711.
Frank MJ, O’Reilly RC (2006) A mechanistic account of striatal dopamine function in human
cognition: psychopharmacological studies with cabergoline and haloperidol. Behav
Neurosci 120:497–517.
Gainetdinov RR, Jones SR, Caron MG (1999) Functional hyperdopaminergia in dopamine
transporter knock-out mice. Biol Psychiatry 46:303–311.
Grandjean J, Schroeter A, He P, Tanadini M, Keist R, Krstic D, Konietzko U, Klohs J, Nitsch
RM, Rudin M (2014) Early alterations in functional connectivity and white matter struc-
ture in a transgenic mouse model of cerebral amyloidosis. J Neurosci 34:13780–13789.
Gurol ME, Viswanathan A, Gidicsin C, Hedden T, Martinez-Ramirez S, Dumas A, Vashkevich
A, Ayres AM, Auriel E, van Etten E, Becker A, Carmasin J, Schwab K, Rosand J, Johnson
KA, Greenberg SM (2013) Cerebral amyloid angiopathy burden associated with leuko-
araiosis: a positron emission tomography/magnetic resonance imaging study. Ann Neurol
73:529–536.
Guzman VA, Carmichael OT, Schwarz C, Tosto G, Zimmerman ME, Brickman AM (2013)
White matter hyperintensities and amyloid are independently associated with entorhinal
cortex volume among individuals with mild cognitive impairment. Alzheimers Dement
9:S124–S131.
Haight TJ, Landau SM, Carmichael O, Schwarz C, DeCarli C, Jagust WJ, Alzheimer’s Disease
Neuroimaging Initiative (2013) Dissociable effects of Alzheimer disease and white matter
hyperintensities on brain metabolism. JAMA Neurol 70:1039–1045.
Hedden T, Gabrieli JDE (2004) Insights into the ageing mind: a view from cognitive neurosci-
ence. Nat Rev Neurosci 5:87–96.
60
Hedden T, Mormino EC, Amariglio RE, Younger AP, Schultz AP, Becker JA, Buckner
RL, Johnson KA, Sperling RA, Rentz DM (2012) Cognitive profile of amyloid bur-
den and white matter hyperintensities in cognitively normal older adults. J Neurosci
32:16233–16242.
Hedden T, Oh H, Younger AP, Patel TA (2013) Meta-analysis of amyloid-cognition relations in
cognitively normal older adults. Neurology 80:1341–1348.
Hedden T, Schultz AP, Rieckmann A, Mormino EC, Johnson KA, Sperling RA, Buckner RL
(2016) Multiple brain markers are linked to age-related variation in cognition. Cereb
Cortex 26:1388–1400.
Hedden T, Van Dijk KRA, Becker JA, Mehta A, Sperling RA, Johnson KA, Buckner RL (2009)
Disruption of functional connectivity in clinically normal older adults harboring amyloid
burden. J Neurosci 29:12686–12694.
Herholz K, Salmon E, Perani D, Baron JC, Holthoff V, Frölich L, Schönknecht P, Ito K, Mielke
R, Kalbe E (2002) Discrimination between Alzheimer Dementia and controls by auto-
mated analysis of multicenter FDG PET. Neuroimage 17:302–316.
Hoffman JM, Welsh-Bohmer KA, Hanson M, Crain B, Hulette C, Earl N, Coleman RE
(2000) FDG PET imaging in patients with pathologically verified dementia. J Nucl Med
41:1920–1928.
Hsu PJ, Shu H, Benzinger T, Marcus D, Durbin T, Morris JC, Sheline YI (2015) Amyloid
burden in cognitively normal elderly is associated with preferential hippocampal subfield
volume loss. J Alzheimers Dis 45:27–33.
Huijbers W, Mormino EC, Wigman SE, Ward AM, Vannini P, McLaren DG, Becker JA, Schultz
AP, Hedden T, Johnson KA, Sperling RA (2014) Amyloid deposition is linked to aberrant
entorhinal activity among cognitively normal older adults. J Neurosci 34:5200–5210.
Ibáñez V, Pietrini P, Furey ML, Alexander GE, Millet P, Bokde ALW, Teichberg D, Schapiro
MB, Horwitz B, Rapoport SI (2004) Resting state brain glucose metabolism is not reduced
in normotensive healthy men during aging, after correction for brain atrophy. Brain Res
Bull 63:147–154.
Inoue K, Ito H, Uchida S, Taki Y, Kinomura S, Tsuji I, Sato S, Horie K, Kawashima R, Ito M,
Fukuda H (2008) Decrease in glucose metabolism in frontal cortex associated with dete-
rioration of microstructure of corpus callosum measured by diffusion tensor imaging in
healthy elderly. Hum Brain Mapp 29:375–384.
Insel PS, Mattsson N, Donohue MC, Mackin RS, Aisen PS, Jack CR, Shaw LM, Trojanowski
JQ, Weiner MW, Alzheimer’s Disease Neuroimaging Initiative (2015) The transitional
association between β-amyloid pathology and regional brain atrophy. Alzheimers Dement
11:1171–1179.
Ishibashi K, Ishii K, Oda K, Kawasaki K, Mizusawa H, Ishiwata K (2009) Regional analysis
of age-related decline in dopamine transporters and dopamine D2-like receptors in human
striatum. Synapse 63:282–290.
Ito H, Shimada H, Shinotoh H, Takano H (2014) Quantitative analysis of amyloid deposi-
tion in Alzheimer Disease using PET and the radiotracer 11C-AZD2184. J Nucl Med
55:932–938.
Jack CR, Wiste HJ, Vemuri P, Weigand SD, Senjem ML, Zeng G, Bernstein MA, Gunter JL,
Pankratz VS, Aisen PS, Weiner MW, Petersen RC, Shaw LM, Trojanowski JQ, Knopman
DS, and the Alzheimer’s Disease Neuroimaging Initiative (2010) Brain beta-amyloid mea-
sures and magnetic resonance imaging atrophy both predict time-to-progression from mild
cognitive impairment to Alzheimer’s disease. Brain 133:3336–3348.
Jack CR, Knopman DS, Weigand SD, Wiste HJ, Vemuri P, Lowe V, Kantarci K, Gunter JL,
Senjem ML, Ivnik RJ, Roberts RO, Rocca WA, Boeve BF, Petersen RC (2012) An
61
operational approach to National Institute on Aging-Alzheimer’s Association criteria for

preclinical Alzheimer disease. Ann Neurol 71:765–775.
Jack CR, Wiste HJ, Weigand SD, Knopman DS, Lowe V, Vemuri P, Mielke MM, Jones DT,
Senjem ML, Gunter JL, Gregg BE, Pankratz VS, Petersen RC (2013) Amyloid-first and
neurodegeneration-first profiles characterize incident amyloid PET positivity. Neurology
81:1732–1740.
Jagust W (2013) Vulnerable neural systems and the borderland of brain aging and neurodegen-
eration. Neuron 77:219–234.
Jagust W, Gitcho A, Sun F, Kuczynski B, Mungas D, Haan M (2006) Brain imaging evidence of
preclinical Alzheimer’s disease in normal aging. Ann Neurol 59:673–681.
Jagust WJ, Landau SM, For the Alzheimer’s Disease Neuroimaging Initiative (2012)
Apolipoprotein E, not fibrillar-amyloid, reduces cerebral glucose metabolism in normal
aging. J Neurosci 32:18227–18233.
Johnson KA, Gregas M, Becker JA, Kinnecom C, Salat DH, Moran EK, Smith EE, Rosand J,
Rentz DM, Klunk WE, Mathis CA, Price JC, DeKosky ST, Fischman AJ, Greenberg SM
(2007) Imaging of amyloid burden and distribution in cerebral amyloid angiopathy. Ann
Neurol 62:229–234.
Jonasson LS, Axelsson J, Riklund K, Braver TS, Ögren M, Bäckman L, Nyberg L (2014)
Dopamine release in nucleus accumbens during rewarded task switching measured by
[¹¹C]raclopride. Neuroimage 99:357–364.
Jones DT, Vemuri P, Murphy MC, Gunter JL, et al. (2012) Non-stationarity in the “resting
brain’s” modular architecture. PLoS ONE 7:e39731.
Kalpouzos G, Chételat G, Baron J-C, Landeau B, Mevel K, Godeau C, Barré L, Constans J-M,
Viader F, Eustache F, Desgranges B (2009) Voxel-based mapping of brain gray matter
volume and glucose metabolism profiles in normal aging. Neurobiol Aging 30:112–124.
Kantarci K, Schwarz CG, Reid RI, Przybelski SA, Lesnick TG, Zuk SM, Senjem ML, Gunter
JL, Lowe V, Machulda MM, Knopman DS, Petersen RC, Jack CR (2014) White matter
integrity determined with diffusion tensor imaging in older adults without dementia: influ-
ence of amyloid load and neurodegeneration. JAMA Neurol 71:1547–1554.
Karlsson S, Nyberg L, Karlsson P, Fischer H, Thilers P, Macdonald S, Brehmer Y, Rieckmann
A, Halldin C, Farde L, Bäckman L (2009) Modulation of striatal dopamine D1 binding by
cognitive processing. Neuroimage 48:398–404.
Karlsson S, Rieckmann A, Karlsson P, Farde L, Nyberg L, Bäckman L (2011) Relationship of
dopamine D1 receptor binding in striatal and extrastriatal regions to cognitive functioning
in healthy humans. Neuroimage 57:346–351.
Kastyak-Ibrahim MZ, Di Curzio DL, Buist R, Herrera SL, Albensi BC, Del Bigio MR, Martin M
(2013) Neurofibrillary tangles and plaques are not accompanied by white matter pathology
in aged triple transgenic-Alzheimer disease mice. Magn Reson Imaging 31:1515–1521.
Katzman R, Terry R, DeTeresa R, Brown T, Davies P, Fuld P, Renbing X, Peck A (1988)
Clinical, pathological, and neurochemical changes in dementia: a subgroup with preserved
mental status and numerous neocortical plaques. Ann Neurol 23:138–144.
Kennedy KM, Rodrigue KM, Devous MDSr., Hebrank AC, Bischof GN, Park DC (2012)
Effects of beta-amyloid accumulation on neural function during encoding across the adult
lifespan. Neuroimage 62:1–8.
Kim J, Basak JM, Holtzman DM (2009) The role of Apolipoprotein E in Alzheimer’s disease.
Neuron 63:287–303.
Kim YJ, Ichise M, Ballinger JR, Vines D, Erami SS, Tatschida T, Lang AE (2002) Combination
of dopamine transporter and D2 receptor SPECT in the diagnostic evaluation of PD, MSA,
and PSP. Mov Disord 17:303–312.
62
Klostermann EC, Braskie MN, Landau SM, O’Neil JP, Jagust WJ (2012) Dopamine and fron-
tostriatal networks in cognitive aging. Neurobiol Aging 33:623–624.
Klunk WE et al. (2004) Imaging brain amyloid in Alzheimer’s disease with Pittsburgh
Compound-B. Ann Neurol 55:306–319.
Klunk WE, Wang Y, Huang G-F, Debnath ML, Holt DP, Shao L, Hamilton RL, Ikonomovic
MD, DeKosky ST, Mathis CA (2003) The binding of 2(4’methylaminophenyl)benzo-
thiazole to postmortem brain homogenates is dominated by the amyloid component. J
Neurosci 23:2086–2092.
Knopman DS, Jack CR, Wiste HJ, Lundt ES, Weigand SD, Vemuri P, Lowe VJ, Kantarci K,
Gunter JL, Senjem ML, Mielke MM, Roberts RO, Boeve BF, Petersen RC (2014) 18F-
fluorodeoxyglucose positron emission tomography, aging, and apolipoprotein E genotype
in cognitively normal persons. Neurobiol Aging 35:2096–2106.
Knopman DS, Jack CR, Wiste HJ, Weigand SD, Vemuri P, Lowe VJ, Kantarci K, Gunter JL,
Senjem ML, Mielke MM, Roberts RO, Boeve BF, Petersen RC (2013) selective worsen-
ing of brain injury biomarker abnormalities in cognitively normal elderly persons with
β-amyloidosis. JAMA Neurol 70:1030–1038.
Kochunov P, Ramage AE, Lancaster JL, Robin DA, Narayana S, Coyle T, Royall DR, Fox P
(2009) Loss of cerebral white matter structural integrity tracks the gray matter metabolic
decline in normal aging. Neuroimage 45:17–28.
Koepp MJ, Gunn RN, Lawrence AD, Cunningham VJ, Dagher A, Jones T, Brooks DJ, Bench
CJ, Grasby PM (1998) Evidence for striatal dopamine release during a video game. Nature
393:266–268.
Landau SM, Harvey D, Madison CM, Koeppe RA, Reiman EM, Foster NL, Weiner MW, Jagust
WJ, Alzheimer’s Disease Neuroimaging Initiative (2011) Associations between cogni-
tive, functional, and FDG-PET measures of decline in AD and MCI. Neurobiol Aging
32:1207–1218.
Landau SM, Lal R, O’Neil JP, Baker S, Jagust WJ (2009) Striatal dopamine and working mem-
ory. Cereb Cortex 19:445–454.
Landau SM, Mintun MA, Joshi AD, Koeppe RA, Petersen RC, Aisen PS, Weiner MW, Jagust
WJ, For the Alzheimer’s Disease Neuroimaging Initiative (2012) Amyloid deposition,
hypometabolism, and longitudinal cognitive decline. Ann Neurol 72:578–586.
Laruelle M (2000) Imaging synaptic neurotransmission with in vivo binding competition tech-
niques: a critical review. J Cereb Blood Flow Metab 20:423–451.
Lim HK, Nebes R, Snitz B, Cohen A, Mathis C, Price J, Weissfeld L, Klunk W, Aizenstein HJ
(2014) Regional amyloid burden and intrinsic connectivity networks in cognitively normal
elderly subjects. Brain 137:3327–3338.
Lim YY, Ellis KA, Pietrzak RH, Ames D, Darby D, Harrington K, Martins RN, Masters CL,
Rowe C, Savage G, Szoeke C, Villemagne VL, Maruff P, AIBL Research Group (2012)
Stronger effect of amyloid load than APOE genotype on cognitive decline in healthy older
adults. Neurology 79:1645–1652.
Lindenberger UU, Oertzen von TT, Ghisletta PP, Hertzog CC (2011) Cross-sectional age vari-
ance extraction: what’s change got to do with it? Psychol Aging 26:34–47.
Lo RY, Jagust WJ, Alzheimer’s Disease Neuroimaging Initiative (2012) Vascular burden and
Alzheimer disease pathologic progression. Neurology 79:1349–1355.
Loessner A, Alavi A, Lewandrowski KU, Mozley D, Souder E, Gur RE (1995) Regional cere-
bral function determined by FDG-PET in healthy volunteers: normal patterns and changes
with age. J Nucl Med 36:1141–1149.
Logan J, Fowler JS, Volkow ND, Wolf AP, Dewey SL, Schlyer DJ, MacGregor RR, Hitzemann R,
Bendriem B, Gatley SJ, Christman DR (1990) Graphical analysis of reversible radioligand
63
binding from time-activity measurements applied to [N-11C-methyl]-(-)-cocaine PET

studies in human subjects. J Cereb Blood Flow Metab 10:740–747.
Lövdén M, Köhncke Y, Laukka EJ, Kalpouzos G, Salami A, Li T-Q, Fratiglioni L, Bäckman L
(2014) Changes in perceptual speed and white matter microstructure in the corticospinal
tract are associated in very old age. Neuroimage 102 Pt 2:520–530.
Lowe VJ, Weigand SD, Senjem ML, Vemuri P, Jordan L, Kantarci K, Boeve B, Jack CR,
Knopman D, Petersen RC (2014) Association of hypometabolism and amyloid levels in
aging, normal subjects. Neurology 82:1959–1967.
MacDonald SWS, Cervenka S, Farde L, Nyberg L, Bäckman L (2009) Extrastriatal dopamine
D2 receptor binding modulates intraindividual variability in episodic recognition and
executive functioning. Neuropsychologia 47:2299–2304.
MacDonald SWS, Karlsson S, Rieckmann A, Nyberg L, Bäckman L (2012) Aging-related
increases in behavioral variability: relations to losses of dopamine D1 receptors. J Neurosci
32:8186–8191.
Madden DJ, Bennett IJ, Song AW (2009) Cerebral white matter integrity and cognitive
aging: contributions from diffusion tensor imaging. Neuropsychol Rev 19:415–435.
Mangialasche F, Solomon A, Winblad B, Mecocci P, Kivipelto M (2010) Alzheimer’s dis-
ease: clinical trials and drug development. Lancet Neurol 9:702–716.
Marchant NL, Reed BR, DeCarli CS, Madison CM, Weiner MW, Chui HC, Jagust WJ
(2012) Cerebrovascular disease, β-amyloid, and cognition in aging. Neurobiol Aging
33:1006.e25–.e36.
Mattsson N, Insel PS, Nosheny R, Tosun D, Trojanowski JQ, Shaw LM, Jack CR, Donohue
MC, Weiner MW, Alzheimer’s Disease Neuroimaging Initiative (2014) Emerging β-
amyloid pathology and accelerated cortical atrophy. JAMA Neurol 71:725–734.
Maurer K, Volk S, Gerbaldo H (1997) Auguste D and Alzheimer’s disease. Lancet
349:1549–1549.
McGeer EG, McGeer PL, Harrop R, Akiyama H, Kamo H (1990) Correlations of regional post-
mortem enzyme activities with premortem local glucose metabolic rates in Alzheimer’s
disease. J Neurosci Res 27:612–619.
McGeer PL, Kamo H, Harrop R, McGeer EG, Martin WR, Pate BD, Li DK (1986) Comparison
of PET, MRI, and CT with pathology in a proven case of Alzheimer’s disease. Neurology
36:1569–1574.
Meltzer CC, Leal JP, Mayberg HS, Wagner HN, Frost JJ (1990) Correction of PET data for
partial volume effects in human cerebral cortex by MR imaging. J Comput Assist Tomogr
14:561–570.
Meltzer CC, Smith G, DeKosky ST, Pollock BG, Mathis CA, Moore RY, Kupfer DJ, Reynolds
CF (1998) Serotonin in aging, late-life depression, and Alzheimer’s disease: the emerging
role of functional imaging. Neuropsychopharmacol 18:407–430.
Mintun MA, Larossa GN, Sheline YI, Dence CS, Lee SY, Mach RH, Klunk WE, Mathis CA,
DeKosky ST, Morris JC (2006) [11C]PIB in a nondemented population: potential anteced-
ent marker of Alzheimer disease. Neurology 67:446–452.
Molinuevo JL, Ripolles P, Simó M, Lladó A, Olives J, Balasa M, Antonell A, Rodriguez-
Fornells A, Rami L (2014) White matter changes in preclinical Alzheimer’s disease: a
magnetic resonance imaging-diffusion tensor imaging study on cognitively normal older
people with positive amyloid β protein 42 levels. Neurobiol Aging 35:2671–2680.
Monchi O, Hyun Ko J, Strafella AP (2006) Striatal dopamine release during performance of
executive functions: A [11C] raclopride PET study. Neuroimage 33:907–912.
Mormino EC, Betensky RA, Hedden T, Schultz AP, Ward A, Huijbers W, Rentz DM, Johnson
KA, Sperling RA, Alzheimer’s Disease Neuroimaging Initiative, Australian Imaging
64
Biomarkers and Lifestyle Flagship Study of Ageing, Harvard Aging Brain Study (2014a)
Amyloid and APOE ε4 interact to influence short-term decline in preclinical Alzheimer
disease. Neurology 82:1760–1767.
Mormino EC, Brandel MG, Madison CM, Marks S, Baker SL, Jagust WJ (2012) Aβ deposition
in aging is associated with increases in brain activation during successful memory encod-
ing. Cereb Cortex 22:1813–1823.
Mormino EC, Kluth JT, Madison CM, Rabinovici GD, Baker SL, Miller BL, Koeppe RA,
Mathis CA, Weiner MW, Jagust WJ, the Alzheimer’s Disease Neuroimaging Initiative
(2009) Episodic memory loss is related to hippocampal-mediated -amyloid deposition in
elderly subjects. Brain 132:1310–1323.
Mormino EC, Knopman DS, Betensky RA, Hedden T, Schultz AP, Ward A, Huijbers W,
Rentz DM, Johnson KA, Sperling RA (2014b) β-Amyloidosis and neurodegeneration in
Alzheimer disease: who’s on first? Neurology 82:1756–1757.
Mormino EC, Smiljic A, Hayenga AO, H Onami S, Greicius MD, Rabinovici GD, Janabi M,
Baker SL, V Yen I, Madison CM, Miller BL, Jagust WJ (2011) Relationships between
beta-amyloid and functional connectivity in different components of the default mode net-
work in aging. Cereb Cortex 21:2399–2407.
Morris JC, Roe CM, Xiong C, Fagan AM, Goate AM, Holtzman DM, Mintun MA (2010)
APOE predicts amyloid-beta but not tau Alzheimer pathology in cognitively normal aging.
Ann Neurol 67:122–131.
Morris JC, Storandt M, McKeel DW, Rubin EH, Price JL, Grant EA, Berg L (1996) Cerebral
amyloid deposition and diffuse plaques in “normal” aging: Evidence for presymptomatic
and very mild Alzheimer’s disease. Neurology 46:707–719.
Mosconi L (2013) Glucose metabolism in normal aging and Alzheimer’s disease:
Methodological and physiological considerations for PET studies. Clin Transl Imaging
1:217–233.
Mosconi L, Berti V, Glodzik L, Pupi A, De Santi S, de Leon MJ (2010) Pre-clinical detection
of Alzheimer’s disease using FDG-PET, with or without amyloid imaging. J Alzheimers
Dis 20:843–854.
Mosconi L, Mistur R, Switalski R, Tsui WH, Glodzik L, Li Y, Pirraglia E, De Santi S, Reisberg
B, Wisniewski T, de Leon MJ (2009) FDG-PET changes in brain glucose metabolism from
normal cognition to pathologically verified Alzheimer’s disease. Eur J Nucl Med Mol
Imaging 36:811–822.
Mosconi L, Tsui WH, Herholz K, Pupi A, Drzezga A, Lucignani G, Reiman EM, Holthoff V, Kalbe
E, Sorbi S, Diehl-Schmid J, Perneczky R, Clerici F, Caselli R, Beuthien-Baumann B, Kurz A,
Minoshima S, De Leon MJ (2008) Multicenter standardized 18F-FDG PET diagnosis of mild
cognitive impairment, Alzheimer’s disease, and other dementias. J Nucl Med 49:390–398.
Mozley LH, Gur RC, Mozley PD, Gur RE (2001) Striatal dopamine transporters and cognitive
functioning in healthy men and women. Am J Psychiatry 158:1492–1499.
Nebes RD, Snitz BE, Cohen AD, Aizenstein HJ, Saxton JA, Halligan EM, Mathis CA, Price JC,
Kamboh MI, Weissfeld LA, Klunk WE (2013) Cognitive aging in persons with minimal
amyloid-β and white matter hyperintensities. Neuropsychologia 51:2202–2209.
Nelson PT et al. (2012) Correlation of Alzheimer disease neuropathologic changes with cog-
nitive status. J Neuropathol Exp Neurol 71:362–381.
Nevalainen N, Riklund K, Andersson M, Axelsson J, Ogren M, Lövdén M, Lindenberger U,
Bäckman L, Nyberg L (2015) COBRA: A prospective multimodal imaging study of dopa-
mine, brain structure and function, and cognition. Brain Res 1612:83–103.
Nosheny RL, Insel PS, Truran D, Schuff N, Jack CR Jr., Aisen PS, Shaw LM, Trojanowski JQ,
Weiner MW (2015) Variables associated with hippocampal atrophy rate in normal aging
and mild cognitive impairment. Neurobiol Aging 36:273–282.
65
Nyberg L, Andersson M, Forsgren L, Jakobsson-Mo S, Larsson A, Marklund P, Nilsson L-

G, Riklund K, Bäckman L (2009) Striatal dopamine D2 binding is related to frontal
BOLD response during updating of long- term memory representations. Neuroimage
46:1194–1199.
Nyberg L, Salami A, Andersson M, Eriksson J, Kalpouzos G, Kauppi K, Lind J, Pudas S,
Persson J, Nilsson L-G (2010) Longitudinal evidence for diminished frontal cortex func-
tion in aging. Proc Natl Acad Sci USA 107:22682–22686.
Oh H, Habeck C, Madison C, Jagust W (2014) Covarying alterations in Aβ deposition, glu-
cose metabolism, and gray matter volume in cognitively normal elderly. Hum Brain Mapp
35:297–308.
Okamura N, Furumoto S, Harada R, Tago T, Yoshikawa T, Fodero-Tavoletti M, Mulligan
RS, Villemagne VL, Akatsu H, Yamamoto T, Arai H, Iwata R, Yanai K, Kudo Y (2013)
Novel 18F-labeled arylquinoline derivatives for noninvasive imaging of tau pathology in
Alzheimer disease. J Nucl Med 54:1420–1427.
Ossenkoppele R, Schonhaut DR, Baker SL, O’Neil JP, Janabi M, Ghosh PM, Santos M, Miller
ZA, Bettcher BM, Gorno-Tempini ML, Miller BL, Jagust WJ, Rabinovici GD (2015) Tau,
amyloid and hypometabolism in a patient with posterior cortical atrophy. Ann Neurol
77:338–342.
Ossenkoppele R, van der Flier WM, Verfaillie SC, Vrenken H, Versteeg A, van Schijndel RA,
Sikkes SA, Twisk J, Adriaanse SM, Zwan MD, Boellaard R, Windhorst AD, Barkhof F,
Scheltens P, Lammertsma AA, van Berckel BN (2014) Long-term effects of amyloid,
hypometabolism, and atrophy on neuropsychological functions. Neurology 82:1768–1775.
Ossenkoppele R, van der Flier WM, Zwan MD, Adriaanse SF, Boellaard R, Windhorst AD,
Barkhof F, Lammertsma AA, Scheltens P, van Berckel BNM (2013) Differential effect
of APOE genotype on amyloid load and glucose metabolism in AD dementia. Neurology
80:359–365.
Parkinson Progression Marker Initiative (2011) The Parkinson Progression Marker Initiative
(PPMI). Prog Neurobiol 95:629–635.
Phelps ME, Huang SC, Hoffman EJ, Selin C, Sokoloff L, Kuhl DE (1979) Tomographic mea-
surement of local cerebral glucose metabolic rate in humans with (F-18)2-fluoro-2-deoxy-
D-glucose: validation of method. Ann Neurol 6:371–388.
Pike KE, Ellis KA, Villemagne VL, Good N, Chételat G, Ames D, Szoeke C, Laws SM, Verdile
G, Martins RN, Masters CL, Rowe CC (2011) Cognition and beta-amyloid in preclinical
Alzheimer’s disease: Data from the AIBL study. Neuropsychologia 49:2384–2390.
Pike KE, Savage G, Villemagne VL, Ng S, Moss SA, Maruff P, Mathis CA, Klunk WE, Masters
CL, Rowe CC (2007) Beta-amyloid imaging and memory in non-demented individu-
als: evidence for preclinical Alzheimer’s disease. Brain 130:2837–2844.
Provenzano FA, Muraskin J, Tosto G, Narkhede A, Ben T Wasserman, Griffith EY, Guzman
VA, Meier IB, Zimmerman ME, Brickman AM, Initiative FTADN (2013) White matter
hyperintensities and cerebral amyloidosis: necessary and sufficient for clinical expression
of Alzheimer disease? JAMA Neurol 70:455–461.
Qin Y-Y, Li M-W, Zhang S, Zhang Y, Zhao L-Y, Lei H, Oishi K, Zhu W-Z (2013) In vivo quan-
titative whole-brain diffusion tensor imaging analysis of APP/PS1 transgenic mice using
voxel-based and atlas-based methods. Neuroradiology 55:1027–1038.
Rabinovici GD, Furst AJ, O’Neil JP, Racine CA, Mormino EC, Baker SL, Chetty S, Patel P,
Pagliaro TA, Klunk WE, Mathis CA, Rosen HJ, Miller BL, Jagust WJ (2007) 11C-PIB
PET imaging in Alzheimer disease and frontotemporal lobar degeneration. Neurology
68:1205–1212.
Racine AM, Adluru N, Alexander AL, Christian BT, Okonkwo OC, Oh J, Cleary CA, Birdsill
A, Hillmer AT, Murali D, Barnhart TE, Gallagher CL, Carlsson CM, Rowley HA, Dowling
66
NM, Asthana S, Sager MA, Bendlin BB, Johnson SC (2014) Associations between white
matter microstructure and amyloid burden in preclinical Alzheimer’s disease: A multi-
modal imaging investigation. Neuroimage Clin 4:604–614.
Rajah MN, D’Esposito M (2005) Region-specific changes in prefrontal function with
age: a review of PET and fMRI studies on working and episodic memory. Brain
128:1964–1983.
Raz N, Lindenberger U, Rodrigue KM (2005) Regional brain changes in aging healthy
adults: general trends, individual differences and modifiers. Cereb Cortex 15:1676–1689.
Reeves S, Bench C, Howard R (2002) Ageing and the nigrostriatal dopaminergic system. Int J
Geriat Psychiatry 17:359–370.
Reeves SJ, Grasby PM, Howard RJ, Bantick RA, Asselin M-C, Mehta MA (2005) A positron
emission tomography (PET) investigation of the role of striatal dopamine (D2) receptor
availability in spatial cognition. Neuroimage 28:216–226.
Resnick SM, Scheltens P (2012) MRI-based biomarkers of preclinical AD: An Alzheimer sig-
nature. Neurology 78:80–81.
Rieckmann A, Hedden T, Younger AP, Sperling RA, Johnson KA, Buckner RL, (2016)
Dopamine transporter availability in clinically normal aging is associated with individual
differences in white matter integrity. Human Brain Mapp 37:621–631.
Rieckmann A, Karlsson S, Fischer H, Bäckman L (2011b) Caudate dopamine D1 receptor den-
sity is associated with individual differences in frontoparietal connectivity during working
memory. J Neurosci 31:14284–14290.
Rieckmann A, Karlsson S, Fischer H, Bäckman L (2012) Increased bilateral frontal connectiv-
ity during working memory in young adults under the influence of a dopamine D1 receptor
antagonist. J Neurosci 32:17067–17072.
Rieckmann A, Karlsson S, Karlsson P, Brehmer Y, Fischer H, Farde L, Nyberg L, Bäckman L
(2011a) Dopamine D1 receptor associations within and between dopaminergic pathways
in younger and elderly adults: links to cognitive performance. Cereb Cortex 21:2023–2032.
Riederer P, Wuketich S (1976) Time course of nigrostriatal degeneration in parkinson’s disease.
A detailed study of influential factors in human brain amine analysis. J Neural Transm
38:277–301.
Riedl V, Bienkowska K, Strobel C, Tahmasian M, Grimmer T, Förster S, Friston KJ, Sorg C,
Drzezga A (2014) Local activity determines functional connectivity in the resting human
brain: a simultaneous FDG-PET/fMRI study. J Neurosci 34:6260–6266.
Rodrigue KM, Kennedy KM, Devous MD, Rieck JR, Hebrank AC, Diaz-Arrastia R, Mathews
D, Park DC (2012) β-Amyloid burden in healthy aging regional distribution and cognitive
consequences. Neurology 78:387–395.
Rönnlund M, Nyberg L, Bäckman L (2005) Stability, growth, and decline in adult life span
development of declarative memory: cross- sectional and longitudinal data from a
population-based study. Psych Aging 20:3–18.
Rutten-Jacobs LCA, de Leeuw F- E, Geurts-van Bon L, Gordinou de Gouberville MC,
Schepens-Franke AN, Dederen PJ, Spliet WGM, Wesseling P, Kiliaan AJ (2011) White
matter lesions are not related to β-amyloid deposition in an autopsy-based study. Curr
Gerontol Geriatr Res 2011:1–5.
Salloway S et al. (2014) Two phase 3 trials of bapineuzumab in mild-to-moderate Alzheimer’s
disease. N Engl J Med 370:322–333.
Salthouse TA (2014) Aging cognition unconfounded by prior test experience. J Gerontol B
Psychol Sci Soc Sci. Epub ahead of print.
Sawaguchi T, Goldman-Rakic P (1991) D1 dopamine receptors in prefrontal cortex: involve-
ment in working memory. Science 251:947–950.
67
Schaie KW (1996) Intellectual development in adulthood: the Seattle longitudinal study.

Cambridge University Press.
Schott BH, Minuzzi L, Krebs RM, Elmenhorst D, Lang M, Winz OH, Seidenbecher CI, Coenen
HH, Heinze H-J, Zilles K, Düzel E, Bauer A (2008) Mesolimbic functional magnetic reso-
nance imaging activations during reward anticipation correlate with reward-related ventral
striatal dopamine release. J Neurosci 28:14311–14319.
Schultz W, Dayan P, Montague PR (1997) A neural substrate of prediction and reward. Science
275:1593–1599.
Schwartz WJ, Smith CB, Davidsen L, Savaki H, Sokoloff L, Mata M, Fink DJ, Gainer H (1979)
Metabolic mapping of functional activity in the hypothalamo-neurohypophysial system of
the rat. Science 205:723–725.
Seeman P, Niznik HB (1990) Dopamine receptors and transporters in Parkinson’s disease and
schizophrenia. FASEB J 4:2737–2744.
Segovia G, Porras A, Del Arco A, Mora F (2001) Glutamatergic neurotransmission in aging: a
critical perspective. Mech Ageing Dev 122:1–29.
Servan-Schreiber D, Printz H, Cohen J (1990) A network model of catecholamine effects: gain,
signal-to-noise ratio, and behavior. Science 249:892–895.
Shankar GM, Li S, Mehta TH, Garcia-Munoz A, Shepardson NE, Smith I, Brett FM, Farrell
MA, Rowan MJ, Lemere CA, Regan CM, Walsh DM, Sabatini BL, Selkoe, DJ (2008)
Amyloid-beta protein dimers isolated directly from Alzheimer’s brains impair synaptic
plasticity and memory. Nat Med 14:837–842.
Shao X, Carpenter GM, Desmond TJ, Sherman P, Quesada CA, Fawaz M, Brooks AF, Kilbourn
MR, Albin RL, Frey KA, Scott PJH (2012) Evaluation of [(11)C]N-Methyl Lansoprazole
as a radiopharmaceutical for PET imaging of tau neurofibrillary tangles. ACS Med Chem
Lett 3:936–941.
Sheline YI, Raichle ME, Snyder AZ, Morris JC, Head D, Wang S, Mintun MA (2010) Amyloid
plaques disrupt resting state default mode network connectivity in cognitively normal
elderly. Biol Psychiatry 67:584–587.
Shokouhi S, Claassen D, Riddle W (2014) Imaging brain metabolism and pathology in
Alzheimer’s disease with positron emission tomography. J Alz Dis Parkinson 4:143.
Small SA, Duff K (2008) Linking Aβ and tau in late-onset Alzheimer’s disease: A Dual Pathway
Hypothesis. Neuron 60:534–542.
Sperling RA et al. (2011) Toward defining the preclinical stages of Alzheimer’s disease:
Recommendations from the National Institute on Aging- Alzheimer’s Association
workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement
7:280–292.
Sperling RA, Dickerson BC, Pihlajamaki M, Vannini P, Laviolette PS, Vitolo OV, Hedden T,
Becker JA, Rentz DM, Selkoe DJ, Johnson KA (2010) Functional alterations in memory
networks in early Alzheimer’s disease. Neuromolecular Med 12:27–43.
Sperling RA, Laviolette PS, O’Keefe K, O’Brien J, Rentz DM, Pihlajamaki M, Marshall G,
Hyman BT, Selkoe DJ, Hedden T, Buckner RL, Becker JA, Johnson KA (2009) Amyloid
deposition is associated with impaired default network function in older persons without
dementia. Neuron 63:178–188.
Sperling RA, Mormino EC, Johnson KA (2014b) The evolution of preclinical Alzheimer’s dis-
ease: Implications for clinical trials. Neuron 84:608–622.
Sperling RA, Rentz DM, Johnson KA, Karlawish J, Donohue M, Dalmon DP, Aisen P (2014a)
The A4 study: stopping AD before symptoms begin? Sci Transl Med 6:228fs13.
Spires-Jones TL, Hyman BT (2014) The intersection of amyloid beta and tau at synapses in
Alzheimer’s disease. Neuron 82:756–771.
68
Suhara T, Fukuda H, Inoue O, Itoh T, Suzuki K, Yamasaki T, Tateno Y (1991) Age-related

changes in human D1 dopamine receptors measured by positron emission tomography.
Psychopharmacology 103:41–45.
Sun SW, Liang HF, Mei J, Xu D, Shi WX (2014) In vivo diffusion tensor imaging of amyloid-
β-induced white matter damage in mice. J Alzheimers Dis 38:93–101.
Swerdlow RH (2007) Is aging part of Alzheimer’s disease, or is Alzheimer’s disease part of
aging? Neurobiol Aging 28:1465–1480.
Takahashi H, Kato M, Takano H, Arakawa R, Okumura M, Otsuka T, Kodaka F, Hayashi
M, Okubo Y, Ito H, Suhara T (2008) Differential contributions of prefrontal and hip-
pocampal dopamine D(1) and D(2) receptors in human cognitive functions. J Neurosci
28:12032–12038.
Terry AV, Buccafusco JJ (2003) The cholinergic hypothesis of age and Alzheimer’s disease-
related cognitive deficits: recent challenges and their implications for novel drug develop-
ment. J Pharmacol Exp Ther 306:821–827.
van Dyck CH, Avery RA, MacAvoy MG, Marek KL, Quinlan DM, Baldwin RM, Seibyl JP,
Innis RB, Arnsten AF (2008) Striatal dopamine transporters correlate with simple reaction
time in elderly subjects. Neurobiol Aging 29:1237–1246.
Vannini P, Hedden T, Becker JA, Sullivan C, Putcha D, Rentz D, Johnson KA, Sperling RA
(2012) Age and amyloid-related alterations in default network habituation to stimulus rep-
etition. Neurobiol Aging 33:1237–1252.
Vannini P, Hedden T, Huijbers W, Ward A, Johnson KA, Sperling RA (2013) The ups
and downs of the posteromedial cortex: age-and amyloid-related functional altera-
tions of the encoding/retrieval flip in cognitively normal older adults. Cereb Cortex
23:1317–1328.
Verhaeghen P, Salthouse TA (1997) Meta- analyses of age- cognition relations in adult-
hood: Estimates of linear and nonlinear age effects and structural models. Psychol Bull
122:231–249.
Villemagne VL, Fodero-Tavoletti MT, Masters CL (2015) Tau imaging: early progress and
future directions. Lancet Neurol 14:114–124.
Villeneuve S, Reed BR, Wirth M, Haase CM, Madison CM, Ayakta N, Mack W, Mungas D,
Chui HC, DeCarli C, Weiner MW, Jagust WJ (2014) Cortical thickness mediates the effect
of β-amyloid on episodic memory. Neurology 82:761–767.
Villien M, Wey H-Y, Mandeville JB, Catana C, Polimeni JR, Sander CY, Zürcher NR, Chonde
DB, Fowler JS, Rosen BR, Hooker JM (2014) Dynamic functional imaging of brain glu-
cose utilization using fPET-FDG. Neuroimage 100:192–199.
Volkow ND, Gur RC, Wang GJ (1998a) Association between decline in brain dopamine activ-
ity with age and cognitive and motor impairment in healthy individuals. Am J Psychiatry
155:344–349.
Volkow ND, Logan J, Fowler JS, Wang GJ, Gur RC, Wong C, Felder C, Gatley SJ, Ding YS,
Hitzemann R, Pappas N (2000) Association between age-related decline in brain dopa-
mine activity and impairment in frontal and cingulate metabolism. Am J Psychiatry
157:75–80.
Volkow ND, Wang GJ, Fowler JS, Ding YS, Gur RC, Gatley J, Logan J, Moberg PJ, Hitzemann
R, Smith G, Pappas N (1998b) Parallel loss of presynaptic and postsynaptic dopamine
markers in normal aging. Ann Neurol 44:143–147.
Walhovd KB, Fjell AM, Reinvang I, Lundervold A, Dale AM, Eilertsen DE, Quinn BT, Salat D,
Makris N, Fischl B (2005) Effects of age on volumes of cortex, white matter and subcorti-
cal structures. Neurobiology Aging 26:1261–1270.
69
Wang GJ, Volkow ND, Logan J, Fowler JS, Schlyer D, MacGregor RR, Hitzemann RJ, Gur RC,
Wolf AP (1995) Evaluation of age-related changes in serotonin 5-HT2 and dopamine D2
receptor availability in healthy human subjects. Life Sci 56:PL249–PL253.
Wang YY, Chan GLG, Holden JEJ, Dobko TT, Mak EE, Schulzer MM, Huser JMJ, Snow BJB,
Ruth TJT, Calne DBD, Stoessl AJA (1998) Age-dependent decline of dopamine D1 recep-
tors in human brain: a PET study. Synapse 30:56–61.
Whitwell JL, Tosakulwong N, Weigand SD, Senjem ML, Lowe VJ, Gunter JL, Boeve BF,
Knopman DS, Dickerson BC, Petersen RC, Jack CR (2013) Does amyloid deposition
produce a specific atrophic signature in cognitively normal subjects? Neuroimage Clin
2:249–257.
Wirth M, Madison CM, Rabinovici GD, Oh H, Landau SM, Jagust WJ (2013) Alzheimer’s dis-
ease neurodegenerative biomarkers are associated with decreased cognitive function but
not β-amyloid in cognitively normal older individuals. J Neurosci 33:5553–5563.
Wong DF, Wagner HN, Dannals RF, Links JM, Frost JJ, Ravert HT, Wilson AA, Rosenbaum
AE, Gjedde A, Douglass KH (1984) Effects of age on dopamine and serotonin receptors
measured by positron tomography in the living human brain. Science 226:1393–1396.
Yanase D, Matsunari I, Yajima K, Chen W, Fujikawa A, Nishimura S, Matsuda H, Yamada M
(2005) Brain FDG PET study of normal aging in Japanese: effect of atrophy correction.
Eur J Nucl Med Mol Imaging 32:794–805.
Yamamoto M, Suhara T, Okubo Y, Ichimiya T, Sudo Y, Inoue M, Takano A, Yasuno F, Yoshikawa
K, Tanada S (2002) Age-related decline of serotonin transporters in living human brain of
healthy males. Life Sci 71:751–757.
70
71
Age Differences in
Structural Connectivity
Diffusion Tensor Imaging and White
Matter Hyperintensities
David J. Madden
Emily L. Parks
T he white matter (myelinated) pathways of the brain are the infrastructure

supporting cortical organization and function. Myelin, a specialization
of glial cells, is produced by oligodendrocytes in the central nervous system (CNS),
and the myelination of axons, first described by Virchow in 1854, was the last major
evolutionary invention within the vertebrate CNS (Hartline and Colman, 2007; Seidl,
2014; Walhovd et al., 2014). Myelination increases axonal conduction velocity by the
propagation of action potentials at the unmyelinated gaps (nodes of Ranvier) between
the myelinated segments of the axon. Many factors influence conduction velocity
including axon diameter, myelin sheath thickness, internode length, and the serial
arrangement of internodes along the axon. In general, for a fixed axon diameter, con-
duction velocity will increase with myelin thickness. For a fixed total diameter (axon
plus myelin sheath), there is an optimal ratio of axon diameter to total fiber diam-
eter (approximately 0.60–0.70) at which conduction velocity is maximized (Waxman,
1980). Early anatomical studies noted age-related differences in myelinated axons.
Corbin and Gardner (1937), for example, counted the number of myelinated fibers in
the 8th and 9th thoracic dorsal and ventral roots of the spinal cord, in 34 human cadav-
ers ranging in age from 1 day to 84 years. The number of myelinated fibers increased
by 187% from infancy to the second and third decades, but decreased by 20%–32%
following the third decade.
71
72
More recently, the rapid development of functional neuroimaging research has

allowed the acquisition of more complex and detailed information about the brain in
vivo, particularly within gray matter. For example, scientists have identified reliable
patterns of regional activation of cerebral gray matter associated with sensory, motor,
and cognitive tasks. The resulting task-related cortical activation is often widespread,
presumably reflecting the operation of distributed neural networks (Mesulam, 1990;
McIntosh, 2000; Glascher et al., 2010). Interpretation of the patterns of cortical activa-
tion has led, in parallel, to the recognition that task-related activation within gray mat-
ter networks must rely, to some degree, on the underlying white matter infrastructure.
Questions regarding the organization, development, and neurobiological properties of
white matter have only recently come to the forefront of the cognitive neuroscience
of aging. These questions are most frequently expressed in terms of a disconnec-
tion model of cognitive aging, in which the age-related loss or deterioration of white
matter within a neural network, even in the absence of significant disease, is viewed
as a loss of structural connectivity within the neural networks comprising cognitive
function (Greenwood, 2000; O’Sullivan et al., 2001; Peters, 2002; Bartzokis, 2004a;
Charlton et al., 2006). The concept of disconnection is also prominent in neurological
models of brain disorders (Catani and Ffytche, 2005; Filley, 2005; Bartolomeo et al.,
2007; He et al., 2007), as proposed originally by Geschwind (1965a, b).
Diffusion tensor imaging (DTI) is a recently developed form of magnetic reso-
nance imaging (MRI) that yields information regarding the macrostructural and
microstructural properties of white matter. In this chapter we review recent research
on age-related differences in structural connectivity, with emphasis on DTI studies of
healthy white matter. Our goal is to highlight some of the critical themes and issues
associated with imaging white matter in the cognitive neuroscience of aging. Here we
do not focus on the methodology of DTI (Basser and Pierpaoli, 1996, 1998; Beaulieu,
2002; Mori and Zhang, 2006; Jones, 2008; Jones, 2011), but rather on the implica-
tions of DTI for understanding age-related differences in brain structure and cogni-
tion. Similarly, we do not review age-related variation in the volume of white matter.
Although this latter topic is important, and relevant for understanding age-related dif-
ferences in cognitive performance (Meier-Ruge et al., 1992; Fjell and Walhovd, 2010;
Walhovd et al., 2011), the volumetric measures do not translate directly to structural
connectivity and are less sensitive than DTI in the detection of age-related effects
(Hugenschmidt et al., 2008; Giorgio et al., 2010). We do consider other forms of
decline in tissue integrity, specifically white matter hyperintensities (WMHs) and
intracellular iron accumulation, which occur during healthy aging. These latter phe-
nomena are typically detected with imaging modalities other than DTI, but WMHs
and iron accumulation have a significant influence on the white matter infrastructure.
Throughout this review, we will refer to four themes that we see illustrated, to
different degrees, in recent research. First, we will emphasize the importance of con-
sidering alternative models of relations among neuroimaging and behavioral vari-
ables. Interpreting the relations among DTI measures of white matter, aging, and
behavioral measures of cognition should involve explicit consideration of not only
the degree to which effects are age-dependent but also alternative models of age-
related influences. The fact that DTI and behavioral measures are correlated in an
older adult sample, for example, does not necessarily imply that white matter has a
â•‡ 73
Age Differences in Structural Connectivityâ•… 73
causal role in the observed age-â•‰related effects in the behavioral measures. Second, to
what extent are different brain regions contributing independently rather than jointly
to the outcome measures? Although different anatomical regions may be of interest
based on established functions, it is an empirical question whether the brain imaging
variables from individual regions within a particular dataset are providing unique
information or instead represent a more general, whole-â•‰brain influence. Third, at
this early stage of neuroimaging research, the relation between the DTI measures of
white matter and MRI measures of cortical activation (both task-â•‰related and restingâ•‰
state) is far from clear. Fourth, as noted previously, we will evaluate the discon-
nection model as an integrative framework for age-â•‰related differences in DTI mea-
sures. Our review will be selective, as several reviews are available that touch on
these themes (Sullivan and Pfefferbaum, 2006, 2007; Madden et al., 2009a; Salat,
2011; Carmichael and Lockhart, 2012; Madden et al., 2012; Bennett and Rypma,
2013; Bennett and Madden, 2014; Salat, 2014b), and we will emphasize more recent
research not covered in these previous reviews.
Diffusion Tensor Imaging Measures of Cerebral White Matter
Diffusion tensor imaging (DTI) is a form of MRI that provides information regarding
white matter by measuring the properties of molecular water (Basser and Pierpaoli,
1996, 1998; Beaulieu, 2002; Mori and Zhang, 2006; Jones, 2008; Jones, 2011).
Variation in the diffusion-â•‰weighted MRI signal reflects the general mobility of water
molecules, which is influenced by temperature, viscosity, the presence of large mol-
ecules, and other factors. The mobility is also influenced by biological barriers to
diffusion, referred to as hindered diffusion (with an upper limit termed restricted dif-
fusion), including cell membranes, myelin sheaths, and microtubules. In fluid-â•‰filled
spaces, and within gray matter, molecular movement of water encounters fewer of
these hindrances, and diffusion is relatively non-â•‰directional (i.e., isotropic) in nature.
In white matter, however, diffusion is more directional (i.e., anisotropic) as micro-
structures such as axonal cell membranes and myelin sheaths bias water movement
to a single, primary direction along the length of the axon. If the diffusion-â•‰weighted
MRI signal is acquired in at least six non-â•‰collinear directions, then it is possible to
estimate a profile of diffusion (i.e., tensor) for each voxel with three perpendicular
directions (Figure 3.1).
The most frequently used DTI metrics obtained from the tensor profile are mean
diffusivity (MD), axial diffusivity (AD), radial diffusivity (RD), and fractional anisot-
ropy (FA). When diffusion is relatively anisotropic, as in white matter, the tensor
modeling assumes that diffusion has an ellipsoid shape, with a primary direction
(eigenvector) that corresponds to the long axis of a similarly aligned axon. Diffusion
along the primary eigenvector is termed AD, whereas RD is diffusion perpendicular
to the primary direction (e.g., the mean of the second and third eigenvectors). Mean
diffusivity is the average of AD and RD and is thus the overall rate of diffusivity,
independent of any directional component. The FA metric is a scalar measure repre-
senting the strength of the directionality of the diffusion, that is, the degree to which
one direction predominates, independent of the rate of diffusion. These metrics are
74
(A) (B) (C) λ1
λ2
λ3
(D) (E)
FA = 0.01
FA = 0.8
(F) (G)
A B C
White Matter Voxels
Diffusion Tensors
Figure 3.1 Contributions to tract anisotropy. (A) Water diffuses more easily along the axis
of a fiber bundle than it does across the axis of the bundle, due to the presence of barriers
such as membranes and myelin. (B) Typically, multiple different diffusion-weighted images are
acquired, with each one sensitized to diffusion along a different direction in space. (C) One can
fit a mathematical model to the measurements in order to estimate certain model parameters
that describe diffusion behavior within each voxel. The most commonly used model, the diffu-
sion tensor model, fits the measurements to a tensor, or ellipsoid, which is fully characterized
by its three orthogonal eigenvectors and their associated lengths, or eigenvalues (k1, k2, k3).
(D) In cerebral spinal fluid (CSF), water diffuses freely in all directions and so FA is close to
zero; in white matter, diffusion is directionally dependent and so FA is closer to one. (E) The
long axis of the diffusion tensor corresponds to the principal diffusion direction. Within a coher-
ent fiber bundle this aligns with the fiber direction. (F) By following these voxel-wise estimates
of principal diffusion directions it is possible to perform diffusion tractography, and reconstruct
estimates of fiber pathways. (G) Variations in diffusion parameters along tracts during norma-
tive development are likely a combination of tract-specific (e.g., myelin content, axonal char-
acteristics) and local environment contributions. Voxel 1 contains a tract of interest (yellow) as
well as a crossing tract (gray), resulting in low anisotropy measurements at this point. Voxel
2 contains only the tract of interest and exhibits high anisotropy. Within voxel 3 axons from
nearby gray matter join the tract and some axons break off, heading toward gray matter targets.
The result would be a drop in anisotropy measurements at this point in the tract. The figure is
reproduced with permission from (Walhovd et al., 2014). (See color plate also)
75
Age Differences in Structural Connectivity 75
modeled at the level of the voxel and can be averaged across voxels within regions of
interest defined by the investigator. Other DTI techniques provide additional infor-
mation regarding the macrostructural organization of white matter. Tractography, for
example, uses the diffusion data across adjacent voxels to estimate the shape of white
matter tracts and the diffusional metrics along each tract (Figure 3.2) (Basser et al.,
2000; Catani et al., 2002; Behrens et al., 2014).
As Jones et al. (2013) and Salat (2014b) have noted, it is critical to recog-
nize that the information provided by DTI refers to properties of the diffusion
of molecular water. Although these properties are influenced by the structure of
white matter, individual DTI metrics typically do not represent specific anatomical
properties. The diffusional information provided by DTI is incomplete and relies
on voxel size, the angular sampling of the diffusion space, the signal to noise ratio,
and many other variables related to the acquisition parameters and to the con-
straints on conducting in vivo imaging of humans. Thus, although variations in
DTI measures such as FA represent the influences from different sources of white
matter integrity, FA is not a direct measure of structural connectivity. Regions with
crossing fibers, for example, may exhibit a paradoxical increase in FA if one of the
two fiber populations undergoes degeneration (Pierpaoli et al., 2001; Douaud et
al., 2011; Concha, 2014). Further, highly organized formations of glial cells sur-
rounding traumatic brain injury may have a greater influence on increasing FA than
axonal or myelin integrity (Budde et al., 2011). Thus, DTI can reveal important
information regarding the microstructural integrity of white matter, and the white
matter tracts described by DTI tractography correspond to known anatomy (Miller
et al., 2011; Hubbard and Parker, 2014), but caution is needed when inferring neu-
robiological mechanisms on the basis of DTI.
(A) (B) (C) (D)
Figure 3.2 Diffusion tensor imaging (DTI) tractography. (A) Seed and target regions.
(B) Estimated fibers in the genu of the corpus callosum. (C) Estimated stream tube. (D) Estimated
stream tube with age group differences in fractional anisotropy (FA) color coded; warm colors
representing age-related decline. Modified and reproduced with permission from Davis et al.
(2009). (See color plate also)
76
The measures RD and AD provide more directionally specific information than

FA, which is relevant for discriminating between biological mechanisms such as
myelin integrity or axonal loss (see the following section, Age-â•‰Related Differences
in Diffusion Tensor Imaging Measures). Animal and human ex vivo studies suggest
that RD, representing diffusion perpendicular to the axon’s primary axis, may be dif-
ferentially sensitive to myelin integrity (Song et al., 2002; Song et al., 2005; Klawiter
et al., 2011). The RD and AD metrics, however, as with all DTI measures, are depen-
dent on data acquisition parameters and modeling assumptions (Wheeler-â•‰Kingshott
and Cercignani, 2009; Beaulieu, 2014). The individual DTI parameters are math-
ematical expressions of the diffusion and the manner in which diffusion is hindered.
Inferring the neurobiological properties of tissue from these parameters requires an
interpretive step.
For example, the underlying model for DTI typically relies on the assumption
of Gaussian diffusion, with a single diffusion tensor within a voxel, which may not
always be valid. To address this, researchers have been using different mathemati-
cal frameworks allowing for non-â•‰Gaussian diffusion (Ennis and Kindlmann, 2006;
Seunarine and Alexander, 2014). These methods can estimate specific neurobiological
properties, such as axonal density (Assaf and Cohen, 2014) and can more accurately
resolve crossing fibers, in some cases resolving multiple axon directions within a
single voxel (Frank, 2002; Liu et al., 2010; Fan et al., 2014).
Age-â•‰Related Differences in Diffusion Tensor Imaging Measures
Histological studies of postmortem data suggest that myelinated axons are lost during
normal aging, at a rate of approximately 10% per decade, with the smaller-â•‰diameter
axons being most vulnerable (Marner et al., 2003). These age-â•‰vulnerable axons are
located primarily within the later-â•‰myelinating, association cortical regions of the brain
and in the white matter tracts connecting these regions. Despite this pattern of overall
decline, axonal repair and even generation of new myelin occurs in adulthood as a
result of the continued generation of oligodendrocytes, the cells within the CNS that
are responsible for the myelination of axons (Peters, 2002; Peters and Sethares, 2003;
Wang and Young, 2014). The remyelination may not be as complete as that of the orig-
inal myelin sheaths, with thinner and shorter remyelinated segments. Additionally, the
newer oligodendrocytes may have inclusions or other structural deficiencies, so that
at some point during the aging process the oligodendrocyte production and myelin
generation do not keep pace with myelin loss. Bartzokis (2004a, 2011) has proposed
that this developmental trajectory of oligodendrocytes is the fundamental mechanism
of age-â•‰related decline in cerebral white matter integrity, which in turn contributes
to a disconnection among the neural networks comprising perception and cognition.
Across the whole lifespan, the development of white matter integrity has a generally
quadratic (inverted-â•‰U) shape (Bartzokis, 2004b, Figure 3.3; Kochunov et al., 2012;
Walhovd et al., 2014) (Figure 3.3), but DTI findings also suggest that a three-â•‰phase
trajectory may be more accurate, with an initial increase in myelination during child-
hood and young adulthood followed by a plateau during early and middle adulthood,
then by decline during later adulthood (Westlye et al., 2010).
77
Fractional anisotropy (FA) Axial (principal) diffusion (AD) Radial diffusion (RD) Mean diffusion (MD)
4 4 –4 –4
2 2 –2 –2
z-score
0 0 0 0
–2 2 2 –2
–4 4 4 –4
–6 6 6 –6
5 25 45 65 85 5 25 45 65 85 5 25 45 65 85 5 25 45 65 85
Years Years Years Years
Age-effects on (FA) White matter volume Cortical myelin montent Cortical volume
4 4 4
2 2 2
0 0 0
–2 –2 –2
–4 –4 –4
–6 –6 –6
Quadratic age-effect 5 25 45 65 85 5 25 45 65 85 5 25 45 65 85
No quadratic age-effect years years years
Figure 3.3 Multimodal imaging of white matter through the lifespan. Results are based on 430 well-screened healthy participants between 8 and 85 years (mean 41.6 years).
Values in the scatterplots are expressed in z-scores (standard deviations) to ease comparison between metrics. Values represent for FA, axial, radial and mean diffusion the mean
of all voxels that were included in the left superior longitudinal fasciculus. The tract-based spatial statistics skeleton represents the middle of the tract for all participants (red and
green voxels in the lower left brain image). White matter volume represents the total volume of all cerebral white matter, and cortical volume represents the volume of all cortical
gray matter, in both cases corrected for total intracranial volume. Cortical myelin content is based on the ratio between T1-and T2-weighted MR images in an overlapping sample
(n = 339, age 8–83 years), sampled 0.2 mm from the white matter/gray matter boundary into the gray matter in the superior frontal cortex. Data modified from Grydeland et al.,
2013 and reproduced with permission from Walhovd et al., 2014. (See color plate also)
78
Anterior-â•‰Posterior Gradient and Myelodegeneration Hypothesis

The initial reports of DTI methods (e.g., Basser and Pierpaoli, 1996) noted that this
imaging modality was well suited to investigating age-â•‰related effects in white matter,
due to the inherent sensitivity of DTI to changes in tissue architecture. Research with
DTI has confirmed the trend from the histological studies that age-â•‰related decline is
most prominent in the later-â•‰myelinating white matter pathways connecting associa-
tion cortical regions, relative to white matter in earlier-â•‰myelinating sensory and motor
regions. One of the first DTI investigations of age-â•‰related differences in white matter
was that of Sullivan et al. (2001), who found that age-â•‰related decline in FA was signifi-
cant for the genu of the corpus callosum (connecting left and right prefrontal regions)
but not the splenium of the corpus callosum (connecting left and right temporal and
parietal regions). Contemporaneously, O’Sullivan et al. (2001) reported that age-â•‰related
decline in FA was more pronounced in anterior versus posterior white matter.
This anterior–posterior gradient of age-â•‰related decline in FA has been replicated
in many studies and is arguably the most consistent theme in DTI studies of healthy,
community-â•‰dwelling individuals (Figure 3.4) (Sullivan and Pfefferbaum, 2006, 2007;
Madden et al., 2009a; Salat, 2011; Carmichael and Lockhart, 2012; Madden et al.,
2012; Bennett and Rypma, 2013; Bennett and Madden, 2014; Salat, 2014b). Sullivan
and Pfefferbaum (2006) stated that “Regardless of method used, one of the most
robust findings describing age-â•‰related differences in regional FA has been a frontal
distribution of low FA selective to frontal white matter” (p. 752). Gong et al. (2009),
using principles of graph theory (Sporns, 2014), proposed that the regional efficiency
of white-matter network connections exhibited a shift, with increased adult age, from
occipital and parietal to frontal and temporal regions. In a DTI comparison of younger
adults, healthy older adults, and older individuals with dementia of the Alzheimer
type, Head et al. (2004) emphasized that the anterior–posterior gradient of FA decline
is a specifically age-â•‰related rather than pathology-â•‰related effect. These authors found
that the anterior–â•‰posterior gradient of decline in FA, relative to younger adults, was
not accentuated in the group with dementia as compared to healthy older adults. The
additional declines in FA associated with dementia were observed in parietal, tempo-
ral, and occipital regions, though not in the posterior corpus callosum.
Subsequent research suggests that the unidimensional concept of an anterior–â•‰pos-
terior gradient is not sufficient to capture all the relevant aspects of age-â•‰related decline
in white matter integrity. Salat et al. (2005), in a DTI study of 38 individuals 21–â•‰76
years of age, generally confirmed greater age-â•‰related decline in frontal regions than
temporal and posterior regions. However, other findings, such as a greater decline in
FA in the posterior limb of internal capsule than in particular regions of frontal white
matter, implied that white matter changes were not unique to prefrontal white mat-
ter, and that these changes were regionally selective. In addition, although FA data
for the corpus callosum suggested an anterior–posterior gradient, other age-â•‰related
differences in FA were not consistent with that theme: for example, a significant age-â•‰
related decline of FA in posterior periventricular but not in anterior periventricular
white matter. Rather, white matter vulnerability seems to be associated with particular
fiber bundles, fiber populations, and regional locations; in particular, fibers within
the prefrontal white matter, anterior callosal fibers, and the corticospinal tracts. In a
79
Left hemisphere Right hemisphere

50 50
40 40
30 30
70
20 20
10 10
60 0 0
0 50 100 150 200 0 50 100 150 200
Fractional anisotropy (%)
50 Midline
10 younger 10 older
40
30
20
10
0.00001
p-value 0.001
0.05
0 genu AC PC splenium
0 50 100 150 200
Anterior Posterior
Slice (mm)
Figure 3.4 The anterior-posterior gradient of age group differences in fractional anisotropy

(FA). Profile of mean ± SEM FA in segmentation-defined supratentorial white matter presented
slice-by-slice from frontal to occipital brain regions in 10 younger and 10 older healthy subjects.
The sagittal brain images are grand averages of FA of the 10 younger and 10 older adults; the
corpus callosum is prominent in cross-section. The large profile is taken from the medial 10 mm
of the brain; the small profiles are taken from the remaining white matter lateral to the midline
sample. Note the systematic FA difference in the anterior regions with higher FA in the younger
compared with older subjects. The black mounds on the x-axis indicate the p-value for group
differences for each slice. The gray horizontal lines over the mounds designate p-values, with the
bottom line being p = .05 and the top line p = .00001. Figure modified from Pfefferbaum et al.,
2005 and reproduced with permission from Sullivan and Pfefferbaum (2006).
similar vein, Sullivan and colleagues have pointed out that when different fiber tracts
are compared, though an anterior–posterior gradient of age-related decline is evident,
a superior–inferior gradient is also detectable, with more superiorly located tracts
exhibiting greater vulnerability to age-related decline (Zahr et al., 2009; Sullivan et
al., 2010b).
As one alternative to an anterior–posterior gradient model, Davis et al. (2009)
emphasized the role of the developmental trajectory of myelin in age-related dif-
ferences in DTI measures. According to their myelodegeneration hypothesis, later-
myelinating white matter regions will be the most vulnerable to age-related decline
(Braak and Braak, 1996; Reisberg et al., 2002; Brickman et al., 2012). These regions
tend to be located anteriorly, but also include language-related association regions at
the border of the temporal and parietal lobes (Yakovlev and Lecours, 1967; Paus et al.,
1999; Sowell et al., 2003). In testing the myelodegeneration hypothesis, Davis et al.
took the additional perspective of analyzing the change along the tract as a continuous
80
variable. Comparing 20 younger and 20 older adults, these authors found that, for
tracts traversing the frontal lobe in an anterior–â•‰posterior direction (e.g., uncinate fas-
ciculus and cingulum), the age group difference in FA declined monotonically across
the tract, consistent with an anterior–â•‰posterior gradient. Critically, however, the age
group difference did not change abruptly at the boundary of the frontal lobe, indicat-
ing that the age-â•‰related effect was not associated specifically with the frontal lobe.
Other tracts, such as the inferior longitudinal fasciculus, did not exhibit a monotonic
gradient of age-â•‰related difference but instead showed a higher degree of age-â•‰related
decline at a specific region, at the boundary of the occipital and temporal lobes, which
is a late-â•‰myelinating region. The Davis et al. data also exhibited stronger age-â•‰related
effects for RD than for FA (cf. also Bhagat and Beaulieu, 2004; Sullivan et al., 2006;
Madden et al., 2009b), which in view of the greater sensitivity of RD to myelin con-
tent (Song et al., 2002; Song et al., 2005; Klawiter et al., 2011) supports a critical
role for myelin integrity as the critical feature of age-â•‰related differences in the DTI
measures.
The myelodegeneration hypothesis, however, cannot account for all relevant find-
ings. Westlye et al. (2010), in a large-â•‰scale study of 430 participants 8–â•‰85 years of age,
found that decline in DTI measures during later adulthood did not conform entirely
to a myelodegeneration model: The corticospinal tract showed early white matter
maturation but was among the first tracts to exhibit decreased integrity during aging;
the dorsal cingulum was late to show maturation and also late to show age-â•‰related
decreased integrity. Both effects are inconsistent with a myelodegeneration model.
Salat (2011) has noted that the spatial location of white matter tracts may be limited
as a theoretical framework for associated age-â•‰related effects. Because extensive evi-
dence exists to indicate the vascular pathophysiologic origins to white matter damage
(see the following section, White Matter Hyperintensities), Salat suggested that future
research may profit from considering the anatomy of the vascular tree, rather than the
spatial location or orientation of the white-matter tracts.
Neurobiological Mechanisms
Converging evidence regarding the neurobiological specificity of age-â•‰related effects
can be derived from the joint consideration of different DTI measures. Bennett et al.
(2010) and Burzynska et al. (2010) proposed that when FA, RD, and AD are consid-
ered simultaneously, distinct patterns of age-â•‰related effects were evident in different
brain regions. The most spatially prominent pattern was a decrease in FA for older
adults relative to younger adults, accompanied by an increase in RD. In a subset of
regions (e.g., genu of the corpus callosum), this pattern also included an age-â•‰related
increase in AD. The combination of increased RD and AD implies pathology related
to both axon fibers and the surrounding myelin sheaths. A more specific disruption of
myelin is implied when an increase in RD occurs without an accompanying increase in
AD, as observed for example in the splenium and posterior dorsal cingulum (Bennett
et al., 2010; Burzynska et al., 2010). Alternatively, in other brain regions, such as
the superior corona radiata, age-â•‰related decreased FA and increased RD tend to be
accompanied by a decreased AD, resulting in no net difference in MD. This type of
â•‡ 81
diffusivity pattern has been observed in secondary (Wallerian) degeneration, which is

the degeneration, over time, of axon fibers distal to the point of transsection or injury.
A decrease in FA and AD with no significant change in RD and MD may reflect dis-
rupted macrostructural reorganization of the fibers, such as less coherent fiber align-
ment (Bennett et al., 2010).
Global vs. Tract-â•‰Specific Effects

In the interpretation of DTI data relevant for structural connectivity, an underlying
assumption is that data from different regions of interest or white-matter tracts are
independent. Alternatively, regions or tracts may be grouped together, for the pur-
poses of analysis, on the basis of their presumed function or regional location, as
defined by the investigator. This assumption of regional independence may be help-
ful for interpreting age-â•‰related effects when they are detected. But because different
white-matter tracts may participate in multiple and overlapping neural networks, the
independence assumption may also hinder the understanding of the coordination of
different tracts and the potential role of any general mechanisms that are common
to all the tracts. That is, within a particular dataset, the DTI data are likely to exhibit
regional covariation that should be taken into consideration.
To date, evidence for both global and tract-â•‰specific DTI effects has emerged. One
of the first studies to examine this issue is that of Penke et al. (2010), who conducted
principal component analysis (PCA) on each of several DTI measures (FA, AD, RD,
MD) from eight white matter tracts, in a sample of 132 older adults (71–â•‰73 years of
age). These authors found that a single principal component explained approximately
45% of the variance in each measure, and that many or most of the individual tracts
loaded highly on each principal component, suggesting a global factor of white matter
integrity. In contrast, other investigators, using independent component analysis (Li et
al., 2012) and hierarchical clustering (Wahl et al., 2010), have reported that DTI data
segregate into identifiable groups of white-matter pathways.
None of these previous investigations, however, was designed to specifically
address age-â•‰related differences in their DTI data. Lovden et al. (2013), in a structural
equation modeling of DTI data from 260 older adults (60–â•‰87 years of age), compared
two models: one positing a general factor (i.e., similar cross–â•‰tract correlations) and
the other positing tract-â•‰specific effects (i.e., each tract-â•‰tract correlation as a separate
factor). Although a global factor was evident in the DTI data, the model that explained
the most variance contained factors for specific tracts. Moreover, age-â•‰related differ-
ences in inter-â•‰tract correlations occurred in only a subset of regions, further indicating
that age-â•‰related effects were not homogeneous across all the tracts. Johnson et al.
(2015) conducted a PCA analysis of the DTI measures FA, MD, AD, and RD, for
a wider age range comprising 52 younger adults (18–â•‰28 years of age) and 64 older
adults (60–â•‰85 years of age). Across eight anatomically defined white-matter tracts,
the authors observed two components for FA, the first with high loadings from the
superior longitudinal fasciculi and corticospinal tracts, and the second with high load-
ings from the optic radiations. In contrast, variation in each of the diffusivity measures
yielded a single-â•‰component solution in each case, with high loadings from most or all
82
tracts. Both the FA and diffusivity components exhibited age-â•‰related decline consis-
tent with the previous literature, suggesting that both global and tract-â•‰specific varia-
tion in white matter were associated with age-â•‰related influences.
Relation Between DTI Measures and Behavioral Measures
As we noted previously (Introduction section), myelination supports the speed and

efficiency of axonal conduction (Waxman, 1980; Seidl, 2014). Thus, in theory, the
age-â•‰related decline in the DTI measures, even in the absence of disease, contributes to
deficits in structural connectivity consistent with a disconnection model of cognitive
aging. A critical question is the functional significance of these deficits. A general
theme of cognitive aging is that fluid (speed-â•‰based) cognitive abilities are more vul-
nerable to age-â•‰related decline than crystallized (knowledge-â•‰based) abilities (Craik and
Salthouse, 2008). If later adulthood is associated with some degree of disconnection
among the neural networks of cognition, are all networks affected equally? Are the
age-â•‰related differences in the DTI measures simply correlated with individual differ-
ences in behavioral measures, or can a causal relation between variables be identified?
What type of behavioral measure is most sensitive to age-â•‰related variation in DTI
measures? Although the literature on age-â•‰related differences in the relation between
DTI measures and behavioral measures of cognition is expanding rapidly, and sub-
stantial information is available, the answers to these questions are not yet definitive.
Network Specificity
The earliest investigations of age-â•‰related differences in DTI measures (O’Sullivan et
al., 2001; Sullivan et al., 2001) also reported correlations between the DTI measures
and some behavioral performance measures, in the direction of declining white mat-
ter integrity associated with declining performance. Since these initial reports, the
literature on the functional correlates of DTI measures in the context of aging has
continued to increase rapidly, and several reviews on this topic now exist (Sullivan and
Pfefferbaum, 2006, 2007; Madden et al., 2009a; Salat, 2011; Carmichael and Lockhart,
2012; Madden et al., 2012; Bennett and Rypma, 2013; Bennett and Madden, 2014;
Salat, 2014b). Madden et al. (2012) limited their review to studies of healthy older
adults that provided both a behavioral measure of cognition and regionally specific
measures of FA from DTI. Across 29 published studies, these authors noted that the
DTI–cognition relation was detectable across several cognitive domains but was more
prominent for measures of fluid cognition, especially perceptual speed and execu-
tive function, relative to memory. However, there is also substantial diversity in both
the types of tasks and the regional location of the correlated DTI measures. In sev-
eral reports correlating DTI measures with fluid cognitive measures, the white-matter
tracts comprising frontoparietal networks, especially the genu and splenium of the cor-
pus callosum, and the superior longitudinal fasciculus yielded significant age-â•‰related
effects (Kennedy and Raz, 2009; Madden et al., 2009b; Gold et al., 2010; Sullivan et
al., 2010a; Kerchner et al., 2012). Fischer et al. (2014), using graph theory analyses of
â•‡ 83
DTI data (Sporns, 2014), proposed that once network connectivity decreases below a
threshold (at approximately 75 years of age), an association between network proper-
ties and general intelligence is evident, such that higher intelligence is associated with
more modular and efficient networks.
Causal Role of DTI Variables

Though research to date suggests an association between a decline in white matter
integrity and age-â•‰related differences in measures of fluid cognition, the issue remains
whether white matter integrity has a causal role in cognitive aging. Previous analyses
have frequently focused on the overall variance in the behavioral outcome measure
associated with the DTI measure, rather than the variance associated specifically with
age. That is, because an age–â•‰cognition correlation is significant, and a DTI–cogni-
tion correlation is significant, that does not necessarily mean that the individual dif-
ferences in white matter that are reflected in the DTI data are responsible for the
age-â•‰related differences in the cognitive measure. Similarly, researchers often report
correlations for younger and older adult age groups separately, without testing the
differences between the correlations directly. Salthouse (2011) pointed out that the
relations among neuroanatomical variables such as DTI, cognition, and aging can be
modeled with different arrangements of the causal relations among the variables, and
that alternative models should be tested empirically. With few tests available for mod-
els that would exclude alternative interpretations, Salthouse (2011) concluded that
“the evidence for a causal role of DTI on age-â•‰cognition relations must be considered
relatively weak at the current time” (p. 773).
Several investigators have addressed the causal influence of white matter integrity
by conducting mediation analyses (Preacher and Hayes, 2008; Hayes, 2009), which
estimate the degree to which a specified variable (e.g., DTI measure) has a causal
(mediating) influence on the relation between age and cognitive measures. White mat-
ter integrity in this context does appear to be a significant mediator of age-â•‰related
decline in fluid cognitive abilities, consistent with the disconnection model (Madden
et al., 2009b; Perry et al., 2009; Gold et al., 2010; Salami et al., 2012; Lu et al., 2013).
Similarly, research using structural equation modeling and related techniques support
models in which white matter integrity has a causal influence on the age–cognition
relation (Charlton et al., 2008; Penke et al., 2010; Burgmans et al., 2011; Voineskos
et al., 2012; Lovden et al., 2014). Kievit et al. (2014) combined gray matter volume
measures of prefrontal regions with DTI measures of white-matter tracts connecting
those regions in a large, population-â•‰based sample of 567 individuals with both behav-
ioral and imaging data. A structural equation modeling analysis of two related mea-
sures of executive function—fluid intelligence and multitasking ability—indicated
that the pattern of mediation differed for the two behavioral domains. Both regional
gray matter volume (Brodmann area 10) and white matter integrity (genu of the cor-
pus callosum) mediated age-â•‰related decline in fluid intelligence, whereas integrity
of the anterior thalamic radiations (frontostriatal tract) mediated multitasking ability.
These previous studies are consistent with the general theme of the disconnec-
tion model of cognitive aging, although as Kievit et al. (2014) have emphasized, the
84
influence of disconnection may vary across behavioral domains and may interact with
changes in regional volume. In future research on mediation effects it will also be
valuable to include a comparison to alternative models in which age has concurrent
but independent effects on the DTI and cognitive variables (Salthouse, 2011; Madden
et al., 2014). That is, if both fluid ability and DTI measures vary with age, alterna-
tive statistical models. Age may be considered a mediator in the sense of being a
proxy variable for accumulated challenges to the CNS, over time, which lead gradu-
ally to decline in both white matter integrity and cognitive performance. Johnson et
al. (2015), for example, conducted mediation analyses of DTI data and cognition for
groups of younger (18–â•‰28 years) and older (60–â•‰85 years) adults, comparing alterna-
tive models in which: a) DTI measures were mediators of the age-â•‰related decline in
perceptual-â•‰motor speed; and b) age was a mediator of the relation between the DTI
variables and speed. The best-â•‰fitting model was the one in which age was the mediator
of the DTI–â•‰speed relation.
Another important issue for future research with mediational models is the dif-
ference between cross-â•‰sectional and longitudinal measurements of change. Cross-â•‰
sectional estimates of age-â•‰related effects may be statistically biased and may not
necessarily reflect individual-â•‰level correlated rates of change (Hofer and Sliwinski,
2001; but cf. Salthouse and Nesselroade, 2002; Ronnlund et al., 2005; Nyberg et al.,
2010; Lindenberger et al., 2011). Longitudinal designs have the advantage of providing
a more direct estimate of change over time. Lovden et al. (2014) measured changes in
DTI variables over a 2.3-year interval in a sample of individuals 81–â•‰103 years of age
and found that estimates of annual longitudinal change tended to be larger than the
cross-â•‰sectional age effect. In some contexts, however, the prior test experience associ-
ated with longitudinal testing may mask actual age-â•‰related decline (Salthouse, 2009).
Variability vs. Central Tendency of Behavioral Measures

Virtually all of the previous research on the relation between measures of DTI and
cognition have defined the cognitive outcome variable in terms of an estimate of cen-
tral tendency, typically each participant’s mean or median reaction time (or accuracy)
within one or more task conditions. Behavioral studies of cognitive aging suggest
that measures of the variability of performance, such as the standard deviation or
coefficient of variation, can provide complementary information that reveals aspects
of age-â•‰related differences not always detected in analyses of central tendency (but
cf. Salthouse and Berish, 2005; MacDonald et al., 2006b; MacDonald et al., 2006a).
Measures of performance variability have also exhibited relations to structural brain
measures (Bunce et al., 2007; Walhovd and Fjell, 2007) and task-â•‰related functional
activation (Garrett et al., 2011).
For example, Fjell et al. (2011), in a study of 270 individuals 20–â•‰83 years of
age, conducted correlational analyses between several DTI measures (FA, MD, AD,
RD) from white-matter tracts and performance variability from a reaction-time task
(Eriksen flanker). These authors found that reaction-time variability was related to all
DTI measures in most white-matter tracts independently of mean reaction time, age,
and gender. Critically, the strength of these relations increased with age, suggesting
â•‡ 85
that age-â•‰related decline in microstructural integrity of white matter modulated the

association between reaction time and the DTI measures. This pattern was clearest in
the less-â•‰demanding (congruent) task condition (Figure 3.5). Grydeland et al. (2013)
further demonstrated that performance variability was related to an MRI measure sen-
sitive to intracortical myelin (the ratio of T1-â•‰weighted and T2-â•‰weighted signal intensi-
ties) and that this relation was more pronounced for individuals in their late 50s and
beyond. Thus, measures of performance variability should be considered when assess-
ing the relation between measures of white matter integrity and cognition.
Relation Between DTI Measures and Functional Measures

of Brain Activity
Structural–â•‰Functional Relations
As the brain’s structural scaffolding, white matter pathways serve to connect widely
dispersed brain regions whose coordinated efforts form the basis of brain function. In
view of the extensive literature on functional brain imaging, it will naturally be impor-
tant to determine the relation between measures of structural connectivity and func-
tional brain activity (Ramnani et al., 2004). To determine the behavioral significance
of this interaction, recent studies have combined structural information from DTI and
functional information from functional magnetic resonance imaging (fMRI), though
relatively few studies have addressed the additional dimension of age. In one of the
earliest multimodal imaging studies to collect both DTI and fMRI data, Werring et al.
(1999) demonstrated the feasibility of this approach by overlaying the FA maps from
DTI and blood oxygen level–â•‰dependent (BOLD) activation maps from fMRI, with-
out spatial registration. These authors found that FA was lower in activated cortical
regions than in white matter, confirming that the BOLD signal was occurring within
the relatively isotropic cortical gray matter. Conturo et al. (1999) demonstrated that
fMRI could be combined with DTI tractography by using fMRI-â•‰activated regions as
endpoints for tractography of the geniculo-â•‰calcarine white-matter tracts connecting
the visual cortex to the thalamus.
More recently, investigators have been concerned with the functional connectivity
among different brain regions, in the time course of the BOLD signal, rather than the
mean level of activation. Functional connectivity can be assessed from resting-â•‰state
fMRI data, without an assigned behavioral task, as well as from task-â•‰related data (see
Chapter 4, this volume). One of the main themes to emerge from the combination of
DTI and fMRI data is that microstructural integrity of white-matter constrains func-
tional connectivity. Using seed regions obtained from resting-â•‰state functional connec-
tivity maps in combination with DTI-â•‰based tractography, Greicius et al. (2009), for
example, revealed white matter tracts that connect the functional hubs of the default
mode network, specifically the cingulum bundle between the posterior cingulate cor-
tex (PCC) and the medial prefrontal cortex, and the descending cingulum bundles
between the PCC and the medial temporal lobe. Van den Heuvel et al. (2008) dem-
onstrated a positive correlation between FA in the default mode network (within the
cingulum bundle) and resting-â•‰state functional connectivity between the PCC and the
86
A
R L
P
FA 0.6
0.5
FA
0.4
25 50 75 100 125
Milliseconds
AD 1350
1300
1250
AD
1200
1150
1100
25 50 75 100 125
Milliseconds
RD
750
700
650
RD
600
550
500
450
25 50 75 100 125
Milliseconds
MD 900
850
MD
800
750
700
25 50 75 100 125
Milliseconds
Young (< 52 years) Elderly (≥ 52 years)
Figure 3.5 Age-related differences in the relation between reaction time standard deviation

(sdRT) and measures of microstructural integrity. Voxels showing a significant age x sdRT inter-
action on diffusion characteristics are displayed (age, sex, mRT, and sdRT were used as covari-
ates). The effects are corrected for multiple comparisons across space by threshold-free cluster
enhancement at p < 0.05. The results are smoothed to ease visualization of effects, and displayed
on top of the WM skeleton (red on green skeleton, FA; copper on green skeleton, AD; blue on
green skeleton, RD; green on red skeleton, MD). Right, Scatterplots illustrating the relationship
between sdRT in milliseconds (x-axis) and diffusion characteristics (y-axis) across all voxels
showing a significant relationship between sdRT and diffusion. As can be seen, weak relation-
ships exist in the younger half of the sample (age < 52 years), while stronger relationships are seen
in the older half (age ≥ 52 years). Reaction time data drawn from the less-demanding (congruent)
condition of the Eriksen flanker task. The more attentionally demanding (incongruent) condition
yielded a similar, though spatially less extensive, pattern. Figure reproduced with permission from
Fjell et al. (2011). (See color plate also)
â•‡ 87
medial prefrontal cortex. Van den Heuvel et al. (2009) employed an analogous method
to other functional networks including the primary visual, primary motor, and fronto-
parietal attention networks. Damoiseaux and Greicius (2009) reviewed eight studies
that compared intrinsic functional and structural connectivity and concluded that DTI
measures reflecting higher integrity of white matter were associated with enhanced
functional connectivity.
Age-â•‰Related Differences in Structural–â•‰Functional Relations

Bennett and Rypma (2013) surveyed 38 multimodal imaging studies combining DTI
and fMRI and concluded that the age of the participants, as well as the spatial prox-
imity of the target regions, were significant moderators of the relation between the
DTI and fMRI measures. These authors proposed that, overall, increased white matter
integrity was predominantly associated with increased neural activity in younger adults
(i.e., positive DTI–â•‰fMRI relationships), but decreased neural activity in older adults
(i.e., negative DTI–â•‰fMRI relationships). With regard to spatial proximity, Bennett and
Rypma found that, for younger adults, white matter integrity was positively related to
neural activity when there was spatial overlap between brain regions from which the
neural measures were extracted (i.e., high spatial proximity), but negative relation-
ships were observed between non-â•‰adjacent regions (i.e., low spatial proximity). For
older adults, data from non-â•‰adjacent regions exhibited a mixed pattern DTI-â•‰fMRI rela-
tionship, and no data for high spatial proximity were available. Bennett and Rypma
stated that “To date, no studies have examined relationships between neural measures
from adjacent brain regions in older adults” (p. 1207). Thus, whereas the proximity
of the neural measures accounted for the direction of DTI–â•‰fMRI relationships for
younger adults (i.e., positive DTI–â•‰fMRI relationships between adjacent brain regions
and negative DTI–â•‰fMRI relationships between non-â•‰adjacent brain regions), it is not
yet clear whether the relation between DTI and fMRI measures varies in a similar
manner, in relation to spatial proximity, for older adults.
One recent study (Salami et al., 2014) has included microstructural integrity and
resting-â•‰state functional connectivity measures from spatially proximate regions (for-
nix and hippocampus). These authors proposed that age-â•‰related decline in the integ-
rity of the fornix was associated with decreased resting-â•‰state functional connectivity
between the hippocampus and other cortical regions, which in turn led to deficits in
active memory encoding.
The concept of compensation has been discussed frequently as a possible expla-
nation for age-â•‰related differences in the relation between structural and functional
brain measures. In particular, because older adults often exhibit fMRI activation that
is either greater in magnitude than the corresponding activation for younger adults, or
located outside of the task-â•‰relevant brain regions, it is possible that this increased acti-
vation may represent a compensatory increase in brain activity that maintains older
adults’ behavioral performance (Cabeza et al., 2002; Reuter-â•‰Lorenz and Cappell,
2008; Park and Reuter-â•‰Lorenz, 2009). Several reports have concluded that compensa-
tion is a potential mechanism of age-â•‰related differences in the relation between DTI
and fMRI measures. Davis et al. (2012), for example, found that older adults were
88
more likely than younger adults to exhibit bilateral, prefrontal fMRI activation during
a word matching task, that the older adults with a higher degree of bilateral activation
also performed better in the task, and that white matter integrity as measured from
DTI was a mediator of the activation–â•‰performance relation for older adults. Critically,
older adults with higher white matter integrity (higher FA and lower RD) exhibited
a higher degree of functional connectivity of activated regions and better behavioral
performance, suggesting the contribution of a compensatory mechanism. Similarly,
Chen et al. (2009) found that increasing resting-â•‰state connectivity was associated with
increasing speed of choice RT for older adults but not for younger adults, and that
increasing FA for a white-matter tract connecting RT-â•‰relevant areas (genu) was cor-
related positively with older adults’ functional connectivity.
The concept of compensation, however, may not successfully account for all
reported patterns of age-â•‰related difference in the combined DTI-â•‰fMRI measures.
Colcombe et al. (2005) found that older adults who exhibited more bilateral activa-
tion tended to perform worse on the behavioral task relative to older adults with less
bilateral activation, contrary to what would be expected on the basis of compensation.
In measures of task-â•‰related fMRI activation, Madden et al. (2007; 2010) found that
age-â•‰related differences in the DTI measures were largely independent of the age-â•‰
related effects in the fMRI measures. Bennett and Rypma (2013) noted that compen-
sation does not characterize the positive relation between DTI and fMRI measures
that is frequently observed for younger adults. In addition, whereas Davis et al. (2012)
and Chen et al. (2009) interpreted a positive DTI-â•‰fMRI relation for older adults as
compensatory, Bennett and Rypma assumed that compensation would be expressed
as a negative DTI-â•‰fMRI relation. Consistent with this latter approach, Daselaar et
al. (2013) have recently interpreted a negative DTI-â•‰fMRI relation for older adults as
compensation, expressed as the less wiring, more firing hypothesis (cf. also Chadick
et al., 2014). Daselaar et al. proposed that it may be necessary to distinguish between
attempted and successful compensation to more completely account for the relations
among age-â•‰related effects in DTI, fMRI, and behavioral performance data. Finally,
brain maintenance, in the sense of the relative lack of brain pathology, rather than
compensation, may be a primary determinant of successful cognitive aging (Nyberg
et al., 2012). Thus, although disparate results have occurred to date, the development
of more explicit models of the causal relations among DTI, fMRI, and behavioral
measures, and testing of alternative models (Salthouse, 2011), should help clarify the
mechanisms by which structural connectivity constrains age-â•‰related differences in the
relation between fMRI measures and cognitive performance.
White Matter Hyperintensities
Diffusion tensor imaging research in aging has focused primarily on the properties
of normal-â•‰appearing white matter. Damaged white matter is typically identified not
through DTI but through T2-â•‰weighted, fluid attenuation inversion recovery (FLAIR)
images, in which damaged white matter appears brighter than healthy tissue. These
WMHs, also referred to as leukoaraioses or white matter lesions, appear most often in
the periventricular regions surrounding the anterior and posterior horns of the lateral
89
ventricles. Hyperintensities occurring outside of the periventricular regions (i.e., in

deep white matter) are highly correlated with periventricular WMHs but are ana-
tomically separable from them (DeCarli et al., 2005; Vannorsdall et al., 2009; Valdes
Hernandez et al., 2014). WMHs have been found to increase in prevalence as a func-
tion of increasing adult age (de Leeuw et al., 2001; Yoshita et al., 2005; Vernooij et
al., 2007). The age-related increase in WMHs occurs even in otherwise healthy indi-
viduals, and the histopathology of WMHs is variable (Figure 3.6). In general, WMHs
appear to represent ischemic damage due to vascular dysfunction (Salat, 2014b, a), as
increasing WMH volume is associated with increasing small-vessel cerebrovascular
disease and levels of hypertension (Longstreth et al., 2005; Raz et al., 2007; Young et
al., 2008; Valdes Hernandez et al., 2014).
An extensive literature has examined the relation between WMHs and cogni-
tive aging (Gunning-Dixon and Raz, 2000; Gunning-Dixon et al., 2009; Salthouse,
(A)
L L L
(B)
L L L
1 19
Figure 3.6 White matter lesion (WML) volume. Panel A: White matter lesions for individual
adults 20, 48, and 65 years of age, in T2-weighted FLAIR images. Participants were healthy,
community-dwelling individuals without any sign of cognitive impairment on neuropsycho-
logical testing or history of cardiovascular disease (other than hypertension). Lesions, as identi-
fied from a semi-automated program separating lesions from normal white matter, appear in
red. Panel B: Voxelwise lesion maps for 23 younger adults (19–39 years of age), 19 middle-
aged adults (40–59 years of age), and 16 older adults (60–79 years of age). Color scale repre-
sents the number of individuals within each group exhibiting a lesion, per voxel, overlaid on a
T1-weighted template. Authors’ data. (See color plate also)
90
2011). The most consistent finding is a stronger relation between increasing WMHs
and decreasing performance on speed-dependent measures of perceptual-motor and
executive function, relative to measures of memory or more global cognitive function.
DeCarli et al. (1995), for example, reported that, in a sample of 51 healthy adults
19–91 years of age, increased age was associated with increasing WMH volume, and
that the most pronounced relation of WMH volume (independent of age) was to tests
involving executive function (trail-making and immediate visual memory). DeCarli
et al. suggested that this pattern may be due to a differential vulnerability of fron-
tal lobe functioning to disruption from WMHs. Other investigators have confirmed
that WMHs in older adults tend to be differentially related to decreased performance
on speed-based tests of executive function (de Groot et al., 2000; Prins et al., 2005;
Rabbitt et al., 2007; Birdsill et al., 2014; Lockhart et al., 2014).
Whereas previous evidence supports a relation between WMHs and cognition
(particularly executive function) in older adults, the nature of that relation is not yet
defined. As Salthouse (2011) has noted, WMHs may have a direct influence on the
age–cognition relation or, alternatively, both WMHs and cognitive function may vary
with age, in parallel, without a mediating influence of WMHs on the age–cognition
relation. Gunning-Dixon and Raz (2003) adopted a model-based approach by using path
analytic techniques to compare different models of the relations among age, regionally
defined WMH volume, and cognitive function for 139 participants 50–81 years of age.
These authors included both WMH volume and regional gray matter volume in the fron-
tal and temporal lobes as predictors of executive function (perseverative errors on the
Wisconsin Card Sorting Test) and working memory. Gunning-Dixon and Raz found that
increasing age was indirectly related to declining executive function, as a result of the
mediating influences from both WMH volume and regional gray matter volume within
the frontal lobe. Further, neither WMH nor gray matter volume in the temporal lobe con-
tributed significantly to the age–executive function path, and the volumetric measures
were not mediators of the age-related decline in working memory performance. Parks et
al. (2011) also conducted a mediational model of the relations among WMHs and cogni-
tion in older adults and proposed that the effects of executive function and memory are
interactive. Specifically, their model suggested an indirect effect of WMHs on episodic
memory, one mediated by executive function and hippocampal volumes. The Parks et al.
model, however, differed from the Gunning-Dixon and Raz approach by not including
age in the model, and the Parks et al. sample was clinically diverse, comprising individu-
als with cognitive impairment as well as healthy individuals.
The metric for WMHs in the majority of previous studies has been relatively
global, typically either WMH tissue volume or the intensity of the MRI signal within
a WMH region, both of which are derived from T2-weighted FLAIR imaging. In
addition, the identification of WMH on FLAIR imaging is based on a categorization
of normal-appearing and abnormal white matter, which may be a dichotomization of
a more continuously distributed dimension of white matter integrity. Further, the rela-
tion between DTI measures of normal-appearing white matter and FLAIR measures
of WMH are not yet understood. From analyses of a population-based sample of 832
older adults, Vernooij et al. (2008) proposed that age-related decline in DTI measures
of white matter were nearly entirely attributable to white matter atrophy or WMH
formation. Maillard et al. (2013), in contrast, detected independent effects of WMHs
â•‡ 91
and DTI data. These authors compared signal intensity from FLAIR and FA from
DTI, in a longitudinal analysis of a clinically diverse (i.e., both cognitively healthy
and impaired) sample of 119 older adults, across 3.75 years. These authors estimated
the likelihood, for each voxel, of a conversion from normal white matter to WMH
across the 3.75-year interval. A model including both FA and FLAIR signal intensity
values at baseline indicated that both lower FA and higher FLAIR signal intensity
baseline values were independently predictive of conversion to WMH. The FA data
were more sensitive, but had lower specificity and accuracy as compared with FLAIR
data. Although FLAIR was the stronger predictor, variation in the diffusional coher-
ence (FA) of normal white matter, assessed from DTI, was informative regarding the
development of WMHs.
In other analyses of this longitudinal dataset, again combining FLAIR and DTI
data, Maillard et al. (2014) proposed that WMHs are surrounded by what they term a
penumbra, a region of white matter that is normal at baseline but differentially vulner-
able to deterioration and thus conversion to WMH. Maillard et al. categorized WMHs
and their penumbra as either constant over time (stagnant), growing over time, or
evident only at follow-â•‰up (incident). The FLAIR and DTI data exhibited different
patterns of longitudinal change: The FLAIR signal was more dichotomous and distin-
guished relatively healthy and damaged tissue, but the change over time was clearest
for the incident WMHs. The DTI data, in contrast, revealed a more fine-â•‰grained order-
ing of average FA across the WMH and penumbra categories that corresponded to the
degree of white matter impairment. The rate of longitudinal decline in FA, however,
was comparable for all the tissue types. Maillard et al. proposed that WMHs and their
penumbra represent a continuous spectrum of white matter injury, and that FLAIR and
DTI yield different sources of information regarding this spectrum. Although Maillard
et al. (2014) did not analyze the relation of the WMHs to longitudinal changes in cog-
nitive performance, other analyses of these data suggest that WMHs that are growing
over time, rather than those that are either stagnant or incident, are differentially asso-
ciated with declining cognitive performance in later adulthood (Maillard et al., 2012).
Iron Accumulation and Magnetic Susceptibility
Recent investigations suggest that the developmental trajectory of iron deposition in

the brain is related to white matter integrity and, more generally, to a disconnection
model of cognitive aging (Bartzokis, 2004b; Bartzokis, 2004a; Zecca et al., 2004;
Bartzokis, 2011; Daugherty and Raz, 2013). Iron is an essential cofactor for many
proteins involved in the normal function of neural tissue. In the brain, the most com-
mon cell type to stain for iron under normal conditions is the oligodendrocyte, which
has a critical role in the production of myelin (Todorich et al., 2009). The excessive
accumulation of iron in specific regions of the brain during later adulthood may be an
indicator of increased oxidative stress, expressing the decreased ability of the CNS to
maintain the homeostatic balance of intracellular iron. Early histochemical analyses of
postâ•‰mortem brain tissue demonstrated that intracellular iron was more prevalent in the
extrapyramidal system (e.g., putamen, caudate, globus pallidus, substantia nigra, red
nuclei, and dentate nuclei) than in other cortical regions or in white matter (Hallgren
92
and Sourander, 1958). The deposition of iron in these deep gray-matter regions, par-
ticularly the putamen, increases with adult age even in the absence of disease, and
different profiles of increased iron accumulation are associated with several neurode-
generative diseases, particularly Parkinson’s disease and Alzheimer’s disease (Zecca
et al., 2004).
Structural MRI studies have confirmed the age-related increase in iron deposition
in vivo. In a meta-analysis of 20 published neuroimaging studies of age-related effects
associated with brain iron in healthy individuals, Daugherty and Raz (2013) reported
a reliable age-related increase in iron deposition in deep gray-matter regions, par-
ticularly the putamen. Bartzokis (2004a; 2004b; 2011) has proposed that age-related
iron accumulation has a critical role in age-related decline in structural connectivity.
Whereas oligodendrocytes continue to differentiate throughout adulthood and contrib-
ute to remyelination and myelin repair processes, later-myelinated axons have thinner
sheaths that are more vulnerable to breakdown. The proliferation of oligodendrocytes
with healthy aging, while assisting with remyelination and repair, also contributes to
the additional iron levels that challenge the homeostatic balance of intracellular iron,
thus contributing to an age-related disconnection of neural networks.
Studies to date have typically estimated iron tissue content on the basis of proper-
ties of the T2-weighted MRI signal that are potentially influenced by background
field inhomogeneity unrelated to iron content (Haacke et al., 2005). Quantitative
susceptibility mapping (QSM) is a new methodology that measures tissue magnetic
susceptibility from the multiecho gradient echo MRI signal phase (Li et al., 2011;
Langkammer et al., 2012; Liu et al., 2015). Although the iron content in white matter
is typically low, current evidence suggests that QSM is informative regarding white
matter as well as gray matter, because the susceptibility contrast is determined primar-
ily by iron content (i.e., paramagnetic) for gray matter but by myelin integrity (i.e.,
diamagnetic) for white matter (Liu et al., 2011). Bilgic et al. (2012) reported that
average regional differences between younger and older adults in iron concentration
estimated from QSM corresponded to postmortem values (Hallgren and Sourander,
1958). Li et al. (2014) obtained QSM data from 191 participants, 1–83 years of age,
and found that magnetic susceptibility values of gray matter and white matter were
associated with different developmental trajectories. Magnetic susceptibility in gray
matter exhibited an exponential increase with increasing age, suggesting a gradual
accumulation of iron content, whereas white matter susceptibility was biphasic (qua-
dratic), consistent with the initial development and subsequent decline in myelin
integrity over the lifespan (Bartzokis, 2004b).
As noted previously, intracellular iron accumulation is associated with several neu-
rodegenerative diseases (Zecca et al., 2004; Bartzokis, 2011), but little information is
available currently on the role of iron in normal aging. Several studies have investigated
an iron–behavior relation, either within a sample of healthy older adults or across a wide
age range. Sullivan et al. (2009) estimated iron content of deep gray-matter structures
from the transverse relaxation rate (R2) of the T2-weighted signal across different sig-
nal strengths (1.5T and 3T). These authors found that, in a sample of 10 older adults,
increased iron estimates in the caudate, globus pallidus, and putamen were associated
with slower motor performance. Penke et al. (2012) used a study-specific method to esti-
mate iron, combining different properties of T1-weighted, T2*-weighted, and FLAIR
â•‡ 93
imaging to distinguish iron from calcium. This study sample comprised 143 individuals
71–â•‰72 years of age, from a longitudinal study with general IQ measures available from
11 years of age. Increasing iron was associated with lower levels of general IQ, even
controlling for IQ at age 11. This relation held for a general IQ factor derived from a
PCA, but not for more specific tests of information processing speed, memory, or verbal
fluency. Li et al. (2015) demonstrated that iron concentration in the globus pallidus,
estimated by QSM, was correlated significantly with decreased manual dexterity, inde-
pendently of age, in a sample of 132 individuals 40–â•‰83 years of age.
These previous studies, however, do not specifically relate the age-â•‰related accu-
mulation of iron to age-â•‰related differences in the behavioral measures. Rodrigue et al.
(2013) focused specifically on the convergence of the age-â•‰related brain and behav-
ioral effects in a structural equation modeling analysis, estimating iron concentration
from a property of the T2-â•‰weighted signal (T2* relaxation time). The Rodrigue et al.
findings suggest that age-â•‰related increases in iron concentration lead to decreased
memory performance by contributing to a decrease in hippocampal volume, presum-
ably as a result of increased oxidative stress associated with the age-â•‰related increases
in free iron concentration. This specific relation among the variables held only for the
hippocampus and was not evident for alternative models based on iron effects in either
the caudate or visual cortex.
Conclusion
Diffusion tensor imaging measures provide an informative, though indirect, character-

ization of the microstructural integrity of the white-matter tracts connecting the neu-
ral networks that are ultimately responsible for perceptual and cognitive functioning.
Substantial findings have accumulated to suggest that individual differences in DTI
measures contribute to age-â•‰related variance in cognitive functioning. In particular, age-â•‰
related differences in fluid cognitive abilities appear to be related to the integrity of
frontoparietal pathways. A disconnection model is useful as a theoretical framework for
integrating the age-â•‰related decline in both DTI measures and cognitive performance.
A valuable direction for future research will be the comparison of different models of
the aging–â•‰cognition relation. To date, investigations of microstructural integrity have
been conducted largely independently of functional neuroimaging, and the relation
between structural and functional measures is not clear. Multimodal imaging, with the
addition of laboratory tasks assessing specific components of cognitive function, will
yield new information regarding the aspects of structural and functional connectivity
most relevant for age-â•‰related differences in cognition.
Acknowledgments
Preparation of this chapter was supported by NIH research grant R01 AG039684.
We are grateful to David Hoagey, Micah Johnson, Sally Cocjin, Maria Boylan,
Catherine Tallman, Max Horowitz, Jesse Honig, Lauren Packard, Rachel Siciliano,
Zach Monge, and Kristin Sundy for their assistance.
94
References
Assaf Y, Cohen Y (2014) Inferring microstructural information of white matter from diffu-
sion MRI. In: Diffusion MRI: From quantitative measurement to in-â•‰vivo neuroanat-
omy (2nd edition) (Johansen-â•‰Berg H, Behrens TEJ, eds), pp 185–â•‰208. San Diego, CA:
Elsevier.
Bartolomeo P, Thiebaut de Schotten M, Doricchi F (2007) Left unilateral neglect as a discon-
nection syndrome. Cereb Cortex 17:2479–â•‰2490.
Bartzokis G (2004a) Age-â•‰related myelin breakdown: A developmental model of cognitive
decline and Alzheimer’s disease. Neurobiol Aging 25:5–â•‰18.
Bartzokis G (2004b) Quadratic trajectories of brain myelin content: Unifying construct for
neuropsychiatric disorders. Neurobiol Aging 25:49–â•‰62.
Bartzokis G (2011) Alzheimer’s disease as homeostatic responses to age-â•‰related myelin break-
down. Neurobiol Aging 32:1341–â•‰1371.
Basser PJ, Pierpaoli C (1996) Microstructural and physiological features of tissues elucidated
by quantitative-â•‰diffusion-â•‰tensor MRI. J Magn Reson B 111:209–â•‰219.
Basser PJ, Pierpaoli C (1998) A simplified method to measure the diffusion tensor from seven
MR images. Magn Reson Med 39:928–â•‰934.
Basser PJ, Pajevic S, Pierpaoli C, Duda J, Aldroubi A (2000) In vivo fiber tractography using
DT-â•‰MRI data. Magn Reson Med 44:625–â•‰632.
Beaulieu C (2002) The basis of anisotropic water diffusion in the nervous system -â•‰a technical
review. NMR Biomed 15:435–â•‰455.
Beaulieu C (2014) The biological basis of diffusion anisotropy. In: Diffusion MRI: From quan-
titative measurement to in-â•‰vivo neuroanatomy (2nd edition) (Johansen-â•‰Berg H, Behrens
TEJ, eds), pp 155–â•‰183. San Diego, CA: Elsevier.
Behrens TEJ, Sotiropoulos SN, Jbabdi S (2014) MR diffusion tractography. In: Diffusion
MRI: From quantitative measurement to in-â•‰vivo neuroanatomy (2nd edition) (Johansen-â•‰
Berg H, Behrens TEJ, eds), pp 429–â•‰451. San Diego, CA: Elsevier.
Bennett IJ, Rypma B (2013) Advances in functional neuroanatomy: A review of combined
DTI and fMRI studies in healthy younger and older adults. Neurosci Biobehav Rev
37:1201–â•‰1210.
Bennett IJ, Madden DJ (2014) Disconnected aging: Cerebral white matter integrity and age-â•‰
related differences in cognition. Neuroscience 276:187–â•‰205.
Bennett IJ, Madden DJ, Vaidya CJ, Howard DV, Howard JH, Jr. (2010) Age-â•‰related differences
in multiple measures of white matter integrity: A diffusion tensor imaging study of healthy
aging. Hum Brain Mapp 31:378–â•‰390.
Bhagat YA, Beaulieu C (2004) Diffusion anisotropy in subcortical white matter and cortical
gray matter: Changes with aging and the role of CSF-â•‰suppression. J Magn Reson Imaging
20:216–â•‰227.
Bilgic B, Pfefferbaum A, Rohlfing T, Sullivan EV, Adalsteinsson E (2012) MRI estimates
of brain iron concentration in normal aging using quantitative susceptibility mapping.
Neuroimage 59:2625–â•‰2635.
Birdsill AC, Koscik RL, Jonaitis EM, Johnson SC, Okonkwo OC, Hermann BP, Larue A, Sager
MA, Bendlin BB (2014) Regional white matter hyperintensities: Aging, Alzheimer’s dis-
ease risk, and cognitive function. Neurobiol Aging 35:769–â•‰776.
Braak H, Braak E (1996) Development of Alzheimer-â•‰related neurofibrillary changes in the
neocortex inversely recapitulates cortical myelogenesis. Acta Neuropathol 92:197–â•‰201.
Brickman AM, Meier IB, Korgaonkar MS, Provenzano FA, Grieve SM, Siedlecki KL,
Wasserman BT, Williams LM, Zimmerman ME (2012) Testing the white matter retrogen-
esis hypothesis of cognitive aging. Neurobiol Aging 33:1699–â•‰1715.
95
Budde MD, Janes L, Gold E, Turtzo LC, Frank JA (2011) The contribution of gliosis to diffu-
sion tensor anisotropy and tractography following traumatic brain injury: Validation in the
rat using Fourier analysis of stained tissue sections. Brain 134:2248–2260.
Bunce D, Anstey KJ, Christensen H, Dear K, Wen W, Sachdev P (2007) White matter hyper-
intensities and within-person variability in community-dwelling adults aged 60–64 years.
Neuropsychologia 45:2009–2015.
Burgmans S, Gronenschild EH, Fandakova Y, Shing YL, van Boxtel MP, Vuurman EF, Uylings
HB, Jolles J, Raz N (2011) Age differences in speed of processing are partially mediated
by differences in axonal integrity. Neuroimage 55:1287–1297.
Burzynska AZ, Preuschhof C, Backman L, Nyberg L, Li SC, Lindenberger U, Heekeren HR
(2010) Age-related differences in white matter microstructure: Region-specific patterns of
diffusivity. Neuroimage 49:2104–2112.
Cabeza R, Anderson ND, Locantore JK, McIntosh AR (2002) Aging gracefully: Compensatory
brain activity in high-performing older adults. Neuroimage 17:1394–1402.
Carmichael O, Lockhart S (2012) The role of diffusion tensor imaging in the study of cognitive
aging. Curr Top Behav Neurosci 11:289–320.
Catani M, Ffytche DH (2005) The rises and falls of disconnection syndromes. Brain
128:2224–2239.
Catani M, Howard RJ, Pajevic S, Jones DK (2002) Virtual in vivo interactive dissection of white
matter fasciculi in the human brain. Neuroimage 17:77–94.
Chadick JZ, Zanto TP, Gazzaley A (2014) Structural and functional differences in medial pre-
frontal cortex underlie distractibility and suppression deficits in ageing. Nat Commun
5:4223.
Charlton RA, Landau S, Schiavone F, Barrick TR, Clark CA, Markus HS, Morris RG (2008)
A structural equation modeling investigation of age-related variance in executive function
and DTI measured white matter damage. Neurobiol Aging 29:1547–1555.
Charlton RA, Barrick TR, McIntyre DJ, Shen Y, O’Sullivan M, Howe FA, Clark CA, Morris
RG, Markus HS (2006) White matter damage on diffusion tensor imaging correlates with
age-related cognitive decline. Neurology 66:217–222.
Chen NK, Chou YH, Song AW, Madden DJ (2009) Measurement of spontaneous signal fluc-
tuations in fMRI: Adult age differences in intrinsic functional connectivity. Brain Struct
Funct 213:571–585.
Colcombe SJ, Kramer AF, Erickson KI, Scalf P (2005) The implications of cortical recruitment
and brain morphology for individual differences in inhibitory function in aging humans.
Psychol Aging 20:363–375.
Concha L (2014) A macroscopic view of microstructure: Using diffusion-weighted images to
infer damage, repair, and plasticity of white matter. Neuroscience 276:14–28.
Conturo TE, Lori NF, Cull TS, Akbudak E, Snyder AZ, Shimony JS, McKinstry RC, Burton H,
Raichle ME (1999) Tracking neuronal fiber pathways in the living human brain. Proc Natl
Acad Sci U S A 96:10422–10427.
Corbin KB, Gardner ED (1937) Decrease in number of myelinated fibers in human spinal roots
with age. Anat Rec 68:63–74.
Craik FIM, Salthouse TA, eds (2008) The handbook of aging and cognition, 3rd ed.
New York: Psychology Press.
Damoiseaux JS, Greicius MD (2009) Greater than the sum of its parts: A review of studies
combining structural connectivity and resting-state functional connectivity. Brain Struct
Funct 213:525–533.
Daselaar SM, Iyengar V, Davis SW, Eklund K, Hayes SM, Cabeza RE (2013) Less wiring, more
firing: Low-performing older adults compensate for impaired white matter with greater
neural activity. Cereb Cortex 25:983–990.
96
Daugherty A, Raz N (2013) Age-related differences in iron content of subcortical nuclei

observed in vivo: A meta-analysis. Neuroimage 70:113–121.
Davis SW, Kragel JE, Madden DJ, Cabeza R (2012) The architecture of cross-hemispheric
communication in the aging brain: Linking behavior to functional and structural connec-
tivity. Cereb Cortex 22:232–242.
Davis SW, Dennis NA, Buchler NG, White LE, Madden DJ, Cabeza R (2009) Assessing the
effects of age on long white matter tracts using diffusion tensor tractography. Neuroimage
46:530–541
de Groot JC, de Leeuw FE, Oudkerk M, van Gijn J, Hofman A, Jolles J, Breteler MM (2000)
Cerebral white matter lesions and cognitive function: The Rotterdam Scan Study. Ann
Neurol 47:145–151.
de Leeuw FE, de Groot JC, Achten E, Oudkerk M, Ramos LM, Heijboer R, Hofman A, Jolles J,
van Gijn J, Breteler MM (2001) Prevalence of cerebral white matter lesions in elderly peo-
ple: A population based magnetic resonance imaging study. The Rotterdam Scan Study. J
Neurol Neurosurg Psychiatry 70:9–14.
DeCarli C, Fletcher E, Ramey V, Harvey D, Jagust WJ (2005) Anatomical mapping of white
matter hyperintensities (WMH): Exploring the relationships between periventricular
WMH, deep WMH, and total WMH burden. Stroke 36:50–55.
DeCarli C, Murphy DG, Tranh M, Grady CL, Haxby JV, Gillette JA, Salerno JA, Gonzales-
Aviles A, Horwitz B, Rapoport SI, et al. (1995) The effect of white matter hyperintensity
volume on brain structure, cognitive performance, and cerebral metabolism of glucose in
51 healthy adults. Neurology 45:2077–2084.
Douaud G, Jbabdi S, Behrens TE, Menke RA, Gass A, Monsch AU, Rao A, Whitcher B,
Kindlmann G, Matthews PM, Smith S (2011) DTI measures in crossing-fibre areas:
Increased diffusion anisotropy reveals early white matter alteration in MCI and mild
Alzheimer’s disease. Neuroimage 55:880–890.
Ennis DB, Kindlmann G (2006) Orthogonal tensor invariants and the analysis of diffusion ten-
sor magnetic resonance images. Magn Reson Med 55:136–146.
Fan Q, Nummenmaa A, Witzel T, Zanzonico R, Keil B, Cauley S, Polimeni JR, Tisdall
D, Van Dijk KR, Buckner RL, Wedeen VJ, Rosen BR, Wald LL (2014) Investigating
the capability to resolve complex white matter structures with high b-value diffusion
magnetic resonance imaging on the MGH-USC Connectom scanner. Brain Connect
4:718–726.
Filley CM (2005) White matter and behavioral neurology. Ann N Y Acad Sci 1064:162–183.
Fischer FU, Wolf D, Scheurich A, Fellgiebel A (2014) Association of structural global brain
network properties with intelligence in normal aging. PLoS One 9:e86258.
Fjell AM, Walhovd KB (2010) Structural brain changes in aging: Courses, causes and cognitive
consequences. Rev Neurosci 21:187–221.
Fjell AM, Westlye LT, Amlien IK, Walhovd KB (2011) Reduced white matter integrity is related
to cognitive instability. J Neurosci 31:18060–18072.
Frank LR (2002) Characterization of anisotropy in high angular resolution diffusion-weighted
MRI. Magn Reson Med 47:1083–1099.
Garrett DD, Kovacevic N, McIntosh AR, Grady CL (2011) The importance of being variable.
J Neurosci 31:4496–4503.
Geschwind N (1965a) Disconnexion syndromes in animals and man. II. Brain 88:585–644.
Geschwind N (1965b) Disconnexion syndromes in animals and man. I. Brain 88:237–294.
Giorgio A, Santelli L, Tomassini V, Bosnell R, Smith S, De Stefano N, Johansen-Berg H (2010)
Age-related changes in grey and white matter structure throughout adulthood. Neuroimage
51:943–951.
97
Glascher J, Rudrauf D, Colom R, Paul LK, Tranel D, Damasio H, Adolphs R (2010) Distributed
neural system for general intelligence revealed by lesion mapping. Proc Natl Acad Sci U
S A 107:4705–4709.
Gold BT, Powell DK, Xuan L, Jicha GA, Smith CD (2010) Age-related slowing of task switch-
ing is associated with decreased integrity of frontoparietal white matter. Neurobiol Aging
31:512–522.
Gong G, Rosa-Neto P, Carbonell F, Chen ZJ, He Y, Evans AC (2009) Age-and gender-related
differences in the cortical anatomical network. J Neurosci 29:15684–15693.
Greenwood PM (2000) The frontal aging hypothesis evaluated. J Int Neuropsychol Soc
6:705–726.
Greicius MD, Supekar K, Menon V, Dougherty RF (2009) Resting-state functional connectiv-
ity reflects structural connectivity in the default mode network. Cereb Cortex 19:72–78.
Grydeland H, Walhovd KB, Tamnes CK, Westlye LT, Fjell AM (2013) Intracortical myelin
links with performance variability across the human lifespan: Results from T1-and T2-
weighted MRI myelin mapping and diffusion tensor imaging. J Neurosci 33:18618–18630.
Gunning-Dixon FM, Raz N (2000) The cognitive correlates of white matter abnormalities in
normal aging: A quantitative review. Neuropsychology 14:224–232.
Gunning-Dixon FM, Raz N (2003) Neuroanatomical correlates of selected executive func-
tions in middle- aged and older adults: A prospective MRI study. Neuropsychologia
41:1929–1941.
Gunning-Dixon FM, Brickman AM, Cheng JC, Alexopoulos GS (2009) Aging of cerebral
white matter: A review of MRI findings. Int J Geriatr Psychiatry 24:109–117.
Haacke EM, Cheng NY, House MJ, Liu Q, Neelavalli J, Ogg RJ, Khan A, Ayaz M, Kirsch W,
Obenaus A (2005) Imaging iron stores in the brain using magnetic resonance imaging.
Magn Reson Imaging 23:1–25.
Hallgren B, Sourander P (1958) The effect of age on the non-haemin iron in the human brain.
J Neurochem 3:41–51.
Hartline DK, Colman DR (2007) Rapid conduction and the evolution of giant axons and
myelinated fibers. Curr Biol 17:R29–R35.
Hayes AF (2009) Beyond Baron and Kenny: Statistical mediation analysis in the new millen-
nium. Commun Monogr 76:408–420.
He BJ, Snyder AZ, Vincent JL, Epstein A, Shulman GL, Corbetta M (2007) Breakdown of
functional connectivity in frontoparietal networks underlies behavioral deficits in spatial
neglect. Neuron 53:905–918.
Head D, Buckner RL, Shimony JS, Williams LE, Akbudak E, Conturo TE, McAvoy M, Morris
JC, Snyder AZ (2004) Differential vulnerability of anterior white matter in nondemented
aging with minimal acceleration in dementia of the Alzheimer type: Evidence from diffu-
sion tensor imaging. Cereb Cortex 14:410–423.
Hofer SM, Sliwinski MJ (2001) Understanding Ageing. An evaluation of research designs for
assessing the interdependence of ageing-related changes. Gerontology 47:341–352.
Hubbard PL, Parker GJM (2014) Validation of tractography. In: Diffusion MRI: From quantita-
tive measurement to in-vivo neuroanatomy (2nd edition) (Johansen-Berg H, Behrens TEJ,
eds), pp 453–480. San Diego, CA: Elsevier.
Hugenschmidt CE, Peiffer AM, Kraft RA, Casanova R, Deibler AR, Burdette JH, Maldjian
JA, Laurienti PJ (2008) Relating imaging indices of white matter integrity and volume in
healthy older adults. Cereb Cortex 18:433–442.
Johnson MA, Diaz MT, Madden DJ (2015) Global versus tract-specific components of cere-
bral white matter integrity: relation to adult age and perceptual-motor speed. Brain Struct
Funct 220: 2705–2720.
98
Jones DK (2008) Studying connections in the living human brain with diffusion MRI. Cortex
44:936–952.
Jones DK (2011) Diffusion MRI: Theory, methods, and applications New York: Oxford
University Press.
Jones DK, Knösche TR, Turner R (2013) White matter integrity, fiber count, and other falla-
cies: The do’s and don’ts of diffusion MRI. Neuroimage 73:239–254.
Kennedy KM, Raz N (2009) Aging white matter and cognition: Differential effects of
regional variations in diffusion properties on memory, executive functions, and speed.
Kerchner GA, Racine CA, Hale S, Wilheim R, Laluz V, Miller BL, Kramer JH (2012) Cognitive
processing speed in older adults: relationship with white matter integrity. PLoS One 7:e50425.
Kievit RA, Davis SW, Mitchell DJ, Taylor JR, Duncan J, Cam CANRt, Henson RNA (2014)
Distinct aspects of frontal lobe structure mediate age-related differences in fluid intelli-
gence and multitasking. Nat Commun 5. DOI: 10.1038/ncomms6658
Klawiter EC, Schmidt RE, Trinkaus K, Liang HF, Budde MD, Naismith RT, Song SK, Cross
AH, Benzinger TL (2011) Radial diffusivity predicts demyelination in ex vivo multiple
sclerosis spinal cords. Neuroimage 55:1454–1460.
Kochunov P, Williamson DE, Lancaster J, Fox P, Cornell J, Blangero J, Glahn DC (2012)
Fractional anisotropy of water diffusion in cerebral white matter across the lifespan.
Neurobiol Aging 33:9–20.
Langkammer C, Schweser F, Krebs N, Deistung A, Goessler W, Scheurer E, Sommer K,
Reishofer G, Yen K, Fazekas F, Ropele S, Reichenbach JR (2012) Quantitative suscepti-
bility mapping (QSM) as a means to measure brain iron? A post mortem validation study.
Neuroimage 62:1593–1599.
Li W, Wu B, Liu C (2011) Quantitative susceptibility mapping of human brain reflects spatial
variation in tissue composition. Neuroimage 55:1645–1656.
Li W, Langkammer C, Chou Y-H, Petrovic K, Schmidt R, Song AW, Madden DJ, Ropele S,
Liu C (2015) Association between increased magnetic susceptibility of deep gray matter
nuclei and decreased motor function in healthy adults. Neuroimage 105:45–52.
Li W, Wu B, Batrachenko A, Bancroft-Wu V, Morey RA, Shashi V, Langkammer C, De Bellis
MD, Ropele S, Song AW, Liu C (2014) Differential developmental trajectories of mag-
netic susceptibility in human brain gray and white matter over the lifespan. Hum Brain
Mapp 35:2698–2713.
Li YO, Yang FG, Nguyen CT, Cooper SR, LaHue SC, Venugopal S, Mukherjee P (2012)
Independent component analysis of DTI reveals multivariate microstructural correlations
of white matter in the human brain. Hum Brain Mapp 33:1431–1451.
Lindenberger U, von Oertzen T, Ghisletta P, Hertzog C (2011) Cross-sectional age variance
extraction: What’s change got to do with it? Psychol Aging 26:34–47.
Liu C, Mang SC, Moseley ME (2010) In vivo generalized diffusion tensor imaging (GDTI)
using higher-order tensors (HOT). Magn Reson Med 63:243–252.
Liu C, Li W, Johnson GA, Wu B (2011) High-field (9.4 T) MRI of brain dysmyelination by
quantitative mapping of magnetic susceptibility. Neuroimage 56:930–938.
Liu C, Li W, Tong KA, Yeom KW, Kuzminski S (2015) Susceptibility-weighted imaging and
quantitative susceptibility mapping in the brain. J Magn Reson Imaging 42:23–41.
Lockhart SN, Roach AE, Luck SJ, Geng J, Beckett L, Carmichael O, DeCarli C (2014) White
matter hyperintensities are associated with visual search behavior independent of general-
ized slowing in aging. Neuropsychologia 52:93–101.
Longstreth WT, Jr., Arnold AM, Beauchamp NJ, Jr., Manolio TA, Lefkowitz D, Jungreis C,
Hirsch CH, O’Leary DH, Furberg CD (2005) Incidence, manifestations, and predictors of
99
worsening white matter on serial cranial magnetic resonance imaging in the elderly: The
Cardiovascular Health Study. Stroke 36:56–61.
Lovden M, Kohncke Y, Laukka EJ, Kalpouzos G, Salami A, Li TQ, Fratiglioni L, Backman L
tract are associated in very old age. Neuroimage 102P2:520–530.
Lovden M, Laukka EJ, Rieckmann A, Kalpouzos G, Li TQ, Jonsson T, Wahlund LO, Fratiglioni
L, Backman L (2013) The dimensionality of between-person differences in white matter
microstructure in old age. Hum Brain Mapp 34:1386–1398.
Lu PH, Lee GJ, Tishler TA, Meghpara M, Thompson PM, Bartzokis G (2013) Myelin break-
down mediates age-related slowing in cognitive processing speed in healthy elderly men.
Brain Cogn 81:131–138.
MacDonald SW, Nyberg L, Backman L (2006a) Intra-individual variability in behavior:
Links to brain structure, neurotransmission and neuronal activity. Trends Neurosci
29:474–480.
MacDonald SW, Hultsch DF, Bunce D (2006b) Intraindividual variability in vigilance per-
formance: Does degrading visual stimuli mimic age-related “neural noise?” J Clin Exp
Neuropsychol 28:655–675.
Madden DJ, Bennett IJ, Song AW (2009a) Cerebral white matter integrity and cognitive aging:
Contributions from diffusion tensor imaging. Neuropsychol Rev 19:415–435.
Madden DJ, Bennett IJ, Burzynska A, Potter GG, Chen NK, Song AW (2012) Diffusion ten-
sor imaging of cerebral white matter integrity in cognitive aging. Biochim Biophys Acta
1822:386–400.
Madden DJ, Spaniol J, Whiting WL, Bucur B, Provenzale JM, Cabeza R, White LE, Huettel SA
(2007) Adult age differences in the functional neuroanatomy of visual attention: A com-
bined fMRI and DTI study. Neurobiol Aging 28:459–476.
Madden DJ, Costello MC, Dennis NA, Davis SW, Shepler AM, Spaniol J, Bucur B, Cabeza
R (2010) Adult age differences in functional connectivity during executive control.
Madden DJ, Parks EL, Davis SW, Diaz MT, Potter GG, Chou YH, Chen NK, Cabeza R (2014)
Age mediation of frontoparietal activation during visual feature search. Neuroimage
102P2:262–274.
Madden DJ, Spaniol J, Costello MC, Bucur B, White LE, Cabeza R, Davis SW, Dennis NA,
Provenzale JM, Huettel SA (2009b) Cerebral white matter integrity mediates adult age
differences in cognitive performance. J Cogn Neurosci 21:289–302.
Maillard P, Carmichael O, Fletcher E, Reed B, Mungas D, DeCarli C (2012) Coevolution of
white matter hyperintensities and cognition in the elderly. Neurology 79:442–448.
Maillard P, Carmichael O, Harvey D, Fletcher E, Reed B, Mungas D, DeCarli C (2013) FLAIR
and diffusion MRI signals are independent predictors of white matter hyperintensities.
AJNR Am J Neuroradiol 34:54–61.
Maillard P, Fletcher E, Lockhart SN, Roach AE, Reed B, Mungas D, DeCarli C, Carmichael OT
(2014) White matter hyperintensities and their penumbra lie along a continuum of injury
in the aging brain. Stroke 45:1721–1726.
Marner L, Nyengaard JR, Tang Y, Pakkenberg B (2003) Marked loss of myelinated nerve fibers
in the human brain with age. J Comp Neurol 462:144–152.
McIntosh AR (2000) Towards a network theory of cognition. Neural Netw 13:861–870.
Meier-Ruge W, Ulrich J, Bruhlmann M, Meier E (1992) Age-related white matter atrophy in the
human brain. Ann N Y Acad Sci 673:260–269.
Mesulam MM (1990) Large-scale neurocognitive networks and distributed processing for
attention, language, and memory. Ann Neurol 28:597–613.
100
Miller KL, Stagg CJ, Douaud G, Jbabdi S, Smith SM, Behrens TE, Jenkinson M, Chance SA,
Esiri MM, Voets NL, Jenkinson N, Aziz TZ, Turner M, Johansen-Berg H, McNab JA
(2011) Diffusion imaging of whole, post-mortem human brains on a clinical MRI scanner.
Mori S, Zhang J (2006) Principles of diffusion tensor imaging and its applications to basic
neuroscience research. Neuron 51:527–539.
Nyberg L, Lovden M, Riklund K, Lindenberger U, Backman L (2012) Memory aging and brain
maintenance. Trends Cogn Sci 16:292–305.
Persson J, Nilsson LG (2010) Longitudinal evidence for diminished frontal cortex func-
tion in aging. Proc Natl Acad Sci U S A 107:22682–22686.
O’Sullivan M, Jones DK, Summers PE, Morris RG, Williams SC, Markus HS (2001) Evidence
for cortical “disconnection” as a mechanism of age-related cognitive decline. Neurology
57:632–638.
Park DC, Reuter-Lorenz P (2009) The adaptive brain: aging and neurocognitive scaffolding.
Annu Rev Psychol 60:173–196.
Parks CM, Iosif A-M, Farias S, Reed B, Mungas D, DeCarli C (2011) Executive function
mediates effects of white matter hyperintensities on episodic memory. Neuropsychologia
49:2817–2824.
Paus T, Zijdenbos A, Worsley K, Collins DL, Blumenthal J, Giedd JN, Rapoport JL, Evans AC
(1999) Structural maturation of neural pathways in children and adolescents: In vivo study.
Science 283:1908–1911.
Penke L, Valdés Hernandéz MC, Maniega SM, Gow AJ, Murray C, Starr JM, Bastin ME, Deary
IJ, Wardlaw JM (2012) Brain iron deposits are associated with general cognitive ability
and cognitive aging. Neurobiol Aging 33:510–517.e512.
Penke L, Munoz Maniega S, Murray C, Gow AJ, Hernandez MC, Clayden JD, Starr JM,
Wardlaw JM, Bastin ME, Deary IJ (2010) A general factor of brain white matter integrity
predicts information processing speed in healthy older people. J Neurosci 30:7569–7574.
Perry ME, McDonald CR, Hagler DJ, Jr., Gharapetian L, Kuperman JM, Koyama AK, Dale
AM, McEvoy LK (2009) White matter tracts associated with set-shifting in healthy aging.
Peters A (2002) The effects of normal aging on myelin and nerve fibers: A review. J Neurocytol
31:581–593.
Peters A, Sethares C (2003) Is there remyelination during aging of the primate central nervous
system? J Comp Neurol 460:238–254.
Pfefferbaum A, Adalsteinsson E, Sullivan EV (2005) Frontal circuitry degradation marks
healthy adult aging: Evidence from diffusion tensor imaging. Neuroimage 26:891–899.
Pierpaoli C, Barnett A, Pajevic S, Chen R, Penix LR, Virta A, Basser P (2001) Water diffu-
sion changes in Wallerian degeneration and their dependence on white matter architecture.
Preacher KJ, Hayes AF (2008) Asymptotic and resampling strategies for assessing and compar-
ing indirect effects in multiple mediator models. Behav Res Methods 40:879–891.
Prins ND, van Dijk EJ, den Heijer T, Vermeer SE, Jolles J, Koudstaal PJ, Hofman A, Breteler
MM (2005) Cerebral small-vessel disease and decline in information processing speed,
executive function and memory. Brain 128:2034–2041.
Rabbitt P, Scott M, Lunn M, Thacker N, Lowe C, Pendleton N, Horan M, Jackson A (2007)
White matter lesions account for all age-related declines in speed but not in intelligence.
Neuropsychology 21:363–370.
Ramnani N, Behrens TE, Penny W, Matthews PM (2004) New approaches for exploring ana-
tomical and functional connectivity in the human brain. Biol Psychiatry 56:613–619.
101
Raz N, Rodrigue KM, Kennedy KM, Acker JD (2007) Vascular health and longitudinal changes
in brain and cognition in middle-aged and older adults. Neuropsychology 21:149–157.
Reisberg B, Franssen EH, Souren LE, Auer SR, Akram I, Kenowsky S (2002) Evidence and
mechanisms of retrogenesis in Alzheimer’s and other dementias: Management and treat-
ment import. Am J Alzheimers Dis Other Demen 17:202–212.
Reuter-Lorenz PA, Cappell KA (2008) Neurocognitive aging and the compensation hypothesis.
Curr Dir Psychol Sci 17:177–182.
Rodrigue KM, Daugherty AM, Haacke EM, Raz N (2013) The role of hippocampal iron con-
centration and hippocampal volume in age-related differences in memory. Cereb Cortex
23:1533–1541.
Ronnlund M, Nyberg L, Backman L, Nilsson LG (2005) Stability, growth, and decline in adult
life span development of declarative memory: Cross-sectional and longitudinal data from
a population-based study. Psychol Aging 20:3–18.
Salami A, Pudas S, Nyberg L (2014) Elevated hippocampal resting- state connectiv-
ity underlies deficient neurocognitive function in aging. Proc Natl Acad Sci U S A
111:17654–17659.
Salami A, Eriksson J, Nilsson LG, Nyberg L (2012) Age-related white matter microstructural
differences partly mediate age-related decline in processing speed but not cognition.
Biochim Biophys Acta 1822:408–415.
Salat DH (2011) The declining infrastructure of the aging brain. Brain Connect 1:279–293.
Salat DH (2014a) Imaging small vessel-associated white matter changes in aging. Neuroscience
276:174–186.
Salat DH (2014b) Diffusion tensor imaging in the study of aging and age-associated neural dis-
iease. In: Diffusion MRI: From quantitative measurement to in-vivo neuroanatomy (2nd
edition) (Johansen-Berg H, Behrens TEJ, eds), pp 257–281. San Diego, CA: Elsevier.
Salat DH, Tuch DS, Greve DN, van der Kouwe AJ, Hevelone ND, Zaleta AK, Rosen BR, Fischl
B, Corkin S, Rosas HD, Dale AM (2005) Age-related alterations in white matter micro-
structure measured by diffusion tensor imaging. Neurobiol Aging 26:1215–1227.
Salthouse TA (2009) When does age- related cognitive decline begin? Neurobiol Aging
30:507–514.
Salthouse TA (2011) Neuroanatomical substrates of age-related cognitive decline. Psychol Bull
137:753–784.
Salthouse TA, Nesselroade JR (2002) An examination of the Hofer and Sliwinski evaluation.
Gerontology 48:18–21; discussion 22–19.
Salthouse TA, Berish DE (2005) Correlates of within-person (across-occasion) variability in
reaction time. Neuropsychology 19:77–87.
Seidl AH (2014) Regulation of conduction time along axons. Neuroscience 276:126–134.
Seunarine KK, Alexander DC (2014) Multiple fibers: Beyond the diffusion tensor. In: Diffusion
MRI: From quantitative measurement to in-vivo neuroanatomy (2nd edition) (Johansen-
Berg H, Behrens TEJ, eds), pp 105–123. San Diego, CA: Elsevier.
Song SK, Sun SW, Ramsbottom MJ, Chang C, Russell J, Cross AH (2002) Dysmyelination
revealed through MRI as increased radial (but unchanged axial) diffusion of water.
Song SK, Yoshino J, Le TQ, Lin SJ, Sun SW, Cross AH, Armstrong RC (2005) Demyelination
increases radial diffusivity in corpus callosum of mouse brain. Neuroimage 26:132–140.
Mapping cortical change across the human life span. Nat Neurosci 6:309–315.
Sporns O (2014) The human connectome: Linking structure and functionin in the human brain.
In: Diffusion MRI: From quantitative measurement to in-vivo neuroanatomy (2nd edition)
(Johansen-Berg H, Behrens TEJ, eds), pp 401–428. San Diego, CA: Elsevier.
102
Sullivan EV, Pfefferbaum A (2006) Diffusion tensor imaging and aging. Neurosci Biobehav
Rev 30:749–761.
Sullivan EV, Pfefferbaum A (2007) Neuroradiological characterization of normal adult ageing.
Br J Radiol 80 Spec No 2:S99–S108.
Sullivan EV, Adalsteinsson E, Pfefferbaum A (2006) Selective age-related degradation of
anterior callosal fiber bundles quantified in vivo with fiber tracking. Cereb Cortex
16:1030–1039.
Sullivan EV, Rohlfing T, Pfefferbaum A (2010a) Quantitative fiber tracking of lateral and
interhemispheric white matter systems in normal aging: Relations to timed performance.
Sullivan EV, Adalsteinsson E, Rohlfing T, Pfefferbaum A (2009) Relevance of iron deposition
in deep gray matter brain structures to cognitive and motor performance in healthy elderly
men and women: Exploratory findings. Brain Imaging Behav 3:167–175.
Sullivan EV, Zahr NM, Rohlfing T, Pfefferbaum A (2010b) Fiber tracking functionally distinct
components of the internal capsule. Neuropsychologia 48:4155–4163.
Sullivan EV, Adalsteinsson E, Hedehus M, Ju C, Moseley M, Lim KO, Pfefferbaum A (2001)
Equivalent disruption of regional white matter microstructure in ageing healthy men and
women. Neuroreport 12:99–104.
Todorich B, Pasquini JM, Garcia CI, Paez PM, Connor JR (2009) Oligodendrocytes and
myelination: The role of iron. Glia 57:467–478.
Valdes Hernandez MC, Piper RJ, Bastin ME, Royle NA, Maniega SM, Aribisala BS, Murray
C, Deary IJ, Wardlaw JM (2014) Morphologic, distributional, volumetric, and intensity
characterization of periventricular hyperintensities. AJNR Am J Neuroradiol 35:55–62.
van den Heuvel M, Mandl R, Luigjes J, Hulshoff Pol H (2008) Microstructural organization
of the cingulum tract and the level of default mode functional connectivity. J Neurosci
28:10844–10851.
van den Heuvel MP, Mandl RC, Kahn RS, Hulshoff Pol HE (2009) Functionally linked resting-
state networks reflect the underlying structural connectivity architecture of the human
brain. Hum Brain Mapp 30:3127–3141.
Vannorsdall TD, Waldstein SR, Kraut M, Pearlson GD, Schretlen DJ (2009) White mat-
ter abnormalities and cognition in a community sample. Arch Clin Neuropsychol
24:209–217.
Vernooij MW, de Groot M, van der Lugt A, Ikram MA, Krestin GP, Hofman A, Niessen WJ,
Breteler MM (2008) White matter atrophy and lesion formation explain the loss of struc-
tural integrity of white matter in aging. Neuroimage 43:470–477.
Vernooij MW, Ikram MA, Tanghe HL, Vincent AJ, Hofman A, Krestin GP, Niessen WJ, Breteler
MM, van der Lugt A (2007) Incidental findings on brain MRI in the general population. N
Engl J Med 357:1821–1828.
Voineskos AN, Rajji TK, Lobaugh NJ, Miranda D, Shenton ME, Kennedy JL, Pollock BG,
Mulsant BH (2012) Age-related decline in white matter tract integrity and cognitive per-
formance: A DTI tractography and structural equation modeling study. Neurobiol Aging
33:21–34.
Wahl M, Li YO, Ng J, Lahue SC, Cooper SR, Sherr EH, Mukherjee P (2010) Microstructural
correlations of white matter tracts in the human brain. Neuroimage 51:531–541.
Walhovd KB, Fjell AM (2007) White matter volume predicts reaction time instability.
Walhovd KB, Johansen-Berg H, Káradóttir RT (2014) Unraveling the secrets of white mat-
ter –Bridging the gap between cellular, animal and human imaging studies. Neuroscience
276:2–13.
103
Walhovd KB, Westlye LT, Amlien I, Espeseth T, Reinvang I, Raz N, Agartz I, Salat DH, Greve
DN, Fischl B, Dale AM, Fjell AM (2011) Consistent neuroanatomical age-related volume
differences across multiple samples. Neurobiol Aging 32:916–932.
Wang S, Young KM (2014) White matter plasticity in adulthood. Neuroscience 276:148–160.
Waxman SG (1980) Determinants of conduction velocity in myelinated nerve fibers. Muscle
Nerve 3:141–150.
Werring DJ, Clark CA, Parker GJ, Miller DH, Thompson AJ, Barker GJ (1999) A direct dem-
onstration of both structure and function in the visual system: Combining diffusion tensor
imaging with functional magnetic resonance imaging. Neuroimage 9:352–361.
Westlye LT, Walhovd KB, Dale AM, Bjornerud A, Due-Tonnessen P, Engvig A, Grydeland H,
Tamnes CK, Ostby Y, Fjell AM (2010) Life-span changes of the human brain white matter:
Diffusion tensor imaging (DTI) and volumetry. Cereb Cortex 20:2055–2068.
Wheeler-Kingshott CA, Cercignani M (2009) About “axial” and “radial” diffusivities. Magn
Reson Med 61:1255–1260.
Yakovlev PI, Lecours AR (1967) The mylogenetic cycles of regional maturation in the brain.
In: Regional development of the brain (Minowski A, ed), pp 3–70. Oxford: Blackwell.
Yoshita M, Fletcher E, DeCarli C (2005) Current concepts of analysis of cerebral white matter
hyperintensities on magnetic resonance imaging. Top Magn Reson Imaging 16:399–407.
Young VG, Halliday GM, Kril JJ (2008) Neuropathologic correlates of white matter hyperin-
tensities. Neurology 71:804–811.
Zahr NM, Rohlfing T, Pfefferbaum A, Sullivan EV (2009) Problem solving, working mem-
ory, and motor correlates of association and commissural fiber bundles in normal
aging: A quantitative fiber tracking study. Neuroimage 44:1050–1062.
Zecca L, Youdim MB, Riederer P, Connor JR, Crichton RR (2004) Iron, brain ageing and neu-
rodegenerative disorders. Nat Rev Neurosci 5:863–873.
104
105
Age Differences in Functional

Connectivity at Rest and During
Cognitive Tasks
Cheryl L. Grady
T he past few years have seen a steep increase in the number of papers
examining functional connectivity (FC) in the brain. In particular, mea-
surement of FC during the resting state, sometimes called “intrinsic FC,” has become
a popular means of studying brain functional organization in both health and disease
(some recent reviews of this topic include Damoiseaux and Greicius, 2009; Fox and
Greicius, 2010; Deco et al., 2011; Hutchison et al., 2013; Vaidya and Gordon, 2013).
In this chapter I will review the literature on age differences in FC, at rest and during
performance of cognitive tasks, to assess the current state of the field and suggest
some ideas for future research (for another recent review of this topic, see Ferreira
and Busatto, 2013).
Before reviewing the literature, however, it is worth noting that there currently is
no “gold standard” for how to identify FC within brain networks, or even a consensus
on what constitutes a network. Most researchers assume that a network consists of a
set of regions that show strong FC with one another, i.e., have activity that covaries
across time, and this has proved to be a useful operational definition. However, we
do not know how many networks there might be in the brain, or what might be the
best way to identify them. A number of approaches have been used to do this, includ-
ing “seed” based methods in which a region of interest is chosen and activity in this
seed is correlated with activity in other brain regions (e.g., Andrews-Hanna et al.,
2010; Campbell et al., 2013). Other studies have assessed voxel-wise correlations
(Hampson et al., 2012) or pair-wise regional connections across the whole brain. This
latter approach involves defining a set of regions of interest (ROIs) and calculating
105
106
pair-â•‰wise correlations using the time courses from these ROIs (e.g., Wang et al., 2012;
Chou et al., 2013). Other approaches are entirely data-â•‰driven and identify multiple
sets of regions across the whole brain that show strong correlations in activity, with
some method needed to choose the number of networks, or components, to keep (e.g.,
Smith et al., 2009; Afshin-â•‰Pour et al., 2014). Graph theory also is gaining use in this
field (Bullmore and Sporns, 2009), and consists of defining a network based on the
correlations or “edges” that exist between regions, and then calculating various met-
rics that describe regional participation in these networks, such as degree (number of
correlations for a given region) and modularity (clustering among groups of regions
within the larger brain network). Unfortunately, the results that one gets can vary
across methods, although some networks, such as the default mode network (DMN),
are robustly identified regardless of the analytical approach that is used (Afshin-â•‰Pour
et al., 2014). Even for the DMN, however, there is some uncertainty about which
regions are reliable nodes within the network (Grigg and Grady, 2010a; Habeck et al.,
2012) or how many subsystems it might include (Andrews-â•‰Hanna et al., 2010; Leech
et al., 2011). Despite the limitations, these methods provide useful windows into brain
complexity that cannot be appreciated by looking at activation patterns alone.
Commonly Studied Brain Networks
The DMN is the most often-â•‰studied brain network and is implicated in normal cog-
nition, as well as a number of disorders (for a review, see Buckner et al., 2008). The
major nodes of the DMN include posterior cingulate cortex (PCC), ventromedial
prefrontal cortex (PFC), the angular gyri and the parahippocampal gyri. This net-
work shows reduced activity during the kind of externally-â•‰driven task typically used
in fMRI experiments (e.g., encoding or recognizing visual stimuli) and increased
activity during rest or fixation (Gusnard et al., 2001). The DMN is thought to under-
lie self-â•‰reference and projection of the self through the past (memory) and future
(planning), as well as having a role in social cognition, such as theory of mind
(Buckner and Carroll, 2007; Harrison et al., 2008; Grigg and Grady, 2010a; Spreng
and Grady, 2010; Andrews-â•‰Hanna et al., 2014). Importantly, DMN modulation is
related to cognition more generally because greater reduction of DMN activity dur-
ing tasks and stronger functional connectivity among DMN nodes both are related
to better performance on a number of tasks (Kelly et al., 2008; Dang et al., 2013).
Another network described in the literature is the frontoparietal control network, or
FPC (Vincent et al., 2008), sometimes known as the central executive network (Seeley
et al., 2007). The nodes of this network include dorsolateral PFC (DLPFC), rostral
frontal cortex near the frontal pole, and the inferior parietal lobes. The FPC is thought
to act as a “switch” to flexibly control the engagement of other brain networks and
thus support the cognitive control processes needed to meet task demands (Cole et al.,
2013; Spreng et al., 2013). The salience network, or SLN (Seeley et al., 2007), also is
involved in the control of behavior, and its major nodes include the anterior cingulate
(ACC), anterior insula/â•‰frontal operculum (aIFO), amygdala and ventral striatum. The
SLN is thought to integrate sensory data with internal states (e.g., visceral, autonomic,
and hedonic “markers”) so that the organism can guide its behavior in the context
â•‡ 107
Age Differences in Functional Connectivity at Rest and During CognitiveÂ€Tasksâ•… 107
of salient events and adapt to changing demands in the environment (Seeley et al.,
2007; Ham et al., 2013). Regions in the FPC and SLN are active during such tasks as
working memory, task switching, planning, and other goal-directed behaviors (Luks
et al., 2002; Owen et al., 2005; Dosenbach et al., 2007; Grady et al., 2010; Spreng et
al., 2010). A third network involved in the control of behavior is the dorsal attention
network (DAN), which includes superior parietal cortex, the frontal eye fields and
lateral temporal regions. The role of this network is thought to lie in the enabling of
responding on tasks in accord with defined goals (Fox et al., 2006; Corbetta et al.,
2008; Shulman et al., 2009). These three control networks are considered to be dis-
tinct, although there is some overlap in nodes, e.g., the aIFO and ACC participate in
both the FPC and SLN.
An interesting property of the DMN and the cognitive control networks mentioned
so far is that activity in these networks is often anti-â•‰correlated. That is, activity in
DMN regions is negatively correlated with activity in the DAN and SLN at rest (Fox
et al., 2005), and the strength of this anti-â•‰correlation is positively related to better
or more consistent performance on tasks (Kelly et al., 2008). The DMN and FPC
also can be negatively correlated at rest, although correlations also can be positive
between these two networks, particularly during tasks that activate the DMN (Spreng
et al., 2010). Thus, the evidence to date suggests that modulation of activity in mul-
tiple networks and functional connectivity within and between networks all reflect the
dynamic range of network activity and are important for cognition.
Other networks appearing in the literature are the sensorimotor network (including
the pre-â•‰and post-â•‰central gyri), auditory and visual networks, the ventral attention net-
work (which overlaps with the SLN), the language network, and various subcortical
and cerebellar networks (Allen et al., 2011; Laird et al., 2011). These networks have
received less attention than the DMN and the cognitive control networks, although
some attempts have been made to link them to specific cognitive tasks (Laird et al.,
2011). In the following sections, I will summarize the trends that have emerged from
the literature on FC in aging with a focus on these specific networks where possible.
Resting Functional Connectivity
Since it was first shown with fMRI that there was strong FC among brain regions at
rest (Biswal et al., 1995), this approach has been used extensively in both younger
and older adults, and in various disorders (Greicius, 2008; Damoiseaux and Greicius,
2009; Fox and Greicius, 2010; van den Heuvel and Hulshoff Pol, 2010; Keilholz,
2014). I will first review the evidence for age differences in resting FC and then dis-
cuss the relevance of these for cognitive performance.
Studies Assessing Resting FC Across the Whole Brain

Some of the work on age differences in functional connectivity at rest has assessed the
whole brain without focusing on any specific network. Using this approach, several stud-
ies have found diffuse reductions with age in many connections, with the exception of
108
those involving subcortical regions, such as caudate and thalamus (Meunier et al., 2009;
Wang et al., 2012; Cao et al., 2014). Others have noted more restricted changes, such as
stronger FC in younger vs. older adults in temporal lobe regions (Achard and Bullmore,
2007; Chou et al., 2013), some frontal regions (Hampson et al., 2012; Tomasi and Volkow,
2012), occipital cortex (Achard and Bullmore, 2007; Meinzer et al., 2013) and regions
associated with the DMN, such as the PCC (Zuo et al., 2012). In contrast to these age
reductions, there are reports that older adults have stronger connections in sensorimotor
cortex (Meunier et al., 2009; Tomasi and Volkow, 2012; Onoda and Yamaguchi, 2013),
subcortical regions (Meinzer et al., 2013; Hampson et al., 2012; Zuo et al., 2012), and
insula (Meinzer et al., 2013; Tomasi and Volkow, 2012). Although the results of these
studies are somewhat diverse, the evidence points to age decreases involving PCC and
frontal regions, with age increases in sensorimotor and subcortical areas.
Published evidence on FC involving the medial temporal lobes (MTL) has been
particularly inconsistent, with some reporting stronger functional connections in
younger than in older adults (Achard and Bullmore, 2007; Chou et al., 2013), and
others finding the opposite (Meinzer et al., 2013; Hampson et al., 2012; Zuo et al.,
2012). A recent study exploring resting FC in a relatively large sample from 20–â•‰80
years of age found results that may help to understand this discrepancy in the litera-
ture (Salami et al., 2014a). In this study, older adults showed reduced functional con-
nectivity between the hippocampus and DMN regions, but stronger FC between the
right and left hippocampi. This result suggests that both age increases and decreases
in MTL coupling are possible, depending on the specific connections that are under
consideration. The inconsistent results found in the MTL also may be related to the
specific subcomponents of this area, such as those found along the long axis of the
hippocampus, which are noted to have different anatomical connections and functions
(e.g., Poppenk et al., 2013; Strange et al., 2014). It may be that some MTL subregions
are more vulnerable to age than others, but this has not been examined systematically
in terms of functional connectivity.
Studies Assessing Resting FC in Specific Networks

A number of studies have examined resting FC in specific brain networks in young
and older adults. These studies have consistently found reduced intrinsic FC in the
DMN with age (Andrews-â•‰Hanna et al., 2007; Damoiseaux et al., 2008; Grady et al.,
2010; Sambataro et al., 2010; Allen et al., 2011; Wu et al., 2011; Grady et al., 2012;
Hampson et al., 2012; Meier et al., 2012; Mowinckel et al., 2012; Onoda et al., 2012;
Sala-â•‰Llonch et al., 2012; Tomasi and Volkow, 2012; Campbell et al., 2013; Mevel et
al., 2013; Chan et al., 2014; Madhyastha and Grabowski, 2014; Salami et al., 2014a;
Schultz et al., 2014; Song et al., 2014; Geerligs et al., 2015; Saverino et al., 2015),
and some have shown weaker coupling between the DMN and other networks as well
(Wu et al., 2011; Meier et al., 2012; Onoda et al., 2012; He et al., 2013). However, age
reductions in DMN connectivity are likely not uniform over all network nodes. For
example, some researchers have found evidence to support the presence of different
subsystems in the DMN (Andrews-â•‰Hanna et al., 2010; Leech et al., 2011), involving
nodes that typically are functionally connected to the DMN, but also can show distinct
109
Age Differences in Functional Connectivity at Rest and During Cognitive Tasks 109
FC patterns. Two such subsystems are the MTL and dmPFC subsystems, thought to
underlie past/future thought and self reference in the present, respectively (Andrews-
Hanna et al., 2010). Subsystems involving the ventral and dorsal PCC also have been
described (Leech et al., 2011), which vary in the degree to which they couple with
purely DMN regions (ventral PCC) or some task-related areas (dorsal PCC). We
showed recently that age reductions are not seen uniformly throughout all DMN sub-
systems (Campbell et al., 2013). Older adults had reduced FC in the ventral PCC and
dmPFC subsystems (which mainly engaged typical DMN regions) but not the MTL
subsystem. In addition, older adults had stronger FC in a subsystem based on the dor-
sal PCC, which consisted of mostly task-related regions, including some areas from
the SLN (Figure 4.1). This result suggests some variability in the effects of aging on
the resting FC of DMN nodes, but confirms the age reduction in the nodes thought to
represent the core of the network (i.e., PCC, ventromedial PFC, angular gyri).
vPCC
Resting functional connectivity

MTL 0.5 * *
0.4
Correlation
0.3
0.2 Young
Older
0.1
0
vPCC MTL dPCC
dPCC
Figure 4.1 Age differences in FC within DMN subsystems are shown. The images on the left
show the patterns of FC with the ventral PCC (vPCC subsystem), the left parahippocampal
gyrus (MTL subsystem) and the dorsal PCC (dPCC subsystem) that were identified in both age
groups. The graph at the right shows that the strength of seed correlation with these FC patterns
differs with age. The mean correlation (across a resting state run) between seed activity and
activity in the relevant brain regions for younger and older adults is plotted. Age differences are
indicated by asterisks. Data are from Campbell et al. (2013). (See color plate also)
110
In terms of other brain networks, reduced resting FC in older adults, compared to

young adults, has been reported in the SLN, FPC, DAN, and visual networks (Andrews-â•‰
Hanna et al., 2007; Meier et al., 2012, Geerligs, 2015; Mowinckel et al., 2012; Onoda
et al., 2012; He et al., 2014; Schultz et al., 2014). Weaker FC in the DAN and FPC
also has been noted in older adults compared to middle aged adults (Madhyastha and
Grabowski, 2014). In contrast, greater FC in older adults has been found in sensorimo-
tor and subcortical networks (Allen et al., 2011; Meier et al., 2012). One interesting
finding is that older adults have increased connectivity between networks despite having
reduced connectivity within networks, indicating age reductions in network segregation.
This has been reported for specific networks, such as the visual and sensorimotor net-
works (Meier et al., 2012; Geerligs et al., 2015), as well more diffusely across multiple
networks (Chan et al., 2014). How these two aspects of age differences in network FC
influence one another, and which one occurs first, will be important topics for future
work to address. In general, these results from studies targeting specific networks are
consistent with the ROI analyses mentioned above; both indicate reduced FC involving
PCC, and other DMN regions, and in prefrontal regions associated with networks like
the FPC, but increased FC within sensorimotor and subcortical regions.
Impact of Resting FC on Behavior

An important aspect of functional connectivity at rest is that it is related to behavior
in both young and older adults. For example, stronger resting FC in the DMN is cor-
related with better scores on autobiographical memory tasks (Mevel et al., 2013), labo-
ratory-â•‰based episodic memory tasks (Wang et al., 2010a; He et al., 2012), and tests of
executive function (Andrews-â•‰Hanna et al., 2007; Duchek et al., 2013), and with more
negative self-â•‰ratings of personality traits (Figure 4.2, Saverino et al., 2015). Resting FC
in the DMN also is associated with faster reaction times on executive function tasks
(Damoiseaux et al., 2008).
An interesting exception to these positive associations is the correlation between
stronger connectivity involving medial temporal lobe regions and poorer performance
in older adults (Westlye et al., 2011; Hafkemeijer et al., 2013), particularly those who
decline in function over time (Salami et al., 2014a). These latter findings indicate that
stronger functional interactions in older adults do not necessarily benefit behavior, but
may be an indication of an adverse process. In addition, carrying out a cognitive task
has been shown to influence subsequent measures of resting FC in younger adults
(Grigg and Grady, 2010b; Stevens et al., 2010; Tambini et al., 2010) and age differ-
ences in post-â•‰task FC also have been noted (Saverino et al., 2015). Thus, intrinsic FC
at rest is dynamic, can reflect prior cognitive processing and is relevant for cognitive
performance. These characteristics, along with its utility in noncompliant populations,
account for the intense interest in resting FC in participants of all ages.
Task-â•‰Based Functional Connectivity
The literature on task-â•‰based FC has typically focused on the specific regions thought
to be important for the particular task and not on networks per se, although inferences
â•‡ 111
0.7
Negative
R2 = 0.31
0.6
0.5
0.4
Self
rating 0.3
0.2
Positive
0.1
0
–0.60 –0.40 –0.20 0.00 0.20 0.40 0.60 0.80
DMN functional connectivity
Young Older
Figure 4.2â•‡ This figure shows data from a study involving judgments about personality traits
regarding the self (Saverino et al., 2015). Shown is a scatterplot of DMN resting state FC and
the proportion of “yes” responses for self-â•‰ratings. In the graph, larger values of the rating indi-
cate more endorsements of negative traits (e.g., “negligent”), and smaller ratings indicate more
endorsements of positive traits (e.g., “likable”). Although younger adults were more likely to
endorse negative traits than older adults, there was a relation between negativity and strength
of DMN FC regardless of age.
can be made about networks based on the regions of interest. This section will focus
on the cognitive domains that have been studied most often, which can be categorized
as follows: 1) cognitive control (executive function); 2) episodic memory; 3) working
memory; and 4) socio-â•‰emotional function.
FC During Cognitive Control

Four experiments examined age differences in FC during tasks tapping into cog-
nitive control. In two of these, the FC between frontal and parietal nodes of the
FPC was stronger in younger than in older adults during task switching (Madden
et al., 2010) and a task involving detection of targets in the presence of interference
(Geerligs et al., 2014). A third study also used an interference task (Salami et al.,
2014b) and found that young adults showed stronger FC among PFC regions and
the ACC, and that activity in these regions was correlated with performance on
the task only in the young group. In contrast, older adults had stronger FC among
parietal regions involved in cognitive control and this pattern was associated with
behavior only in the older group. Thus, there were age-â•‰unique patterns of FC that
correlated differently with performance across groups, as others have reported for
task-â•‰related activation (McIntosh et al., 1999; Grady et al., 2003). The fourth study
112
examined functional coupling between the FPC and the DAN or DMN during a
visuospatial planning task that activated the DAN and an autobiographical plan-
ning task that activated the DMN (Spreng and Schacter, 2012). In young adults
the FPC was functionally coupled to the DAN only during the task that activated
the DAN, and was coupled to the DMN only during the task that activated the
DMN, suggesting that the control processes of the FPC, and its between-â•‰network
coupling, switched depending on the task demands. However, in older adults the
FPC remained coupled to the DMN in both tasks, suggesting less flexible network
interactivity in older adults, and perhaps reduced ability of the FPC to organize this
flexible switching of network activity.
FC During Episodic Memory
Most of the experiments looking at episodic memory have used seed regions in the
MTL, given the prominent role of the MTL in both memory encoding and recogni-
tion (e.g., Squire, 1992; Nadel et al., 2000). Although some have reported greater
coupling involving the MTL and other brain regions in younger vs. older adults
(Dennis et al., 2008; Dennis and Cabeza, 2011; St Jacques et al., 2012) others
have found a more complex pattern involving the FC of this region. For example,
Dew et al. (2012) used a cued memory paradigm to assess source memory and
measured FC during the cues, which indicated whether item or source memory
would be required, as well as during the correct retrieval of source information.
Young adults had stronger FC than older adults between the hippocampus and
PFC during the cues, whereas older adults had stronger FC between these regions
during the retrieval phase. This result was interpreted as an age difference in the
timing of the engagement of cognitive control processes in the service of memory
retrieval. Daselaar et al. (2006) reported stronger FC in young adults between hip-
pocampus and parietal regions, but stronger FC in older adults between perirhinal
cortex and PFC, during a word recognition task. This result is consistent with the
idea that younger adults rely more on recollection during recognition memory,
which is mediated by the hippocampus, and older adults rely more on familiarity,
which depends more on the perirhinal cortex (Yonelinas et al., 2007). A similar
result was reported in a study of encoding (Grady et al., 2003) in which younger
adults had stronger FC between hippocampus and inferior frontal cortex, whereas
older adults had stronger FC between hippocampus and DLPFC, but each age-â•‰
specific pattern was correlated with better memory in the respective groups. These
studies, taken together with the cognitive control study mentioned above (Salami
et al., 2014b), show that age differences exist in the patterns of FC involving
specific task-â•‰related regions and that these age-â•‰unique patterns of FC can sup-
port cognitive performance. Finally, one study (Wang et al., 2010b) assessed FC
among all brain voxels during encoding and recognition tasks and found age dif-
ferences in long-â•‰range functional connections, consisting of stronger FC in young
vs. older adults in PFC regions, but stronger FC in older adults among parietal
regions. Older adults also had longer path lengths in their brain networks, but this
measure was not correlated with speed of performance after accounting for age.
â•‡ 113
FC During Working Memory
Given the role of DMN deactivation in cognitive performance, several studies of
working memory have looked at DMN FC. One used an n-â•‰back task with numbers
and found that FC within the DMN was higher in young vs. older adults, as was the
negative coupling between the DMN and regions of the FPC (Sambataro et al., 2010).
Interestingly, the strength of coupling within the DMN was associated with better
performance on the working memory task across younger and older adults, an effect
also noted in younger adults in another study (Dang et al., 2013). Similarly, stronger
FC between the DMN and a number of other networks in younger than older adults
was reported during a Sternberg letter task (Steffener et al., 2012) and FC of the DMN
mediated the relation between age and RT on the task.
Other studies of working memory have focused on regions not associated with the
DMN. In a study using a delayed match-â•‰to-â•‰sample task for faces, stronger FC between
a seed in the fusiform face area (FFA) and PFC was noted in the young adults, rela-
tive to older adults (Bollinger et al., 2011). In another study from this group, FC in
the parahippocampal place area (PPA) was examined during working memory for
scenes, in which an interfering face was interposed between the stimuli to-â•‰be-â•‰remem-
bered and the memory probe (Clapp et al., 2011). Young adults showed a drop in FC
between the PPA and PFC during the interfering face, but this FC returned to its for-
mer level after the face was removed. Older adults also showed a drop in FC between
the PPA and PFC during face presentation, but FC remained low, even after the face
was removed. In addition, although both groups showed increased FC between the
FFA and PFC when viewing the interfering face, this FC declined in younger adults
when the face disappeared, but persisted in older adults during the post-â•‰interference
interval. This result suggests a greater disruption of brain function from interference
in older adults, consistent with behavioral studies (e.g., Lustig et al., 2001; Healey
et al., 2008) and other fMRI work (Campbell et al., 2012). A remaining study (Sala-â•‰
Llonch et al., 2012) looked at FC in the FPC during working memory tasks and found
an increase with age in the FC among prefrontal regions in the FPC during the task,
but only in those older adults who performed better on the task. This result differs
from the resting state data, in which younger adults typically show stronger FC in the
FPC, but is in line with many reports of over-â•‰recruitment of prefrontal regions dur-
ing cognitive tasks in older adults (e.g., Cabeza, 2002; Rajah and D’Esposito, 2005;
Grady, 2012). In addition, this latter study suggests that over-â•‰recruitment of network
connectivity can be beneficial for cognition in older adults (for a different view, see
de Chastelaine et al., 2011).
FC During Social Cognition

In the final cognitive domain, socio-â•‰emotional function, three studies have examined
FC during viewing of positively and negatively valenced visual images. All of these
found stronger FC in older adults, compared to young adults, between ventromedial
PFC and other regions, including the amygdala (Addis et al., 2010; St Jacques et al.,
2010) and ventral striatum (Ritchey et al., 2011). It is interesting that this increase
was seen for positive images, and not negative images, in two of these studies (Addis
114
et al., 2010; Ritchey et al., 2011), given the greater attention to positively valenced
information with increasing age (Mather and Carstensen, 2005) and more positive
outlook in older adults (Stone et al., 2010; Carstensen et al., 2011), sometimes called
the “positivity effect” in aging.
Two studies have examined age differences in FC during tasks involving faces,
which are highly salient social stimuli. One study found an age increase in FC between
ventromedial PFC and the right fusiform gyrus during face viewing (Burianova et al.,
2013). Of particular interest, functional coupling among these two regions and a dis-
tributed set of frontal and parietal areas was related to better face discrimination, but
only in the older group (Figure 4.3). In contrast, FC between FFA and lateral frontal
cortex was decreased with age during face processing in another study (Kalkstein
et al., 2011), suggesting that the age increase may be specific for ventromedial PFC.
A final study examined FC within the DMN and reward network during trait judg-
ments about the self and a close other, and found that older adults showed reduced FC
in both networks, compared to younger adults during these social decisions (Grady
et al., 2012). A subsequent analysis revealed that stronger FC in the DMN predicted
more negative self-judgments in younger and older adults (Saverino et al., 2015),
consistent with the reported link between DMN FC and depressive symptoms (e.g.,
Whitfield-Gabrieli and Ford, 2012; Zeng et al., 2012).
Interpreting Task-Based FC in Light of Current Theories of Cognitive

Aging
A critical question to be asked is how these age differences in task-based FC can inform
various theories of cognitive aging. The age-unique patterns for encoding (Grady et
al., 2003) and the relation between face-network activity and performance only in
older adults (Burianova et al., 2013) are consistent with the idea of compensation
(Grady, 2012) because the relations between activity and performance were specific to
the older groups and greater FC was associated with better performance. However, in
the work by Salami et al. on cognitive control (2014b), the distinct pattern of FC and
correlation with performance was seen only in low performing older adults, whereas
high performing elderly looked like young adults, a result that is not consistent with
compensation. This is similar to conflicting results found for the relation between
task-related activation and relation to performance in older adults (Grady, 2012), so
this issue remains to be settled.
Another idea proposed to explain the neural correlates of cognitive aging is dedif-
ferentiation, or less selective patterns of activity or FC across tasks in older relative to
younger adults (Carp et al., 2011; Grady et al., 2011). Some of the studies reviewed
here have found evidence for dedifferentiation. For example, the reduced anti-correla-
tions between the DMN and task regions (Sambataro et al., 2010), as well as the fail-
ure to uncouple the FPC and DMN during a task that engages the DAN (Spreng and
Schacter, 2012), support the idea that task-related FC across experimental conditions
is less distinctive in older adults. The study by Chan et al. (2014) showing greater
between-network connectivity with age is also consistent with the idea of dedifferen-
tiation and shows that this can also occur at rest.
115
(A)
(B) 1.00
0.80
0.60
0.40
Correlations
0.20
Older
0.00
Young
–0.20
–0.40
–0.60
–0.80
–1.00
RFG LFG medOFC RT ACC
Figure 4.3 (A) The colored brain regions represent a set of areas active in young and older adults during a face-matching task. (B) The graph shows
the correlations between activity in the regions in (a) and activity in right fusiform, left fusiform, medial orbitofrontal cortex (medOFC). Correlations
between activity in the regions seen in (a) and behavior also are shown (reaction time, RT; accuracy, acc). Young adults show correlations between right
and left fusiform and the regions seen in (a), but not with medOFC or behavior. In contrast, older adults show reliable positive correlations between
the pattern of activity, right fusiform activity, and accuracy on the task, indicating age differences in FC as well as how the FC pattern relates to perfor-
mance. Error bars denote 95% confidence intervals for the correlations. Data are from Burianova et al. (2013). (See color plate also)
116
Finally, Nyberg et al. (2012) have suggested that much of the aging literature is
consistent with the idea that those older adults who age more successfully are those
who also maintain patterns of brain activity similar to those seen in younger adults.
The recent results of Salami et al. (Salami et al., 2014a; Salami et al., 2014b) are con-
sistent with this maintenance theory. Clearly, further study of task-â•‰based FC will be
needed to determine which of these proposed mechanisms of aging best accounts for
the link between FC and behavior in older adults.
Common Themes Across Resting and Task Studies
Although the approaches taken in the studies of resting and task-â•‰based FC are rather
different, some common themes have emerged. Figure 4.4 summarizes the studies, both
task and resting, in which the age differences can be attributed clearly to regions consis-
tent with a specific network, or to subcortical regions (such as the striatum and thalamus).
This figure shows the strong evidence for age reductions in the DMN, although this is
likely due to some extent to the popularity of this network in the FC literature in general.
This trend for age reductions also is seen in the FPC, SLN, visual network, and DAN,
although fewer studies looking at these networks are available currently. Interestingly,
for the sensorimotor network and subcortical regions the trend is reversed, with older
adults showing stronger FC in these regions in the majority of papers. In particular the
subcortical regions (i.e., putamen, caudate, cerebellum, and thalamus) seem to show
the largest effect of greater FC in older adults. The reasons for this are unknown, but 6
of the 7 studies reporting stronger subcortical FC with age are resting state studies that
make use of lifespan samples (from childhood or young adulthood to older adults) and it
may be that this age effect is relatively small in magnitude but can be detected with large
samples and wide age ranges. These trends are interesting, although more work needs to
be done, particularly for networks other than the DMN.
Reductions in FC with aging are not surprising, especially those found in networks
involving cognitive control, considering that there are numerous reports of reduced
cognitive control in older adults (e.g., Hasher et al., 1999; West et al., 2002; De Luca
et al., 2003). The age reduction in FC of the DMN is intriguing from two points of
view. First, as noted above, older adults tend to show a “positivity effect” in emotional
processing, and greater emotional negativity has been associated with stronger FC in
the DMN in both healthy adults (Saverino et al., 2015) and in people with depression
(Andreescu et al., 2011; Berman et al., 2011). Second, there is evidence that stronger
DMN FC is associated with better performance on cognitive tasks in both young (Kelly
et al., 2008; Dang et al., 2013) and older adults (Andrews-â•‰Hanna et al., 2007). These
two lines of evidence suggest that weaker DMN FC in older adults may underlie emo-
tional function in late life, perhaps even aiding the improvement noted in emotional
experience with aging (Carstensen et al., 2011), but may also have a more deleterious
effect on other aspects of cognitive function. The connections between cognitive func-
tion and the increases in FC with age seen in sensorimotor and subcortical regions are
less clear, and have yet to be systematically investigated. However, it is possible that
maintained or relatively higher FC in these areas is related to the less marked struc-
tural changes seen over time in the volume of sensory and subcortical regions in older
â•‡ 117
25
20
Number of papers
15
10
0
DMN FPC DAN SLN Visual SM SubC
FigureÂ€4.4â•‡ Summary figure showing the number of studies reporting age differences in regions
consistent with a set of identified brain networks, including both task and resting state studies.
Subcortical regions have been included even if a distinct network involving these regions is
not specified in the paper. Studies cited: DMN, default mode network (Andrews-â•‰Hanna et al.,
2007; Damoiseaux et al., 2008; Grady et al., 2010; Sambataro et al., 2010; Allen et al., 2011;
Wu et al., 2011; Grady et al., 2012; Hampson et al., 2012; Meier et al., 2012; Mowinckel et al.,
2012; Onoda et al., 2012; Sala-â•‰Llonch et al., 2012; Tomasi and Volkow, 2012; Campbell
et al., 2013; Mevel et al., 2013; Chan et al., 2014; Madhyastha and Grabowski, 2014; Salami
et al., 2014a; Schultz et al., 2014; Song et al., 2014; Geerligs et al., in press; Saverino et al., in
press); FPC, frontoparietal control network (Grady et al., 2010; Allen et al., 2011; Campbell
et al., 2012; Sala-â•‰Llonch et al., 2012; Chan et al., 2014; Geerligs et al., 2014; Madhyastha and
Grabowski, 2014; Schultz et al., 2014; Geerligs et al., in press); DAN, dorsal attention net-
work (Andrews-â•‰Hanna et al., 2007; Grady et al., 2010; Allen et al., 2011; Tomasi and Volkow,
2012; Madhyastha and Grabowski, 2014; Schultz et al., 2014); SLN, salience network (Allen
et al., 2011; Meier et al., 2012; Onoda et al., 2012; Campbell et al., 2013; He et al., 2013,
2014; Geerligs et al., in press); Visual, networks including lingual and fusiform gyri (Hampson
et al., 2012; Mowinckel et al., 2012; Onoda et al., 2012; Meinzer et al., 2013; Chan et al., 2014;
Schultz et al., 2014); SM, sensorimotor networks, involving pre-â•‰and post-â•‰central gyri (Meier
et al., 2012; Mowinckel et al., 2012; Tomasi and Volkow, 2012; Onoda and Yamaguchi, 2013;
Cao et al., 2014; Geerligs et al., 2014; Song et al., 2014); SubC, subcortical regions, including
putamen, thalamus, caudate and cerebellum (Achard and Bullmore, 2007; Allen et al., 2011;
Hampson et al., 2012; Tomasi and Volkow, 2012; Wang et al., 2012; Zuo et al., 2012; Chou et al.,
2013; Meinzer et al., 2013; Cao et al., 2014).
adults, compared to those seen in frontal and parietal cortices (Kemper, 1984; Resnick
et al., 2003; Raz et al., 2010).
Factors Influencing Functional Connectivity in Older Adults
As with any measure of brain activity, a number of factors can influence FC in older
adults. In line with evidence that a higher level of education is associated with slower
118
cognitive decline (Alexander et al., 1997; Stern et al., 1999; Bennett et al., 2003),
intrinsic DMN FC is stronger in more highly educated older adults, relative to their
less educated peers (Arenaza-Urquijo et al., 2013). Similar results have been found for
lifelong bilingualism, which protects against AD symptoms and is associated with bet-
ter cognitive control in older adults, relative to monolinguals (Bialystok et al., 2014).
Bilingual older adults also have stronger FC in the DMN and FPC (Grady et al., 2015),
which may be facilitated by better maintained white matter connections between the
hemispheres and in anterior–posterior white matter tracts (Luk et al., 2011). Both edu-
cation and language experience may act as a type of cognitive reserve (Stern, 2002),
with stronger FC as a potential brain mechanism underlying this reserve.
Not surprisingly, brain changes associated with Alzheimer disease (AD) also affect
resting FC. Higher amyloid burden in the brains of otherwise healthy elderly indi-
viduals has been associated with disrupted FC within the DMN (Hedden et al., 2009;
Sheline et al., 2010b; Mormino et al., 2011; Elman et al., 2016) and a similar reduced
FC has been shown in those who are APOE4 carriers and thus at higher risk for AD
(Sheline et al., 2010a; Machulda et al., 2011). Once clinical decline has been identi-
fied, there is further disruption of DMN FC in AD (Greicius et al., 2004; Buckner et
al., 2005; Lehmann et al., 2013) and decline in FC of the DMN also is a good pre-
dictor of cognitive decline in those individuals with mild cognitive impairment who
progress to dementia (Petrella et al., 2011).
Brain structural factors can influence FC in general (Greicius et al., 2009; Honey et
al., 2009; van den Heuvel and Sporns, 2013), and, as there are marked declines with age
in many brain regions (Resnick et al., 2003; Raz et al., 2010), structure is likely to influ-
ence both network FC and cognitive performance in older adults. A number of studies
have found this to be the case (Teipel et al. 2010; Davis et al., 2009; Madden et al., 2009;
Burgmans et al., 2011; Davis et al., 2012; Fling et al., 2012). Interestingly, one recent
study of a sample of adults aged 65–90 years found that FC of the DMN and FPC did
not add significantly to the ability to predict cognitive function when structural mea-
sures were included in the model (Hedden et al., 2016). Although only two networks
were examined, this finding raises several issues. First, the inclusion only of older adults
raises the possibility that FC measures might be more predictive of cognitive perfor-
mance in a sample that also included younger and middle-aged adults. Second, it may be
that other measures of network function, besides FC per se, might be more sensitive to
age differences. For example, Chan et al. (2014) found a relation between network seg-
regation (a measure of within vs. between-network connections) and episodic memory
performance, after accounting for age. Thus, measures assessing the balance between
intra-and inter-network connectivity also can predict behavior, and may be more sensi-
tive than measures of FC within specific networks, although future studies will need to
compare these sensitivities directly. Finally, the study by Hedden et al. (2016) raises the
question of which brain markers, whether structural or functional, will ultimately be the
best predictors of behavior and the best targets for intervention in older adults.
In addition to the factors listed above, fundamental changes in neural function
also may influence FC in older adults. For example, age differences have been identi-
fied in dopamine D1 receptor density, which in turn influences FC between prefron-
tal and parietal regions (Rieckmann et al., 2011). In addition, polymorphisms in the
catechol O-methyltransferase gene, which is involved in the metabolic degradation
â•‡ 119
of dopamine, influence cognitive decline in older adults (de Frias et al., 2005), and
modulate FC between PFC and parietal regions, an effect that is enhanced in older
adults (Sambataro et al., 2009). Basic physiological processes, such as sleep (Mander
et al., 2013) and cardiovascular fitness (Voss et al., 2010a), also can influence FC.
Lastly, there is recent evidence of age reductions in the within-â•‰person variability of
the BOLD signal (Garrett et al., 2010, 2011). Interestingly, although younger adults
have greater variability in many cortical regions, older adults have more variability
in striatum and cerebellum, mirroring the age increases that have been found in rest-
ing FC of subcortical areas, suggesting that neural variability may underlie some age
differences in FC.
The Future
One future goal for this field should be to better understand how age differences in FC
might shift between rest and tasks. Several papers have shown that carrying out tasks
can result in FC within brain networks that differs somewhat from that seen at rest
(Mennes et al., 2013; Spreng et al., 2013; Cole et al., 2014), but how the shift from rest
to task influences age differences in FC is far from clear. In addition, if the FPC acts as a
switch that engages and disengages other networks, then this switch may be particularly
impacted by aging, and its inter-â•‰network FC may influence multiple cognitive domains
in older adults. There is some evidence that this is the case (Spreng and Schacter, 2012;
Grady et al., 2016), but further experiments are necessary to fully understand age differ-
ences in the interplay between networks during cognitive processing.
In addition to exploring the impact of various factors on FC in older adults,
it will also be important to study the role of interventions on FC in older adults.
A few studies have taken this approach with interesting results. One study utilized
several types of cognitive training procedures as interventions in an older sample
and found that they influenced FC involving superior parietal cortex, a node of the
DAN (Strenziok et al., 2014). Another study found that physical fitness training
increased FC within both the DMN and FPC, and that increased FC was associated
with better performance on cognitive tasks (Voss et al., 2010b). These results are
promising and it would be interesting to see if one could improve cognition in older
adults using interventions that targeted FC within or between specific networks.
One possibility in this regard would be the use of transcranial magnetic stimula-
tion, which was recently used to enhance FC within the FPC via stimulation of PFC
nodes of this network (Chen et al., 2013).
Conclusion
The study of FC in aging appears to be growing rapidly both in terms of resting and
task-â•‰based FC, but has been somewhat hampered by the use of differing approaches
and a focus on individual regions rather than networks per se. The study of resting
FC has focused mainly on specific brain networks, whereas work on task-â•‰based FC
has focused on disparate regions, making it difficult to synthesize the results across
120
these fields. Nevertheless some trends are apparent. Younger adults consistently show
stronger FC in the DMN compared to older adults, and often have stronger FC in
other cortical networks as well, although not as consistently. Older adults, on the other
hand, are more likely to have stronger FC in sensorimotor or subcortical networks.
Some of this age difference could be due to patterns of atrophy with age, which are
less pronounced in sensorimotor and subcortical areas, but also is undoubtedly the
result of as yet unknown functional changes in the systems involved. Future work on
how FC within and between networks changes with age should be very useful for fur-
thering our understanding of cognition in older age, both in terms of those processes
that decline as well as those that are maintained.
References
Achard S, Bullmore E (2007) Efficiency and cost of economical brain functional networks.
PLoS Comput Biol 3:174–â•‰183.
Addis DR, Leclerc CM, Muscatell KA, Kensinger EA (2010) There are age-â•‰related changes in
neural connectivity during the encoding of positive, but not negative, information. Cortex
46:425–â•‰433.
Afshin-â•‰Pour B, Grady C, Strother S (2014) Evaluation of spatio-â•‰temporal decomposition tech-
niques for group analysis of fMRI resting state data sets. Neuroimage 87:363–â•‰382.
Alexander GE, Furey ML, Grady CL, Pietrini P, Brady DR, Mentis MJ, Schapiro MB (1997)
Association of premorbid intellectual function with cerebral metabolism in Alzheimer’s
disease: implications for the cognitive reserve hypothesis. American Journal of Psychiatry
154:165–â•‰172.
Allen EA, Erhardt EB, Damaraju E, Gruner W, Segall JM, Silva RF, Havlicek M, Rachakonda
S, Fries J, Kalyanam R, Michael AM, Caprihan A, Turner JA, Eichele T, Adelsheim S,
Bryan AD, Bustillo J, Clark VP, Feldstein Ewing SW, Filbey F, Ford CC, Hutchison K,
Jung RE, Kiehl KA, Kodituwakku P, Komesu YM, Mayer AR, Pearlson GD, Phillips JP,
Sadek JR, Stevens M, Teuscher U, Thoma RJ, Calhoun VD (2011) A baseline for the mul-
tivariate comparison of resting-â•‰state networks. Front Syst Neurosci 5:2.
Andreescu C, Wu M, Butters MA, Figurski J, Reynolds CF, 3rd, Aizenstein HJ (2011) The
default mode network in late-â•‰life anxious depression. Am J Geriatr Psychiatry 19:980–â•‰983.
Andrews-â•‰Hanna JR, Reidler JS, Sepulcre J, Poulin R, Buckner RL (2010) Functional-â•‰anatomic
fractionation of the brain’s default network. Neuron 65:550–â•‰562.
Andrews-â•‰Hanna JR, Smallwood J, Spreng RN (2014) The default network and self-â•‰generated
thought: component processes, dynamic control, and clinical relevance. Ann N Y Acad Sci
1316:29–â•‰52.
Andrews-â•‰Hanna JR, Snyder AZ, Vincent JL, Lustig C, Head D, Raichle ME, Buckner RL
(2007) Disruption of large-â•‰scale brain systems in advanced aging. Neuron 56:924–â•‰935.
Arenaza-â•‰Urquijo EM, Landeau B, La Joie R, Mevel K, Mezenge F, Perrotin A, Desgranges
B, Bartres-â•‰Faz D, Eustache F, Chetelat G (2013) Relationships between years of educa-
tion and gray matter volume, metabolism and functional connectivity in healthy elders.
Bennett DA, Wilson RS, Schneider JA, Evans DA, Mendes de Leon CF, Arnold SE, Barnes LL,
Bienias JL (2003) Education modifies the relation of AD pathology to level of cognitive
function in older persons. Neurology 60:1909–â•‰1915.
Berman MG, Peltier S, Nee DE, Kross E, Deldin PJ, Jonides J (2011) Depression, rumination
and the default network. Soc Cogn Affect Neurosci 6:548–â•‰555.
121
Bialystok E, Craik FIM, Binns MA, Ossher L, Freedman M (2014) Effects of bilingualism on
the age of onset and progression of MCI and AD: Evidence from executive function tests.
Biswal B, Yetkin FZ, Haughton VM, Hyde JS (1995) Functional connectivity in the motor
cortex of resting human brain using echo-planar MRI. Magnetic Resonance in Medicine
34:537–541.
Bollinger J, Rubens MT, Masangkay E, Kalkstein J, Gazzaley A (2011) An expectation-based
memory deficit in aging. Neuropsychologia 49:1466–1475.
Buckner RL, Andrews-Hanna JR, Schacter DL (2008) The brain’s default network: anatomy,
function, and relevance to disease. Ann N Y Acad Sci 1124:1–38.
Buckner RL, Carroll DC (2007) Self-projection and the brain. Trends in Cognitive Science
11:49–57.
Buckner RL, Snyder AZ, Shannon BJ, LaRossa G, Sachs R, Fotenos AF, Sheline YI, Klunk
WE, Mathis CA, Morris JC, Mintun MA (2005) Molecular, structural, and functional char-
acterization of Alzheimer’s disease: evidence for a relationship between default activity,
amyloid, and memory. J Neurosci 25:7709–7717.
Bullmore E, Sporns O (2009) Complex brain networks: graph theoretical analysis of structural
and functional systems. Nat Rev Neurosci 10:186–198.
by differences in axonal integrity. Neuroimage 55:1287–1297.
Burianova H, Lee Y, Grady CL, Moscovitch M (2013) Age-related dedifferentiation and com-
pensatory changes in the functional network underlying face processing. Neurobiol Aging
34:2759–2767.
Cabeza R (2002) Hemispheric asymmetry reduction in older adults: The HAROLD model.
Campbell KL, Grady CL, Ng C, Hasher L (2012) Age differences in the frontoparietal cognitive
control network: Implications for distractibility. Neuropsychologia 50:2212–2223.
Campbell KL, Grigg O, Saverino C, Churchill N, Grady CL (2013) Age differences in the
intrinsic functional connectivity of default network subsystems. Front Aging Neurosci
5:Article 73.
Cao M, Wang JH, Dai ZJ, Cao XY, Jiang LL, Fan FM, Song XW, Xia MR, Shu N, Dong
Q, Milham MP, Castellanos FX, Zuo XN, He Y (2014) Topological organization of the
human brain functional connectome across the lifespan. Dev Cogn Neurosci 7:76–93.
Carp J, Park J, Polk TA, Park DC (2011) Age differences in neural distinctiveness revealed by
multi-voxel pattern analysis. Neuroimage 56:736–743.
Carstensen LL, Turan B, Scheibe S, Ram N, Ersner-Hershfield H, Samanez-Larkin GR, Brooks
KP, Nesselroade JR (2011) Emotional experience improves with age: evidence based on
over 10 years of experience sampling. Psychol Aging 26:21–33.
Chan MY, Park DC, Savalia NK, Petersen SE, Wig GS (2014) Decreased segregation of brain
systems across the healthy adult lifespan. Proc Natl Acad Sci U S A 111:E4997–5006.
Chen AC, Oathes DJ, Chang C, Bradley T, Zhou ZW, Williams LM, Glover GH, Deisseroth K,
Etkin A (2013) Causal interactions between fronto-parietal central executive and default-
mode networks in humans. Proc Natl Acad Sci U S A 110:19944–19949.
Chou YH, Chen NK, Madden DJ (2013) Functional brain connectivity and cognition: effects of
adult age and task demands. Neurobiol Aging 34:1925–1934.
Clapp WC, Rubens MT, Sabharwal J, Gazzaley A (2011) Deficit in switching between func-
tional brain networks underlies the impact of multitasking on working memory in older
adults. Proc Natl Acad Sci U S A 108:7212–7217.
122
Cole MW, Bassett DS, Power JD, Braver TS, Petersen SE (2014) Intrinsic and task-evoked
network architectures of the human brain. Neuron 83:238–251.
Cole MW, Reynolds JR, Power JD, Repovs G, Anticevic A, Braver TS (2013) Multi-task con-
nectivity reveals flexible hubs for adaptive task control. Nat Neurosci 16:1348–1355.
Corbetta M, Patel G, Shulman GL (2008) The reorienting system of the human brain: from
environment to theory of mind. Neuron 58:306–324.
Damoiseaux JS, Beckmann CF, Sanz Arigita EJ, Barkhof F, Scheltens P, Stam CJ, Smith SM,
Rombouts SA (2008) Reduced resting-state brain activity in the “default network” in nor-
mal aging. Cereb Cortex 18:1856–1864.
Damoiseaux JS, Greicius MD (2009) Greater than the sum of its parts: a review of studies
combining structural connectivity and resting-state functional connectivity. Brain Struct
Funct 213:525–533.
Dang LC, O’Neil JP, Jagust WJ (2013) Genetic effects on behavior are mediated by neurotrans-
mitters and large-scale neural networks. Neuroimage 66:203–214.
Daselaar SM, Fleck MS, Dobbins IG, Madden DJ, Cabeza R (2006) Effects of Healthy Aging
on Hippocampal and Rhinal Memory Functions: An Event-Related fMRI Study. Cereb
Cortex 16:1771–1782.
Davis SW, Dennis NA, Buchler NG, White LE, Madden DJ, Cabeza R (2009) Assessing the
effects of age on long white matter tracts using diffusion tensor tractography. Neuroimage
46:530–541.
communication in the aging brain: linking behavior to functional and structural connectiv-
ity. Cereb Cortex 22:232–242.
de Chastelaine M, Wang TH, Minton B, Muftuler LT, Rugg MD (2011) The effects of age,
memory performance, and callosal integrity on the neural correlates of successful associa-
tive encoding. Cereb Cortex 21:2166–2176.
de Frias CM, Annerbrink K, Westberg L, Eriksson E, Adolfsson R, Nilsson LG (2005) Catechol
O-methyltransferase Val158Met polymorphism is associated with cognitive performance
in nondemented adults. J Cogn Neurosci 17:1018–1025.
De Luca CR, Wood SJ, Anderson V, Buchanan JA, Proffitt TM, Mahony K, Pantelis C (2003)
Normative data from the CANTAB. I: development of executive function over the lifes-
pan. J Clin Exp Neuropsychol 25:242–254.
Deco G, Jirsa VK, McIntosh AR (2011) Emerging concepts for the dynamical organization of
resting-state activity in the brain. Nat Rev Neurosci 12:43–56.
Dennis NA, Cabeza R (2011) Age-related dedifferentiation of learning systems: an fMRI study
of implicit and explicit learning. Neurobiol Aging 32:2318 e2317–e2330.
Dennis NA, Hayes SM, Prince SE, Madden DJ, Huettel SA, Cabeza R (2008) Effects of aging
on the neural correlates of successful item and source memory encoding. J Exp Psychol
Learn Mem Cogn 34:791–808.
Dew IT, Buchler N, Dobbins IG, Cabeza R (2012) Where is ELSA? The early to late shift in
aging. Cereb Cortex 22:2542–2553.
Dosenbach NU, Fair DA, Miezin FM, Cohen AL, Wenger KK, Dosenbach RA, Fox MD,
Snyder AZ, Vincent JL, Raichle ME, Schlaggar BL, Petersen SE (2007) Distinct brain
networks for adaptive and stable task control in humans. Proc Natl Acad Sci U S A
104:11073–11078.
Duchek JM, Balota DA, Thomas JB, Snyder AZ, Rich P, Benzinger TL, Fagan AM, Holtzman
DM, Morris JC, Ances BM (2013) Relationship between Stroop performance and rest-
ing state functional connectivity in cognitively normal older adults. Neuropsychology
27:516–528.
123
Elman JA, Madison CM, Baker SL, Vogel JW, Marks SM, Crowley S, O’Neil JP, Jagust WJ
(2016) Effects of Beta-Amyloid on Resting State Functional Connectivity Within and
Between Networks Reflect Known Patterns of Regional Vulnerability. Cereb Cortex
26:695–707.
Ferreira LK, Busatto GF (2013) Resting-state functional connectivity in normal brain aging.
Neurosci Biobehav Rev 37:384–400.
Fling BW, Kwak Y, Peltier SJ, Seidler RD (2012) Differential relationships between transcal-
losal structural and functional connectivity in young and older adults. Neurobiol Aging
33:2521–2526.
Fox MD, Corbetta M, Snyder AZ, Vincent JL, Raichle ME (2006) Spontaneous neuronal activ-
ity distinguishes human dorsal and ventral attention systems. Proc Natl Acad Sci U S A
103:10046–10051.
Fox MD, Greicius M (2010) Clinical applications of resting state functional connectivity. Front
Syst Neurosci 4:19.
Fox MD, Snyder AZ, Vincent JL, Corbetta M, Van Essen DC, Raichle ME (2005) The human
brain is intrinsically organized into dynamic, anticorrelated functional networks. Proc Nat
Acad Sci 102:9673–9678.
Garrett DD, Kovacevic N, McIntosh AR, Grady CL (2010) Blood oxygen level-dependent sig-
nal variability is more than just noise. J Neurosci 30:4914–4921.
Garrett DD, Kovacevic N, McIntosh AR, Grady CL (2011) The importance of being variable.
J Neurosci 31:4496–4503.
Geerligs L, Renken RJ, Saliasi E, Maurits NM, Lorist MM (2015) A Brain-Wide Study of Age-
Related Changes in Functional Connectivity. Cereb Cortex 25:1987–1999.
Geerligs L, Saliasi E, Maurits NM, Renken RJ, Lorist MM (2014) Brain mechanisms
underlying the effects of aging on different aspects of selective attention. Neuroimage
91:52–62.
Grady C (2012) The cognitive neuroscience of ageing. Nat Rev Neurosci 13:491–505.
Grady CL, Charlton R, He Y, Alain C (2011) Age differences in FMRI adaptation for sound
identity and location. Front Hum Neurosci 5:24.
Grady CL, Grigg O, Ng C (2012) Age differences in default and reward networks during pro-
cessing of personally relevant information. Neuropsychologia 50:1682–1697.
Grady CL, Luk G, Craik FI, Bialystok E (2015) Brain network activity in monolingual and
bilingual older adults. Neuropsychologia 66:170–181.
Grady CL, McIntosh AR, Craik FI (2003) Age-Related differences in the functional connectiv-
ity of the hippocampus during memory encoding. Hippocampus 13:572–586.
Grady CL, Protzner AB, Kovacevic N, Strother SC, Afshin-Pour B, Wojtowicz M, Anderson
JA, Churchill N, McIntosh AR (2010) A multivariate analysis of age-related differences in
default mode and task-positive networks across multiple cognitive domains. Cereb Cortex
20:1432–1447.
Grady C, Sarraf S, Saverino C, Campbell K (2016) Age differences in the functional interac-
tions among the default, frontoparietal control, and dorsal attention networks. Neurobiol
Aging 41:159–172.
Greicius M (2008) Resting-state functional connectivity in neuropsychiatric disorders. Curr
Opin Neurol 21:424–430.
Greicius MD, Srivastava G, Reiss AL, Menon V (2004) Default-mode network activity distin-
guishes Alzheimer’s disease from healthy aging: evidence from functional MRI. Proc Natl
Acad Sci U S A 101:4637–4642.
Greicius MD, Supekar K, Menon V, Dougherty RF (2009) Resting-state functional connectiv-
ity reflects structural connectivity in the default mode network. Cereb Cortex 19:72–78.
124
Grigg O, Grady CL (2010a) The default network and processing of personally relevant infor-
mation: Converging evidence from task-related modulations and functional connectivity.
Grigg O, Grady CL (2010b) Task-related effects on the temporal and spatial dynamics of
resting-state functional connectivity in the default network. PLoS One 5:e13311.
Gusnard DA, Akbudak E, Shulman GL, Raichle ME (2001) Medial prefrontal cortex and self-
referential mental activity: Relation to a default mode of brain function. Proc Nat Acad Sci
USA 98:4259–4264.
Habeck C, Steffener J, Rakitin B, Stern Y (2012) Can the default-mode network be described
with one spatial-covariance network? Brain Research 1468:38–51.
Hafkemeijer A, Altmann-Schneider I, Oleksik AM, van de Wiel L, Middelkoop HAM, van
Buchem MA, van der Grond J, Rombouts SARB (2013) Increased Functional Connectivity
and Brain Atrophy in Elderly with Subjective Memory Complaints. 3:353–362.
Ham TE, de Boissezon X, Leff A, Beckmann C, Hughes E, Kinnunen KM, Leech R, Sharp
DJ (2013) Distinct frontal networks are involved in adapting to internally and externally
signaled errors. Cereb Cortex 23:703–713.
Hampson M, Tokoglu F, Shen X, Scheinost D, Papademetris X, Constable RT (2012)
Intrinsic brain connectivity related to age in young and middle aged adults. PLoS One
7:e44067.
Harrison BJ, Pujol J, Lopez-Sola M, Hernandez-Ribas R, Deus J, Ortiz H, Soriano-Mas C,
Yucel M, Pantelis C, Cardoner N (2008) Consistency and functional specialization in the
default mode brain network. Proc Natl Acad Sci U S A 105:9781–9786.
Hasher L, Zacks RT, May CP (1999) Inhibitory control, circadian arousal, and age. In: Attention
and Performance XVII: Cognitive regulation of performance Interaction of theory and
application(Gopher, D. and Koriat, A., eds), pp 653–675 Cambridge, MA: MIT Press.
He J, Carmichael O, Fletcher E, Singh B, Iosif AM, Martinez O, Reed B, Yonelinas A, Decarli
C (2012) Influence of functional connectivity and structural MRI measures on episodic
memory. Neurobiol Aging 33:2612–2620.
He X, Qin W, Liu Y, Zhang X, Duan Y, Song J, Li K, Jiang T, Yu C (2013) Age-related decrease
in functional connectivity of the right fronto-insular cortex with the central executive and
default-mode networks in adults from young to middle age. Neurosci Lett 544:74–79.
He X, Qin W, Liu Y, Zhang X, Duan Y, Song J, Li K, Jiang T, Yu C (2014) Abnormal salience
network in normal aging and in amnestic mild cognitive impairment and Alzheimer’s dis-
ease. Hum Brain Mapp 35:3446–3464.
Healey MK, Campbell KL, Hasher L (2008) Cognitive aging and increased distractibil-
ity: Costs and potential benefits. In: Progress in Brain Research Vol 169: The Essence of
Memory(Sossin, W. et al., eds) Amsterdam: Elsevier.
(2016) Multiple Brain Markers are Linked to Age-Related Variation in Cognition. Cereb
Cortex 26:1388–1400.
Hedden T, Van Dijk KR, Becker JA, Mehta A, Sperling RA, Johnson KA, Buckner RL (2009)
burden. J Neurosci 29:12686–12694.
Honey CJ, Sporns O, Cammoun L, Gigandet X, Thiran JP, Meuli R, Hagmann P (2009)
Predicting human resting- state functional connectivity from structural connectivity.
106:2035–2040.
Hutchison RM, Womelsdorf T, Allen EA, Bandettini PA, Calhoun VD, Corbetta M, Della
Penna S, Duyn J, Glover G, Gonzalez-Castillo J, Handwerker DA, Keilholz S, Kiviniemi
V, Leopold DA, de Pasquale F, Sporns O, Walter M, Chang C (2013) Dynamic functional
connectivity: Promises, issues, and interpretations. Neuroimage 80:360–378.
125
Kalkstein J, Checksfield K, Bollinger J, Gazzaley A (2011) Diminished top-down control

underlies a visual imagery deficit in normal aging. J Neurosci 31:15768–15774.
Keilholz SD (2014) The Neural Basis of Time-Varying Resting-State Functional Connectivity.
Brain Connect 4:769–779.
Kelly AM, Uddin LQ, Biswal BB, Castellanos FX, Milham MP (2008) Competition between
functional brain networks mediates behavioral variability. Neuroimage 39:527–537.
Kemper T (1984) Neuroanatomical and neuropathological changes in normal aging and in
dementia. In: Clinical Neurology of Aging(Albert, M., ed), pp 9–52 New York: Oxford
Univ. Press.
Laird AR, Fox PM, Eickhoff SB, Turner JA, Ray KL, McKay DR, Glahn DC, Beckmann CF,
Smith SM, Fox PT (2011) Behavioral interpretations of intrinsic connectivity networks. J
Cogn Neurosci 23:4022–4037.
Leech R, Kamourieh S, Beckmann CF, Sharp DJ (2011) Fractionating the default mode net-
work: distinct contributions of the ventral and dorsal posterior cingulate cortex to cogni-
tive control. J Neurosci 31:3217–3224.
Lehmann M, Madison CM, Ghosh PM, Seeley WW, Mormino E, Greicius MD, Gorno-Tempini
ML, Kramer JH, Miller BL, Jagust WJ, Rabinovici GD (2013) Intrinsic connectivity net-
works in healthy subjects explain clinical variability in Alzheimer’s disease. Proc Natl
Acad Sci U S A 110:11606–11611.
Luk G, Bialystok E, Craik F, Grady C (2011) Lifelong bilingualism maintains white matter
integrity in older adults. J Neurosci 31 16808–16813.
Luks TL, Simpson GV, Feiwell RJ, Miller WL (2002) Evidence for anterior cingulate cortex
involvement in monitoring preparatory attentional set. Neuroimage 17:792–802.
Lustig C, May CP, Hasher L (2001) Working memory span and the role of proactive interfer-
ence. J Exp Psychol Gen 130:199–207.
Machulda MM, Jones DT, Vemuri P, McDade E, Avula R, Przybelski S, Boeve BF, Knopman
DS, Petersen RC, Jack CR, Jr. (2011) Effect of APOE epsilon4 status on intrinsic network
connectivity in cognitively normal elderly subjects. Arch Neurol 68:1131–1136.
Madden DJ, Costello MC, Dennis NA, Davis SW, Shepler AM, Spaniol J, Bucur B, Cabeza
R (2010) Adult age differences in functional connectivity during executive control.
Madden DJ, Spaniol J, Costello MC, Bucur B, White LE, Cabeza R, Davis SW, Dennis NA,
Provenzale JM, Huettel SA (2009) Cerebral white matter integrity mediates adult age dif-
ferences in cognitive performance. J Cogn Neurosci 21:289–302.
Madhyastha TM, Grabowski TJ (2014) Age-related differences in the dynamic architecture of
intrinsic networks. Brain Connect 4:231–241.
Mander BA, Rao V, Lu B, Saletin JM, Lindquist JR, Ancoli-Israel S, Jagust W, Walker MP
(2013) Prefrontal atrophy, disrupted NREM slow waves and impaired hippocampal-
dependent memory in aging. Nat Neurosci 16:357–364.
Mather M, Carstensen LL (2005) Aging and motivated cognition: the positivity effect in atten-
tion and memory. Trends Cogn Sci 9:496–502.
McIntosh AR, Sekuler AB, Penpeci C, Rajah MN, Grady CL, Sekuler R, Bennett PJ (1999)
Recruitment of unique neural systems to support visual memory in normal aging. Current
Biology 9:1275–1278.
Meier TB, Desphande AS, Vergun S, Nair VA, Song J, Biswal BB, Meyerand ME, Birn RM,
Prabhakaran V (2012) Support vector machine classification and characterization of age-
related reorganization of functional brain networks. Neuroimage 60:601–613.
Meinzer M, Lindenberg R, Antonenko D, Flaisch T, Floel A (2013) Anodal transcranial direct
current stimulation temporarily reverses age-associated cognitive decline and functional
brain activity changes. J Neurosci 33:12470–12478.
126
Mennes M, Kelly C, Colcombe S, Castellanos FX, Milham MP (2013) The extrinsic and
intrinsic functional architectures of the human brain are not equivalent. Cereb Cortex
23:223–229.
Meunier D, Achard S, Morcom A, Bullmore E (2009) Age-related changes in modular organi-
zation of human brain functional networks. Neuroimage 44:715–723.
Mevel K, Landeau B, Fouquet M, La Joie R, Villain N, Mezenge F, Perrotin A, Eustache F,
Desgranges B, Chetelat G (2013) Age effect on the default mode network, inner thoughts,
and cognitive abilities. Neurobiol Aging 34:1292–1301.
Mormino EC, Smiljic A, Hayenga AO, Onami SH, Greicius MD, Rabinovici GD, Janabi M,
Baker SL, Yen IV, Madison CM, Miller BL, Jagust WJ (2011) Relationships between beta-
amyloid and functional connectivity in different components of the default mode network
in aging. Cereb Cortex 21:2399–2407.
Mowinckel AM, Espeseth T, Westlye LT (2012) Network-specific effects of age and in-
scanner subject motion: a resting-state fMRI study of 238 healthy adults. Neuroimage
63:1364–1373.
Nadel L, Samsonovich A, Ryan L, Moscovitch M (2000) Multiple trace theory of human
memory: computational, neuroimaging, and neuropsychological results. Hippocampus
10:352–368.
Onoda K, Ishihara M, Yamaguchi S (2012) Decreased functional connectivity by aging is asso-
ciated with cognitive decline. J Cogn Neurosci 24:2186–2198.
Onoda K, Yamaguchi S (2013) Small-worldness and modularity of the resting-state functional
brain network decrease with aging. Neurosci Lett 556:104–108.
Owen AM, McMillan KM, Laird AR, Bullmore E (2005) N-back working memory paradigm: a
meta-analysis of normative functional neuroimaging studies. Hum Brain Mapp 25:46–59.
Petrella JR, Sheldon FC, Prince SE, Calhoun VD, Doraiswamy PM (2011) Default mode
network connectivity in stable vs progressive mild cognitive impairment. Neurology
76:511–517.
Poppenk J, Evensmoen HR, Moscovitch M, Nadel L (2013) Long-axis specialization of the
human hippocampus. Trends Cogn Sci 17:230–240.
Rajah MN, D’Esposito M (2005) Region-specific changes in prefrontal function with age: a
review of PET and fMRI studies on working and episodic memory. Brain 128:1964–1983.
Raz N, Ghisletta P, Rodrigue KM, Kennedy KM, Lindenberger U (2010) Trajectories of brain
aging in middle-aged and older adults: regional and individual differences. Neuroimage
51:501–511.
Resnick SM, Pham DL, Kraut MA, Zonderman AB, Davatzikos C (2003) Longitudinal
magnetic resonance imaging studies of older adults: a shrinking brain. J Neurosci
23:3295–3301.
Rieckmann A, Karlsson S, Fischer H, Backman L (2011) Caudate dopamine D1 receptor den-
sity is associated with individual differences in frontoparietal connectivity during working
memory. J Neurosci 31:14284–14290.
Ritchey M, Bessette-Symons B, Hayes SM, Cabeza R (2011) Emotion processing in the aging
brain is modulated by semantic elaboration. Neuropsychologia 49:640–650.
Sala-Llonch R, Arenaza-Urquijo EM, Valls-Pedret C, Vidal-Pineiro D, Bargallo N, Junque C,
Bartres-Faz D (2012) Dynamic functional reorganizations and relationship with working
memory performance in healthy aging. Front Hum Neurosci 6:152.
Salami A, Pudas S, Nyberg L (2014a) Elevated hippocampal resting- state connectiv-
ity underlies deficient neurocognitive function in aging. Proc Natl Acad Sci U S A
111:17654–17659.
127
Salami A, Rieckmann A, Fischer H, Backman L (2014b) A multivariate analysis of age-

related differences in functional networks supporting conflict resolution. Neuroimage
86:150–163.
Sambataro F, Murty VP, Callicott JH, Tan HY, Das S, Weinberger DR, Mattay VS (2010) Age-
related alterations in default mode network: Impact on working memory performance.
Sambataro F, Reed JD, Murty VP, Das S, Tan HY, Callicott JH, Weinberger DR, Mattay VS
(2009) Catechol- O-methyltransferase valine(158)methionine polymorphism modulates
brain networks underlying working memory across adulthood. Biological Psychiatry
66:540–548.
Saverino C, Grigg O, Churchill N, Grady CL (2015) Age differences in the default network at
rest and the relation to self-referential processing. Soc Cogn Affect Neurosci 10:231–239.
Schultz AP, Chhatwal JP, Huijbers W, Hedden T, van Dijk KR, McLaren DG, Ward AM,
Wigman S, Sperling RA (2014) Template based rotation: A method for functional con-
nectivity analysis with a priori templates. Neuroimage 102 Pt 2:620–636.
Seeley WW, Menon V, Schatzberg AF, Keller J, Glover GH, Kenna H, Reiss AL, Greicius MD
(2007) Dissociable intrinsic connectivity networks for salience processing and executive
control. J Neurosci 27:2349–2356.
Sheline YI, Morris JC, Snyder AZ, Price JL, Yan Z, D’Angelo G, Liu C, Dixit S, Benzinger
T, Fagan A, Goate A, Mintun MA (2010a) APOE4 allele disrupts resting state fMRI
connectivity in the absence of amyloid plaques or decreased CSF Abeta42. J Neurosci
30:17035–17040.
Sheline YI, Raichle ME, Snyder AZ, Morris JC, Head D, Wang S, Mintun MA (2010b) Amyloid
plaques disrupt resting state default mode network connectivity in cognitively normal
elderly. Biological Psychiatry 67:584–587.
Shulman GL, Astafiev SV, Franke D, Pope DL, Snyder AZ, McAvoy MP, Corbetta M (2009)
Interaction of stimulus-driven reorienting and expectation in ventral and dorsal frontopari-
etal and basal ganglia-cortical networks. J Neurosci 29:4392–4407.
Smith SM, Fox PT, Miller KL, Glahn DC, Fox PM, Mackay CE, Filippini N, Watkins KE, Toro
R, Laird AR, Beckmann CF (2009) Correspondence of the brain’s functional architecture
during activation and rest. Proc Natl Acad Sci U S A 106:13040–13045.
Song J, Birn RM, Boly M, Meier TB, Nair VA, Meyerand ME, Prabhakaran V (2014) Age-
related reorganizational changes in modularity and functional connectivity of human brain
networks. Brain Connect 4:662–676.
Spreng RN, Grady CL (2010) Patterns of brain activity supporting autobiographical memory,
prospection, and theory of mind, and their relationship to the default mode network. J
Spreng RN, Schacter DL (2012) Default Network Modulation and Large-Scale Network
Interactivity in Healthy Young and Old Adults. Cereb Cortex 22:2610–2621.
Spreng RN, Sepulcre J, Turner GR, Stevens WD, Schacter DL (2013) Intrinsic architecture
underlying the relations among the default, dorsal attention, and frontoparietal control
networks of the human brain. J Cogn Neurosci 25:74–86.
Spreng RN, Stevens WD, Chamberlain JP, Gilmore AW, Schacter DL (2010) Default network
activity, coupled with the frontoparietal control network, supports goal-directed cognition.
Squire LR (1992) Memory and the hippocampus: A synthesis from findings with rats, monkeys,
and humans. Psychol Rev 99:195–231.
St Jacques P, Dolcos F, Cabeza R (2010) Effects of aging on functional connectivity of the
amygdala during negative evaluation: a network analysis of fMRI data. Neurobiol Aging
31:315–327.
128
St Jacques PL, Rubin DC, Cabeza R (2012) Age-related effects on the neural correlates of auto-
biographical memory retrieval. Neurobiol Aging 33:1298–1310.
Steffener J, Habeck CG, Stern Y (2012) Age-related changes in task related functional network
connectivity. PLoS One 7:e44421.
Stern Y (2002) What is cognitive reserve? Theory and research application of the reserve con-
cept. J Int Neuropsychol Soc 8:448–460.
Stern Y, Albert S, Tang MX, Tsai WY (1999) Rate of memory decline in AD is related to educa-
tion and occupation: cognitive reserve? Neurology 53:1942–1947.
Stevens WD, Buckner RL, Schacter DL (2010) Correlated low-frequency BOLD fluctuations in
the resting human brain are modulated by recent experience in category-preferential visual
regions. Cereb Cortex 20:1997–2006.
Stone AA, Schwartz JE, Broderick JE, Deaton A (2010) A snapshot of the age distribution of
psychological well-being in the United States. Proc Nat Acad Sci 107:9985–9990.
Strange BA, Witter MP, Lein ES, Moser EI (2014) Functional organization of the hippocampal
longitudinal axis. Nat Rev Neurosci 15:655–669.
Strenziok M, Parasuraman R, Clarke E, Cisler DS, Thompson JC, Greenwood PM (2014)
Neurocognitive enhancement in older adults: comparison of three cognitive training
tasks to test a hypothesis of training transfer in brain connectivity. Neuroimage 85 Pt
3:1027–1039.
Tambini A, Ketz N, Davachi L (2010) Enhanced brain correlations during rest are related to
memory for recent experiences. Neuron 65:280–290.
Teipel SJ, Bokde AL, Meindl T, Amaro E, Jr., Soldner J, Reiser MF, Herpertz SC, Moller HJ,
Hampel H (2010) White matter microstructure underlying default mode network connec-
tivity in the human brain. 49:2021–2032.
Tomasi D, Volkow ND (2012) Aging and functional brain networks. Mol Psychiatry
17:549–558.
Vaidya CJ, Gordon EM (2013) Phenotypic variability in resting-state functional connectiv-
ity: current status. Brain Connect 3:99–120.
van den Heuvel MP, Hulshoff Pol HE (2010) Exploring the brain network: a review on resting-
state fMRI functional connectivity. Eur Neuropsychopharmacol 20:519–534.
van den Heuvel MP, Sporns O (2013) An anatomical substrate for integration among functional
networks in human cortex. J Neurosci 33:14489–14500.
Vincent JL, Kahn I, Snyder AZ, Raichle ME, Buckner RL (2008) Evidence for a frontopa-
rietal control system revealed by intrinsic functional connectivity. J Neurophysiol
100:3328–3342.
Voss MW, Erickson KI, Prakash RS, Chaddock L, Malkowski E, Alves H, Kim JS, Morris
KS, White SM, Wojcicki TR, Hu L, Szabo A, Klamm E, McAuley E, Kramer AF (2010a)
Functional connectivity: a source of variance in the association between cardiorespiratory
fitness and cognition? Neuropsychologia 48:1394–1406.
Voss MW, Prakash RS, Erickson KI, Basak C, Chaddock L, Kim JS, Alves H, Heo S, Szabo
AN, White SM, Wojcicki TR, Mailey EL, Gothe N, Olson EA, McAuley E, Kramer AF
(2010b) Plasticity of brain networks in a randomized intervention trial of exercise training
in older adults. Front Aging Neurosci 2:doi: 10.3389/fnagi.2010.00032.
Wang L, Laviolette P, O’Keefe K, Putcha D, Bakkour A, Van Dijk KR, Pihlajamaki M,
Dickerson BC, Sperling RA (2010a) Intrinsic connectivity between the hippocampus and
posteromedial cortex predicts memory performance in cognitively intact older individuals.
Wang L, Li Y, Metzak P, He Y, Woodward TS (2010b) Age-related changes in topological pat-
terns of large-scale brain functional networks during memory encoding and recognition.
129
Wang L, Su L, Shen H, Hu D (2012) Decoding lifespan changes of the human brain using
resting-state functional connectivity MRI. PLoS One 7:e44530.
West R, Murphy KJ, Armilio ML, Craik FI, Stuss DT (2002) Lapses of intention and perfor-
mance variability reveal age-related increases in fluctuations of executive control. Brain
Cogn 49:402–419.
Westlye ET, Lundervold A, Rootwelt H, Lundervold AJ, Westlye LT (2011) Increased hip-
pocampal default mode synchronization during rest in middle-aged and elderly APOE
epsilon4 carriers: relationships with memory performance. J Neurosci 31:7775–7783.
Whitfield-Gabrieli S, Ford JM (2012) Default mode network activity and connectivity in psy-
chopathology. Annu Rev Clin Psychol 8:49–76.
Wu JT, Wu HZ, Yan CG, Chen WX, Zhang HY, He Y, Yang HS (2011) Aging-related changes
in the default mode network and its anti-correlated networks: a resting-state fMRI study.
Neurosci Lett 504:62–67.
Yonelinas AP, Widaman K, Mungas D, Reed B, Weiner MW, Chui HC (2007) Memory in the
aging brain: doubly dissociating the contribution of the hippocampus and entorhinal cor-
tex. 17:1134–1140.
Zeng LL, Shen H, Liu L, Wang L, Li B, Fang P, Zhou Z, Li Y, Hu D (2012) Identifying major
depression using whole-brain functional connectivity: a multivariate pattern analysis.
Brain 135:1498–1507.
Zuo XN, Ehmke R, Mennes M, Imperati D, Castellanos FX, Sporns O, Milham MP (2012)
Network centrality in the human functional connectome. Cereb Cortex 22:1862–1875.
130
131
Multimodal Imaging
of the Aging Brain
Anders M. Fjell
Kristine B. Walhovd
I magine that you were to pinpoint the ongoing change of some critical param-
eter, and you were presented with a set of widely different descriptions—for
instance, one depicting linear decrease, versus a U-shaped function, an inverse U-
shaped function, or a relatively straight line showing stability. Surely you would try to
figure out which is the right one, especially if the question was of vital importance, like
access to food resources or the extent of global warming. We will argue that in study-
ing the aging brain with different neuroimaging modalities, one is actually faced with
equally divergent alternatives and choices—and a question that is of vital importance.
How the human brain changes through life is really a question of how people and their
premises for cognitive capacity change, and how society can accommodate their needs.
Yet, all of the above-mentioned trajectories can be observed in neuroimaging studies of
aging, demonstrating that widely different types of age changes are seen. For instance,
across the adult lifespan, the following trajectories have been observed:
• monotonic, almost linear decrease in measures of cortical thickness (Salat et al.,

2004, Fjell et al., 2014b, Storsve et al., 2014) as measured by T1-weighted
(T1w) MRI
• U-shaped age function for white matter (WM) mean diffusivity (MD) and
inverse U-shaped functions for WM fractional anisotropy (FA) as measured by
diffusion tensor imaging (DTI) (Westlye et al., 2010b)
• U-shaped age function for intracortical myelination as measured by T1w/T2w
ratio (Grydeland et al., 2013)
131
132
• relative stability in aging has been found for activity in parts of the default mode
network (DMN) as studied by positron emission tomography (PET) (Beason-
Held et al., 2009) and functional MRI (fMRI) (Persson et al., 2014), while oth-
ers found marked effects of aging (Lustig et al., 2003)
It will be impossible and outright wrong to pick any one of these descriptions and say
that this is it, this is how to best describe the aging of the brain. There are numerous
structural and functional brain changes in aging, and different neuroimaging modali-
ties are sensitive to different features. As a consequence, multimodal neuroimaging
has grown in popularity in studies of the aging brain (Nevalainen et al., 2015). A
pressing question, then, is how to best make use of it.
Most comprehensive aging studies now include multiple types of brain imaging
scans in their data collection protocols. While multimodal neuroimaging approaches
may yield critical information about processes ongoing in the aging brain, the use of
a multimodal approach does not come without costs. These costs relate to money,
time, and hence possibly also attrition rates in longitudinal studies. Much more effort
needs to be invested to properly analyze the data, also leading to more complex and
sometimes less easily communicable results. Here, we will discuss different benefits
of using a multimodal neuroimaging approach to study the aging brain, and briefly
review some ways of analyzing complex multimodal datasets. We will critically eval-
uate how brain aging can be investigated by multimodal imaging, and give exam-
ples of recent studies that have informed our understanding of aging by use of such
approaches. Also, we will discuss to what extent multimodal imaging can be used to
inform theoretical views on brain aging. We will conclude by discussing whether any
given measure can be seen as temporally prior to the others, more directly related to
cognition, or relatively more diagnostic of disorders of the aging brain. In this, we dis-
cuss challenges associated with a multimodal neuroimaging approach to brain aging
and give some recommendations for future research.
First, we need to explicate our view on what constitutes multimodal neuroimaging.
In this chapter, the term multimodal can refer to research on at least three different
levels: First, different neuroimaging hardware can be used, such as magnetic reso-
nance imaging (MRI) versus positron emission tomography (PET) scanners. Second,
there are different ways of acquiring data on the same hardware, such as T1-weighted
(T1w) sequences versus diffusion tensor imaging (DTI) or blood oxygenation level-
dependent (BOLD) contrast for MRI, and use of ligands for glucose metabolism vs.
ligands for amyloid accumulation for PET. Finally, here we also regard as multimodal
studies those where the same scans are used to quantify different tissue properties. For
instance, T1w scans can be used to estimate cortical thickness, and also signal inten-
sity and contrast between different tissue classes. Multimodal studies of brain aging
have been undertaken on all these three levels. A quick overview of some common
multimodal approaches is given in Box 5.1.
Strictly speaking, aging studies should either investigate change over time with
longitudinal designs, or use age as a variable of interest in cross-sectional studies.
Studies testing, for instance, brain–cognition relationships only in groups of older per-
sons cannot make inferences about the aging process per se, because we do not have
information about whether the reported relationships are dependent on the influence
â•‡ 133
Multimodal Imaging of the AgingÂ€Brainâ•… 133
Box 5.1â•‡ Examples ofÂ€imaging modalities used withÂ€success toÂ€study theÂ€agingÂ€brain.
(1) Different hardware (scanners)

• Magnetic resonnance imaging (MRI)
• Positron emission tomography (PET)
(2) Acquisition of different data on the same hardware
MRI
• T1‐weighted (T1w)
• T2‐weighted (T2w)
• Diffusion tensor imaging (DTI)
• BOLD (functional activity)
PET
• FDG—â•‰Fludeoxyglucose [18F] (glucose metabolism)
• PIB—â•‰Pittsburgh Compund B [11C]/â•‰Florbetapir [18F]/â•‰Florbetapen
[18F] (amyloid)
• SCH 23390 [11C] (dopamine D1 receptors)/â•‰raclopride [11C] (D2/â•‰D3
receptors)
(3) Quantification of different properties from the same data
• T1w for morphometry vs. signal intensity and tissue class contrast
(e.g. GM/â•‰WM)
• BOLD for level of activity change between conditions vs. functional
connectivity
Examples of imaging modalities used with success to study the aging brain.
of age in any way. In the present chapter, however, we will also refer to discuss stud-
ies of older persons that do not fulfill these criteria when relevant for the discussion
of multimodal imaging.
We will focus on the most-used imaging methods in aging studies—â•‰based on MRI
and PET. Other methods, such as electroencephalography/â•‰event-â•‰related potentials
(EEG/â•‰ERP), single photon emission computed tomography (SPECT), and near-â•‰infra-
red spectroscopy (NIRS), are also important in aging studies but will not be topics for
the present chapter.
Why Use a Multimodal Imaging Approach to Study Brain Aging?
A multiâ•‰modal imaging approach can be beneficial for several reasons that can be
divided into at least four broad but overlapping classes—â•‰the power to explain an
increased amount of aging-â•‰related variance in a measure of interest, such as cognitive
function; to control for confounding variables in studying a particular phenomenon; to
provide better illumination of underlying tissue or functional changes; and to facilitate
134
the discovery of heterogeneities in aging trajectories between brain tissues, regions,

and functions. All of these contribute to give multimodal approaches the potential to
yield a more accurate and complete picture of the aging brain.
Multimodal Imaging Can Increase Amount of Explained Variance

Although it is a futile task to fully explain the phenomenon of brain aging, adding
imaging modalities to an analysis has the potential for allowing more variance to be
explained. For instance, volumetric and microstructural changes of WM are relatively
weakly or moderately correlated (Fjell et al., 2008), yet both have been shown to
relate to cognitive performance (Fjell et al., 2011, Westlye et al., 2011, Borghesani et
al., 2013, Laukka et al., 2013, Bennett and Madden, 2014). Including both volumetric
and microstructural measures may thus increase amount of explained variance (Fjell
et al., 2012). The same goes for other combinations of imaging measures (Hedden et
al., 2016). Enrichment strategies including multiple neuroimaging modalities along
with other biomarkers and genetic information are often used to increase diagnostic
and predictive accuracy within research on aging-â•‰related cognitive disorders such as
Alzheimer’s disease (AD) (Villain et al., 2010, Walhovd et al., 2010, Weiner et al.,
2012). Studies using similar approaches are now emerging also focusing on normal
aging.
One such recent study combined measures of gray matter volume (GM), metab-
olism and functional connectivity (FC) in healthy elderly (Arenaza-â•‰Urquijo et al.,
2013). First, they found that higher education was related to greater volume and
metabolism in the anterior cingulate. Next, they demonstrated that resting-state func-
tional connectivity (RSFC) between the anterior cingulate and several other regions,
including the hippocampus, was positively related to education. Finally, higher RSFC
between these areas was related to cognitive performance. Thus, by simultaneously
assessing morphometry, metabolism, and RSFC, this study was able to give a much
more detailed account of the relationship between education and features of the
aged, non-â•‰demented brain. In a multimodal study of age-â•‰effect on memory, Ward et
al. compared hippocampal volume, entorhinal thickness, and default-â•‰mode network
(DMN) activity across young and older adults, and tested the relationship to memory
performance (Ward et al., 2015). The results showed that each imaging marker sig-
nificantly mediated the relationship between age and memory performance, and col-
lectively accounted for 68.8% of the variance in age-â•‰related memory performance.
Thus, age-â•‰related decline in memory could be best explained by combining multi-
ple neuroimaging markers. Interestingly, the relationships were stronger among the
older participants than among the younger, which could indicate that they emerge as
a result of aging-â•‰related brain changes, although caution of course needs to be taken
when inferring processes of change from cross-â•‰sectional data. Similarly, Hedden et
al. found that 70%–â•‰80% of the age-â•‰related variance in cognition could be explained
by combining a range of neuroimaging biomarkers, that is, gray matter volume and
thickness, WM hyperintensities, FA, RSFC, PET markers of glucose metabolism, as
well as amyloid burden (Hedden et al., 2016). However, the authors also suggested
that most age-â•‰related variation in cognition is shared among multiple markers, which
135
Multimodal Imaging of the Aging Brain 135
could lead to questions about the use of several markers in the same study. Still, up to
17.5% of the age-related variance in cognitive function could be explained uniquely
by single imaging markers alone when the effects of other markers were accounted
for. Also of importance, different markers were most potent in explaining variance
across the three cognitive domains tested—processing speed, executive function, and
episodic memory. Finally, the authors chose summary measures with little regional
variability, such as a single measure of FA, to optimize generalizability. It is possible
that this approach also maximized the amount of shared variance between measures
and modalities at the cost of higher specificity in the contribution from single mea-
sures (Hedden et al., 2016).
In sum, there are good examples of studies taking advantage of more than one
neuroimaging modality to explain a phenomenon such as the relationship between
education and brain aging, and between cognition in aging and brain structure and
function, respectively, increasing the amount of variance that can be accounted for.
Multimodal Imaging Can Ease Interpretation via Increased Control for

Confounding Variables
Since there are numerous processes in the brain that are affected by aging, covariance
between different neuroimaging measures can often be expected and observed. For
instance, reduction of cortical thickness in aging may, in principle, cause reduction
in the amplitude of the PET FDG signal solely due to an increase of partial voluming
from surrounding CSF and underlying WM, where glucose metabolism is much lower
(Harris and Attwell, 2012). If the FDG signal generated in GM is three times higher
than in WM (Heiss et al., 2004), a 10% reduction of thickness could cause a > 5%
reduction in PET signal due to partial voluming alone, given a voxel size of 5×5×5
mm. In such cases, observed differences in metabolism as a function of age could, in
principle, be completely accounted for by changes in cortical thickness. Thus, adding
cortical thickness as a covariate in analyses of PET images will ensure that effects
of age or relationships between metabolism and cognitive function are not merely
products of atrophy (see e.g., Westlye et al., 2010b). This is illustrated in Figure 5.1.
In this case, cognitively normal adults and patients with mild cognitive impairment
(MCI) from the ADNI show significantly different glucose metabolism in hippocam-
pus as measured with FDG PET. When the differences in hippocampal volume are
taken into account to adjust the metabolic measure, the group difference is no longer
significant. This is one simple example of how multimodal imaging allows controlling
for potentially important confounds that may critically affect the results of a study and
the conclusions that can be drawn.
Controlling for volumetric effects has also gradually become more common in
fMRI studies of aging, for instance when testing correlations between RSFC and
age. Using this approach, several cross-sectional studies found that relationships
between age and RSFC survived correction for GM volume (Damoiseaux et al., 2008,
Mowinckel et al., 2012, Onoda et al., 2012). A longitudinal study demonstrated a rela-
tionship between change in DMN connectivity and change in memory scores, with
no major changes in brain volume observed in regions overlapping with the DMN,
136
FGD PET Volume PET corrected (z)
p < .05 p < .05 n.s.

1,25 3600 1
0,8
1,2 3500
0,6
3400
1,15 0,4
3300 0,2
1,1
0
3200
Older adults MCI
1,05 –0,2
3100
–0,4
1
3000 –0,6
–0,8
0.95 2900
Older adults MCI Older adults MCI –1
Figure 5.1 Controlling for confounding variables

One application of multimodal imaging in aging studies is better control for confounding variables. Left plot: FDG PET (normalized to the pons) in the left hippocampus
from 101 cognitively normal older adults and 200 MCI patients from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Middle plot: Left hippocampal volume. Right
plot: FDG PET in left hippocampus, corrected for group differences in volume. When hippocampal volume differences were used to adjust the FDG PET values, group differ-
ences were no longer significant. This illustrates that multimodal imaging can be used to control for confounding variables that substantially may affect the observed results.
137
indicating that connectivity changes could not be explained by morphometric effects

(Persson et al., 2014). In these studies, inclusion of volumetric measures strengthens
the conclusion that functional effects of aging do not only reflect gross morphometric
changes.
A related issue concerns the relationship between WM microstructure and GM
atrophy. Changes in WM integrity, as quantified by DTI, have been reported in age-
related cognitive decline. However, there is debate whether WM degeneration in such
degenerative conditions is secondary to GM injury, especially through Wallerian
degeneration, or whether WM changes in and of themselves can be independent of
processes in GM (Amlien and Fjell, 2014). By using measures of GM structure as
covariate, microstructural differences between healthy elderly and patients with AD
that could not be accounted for by volumetric GM differences and vice versa have
been shown (Canu et al., 2010). Similarly, Stricker et al. tested whether reduced WM
integrity in patients with MCI would still be seen if cortical thickness was added as
covariate in the analyses (Stricker et al., 2013). The results showed that medial tem-
poral FA was related to memory, and parietal FA was related to executive functioning,
and suggested that changes in WM may be independent of GM changes in early AD.
Since neuronal (Peters et al., 1998, Freeman et al., 2008) and even glial (Pakkenberg
et al., 2003) death is limited in normal aging, less focus has been placed on Wallerian
degeneration and the question of independence of WM vs. GM changes in healthy
older adults.
Multimodal Imaging can Contribute to Illuminate Biological

Mechanisms Better than Any One Method Alone
Aging affects a multitude of different brain characteristics—functional and structural,
microstructural and macrostructural, which should, as discussed above, be seen in
relation to each other. It is given that one cannot understand many of these differ-
ent age processes without simultaneously using multiple modalities. Over the last
years, a great number of studies have taken advantage of different neuroimaging tech-
niques to yield a better picture of the neurobiological mechanisms that characterize
the aging of the brain, and new studies are ongoing, exploring an even greater range
of possible mechanisms (Nevalainen et al., 2015). By applying different methods to
study the same tissue class, we have gained increased understanding of the effects of
aging on the brain. For instance, Leritz et al. explored the role of WM lesions, which
typically increase with age, in cognitive decline (Leritz et al., 2014). They examined
whether major macrostructural WM damage, detectable on T1-weighted MRI scans,
was associated with microstructural integrity of normal-appearing WM as assessed
by DTI. Indeed, the volume of major WM damage was found to be associated with
lower microstructural integrity in widespread brain regions as indexed both by FA and
radial diffusivity, but to a lesser extent by axial diffusivity. However, controlling for
the amount of major WM lesions did not remove the relationship between FA and age,
indicating that the etiology of major WM lesions could not fully account for all age-
associated WM deterioration. Another multimodal study of WM found that micro-
structural characteristics at baseline, in this case FA as indexed by DTI, predicted
138
degree of volumetric shrinking of WM over a four-year period in middle-aged healthy

adults (Ly et al., 2014). Naturally, only parts of the observed WM atrophy could be
predicted from baseline microstructure, but the results still suggest an interesting link
between events on a micro-and macrostructural level.
A multitude of other measures have also been combined to study different aspects
of tissue classes, such as combining microstructural properties of WM with struc-
tural changes in GM. An example of this is given by He et al. (2012). They showed
that in a group of cognitively normal older adults, GM volume, WM hyperintensi-
ties, and structural connectivity were independently associated with episodic memory
performance. It was further found that greater connectivity was associated with better
memory performance only within the group of participants with GM volumes below
the median. This, they argued, suggested compensation against the effects of neuronal
injury.
In addition to structural connectivity, FC has also been related to GM structure
in aging. Meunier et al. directly related RSFC and GM density and found that lower
GM density was associated with lower connectivity within key language regions, but
higher overall FC (Meunier et al., 2014). They suggested that this could be interpreted
as less efficient network organization with greater age, and that engagement of a more
distributed network in aging might be triggered by reduced connectivity within spe-
cialized networks.
Studies have also focused on understanding whether Aß deposition, as measured
by amyloid PET imaging, is related to structural and functional characteristics of the
aged, non-demented brain. For instance, Wirth et al. found in a sample of healthy
adults that PiB retention and cortical neurodegeneration were independent measures,
but PiB-positive participants with more neurodegeneration showed the most longitu-
dinal cognitive decline (Wirth et al., 2013b). In a study of cognitively normal older
adults, Glodzik et al. combined T2-FLAIR MRI for mapping of WM lesions with FDG
and PIB PET to measure glucose metabolism and amyloid accumulation, respectively
(Glodzik et al., 2014). They hypothesized that WM lesions would disrupt connections
between GM regions, thereby changing their activation patterns. Interestingly, they
observed that brain regions with more lesions in connecting WM had lower glucose
metabolism and lower Aß deposition. They suggested that this could be caused by Aβ
being positively related to synaptic activity (Cirrito et al., 2005; Bero et al., 2011),
and disruption of activity by lesions in connecting WM regions would thus reduce
metabolism and thereby Aβ accumulation. The latter has been proposed as a hypoth-
esis to explain the spatial overlap between DMN and amyloid accumulation, taking
into account the possibility that not only may Aß impact cognitive activity, but brain
activity itself may be instrumental in causing Aß accumulation (for further discus-
sions of this issue, see Buckner, 2012; Jagust, 2013; Fjell et al., 2014a).
After the initial discovery of a spatial overlap between accumulation of amyloid as
measured by PET and the brain’s default-mode network (DMN; see Buckner, 2012),
several studies focused on the relationship between RSFC within the DMN and amy-
loid imaging, both in terms of spatial distribution and in terms of temporal evolution.
Sheline and Raichle (2013) conclude that some studies have shown RSFC effects
before evidence of amyloid deposition, such as one study demonstrating that APOE4-
positive subjects who were PIB-negative and had normal CSF levels of Aβ42 had
139
lower RSFC between precuneus, hippocampus, and anterior cingulate cortex (Sheline
et al., 2010). Sheline and Raichle stated that “It is not clear whether these additional
RSFC differences reflect an effect of amyloid detectable before PET scan or, more
likely, whether these additional rs-fMRI differences reflect other effects of genetically
induced neurodevelopmental brain differences (Sheline and Raichle, 2013).” Lower
DMN connectivity and lower connectivity from the precuneus to the hippocampus
have been seen also in other studies of cognitively normal older adults (Hedden et al.,
2009). Drzega et al. found that older adults with high levels of amyloid but without
clinical symptoms still showed disrupted RSFC in cortical hubs, and that these dis-
ruptions were associated with hypometabolism as measured by FDG PET (Drzezga
et al., 2011).
This approach of focusing on networks and connections between regions of assumed
importance is very interesting, and related disturbances and dysfunctions of specific
functional networks to structural brain characteristics, cognitive skills, and different
risk factors for cognitive decline will likely contribute to a better understanding of
the correlates of successful neurocognitive aging. Naturally, this has become a huge
interest in current neuroimaging, along with outside research on pre-dementing states
and the role of amyloid with special focus on integration of structural and functional
brain measures. Several researchers have pointed to the need to integrate functional
and structural connectivity in aging studies. This should ideally also be combined
with detailed analysis of behavioral performance (Antonenko and Floel, 2014). Such
analyses go beyond controlling for atrophy as a confounding variable. Relationships
between DTI and RSFC in aging have been reported (Andrews-Hanna et al., 2007,
Chen et al., 2009, Damoiseaux and Greicius, 2009, Teipel et al., 2010, Davis et al.,
2012). However, even though MCI and AD patients generally seem to have lower
RSFC than cognitively healthy elderly (Sheline and Raichle, 2013), a less straightfor-
ward picture emerges from aging studies per se. Both positive and negative RSFC–age
correlations are observed, depending on, among other things, the network under study.
The DMN, a so-called “task-negative network,” has been the most studied, and it
seems that aging generally is negatively related to connectivity within DMN. Based
on a very large sample (n = 913) from the Functional Connectome Project, Tomasi and
Volkow found decline in DMN and the dorsal attention network and suggested that
long-range connections are more prone to age effects than short-range connections
(Tomasi and Volkow, 2012, Ferreira and Busatto, 2013). Age-related changes in RSFC
are likely partly caused by related changes in WM integrity, dopamine transmission,
and/or amyloid deposition (Sheline and Raichle, 2013). Thus, better understanding of
functional and structural connectivity changes in aging is pivotal for a better under-
standing of both phenomena.
However, integrating structural and functional connectivity is challenging, as
structural connectivity measures usually show fairly consistent effects of age in terms
of direction, while FC is related to age in a much more complex way. While mean and
radial diffusivity increases and FA decreases in aging (Salat et al., 2005, Walhovd et
al., 2005, Westlye et al., 2010b, Bennett and Madden, 2014, Lockhart and DeCarli,
2014, Sexton et al., 2014), RSFC does not uniformly decrease with age (Antonenko
and Floel, 2014). In the simplest model, RSFC connectivity depends on effective trans-
fer of information across distant brain regions, which again necessitates myelinated,
140
well-functioning tracts. Thus, one could expect a close relationship between functional
and structural connectivity in aging, but this is generally not seen. In a recent review
paper, it was even suggested that the relationship between WM integrity and neural
activity varied as a function of age, so that positive DTI-fMRI relationships char-
acterize young, and negative relationships characterize older samples (Bennett and
Rypma, 2013). The complexity of the relationship becomes even more evident when
comparing lifespan trajectories of functional vs. structural connectivity measures. An
important point in this regard is that although connectivity can be operationalized both
at a functional and a structural level, it is often measured from different tissue types;
that is, GM vs. WM. As described above, GM and WM macrostructure also show very
different age trajectories. This situation makes it evident that multimodal approaches
have the potential to contribute critically. Ideally, structural and functional connectiv-
ity measures should inform each other in cases where they are differentially affected
by age. For instance, Hawelka et al. suggested that when increased FC is accompanied
by structural disconnectivity this may be caused by (1) less-differentiated patterns of
neural activity, (2) reduced cognitive efficiency, and (3) anatomical disconnectivity
leading to loss of functional diversity among brain networks (Hawellek et al., 2011).
Effects of age on functional MRI per se are more thoroughly discussed in another
chapter in this book, but the take-home message here is that multimodal imaging, such
as the combination of FC with different structural brain measures, can contribute to
illuminate biological mechanisms better than any one method alone
Discover Heterogeneities in Aging Trajectories Between Brain Tissues,

Regions, and Functions
Aging does not start when one reaches 60, and most researchers now agree that it
is beneficial to study samples with wide age distributions, both with cross-sectional
and longitudinal designs (see discussion below). Even if groups of young and older
participants show the same brain characteristics, or the same rate of change, it is dif-
ficult to distinguish a phenomenon of apparent preservation through adult life from a
U-shaped life span curve if only older and younger adults are included. For instance,
cross-sectional studies indicate that WM increases in volume until about middle age
before reductions kick in (Allen et al., 2005, Walhovd et al., 2005). Thus, early studies
of WM volume comparing younger with older adults would often observe no or only
small effects of age, and only studies sampling more continuously across at least adult
life would observe the inverted U-shaped age function. In contrast, cortical thick-
ness shows monotonic reductions from early childhood (Westlye et al., 2010a), and
volume declines from adolescence (Brown and Jernigan, 2012, Amlien et al., 2014)
(see Figure 5.2). Such differences in shape of the age-curves across tissue types reveal
important information relevant for aging that cannot easily be uncovered by samples
representing older adults only. Thus, studying different neuroimaging parameters over
a wider age range is, in our opinion, one of the most promising approaches for mul-
timodal studies.
One recent multimodal study used this approach by combining DTI with T1w and
T2w MR scans and tracked age trajectories from childhood to old age (Grydeland et al.,
141
GM volume WM volume
4,0
2,0
3,0
Z-scores
–2,0
–4,0
–6,0
20 40 60 80 100 20 40 60 80 100
Years
Figure 5.2 Life-span trajectories of cortical GM volume and WM volume. Plotted as a func-
tion of age, cortical GM volume and WM volume show very different trajectories. The mono-
tonic decline of cortical volume is accompanied by protracted growth of WM volume, before
both measures decline after middle age. Data modified from Fjell et al. (2013b).
2013). MD was calculated from the DTI scans, and the ratio between signal intensity
in T1w and T2w scans was also measured. MD and the T1w/T2w ratio are measures of
microstructural properties of brain tissue. Although especially MD is affected by many
different aspects of brain tissue (Concha et al., 2010, Zatorre et al., 2012), both MD
and the T1w/T2w ratio have been related to myelin organization or content (Song et
al., 2003, Glasser and Van Essen, 2011, Glasser et al., 2014). Grydeland et al. showed
that myelin content, as indexed by the T1w/T2w ratio from regions within the cerebral
cortex, showed inverted U-shaped age trajectories. Myelin maturation seemed to be
ongoing until the late 30s, followed by relative stability for 20 years, before declining
from the late 50s (Grydeland et al., 2013). MD showed U-shaped age trajectories, with
reductions of diffusion ongoing until middle age before gradual increases were seen.
Although the T1w/T2w measure tended to be more sensitive in distinguishing between
maturation-related and aging-related processes, the two methods generally converged.
Thus, in this case, two different imaging modalities yielded partly overlapping infor-
mation about processes affecting WM properties in the aging brain, which became evi-
dent when studied across the lifespan. However, another DTI metric—FA—measured
in the underlying WM of the same participants showed quite different curve shapes.
Here, age at peak development was around 28 years (Westlye et al., 2010b). This dif-
fered from the finding of peaks generally occurring later than 50 years when the T1w/
T2w ratio measure was used, even when measured in WM. Thus, in this case, two dif-
ferent imaging modalities yielded non-overlapping information about processes affect-
ing WM properties in the aging brain, which, again, became evident when mapped
across a wide age range (see also Figure 5.3).
A telling example of two imaging modalities yielding complementary information
is the age-curves of cortical thickness compared to signal intensity. While cortical
thickness decreases monotonically, T1w intensity increases until about 30 years before
decline begins (Westlye et al., 2010b) (see Figure 5.4). Interestingly, development of
142
Fractional anisotropy Myelin content

(Superior longitudinal fasciculus) (Superior frontal gray matter)
3,0 ~28 years ~50 years
2,0
1,0
,0
Z-scores
–1,0
–2,0
–3,0
–4,0
–5,0
5 25 45 65 85 5 25 45 65 85
Years Years
Figure 5.3 Lifespan trajectories of different measures. Using multimodal imaging to study

brain characteristics that are presumably related can uncover important knowledge about brain
aging. In this example, fractional anisotropy in the superior longitudinal fasciculus as measured
by DTI indicates a positive age-function until around 28 years, after which a relatively linear
function explain the trend in the data. A proposed measure of myelin content in vivo, the ratio
between T1w and T2w MRI images taken from the superior frontal gray matter shows the same
general inverted U-shaped function, but the curve peaks at much later age (around 50 years).
The same is seen when myelin is measured within WM. Thus, two different imaging modalities
yield non-overlapping information about processes affecting WM properties in the aging brain,
which becomes evident when studied across the lifespan. Data for the FA plot is modified from
Westlye et al. (2010b) and data for the myelin plot from Grydeland et al. (2013).
the T1 intensity profile corresponds remarkably well to the inflection point of the age-
function for cortical thickness. Thus, combining the two sources of information yields
compelling hints about when cortical development typically may come to an end and
when aging-related processes begin. In this case, both measures could be obtained
from the same T1w scan, but they are still fundamentally different, thus constituting
an example of multimodal imaging in the third sense outlined above.
Ways of Analyzing Multimodal Imaging Data in Brain

Aging Studies
Multimodal neuroimaging in aging studies imposes challenges related to how data

analysis should be optimized. Each modality increases the information load, and there
is currently no agreement on the best way to deal with the risks of information over-
load in multimodal studies. Typically, data from each modality are analyzed through
separate pipelines, and the multimodal aspect is involved after the main preprocessing
has been done. Often these approaches also involve data reduction in one or sev-
eral modalities, typically extracting values from specific regions of interest (ROIs) or
tracts of interest (TOIs), which can be entered into some statistical framework. Often,
143
Cortical thickness
T1 GM intensity
3.2
Cortical thickness (mm)

T1 GM/ventricular
3.0
intensity ratio
4.4
2.8
4.2
2.6
20 40 60 80
Years
Figure 5.4 Intensity versus thickness of the cortex. From the superior frontal cortex, thick-
ness was measured along with the ratio between signal intensity in the T1w image sampled
1mm into the cortex, and the intensity of the ventricles plotted against age (dotted line). As
can be seen, these two different tissue properties from the same brain region showed markedly
different age functions. While cortical thickness decreases monotonically across the age-span
samples, T1 intensity increases until about 30 years before it starts to decline. Interestingly,
the peak of the T1 intensity curve corresponds remarkably well with the inflection point of the
thickness graph. Data modified from Westlye et al. (2010b).
each imaging modality is reduced to one or a few variables of interest, which then are
used in statistical analyses. Alternatively, data reduction is done for one modality only
and then tested against another modality in a less constrained manner. This method
has been used in many excellent studies and is a simple, yet powerful, approach.
Versions of this strategy were used in most of the studies described above.
Another approach is to use the results from one modality to guide the analyses of
the next modality. One can use results from fMRI to define regions of activation, con-
nectivity change, etc., and then test structural characteristics within these data-driven
regions, use tractography to define tracts from the activated regions, etc. Such an
approach was, for instance, used by Arenaza-Urquijo et al. when they first tested the
relationship between education, volume, and metabolism, and then used the resulting
regions as seeds for their FC analysis (Arenaza-Urquijo et al., 2013).
Recently, attempts have also been made to integrate different modalities earlier in
the analysis stream, at the pre-statistical level. Groves and colleagues used so-called
“multimodal fusion analysis” to study lifespan patterns in cortical morphometry
and WM microstructure (Groves et al., 2012). Here, linked independent component
analysis (LICA) was used on preprocessed data to model multimodal covariances in
a completely data-driven way, including six different morphometric and DTI mea-
sures. Of 100 components, several were modality-specific, meaning that one modality
dominated the component. However, other components represented a mix of DTI and
morphometric measures. Interestingly, the component most strongly related to cross-
sectional age (r = .95) represented multimodal information from cortical thickness and
voxel-based morphometry as well as the DTI indexes FA, MD, and MO (mode of the
diffusion tensor). Also, other multimodal components showed a distinct relationship
144
to age. One component showing a U-shaped curve, receiving contributions from WM

and localized GM effects. This represents a fundamentally different approach to data
reduction and integration of multimodal data in aging studies. The same approach was
later used to identify components with different age trajectories, which again could
be related to neurodevelopmental, neuropsychiatric, and neurodegenerative condi-
tions, and could be shown to correlate with general intellectual function and episodic
memory (Douaud et al., 2014).
We will not make any strong recommendations as to which approach generally is
the best in multimodal aging studies, because they all have their strengths. Multimodal
fusion analysis has the advantage of integrating different modalities at a pre-statistical
level, enabling identification of cross-modal covariance patterns, and it is very true to
data. However, this approach may not be the best for maximizing reproducibility or
interpretability. Preprocessing of each modality by separate streams and merging data
by uni-or multivariate statistics at the last stage requires data reduction before the
multimodal nature of data can be taken advantage of, but it is still a simple and power-
ful framework, yielding results that often are easier to interpret. Which approach to
use will ultimately depend on the data available and the research questions of interest.
Theoretical Perspectives—Multimodal Imaging Methods Converge

to Reveal General Features of the Aging Brain
How can multimodal imaging inform theoretical views on aging? Studies of age
differences using only measures of gross neuroanatomical volumes, such as from
T1-weighted MR scans, could easily fit with a picture of brain aging as one of uni-
versal and more or less linear decline. For instance, cortical thickness and volume is
seen to decrease monotonically through adulthood (Storsve et al., 2014). However,
this picture is refined when combining multiple imaging modalities, showing that
in young adulthood, GM signal intensities (Westlye et al., 2010a) and T1/T2 ratios
are simultaneously increasing, likely due to ongoing myelination (Grydeland et al.,
2013). Likewise, microstructural properties of WM show ongoing maturation and an
apparent stability at the same time as WM volumes are fluctuating (Fjell et al., 2008,
Westlye et al., 2010b). Further adding to and complicating the picture, functional
connectivity does not necessarily decline (Persson et al., 2014) in synchrony with
structural connectivity. Multimodal imaging helps to create a fuller picture of the
aging brain, namely one where degenerative and positive processes may be ongoing
simultaneously in different tissue classes and affecting different imaging parameters.
This also underscores the necessity of a lifespan perspective on brain aging, where
changes are continuously ongoing. Chances are, if no change is observed, one has
just not included the measure to capture it. Given the vast normal changes that occur,
it is also necessary to consider the true dimensionality of brain imaging differences
in older age. A brain in constant flux of change may not fit well into strict confines
of normality and disease, as such borders can be hard to define (Jagust, 2013, Fjell et
al., 2014a). Data from multiple imaging modalities now point to differences along a
continuum from normality to disease, and one may question whether any one method
holds promise to be more diagnostic, either alone or in combination.
145
The multidimensionality of the aging brain uncovered by multimodal imaging could

lead one to ask whether anything can be concluded regarding the temporal sequence
of brain age changes to be seen in the different imaging modalities? Can we expect to
see, for instance, microstructural changes prior to volume loss? Or functional activity
changes prior to cognitive decline? These questions might seem naïve, yet we believe
them to be central to multimodal imaging studies. For AD, a sequential model has
been proposed based on the amyloid cascade hypothesis (Jack et al., 2010, Jack et al.,
2013), where alterations in CSF Aβ42 and amyloid aggregation as indexed by PET
is followed by MRI volume loss and then cognitive decline in an orderly temporal
sequence (Vandenberghe et al., 2013). It would be most welcome if we could detail a
sequence of changes that are measureable in vivo in normal aging, perhaps predictive
of extent of cognitive decline. There are multiple reasons why this in some ways lucra-
tive option is not likely. First, even in AD, this model cannot be as straightforwardly
applied as implied (Chetelat, 2013a). As increasingly acknowledged, alternatives exist
to such linear sequential models, and the different in vivo imaging modalities may
rather reflect partly independent events (Vandenberghe et al., 2013). For instance, neu-
rodegeneration has been shown in the absence of Aβ deposition, before Aβ deposition,
and after Aβ deposition (Fjell et al., 2010, Fleisher et al., 2012, Reiman et al., 2012,
Jack et al., 2014), with medial temporal lobe atrophy having a multitude of different
causes (Barkhof et al., 2007). Even more so, perhaps, than AD, normal aging needs to
be imaged in a multidimensional space where different measures co-determine cogni-
tive outcome. It follows from the vastly different trajectories of the various imaging
modalities and measures that neither is likely related to cognitive performance in a
simple linear fashion. Two independent studies recently demonstrated neurodegenera-
tion within typical AD regions regardless of Aβ deposition, with amount of degenera-
tion still having detrimental effects on cognitive functions such as episodic memory
(Fjell et al., 2013a, Wirth et al., 2013a). This means that neuroimaging biomarkers
and other biomarkers sensitive to brain changes normal both in aging and AD may
operate in additive rather than sequential ways when it comes to impacting cognition
(Chetelat, 2013a, b), that the sequence of accelerated decline reflected by different
imaging markers may differ between individuals, and that the brain events indexed by
different markers may progress partly independently and partly in synchronicity with
each other (Fjell et al., 2014a). Thus, a given brain event may represent one, but not the
only, pathway to cognitive decline (Herrup, 2010).
We believe that this is exemplified in that the context of multimodal imaging has
been pivotal to a vast empirical-theoretical work over the last two decades, and to
continuously developing current research trends to understand mechanisms underly-
ing better or worse cognitive performance in older adults. Specifically, the focus on
compensatory mechanisms (Grady, 2012) relative to brain maintenance (Nyberg et
al., 2012) as critical to good cognitive function in older age relies heavily on seeing
different aspects of brain function and structure in concert with behavioral perfor-
mance. The concepts of posterior–anterior shift in aging (PASA)(Davis et al., 2008),
hemispheric-asymmetry reduction in older adults (HAROLD) (Cabeza, 2002), com-
pensation-related utilization of neural circuits hypothesis (CRUNCH) (Reuter-Lorenz
and Cappell, 2008), along with dedifferentiation (Goh et al., 2010) are inspired by,
developed, and/or criticized based on evidence from multiple imaging modalities.
146
Hence, brain differences in normal aging can now be seen within a multidimensional
space, and effects of individual differences within multiple modalities and their rela-
tion to each other and to cognition are beginning to be uncovered. For example, it was
recently found that decreased GM integrity was associated with decreased connectiv-
ity within key language regions, but increased overall FC (Meunier et al., 2014). As
this network organization was less efficient, it was suggested that engagement of a
more distributed network in aging might be triggered by reduced connectivity within
specialized networks.
As cognition reflects a number of structural and functional characteristics and their
interactions, we can only hope to capture a fraction of this important interplay even
when using multiple imaging modalities in longitudinal studies. Importantly, it has
become abundantly clear that inter-and intra-individual variability in brain structure
and function as measured by multimodal imaging must be understood in the context
of past and previous characteristics of those individuals. This regards both early influ-
ences on an individual’s trajectory of change throughout life and different offsets for
different persons, similar to what is often conceptualized as cognitive or brain reserve.
Thus, there is likely an early foundation that will affect the state of the brain as one
enters older age, which again will be critical for how much decline one can endure
before functional impairment is detectable and a clinical diagnosis is given. In addi-
tion, there may be early influences on the individual’s path through life which, in
principle, are different from the effects of the offset.
Possibly the greatest predictors of characteristics measured by neuroimaging
in aging is found at a very early stage of life. For instance, birth weight has been
found to predict neuroanatomical volumes and cortical surface area in later child-
hood, adolescence, and early adulthood (Walhovd et al., 2012). For the fat massand
obesity (FTO)–associated gene, a commonly occurring variation has been associated
with reduced brain volumes in healthy aging (Ho et al., 2010) as well as risk of AD
(Reitz et al., 2012). Recently, it was shown that this FTO-variance was associated
with smaller brain volumes in adolescents as well (Melka et al., 2013), meaning that
the differences observed in aging may be present throughout life, including at the
embryonic stage. Likewise, APOE ε4 carriers were recently found to show lesser
temporal lobe volumes at birth (Knickmeyer et al., 2014). In sum, current data point
to many factors associated with brain aging that could be revealed by imaging stud-
ies from the beginning. They may represent stable risk factors, and sometimes have
accumulative consequences throughout the lifespan. While we know by now some
of the developmental and aging trajectories of different imaging modality measures,
knowledge of whether and how their interrelations change across life is still scarce.
To some degree, combined cross-sectional and longitudinal designs can alleviate
this. An ideal approach to reproduce the dynamic process of change would be longi-
tudinal studies with high density of measures and assessment of multiple time win-
dows across the lifespan (Raz & Lindenberger, 2011). In multimodal imaging of the
aging brain, taking account of earlier life factors constitutes taking account of the
entire lifespan of individuals. As of yet, we know of no studies including brain imag-
ing data, much less multimodal brain imaging data, using a design that covers the
span from birth into old age.
â•‡ 147
Multimodal Imaging of the AgingÂ€Brainâ•… 147
Conclusion: When Is the Whole More than

the Sum of its Methods?
Running the same participants through several imaging modalities, and using greater
effort to analyze and present the results, comes at a cost. To justify using a multiâ•‰modal
neuroimaging approach in studying the aging brain instead of running separate stud-
ies with fewer modalities, the benefits need to outweigh these costs. In the present
chapter, we have argued that a multiâ•‰modal imaging study can give unique and supe-
rior information through the ability to control for confounds, and by offering more
complete explanations of biological substrates as well as mechanisms of change. The
latter is especially interesting in a lifeâ•‰span perspective, where different modalities
and brain features often show strikingly divergent trajectories, and the relationships
between them are not invariant through life. In those cases, multiâ•‰modal neuroimaging
provides a truly unique opportunity to gain critical insights into the aging brain and
the resulting changes in cognitive function.
References
Allen JS, Bruss J, Brown CK, Damasio H (2005) Normal neuroanatomical variation due to
age: the major lobes and a parcellation of the temporal region. Neurobiology of Aging
26:1245–â•‰1260; discussion 1279–â•‰1282.
Amlien I, Fjell AM, Tamnes CK, Grydeland H, Krogsrud SK, Chaplin T, Rosa M, Walhovd KB
(2014) Organizing principles of human cortical development –â•‰thickness and area from
four to thirty years: Insights from comparative primate neuroanatomy. Cerebral Cortex
26(1):257–267.
Amlien IK, Fjell AM (2014) Diffusion tensor imaging of white matter degeneration in
Alzheimer’s disease and mild cognitive impairment. Neuroscience 276:206–215.
Andrews-â•‰Hanna JR, Snyder AZ, Vincent JL, Lustig C, Head D, Raichle ME, Buckner RL
(2007) Disruption of large-â•‰scale brain systems in advanced aging. Neuron 56:924–â•‰935.
Antonenko D, Floel A (2014) Healthy aging by staying selectively connected: a mini-â•‰review.
Gerontology 60:3–â•‰9.
Arenaza-â•‰Urquijo EM, Landeau B, La Joie R, Mevel K, Mezenge F, Perrotin A, Desgranges
B, Bartres-â•‰Faz D, Eustache F, Chetelat G (2013) Relationships between years of educa-
tion and gray matter volume, metabolism and functional connectivity in healthy elders.
NeuroImage 83:450–â•‰457.
Barkhof F, Polvikoski TM, van Straaten EC, Kalaria RN, Sulkava R, Aronen HJ, Niinisto L,
Rastas S, Oinas M, Scheltens P, Erkinjuntti T (2007) The significance of medial temporal
lobe atrophy: a postmortem MRI study in the very old. Neurology 69:1521–â•‰1527.
Beason-â•‰Held LL, Kraut MA, Resnick SM (2009) Stability Of Default-â•‰Mode Network Activity
In The Aging Brain. Brain Imaging and Behavior 3:123–â•‰131.
Bennett IJ, Madden DJ (2014) Disconnected aging: Cerebral white matter integrity and age-â•‰
related differences in cognition. Neuroscience 276C:187–â•‰205.
Bennett IJ, Rypma B (2013) Advances in functional neuroanatomy: a review of combined DTI
and fMRI studies in healthy younger and older adults. Neuroscience and Biobehavioral
Reviews 37:1201–â•‰1210.
Bero AW, Yan P, Roh JH, Cirrito JR, Stewart FR, Raichle ME, Lee JM, Holtzman DM (2011)
Neuronal activity regulates the regional vulnerability to amyloid-â•‰beta deposition. Nature
Neuroscience 14:750–â•‰756.
148
Borghesani PR, Madhyastha TM, Aylward EH, Reiter MA, Swarny BR, Schaie KW, Willis
SL (2013) The association between higher order abilities, processing speed, and age
are variably mediated by white matter integrity during typical aging. Neuropsychologia
51:1435–1444.
Brown TT, Jernigan TL (2012) Brain development during the preschool years. Neuropsychology
Review 22:313–333.
Buckner RL (2012) The serendipitous discovery of the brain’s default network. NeuroImage
62:1137–1145.
Cabeza R (2002) Hemispheric asymmetry reduction in older adults: the HAROLD model.
Psychology and Aging 17:85–100.
Canu E, McLaren DG, Fitzgerald ME, Bendlin BB, Zoccatelli G, Alessandrini F, Pizzini FB,
Ricciardi GK, Beltramello A, Johnson SC, Frisoni GB (2010) Microstructural diffusion
changes are independent of macrostructural volume loss in moderate to severe Alzheimer’s
disease. Journal of Alzheimer’s Disease: JAD 19:963–976.
Chen NK, Chou YH, Song AW, Madden DJ (2009) Measurement of spontaneous signal fluctua-
tions in fMRI: adult age differences in intrinsic functional connectivity. Brain Structure &
Function 213:571–585.
Chetelat G (2013a) Alzheimer disease: Abeta-independent processes-rethinking preclinical
AD. Nature reviews Neurology 9:123–124.
Chetelat G (2013b) Reply: The amyloid cascade is not the only pathway to AD. Nature Reviews
Neurology 9:356.
Cirrito JR, Yamada KA, Finn MB, Sloviter RS, Bales KR, May PC, Schoepp DD, Paul SM,
Mennerick S, Holtzman DM (2005) Synaptic activity regulates interstitial fluid amyloid-
beta levels in vivo. Neuron 48:913–922.
Concha L, Livy DJ, Beaulieu C, Wheatley BM, Gross DW (2010) In vivo diffusion tensor
imaging and histopathology of the fimbria-fornix in temporal lobe epilepsy. The Journal of
Neuroscience: The Official Journal of the Society for Neuroscience 30:996–1002.
Damoiseaux JS, Beckmann CF, Arigita EJ, Barkhof F, Scheltens P, Stam CJ, Smith SM,
Rombouts SA (2008) Reduced resting-state brain activity in the “default network” in nor-
mal aging. Cerebral Cortex 18:1856–1864.
Damoiseaux JS, Greicius MD (2009) Greater than the sum of its parts: a review of studies com-
bining structural connectivity and resting-state functional connectivity. Brain Structure &
Function 213:525–533.
Davis SW, Dennis NA, Daselaar SM, Fleck MS, Cabeza R (2008) Que PASA? The posterior-
anterior shift in aging. Cerebral Cortex 18:1201–1209.
communication in the aging brain: linking behavior to functional and structural connectiv-
ity. Cerebral Cortex 22:232–242.
Douaud G, Groves AR, Tamnes CK, Westlye LT, Duff EP, Engvig A, Walhovd KB, James A,
Gass A, Monsch AU, Matthews PM, Fjell AM, Smith SM, Johansen-Berg H (2014) A
common brain network links development, aging, and vulnerability to disease. Proceedings
of the National Academy of Sciences of the United States of America 111:17648–17653.
Drzezga A, Becker JA, Van Dijk KR, Sreenivasan A, Talukdar T, Sullivan C, Schultz AP,
Sepulcre J, Putcha D, Greve D, Johnson KA, Sperling RA (2011) Neuronal dysfunction
and disconnection of cortical hubs in non-demented subjects with elevated amyloid bur-
den. Brain: A Journal of Neurology 134:1635–1646.
Ferreira LK, Busatto GF (2013) Resting-state functional connectivity in normal brain aging.
Neuroscience and Biobehavioral Reviews 37:384–400.
Fjell AM, McEvoy L, Holland D, Dale AM, Walhovd KB, Alzheimer’s Disease Neuroimaging
I (2013a) Brain changes in older adults at very low risk for Alzheimer’s disease.
149
The Journal of neuroscience: the official journal of the Society for Neuroscience
33:8237–8242.
Fjell AM, McEvoy L, Holland D, Dale AM, Walhovd KB, Alzheimer’s Disease Neuroimaging
I (2014a) What is normal in normal aging? Effects of aging, amyloid and Alzheimer’s
disease on the cerebral cortex and the hippocampus. Progress in Neurobiology 117:20–40.
Fjell AM, Walhovd KB, Brown TT, Kuperman JM, Chung Y, Hagler DJ, Jr., Venkatraman V,
Roddey JC, Erhart M, McCabe C, Akshoomoff N, Amaral DG, Bloss CS, Libiger O,
Darst BF, Schork NJ, Casey BJ, Chang L, Ernst TM, Gruen JR, Kaufmann WE, Kenet
T, Frazier J, Murray SS, Sowell ER, van Zijl P, Mostofsky S, Jernigan TL, Dale AM,
Pediatric Imaging N, Genetics S (2012) Multimodal imaging of the self-regulating devel-
oping brain. Proceedings of the National Academy of Sciences of the United States of
America 109:19620–19625.
Fjell AM, Walhovd KB, Fennema-Notestine C, McEvoy LK, Hagler DJ, Holland D, Blennow
K, Brewer JB, Dale AM, Alzheimer’s Disease Neuroimaging I (2010) Brain atrophy in
healthy aging is related to CSF levels of Abeta1–42. Cerebral Cortex 20:2069–2079.
Fjell AM, Westlye LT, Amlien IK, Walhovd KB (2011) Reduced white matter integrity is related
to cognitive instability. The Journal of Neuroscience: The Official Journal of the Society
for Neuroscience 31:18060–18072.
Fjell AM, Westlye LT, Greve DN, Fischl B, Benner T, van der Kouwe AJ, Salat D, Bjornerud A,
Due-Tonnessen P, Walhovd KB (2008) The relationship between diffusion tensor imaging
and volumetry as measures of white matter properties. NeuroImage 42:1654–1668.
Walhovd KB, Alzheimer Disease Neuroimaging I (2014b) Accelerating cortical thinning:
unique to dementia or universal in aging? Cerebral Cortex 24:919–934.
Fleisher AS, Chen K, Quiroz YT, Jakimovich LJ, Gomez MG, Langois CM, Langbaum JB,
Ayutyanont N, Roontiva A, Thiyyagura P, Lee W, Mo H, Lopez L, Moreno S, Acosta-
Baena N, Giraldo M, Garcia G, Reiman RA, Huentelman MJ, Kosik KS, Tariot PN,
Lopera F, Reiman EM (2012) Florbetapir PET analysis of amyloid-beta deposition in the
presenilin 1 E280A autosomal dominant Alzheimer’s disease kindred: a cross-sectional
study. The Lancet Neurology 11:1057–1065.
Freeman SH, Kandel R, Cruz L, Rozkalne A, Newell K, Frosch MP, Hedley-Whyte ET,
Locascio JJ, Lipsitz LA, Hyman BT (2008) Preservation of neuronal number despite age-
related cortical brain atrophy in elderly subjects without Alzheimer disease. Journal of
Neuropathology and Experimental Neurology 67:1205–1212.
Glasser MF, Goyal MS, Preuss TM, Raichle ME, Van Essen DC (2014) Trends and properties
of human cerebral cortex: correlations with cortical myelin content. NeuroImage 93 Pt
2:165–175.
Glasser MF, Van Essen DC (2011) Mapping human cortical areas in vivo based on myelin con-
tent as revealed by T1-and T2-weighted MRI. The Journal of Neuroscience: The Official
Journal of the Society for Neuroscience 31:11597–11616.
Glodzik L, Kuceyeski A, Rusinek H, Tsui W, Mosconi L, Li Y, Osorio RS, Williams S, Randall
C, Spector N, McHugh P, Murray J, Pirraglia E, Vallabhajosula S, Raj A, de Leon MJ
(2014) Reduced glucose uptake and Abeta in brain regions with hyperintensities in con-
nected white matter. NeuroImage 100:684–691.
Goh JO, Suzuki A, Park DC (2010) Reduced neural selectivity increases fMRI adaptation with
age during face discrimination. NeuroImage 51:336–344.
Grady C (2012) The cognitive neuroscience of ageing. Nature Reviews Neuroscience
13:491–505.
Groves AR, Smith SM, Fjell AM, Tamnes CK, Walhovd KB, Douaud G, Woolrich MW, Westlye
LT (2012) Benefits of multi-modal fusion analysis on a large-scale dataset: life-span
150
patterns of inter-subject variability in cortical morphometry and white matter microstruc-

ture. NeuroImage 63:365–380.
Grydeland H, Walhovd KB, Tamnes CK, Westlye LT, Fjell AM (2013) Intracortical myelin links
with performance variability across the human lifespan: results from T1-and T2-weighted
MRI myelin mapping and diffusion tensor imaging. The Journal of Neuroscience: The
Official Journal of the Society for Neuroscience 33:18618–18630.
Harris JJ, Attwell D (2012) The energetics of CNS white matter. The Journal of neurosci-
Hawellek DJ, Hipp JF, Lewis CM, Corbetta M, Engel AK (2011) Increased functional con-
nectivity indicates the severity of cognitive impairment in multiple sclerosis. Proceedings
of the National Academy of Sciences of the United States of America 108:19066–19071.
He J, Carmichael O, Fletcher E, Singh B, Iosif AM, Martinez O, Reed B, Yonelinas A, Decarli
C (2012) Influence of functional connectivity and structural MRI measures on episodic
memory. Neurobiology of Aging 33:2612–2620.
(2016) Multiple Brain Markers are Linked to Age-Related Variation in Cognition. Cerebral
Cortex 26(4):1388–400
Hedden T, Van Dijk KR, Becker JA, Mehta A, Sperling RA, Johnson KA, Buckner RL (2009)
burden. The Journal of neuroscience: the official journal of the Society for Neuroscience
29:12686–12694.
Heiss WD, Habedank B, Klein JC, Herholz K, Wienhard K, Lenox M, Nutt R (2004) Metabolic
rates in small brain nuclei determined by high- resolution PET. Journal of Nuclear
Medicine: Official Publication, Society of Nuclear Medicine 45:1811–1815.
Herrup K (2010) Reimagining Alzheimer’s disease--an age-based hypothesis. The Journal of
neuroscience: the official journal of the Society for Neuroscience 30:16755–16762.
Ho AJ, Stein JL, Hua X, Lee S, Hibar DP, Leow AD, Dinov ID, Toga AW, Saykin AJ, Shen L,
Foroud T, Pankratz N, Huentelman MJ, Craig DW, Gerber JD, Allen AN, Corneveaux JJ,
Stephan DA, DeCarli CS, DeChairo BM, Potkin SG, Jack CR, Jr., Weiner MW, Raji CA,
Lopez OL, Becker JT, Carmichael OT, Thompson PM, Alzheimer’s Disease Neuroimaging
I (2010) A commonly carried allele of the obesity-related FTO gene is associated with
reduced brain volume in the healthy elderly. Proceedings of the National Academy of
Sciences of the United States of America 107:8404–8409.
Jack CR, Jr., Knopman DS, Jagust WJ, Petersen RC, Weiner MW, Aisen PS, Shaw LM, Vemuri
P, Wiste HJ, Weigand SD, Lesnick TG, Pankratz VS, Donohue MC, Trojanowski JQ
(2013) Tracking pathophysiological processes in Alzheimer’s disease: an updated hypo-
thetical model of dynamic biomarkers. The Lancet Neurology 12:207–216.
Jack CR, Jr., Knopman DS, Jagust WJ, Shaw LM, Aisen PS, Weiner MW, Petersen RC,
Trojanowski JQ (2010) Hypothetical model of dynamic biomarkers of the Alzheimer’s
pathological cascade. The Lancet Neurology 9:119–128.
Jack CR, Jr., Wiste HJ, Weigand SD, Rocca WA, Knopman DS, Mielke MM, Lowe VJ, Senjem
ML, Gunter JL, Preboske GM, Pankratz VS, Vemuri P, Petersen RC (2014) Age-specific
population frequencies of cerebral beta-amyloidosis and neurodegeneration among people
with normal cognitive function aged 50–89 years: a cross-sectional study. The Lancet
Neurology 13:997–1005.
Jagust W (2013) Vulnerable neural systems and the borderland of brain aging and neurodegen-
eration. Neuron 77:219–234.
Knickmeyer RC, Wang J, Zhu H, Geng X, Woolson S, Hamer RM, Konneker T, Lin W, Styner
M, Gilmore JH (2014) Common Variants in Psychiatric Risk Genes Predict Brain Structure
at Birth. Cerebral Cortex 24(5):1230–1246.
151
Laukka EJ, Lovden M, Kalpouzos G, Li TQ, Jonsson T, Wahlund LO, Fratiglioni L, Backman
L (2013) Associations between white matter microstructure and cognitive performance in
old and very old age. PloS One 8:e81419.
Leritz EC, Shepel J, Williams VJ, Lipsitz LA, McGlinchey RE, Milberg WP, Salat DH (2014)
Associations between T1 white matter lesion volume and regional white matter micro-
structure in aging. Human Brain Mapping 35:1085–1100.
Lockhart SN, DeCarli C (2014) Structural imaging measures of brain aging. Neuropsychology
Review 24:271–289.
Lustig C, Snyder AZ, Bhakta M, O’Brien KC, McAvoy M, Raichle ME, Morris JC, Buckner
RL (2003) Functional deactivations: change with age and dementia of the Alzheimer
type. Proceedings of the National Academy of Sciences of the United States of America
100:14504–14509.
Ly M, Canu E, Xu G, Oh J, McLaren DG, Dowling NM, Alexander AL, Sager MA, Johnson
SC, Bendlin BB (2014) Midlife measurements of white matter microstructure predict
subsequent regional white matter atrophy in healthy adults. Human Brain Mapping
35:2044–2054.
Melka MG, Gillis J, Bernard M, Abrahamowicz M, Chakravarty MM, Leonard GT, Perron
M, Richer L, Veillette S, Banaschewski T, Barker GJ, Buchel C, Conrod P, Flor H,
Heinz A, Garavan H, Bruhl R, Mann K, Artiges E, Lourdusamy A, Lathrop M, Loth E,
Schwartz Y, Frouin V, Rietschel M, Smolka MN, Strohle A, Gallinat J, Struve M, Lattka
E, Waldenberger M, Schumann G, Pavlidis P, Gaudet D, Paus T, Pausova Z (2013) FTO,
obesity and the adolescent brain. Human Molecular Genetics 22:1050–1058.
Meunier D, Stamatakis EA, Tyler LK (2014) Age-related functional reorganization, structural
changes, and preserved cognition. Neurobiology of Aging 35:42–54.
Mowinckel AM, Espeseth T, Westlye LT (2012) Network-specific effects of age and in-scanner
subject motion: a resting-state fMRI study of 238 healthy adults. NeuroImage 63:1364–1373.
Nevalainen N, Riklund K, Andersson M, Axelsson J, Ogren M, Lovden M, Lindenberger
U, Backman L, Nyberg L (2015) COBRA: A prospective multimodal imaging study of
dopamine, brain structure and function, and cognition. Brain Research 1612:83–103. doi:
10.1016/j.brainres.2014.09.010.
maintenance. Trends in Cognitive Sciences 16:292–305.
Onoda K, Ishihara M, Yamaguchi S (2012) Decreased functional connectivity by aging is asso-
ciated with cognitive decline. Journal of cognitive neuroscience 24:2186–2198.
Pakkenberg B, Pelvig D, Marner L, Bundgaard MJ, Gundersen HJ, Nyengaard JR, Regeur L
(2003) Aging and the human neocortex. Experimental Gerontology 38:95–99.
Persson J, Pudas S, Nilsson LG, Nyberg L (2014) Longitudinal assessment of default-mode
brain function in aging. Neurobiology of Aging 35:2107–2117.
Peters A, Morrison JH, Rosene DL, Hyman BT (1998) Feature article: are neurons lost from the
primate cerebral cortex during normal aging? Cerebral Cortex 8:295–300.
Raz N, Lindenberger U (2011) Psychol Bull 137(5):790–795.
Reiman EM, Quiroz YT, Fleisher AS, Chen K, Velez-Pardo C, Jimenez-Del-Rio M, Fagan AM,
Shah AR, Alvarez S, Arbelaez A, Giraldo M, Acosta-Baena N, Sperling RA, Dickerson B,
Stern CE, Tirado V, Munoz C, Reiman RA, Huentelman MJ, Alexander GE, Langbaum
JB, Kosik KS, Tariot PN, Lopera F (2012) Brain imaging and fluid biomarker analysis in
young adults at genetic risk for autosomal dominant Alzheimer’s disease in the presenilin
1 E280A kindred: a case-control study. The Lancet Neurology 11:1048–1056.
Reitz C, Tosto G, Mayeux R, Luchsinger JA, Group N- LNFS, Alzheimer’s Disease
Neuroimaging I (2012) Genetic variants in the Fat and Obesity Associated (FTO) gene
and risk of Alzheimer’s disease. PloS One 7:e50354.
152
Salat DH, Buckner RL, Snyder AZ, Greve DN, Desikan RS, Busa E, Morris JC, Dale AM,
Fischl B (2004) Thinning of the cerebral cortex in aging. Cerebral cortex 14:721–730.
Salat DH, Tuch DS, Hevelone ND, Fischl B, Corkin S, Rosas HD, Dale AM (2005) Age-related
changes in prefrontal white matter measured by diffusion tensor imaging. Annals of the
New York Academy of Sciences 1064:37–49.
Sexton C, Walhovd KB, Storsve AB, Tamnes CK, Westlye LT, Johansen- Berg H, Fjell
AM (2014) Accelerated Changes in White Matter Microstructure During Ageing: A
Longitudinal Diffusion Tensor Imaging Study. J Neurosci. 34(46):15425–15436.
Sheline YI, Morris JC, Snyder AZ, Price JL, Yan Z, D’Angelo G, Liu C, Dixit S, Benzinger
T, Fagan A, Goate A, Mintun MA (2010) APOE4 allele disrupts resting state fMRI con-
nectivity in the absence of amyloid plaques or decreased CSF Abeta42. The Journal of
Neuroscience: The Official Journal of the Society for Neuroscience 30:17035–17040.
Sheline YI, Raichle ME (2013) Resting state functional connectivity in preclinical Alzheimer’s
disease. Biological Psychiatry 74:340–347.
Song SK, Sun SW, Ju WK, Lin SJ, Cross AH, Neufeld AH (2003) Diffusion tensor imaging
detects and differentiates axon and myelin degeneration in mouse optic nerve after retinal
ischemia. NeuroImage 20:1714–1722.
Storsve AB, Fjell AM, Tamnes CK, Westlye LT, Overbye K, Aasland HW, Walhovd KB (2014)
Differential longitudinal changes in cortical thickness, surface area and volume across the
adult life span: regions of accelerating and decelerating change. The Journal of neurosci-
Stricker NH, Salat DH, Foley JM, Zink TA, Kellison IL, McFarland CP, Grande LJ, McGlinchey
RE, Milberg WP, Leritz EC (2013) Decreased white matter integrity in neuropsychologi-
cally defined mild cognitive impairment is independent of cortical thinning. Journal of the
International Neuropsychological Society: JINS 19:925–937.
Teipel SJ, Bokde AL, Meindl T, Amaro E, Jr., Soldner J, Reiser MF, Herpertz SC, Moller HJ,
tivity in the human brain. NeuroImage 49:2021–2032.
Tomasi D, Volkow ND (2012) Aging and functional brain networks. Molecular Psychiatry
17:471, 549–558.
Vandenberghe R, Adamczuk K, Dupont P, Laere KV, Chetelat G (2013) Amyloid PET in clini-
cal practice: Its place in the multidimensional space of Alzheimer’s disease. NeuroImage
Clinical 2:497–511.
Villain N, Fouquet M, Baron JC, Mezenge F, Landeau B, de La Sayette V, Viader F, Eustache F,
Desgranges B, Chetelat G (2010) Sequential relationships between grey matter and white
matter atrophy and brain metabolic abnormalities in early Alzheimer’s disease. Brain: A
Journal of Neurology 133:3301–3314.
Walhovd KB, Fjell AM, Brewer J, McEvoy LK, Fennema-Notestine C, Hagler DJ, Jr.,
Jennings RG, Karow D, Dale AM, Alzheimer’s Disease Neuroimaging I (2010)
Combining MR imaging, positron-emission tomography, and CSF biomarkers in the diag-
nosis and prognosis of Alzheimer disease. AJNR American Journal of Neuroradiology
31:347–354.
Walhovd KB, Fjell AM, Brown TT, Kuperman JM, Chung Y, Hagler DJ, Jr., Roddey JC,
Erhart M, McCabe C, Akshoomoff N, Amaral DG, Bloss CS, Libiger O, Schork
NJ, Darst BF, Casey BJ, Chang L, Ernst TM, Frazier J, Gruen JR, Kaufmann WE,
Murray SS, van Zijl P, Mostofsky S, Dale AM, Pediatric Imaging N, Genetics S (2012)
Long-term influence of normal variation in neonatal characteristics on human brain
153
development. Proceedings of the National Academy of Sciences of the United States of

America 109:20089–20094.
cal structures. Neurobiology of Aging 26:1261–1270; discussion 1275–1268.
Ward AM, Mormino EC, Hujbers W, Schultz AP, Hedden T, Sperling RA (2015) Relationships
between default-mode network connectivity, medial temporal lobe structure, and age-
related memory deficits. Neurobiol Aging 36(1):265–272.
Weiner MW, Veitch DP, Aisen PS, Beckett LA, Cairns NJ, Green RC, Harvey D, Jack CR,
Jagust W, Liu E, Morris JC, Petersen RC, Saykin AJ, Schmidt ME, Shaw L, Siuciak
JA, Soares H, Toga AW, Trojanowski JQ, Alzheimer’s Disease Neuroimaging I (2012)
The Alzheimer’s Disease Neuroimaging Initiative: a review of papers published since
its inception. Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association
8:S1–S68.
Westlye LT, Grydeland H, Walhovd KB, Fjell AM (2011) Associations between regional corti-
cal thickness and attentional networks as measured by the attention network test. Cerebral
Cortex 21:345–356.
Tamnes CK, Ostby Y, Fjell AM (2010a) Differentiating maturational and aging-related
changes of the cerebral cortex by use of thickness and signal intensity. NeuroImage
52:172–185.
Tamnes CK, Ostby Y, Fjell AM (2010b) Life-span changes of the human brain white mat-
ter: diffusion tensor imaging (DTI) and volumetry. Cerebral Cortex 20:2055–2068.
Wirth M, Madison CM, Rabinovici GD, Oh H, Landau SM, Jagust WJ (2013a) Alzheimer’s dis-
ease neurodegenerative biomarkers are associated with decreased cognitive function but
not beta-amyloid in cognitively normal older individuals. The Journal of neuroscience: the
official journal of the Society for Neuroscience 33:5553–5563.
Wirth M, Oh H, Mormino EC, Markley C, Landau SM, Jagust WJ (2013b) The effect of
amyloid beta on cognitive decline is modulated by neural integrity in cognitively nor-
mal elderly. Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association
9:687–698 e681.
Zatorre RJ, Fields RD, Johansen-Berg H (2012) Plasticity in gray and white: neuroimaging
changes in brain structure during learning. Nature Neuroscience 15:528–536.
154
â•‡ 155
Structural and Functional

Imaging ofÂ€Aging
Longitudinal Studies
Lars Nyberg
Sara Pudas
Anders Lundquist
Introduction—â•‰Why Longitudinal Studies?
Aging is in many ways a highly complex process to study. From a measurement per-
spective, one of the main challenges is its temporal extension. Important insights into
longevity have been provided by studies of, for example, flies and worms where the
lifespan is measured in days (see e.g., Vaupel, 1998). In humans, by contrast, where
many individuals in countries like Sweden and Japan, but also in an increasingly
number of additional countries, can expect to live 100 years or more (Vaupel, 2010),
the temporal extension of aging poses greater challenges to researchers. Therefore,
it is not surprising that the vast majority of studies pertaining to examine aspects
of human aging rely on a cross-â•‰sectional design. No doubt, many cross-â•‰sectional
studies, spanning diverse scientific fields, have yielded important knowledge. At the
same time, by now, it is generally agreed that attempts to make inferences about
age-â•‰related changes on basis of cross-â•‰sectional studies of age differences may lead
to erroneous conclusions (e.g., Lindenberger, von Oertzen, Ghisletta, & Hertzog,
2011). One example of a marked discrepancy between cross-â•‰sectional and longitudi-
nal results comes from a study of age-â•‰related decline in episodic memory (Rönnlund
et al., 2005). As shown in Figure 6.1, analyses of cross-â•‰sectional data from over 800
individuals indicated an early â•‰onset of decline (cf., e.g., Park et al., 2002; Salthouse,
2004; Singh-â•‰Manoux et al., 2012), whereas corresponding analyses of longitudinal
data revealed relative stability of episodic memory until age 60 years (cf., Schaie,
1994). It is well known that cohort effects can severely influence cross-â•‰sectional
155
156
Episodic memory performance
30 60 90
age
Figure 6.1â•‡Schematic illustration of cross-â•‰sectional (dashed) and longitudinal (solid) esti-

mates of age-â•‰related reductions in episodic memory (cf., Rönnlund et al., 2005)
comparisons, and Rönnlund and colleagues (2005) showed that cohort â•‰differences in
educational attainment largely accounted for the discrepancy between cross-â•‰sectional
and longitudinal data in this instance.
Thus, longitudinal analyses can yield markedly different conclusions on epi-
sodic-â•‰memory change compared to what a cross-â•‰sectional analysis would suggest.
Needless to say, discrepancies are far from always as striking, and as will be discussed
later, longitudinal studies are associated with their own methodological challenges.
Nevertheless, there is an emerging call for longitudinal designs in aging studies, not
least in the field of cognitive neuroscience and brain imaging where such studies are
particularly scarce. A major purpose of the present chapter is to review structural
and functional imaging studies of brain aging, with the goal of extracting representa-
tive longitudinal age gradients where possible. A second purpose is to consider key
methodological issues in longitudinal imaging, including influences on the findings
of attrition/â•‰dropâ•‰out, change in hardware over time, and choice of statistical models.
Finally we present some suggestions of what will be important to consider in future
longitudinal brain-â•‰imaging studies, including the topic of joint analyses of age-â•‰related
brain-â•‰behavior changes.
Longitudinal Brain Imaging Studies
Most results to be presented and discussed in this section emanate from functional and
structural MRI studies. The non â•‰invasive nature of MRI lends itself well to designs
with repeated measurements over time. However, as presented below, some influential
longitudinal studies, notably the Baltimore Longitudinal Study of Aging (Resnick et al.,
2000), involve repeated measurement of brain blood flow with PET. PET also offers
unique opportunities to study how various neurotransmitter systems are altered in aging,
such as the dopamine system (e.g., Bäckman et al., 2000, 2010). In addition, PET allows
quantification of the accumulation of beta-â•‰amyloid in the brain (Klunk et al., 2004;
â•‡ 157
Structural and Functional Imaging ofÂ€Agingâ•… 157
Mormino et al., 2009; Rodrigue et al., 2009). However, with only a few exceptions (e.g.,
Sojkova et al., 2011), invasive PET applications have not been used in longitudinal stud-
ies of normal aging, and will therefore not be reviewed here.
Structural Brain Imaging

Gray â•‰Matter Changes
Cross-â•‰sectional studies of age differences in (regional) brain volumes appeared a cou-
ple of decades ago (e.g., Blatter et al., 1995), and were soon followed by longitudinal
studies. Initially, the longitudinal volumetric studies were related to early prediction
of dementia (e.g., Fox et al., 1996), but subsequent studies focused on changes in
normal aging (e.g., Tang et al., 2001). By now, numerous cross-â•‰sectional (e.g., Good
et al., 2001; Walhovd et al., 2005; Fotenos et al., 2008; Fjell et al., 2009; Kalpouzos
et al., 2009) and longitudinal (e.g., Resnick et al., 2000, 2003; Rodrigue and Raz,
2004; Raz et al., 2005, 2007; Taki et al., 2013; Jiang et al., 2014) studies of how brain
volume is altered with increasing age have been published, and the findings of these
studies have been thoroughly reviewed elsewhere (e.g., Raz and Rodrigue, 2006; Fjell
and Walhovd, 2010; Salthouse, 2011).
Findings from cross-â•‰ sectional studies of regional or whole-â•‰ brain volumetric
changes have been taken as evidence for early onset shrinkage, with an approximately
linear negative age relation from early adulthood in whole-â•‰brain and gray matter vol-
ume (Fig. 6.2 left; for a review and discussion, see e.g., Salthouse, 2011). In one
large-â•‰scale study, Sowell et al. (2003) mapped gray-matter density in 176 individuals
between 7–â•‰87 years of age. Several different cortical areas were examined, and with
only one exception (left posterior temporal cortex) the gray-matter density loss was
most rapid during the first six decades of life. This pattern might be seen as consis-
tent with the notion of an early-â•‰onset of gray-â•‰matter changes. However, as noted by
Sowell and colleagues, the driving factor behind the apparent rapid early decline in
gray-â•‰matter density was likely an increase in cerebral myelination (see further the
below section on white-â•‰matter changes). In another study, Fotenos et al., (2005) used
1.0 Cross-sectional 1.0 Longitudinal
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
20 30 40 50 60 70 80 90 20 30 40 50 60 70 80 90
Age Age
Figure 6.2â•‡ Schematic age—â•‰gray-â•‰matter mappings based on cross-â•‰sectional (left) and longitu-
dinal (right) analyses.
158
MRI to cross-sectionally examine the brains of 370 adults between 18 to 97 years at

baseline. They found that whole-brain volume differences as well as normalized gray-
matter differences were detected by age 30. Importantly, though, greater whole-brain
reductions were noted in older adults, and the age-by gray-matter-volume correlation
was much weaker when the age range was restricted to 18 to 30 (r = −0.20) than when
the 30–95 year olds also were included (r = −0.91).
In a review of 56 longitudinal MRI studies, Hedman and colleagues (2012)
showed whole-brain increases throughout the 20’s, and stability until age 35. Later,
the longitudinal data indicated modest shrinkage between age 35 to 60 (0.2%–0.5%
per year), with a more pronounced decline thereafter (> 0.5% annual volume loss).
Corresponding analyses of gray-matter changes were based on 10 longitudinal studies,
and broadly converged with the whole-brain data in showing minor changes between
20 to approximately 50, and marked changes thereafter (Fig. 6.2 right; cf., Hedman
et al., 2012, Fig. 4A). Taken together, for both whole-brain and regional gray-matter
volumes, the review of longitudinal MRI studies indicated a later onset of shrinkage
than was suggested by some previous cross-sectional studies.
A few reports have presented direct, within-study comparisons of cross-sectional
and longitudinal estimations of how the brain is shrinking with advancing age (Raz
et al., 2005; Nyberg et al., 2010; Fjell et al., 2014). Consistently, these comparisons
found that the longitudinal estimates of age changes were substantially higher than
the corresponding cross-sectional estimates, which likely reflected less noise in the
longitudinal, within-person, analyses. Qualitatively, with a noteworthy exception for
a cross-sectional finding of preservation or even thickening of the anterior cingulate
cortex along with a corresponding longitudinal observation of thinning in the same
region (Fjell et al., 2014), there were marked similarities in the findings from the
cross-sectional and longitudinal analyses reported in these studies. In the study by
Fjell and colleagues (2014), longitudinal data were restricted to participants between
60 to 93 years. For this age-segment, both the cross-sectional and longitudinal esti-
mates suggested that cortical thickness in most areas declines linearly with age. This
is consistent with the observation by Rönnlund et al (2005) that cross-sectional and
longitudinal estimates of episodic-memory change were similar for groups 60 to
85 years old. Critically, in the Rönnlund et al. study, the deviation between cross-
sectional and longitudinal estimates was seen for the 35 to 60 age span. Similarly,
in the Raz et al. (2005) study, in which participants with a baseline age of 20 to 77
years were included, the longitudinal estimates indicated nonlinear regional brain
aging with the mid-fifties being an approximate point of inflection of age trends.
Accelerated shrinking was observed in several regions, including the entorhinal cor-
tex, an effect that also was seen in the Fjell et al. (2014) study (see also Du et al.,
2006).
In summary, the findings from (predominantly) cross-sectional studies have been
taken as evidence for approximately linear negative age relations from early adulthood
in gray matter volume (see e.g., Salthouse, 2011), although some cross-sectional stud-
ies found evidence for nonlinear trajectories over the adult lifespan (see e.g., Ziegler
et al., 2012). By contrast, longitudinal studies strongly suggest a nonlinear pattern of
gray-matter changes across the adult lifespan with relative stability between approxi-
mately 35 to 55 years. At older ages, there seems to be quite good correspondence
â•‡ 159
between both analytic approaches, with an approximately linear age-â•‰volume mapping

function (cf., Fjell et al., 2014; Fotenos et al., 2005).
White â•‰Matter (WM) Changes

As reviewed by Fjell and Walhovd (2010), some studies found that brain WM, as
measured by volumetric and diffusion tensor imaging (DTI) methods, is fairly stable
in adulthood and aging (e.g., Good et al., 2001; Sullivan, 2004; Abe et al., 2008),
whereas others suggested a linear decline in white â•‰matter volume. Additional larger-â•‰
scale studies provided evidence for an inverted U-â•‰shape pattern with a long matura-
tional phase of WM increases, a stable phase with no or little decline in WM, and
a phase characterized by accelerated decline with increasing age (e.g., Salat et al.,
2009; Walhovd et al., 2009; Kochunov et al., 2010; Westlye et al., 2010; Lebel et al.,
2012; Salami et al., 2012a). While there is some variation among these cross-â•‰sectional
studies regarding regional/â•‰tract-â•‰specific patterns of change and when the develop-
mental trajectory peaks, as well as in different MRI-â•‰based measures of WM â•‰integrity
(Burzynska et al., 2010), these studies converge on a lifeâ•‰span pattern of WM change
as schematically outlined in Figure 6.3—â•‰left (for a review, see Yap et al., 2013).
The number of longitudinal studies of WM change is still limited, and most
studies had a fairly short time interval between measurement points (for relevant
examples, see Table 6.1). With few exceptions, the results from longitudinal studies
converge with those from cross-â•‰sectional studies in showing significant age changes
in various aspects of brain WM (see also Fletcher et al., 2013), and several of the
studies summarized in Table 6.1 found that the longitudinal method was more sensi-
tive in detecting an effect. In relation to the accelerated WM reduction with increas-
ing age that was suggested by cross-â•‰sectional studies (cf., Fig 6.3 left), Barrick et
al. (2010) found no evidence for accelerated longitudinal frontal decline. One study
even observed decelerated change with increasing age for one DTI measure (MD;
Lövdén et al., 2014), but that sample was rather old and not evenly distributed over
0.6
0.5
0.5
0.4 0.4
0.3
0 10 20 30 40 50 60 70 80 90 20 40 60 80
Age Age
Figure 6.3â•‡Left—â•‰schematic illustration of cross-â•‰sectional age-â•‰related differences in brain

white matter (fractional anisotropy [FA]; based on Westlye et al., 2010). Right—â•‰example of
longitudinal change in frontal white matter (FA; based on Sexton et al., 2014).
160
Table 6.1 Longitudinal studies of age-related change in brain white matter
Author MRI methods Sample size Age (baseline) Study Sessions & Main findings
interval
Sullivan et al. (2002) 1.5T 215 baseline & 71.9 +/- 2.7 NHLBI 2, 4 years Significant change for callosal regions (0.9% annual
Callosal size follow-up Twin thinning)
Barrick et al. (2010) 1.5T 99 baseline 50–90 GENIE 2, 2 years Longitudinal change throughout the brain; no evidence
DTI, 12 73 follow-up M = 69 for accelerated decline in frontal regions. Longitudinal
directions effect > cross-sectional
Charlton et al. (2010) 1.5 T 99 baseline 50–90 GENIE 2, 2 years Significant change in all MRI measures
DTI, 12 dir. 73–80 follow-up M = 69
WMH, volume
Nyberg et al. (2010) 1.5T 60 baseline 49–79 Betula 2, 6 years Longitudinal change throughout the brain, notably in
DTI, 6 dir. 38 follow-up M = 66 corpus callosum. Longitudinal effect > cross-sectional
Sullivan et al. (2010) 3T 20 baseline 20–30 and — 2, 2 years Cross-sectional DTI difference;
DTI 6 dir. 16 follow-up 60–70 No longitudinal change
volume (8/age group)
Teipel et al. (2010) 3T 11 (+ 14 with MCI) 59–83 — 2, 13–16 months Significant decline in intracortical projecting fiber tracts
DTI, 12 dir. M = 67
Ly et al. (2014) 3T 44 baseline and 42–75 — 2, 3.5 years Volume reduction on frontal, temporal, and cerebellar
volume, DTI follow-up M = 56 regions; reduction predicted by baseline FA
12 dir.
Lövdén et al. (2014) 1.5T 77 baseline 81–103 SNAC-K 2, 2.3 years Analysis of 6 ROI’s revealed decreases for most tracts;
DTI 6 dir. 40 follow-up Longitudinal effect > cross-sectional
Sexton et al. (2014) 1.5T 203 baseline and 20–84 CPtL 2, 3.6 years Decline began in the 5th decade and increased with age
DTI 30 dir. follow-up
Note: T = tesla; DTI = diffusion tensor imaging; M = mean; WMH = white matter hyperintensities; MRI = magnetic resonance imaging; dir. = directions; MCI = mild cognitive
impairment; FA = fractional anisotrophy; ROI = region of interest. Abbreviations of study names: NHLBI twin = National Heart, Lung, and Blood Institute Twin Study;
GENIE = St George’s Neuropsychology and Imaging in the Elderly; SNAC-K = Swedish National study on Aging and Care in Kungsholmen; CPtL = Cognition and Plasticity
through the Lifespan. See the respective reference for full study information.
â•‡ 161
the age range, which might have contributed to this rare pattern of change. A recent
longitudinal study by Sexton and (2014) examined WM changes (FA, AD, RD, MD)
in a sample of 203 adults between 20 and 84 years, and found that age-â•‰related decline
became apparent in the fifth decade and thereafter showed accelerated change with
increasing age (Fig 6.3—â•‰right).
Taken together, as was seen for gray matter, longitudinal analyses of WM appear
to be more sensitive than cross-â•‰sectional WM analyses, but qualitatively these ana-
lytic approaches broadly converge on relative stability in middle age
â•‰ and accelerated
decline in older age.
Functional Brain Imaging

In contrast to the literature on structural brain changes in aging, there are relatively
few longitudinal functional-â•‰activation studies published to date. Many of the exist-
ing ones originate from two samples; the neuroimaging sub-â•‰studies of the Baltimore
Longitudinal Study on Aging (Resnick et al., 2000) and the Betula study (Nyberg
et al., 2010). Since longitudinal age gradients for functional activation have generally
not been reported (but see Nyberg et al., 2010), and due to the scarcity and diver-
sity of studies, the findings will be discussed in terms of age-â•‰related increases and
decreases over time rather than presented as age-â•‰gradients in the same manner as for
the structural studies. A full literature review will not be attempted; instead this sec-
tion highlights representative and large-â•‰scale cross-â•‰sectional age gradients of brain
activity, and contrasts them with findings from the few published longitudinal studies
to date. The focus will be on the hippocampus and prefrontal cortex (PFC), as well as
the default mode network (DMN), as these regions have been the topic for the major-
ity of published longitudinal studies to date.
Task-â•‰Based Activation Studies

Figure 6.4 illustrates average cross-â•‰sectional decline in hippocampal activity in a large
sample of representative older adults (N = 292; Salami et al., 2012b). Although such
age-â•‰related hippocampal underactivation has been demonstrated in cross-â•‰sectional
age-â•‰comparisons (e.g., Gutchess et al., 2005), there are also numerous reports of
equivalent activation in younger and older individual (e.g., Persson et al., 2011), as
well as hippocampal overactivation in older individuals with mild cognitive impair-
ment (Celone et al., 2006). Two early longitudinal functional imaging studies sug-
gested modestly increased hippocampal activation over a 9-â•‰year period (Beason-â•‰Held
et al., 2008a, 2008b). But the small (N = 25) and highly selected (cognitively stable
and physically healthy) sample in these studies cannot be taken to represent a gen-
eral age-â•‰trend. Two more recent longitudinal studies, with more cognitively diverse
samples (O’Brien et al., 2010; Persson et al., 2012) have demonstrated significant
associations between hippocampal decline and cognitive decline over time, although
no overall reductions in the hippocampus were detected across a sample overlapping
that of the second study (Nyberg et al., 2010). Taken together, the longitudinal studies
on hippocampal function in aging have yielded varying findings, a likely reflection
of the large inter-â•‰individual differences present in older samples (e.g., Wilson et al.,
162
Left hippocampus Right hippocampus

0.20 0.25
0.15 0.20
0.10 0.15
0.05 0.10
0 0.05
–0.05 0
–0.10 –0.05
20 30 40 50 60 70 80 20 30 40 50 60 70 80
Age Age
Figure 6.4 Cross-sectional estimates of age-related decline in hippocampal activity during epi-
sodic memory encoding. The sample comprised 292 individuals aged 20–80 years, divided into
8 age groups. The blue clusters show the main effect of encoding across the sample, whereas
the age-effect is shown in red. Reproduced with data from Salami et al., 2012, Journal of
Neuroscience. (See color plate also)
2002). A larger degree of cognitive decline appears to be associated with more hip-
pocampal decline, whereas carefully characterized successful agers have been shown
to have spared hippocampal function into old age (Pudas et al., 2013). Thus, the cross-
sectional age gradients in Figure 6.4 represent the best available approximation of
average age-related changes in hippocampal function to date, but remain to be quali-
fied by larger scale longitudinal studies with representative samples.
When it comes to the prefrontal cortex, prior reviews show that the literature
on cross-sectional age differences in frontal cortex activation is characterized by
findings of both underrecruitment and over-recruitment in older compared to young
adults, across a variety of cognitive domains (for reviews, see Rajah and D’Esposito,
2005; Persson and Nyberg, 2006; Dennis and Cabeza, 2008). Over-recruitment has
received more attention, due to the possibility that it may reflect a compensatory
response in the older brain (Cabeza et al., 2002; Park and Reuter-Lorenz, 2009),
but it has also been suggested to be a reflection of more nonselective recruitment
(Logan et al., 2002) and has sometimes been found to be associated with declining
cognition (Persson et al., 2006). A more recent large-scale cross-sectional study sug-
gested that increased and decreased activation with age may co-occur in the frontal
cortex (Salami et al., 2012b). The picture has also been refined by cross-sectional
findings indicating that older adults may display over-recruitment during easier cog-
nitive tasks, whereas more difficult conditions elicit age-related under-recruitment
(Mattay et al., 2006; Reuter-Lorenz and Cappell, 2008; Nyberg et al., 2014). Figure
6.5 shows cross-sectional age gradients in a dorsolateral PFC region, which are posi-
tive during an easier task condition and negative during a more demanding one. The
163
Structural and Functional Imaging of Aging 163
(A)
(B) 0.60
0.50
0.25
% signal change
0.40
0.20
% signal change
25–50 0.30
0.15 55–60
65–70 0.20
0.10
75–80
0.05 0.10
0.00 0.00
Maintenance Manipulation vs. Maintenance
Figure 6.5 Cross-sectional estimates of age-effects in the frontal cortex in a right-sided dorso-
lateral cluster, across a representative sample of 287 participants. During an easier task condi-
tion, requiring only maintenance of information in working memory, increased activity with
age is seen (left panel). A more difficult condition, also involving manipulation of information,
instead elicited diminishing activity with age. Reprinted with permission from Nyberg et al.,
2013, Journal of Cognitive Neuroscience.
co-occurrence of frontal increases and decreases over time within a cognitively stable
sample was supported by two early longitudinal studies (Beason-Held et al., 2008a,
2008b), but a subsequent longitudinal study by Nyberg et al., (2010) raised doubts
about the validity of cross-sectional over-recruitment in older adults. The latter study
is one of few studies to report a within-study comparison between longitudinal and
cross-sectional analyses of the same dataset, which revealed a marked discrepancy
between cross-sectional and longitudinal estimates. Six-year longitudinal fMRI data
suggested an age-related decline in right frontal recruitment, which accelerated with
increasing age (Figure 6.6A). By contrast, cross-sectional results from two time
points indicated an age-related increase in frontal cortex activity during incidental
episodic encoding (Figure 6.6B). Further analyses provided evidence that the inclu-
sion of a high-functioning elderly sample biased the cross-sectional results in the
direction of frontal over-recruitment—but when the same high-functioning elderly
164
(A)
(B) (C)
0.6
0.2
Beta change 2003–2008
0 0.4
–0.2
Beta values
0.2
–0.4
0
–0.6
–0.2
–0.8
–0.4
45 50 55 60 65 70 75 80 85 45 50 55 60 65 70 75 80 85
Age Age
Figure 6.6 Discrepancy between longitudinal (panel A) and cross-sectional (panel B) age-effects
on activity in the right dorsolateral frontal cortex. Reproduced with permission from Nyberg et al.,
2010, Proceedings of the National Academy of Sciences of the United States of America. (See
color plate also)
were examined longitudinally an activity reduction was apparent. Hence, this study
suggests that some cross-sectional demonstrations of age-related over-recruitment
may in fact be artifacts of sampling bias.
Subsequently, increases in frontal recruitment over time have been demonstrated
within a large longitudinal dataset (N = 99; Goh et al., 2013), and were found to
be associated with declining cognitive functions over time. Although the direction
of the effect is opposite to that reported by Nyberg et al. (2010), it should be noted
that the negative association with cognitive-performance change appears to contradict
the common notion of beneficial compensatory age-related frontal increases from the
cross-sectional literature. The discrepancy between the Nyberg et al. and Goh et al.
studies could be partially explained by the fact that Goh et al. only reported correla-
tions between change in cognitive function and change in brain activity, and did not
investigate or report a main effect of time across the whole sample. These two studies
also differ in their conclusions regarding a common issue in longitudinal datasets,
namely practice effects. While Nyberg et al. (2010) specifically tested and rejected
the possibility that the frontal activity reduction over time was a result of practice
â•‡ 165
effects, Goh et al. (2013) on the other hand explained their association between pre-
served/â•‰improving cognitive performance and decline in frontal recruitment as a result
of practice. The reasoning was that less neural resources were needed on subsequent
task executions due to practice, and that this effect was more pronounced among those
with preserved cognition. Taken together, it appears as if both increases and decreases
in frontal recruitment with age are possible. To fully understand their significance,
factors such as task difficulty, performance level, and practice effects need to be sys-
tematically investigated in longitudinal, rather than cross-â•‰sectional designs, in order
to assure that true age-â•‰related changes are captured.
Default Mode Network Studies

Cross-â•‰sectional studies have indicated rather prominent effects of age within the
default mode network (DMN; Lustig et al., 2003; Grady et al., 2006), with decreased
magnitude of deactivation with increased age (see Figure 6.7). In contrast, a longitudi-
nal study by Beason-â•‰Held et al. (2009) emphasized 9-â•‰year stability in both the spatial
extent and magnitude of deactivation in the major components of the DMN regions
in a sample of healthy and cognitively stable older adults. A great degree of stability
of deactivations was also found in a subsequent longitudinal study by Persson et al.
(2014), but a decrease in the magnitude of deactivation was found in the lateral parietal
cortex across 6 years. Interestingly, Persson et al. also reported a within-â•‰study com-
parison of cross-â•‰sectional and longitudinal data, which suggested improved sensitiv-
ity for the longitudinal approach. Specifically, the cross-â•‰sectional analyses (across the
age-â•‰span 55–â•‰84 years) failed to detect any effects of age, although this may partially be
attributed to averaging deactivation magnitudes across four regions-â•‰of-â•‰interest, poten-
tially attenuating associations between age and deactivation. The apparent discrepancy
between these cross-â•‰sectional results and prior studies (Lustig et al., 2003; Grady et
al., 2006) may be explained by the longer age-â•‰span covered in the earlier studies, with
more than 50 years separating the youngest and oldest participants, compared to 30
years in the study by Persson et al. The greater length of the age-â•‰span may also be one
explanation of why the cross-â•‰sectional studies reported more prominent age-â•‰effects,
while the longitudinal studies with 6–â•‰9 year follow-â•‰up intervals instead emphasized
stability. Larger age-â•‰effects would presumably be observed across 50 years than 6–â•‰9
years. Another explanation of smaller age-â•‰effects in longitudinal studies could be selec-
tive attrition, which may render the samples healthier and more well-â•‰functioning than
average, hence obstructing detection of significant age-â•‰effects. However, despite these
obstacles, significant longitudinal age-â•‰effects were detected by Persson et al. (2014),
which may speak for the sensitivity of the longitudinal approach.
In summary, this section contrasted cross-â•‰sectional and longitudinal findings
from functional neuroimaging studies of aging. The diversity of the longitudinal
findings underscore the complexity of the subject matter, and extracting general
age-â•‰gradients based on the few available studies would be premature. Larger-â•‰scale
longitudinal studies with representative samples are needed to target such questions,
given that individual differences in cognitive aging appears to be a critical factor
determining age-â•‰related changes in brain activity (O’Brien et al., 2010; Persson et
al., 2012; Beason-â•‰Held et al., 2013; Goh et al., 2013). Most critically, given evi-
dence of discrepancies between cross-â•‰sectional and longitudinal analyses of the
166
z=8 z = 24 z = 32
(A) (B)
30 20
20 10
Brain score
Brain score
10 0
0
–10
–10
–20
–20
–30
10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90
Encoding tasks Recognition tasks
Figure 6.7â•‡ Scatter plots indicate significant cross-â•‰sectional age-â•‰related increases in a brain
network comprising medial frontal gyrus, precuneus, posterior cingulate cortex, cuneus, and
middle occipital gyrus (yellow/â•‰orange clusters), as indicated by higher brain scores. Conversely,
younger adults had higher activity in a network comprising middle and inferior frontal gyrus,
caudate nucleus, putamen and fusiform gyrus (blue brain clusters), as indicated by lower brain
scores. The changes are common to various task conditions during memory encoding and
retrieval. Reprinted by permission of MIT Press Journals, from Grady et al. (2006), Journal of
Cognitive Neuroscience. (See color plate also)
same dataset (Nyberg et al., 2010), it is clearly inappropriate to rely on cross-â•‰sec-

tional data alone.
Methodological Issues in Longitudinal Imaging
Naturally, there are many methodological aspects to consider when designing a lon-
gitudinal imaging study, so the topics presented here constitute a selection. We first
â•‡ 167
give a short background and motivation for our chosen topics. To begin with, any
longitudinal study involving human subjects and a follow-â•‰up time which is counted
in months or years rather than days or weeks will most likely experience subjects
who drop out of the study for various reasons, and thus the first order of business
is to consider dropout. Hopefully, though, most subjects remain in the study for the
follow-â•‰ups, going through imaging and commonly also some behavioral testing at
each follow-â•‰up. Results on repeated testing may be affected by the test-â•‰retest effect,
so some attention is given to the test-â•‰retest effects. After data collection follows data
analysis, in research studies as well as this chapter, and we review some key aspects of
the statistical analysis of longitudinal (imaging) studies. Finally, not only the subjects
may change or drop out during the course of a study. Rather, it may well be the case
that the scanner itself is upgraded or simply replaced, which gives us reason to discuss
findings regarding the reliability of repeated measurements when there is a change in
the scanning equipment.
Dropout and Re-â•‰test Effects
When performing a longitudinal study, dropout (sometimes called attrition, the terms
are used interchangeably) will almost certainly occur, unless the time to follow-â•‰up is
very short. Dropout is a special case of more general data missingness; it is possible
that a subject misses one follow-â•‰up but comes back again for the next one. When
a subject drops out, he or she does not reâ•‰enter the study at a later time. Dropout is
the most common structure of missing data in longitudinal studies involving human
subjects. We will normally use the more general term missingness, since definitions
and methods are valid for any missing data structure, not just dropout. Further, miss-
ingness can only arise from subjects who have entered into the study at baseline. If
subjects are excluded at baseline, e.g., for not meeting screening criteria, this is nor-
mally not regarded as missingness but rather a part in defining the study’s baseline
target population, e.g., “healthy 60-â•‰year-olds.” Results from a study with a strictly
selected baseline target population are most likely biased with respect to the general
population, which must be acknowledged when discussing the results. Such a bias
will however be present even if there is no missingness in the study itself, so it not a
problem related to missingness. As we shall see, missingness may cause results to be
biased even with respect to the baseline target population. Once a subject has entered
into the study, failure to meet screening criteria at follow-â•‰up is, however, a source of
missingness.
In general, missing data can cause many issues for subsequent data analysis, here
we mention some of the more important ones. To begin with simple analysis proce-
dures requiring balanced data (i.e., the same number of observations for all subjects),
e.g., some ANOVA-â•‰methods, are not directly applicable unless subjects with missing
data are removed (not generally recommended) or the missing data is somehow filled
in (more on that later). When we have missing data, there will inevitably be a loss in
precision when estimating quantities of interest (e.g., change over time) compared
to complete data. Put simply, we will have larger standard errors since we have less
information. Perhaps more importantly, missing data may cause results to be biased,
depending on the nature of the missingness and how it is handled in the analysis. We
168
begin by exploring different types of missingness, using the classification and termi-
nology coined by Rubin (1976). Later we will return to how to deal with missingness
when analyzing data. We also note that it is very rare in imaging papers to explicitly
state which type of missingness is assumed; it must instead be inferred from the analy-
sis choices.
Missing Completely at Random (MCAR)

When the missingness mechanism is such that, for all time points, it depends only
on observable covariates or just pure chance, we have MCAR missingness. A simple
example of MCAR would be that a subject’s scan is missing due to some techni-
cal problem with the scanner, or that dropout only depends on the height of the
subject (say e.g., if tall persons find the scanner environment more uncomfortable
and drop out at a higher rate). MCAR implies that subjects sharing the same values
of covariates also share the same probability distribution for the outcome, whether
they drop out or not. If we condition on the covariates using e.g., stratification or
regression modeling, dropouts and completers follow the same probability distribu-
tion for the outcome parts which are not observed for the dropouts. In practice, if
we have MCAR we may view the observed data as a random sample of the complete
data (after conditioning on the covariates), with a subsequent loss in precision due
to a smaller sample size but basically no other major issues. Subjects with missing
data may then be discarded from the analysis without introducing bias in the results,
just a loss in precision (larger Standard Errors). Observed data can be used to test
the MCAR assumption (Diggle, 1989; Little, 1988). We finally note that MCAR is a
strong assumption and often unlikely to be true in practice; see e.g., Glymour et al.
(2012) and Lo and Jagust (2012), who conclude that MCAR cannot be assumed for
the investigated imaging measures in the Three-â•‰City Dijon study and ADNI, respec-
tively. Nevertheless, longitudinal imaging data are often analyzed under MCAR
assumptions since complete case data are used. If MCAR doesn’t hold, results are
biased, but the MCAR assumption is usually not even commented, and even more
rarely investigated further, which leaves us guessing with respect to the validity of
results.
Missing at Random (MAR)

MAR occurs when the missingness depends not only on covariates but also on
the observed values of the outcome. If we have MAR, subjects sharing the same
observed values of the outcome as well as the covariates share the same probability
distribution for the outcome, whether they drop out or not. Given the covariates and
the outcome which we have observed, dropouts and completers follow the same
probability distribution regarding the outcome parts which are not observed for the
dropouts. For instance, consider a cognition study with a baseline measurement and
one follow-â•‰up. If subjects with low cognitive scores at baseline drop out more fre-
quently before follow-â•‰up, MCAR is out of the question, but we may assume MAR.
â•‡ 169
If we assume MAR, we assume that subjects dropping out before follow-â•‰up would,
on average, have performed like subjects with the same baseline score but who
completed the follow-â•‰up. Since missingness depends on the available observations
of the outcome of interest, simply excluding subjects with missing data from the
analysis will lead to biased results, and thus cannot be recommended (Fitzmaurice
et al., 2011, p. 496). In light of that remark, it is unfortunate that MAR is sometimes
confused with MCAR, resulting in the removal of subjects having missing data. The
established terminology, though, is not ideally suited for avoiding confusion, which
is unsatisfactory. Since MAR is less restrictive than MCAR regarding assumptions
on the missingness mechanism, it is more realistic and often recommended as the
default assumption for missing data (Fitzmaurice et al., 2011, p. 498). Conversely,
MCAR can be seen as a restrictive special case of MAR. It should be noted that,
when analyzing data with MAR missingness using likelihood-â•‰based methods (e.g.,
mixed-â•‰model approaches), we do not need to specify an explicit model for the
missingness mechanism (Fitzmaurice et al., 2011, p. 497). For this reason, MAR
missingness is often called “ignorable” (or non-â•‰informative). This is yet another
questionable choice of terminology, since it may be perceived as the missing data
itself being ignorable and subjects with missing data consequently removed. We
stress that the term ignorable refers to the fact that an exact model for missingness
need not be specified. Observed data cannot be used to distinguish between MAR
and NMAR (see below), and thus it is up to the researcher to make a judgment
regarding the plausibility of assuming MAR. In the imaging literature, e.g., Bernal-â•‰
Rusiel et al. (2013a, 2013b) and Chen et al. (2013) advocate the use of methods,
e.g., mixed models, which are suitable for MAR missingness, even though they do
not mention MAR explicitly but instead “missing data” without further specifica-
tion of the missingness type.
Not Missing at Random (NMAR)

If missingness depends on not only the covariates and observed data on the outcome,
but also on the missing outcome values, we have NMAR. If NMAR is the case, con-
ditioning on observed quantities will, unlike MAR/â•‰MCAR, not ensure that dropouts
and completers follow the same probability distribution regarding the outcome parts
which are not observed for the dropouts. In short, we cannot save the situation by con-
ditioning on what we have observed. If we again consider the hypothetical cognition
study with one baseline and one follow-â•‰up measurement where low scorers at baseline
drop out more frequently, NMAR implies that subjects dropping out before follow-â•‰
up would, on average, not have performed similar to subjects with the same baseline
score who completed the follow-â•‰up (in fact, not similar to any subjects who have com-
pleted the follow-â•‰up). Glymour et al. (2012; p.1345) conclude that “missing at random
(…) is unlikely to be the case in studies of dementia or cognitive decline,” indicating
that missingness should be treated as NMAR in such studies. If we have NMAR, we
need to explicitly consider the missingness mechanism and take it into account in the
analysis. There are (at least) two major concerns with that approach: The analysis will
be more complicated, both technically and, more importantly, in terms of interpreting
170
results, since they will depend on the particular choice of missingness model. Further,
there is no way of assessing if the missingness model is appropriate since that would
require observing the missing data. The results depend on model assumptions that we
cannot check using the observed data, and thus a sensitivity analysis should be per-
formed using different dropout models (Fitzmaurice et al., 2011, p. 500). To the best
of our knowledge, imaging data assuming NMAR missingness have not been studied
in the literature.
Test-â•‰Retest Effects
Besides dropout, there are other potential issues when estimating longitudinal
development. One, which requires some attention and thought, is the test-â•‰retest
or practice effect. If the same test or task is used repeatedly, a test-â•‰retest effect
may be present and affect the data. To make things even more difficult, dropout
sometimes affects the result in the same direction as the test-â•‰retest effect, mak-
ing it difficult to tease these effects apart. In imaging studies where the test-â•‰
retest effect was investigated, most authors associate the test-â•‰retest effect with
decreased brain activation. Examples include Sayres and Grill-â•‰Spector (2006),
who found repeated exposure to images of animals to be associated with reduced
activity in the occipital cortex, and Wagner et al. (2000), who showed decreases
in frontal areas during a word processing task (see also a review by Grill-â•‰Spector
et al., 2006). If a reduction in brain activity due to test-â•‰retest effects is present
in older subjects and is not accounted for, it may be mistaken for an age-â•‰related
decline. Some authors, on the other hand, have found increases in activation over
repeated performance of the same task, e.g., Kirschen et al. (2005), who report
increased cerebellar and parietal activity within-â•‰session for a verbal working
memory task. Jolles et al. (2010) interpreted increased DLPFC activity during a
working memory task in a control group as being a test-â•‰retest effect, although it
was only seen for the most demanding task. One should also keep in mind that
most imaging studies investigating the test-â•‰retest effect had a rather short time
between repetitions, from within-â•‰session to about six weeks, and the subjects
are usually quite young (under 40 years old). In a study having the rare combi-
nation (at least for imaging studies) of both older subjects (up to 80 years) and
a longer time between task-â•‰repetition (6 years), Nyberg et al. (2010) found no
evidence of a test-â•‰retest effect on brain activity. However, in behavioral studies
such as Rönnlund et al. (2005) and Rabbitt et al. (2004), test-â•‰retest effects were
seen using 5–â•‰7 years between tests, so it is not unthinkable that a test-â•‰retest
effect is present in “long-â•‰term” imaging data, although not yet demonstrated.
In their study, Rabbit et al. (2004) further report that the estimated test-â•‰retest
effect was about twice as large for 70-â•‰year-â•‰olds compared to 50-â•‰year-â•‰olds, effec-
tively counteracting some of the longitudinal decline in cognitive performance.
Further, Suchy et al. (2011) suggest that persons in early stages of cognitive
decline exhibit larger test-â•‰retest effects. It is unknown if such age-â•‰practice or
decline-â•‰practice interaction effects are present in brain activation data, which
presents an interesting avenue for future research.
â•‡ 171
Statistical Analysis
The key distinguishing element in the statistical analysis of longitudinal data com-
pared to cross-â•‰sectional is the multiple measurements per subject. This is handled
in different ways in the literature, and we give a brief account of some common
analysis methods encountered in imaging studies, along with their advantages and
disadvantages.
Summary measures: The simplest approach to longitudinal data analysis is to col-
lapse the within-â•‰subject data into a summary measure for that subject, see e.g., Clark
et al. (2012), who study structural and functional changes using structural MRI, PET,
and cerebral blood flow measures on healthy subjects, or Frings et al. (2012), who
investigate gray â•‰matter volume change for a group of patients. Using summary mea-
sures takes us back to the (cross-â•‰sectional) case with one observation per subject, and
we may use our favorite method for such data, e.g., t-â•‰test or ANOVA/â•‰Regression. The
summary measure might be, e.g., the average over all time points (Clark et al., 2012)
or the average difference between successive time points (Frings et al., 2012). Having
conceptual simplicity as their main virtue, summary measures also have drawbacks,
including the following ones. Firstly, collapsing data over time means that informa-
tion is lost; response profiles that look very different may yield the same value for the
summary measure. We cannot have time-â•‰varying covariates, and it might be unclear
how to handle missing data (especially subjects with only baseline data), unless those
subjects can be discarded.
Repeated Measures (RM-â•‰) ANOVA

Historically, RM-â•‰ANOVA was one of the first methods aimed at analyzing a complete
longitudinal dataset without using summary measures. It builds on concepts that are
well-â•‰understood and known to a broad audience, hence its still-â•‰standing popularity,
see e.g., Asami et al. (2012), where gray â•‰matter volume in schizophrenia patients was
studied, or the Frings et al. (2012) study mentioned above, which used RM-â•‰ANOVA
(actually RM-â•‰ANCOVA) in a subgroup analysis. The main difference from an ordi-
nary ANOVA is the addition of a random effect to model the within-â•‰subject varia-
tion, in addition to the between-â•‰subject variance term already present in the simple
ANOVA model. These two random terms are assumed independent. The RM-â•‰ANOVA
model yields a particular structure for the covariance between measurements on the
same subject. The structure is called “compound symmetry” and implies that that the
covariance between any two measurements on one subject is positive and constant. In
practice, we usually expect the strength of association between two observations to
decay as we consider time points further and further apart, which is not captured by
the compound symmetry structure. Of course, if we only have two time points, as e.g.,
Frings et al. (2012) in their subgroup analysis, the problem with compound symme-
try disappears. The straight â•‰forwardness and simplicity of RM-â•‰ANOVA is somewhat
diminished if we have missing data, since a complete dataset is required. Thus, we
need to either discard subjects (as in Asami et al., 2012) or use some kind of imputa-
tion scheme to make the data complete (see Josefsson et al., in preparation).
172
Linear Mixed Effect Model (LME)

While scarcely used in the imaging literature (there are exceptions though, see e.g.,
Thambisetty et al., 2010), the LME formulation allow for a very general approach
to analyze longitudinal data (as noted in e.g., Bernal-â•‰Rusiel et al., 2013a and Chen
et al., 2013). We are quite free to choose how to model both mean (time trajectory)
and covariance structures. For the mean, we may choose e.g., a linear (Bernal-â•‰Rusiel
et al., 2013a, 2013b) or polynomial trend, or to actually not assume any particular
shape at all using e.g., profile analysis or splines. Regarding the covariance struc-
ture, we have a variety of choices ranging from the most general (“Unstructured”
where every pair of time points has a unique covariance) to more restrictive choices
like compound symmetry, i.e., we can obtain RM-â•‰ANOVA as a special case. We
refer the reader to Fitzmaurice et al. (2011, ch. 7.4) for a more detailed overview
of the different options for modeling the covariances. The random effects may be
used for modeling intercept and time slope (Bernal-â•‰Rusiel et al., 2013a, 2013b),
as well as (regional) spatial effects (DuBois Bowman et al., 2008). It is also pos-
sible to include spatial covariance structures directly (Bernal-â•‰Rusiel et al., 2013b)
which however may greatly increase model complexity (i.e., number of parameters
to estimate). Bernal-â•‰Rusiel et al. (2013a) demonstrate, using a simulation study
based on ADNI data, that the detection power for different atrophy rates between
healthy controls and AD patients is greatest when using LME, compared to using
RM-â•‰ANOVA or a summary rate measure. Chen et al (2013) name six situations
where LME is preferred over ANOVA-â•‰methods, including longitudinal studies
with time-â•‰varying covariates and/â•‰or missing data. Historically, the biggest draw-
back of LME models compared to RM-â•‰ANOVA and summary measures has been
computational. Given the current computational resources, LME models should
be possible to use in most situations. Another reason preventing the widespread
use of LME models in neuroimaging is simply that the possibility of using LME
models has so far been lacking in most software aimed at analyzing neuroimaging
data, with AFNI (Cox, 1996) being the exception. Recently, LME capability has
been added to FreeSurfer (Reuter et al., 2012), making LME modeling increasingly
available in neuroimaging.
Putting together the conclusions from the missingness and longitudinal analysis
sections, we make some summary remarks on analyzing longitudinal data with
missingness present in the data. Starting with RM-â•‰ANOVA and summary mea-
sures, they require complete datasets. If we have missing data there are basically
two ways of artificially obtaining a complete dataset; to discard subjects having
missing data, i.e., do a complete-â•‰case (CC) analysis, which is by far the most com-
mon approach seen in the neuroimaging literature. We emphasize that results of a
CC analysis is valid only if missingness is MCAR, otherwise results will be biased.
The other way to construct a complete dataset from an incomplete one is to use
some imputation technique to fill in the missing data. Imputation has been shown
to work under MAR missingness (see e.g., Carpenter and Kenward, 2013, ch.2).
However, when imputing we need to make some modeling assumption(â•‰s) and the
validity (unbiasedness) of the results hinges on those assumptions being satis-
fied. As an example, one widely used imputation method for dropout is the “last
â•‡ 173
value carried forward” (LVCF) method. Missing values are replaced with the last
value observed before dropout. This is very easy to perform, but the assumption
is that the variable of interest remains constant starting from the last observation.
This is a very strong and in many settings quite unrealistic assumption (see e.g.,
Kenward and Molenberghs, 2009). Other ways include Nearest Neighbor imputa-
tion, where the observation(â•‰s) from the most similar completer is imputed, or using
some regression model for predicting the missing values, where one can form one
or several predictions (single/â•‰multiple imputation), the final result depends on the
prediction model being correctly specified (see Josefsson et al., in preparation for
a review and comparison of imputation methods in a longitudinal fMRI setting).
In the Josefsson study, the multiple imputation method was found to perform best.
The LME approach does not require a complete dataset, and if the missingness is
MAR/â•‰MCAR we do not need to specify the exact missingness mechanism. The
results are valid, as long as both the mean and covariance structures of the LME
model have been correctly specified (see e.g., Fitzmaurice et al., 2011, p. 496). If
missingness is NMAR (suggested by Glymour et al., 2012 to be the case for studies
of dementia/â•‰cognitive decline) we need to specify the missingness mechanism. The
analysis results will depend on that specification, which cannot be checked using
observed data. It is recommended to do some sensitivity analyses, e.g., using dif-
ferent missingness specifications and compare results or include some sensitivity
parameters (which can be varied) using a Pattern-Mixture-Model approach (see
e.g., Daniels and Hogan, 2008, ch. 8.4).
Test-â•‰Retest Reliability
When designing a longitudinal study, we expect that what we measure using the scan-
ner will change over time, but we also must consider that properties of the scanner
itself may change over time. Scanners may be replaced or upgraded with new hard-
ware/â•‰software. If the longitudinal perspective stretches over several years, we can
expect that some aspects of the scanning equipment (hardware and/â•‰or software) will
change, especially if the scanning device is shared with clinicians (see e.g., Ekman
et al., 2012). Different scanner characteristics may also be encountered in data from
multi-â•‰center studies (see e.g., Kochunov et al., 2014).
The main message is to change as little as possible, preferably nothing, in the scan-
ner configurations (see e.g., Han et al., 2006 or Takao et al., 2013), since the magni-
tude of scanner effects are possibly quite substantial. Takao et al. (2013) estimated the
scanner effect (using two 3T scanners of the same brand) when measuring a 1-â•‰year
change in gray-â•‰matter (GM) volume to be on par with the typical 1-â•‰year GM atrophy
rate in AD patients, e.g., those reported by Fjell et al. (2009a), based on ADNI data
where “a variety of scanners” were used. Even the within-â•‰scanner effect of a software
update was comparable to the atrophy rates reported in Fjell et al. (2009a).
Even assuming no expected effect from changing hardware and/â•‰or software, reliabil-
ity issues may be encountered. Zanto et al. (2014) investigated the reliability of brain
activation during a memory task over a 3-â•‰months interval. Using the exact same scanner
and software, they reported the reliability of activations to be in the moderate to good
174
range (IntraClass Correlation coefficients between 0.3–â•‰0.7). This is slightly discourag-

ing, since even an ICC of 0.7 means that only about half the IC variance is accounted
for. Notably, it seems possible to obtain better reliability cross-â•‰sectionally using the
same scanner brand and model but in different locations, as Forsyth et al. (2014) reaches
ICC:s of 0.81–â•‰0.95. Also, Han et al. (2006) report quite good reliability using four scans
over an eight-â•‰week period (no ICC:s reported, but small between-â•‰session variability). In
their discussion, Takao et al. (2013) mention the issue of drift in scanners as a source of
variation that becomes non-â•‰negligible over longer time periods, which might contribute
to the difference in results from long and short/â•‰cross-â•‰sectional time-â•‰spans.
To evaluate immediate effects of a change in scanner configuration, both a small
sample of subjects (e.g., Han et al., 2006) or phantoms (e.g., Teipel et al., 2010) have
been used. If long-â•‰term (say around 6 years, like in e.g., Nyberg et al., 2010, or lon-
ger) scanner effects are to be evaluated, human subjects can probably not be used for
evaluating scanner effects, but the use of phantoms seems like a viable option.
Conclusion: Where Next? Change-â•‰Change Studies
Taken together, longitudinal studies of age-â•‰related changes in episodic memory (e.g.,

Rönnlund et al., 2005), fMRI activation patterns (e.g., Nyberg et al., 2010), gray-â•‰mat-
ter volume (e.g., Hedman et al., 2012), and white-â•‰matter tracts (e.g., Sexton et al.,
2014) converge on stability or very marginal changes until age 55–â•‰60, and acceler-
ated decline thereafter. The apparent similarities in these curvilinear change functions
indicate that decline in episodic memory is related to corresponding brain changes.
Indeed, this is a fundamental assumption that (explicitly or implicitly) underlies most
studies in the cognitive neuroscience of aging field. Tentatively, it has been estimated
that age-â•‰related structural brain changes account for between 25%–â•‰100% of the vari-
ance in cognitive functions (Fjell & Walhovd, 2010), but there is marked variability
between cross-â•‰sectional studies and the brain-â•‰cognition mapping is still not resolved
(see e.g., Salthouse, 2011; Hedden et al., 2014).
To shed further light on the brain-â•‰cognition link in aging, examination of conjoint
longitudinal changes in memory and brain parameters may prove to be informative,
but only a limited number of such studies have been published to date. In a recent
study, age-â•‰related changes in brain white matter could be linked to changes in sen-
sorimotor speed but not to higher-â•‰order cognitive functions such as episodic memory
(Lövdén et al., 2014; for a similar cross-â•‰sectional finding, see Salami et al., 2012a;
see also Garde et al., 2000; Charlton et al., 2010). These findings contradict the notion
of “cortical disconnection” as a cause of episodic memory decline in aging (see e.g.,
O’Sullivan et al., 2001), but future larger-â•‰scale studies, targeting “episodic-â•‰memory
relevant tracts” such as the fornix (Fletcher et al., 2013), will be needed to more
thoroughly examine the role of age-â•‰related white â•‰matter changes in accounting for
episodic memory decline.
Gray-matter changes, notably in medial-â•‰temporal lobe regions, have since long
been discussed in the context of age-â•‰related episodic â•‰memory decline, but proven dif-
ficult to conclusively address in cross-â•‰sectional studies, partly due to marked inter-â•‰
individual volumetric differences (e.g., Lupien et al., 2007). In a longitudinal study
â•‡ 175
of 6-â•‰year within-â•‰person changes in hippocampus volume and episodic memory, we

found strong support for a link between hippocampus shrinkage and memory decline
(Persson et al., 2012; cf., Golomb et al., 1996; Kramer et al., 2007).
As reviewed in the section on functional imaging, decline in hippocampal activ-
ity has been demonstrated to be reliably associated with memory decline in two stud-
ies (O’Brien et al., 2010; Persson et al., 2014), which in both cases appeared to be
independent of volumetric gray matter losses. Thus, functional imaging provides both
converging and complementary information on the importance of the hippocampus for
age-â•‰related memory decline. The one available change-â•‰change study for frontal-â•‰cortex
function observed declining frontal recruitment over time to be associated with main-
tained cognitive functions, highlighting the need to consider potential practice effects
on functional imaging measures. Today, little is known regarding any long-â•‰term practice
effects, studies aimed at investigating such effects typically have times between scan-â•‰
rescan ranging from within-â•‰session (Wagner et al., 2000) to around six weeks (Jolles
et al., 2010). To further complicate matters, there are indications from recent research
(Takao et al., 2013; Zanto et al., 2014) that the test-â•‰retest reliability of the scanning envi-
ronment depends quite heavily on both hardware as well as software changes/â•‰updates.
This makes the planning of long-â•‰term longitudinal scanning experiments a delicate task;
it is usually quite hard to keep both hardware and software identical over several years.
Scanning a standardized phantom regularly may alleviate this concern.
Reviewing the longitudinal neuroimaging literature, a frequent practice is to use
data only from subjects that complete the study. Using completers’ data only implies
that Missing Completely at Random (MCAR) is assumed to be the missingness mech-
anism. This assumption is not often stated explicitly, which makes it unclear if the
authors have considered the missingness and decided that MCAR is an appropriate
assumption, or used completers’ data by force of habit. Further, if MCAR is assumed,
there exist tests (Diggle 1989, Little 1988) for investigating that assumption, and they
should be utilized to a greater extent. Performing the analysis using the mixed-â•‰model
approach (Bernal-â•‰Rusiel et al., 2013a 2013b, Chen et al., 2013) yields valid results
without needing to specify the missingness mechanism beyond assuming that it its
MCAR or Missing at Random (MAR), where MAR is often argued to be a reasonable
assumption in practice (Fitzmaurice et al., 2011, p.498). This makes the mixed-â•‰effect
approach attractive, and its results can serve as a point of departure for more intricate
analyses involving non-â•‰random missingness.
In conclusion, longitudinal imaging of the aging brain holds great promise, but also
poses challenges. It is clear that issues pertaining to non-â•‰random drop â•‰out, practice
effects, and reliability issues such as scanner upgrades must be appropriately handled
to assure the validity of findings. Nonetheless, we argue that future longitudinal stud-
ies will be a key to a solid understanding of independent and interrelated age-â•‰related
changes of brain and mind.
Acknowledgments
Supported by the Knut and Alice Wallenberg Foundation (KAW), the Söderberg’s
Fundation, and the Swedish Research Council.
176
References
Abe O, Yamasue H, Aoki S, Suga M, Yamada H, Kasai K, Masutani Y, Kato N, Kato N, Ohtomo
K (2008) Aging in the CNS: comparison of gray/â•‰white matter volume and diffusion tensor
data. Neurobiol Aging 29:102–â•‰116.
Asami T, Bouix S, Whitford TJ, Shenton ME, Salisbury DF, McCarley RW (2012) Longitudinal
loss of gray matter volume in patients with first-â•‰episode schizophrenia: DARTEL auto-
mated analysis and ROI validation. Neuroimage 59:986–â•‰996.
Barrick TR, Charlton RA, Clark CA, Markus HS (2010) White matter structural decline in
normal ageing: a prospective longitudinal study using tract-â•‰ based spatial statistics.
Beason-â•‰Held LL, Goh JO, An Y, Kraut M A, O’Brien RJ, Ferrucci L, Resnick SM (2013)
Changes in brain function occur years before the onset of cognitive impairment. J Neurosci
33:18008–â•‰18014.
Beason-â•‰Held LL, Kraut MA, Resnick SM (2008a) II. Temporal patterns of longitudinal change
in aging brain function. Neurobiol Aging 29:497–â•‰513.
Beason-â•‰Held LL, Kraut MA, Resnick SM (2008b) I. Longitudinal changes in aging brain func-
tion. Neurobiol Aging 29:483–â•‰496.
Beason-â•‰Held LL, Kraut MA, Resnick SM (2009) Stability of Default-Mode Network Activity
in the Aging Brain. Brain Imaging Behav 3:123–â•‰131.
Bernal-â•‰Rusiel JL, Greve DN, Reuter M, Fischl B, Sabuncu MR (2013a) Statistical analy-
sis of longitudinal neuroimage data with Linear Mixed Effects models. Neuroimage
66:249–â•‰260.
Bernal-â•‰Rusiel JL, Reuter M, Greve DN, Fischl B, Sabuncu MR (2013b) Spatiotemporal linear
mixed effects modeling for the mass-â•‰univariate analysis of longitudinal neuroimage data.
Blatter DD, Bigler ED, Gale SD, Johnson SC, Anderson C V, Burnett BM, Parker N, Kurth S,
Horn SD (1995) Quantitative volumetric analysis of brain MR: normative database span-
ning 5 decades of life. Am J Neuroradiol 16:241–â•‰251.
Burzynska AZ, Preuschhof C, Bäckman L, Nyberg L, Li S-â•‰C, Lindenberger U, Heekeren HR
(2010) Age-â•‰related differences in white matter microstructure: region-â•‰specific patterns of
diffusivity. Neuroimage 49:2104–â•‰2112.
Bäckman L, Ginovart N, Dixon RA, Wahlin T, Wahlin A, Halldin C, Farde L (2000) Age-â•‰related
cognitive deficits mediated by changes in the striatal dopamine system. Am J Psychiatry
157:635–â•‰637.
Bäckman L, Lindenberger U, Li S-â•‰C, Nyberg L (2010) Linking cognitive aging to alterations
in dopamine neurotransmitter functioning: Recent data and future avenues. Neurosci
Biobehav Rev 34:670–â•‰677.
Cabeza R, Anderson ND, Locantore JK, Mcintosh AR (2002) Aging gracefully: Compensatory
brain activity in high-â•‰performing older adults. Neuroimage 17:1394–â•‰1402.
Carpenter JR, Kenward MG (2013) Multiple Imputation and its Application. Wiley.
Celone KA, Calhoun VD, Dickerson BC, Atri A, Chua EF, Miller SL, DePeau K, Rentz DM,
Selkoe DJ, Blacker D, Albert MS, Sperling RA (2006) Alterations in memory networks in
mild cognitive impairment and Alzheimer’s disease: an independent component analysis.
J Neurosci 26:10222–â•‰10231.
Charlton RA, Schiavone F, Barrick TR, Morris RG, Markus HS (2010) Diffusion tensor imag-
ing detects age related white matter change over a 2 year follow-â•‰up which is associated
with working memory decline. J Neurol Neurosurg Psychiatry 81:13–â•‰19.
Chen G, Saad ZS, Britton JC, Pine DS, Cox RW (2013) Linear mixed-â•‰effects modeling
approach to FMRI group analysis. Neuroimage 73:176–â•‰190.
177
Clark VH, Resnick SM, Doshi J, Beason-Held LL, Zhou Y, Ferrucci L, Wong DF, Kraut M A,
Davatzikos C (2012) Longitudinal imaging pattern analysis (SPARE-CD index) detects
early structural and functional changes before cognitive decline in healthy older adults.
Cox RW (1996) AFNI: software for analysis and visualization of functional magnetic reso-
nance neuroimages. Comput Biomed Res 29:162–173.
Daniels MJ, Hogan JW (2008) Missing Data in Longitudinal Studies: Strategies for Bayesian
Modeling and Sensitivity Analysis. Chapman & Hall.
Dennis NA, Cabeza R (2008) Neuroimaging of healthy cognitive aging. In: The
Handbook of Aging and Cognition, 3rd ed. (Craik FIM, Salthouse TA, eds), pp 1–54.
New York: Psychology Press.
Diggle P (1989) Testing for random dropouts in repeated measurement data. Biometrics
47:1617–1619; discussion 1619–1621.
Du A-T, Schuff N, Chao LL, Kornak J, Jagust WJ, Kramer JH, Reed BR, Miller BL, Norman
D, Chui HC, Weiner MW (2006) Age effects on atrophy rates of entorhinal cortex and hip-
pocampus. Neurobiol Aging 27:733–740.
DuBois Bowman F, Caffo B, Bassett SS, Kilts C (2008) A Bayesian hierarchical framework for
spatial modeling of fMRI data. Neuroimage 39:146–156.
Ekman U, Eriksson J, Forsgren L, Mo SJ, Riklund K, Nyberg L (2012) Functional brain activity
and presynaptic dopamine uptake in patients with Parkinson’s disease and mild cognitive
impairment: a cross-sectional study. Lancet Neurol 11:679–687.
Fitzmaurice GM, Laird NM, Ware JH (2011) Applied Longitudinal Analysis, 2nd ed. Wiley.
Fjell A, Walhovd K (2010) Structural brain changes in aging: courses, causes and cognitive
consequences. Rev Neurosci 221:187–221.
Fjell AM, Walhovd KB, Fennema-Notestine C, McEvoy LK, Hagler DJ, Holland D, Brewer
JB, Dale AM (2009a) One- year brain atrophy evident in healthy aging. J Neurosci
29:15223–15231.
Fjell AM, Westlye LT, Amlien I, Espeseth T, Reinvang I, Raz N, Agartz I, Salat DH, Greve DN,
Fischl B, Dale AM, Walhovd KB (2009b) High consistency of regional cortical thinning
in aging across multiple samples. Cereb Cortex 19:2001–2012.
Walhovd KB (2014) Accelerating cortical thinning: unique to dementia or universal in
aging? Cereb Cortex 24:919–934.
Fletcher E, Raman M, Huebner P, Liu A, Mungas D, Carmichael O, DeCarli C (2013) Loss of
fornix white matter volume as a predictor of cognitive impairment in cognitively normal
elderly individuals. JAMA Neurol 70:1389–1395.
Forsyth JK et al. (2014) Reliability of functional magnetic resonance imaging activation dur-
ing working memory in a multi-site study: Analysis from the North American Prodrome
Longitudinal Study. Neuroimage 97C:41–52.
Fotenos A, Mintun M, Snyder A, Morris J, Buckner R (2008) Brain Volume Decline in Aging.
Arch Neurol 65:113–120.
Fotenos A, Snyder A, Girton L (2005) Normative estimates of cross-sectional and longitudinal
brain volume decline in aging and AD. Neurology 64:1032–1039.
Fox NC, Warrington EK, Freeborough PA, Hartikainen P, Kennedy AM, Stevens JM, Rossor
MN (1996) Presymptomatic hippocampal atrophy in Alzheimer’s disease: A longitudinal
MRI study. 119:2001–2007.
Frings L, Mader I, Landwehrmeyer BG, Weiller C, Hüll M, Huppertz H-J (2012) Quantifying
change in individual subjects affected by frontotemporal lobar degeneration using auto-
mated longitudinal MRI volumetry. Hum Brain Mapp 33:1526–1535.
178
Garde E, Mortensen EL, Krabbe K, Rostrup E, Larsson HB (2000) Relation between age-
related decline in intelligence and cerebral white-matter hyperintensities in healthy octo-
genarians: a longitudinal study. Lancet 356:628–634.
Glymour MM, Chêne G, Tzourio C, Dufouil C (2012) Brain MRI markers and dropout in a lon-
gitudinal study of cognitive aging: the Three-City Dijon Study. Neurology 79:1340–1348.
Goh JO, Beason-Held LL, An Y, Kraut MA, Resnick SM (2013) Frontal function and executive
processing in older adults: Process and region specific age-related longitudinal functional
changes. Neuroimage 69:43–50.
Golomb J, Kluger A, de Leon MJ, Ferris SH, Mittelman MP, Cohen J, George AE (1996)
Hippocampal formation size predicts declining memory performance in normal aging.
Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS (2001) A voxel-
based morphometric study of ageing in 465 normal adult human brains. Neuroimage
14:21–36.
Grady CL, Springer M V, Hongwanishkul D, McIntosh AR, Winocur G (2006) Age-related
changes in brain activity across the adult lifespan. J Cogn Neurosci 18:227–241.
Grill-Spector K, Henson R, Martin A (2006) Repetition and the brain: neural models of
stimulus-specific effects. Trends Cogn Sci 10:14–23.
Gutchess AH, Welsh RC, Hedden T, Bangert A, Minear M, Liu LL, Park DC (2005) Aging and
the neural correlates of successful picture encoding: Frontal activations compensate for
decreased medial-temporal activity. J Cogn Neurosci 17:84–96.
Han X, Jovicich J, Salat D, van der Kouwe A, Quinn B, Czanner S, Busa E, Pacheco J, Albert
M, Killiany R, Maguire P, Rosas D, Makris N, Dale A, Dickerson B, Fischl B (2006)
Reliability of MRI-derived measurements of human cerebral cortical thickness: the effects
of field strength, scanner upgrade and manufacturer. Neuroimage 32:180–194.
Hedman AM, van Haren NEM, Schnack HG, Kahn RS, Hulshoff Pol HE (2012) Human brain
changes across the life span: a review of 56 longitudinal magnetic resonance imaging stud-
ies. Hum Brain Mapp 33:1987–2002.
(2016) Multiple Brain Markers are Linked to Age-Related Variation in Cognition. Cereb
Cortex 26:1388–1400.
Jiang J, Sachdev P, Lipnicki DM, Zhang H, Liu T, Zhu W, Suo C, Zhuang L, Crawford
J, Reppermund S, Trollor J, Brodaty H, Wen W (2014) A longitudinal study of
brain atrophy over two years in community-dwelling older individuals. Neuroimage
86:203–211.
Jolles DD, Grol MJ, Van Buchem MA, Rombouts SARB, Crone EA (2010) Practice effects in
the brain: Changes in cerebral activation after working memory practice depend on task
demands. Neuroimage 52:658–668.
Josefsson M, Lundquist A, Nyberg L (in preparation) A simulation study of missing data meth-
ods for longitudinal fMRI data.
Kalpouzos G, Chételat G, Baron J-C, Landeau B, Mevel K, Godeau C, Barré L, Constans
J-M, Viader F, Eustache F, Desgranges B (2009) Voxel-based mapping of brain gray
matter volume and glucose metabolism profiles in normal aging. Neurobiol Aging
30:112–124.
Kenward MG, Molenberghs G (2009) Last observation carried forward: a crystal ball?
J Biopharm Stat 19:872–888.
Kirschen MP, Chen SHA, Schraedley-Desmond P, Desmond JE (2005) Load-and practice-
dependent increases in cerebro-cerebellar activation in verbal working memory: an fMRI
study. Neuroimage 24:462–472.
179
Klunk WE, Engler H, Nordberg A, Wang Y, Blomqvist G, Holt DP, Bergstro M, Savitcheva
I, Debnath ML, Barletta J, Price JC, Sandell J, Lopresti BJ, Wall A, Koivisto P, Antoni
G, Mathis CA, Långstrom B (2004) Imaging brain amyloid in Alzheimer’s disease with
Pittsburgh Compound‐B. Ann Neurol 55:306–319.
Kochunov P et al. (2014) Multi-site study of additive genetic effects on fractional anisotropy of
cerebral white matter: Comparing meta and megaanalytical approaches for data pooling.
Kochunov P, Glahn DC, Lancaster JL, Winkler a M, Smith S, Thompson PM, Almasy L,
Duggirala R, Fox PT, Blangero J (2010) Genetics of microstructure of cerebral white mat-
ter using diffusion tensor imaging. Neuroimage 53:1109–1116.
Kramer JH, Mungas D, Reed BR, Wetzel ME, Burnett MM, Miller BL, Weiner MW, Chui
HC (2007) Longitudinal MRI and cognitive change in healthy elderly. Neuropsychology
21:412–418.
Lebel C, Gee M, Camicioli R, Wieler M, Martin W, Beaulieu C (2012) Diffusion tensor imag-
ing of white matter tract evolution over the lifespan. Neuroimage 60:340–352.
extraction: What’s change got to do with it? Psychol Aging 26:34–47.
Little R (1988) A test of missing completely at random for multivariate data with missing val-
ues. J Am Stat Assoc 83:1198–1202.
Lo RY, Jagust WJ (2012) Predicting missing biomarker data in a longitudinal study of Alzheimer
disease. Neurology 78:1376–1382.
Logan JM, Sanders AL, Snyder AZ, Morris JC, Buckner RL (2002) Under-recruitment and
nonselective recruitment: Dissociable neural mechanisms associated with aging. Neuron
33:827–840.
Lupien SJ, Evans A, Lord C, Miles J, Pruessner M, Pike B, Pruessner JC (2007) Hippocampal
volume is as variable in young as in older adults: implications for the notion of hippocam-
pal atrophy in humans. Neuroimage 34:479–485.
RL (2003) Functional deactivations: change with age and dementia of the Alzheimer type.
Proc Natl Acad Sci U S A 100:14504–14509.
Ly M, Canu E, Xu G, Oh J, McLaren DG, Dowling NM, Alexander AL, Sager M a, Johnson
SC, Bendlin BB (2014) Midlife measurements of white matter microstructure pre-
dict subsequent regional white matter atrophy in healthy adults. Hum Brain Mapp
35:2044–2054.
Lövdén M, Köhncke Y, Laukka EJ, Kalpouzos G, Salami A, Li T-Q, Fratiglioni L, Bäckman L
tract are associated in very old age. Neuroimage 102:520–530.
Mattay VS, Fera F, Tessitore A, Hariri AR, Berman KF, Das S, Meyer-Lindenberg A, Goldberg
TE, Callicott JH, Weinberger DR (2006) Neurophysiological correlates of age-related
changes in working memory capacity. Neurosci Lett 392:32–37.
Mormino EC, Kluth JT, Madison CM, Rabinovici GD, Baker SL, Miller BL, Koeppe RA, Mathis
CA, Weiner MW, Jagust WJ (2009) Episodic memory loss is related to hippocampal-
mediated beta-amyloid deposition in elderly subjects. Brain 132:1310–1323.
Nyberg L, Andersson M, Kauppi K, Lundquist A, Persson J, Pudas S, Nilsson L-G (2014)
Age-related and Genetic Modulation of Frontal Cortex Efficiency. J Cogn Neurosci
26:746–754.
tion in aging. Proc Natl Acad Sci U S A 107:22682–22686.
180
O’Brien JL, O’Keefe KM, LaViolette PS, DeLuca AN, Blacker D, Dickerson BC, Sperling RA
(2010) Longitudinal fMRI in elderly reveals loss of hippocampal activation with clinical
decline. Neurology 74:1969–1976.
O’Sullivan M, Jones DK, Summers PE, Morris RG, Williams SCR, Markus HS (2001) Evidence
57:632–638.
Park DC, Lautenschlager G, Hedden T, Davidson NS, Smith AD, Smith PK (2002) Models of
visuospatial and verbal memory across the adult life span. Psychol Aging 17:299–320.
Persson J, Kalpouzos G, Nilsson L-G, Ryberg M, Nyberg L (2011) Preserved hippocampus
activation in normal aging as revealed by fMRI. Hippocampus 21:753–766.
Persson J, Nyberg L (2006) Altered brain activity in healthy seniors: What does it mean? Prog
Brain Res 157:45–57.
Persson J, Nyberg L, Lind J, Larsson A, Nilsson L-G, Ingvar M, Buckner RL (2006) Structure-
function correlates of cognitive decline in aging. Cereb Cortex 16:907–915.
Persson J, Pudas S, Lind J, Kauppi K, Nilsson L-G, Nyberg L (2012) Longitudinal structure-
function correlates in elderly reveal MTL dysfunction with cognitive decline. Cereb
Cortex 22:2297–2304.
Persson J, Pudas S, Nilsson L-G, Nyberg L (2014) Longitudinal assessment of default-mode
brain function in aging. Neurobiol Aging 35:2107–2117.
Pudas S, Persson J, Josefsson M, de Luna X, Nilsson L-G, Nyberg L (2013) Brain character-
istics of individuals resisting age-related cognitive decline over two decades. J Neurosci
33:8668–8677.
Rabbitt P, Diggle P, Holland F, McInnes L (2004) Practice and drop-out effects during a 17-year
longitudinal study of cognitive aging. J Gerontol B Psychol Sci Soc Sci 59: 84–97.
Rajah MN, D’Esposito M (2005) Region- specific changes in prefrontal function with
age: A review of PET and fMRI studies on working and episodic memory. Brain
128:1964–1983.
D, Acker JD (2005) Regional brain changes in aging healthy adults: General trends, indi-
vidual differences and modifiers. Cereb Cortex 15:1676–1689.
Raz N, Rodrigue KM (2006) Differential aging of the brain: patterns, cognitive correlates and
modifiers. Neurosci Biobehav Rev 30:730–748.
Resnick SM, Goldszal AF, Davatzikos C, Golski S, Kraut MA, Metter EJ, Nick R, Zonderman
AB (2000) One-year Age Changes in MRI Brain Volumes in Older Adults. Cereb Cortex
10:464–472.
Resnick SM, Pham DL, Kraut MA, Zonderman AB, Davatzikos C (2003) Longitudinal magnetic
resonance imaging studies of older adults: A shrinking brain. J Neurosci 23:3295–3301.
Reuter M, Schmansky NJ, Rosas HD, Fischl B (2012) Within-subject template estimation for
unbiased longitudinal image analysis. Neuroimage 61:1402–1418.
Rodrigue KM, Kennedy KM, Park DC (2009) Beta-amyloid deposition and the aging brain.
Neuropsychol Rev 19:436–450.
Rodrigue KM, Raz N (2004) Shrinkage of the entorhinal cortex over five years predicts mem-
ory performance in healthy adults. J Neurosci 24:956–963.
Rubin D (1976) Inference and missing data. Biometrika 63:581–592.
181
Rönnlund M, Nyberg L, Bäckman L, Nilsson L-G (2005) Stability, growth, and decline in adult
life span development of declarative memory: Cross-sectional and longitudinal data from
a population-based study. Psychol Aging 20:3–18.
Salami A, Eriksson J, Nilsson L-G, Nyberg L (2012a) Age-related white matter microstruc-
tural differences partly mediate age-related decline in processing speed but not cognition.
Salami A, Eriksson J, Nyberg L (2012b) Opposing effects of aging on large-scale brain systems
for memory encoding and cognitive control. J Neurosci 32:10749–10757.
Salat DH, Greve DN, Pacheco JL, Quinn BT, Helmer KG, Buckner RL, Fischl B (2009)
Regional white matter volume differences in nondemented aging and Alzheimer’s disease.
Salthouse TA (2004) What and when of cognitive aging. Curr Dir Psychol Sci 13:140–144.
Salthouse TA (2011) Neuroanatomical substrates of age-related cognitive decline. Psychol Bull
137:753–784.
Sayres R, Grill-Spector K (2006) Object-selective cortex exhibits performance-independent
repetition suppression. J Neurophysiol 95:995–1007.
Schaie KW (1994) The course of adult intellectual development. Am Psychol 49:304–313.
Sexton CE, Walhovd KB, Storsve AB, Tamnes CK, Westlye LT, Johansen-Berg H, Fjell AM
(2014) Accelerated changes in white matter microstructure during aging: A longitudinal
diffusion tensor imaging study. The Journal of Neuroscience 34: 15425–15436.
Singh-Manoux A, Kivimaki M, Glymour MM, Elbaz A, Berr C, Ebmeier KP, Ferrie JE,
Dugravot A (2012) Timing of onset of cognitive decline: results from Whitehall II pro-
spective cohort study. Br Med J 344:d7622.
Sojkova J, Zhou Y, An Y, Kraut MA, Ferrucci L, Wong DF, Resnick SM (2011)
Longitudinal patterns of β-amyloid deposition in nondemented older adults. Arch
Neurol 68:644–649.
Mapping cortical change across the human life span. Nat Neurosci 6:309–315.
Suchy Y, Kraybill ML, Franchow E (2011) Practice effect and beyond: reaction to novelty as
an independent predictor of cognitive decline among older adults. J Int Neuropsychol Soc
17:101–111.
Sullivan E (2004) Effects of age and sex on volumes of the thalamus, pons, and cortex.
Sullivan E V, Rohlfing T, Pfefferbaum A (2010) Longitudinal study of callosal microstructure
in the normal adult aging brain using quantitative DTI fiber tracking. Dev Neuropsychol
35:233–256.
Takao H, Hayashi N, Ohtomo K (2013) Effects of the use of multiple scanners and of scan-
ner upgrade in longitudinal voxel-based morphometry studies. J Magn Reson Imaging
38:1283–1291.
Taki Y, Thyreau B, Kinomura S, Sato K, Goto R, Wu K, Kawashima R, Fukuda H (2013) A
longitudinal study of age-and gender-related annual rate of volume changes in regional
gray matter in healthy adults. Hum Brain Mapp 34:2292–22301.
Tang Y, Whitman GT, Lopez I, Baloh RW (2001) Brain Volume Changes on Longitudinal
Magnetic Resonance. J Neuroimaging 11:393–400.
Teipel SJ, Meindl T, Wagner M, Stieltjes B, Reuter S, Hauenstein K-H, Filippi M, Ernemann
U, Reiser MF, Hampel H (2010) Longitudinal changes in fiber tract integrity in healthy
aging and mild cognitive impairment: a DTI follow- up study. J Alzheimer’s Dis
22:507–522.
Thambisetty M, Wan J, Carass A, An Y, Prince JL, Resnick SM (2010) Longitudinal changes in
cortical thickness associated with normal aging. Neuroimage 52:1215–1223.
182
Wagner AD, Koutstaal W, Maril A, Schacter DL, Buckner RL (2000) Task-specific repetition
priming in left inferior prefrontal cortex. Cereb Cortex 10:1176–1184.
Walhovd KB, Fjell AM, Amlien I, Grambaite R, Stenset V, Bjørnerud A, Reinvang I, Gjerstad
L, Cappelen T, Due-Tønnessen P, Fladby T (2009) Multimodal imaging in mild cognitive
impairment: Metabolism, morphometry and diffusion of the temporal-parietal memory
network. Neuroimage 45:215–223.
cal structures. Neurobiol Aging 26:1261–1270.
Vaupel JW (1998) Biodemographic Trajectories of Longevity. Science 280:855–860.
Vaupel JW (2010) Biodemography of human ageing. Nature 464:536–542.
Westlye LT, Walhovd KB, Dale AM, Bjørnerud A, Due-Tønnessen P, Engvig A, Grydeland H,
Tamnes CK, Ostby Y, Fjell AM (2010) Life-span changes of the human brain white mat-
ter: diffusion tensor imaging (DTI) and volumetry. Cereb Cortex 20:2055–2068.
Wilson RS, Beckett LA, Barnes LL, Schneider JA, Bach J, Evans DA, Bennett DA (2002)
Individual differences in rates of change in cognitive abilities of older persons. Psychol
Aging 17:179–193.
Yap QJ, Teh I, Fusar-Poli P, Sum MY, Kuswanto C, Sim K (2013) Tracking cerebral white mat-
ter changes across the lifespan: insights from diffusion tensor imaging studies. J Neural
Transm 120:1369–1395.
Zanto TP, Pa J, Gazzaley A (2014) Reliability measures of functional magnetic resonance imag-
ing in a longitudinal evaluation of mild cognitive impairment. Neuroimage 84:443–452.
Ziegler G, Dahnke R, Jäncke L, Yotter RA, May A, Gaser C (2012) Brain structural trajectories
over the adult lifespan. Hum Brain Mapp 33:2377–2389.
183
Interpreting Age-Related Differences

in Memory-Related Neural Activity
Michael D. Rugg
L ike several other cognitive abilities, episodic memory—

accessible memory for unique events—declines with increasing age
consciously
(Nilson, 2003; Salthouse, 2010). Over the past two decades or so, several accounts
have been proposed to explain the cognitive and neural bases of this decline.
According to “single-factor” accounts, for example, episodic memory is just one of
several cognitive abilities impacted by aging by virtue of its dependence on a com-
mon age-sensitive process or set of processes (e.g., Luo and Craik, 2008; Salthouse,
1996). By contrast, other accounts focus on age-sensitive processes associated spe-
cifically with memory. For example, one such account proposes that age-related epi-
sodic memory decline reflects impairment of a “pattern separation process” that is
critically dependent on the hippocampus (Yassa and Stark, 2012; see Chapter 11).
An important source of evidence relevant to these and other accounts comes from
investigations of the effects of age on the neural correlates of memory processing
identified by noninvasive measures of brain activity. In recent years this evidence has
come largely from studies employing functional magnetic resonance imaging (fMRI)
(Daselaar and Cabeza, 2013; Maillet and Rajah, 2014; see chapter 12) as well as, to
a lesser extent, studies using electrophysiological measures such as event-related
potentials (ERPs) (Friedman, 2013).
In the present chapter a distinction is drawn between age effects and aging effects.
An age effect (equivalently, age-related effect, age-related difference) exists when
a dependent variable differs between samples of people with different mean ages
(typically, young adults aged around 18–30 years, and older adults aged around
65–75 years). An age effect is a descriptive term, and implies nothing about the rea-
sons for the effect. By contrast, an aging effect refers to an age-related difference in a
183
184
184â•… Methods and Issues
dependent variable that can be attributed to a time-â•‰dependent change in the variable,

that is, to the “process” of aging. It is argued here that while the advent of functional
neuroimaging has allowed us to learn much over the past two decades or so about
the effects of age on the neural correlates of episodic memory, we know considera-
bly less about the effects of aging. The reasons for this rather pessimistic conclusion
are largely methodological. They include inferential problems arising from the use
of cross-â•‰sectional experimental designs, the failure to separate neural correlates of
aging from those associated with age-â•‰related differences in task performance, and the
employment of tasks where performance can be supported by more than one cognitive
mechanism. Although these issues are by no means specific to the cognitive neurosci-
ence of memory and, in the case of the inferences that can legitimately be drawn from
cross-â•‰sectional data, have been widely discussed and debated in the broader aging
literature, the present chapter will focus on their relevance for the understanding of
age-â•‰related memory decline. The chapter concludes on a more optimistic note with a
discussion of how cross-â•‰sectional data from functional neuroimaging studies can con-
tribute to the understanding of the relationship between memory and the brain, and
how this relationship might vary across the healthy adult lifespan.
Limitations of Cross-â•‰Sectional Designs
With only a few exceptions (e.g., Nyberg, et al., 2010; Persson et al., 2012; 2014),
functional neuroimaging studies of cognitive aging have employed cross-â•‰sectional
(contrasts between two or more age groups at a single time-â•‰point) rather than lon-
gitudinal (contrasts within-â•‰subjects at two or more time-â•‰points) designs to examine
the effects of age on the neural correlates of memory processing. The interpretational
problems that arise with such designs have been much discussed (e.g., Hofer and
Sliwinkski, 2001; Raz and Lindenburger, 2011; but see also Salthouse, 2010, 2014),
and will not be rehearsed in detail here. Four issues with particular relevance for the
cognitive neuroscience of memory and aging, and which have received relatively little
attention in the field, are however discussed below.
Cross-Sectional Designs Confound Aging, “Survivor,”

and Birth Cohort Effects
When groups of healthy individuals differing in age are contrasted, it is tempting
(and, in cognitive neuroscience, very common) to attribute any differences observed
in the dependent variables to the effects of aging. There are however at least two other
sources of age-â•‰related variance that should be considered before accepting such a
conclusion. Survivor effects refer to the fact that the inclusion and exclusion criteria
for eligibility in a typical functional neuroimaging study of aging make it almost inev-
itable that older adults are recruited more selectively than are young adults. For exam-
ple, few young adults will be excluded from a study by virtue of their health status,
but this is a fairly common reason for exclusion among older adults. In other words,
cross-â•‰sectional designs confound chronological age with the propensity for successful
185
Interpreting Age-Related Differences in Memory-Related Neural Activity 185
aging: whereas older subjects are typically recruited only from the subset of people
who aged successfully, young subject samples are more heterogeneous, comprising
people who are destined to vary widely in how successfully they age. To truly match
young and older samples, it would be necessary to select only those young subjects
who will go on to age sufficiently successfully to still be eligible for inclusion if they
were to reach the same age as the older sample; an impossible task at present. Thus, an
unknown proportion of the age-related variance identified in a typical cross-sectional
study will reflect the fact that older subject samples are more highly selected for the
capacity to age successfully than are young samples.
This is not merely an academic point: on the basis of a dissociation between
cross-sectional and longitudinal findings, Nyberg et al. (2010; see also Chapter 6)
argued that the often-reported finding of age-related right frontal over-recruitment—
greater activation in right prefrontal cortex during memory encoding in older than
in younger subjects (e.g., Morcom et al., 2003)—is a consequence of selection bias
rather than an age-related change in a pattern of brain activation. The authors reported
that while older individuals demonstrated greater right prefrontal activity than young
adults at the group level, longitudinal assessment showed that right prefrontal activity
decreased, rather than increased, with increasing age. Nyberg et al. (2010) argued
that their older subjects represented a successfully aging sub-group of the population
in whom right prefrontal “over-recruitment” is always present, albeit declining over
time. A subsequent study from the same group (Pudas et al., 2013) provided further
evidence in support of this proposal. To the extent that the proposal is correct, the
hypothesis that “right frontal over-recruitment” reflects a compensatory response to
deleterious effects of aging (e.g., Cabeza et al., 2002; Park and Reuter-Lorenz, 2009)
will be difficult to sustain.
A related issue is the more general one of cohort effects. Even when their socioeco-
nomic and educational status are matched, people born at different times inevitably
grow up in different circumstances; for example, most of today’s 25-year-olds will
have been exposed for much of their childhoods to computer-based technologies that
simply did not exist when today’s 70-year-olds were children. It is difficult to gauge
the specific impact on cognitive performance or brain organization of each of the
myriad of such factors that differ across generations. Nonetheless, birth cohort is an
important source of variance on psychometric test scores, as evidenced most famously
by the “Flynn Effect”—the finding that, as assessed by periodic re-norming of stan-
dardized intelligence tests, full-scale IQ has ostensibly risen by around 3 points per
decade for the past 50 years or so (Trahan et al., 2014).
Importantly, the Flynn Effect is not confined to performance on tests of intelli-
gence. Comparison of the performance of similarly aged groups of individuals tested
at different times over a 15-year period demonstrated that episodic memory, whether
assessed by recognition or recall, also shows a marked birth cohort effect (Ronnlund
and Nilsson, 2009; Baxendale, 2010). On the basis of a hierarchical regression analy-
sis Ronnlund and Nilsson (2009) concluded that three variables—childhood nutrition,
family size, and educational experience—were the likely mediators of the effects that
they reported. Whatever the explanation, it seems likely that birth cohort effects con-
tribute to the different trajectories of age-related episodic memory decline that are
estimated from cross-sectional studies (where a decline over the lifespan beginning
186
around age 30 has been reported; e.g., Park et al., 2002) and longitudinal studies
(where age effects may not emerge until age 60 or so; e.g., Ronnlund et al., 2005, but
see Salthouse, 2014, for an alternative perspective).
The finding that episodic memory performance likely varies by birth cohort is
of considerable relevance for cognitive neuroscience studies of aging and memory.
Age-related reductions in memory performance are almost invariably reported in
such studies, but the extent to which these reductions should be attributed to cohort
rather than aging effects (as well as the extent that reductions might be under-
estimated by virtue of survivor effects, see above) is difficult to assess. Clearly,
though, if the aim of a study is to identify age-related differences in functional brain
activity (or in brain structure) that underlie the effects of aging on memory, this
is a crucial question. To anticipate experimental findings discussed in more detail
later: we recently described an fMRI study of associative memory encoding con-
trasting groups of young, middle-aged and older individuals. Performance on the
experimental memory task was worse in the older subjects than in the middle-aged
individuals, who in turn were out-performed by the younger group. The findings of
Ronnlund and Nilsson (2009) raise the possibility that these differences in memory
performance reflect, at least partially, when the subjects in the three age groups
were born rather than their age when they were tested. Disregarding the additional
complications arising from possible survivor effects, the inability to partition
between-group variance in memory performance between cohort and aging effects
greatly complicates the interpretation of the age-related differences in functional
brain activity that we observed. Put simply, we cannot know the extent to which the
neural differences are correlates of the deleterious effects of aging on memory if we
don’t know the size of these effects.
Cross-Sectional Designs Do Not Permit Individual Variation in

Cognitive Performance or Brain Activity to Be Partitioned
Between Age-Invariant and Age-Related Factors
Up to 50% of the variance in IQ test scores among healthy people who reach their
70s is shared with variance in their IQ at age 11 (e.g., Deary et al., 2004). There
seems little reason to suppose that things are much different for episodic memory,
although to our knowledge data directly relevant to this issue have not been reported.
The stability across the lifespan in estimates of cognitive ability has an important
implication for the interpretation of individual differences in memory performance
among older individuals: it cannot be assumed that the individuals with the lowest
memory performance are those for whom memory has been the most affected by
aging. Exemplifying this point, Pudas et al. (2014) reported that the correlation in a
large sample of healthy adults between their memory performance in “mid-life” (age
35–60 years) and their performance 15–20 years later was almost 0.7. The authors
further reported that, in combination, mid-life memory performance and the change in
performance between the earlier and the current test occasions accounted for as much
variance in encoding-related neural activity (assessed with fMRI) as did the current
level of memory performance.
187
The advantages of adopting a longitudinal perspective when examining the neural

correlates of individual differences in cognitive performance are further illustrated
by a study in which “task-related deactivation” in groups of middle-aged adults who
had aged “successfully” or “unsuccessfully” was contrasted (Hansen et al., 2014).
Rather than just ranking the adults by their current cognitive performance, Hansen et
al. (2014) used cognitive test scores from the same individuals when they were aged
around 19 years to identify people who were now scoring either higher or lower than
would have been predicted from their performance on the earlier test (the correlation
between the two sets of scores was around 0.7; see Figure 7.1). Thus, the authors were
able to identify two groups of middle-aged individuals who were matched for cogni-
tive ability at age 19, but who now differed in ability, presumably as a consequence
of differential aging. Hansen et al. (2014) reported that, as assessed with fMRI, task-
related deactivation of the default mode network was greater in the high performers.
(A) 60
A
Cognitive ability in middle-age
40
B
20
0
0 20 40 60 80
Cognitive ability in youth
(B)
1.5
1
*
0.5
–0.5
Gr A Gr B
Figure 7.1 A: Relationship between cognitive ability at approximate age 20 years (youth), and
cognitive ability at approximate age 46 years (middle-age). Solid and open circles represent
individuals identified as demonstrating well-preserved and declining cognitive ability respec-
tively. B: Mean task-related deactivation in the indicated regions for individuals with well-
reserved (Gr A) and declining (Gr B) cognitive ability respectively. Re-drawn with permission
of the authors from f igures 2 and f igure 4 of Hansen et al. (2014).
188
Because of how the groups were selected, this difference in functional brain activity
can likely be attributed to factors associated with differential aging. Of course, this
does not mean that the neural differences are a consequence of different aging trajec-
tories; the data are equally compatible with the possibility that the differences were
also present earlier in life when cognitive performance was matched between the two
groups, and that magnitude of task-related deactivation is predictive, rather than a
consequence, of “successful” aging (cf. Nyberg et al., 2010).
Inferences About Age-Related Changes in Cognitive Performance

and Their Neural Correlates are Difficult, If Not Impossible,
to Make from Cross-Sectional Data
As has been discussed previously (e.g., Lindenberger et al., 2014) there is no neces-
sary relationship between an age effect identified in a cross-sectional design and an
effect of age on the same dependent variable that is identified longitudinally—the two
classes of effect can be positively correlated, orthogonal, or even, as in the study of
Nyberg et al., (2010) discussed above, negatively correlated. Importantly, the absence
of a fixed relation between cross-sectional and longitudinal findings cautions against
the use of cross-sectional findings to develop models in which age is treated as a medi-
ating or a causal variable to explain age-related variance in functional brain activity
(Raz and Lindenberger, 2011).
Comparison of the fMRI BOLD Signal Across Age Groups Are

Potentially Confounded by Differences in the Hemodynamic
Response Function
Finally, it is worth mentioning one other variable with the potential to complicate the
interpretation of age-related effects on task-related neural activity, albeit one relevant
largely for studies employing the fMRI BOLD signal (and no less relevant for studies
employing longitudinal than cross-sectional designs). The variable in question is cere-
brovascular reactivity (CVR), a measure of vasodilation. CVR is an important non-
neural determinant of BOLD signal magnitude (for the same level of neural activity,
greater CVR gives a larger BOLD signal; see for example Liu et al., 2013 and c hapter
1). Crucially, CVR declines with increasing age in much of the brain (Figure 7.2).
Therefore, other factors being equal, an event eliciting an equivalent neural response
in young and older subjects is likely to elicit a smaller BOLD response in the latter
group, potentially leading to an erroneous finding of age-related neural under-recruit-
ment. Analogously, without correction for differences in CVR, the magnitude of age-
related over-recruitment is likely to be under-estimated. Both of these sources of error
have been identified in studies where age effects in fMRI data were compared before
and after the data had been corrected for between-subject differences in CVR (e.g.,
Lui et al., 2013; Tsvetanov, et al., 2015).
The broad impact of age-related differences in CVR on fMRI studies of the effects
of age on neural correlates of episodic memory remains to be established. Nonetheless,
189
V1 area MTL
CVR (%/mmHg) 0.4 0.4
CVR (%/mmHg)
0.3 0.3
0.2 0.2
0.1 p = 0.0021 0.1

p = 0.0081
0 0
20s 30s 40s 50s 60s 70s 80s 20s 30s 40s 50s 60s 70s 80s
Left IFG Right IFG

0.4 0.4
CVR (%/mmHg)
CVR (%/mmHg)
0.3 0.3
0.2 0.2
0.1 0.1
p < 0.0001 p = 0.0001
0 0
20s 30s 40s 50s 60s 70s 80s 20s 30s 40s 50s 60s 70s 80s
Figure 7.2 Cerebrovascular reactivity (CVR) as a function of age in V1, medial temporal lobe
(MTL) and left and right inferior frontal gyrus (LIFG, RIFG). Data from figure 4 of Liu et al.
(2013). Reproduced with permission of the authors.
the findings noted above raise an important caveat in the interpretation of fMRI data
uncorrected for between-subject differences in CVR. This caveat extends to the find-
ings from our own laboratory discussed below.
What Can Be Concluded from Cross-Sectional Findings?
In light of the foregoing discussion, it might seem that there is little to be gained from
the employment of cross-sectional designs to study cognitive aging and its neural
correlates. Despite the limitations of these designs, this is not the case. Above all, the
simple pragmatics of the study of aging must be acknowledged: it is just not feasi-
ble, either on logistical or economic grounds, to address every question we need to
answer about the effects of age on cognition and the brain with a purely longitudinal
approach. At the least, cross-sectional data can play a heuristic role in identifying
putative changes in task-related neural activity that are associated with age-related
memory decline (or, perhaps, with resistance to decline). Below, this and two other
ways that cross-sectional findings can be of value are discussed.
190
True Age Effects on the Neural Correlates of a Memory Process Require

an Explanation, and Motivate Hypotheses about the Effects of Aging
on Memory that Can Be Tested in Longitudinal Studies
Whereas a difference between age groups in the neural correlates of a memory pro-
cess does not constitute direct evidence of an aging effect, the finding still requires an
explanation and can be of value in shaping the questions and hypotheses examined
in more resource-intensive longitudinal studies. One obvious hypothesis is that such
a difference does in fact reflect an age-related change in the relationship between a
memory process and its neural substrates (i.e., an aging effect). Alternative possibil-
ities are not necessarily any less interesting, however, despite their more tangential
relevance to understanding how aging affects the brain. For example, the finding of
a pattern of task-related brain activity that is shared by older adults and a subset of
young individuals who go on to age “successfully” would be of considerable interest,
motivating research to understand what it is about this pattern that makes it a predictor
of successful aging. As discussed above, right prefrontal over-recruitment might cor-
respond to such a pattern (Nyberg et al., 2010; although see Duverne et al., 2009, and
de Chastelaine et al., 2011, for evidence that right prefrontal over-recruitment, at least
when in the form of “subsequent memory effects,” is negatively correlated with mem-
ory performance). In a similar vein, an age-related difference in a neural correlate
of memory (or any other cognitive function) that turned out to reflect a birth cohort
effect rather than an aging effect would potentially offer insight into how different life
experiences influence brain-behavior relationships in adulthood.
The heading to the present section refers to “true” age effects on a neural correlate
of memory. Here, a true effect is one that does not result from the confounding of the
variables of age and memory performance. When older individuals perform less well
than young individuals on a test of episodic memory, the interpretation of age-related
differences in associated neural measures becomes ambiguous (Rugg and Morcom,
2004): simply put, do the neural differences reflect the effects of age or performance?
For reasons discussed previously (Rugg and Morcom, 2004), other things being equal
(admittedly, they seldom are), the difference in the neural activity elicited by later
remembered and later forgotten study items, or by test items associated with success-
ful and unsuccessful retrieval, will be inversely proportional to memory accuracy,
with an expectation of zero difference when performance is at chance. Thus, when
older subjects demonstrate lower memory performance than young subjects (which
they almost invariably do), the finding that older subjects also demonstrate smaller
encoding-or retrieval-related neural effects in one or more brain regions need not
necessarily reflect the effects of age, or at least, not in a manner that is directly infor-
mative about how aging affects brain function.
One way to disambiguate age and performance effects is to employ experimental
designs in which performance is matched across age groups. For example, in a study
of the effects of age on memory encoding (Morcom et al., 2003), we manipulated the
variable of study-test interval to allow subsequent memory effects to be compared
across age groups when memory performance was equivalent (by comparing subse-
quent memory effects predictive of performance after a short delay in older subjects,
and after a longer delay in the young sample). Additionally, in several studies of the
191
effects of age on neural correlates of memory retrieval, we manipulated the number

of study presentations to vary the difficulty of subsequent retrieval judgments, again
permitting retrieval-related neural activity to be contrasted across age groups when
performance was matched (e.g., Morcom et al., 2007; Duverne et al., 2008; see also
Angel et al., 2013). In another study (Wang et al., 2012), we matched performance by
selecting subsets of high-and low-performing older and young subjects.
Of course, none of these approaches is without its problems. For instance, the
manipulation of study-test delay depends on the assumption that the encoding pro-
cesses supporting memory at the different delays are functionally and neurally equiv-
alent, whereas matching performance by varying number of study presentations is
valid only if the study manipulation does not lead to the formation of qualitatively
different memory representations. And the strategy of selecting matched sub-groups
depends on the assumptions that the sub-groups are representative of the groups as a
whole (cf. Nyberg et al., 2010), and that the reliability of the performance measure
used to stratify each group is sufficiently high to prevent “illusory” matching caused
by regression to the mean (Barnett et al., 2005). Whereas each of these assumptions
can be tested to some degree (in the cases of delay and study presentation manip-
ulations, for example, by using fully crossed designs that permit the effects of the
manipulation to be examined within each group), the ideal approach would be to look
for convergence across different matching methods.
Another approach to dealing with age-related differences in performance is to sta-
tistically control for the effects of variation in performance across subjects. Of course,
the validity of this approach also rests on crucial assumptions, not the least of which is
that any relationship between performance and the neural measures of interest is linear.
To illustrate the value of the approach, we describe two examples of findings where,
on the face of it, age modulated the magnitude of an fMRI memory effect but where
further analyses led to quite different conclusions in the two cases. The first example
comes from a study conducted on 136 healthy individuals (36 young (20–30 years
of age), 36 middle-aged (45–55 years old), and 64 old (aged 63–77 years); see de
Chastelaine at al., 2015, for full details). The subjects were scanned while they per-
formed an associative encoding task and, subsequently, while they undertook a test
of associative recognition. Each encoding trial comprised the presentation of a pair
of concrete words with the requirement to decide which denoted object would “fit”
into which. The associative recognition test required discrimination between “intact”
pairs (word pairs repeated from the study phase), “rearranged” pairs (pairs comprising
studied words, but coming from different study trials), and “new” pairs (pairs com-
prising two unstudied words). The neural correlates of successful recollection were
operationalized by the contrast between intact test pairs that were either correctly
judged intact or wrongly endorsed as rearranged.
In line with prior findings (Old and Naveh-Benjamin, 2008), we found that asso-
ciative recognition performance in the young subject group exceeded that in the
old group, with the performance of the middle-aged subjects falling in between (de
Chastelaine et al., 2015). Turning to the fMRI findings from the associative recogni-
tion test, Figure 7.3A illustrates the associative recognition (“recollection success”)
effects identified in the hippocampus in each of the three groups: as is evident from
the figure, there was a significant main effect of group, and follow-up tests revealed
192
(A) 1.5 **
Hippo recollection effect

*
0.5
0
Young Middle Old
(B)
Hippo recollection effect (adj)
1.5 ns
ns
1
0.5
0
Young Middle Old
Figure 7.3 A: Recollection-related activity (successful recollection > unsuccessful recollec-

tion) in the anterior hippocampus of sample of young (N = 36), middle-aged (N = 36) and older
(N = 72) subjects. Middle-aged and older subjects demonstrated smaller recollection effects
than did the young. B: The same data after controlling for recollection performance. Differences
between the groups are nonsignificant. Figure based on data reported in de Chastelaine et al.
(2016).
that the recollection effects were larger in the young group than in either of the other
two groups, consistent with some prior findings (e.g., Daselaar et al., 2006). Figure
7.3B illustrates the data after controlling for between-subjects differences in associa-
tive recognition performance: differences between the age groups in the fMRI effect
are now far from significant. A parsimonious explanation of these findings is that
the age-related differences in hippocampal recollection effects illustrated in Figure
7.3A are not “true” age effects, as these were defined above, but instead reflect an
age-independent relationship between hippocampal recollection-related activity and
memory performance (a conclusion bolstered by the finding that, after controlling
for age, hippocampal recollection effects and memory performance were significant
correlated across the entire sample; r = 0.25, p < .005). Thus, the findings illustrated
in Figure 7.3A do not support the conclusion that recollection-related hippocampal
activity is modulated by age. Rather, the magnitude of the activity in older subjects is
no different from that expected in young subjects with comparable levels of memory
performance.
Examination of the effects of age on a neural correlate of successful associa-
tive encoding leads to a quite different conclusion. The relevant data come from
two studies—one the study just described, and an earlier study (de Chastelaine
193
et al., 2011) that employed the same experimental procedure but different samples
of young (n = 18) and older (n = 36) subjects. A near-ubiquitous finding in fMRI
studies of episodic encoding is that study items that go on to receive accurate judg-
ments on a subsequent memory test elicit lower activity than items that go on to
receive inaccurate judgments in a characteristic set of brain regions that include
members of the default mode network (Rugg et al., 2015). It has consistently been
reported that these “negative subsequent memory effects” are smaller in older than
in younger individuals (see Maillet and Rajah 2014 and Chapter 12 for reviews).
Figure 7.4 illustrates the effects identified in the experiment of de Chastelaine et
al. (2011). As is evident from the figure, and consistent with the prior literature,
the effects were markedly greater in the young than in the old group; indeed, in
this study the older subjects’ effects did not differ significantly from zero (unlike in
de Chastelaine et al., 2015, when the older subjects’ negative effects were reliably
different from zero). Crucially, these age-related differences in negative subsequent
memory effects remained when between-subjects differences in subsequent mem-
ory performance were partialled out. Together with the findings of de Chastelaine
et al. (2015), where age-related differences in negative subsequent memory effects
also remained after controlling for memory performance, these findings suggest
that age-related attenuation of negative subsequent memory effects reflects a “true”
effect of age on a neural correlate of memory processing. While the findings do not
necessarily imply that negative effects decline in magnitude over the lifespan (see
section 1), they do suggest that this is a hypothesis worth pursuing. Additionally,
and perhaps no less important, the findings present a challenge to any proposal
that negative subsequent memory effects reflect processes that are necessary for
successful encoding: the data depicted in Figure 7.4 suggest that some people (here,
high-functioning older adults) are able to successfully encode episodic memories
seemingly without the benefit of these processes.
An analogous conclusion can be drawn from an ERP study of recognition mem-
ory by Wang et al. (2012). The authors reported that a neural correlate of familiarity
strength that was robustly present in a sample of young subjects was undetectable
in a group of older individuals matched for the strength of their familiarity-driven
0.4 Young Old
–0.0
Parameter estimate
–0.4
–0.8 Assoc. hits

Assoc. misses
–1.2
–1.6
Figure 7.4 Negative subsequent memory effects in the young subject sample of de Chastelaine
et al. (2011), and the encoding-related activity elicited the older subject sample in the same
cortical regions. Figure 4 of de Chastelaine et al. (2011).
194
(A)
(B) 0.4
Recollection estimate
0.25
0.1
–0.05
–0.2
–0.35
–3 –1.5 0 1.5 3
K-R monitoring effect
Figure 7.5 (A) Right frontal and anterior cingulate “monitoring effects” (Know > Remember)
collapsed across the young and older subject sample from Wang et al. (2016). (B) Relationship
between recollection performance and the anterior cingulate monitoring effect after controlling
for age. Blue circles = young subjects, red circles = older subjects. Re-drawn from Wang et al.
(2016).
recognition (see also Duarte et al., 2006).This finding places a strong constraint on
hypotheses about the functional significance of the ERP effect.
Cross-Sectional Studies Can Identify Brain-Behavior Relationships

Sufficiently Fundamental that They Generalize Across the Lifespan
(or Birth Cohorts)
Although it might seem odd to extol null findings, the implications of a null effect
of age in an adequately powered cross-sectional study of healthy individuals should
not be overlooked. The finding that the relationship between a cognitive process and
a pattern of brain activity is age-invariant can serve to strengthen confidence that the
relationship represents a fundamental aspect of functional brain organization. This is
even more so if individual differences in performance on a task that is dependent on
the cognitive process in question co-vary with its putative neural correlate in an age-
invariant manner.
An example of such findings is found in an fMRI study that examined the effects
of age on neural correlates of episodic memory retrieval (Wang et al., 2016). The
study employed two groups of healthy subjects with mean ages of approximately 24
195
years and 68 years (Ns of 24 in each case). After a study phase in which randomly
intermixed pictures and words were presented, the subjects performed a recognition
memory task in which words denoting studied and unstudied items were subjected to
“Remember/Know/New” judgments (Tulving, 1985). As is frequently reported (Koen
and Yonelinas, 2014), measures of both “recollection” (derived from Remember judg-
ments) and “familiarity” (derived from Know judgments) were significantly lower
in the older group. For the present purpose, the critical fMRI contrast is one that is
held to identify the neural correlates of “post-retrieval monitoring,” a set of control
processes that, collectively, monitor and evaluate the products of a retrieval attempt
(Rugg, 2004). Monitoring is thought to be especially important when a test item elic-
its a weak or an ambiguous “memory signal,” as is hypothesized to occur more fre-
quently for items attracting Know than for Remember judgments (Henson et al., 1999;
see also Henson et al., 2000). Thus, it has been argued that regions where neural activ-
ity is greater for items accorded Know rather than Remember judgments include those
that support monitoring (Henson et al., 1999).
Shown in Figure 7.5A is the outcome of the main effect of the contrast between the
fMRI responses elicited by items attracting accurate Know and accurate Remember
judgments. Replicating prior findings (Henson et al., 1999), the contrast identified
enhanced activity for Know judgments in, among other regions, anterior cingulate and
right dorsolateral prefrontal cortex. The magnitude of these “monitoring effects,” both
in these two regions and elsewhere, did not differ between the age groups. Figure 7.5B
illustrates the relationship between the effects in the anterior cingulate and recollec-
tion accuracy (indexed as Remember hit rate—Remember false alarm rate). The two
variables were significantly correlated (r = .54, after partialling out age), and were
also significantly and equivalently correlated in each group separately. These findings
suggest that engagement of neural regions that support post-retrieval monitoring does
not necessarily vary with age (see McDonough et al., 2013, for an alternate finding).
More interestingly, the findings also suggest that the ability to engage these regions
might be an important determinant of memory accuracy across much of the healthy
adult lifespan (at least for relatively high-functioning adults).
These conclusions are of course subject to important caveats. First, and most
obviously, they rest on an acceptance of null results. It is possible that in a study
with greater statistical power age-related differences in the magnitude of monitoring
effects, or in the strength of their relationship with memory performance, would be
detected (that said, we found essentially identical results—no age differences in the
magnitude of monitoring effects, and a robust relationship with accuracy regardless
of age—in the aforementioned study of associative recognition memory that was sub-
stantially more highly powered (N = 136 rather than N = 48) than that of Wang et al.,
in press). Second, and relatedly, the conclusions are constrained by the methods that
were employed to measure monitoring-related neural activity. For example, it remains
to be seen whether studies employing alternative and perhaps more sensitive fMRI
analyses (e.g., multivariate rather than univariate approaches, Davis and Poldrack,
2013, or analyses that correct for age-related differences in the hemodynamic transfer
function, Liu et al., 2013, and section 1(iv)), will yield findings congruent with those
illustrated in Figure 7.5. It also remains to be seen whether the findings extend to
alternate methods for measuring task-related brain activity such as ERPs.
196
Brain-Memory Relationships that Differ According to Age Pose an

Important Challenge to “Generic” Cognitive Neuroscience Models of
Memory
The human adult lifespan extends over some seven decades, and the proportion of
the population in the sixth and seventh decades is expanding rapidly (Harper, 2014).
Nonetheless, the overriding majority of cognitive neuroscience studies aimed at elu-
cidating the neural bases of memory continue to exclusively employ subjects in their
first decade of adult life. It is increasingly clear however that the findings from studies
of young adults do not necessarily generalize to older individuals. Regardless of why
an age-related difference might exist in the relationship between brain activity and a
cognitive process, such differences pose a challenge for what one might call “one size
fits all” neurocognitive models, for example, the idea that negative subsequent mem-
ory effects reflect processes necessary for successful episodic encoding (see above).
Below, we extend this theme to examples of age effects on the relationship between
individual differences in memory performance and measures of brain activity and
structure.
Before describing these examples, it is important to recognize an important caveat
to the argument that age-related differences in the relationship between brain activity
and cognitive performance are problematic for “generic” neurocognitive theories. The
argument depends crucially on the assumption that subjects belonging to different age
groups perform an experimental task in the same way; that is, that they do not “solve”
the task by using different cognitive strategies and hence by engaging different cogni-
tive operations. To give a well-worn example (Rugg and Morcom, 2004), there is good
evidence that compared with young individuals, older subjects perform disproportion-
ately worse on memory tests that depend on recollection—the retrieval of qualitative
details about a study episode—than they do on tests that can be performed on the
basis of a general sense of familiarity, such as simple recognition memory (Koen and
Yonelinas, 2014). Additionally, there is abundant evidence that the neural correlates
of recollection and familiarity are qualitatively distinct (e.g., Johnson et al., 2013).
Thus findings that the neural correlates of successful recognition memory differ with
age (e.g., Tisserand et al., 2005) may merely reflect the fact that older people rely on
familiarity more heavily than younger people do. The findings may therefore have
little or no bearing on the question of whether one or more of the component processes
supporting recognition memory demonstrate age-related differences in their neural
correlates. An argument along these lines was advanced by Mattson et al. (2014) to
account for their finding that negative subsequent memory effects associated with
simple recognition memory judgments (which can be supported by both familiarity
and recognition) were attenuated in older relative to young adults, whereas negative
subsequent memory effects associated with accurate, highly confident source memory
judgments (which necessitate retrieval of qualitative information) were equivalent in
the two groups.
The specific example just described is fairly straightforward, and the problems that
it illustrates are seemingly easily obviated by the judicious choice of a memory test.
It is arguable however that even the most “process-pure” test may not be completely
immune to the engagement of different cognitive strategies. For example, it is widely
197
held that successful episodic retrieval is associated with the “reinstatement” of the
processes and representations that were active at the time the episode was experienced
(Danker and Anderson, 2010). Thus, something as simple as an age-related difference
in the cognitive processes preferentially engaged by a given study task may lead to
age-related differences not only in encoding-related neural activity, but in retrieval-
related activity also, reflecting in this latter case patterns of reinstatement effects that
reflect the different cognitive operations engaged by young and older subjects during
study (cf. McDonough et al., 2014; Wang et al., 2016). These differences might shed
light on the cognitive strategies characteristically engaged by people of different ages,
but they offer little insight into whether or how the neural correlates of memory pro-
cesses engaged in common differ with age.
With this caveat in mind, we give two examples of findings where it can be argued
that an age-related dissociation in a brain-behavior relationship suggests that the func-
tional significance of a neural measure varies in an age-dependent manner. The first
example comes from the study of de Chastelaine et al. (2015) described previously. In
addition to examining negative subsequent memory effects as a function of age group
(young, middle, and older), we also assessed what are sometimes referred to as “task
effects”—the simple effects of stimulus presentation with respect to the inter-stimulus
baseline—in the same brain regions. The effects took the form of the well-known
“task-related deactivation” that is characteristic of the components of the default
mode network (Gusnard and Raichle, 2001). Replicating numerous prior findings, the
effects were reliably larger in the young group than in either of the other age groups (a
finding that remained reliable after between-subjects differences in performance were
partialled out). Crucially, the relationship across subjects between associative recog-
nition performance and the magnitude of task-negative effects differed between the
age groups. As is illustrated in Figure 7.6, whereas there was a nonsignificant trend
toward a negative relationship between task effects and performance in the young
group, and no sign of any relationship in the middle-aged subjects, there was a sig-
nificant positive relationship between the two variables in the older group, such that
larger task-negative effects were associated with higher performance (see Miller et
al., 2008, for similar findings). Whatever the explanation for these findings (we spec-
ulated that they might reflect individual differences among our older subjects in beta-
amyloid deposition, cf. Sperling et al., 2009), they complicate any general account of
the functional significance of the phenomenon of task-related deactivation in respect
of associative memory encoding.
The second example of an age-related dissociation between a neural measure and
memory performance derives from the same study, but involves a measure of brain
structure, rather than function. The significance of the example lies in the fact that
it constitutes a clear age-related double dissociation, such that the direction of the
relationship between the neural measure and memory performance is reversed as a
function of age. The relevant findings are summarized in Figure 7.7, which depicts
the across-subjects relationship between mean cortical thickness and associative rec-
ognition performance in each of the three age groups. As is evident from the figure,
whereas this relationship was positive in older subjects (consistent with prior reports;
Fjell and Walhovd, 2010), the relationship was negative in the young group, where
a thinner cortex was associated with higher associative recognition performance.
198
0.5 r = .284, n.s.
pR 0.0
–0.5
–1.6 0.0 1.6
0.5 r = –.307, p < .025
0.0
pR
–0.5
–1.6 0.0 1.6
Mean task-negative effect
Figure 7.6 Relationship between recollection performance (pR) and task-negative effects

(after controlling for age) in the young (upper panel) and older (lower panel) subjects of de
Chastelaine et al. (2015).
A closely analogous dissociation was reported for the relationship between cortical
thickness and full-scale IQ in an elegant study that combined cross-sectional and lon-
gitudinal approaches (Schnack et al., 2015), and the present finding is also reminiscent
of findings that hippocampal volume and memory performance have been reported
to be negatively correlated in young adults (Van Petten, 2004). Following Schnack
et al. (2015), we speculate that the dissociation illustrated in Figure 7.7 reflects the
dynamic character of brain-behavior relationships across the lifespan. Specifically,
whereas greater cortical “pruning”—resulted in thinning over the course of childhood
and adolescence—leads to greater neural efficiency early in life (Kharitonova et al.,
2013), efficiency in later life depends on how successfully this “pruned” pattern of
neural circuitry can be maintained. Regardless of the merit of this account, findings
such as those illustrated in Figure 7.7 and reported by Schnack et al. (2015) empha-
size just how dramatically the relationship between a neural variable and a cognitive
measure can vary as a function of age. Obviously, the findings rule out any account
of the relationship between cortical thickness and cognitive performance that fails to
take age into account. It remains to be seen whether, and how frequently, analogous
â•‡ 199
Interpreting Age-Related Differences in Memory-Related Neural Activityâ•… 199
2.8
Mean cortical thickness (mm)

2.6
2.4
r = –.393, p < .025

2.2
0 0.2 0.4 0.6 0.8
pR
2.6
Mean cortical thickness (mm)
2.4
2.2
r = .324, p = .01
2
0 0.2 0.4 0.6 0.8
pR
FigureÂ€7.7â•‡Relationship between recollection performance (pR) and mean cortical thick-

ness (after controlling for age) in young (upper panel) and older (lower panel) subjects
(unpublished data).
findings will emerge with respect to measures of brain function rather than brain
structure; a tantalizing hint that these effects might exist is given in Figure 7.6. If
they can be found, such double dissociations would make a compelling case for the
importance of routinely including subject samples drawn from across the lifespan in
functional neuroimaging studies of memory and other cognitive domains.
Conclusion
Understanding how episodic memory is affected by aging, and why some people’s
memories are seemingly affected more than others, has been a major goal of the
cognitive neuroscience of aging for more than two decades. It is argued here that
200
200â•… Methods and Issues
findings from these studies have been highly informative about age-â•‰related differ-
ences in memory and their underlying neural correlates, but less so about the rela-
tionship between age-â•‰related changes in memory and concomitant changes in brain
function. This latter question is difficult to address within the cross-â•‰sectional para-
digm that has predominated in the field from the earliest functional neuroimaging
studies of aging. Although they are logistically and economically challenging, it
seems clear that the field stands to benefit hugely from the adoption of longitudinal
approaches, as is evidenced by the few reports that have emerged from longitudinal
studies to date.
It must be emphasized however that the present chapter is not intended to deride cross-â•‰
sectional studies of aging. Despite their limitations, for the reasons discussed in section
2 above the findings from such studies can be highly informative. This is especially so
in the broader context of attempts to develop accounts of the relationship between mem-
ory processes and their neural underpinnings, where the evidence indicates that such
accounts may not always generalize across people of different ages. An important chal-
lenge for the future is the development of theoretical models that can smoothly account
for age-â•‰related variation in the relationship between memory and the brain.
Acknowledgments
Some of the research described within this chapter was supported by the National
Institute on Aging (grants R01AG039103 and 5P50AG16573). The author gratefully
acknowledges the contributions of Marianne de Chastelaine and Brian Donley to the
collection and analysis of the unpublished data illustrated in Figure 7.7.
References
Angel L, Bastin C, Genon S, Balteau E, Phillips C, Luxen A, Maquet P, Salmon E, Collette F
(2013) Differential effects of aging on the neural correlates of recollection and familiarity.
Cortex 49:1585–â•‰1597.
Barnett AG, van der Pols JC, Dobson AJ (2005) Regression to the mean: what it is and how to
deal with it. Internat J Epidemiol 2005, 34:215–â•‰220
Baxendale, S. The Flynn effect and memory function (2010) J Clin Exp Neuropsychol. 32: 699–â•‰703.
Cabeza R, Anderson ND, Locantore JK, McIntosh AR. (2002) Aging gracefully: compensatory
brain activity in high-â•‰performing older adults. Neuroimage 17:1394–â•‰1402.
Danker JF, Anderson JR (2010) The ghosts of brain states past: remembering reactivates the
brain regions engaged during encoding. Psychol Bull 136:87–â•‰102.
Daselaar SM, Fleck MS, Dobbins IG, Madden DJ, Cabeza R (2006) Effects of healthy aging
on hippocampal and rhinal memory functions: an event-â•‰related fMRI study. Cereb Cortex
16:1771–â•‰1782.
Daselaar S, Cabeza R (2013) Age-â•‰related decline in working memory and episodic memory. In
K.N. Ochsner and S. Kosslyn (eds.) Oxford Handbook of Cognitive Neuroscience, Vol. 1.
Davis T, Poldrack RA (2013) Measuring neural representations with fMRI: practices and pit-
falls. Ann NY Acad Sci. 1296:108–â•‰134.
de Chastelaine M., Wang TH, Minton B, Muftuler LT, Rugg MD (2011) The effects of age,
memory performance and callosal integrity on the neural correlatesof successful associa-
tive encoding. Cereb Cortex 21:2166–â•‰2176.
201
de Chastelaine M, Mattson JT, Wang TH, Donley BE. Rugg MD. (2015) Sensitivity of
Negative Subsequent Memory and Task-Negative Effects to Age and Associative Memory
Performance. Brain Res. 1612: 16–29.
Deary IJ, Whiteman MC, Starr JM, Whalley LJ, Fox HC (2004) The impact of childhood intel-
ligence on later life: following up the Scottish mental surveys of 1932 and 1947. J Pers
Soc Psychol 86:130–147.
Duarte A, Ranganath C, Trujillo C, Knight RT (2006) Intact recollection memory in high-
performing older adults: ERP and behavioral evidence. J Cogn Neurosci 18:33–47.
Duverne, S, Habibi A, Rugg MD (2008) Regional specificity of age effects on the neural cor-
relates of episodic retrieval. Neurobiology of Aging 29:1902–1916.
Duverne S, Motamedinia S, Rugg, MD (2009) The relationship between aging, performance,
and the neural correlates of successful memory encoding. Cereb Cortex 19:733–744.
Fjell AM, Walhovd KB (2010) Structural brain changes in aging: courses, causes and cognitive
consequences. Rev Neurosci 21:187–221
Friedman, D (2013) The cognitive aging of episodic memory: a view based on the event-related
brain potential. Front Behav Neurosc 7:111.
Gusnard DA, Raichle ME (2001) Searching for a baseline: functional imaging and the resting
human brain. Nat Rev Neurosci 2:685–694.
Hansen NL, Lauritzen M, Mortensen EL, Osler M, Avlund K, Fagerlund B, Rostrup E (2014)
Subclinical cognitive decline in middle-age is associated with reduced task-induced deac-
tivation of the brain’s default mode network. Hum Brain Mapp 35:4488–4498.
Harper S (2014) Economic and social implications of aging societies. Science 346:587–591.
Henson RNA, Rugg MD, Shallice T, Josephs O, Dolan RJ (1999) Recollection and familiarity
in recognition memory: an event-related functional magnetic resonance imaging study. J
Neurosci. 19:3962–3972.
Henson RN, Rugg MD, Shallice T, Dolan RJ (2000) Confidence in recognition memory
for words: dissociating right prefrontal roles in episodic retrieval. J Cogn Neurosci
12:913–923.
Hofer SM, Sliwinski MJ (2001) Understanding ageing. An evaluation of research designs for
assessing the interdependence of ageing-related changes. Gerontology 47:341–352.
Johnson JD, Suzuki M, Rugg MD (2013) Recollection, familiarity, and content-sensitivity in
lateral parietal cortex: A high-resolution fMRI study. Front Hum Neurosci 7:219.
Kharitonova M, Martin RE, Gabrieli JD, Sheridan MA (2013) Cortical gray-matter thinning is
associated with age-related improvements on executive function tasks. Dev Cogn Neurosci
6:61–71
Koen JD, Yonelinas AP (2014) Recollection and familiarity in healthy aging, amnestic Mild
Cognitive Impairment, and Alzheimer’s disease: a review. Neuropsychol Rev 24:332–354
extraction: what’s change got to do with it? Psychol Aging 26:34–47.
Liu P, Hebrank AC, Rodrigue KM, Kennedy KM, Section J, Park DC, Lu H (2013) Age-related
differences in memory-encoding fMRI responses after accounting for decline in vascular
reactivity. Neuroimage 78:415–425.
Luo L, and Craik FI (2008) Aging and memory: a cognitive approach. Can J Psychiatry
53:346–353.
McDonough IM, Wong JT, Gallo DA (2013) Age-related differences in prefrontal cortex activ-
ity during retrieval monitoring: testing the compensation and dysfunction accounts. Cereb
Cortex 23:1049–1060.
McDonough IM, Cervantes SN, Gray SJ, Gallo DA (2014) Memory’s aging echo: age-related
decline in neural reactivation of perceptual details during recollection. Neuroimage
98:346–358
202
Maillet D. and Rajah MN (2014) Age-related differences in brain activity in the subsequent
memory paradigm: a meta-analysis. Neurosci Biobehav Rev 45:246–257.
Mattson JT, Wang TH, de Chastelaine M, Rugg MD (2014) Effects of age on negative subse-
quent memory effects associated with the encoding of item and item-context information.
Miller SL, Celone K, DePeau K, Diamond E, Dickerson BC, Rentz D, Pihlajamäki M, Sperling
RA (2008) Age-related memory impairment associated with loss of parietal deactivation
but preserved hippocampal activation. Proc Natl Acad Sci U S A. 105:2181–2186.
Morcom AM, Good CD, Frackowiak RSJ, Rugg MD (2003) Age effects on the neural cor-
relates of successful memory encoding. Brain 126:213–229.
Morcom AM, Li J, Rugg MD (2007) Age effects on the neural correlates of episodic retrieval:
Increased cortical recruitment with matched performance. Cereb Cortex 17:2491–2506
Nilson, L-G (2003) Memory function in normal aging. Acta Neurol Scand 107 (suppl 179):7–13.
tion in aging. Proc Natl Acad Sci USA 107:22682–22686.
Old SR, Naveh-Benjamin M (2008) Differential effects of age on item and associative measures
of memory: a meta-analysis. Psychol Aging 23:104–118.
Park DC, Lautenschlager G, Hedden T, Davidson NS, Smith AD, Smith PK. Models of visuo-
spatial and verbal memory across the adult life span (2002) Psychol Aging 17:299–320.
Persson J, Pudas S, Lind J, Kauppi K, Nilsson LG, Nyberg L (2012) Longitudinal structure-
function correlates in elderly reveal MTL dysfunction with cognitive decline. Cereb
Cortex 10:2297–3304.
Persson J, Pudas S, Nilsson LG, Nyberg L (2014) Longitudinal assessment of default-mode
brain function in aging. Neurobiol Aging 35:2107–2017.
Pudas S, Persson J, Josefsson M, de Luna X, Nilsson LG, Nyberg L (2013) Brain character-
istics of individuals resisting age-related cognitive decline over two decades. J Neurosci
33:8668–8677.
Pudas S, Persson J, Nilsson LG, Nyberg L (2014) Midlife memory ability accounts for brain
activity differences in healthy aging. Neurobiol Aging 35:2495–2503.
Raz N, and Lindenburger U (2011) Only time will tell: cross-sectional studies offer no solution to
the age-brain-cognition triangle: comment on Salthouse (2011). Psychol Bull 37:790–795.
Rönnlund M, Nyberg L, Bäckman L, Nilsson LG (2005) Stability, growth, and decline in adult
life span development of declarative memory: cross-sectional and longitudinal data from a
population-based study. Psychol Aging 20: 3–18.
Rönnlund M, Nilsson LG (2009) Flynn effects on sub-factors of episodic and semantic mem-
ory: parallel gains over time and the same set of determining factors. Neuropsychologia
47:2174–2180.
Rugg MD (2004) Retrieval processing in human memory: Electrophysiological and fMRI evi-
dence. In M.S. Gazzaniga (Ed.) The Cognitive Neurosciences 3rd Ed., MIT press.
Rugg MD, Morcom AM (2004) The relationship between brain activity, cognitive performance
and aging: The case of memory. In Cabeza, R., Nyberg, L., and Park, D. (Eds.) Cognitive
Neuroscience of Aging: Linking Cognitive and Cerebral Aging. Oxford University Press.
Rugg MD, Johnson JD, Uncapher MR (2015) Encoding and retrieval in episodic mem-
ory: Insights from fMRI. In Duarte, A., Barense, M., Addis, D.R. (Eds.), Handbook on the
Cognitive Neuroscience of Memory. Wiley-Blackwell.
Salthouse TA (1996) The processing-speed theory of adult age differences in cognition. Psychol
Rev 103:403–428.
203
Salthouse TA (2010) Major Issues in Cognitive Aging. Oxford University Press.

Salthouse TA (2014) Why are there different age relations in cross-sectional and longitudinal
comparisons of cognitive functioning? Psychol Sci 23:252–256.
Schnack HG, van Haren NE, Brouwer RM, Evans A, Durston S, Boomsma DI, Kahn RS,
Hulshoff Pol HE (2015) Changes in thickness and surface area of the human cortex and
their relationship with intelligence. Cereb Cortex. 25: 1608–1617.
Sperling RA, Laviolette PS, O’Keefe K, O’Brien J, Rentz DM, Pihlajamaki M, Marshall G,
Trahan LH, Stuebing KK, Fletcher JM, Hiscock M (2014) The Flynn effect: a meta-analysis.
Psychol Bull 140:1332–1360.
Tsvetanov KA, Henson RN, Tyler LK, Davis SW, Shafto MA, Taylor JR, Williams N, Cam-
Can, Rowe JB (2015) The effect of ageing on fMRI: Correction for the confounding
effects of vascular reactivity evaluated by joint fMRI and MEG in 335 adults. Hum Brain
Mapp 36:2248–2269.
Tisserand DJ, McIntosh AR, van der Veen FM, Backes WH, Jolles J (2005) Age-related reorga-
nization of encoding networks directly influences subsequent recognition memory. Brain
Res Cogn Brain Res 25:8–18.
Tulving E (1985) Memory and consciousness. Can Psych 40:1–12.
Van Petten C. Relationship between hippocampal volume and memory ability in healthy
individuals across the lifespan: review and meta- analysis (2004) Neuropsychologia
42:1394–1413
Wang TH, de Chastelaine M, Minton B, Rugg MD (2012) Effects of age on the neural correlates
of familiarity as indexed by ERPs. J Cog Neurosci 24:1055–1068.
Wang TH, Johnson JD, de Chastelaine M, Donley BE, Rugg MD (2016) The effects of age
on the neural correlates of recollection success, recollection-related cortical reinstatement
and post-retrieval monitoring. Cereb Cortex 26:1698–1714.
Yassa M, Stark CEL (2011) Pattern separation in the hippocampus. Trends Neurosci
34:515–525.
204
205
II
COGNITIVE PROCESSES
206
207
Selective Attention and Inhibitory

Control in the Aging Brain
Theodore P. Zanto
Adam Gazzaley
T he term attention is multifaceted and often refers to a set of cognitive

processes that transcends a single definition or overarching theory
(Parasuraman, 1998). The goal of this chapter is to review aspects of selective atten-
tion and its neural substrates in the context of changes that occur with normal aging.
Selective attention refers to goal-directed focus on task-relevant information while
ignoring other irrelevant information. This chapter is subdivided according to the type
of information that is selectively attended/ignored: spatial location, features, objects,
and internally-based items. Although there are several critical reviews on attention
and aging within the last few years (Madden, 2007; Kramer and Madden, 2008; Drag
and Bieliauskas, 2010; Zanto and Gazzaley, 2014), this prior work often focused on
behavioral paradigms with less discussion of the underlying neural mechanisms. As
the literature on the cognitive neuroscience of aging continually grows, this chap-
ter will focus almost exclusively on addressing age-related changes in the brain that
relate to selective attention. It should be noted that the majority of this research has
been conducted in the visual domain and cited research herein will be assumed to be
in this modality unless otherwise stated.
To address the cognitive neuroscience of aging, studies reported throughout this
chapter typically use either functional magnetic resonance imaging (fMRI) or elec-
troencephalography (EEG) due to their high spatial (millimeter) or temporal (milli-
second) resolution, respectively. When results from fMRI research are discussed, a
distinction will be made between activity and functional connectivity. fMRI activity
will refer to blood oxygen level dependent (BOLD) measures in select regions of gray
matter, while functional connectivity will refer to how BOLD activity fluctuates (or
207
208
208â•… Cognitive Processes
correlates) between two regions of gray matter. Results from EEG research reported
below fall into one of two categories, event related potentials (ERPs; most common)
and spectral measures (less common). ERP results are typically described by the
polarity of the waveform (i.e., “P” for positive and “N” for negative) and the latency
of the peak. ERP components discussed include (in temporal order) the P1, N1, N2,
and P3, where the P1 occurs closest to stimulus onset and reflects early neural pro-
cesses such as those related to perception, whereas the P3 is further from stimulus
onset and reflects late neural processes related to higher cognition such as updating
memory traces. Importantly, the amplitude and/â•‰or latency of these ERP components
are altered by selective attention and serve as a means to assess attentional modulation
(biasing) of neural activity in aging based on task goals (i.e., attend or ignore stimuli).
When EEG spectral measures are discussed below, it is referring to a decomposition
of the EEG data into its component frequency bands, such as theta (4–â•‰7 Hz) and alpha
(8–â•‰12 Hz), which are also altered by selective attention and serve as neural markers
to assess age-â•‰related changes in selective attention and inhibition. Regardless of the
methodology, differences in neural activity between attended and ignored stimuli will
be referred to as attentional modulation, and will be used throughout this chapter as a
metric by which selective attention and inhibitory control is affected by aging.
Models of Cognitive Aging
Although the chapter is organized by the type of information that is selectively

attended/â•‰ignored (spatial location, features, objects, internally-â•‰based items), general
themes of age-â•‰related changes emerge from the literature. Therefore, before discuss-
ing details of the research on selective attention, inhibition, and aging, it is worthwhile
to orient the reader with these themes and a few of the theories/â•‰models that have been
formed in response to some of this research. Notably, older adults may exhibit task
performance that is comparable to younger adults but utilize different neural mech-
anisms. To account for this, the posterior-â•‰to-â•‰anterior shift in aging (PASA) model
(Davis et al., 2008) suggests older adults exhibit decreased occipital neural activity
and increased prefrontal activity as a means to compensate for declines in perceptual
processes that occur in occipital regions. Yet, when both younger and older adults uti-
lize prefrontal cortex (PFC) during selective attention tasks, older adults may exhibit
comparable performance by recruiting contralateral PFC regions as described by the
hemispheric asymmetry reduction in aging (HAROLD) model (Cabeza, 2002).
Despite these compensatory mechanisms in aging, there are many scenarios as
described below where older adults exhibit lower task performance compared to
younger adults. For example, tasks that incorporate distractors often exhibit age-â•‰
related performance declines and has led to the hypothesis that older adults have a
deficit in inhibiting irrelevant information (Hasher and Zacks, 1988; Hasher et al.,
2007), which may arise due to changes in PFC structure and function (Dempster,
1992; West, 1996). In the absence of distraction, age-â•‰related performance declines
may still be observed and has been associated with slowed processing of information
in older adults (Salthouse, 1996). Incorporating these models, it has been proposed that
declines in PFC structure and function result in deficient top-â•‰down attention-â•‰related
â•‡ 209
Selective Attention and Inhibitory Control in the AgingÂ€Brainâ•… 209
abilities to modulate neural activity in sensory cortex (Gazzaley, 2012), which link
theories of inhibition and processing speed (Gazzaley et al., 2008), and helps explain
widespread impairment associated with cognitive aging.
Interestingly, age-â•‰related declines in performance become most pronounced as
task difficulty increases and has given rise to models of cognitive aging that suggest
older adults may utilize compensatory neural mechanisms until a cognitive resource
limit is reached, as described by the cognitive reserve hypothesis (Stern, 2002) and
the compensation-â•‰related utilization of neural circuits hypothesis (CRUNCH)(Reuter-â•‰
Lorenz and Lustig, 2005; Reuter-â•‰Lorenz and Cappell, 2008). Similarly, the scaffold-
ing theory of aging and cognition (STAC) suggests increased PFC activity with age
serves as compensatory scaffolding in response to increased challenges posed by
declining neural structures and function (Park and Reuter-â•‰Lorenz, 2009). Although
scaffolding is thought to be more prevalent in older adults due to increased challenge
with lower task demands, scaffolding in older adults is less plastic and may be less
efficient than in younger adults, indicating why age-â•‰related performance declines are
most pronounced with increased task demands.
Spatial Selective Attention
Spatial selective attention refers to allocating attention to one location while ignoring
another. This section will focus on the three most common paradigms used to study
the cognitive neuroscience of spatial selective attention in aging: stimulus detection,
target detection, and change detection.
Stimulus Detection
To study spatial attention, stimulus detection paradigms often use a cue to indi-
cate where in space a probe will subsequently appear and participants may use this
cued, predictive information to guide where attention should be covertly allocated.
Importantly, the response time difference between valid and invalid (or neutral) cues
provide a measure of the spatial attention orienting response and is thought to be
largely preserved in aging (Nissen and Corkin, 1985; Hartley et al., 1990; Gottlob
and Madden, 1998). Despite comparable cue-â•‰based benefits between younger and
older adults, the neural processes involved in sensory detection paradigms often differ
between these age groups. Early visual event related potential (ERP) components (e.g.,
P1 that occurs 100 ms post stimulus onset) arise in sensory cortex in response to visual
probe stimuli and are delayed in older adults, suggesting a decline in early sensory
(bottom-â•‰up) processes (Yamaguchi et al., 1995; Curran et al., 2001; Lorenzo-â•‰Lopez et
al., 2002). Importantly, these early sensory processes are modulated by attention (i.e.,
differentially responsive to attended and ignored stimuli) and the magnitude of these
neural responses decline with repeated sound stimuli in younger, but not older, adults
suggesting an age-â•‰related decline in attentional (top-â•‰down inhibitory) modulation of
early sensory processing (Leung et al., 2013). Although age-â•‰related differences in early
top-â•‰down and bottom-â•‰up sensory processing are observed, older adults may exhibit
210
increased attentional modulation of neural activity at later processing stages reflect-

ing increased stimulus selection (occipital N1, ~170 ms post-stimulus) and updating
memory traces (frontal P3, ~300 ms post stimulus onset) (Talsma et al., 2006). Based
on these results, it may be speculated that older adults increase attentional modulation
to spatial locations during later processing stages in order to compensate for declines
during early sensory processing, thereby yielding detection performance comparable
to younger adults. However, a direct relationship between these neural differences and
performance were not observed and therefore may not be considered compensatory.
Nonetheless, compensatory neural activity in aging has been demonstrated using many
other paradigms as detailed below, and so these results remain suggestive. Together,
attentional orienting to a location for stimulus detection appears largely unaffected in
aging, although different functional mechanisms may be utilized.
Target Detection
Spatial target detection paradigms typically require participants to detect a target in
the presence of distracting non-â•‰targets and commonly involves search paradigms.
Target detection differs from stimulus detection not only in the presence of distractors,
but that the to-â•‰be-â•‰identified target is disclosed to the participant prior to beginning the
experiment, and as such, is held in long-â•‰term memory for the duration of the exper-
iment. Consistent with previous results from stimulus detection paradigms, older
adults are just as effective as younger adults in using pre-â•‰cued distractor locations to
improve target detection (search) performance, but utilize different neural networks
as indicated by differential fMRI BOLD activity and functional connectivity (Allen
and Payne, 2012). Notably, older adults’ performance improved in participants who
exhibited increased activity in frontal regions (frontal eye fields), whereas younger
adults’ performance correlated with posterior visual cortex activity, suggesting that
older adults may utilize frontal, attention related, regions to compensate for declines
in sensory cortex. Indeed, age-â•‰related declines in attentional modulation of sensory
cortical activity during spatial target detection are often accompanied by increased
prefrontal cortical (PFC) activity as indexed by increased regional cerebral blood flow
(rCBF) measured with positron emission tomography (PET)(Madden and Hoffman,
1997), increased BOLD activity (Madden et al., 2007; Geerligs et al., 2014), and
an increased frontal P3 amplitude (Li et al., 2013). Importantly, increases in PFC
activity correlate with improved target detection performance, indicating this activity
serves as a compensatory mechanism in aging (Madden et al., 1997; Madden et al.,
2007) to overcome disinhibition and distraction in this population (Li et al., 2013).
Although older adults may compensate for declines in bottom-â•‰up sensory processes
with enhanced PFC activity (Li et al., 2013), the PFC activity in older adults arises in
a nonselective manner for different search arrays, indicating less specialized neural
mechanisms for task execution or perhaps indicative of a more effortful search strat-
egy that relies on attentional control processes mediated by the PFC (Lorenzo-â•‰Lopez
et al., 2008b). Nonetheless, the differential networks utilized by older adults to retain
target detection performance support the posterior-â•‰to-â•‰anterior shift in aging (PASA)
model (Davis et al., 2008).
â•‡ 211
When the target location is not cued, older adults exhibit performance declines (i.e.,
slower response time and lower accuracy) concomitant with deficient neural activity
in occipital-â•‰temporal cortex related to an impaired allocation of visuospatial attention
(i.e., a slower and attenuated posterior N2pc to targets) (Lorenzo-â•‰Lopez et al., 2008a;
Lorenzo-â•‰Lopez et al., 2011). These results support research suggesting older adults
require more time to orient attention to the target (Amenedo et al., 2012), which may
not be fully attributed to a generalized slowing of neural processing (Lorenzo-â•‰Lopez
et al., 2008a). Moreover, these results provides a more specific cognitive function,
attentional orientation, for theories that suggest age-â•‰related declines stem from slowed
processing speed (e.g., Salthouse, 1996).
Whereas spatial target detection tasks consistently reveal different neural activity pro-
files in aging, age-â•‰related performance declines are not consistently observed, which in
addition to compensatory mechanisms, may reflect influences of task difficulty. Under
conditions of increased task difficulty due to a high perceptual load (i.e., distractors with
similar target features), older adults’ target detection performance declines (slower detec-
tion and lower accuracy) and exhibit neural activity that suggests increased demands of
attentional selection in aging (i.e., larger and slower occipital N1)(Wang et al., 2012).
These age-â•‰related declines in target detection due to increased task difficulty have been
associated with anatomical changes such as gray matter volume shrinkage in frontopari-
etal regions known to be involved in attentional control (Muller-â•‰Oehring et al., 2013) as
well as declines in white matter tract integrity connecting frontoparietal attention net-
works (superior and inferior longitudinal fasciculi)(Bennett et al., 2012). Together, these
results support models of cognitive aging that suggest increased task difficulty results in
age-â•‰related performance declines due to limited resources that stem from functional and
anatomical changes throughout the lifespan (Stern, 2002; Reuter-â•‰Lorenz and Lustig,
2005; Reuter-â•‰Lorenz and Cappell, 2008; Park and Reuter-â•‰Lorenz, 2009).
Although it is plausible that age-â•‰related declines in target detection performance
may be observed only when task difficulty exceeds limited resources, another possi-
bility may stem from motivational differences between age groups. It was recently
shown that monetary reward speeds visual target detection in both younger and older
adults, but more so in younger adults, who also exhibit decreased trialwise variability
(Stormer et al., 2014). These results suggest that when attentional or perceptual lim-
itations are not exceeded by task difficulty, older adults may exhibit declines in target
detection performance when younger adults bias attention to motivationally salient
events. This implies that previous studies reporting age-â•‰equivalence may observe age
differences if younger adults are properly motivated. Overall, results from visuospa-
tial target detection paradigms indicate that older adults recruit additional PFC and
some parietal regions to compensate for declines in allocating visuospatial attention
to bias sensory cortical activity, and that age-â•‰related performance declines are most
pronounced when task difficulty increases.
Change Detection
Change detection paradigms generally incorporate working memory processes
to address whether a previously presented object or feature has changed its spatial
212
212 Cognitive Processes
location or orientation. As such, change detection tasks may also be referred to as

location matching or delayed match to sample paradigms. Similar to visual target
detection tasks, age-related performance declines in visual change detection become
more pronounced as task difficulty increases (Piefke et al., 2012; Sander et al., 2012).
With increasing task difficulty, older adults exhibit less lateralized parietal-occipital
alpha (8–12 Hz) activity in anticipation of the probe (Sander et al., 2012). Because
increased alpha activity has been associated with increased inhibitory neural activity,
these results suggest older adults have difficulty sustaining inhibition of the unattended
hemisphere at high loads. However, other research indicates that age-related differ-
ences in inhibiting parietal-occipital neural activity may be the product of different
mechanisms used by older adults to inhibit distraction in order to retain performance
ability (Vaden et al., 2012). This is in line with previous research indicating that visual
change detection performance is associated with different neural networks across age
groups, such that younger adults rely more on posterior sensory regions, while older
adults engage PFC (Solbakk et al., 2008). The idea that older adults may utilize dif-
ferent neural mechanisms (e.g., PFC rather than sensory cortex) to inhibit distracting
information is certainly credible given these studies agree with research from sensory
and target detection paradigms in supporting the PASA model and indicating that age-
related declines in performance are most pronounced during increased task difficulty.
When change detection task difficulty is decreased (e.g., low working memory load),
younger adults exhibit left lateralized PFC BOLD activity, whereas older adults exhibit
bilateral PFC activity (Piefke et al., 2012), in line with the hemispheric asymmetry reduc-
tion in aging (HAROLD) model (Cabeza, 2002). Indeed, increased bilateral PFC activity
in older adults during working memory encoding is associated with improved spatial
change detection performance (Macpherson et al., 2014), while high performing older
adults exhibit more “youth-like” load-dependent modulation of BOLD activity across
frontoparietal regions (Nagel et al., 2009). However, when task difficulty increases, older
adults exhibit right PFC activity (contralateral to younger adults) as well as parieto-
occipital reorganization (Piefke et al., 2012) that may reflect failed compensatory mech-
anisms, as age-related performance declines persist. These results are consistent with
more recent findings that suggest dorsolateral prefrontal cortex (DLPFC) serves to com-
pensate for older adults’ diminishing performance in spatial change detection tasks until
a resource ceiling is reached (Toepper et al., 2014). Together, these results corroborate
research from spatial target detection tasks in supporting models of cognitive aging that
indicate compensatory neural activity may be recruited in aging to retain performance
but fail to help once capacity limitations are exceeded (Stern, 2002; Reuter-Lorenz and
Lustig, 2005; Reuter-Lorenz and Cappell, 2008; Park and Reuter-Lorenz, 2009).
Due to the multiple cognitive functions engaged during change detection tasks, and
the multiple frontoparietal regions affected in aging, it is difficult to attribute these
age-related differences to one cognitive operation. Piefke et al. (2012) observed com-
parable BOLD activity between age groups in the superior parietal cortex, which is
known to be involved in visuospatial attention. This was used to attribute age-related
alterations in frontoparietal regions (i.e., DLPFC, precuneous, inferior parietal lobule)
to changes in executive capacity, mental imagery, and visual processing, but not visu-
ospatial attention. Yet, this (and most) fMRI studies of spatial change detection cannot
differentiate between distinct task components (i.e., encoding, maintenance, probe)
â•‡ 213
due to the short trial duration. Thus, age-â•‰related differences observed in change detec-
tion fMRI studies likely reflect alterations in multiple cognitive domains that include
visuospatial attention, cognitive control, response preparation, and working memory
maintenance (Nagel et al., 2009). Fortunately, this has somewhat been addressed by
using EEG to show age-â•‰related declines in change detection performance is associ-
ated with differences in attention modulating working memory encoding processes
(Stormer et al., 2013), supporting the notion that selective attention and working
memory processes are overlapping (Cowan, 1995; Gazzaley and Nobre, 2012).
Specifically, Stormer et al. (2013) showed attentional modulation of neural activity
in sensory cortical regions occurred during earlier perceptual, stimulus selection, pro-
cessing stages (i.e., N1) in older adults compared to younger adults who modulated
neural activity during later visual processing stages that are associated with visuospa-
tial attention (i.e., N2pc). Moreover, attentional modulation of neural activity in older
adults was nonspecific (i.e., did not differentiate between the presence or absence of
distractors), indicating this was a general boost in perceptual processing and not an
attempt to inhibit neural activity in response to irrelevant stimuli. These results pro-
vide supporting evidence that visuospatial attentional processes are affected in aging
and operate on different cognitive stages in a nonâ•‰specific manner, likely due to differ-
ent neural regions (e.g., PFC) utilized by older adults.
Whereas the nonâ•‰specific attentional modulation of neural activity in aging hints
at the possibility of dedifferentiation (i.e., reduced functional specialization), it does
not provide strong evidence. During a change detection task with objects and spatial
locations, both younger and older adults exhibit occipitotemporal activation during
object change detection and superior parietal activation during spatial location change
detection (Grady et al., 1992; Grady et al., 1994). However, older adults exhibit more
activity in these regions during the opposite tasks, indicating less functional special-
ization (dedifferentiation) of the dorsal and ventral visual pathways. Furthermore,
older adults utilize more attention-â•‰based networks (frontoparietal areas) during spa-
tial location change detection, indicating less efficient sensory cortical processing
(Grady et al., 1994). Thus, age-â•‰related dedifferentiation in sensory processing regions
may be interpreted more generally as bottom-â•‰up sensory processing declines and sup-
port models (e.g., PASA) that suggest the increased use of top-â•‰down attention related
regions (e.g., PFC) in older adults serve to compensate for these declines.
Overall, current cognitive neuroscience research on spatial attention in aging con-
verge to show that older adults tend to recruit additional PFC regions to compensate for
declines in sensory cortical processing in order to retain performance levels. However,
as task difficulty increases, age-â•‰related performance declines in spatial attention tasks
become apparent. It remains unclear whether this decline in performance represents a
deficit in top-â•‰down attentional mechanisms or a limitation in compensatory top-â•‰down
processes to overcome bottom-â•‰up deficits.
Feature Selective Attention
Feature selective attention refers to attending and ignoring elementary parts of a stim-
ulus, such as a color or shape. Feature selective attention paradigms typically fall
214
within one of four main categories: stimulus detection, target detection, change detec-
tion, and discrimination.
Stimulus Detection
In contrast to spatial stimulus detection, there are relatively few feature-â•‰based stimu-
lus detection studies that address the cognitive neuroscience of aging. Feature-â•‰based
stimulus detection paradigms are very basic experimental paradigms that may or may
not even require participants to respond when a stimulus is presented. For example,
passively viewing a colored checkerboard resulted in increased lateral occipital activ-
ity, which was interpreted as a signature of age-â•‰based declines in top-â•‰down inhibitory
control of bottom-â•‰up sensory cortical activity (Tse et al., 2010). Similar age-â•‰based
declines in inhibitory control have been observed in the auditory domain, such that
during dichotic listening, older adults are more negatively affected by increased dis-
tractor saliency as indexed by decreased attentional modulation of frontoparietal neu-
ral activity (Passow et al., 2014). Thus, age-â•‰related declines in attentional modulation
of neural activity during feature sensory detection appear independent of sensory
modality. Together, this supports research in spatial attention to suggest age-â•‰related
declines in feature-â•‰based attention may also stem from deficient attentional modu-
lation of sensory cortex, or more specifically, declines in inhibiting sensory cortical
processing of irrelevant information.
Although most research has identified age-â•‰related changes in attentional mod-
ulation of neural activity during stimulus presentation, recent evidence indicates
age-â•‰related deficits in attentional biasing of sensory cortex may stem from impair-
ments in expectation processes formed prior to stimulus onset (Zanto et al., 2013).
Specifically, when instructed to respond to the onset of motion, older adults’ ability
to detect any direction of motion is negatively impacted when older adults expected
horizontal, but not vertical, motion concomitant with less parietal-â•‰occipital neu-
ral activity. These results were interpreted as age-â•‰related declines in attentional
biasing of neural networks underlying horizontal motion processing, which rely
more heavily on interhemispheric communication compared to vertical motion
processes. As white matter tracts are known to degrade with age (Rabbitt et al.,
2007; Turken et al., 2008), this suggests an interesting possibility that attentional
decline in age may be most pronounced when attentional operations utilize longer
(or more distributed) anatomical networks. Furthermore, this would fall in line with
the scaffolding theory of cognitive aging (STAC) that indicates neural networks
in aging are less efficient (Park and Reuter-â•‰Lorenz, 2009). Although additional
research will be required to test this hypothesis, it could help explain age-â•‰related
declines in sensory cortical modulation from prefrontal cortex, as well as why
age-â•‰based changes are most prominent during increased task complexity. Despite
limited studies on feature stimuli detection, results indicate age-â•‰based attentional
declines are related to difficulties in inhibiting sensory cortical activity, which may
arise in anticipation of stimulus onset.
â•‡ 215
Target Detection
Target detection paradigms inform participants prior to the experiment of a specific to-â•‰
be-â•‰detected target stimulus that is presented along with distracting irrelevant stimuli.
In line with studies addressing spatial selective attention and the PASA model, older
adults may yield comparable performance to younger adults, but exhibit a posterior to
anterior shift in neural activity during feature target detection (O’connell et al., 2012;
Alperin et al., 2014; Zhuravleva et al., 2014). Although it is hypothesized that older
adults recruit frontal regions to compensate for declines in bottom-â•‰up sensory process-
ing, there is evidence that older adults may exhibit comparable detection performance
as younger adults along with equivalent neural measures of bottom-â•‰up stimulus pro-
cessing in visual cortex (Quigley et al., 2010; Quigley and Muller, 2014). Importantly,
age-â•‰related performance declines are observed when distracting stimuli are presented,
and this occurs concomitant with declines in attentional modulation of sensory cortical
activity. Therefore, these results provide evidence that top-â•‰down attentional modulation
of sensory cortex declines in age, which may not be attributed to differences in bottom-â•‰
up visual processing. Furthermore, this data supports models of cognitive aging that
suggest performance declines result from deficient top-â•‰down signals from PFC that fail
to bias sensory cortex activity in the service of task goals (Gazzaley, 2012).
Because measures of attentional modulation of neural activity reflect the difference
between attended and ignored stimuli, it is difficult to attribute age-â•‰related changes
in neural modulation to alterations in enhancing or suppressing activity to relevant or
irrelevant stimuli, respectively. By using a neutral baseline where participants pas-
sively view stimuli, older adults exhibit declines in suppressing PFC activity to irrele-
vant features (i.e., comparable activity between passively viewed and ignored stimuli),
but exhibit increased enhancement of PFC activity to attended features (i.e., attended
> passively viewed stimuli), which indicates older adults rely more heavily on atten-
tional resources (Haring et al., 2013). It should be noted that these differences in
enhancement and suppression of PFC activity occurred around 200 ms post stimulus
onset and are thought to influence later processing stages (between 300–â•‰500 ms post
stimulus onset), such that older adults allocate excessive attentional resources to both
relevant and irrelevant stimuli, as indexed by increased PFC (Cona et al., 2013) and
increased parietal-â•‰occipital (Alperin et al., 2013; Daffner et al., 2014) activity, respec-
tively. Interestingly, older adults exhibit less neural processing in PFC and sensory
cortex to stimuli following a target, suggesting an age-â•‰related decline in shifting atten-
tion (disengaging) from targets (Cona et al., 2013), which may be the consequence
of allocating excessive attentional resources to target stimuli. Together, these results
support models that suggest aging is associated with declines in inhibition (Hasher
and Zacks, 1988; Hasher et al., 2007) due to changes in PFC structure and function
(Dempster, 1992; West, 1996).
Overall, research using feature target detection paradigms agree with spatial
target detection paradigms in that older adults shift neural processes from poste-
rior to anterior regions. Despite this increased PFC activity, older adults exhibit
declines in attentional modulation of sensory cortex activity that is thought to arise
216
from PFC regions, and that such age-â•‰related declines in attentional modulation
may occur even in the absence of bottom-â•‰up sensory deficits. Due to deficient
attentional control in aging, older adults exhibit excessive attentional allocation to
both relevant and irrelevant stimuli, which may contribute to difficulty disengaging
from a stimulus.
Change Detection
During feature change detection paradigms, participants invoke working memory
processes to identify whether a cue stimulus matches a probe after a delay period.
Similar to the previously described feature and spatial attention paradigms, older
adults exhibit increased PFC activity during an auditory feature change detection task,
concomitant with performance declines (Grady et al., 2008). In addition to differential
PFC activity, older adults exhibit a deficit in suppressing neural activity in primary
auditory cortex to distractor tones that occur during long (> 9 sec), but not short (< 9
sec), delays and this decline was related to deficient performance measures of frontal
lobe functioning (Wisconsin card sorting test; Chao and Knight, 1997). Because this
aging effect was not observed during short delays, this suggests the age-â•‰based sup-
pression deficit is due to sustained attention processes from the prefrontal cortex, and
not due to bottom-â•‰up sensory deficits.
The idea that bottom-â•‰up declines in aging cannot fully account for attentional dif-
ferences in aging has received additional support in the visual domain. Once per-
ceptual differences and motoric slowing have been equated across age groups, older
adults exhibit working memory performance declines along with decreases in atten-
tional modulation of visual cortex (Zanto et al., 2010a). Such changes in attentional
modulation of visual cortex have been attributed to declines in functional interac-
tions with the PFC (Zanto et al., 2010b; Zanto et al., 2011a), supporting models that
posit age-â•‰based declines in attentional modulation of neural activity may arise from
changes in communication between PFC and sensory cortex (Gazzaley, 2012).
Although attentional declines are observed in aging that are distinct from bottom-â•‰
up sensory changes, perceptual training improves working memory for motion during
a visual change detection task along with reduced visual cortex activity related to
stimulus selection (i.e., N1; Berry et al., 2010). This reduced activity in older adults
post-â•‰training may reflect less neurons firing in response to sharpened perceptual tun-
ing. Given that these measures of visual cortex activity are under top-â•‰down influences
(Berry et al., 2009; Zanto et al., 2013), and earlier perceptual processes did not exhibit
such gains (i.e., P1; Berry et al., 2010), this suggests that less attentional control may
have been required. Yet, additional research is necessary to determine whether per-
ceptual training will diminish age-â•‰related declines in feature change detection due to
increased task difficulty.
In line with spatial attention paradigms, older adults may functionally reorganize
neural network activity to accommodate increasing task demands (Schulte et al.,
2011). For example, older adults exhibit performance declines during feature change
detection that become more pronounced with task difficulty (Ansado et al., 2012;
Ansado et al., 2013). Neurally, older adults exhibit a posterior to anterior shift in
activity, in line with the PASA model, and with increasing task demands, additional
â•‡ 217
regions are recruited in anterior cingulate cortex as well as superior parietal lobe,
indicating an increased engagement with attentional networks.
Together, studies on feature change detection suggest that attentional declines in
aging are distinct from bottom-â•‰up sensory changes. Furthermore, there is evidence
that functional networks are plastic throughout the lifespan, such that frontoparietal
networks may be recruited as needed in response to task demands and that cogni-
tive training may enhance change detection performance by lowering the demands on
attentional processes.
Discrimination
Discrimination paradigms instruct participants to make a categorical decision and
response about a particular stimulus. One of the most common feature discrimination
paradigms is the Stroop task. Participants are required to report the color of a word
and not the word itself, which is the name of a different color. During this feature
discrimination task, older adults exhibit fewer successful inhibitions of the irrelevant
word along with increased activity and increased extent of activity in frontal and pari-
etal regions (Nielson et al., 2002; Langenecker et al., 2004; Nielson et al., 2004), sug-
gesting an increased reliance on attentional networks. Furthermore, these age-â•‰related
declines in interference have been associated with declines in white matter integrity in
anterior corpus callosum (Sullivan et al., 2006) as well as within parietal and occipital
regions (Kennedy and Raz, 2009). Controlling for general slowing, age-â•‰related Stroop
interference effects remain and are associated with declines in white matter integrity
in the genu of corpus callosum, anterior corona radiata, and anterior limb of capsula
interna (Wolf et al., 2014). These pathways are associated with frontal regions by con-
necting DLPFC/â•‰anterior cingulate cortex with other frontal and subcortical regions.
Together, these results support models that suggest aging is associated with declines
in inhibition (Hasher and Zacks, 1988; Hasher et al., 2007) due to changes in PFC
structure and function (Dempster, 1992; West, 1996).
Although older adults exhibit increased BOLD activity during Stroop tasks, this
occurs without changing cerebral blood flow, indicating less oxygen metabolism
response and may be interpreted as reduced neural activity in older adults (Mohtasib
et al., 2012). It should be noted that some research shows age-â•‰equivalent performance
during a Stroop task (Milham et al., 2002). Despite comparable performance, older
adults exhibit declines in activity modulation in DLPFC and posterior parietal cortex
between congruent and incongruent trials, suggesting declines in attentional control.
Similar to spatial attention and other feature attention paradigms, age-â•‰equivalent per-
formance may be related to task difficulty such that age-â•‰related performance declines
in inhibition during a Stroop task become more pronounced with increasing task
difficulty (Prakash et al., 2009). Importantly, frontoparietal activity in older adults
during low-â•‰demand conditions appear similar to younger adults during high demand
conditions, which is thought to reflect an age-â•‰related limitation in flexibly recruiting
additional attentional network regions in response to increasing cognitive demands.
Results from non-â•‰Stroop feature discrimination paradigms support other research
from feature and spatial attention paradigms. Specifically, older adults are nega-
tively influenced by distraction, requiring more time to disengage from distraction
218
(Cashdollar et al., 2013), recruit frontoparietal attentional network regions to retain

feature discrimination performance (Huang et al., 2012) and these recruited regions
increase in activity with increasing task demands (Hedden et al., 2012). However,
older adults who do not exhibit increased activity with increasing task demands dis-
play lower task performance, increased white matter hyperintensities, as well as lower
neuropsychological scores in executive function and processing speed (Hedden et al.,
2012). These results indicate that declines in neural networks underlying attentional
control is heterogeneous in the aged population and that those with white matter
hyperintensities may not be able to utilize increased frontoparietal activity as a com-
pensatory mechanism.
Whereas the majority of aging research compares young adults (younger than
35 years) to older adults (older than 65 years), middle aged adults have also been
shown to exhibit increased PFC activity as a potential compensatory mechanism
during visual (Zysset et al., 2007) and tactile (Reuter et al., 2013) feature discrimina-
tion. Interestingly, both younger and older adults with low attentional control measures
(operation span score) display a posterior to anterior shift in neural activity during
feature discrimination (Peltz et al., 2011). As this shift in processing was observed in
younger adults with low attentional control, this suggests that individual differences
in cognitive aging may begin early in life (Peltz et al., 2011). Together, age-â•‰related
declines in feature discrimination stem from functional and structural changes in the
PFC underlying attentional control, which is most pronounced with increasing task
difficulty and these declines may begin early in adulthood.
Overall, studies of feature-â•‰based selective attention corroborate research from
spatial-â•‰based selective attention to show that performance from older adults is more
negatively influenced by distraction and task difficulty, with a coincident increase in
PFC activity. Yet, due to declines in PFC structure and function, older adults exhibit
deficient top-â•‰down attentional control, which results in less modulation of activity in
sensory cortex based on task goals. Thus, older adults may allocate excessive atten-
tion to both relevant and irrelevant stimuli, resulting in increased interference from
irrelevant information and contributes to a difficulty disengaging from a stimulus.
Importantly, there is mounting evidence that cognitive declines associated with aging
may become manifest in early and mid-â•‰adulthood.
Object Selective Attention
Object selective attention refers to attending and ignoring stimuli from a holistic per-
spective, such as everyday items including faces or tools, and even whole scenes,
which may differ within object category based on component features. Object selec-
tive attention paradigms typically fall within one of three main categories: stimulus
detection, change detection, and discrimination.
Stimulus Detection
In line with feature-â•‰based sensory detection, object-â•‰based sensory detection tasks
show age-â•‰related declines in utilizing predictive information to enhance performance,
â•‡ 219
as indicated neurally by decreased anticipatory activity in visual cortex (Zanto et al.,

2011b). Interestingly, in addition to the stimulus detection task, this expectation defi-
cit in aging was replicated twice in the same participants using two distinct object
discrimination tasks (see Discrimination section below). These results support mod-
els indicating age-â•‰related declines in cognition may stem from deficient expectation
mechanisms (Zanto et al., 2011b) or deficient goal maintenance abilities that result in
a proactive to reactive shift in cognitive control (Paxton et al., 2008; Dew et al., 2012).
Change Detection
Both younger and older adults exhibit occipitotemporal activation during a visual
change detection task, but older adults exhibit more activity in visuospatial regions
(superior parietal cortex), indicating less functional specialization of the dorsal and
ventral visual pathways (dedifferentiation; Grady et al., 1992; Grady et al., 1994). Yet,
older adults may utilize frontoparietal attentional networks to compensate for visual
processing deficiencies (Lee et al., 2011; Burianova et al., 2013). Despite the poten-
tial for compensatory mechanisms, older adults exhibit a disproportionate decline in
working memory performance in the presence of irrelevant stimuli during the encod-
ing period (Gazzaley et al., 2005; Gazzaley et al., 2008). This age-â•‰related decline
in performance is related to a deficit in suppressing neural activity associated with
processing irrelevant stimuli as assessed by BOLD activity in sensory selective cor-
tex (Figure 8.1)(Gazzaley et al., 2005). This selective decline in suppressing neural
activity to irrelevant stimuli has been observed during both early and later stages of
stimulus processing (i.e., between 100 and 500 ms post stimulus onset; Figure 8.2)
(Gazzaley et al., 2008; Deiber et al., 2010; Finnigan et al., 2011). Importantly, this
suppression deficit in aging was observed as alterations in both amplitude and latency
of neural activity to irrelevant information, thereby helping to reconcile models that
suggest age-â•‰related declines in cognition result from deficient inhibition (Hasher and
Zacks, 1988; Hasher et al., 2007) or slowed processing speed (Salthouse, 1996).
This suppression deficit of irrelevant sensory information not only predicts work-
ing memory decline due to distraction, but is also related to decreased activity in
PFC as well as decreased functional connectivity between sensory cortex and PFC
(Campbell et al., 2012; Chadick et al., 2014). Furthermore, with increasing task
demands, older adults do not increase frontoparietal attention network activity as
do younger adults (Prakash et al., 2012). Anatomically, the magnitude of age-â•‰based
distractibility and suppression deficit in sensory cortex is related to decreased gray
matter volume (Figure 8.3) in medial PFC as well as decreased white matter integrity
(Figure 8.4) in the medial PFC (genu of the corpus callosum) and tracts subserving
frontoposterior attentional networks (superior longitudinal fasciculus)(Chadick et al.,
2014). Together, these results support models that suggest age-â•‰related declines in cog-
nition stem from deficient top-â•‰down control mechanisms in the PFC that fail to mod-
ulate sensory cortical activity based on task goals (Gazzaley, 2012).
During a change detection task with an interfering stimulus presented in the middle
of the delay period (i.e., during working memory maintenance), older adults exhibit
a larger working memory decline compared to younger adults due to interference
(Clapp et al., 2011; Clapp and Gazzaley, 2012; Ku et al., Submitted). The magnitude
220
(A) Younger subjects (B) Older subjects

5.0 5.0 *
4.5 * 4.5
4.0 4.0
3.5 3.5
3.0 3.0
2.5 2.5
∆β
β
2.0 2.0
1.5 1.5
1.0 1.0
0.5 0.5
0 0
Remember Passive Ignore Remember Passive Ignore
scenes view scenes scenes view scenes
(C) Enhancement index (D) Suppression index

2.5 2.5 *
2.0 2.0
1.5 1.5
1.0 1.0
∆β
β
0.5 0.5
0 0
–0.5 Younger Older –0.5 Younger Older
subjects subjects subjects subjects
Figure 8.1 fMRI data showing a selective deficit of top-down suppression in older adults.
(a,b) Within-group comparisons of the BOLD signal magnitude in the scene-selective region of
interest from visual cortex during the “remember scenes,” “passive view” and “ignore scenes”
conditions for the (a) younger and (b) older age groups. (c,d) Across-group comparisons of
(c) enhancement (“remember scenes”—“passive view”) and (d) suppression (“passive view”—
“ignore scenes”) indices. Error bars indicate standard error of the mean. Previously published
(Gazzaley et al., 2005).
of this age-related decline is contingent on the type of interference, such that inter-
ruptions (stimuli requiring attention) negatively affect working memory more in older
adults compared to distractions (stimuli that are to be ignored). Prior to distraction,
older adults exhibit less inhibition-related neural activity (i.e., alpha oscillations) in
parietal-occipital regions as well as decreased functional connectivity (i.e., alpha
coherence) between PFC and sensory cortex, suggesting older adults are not pre-
pared to inhibit neural processing of the impending distraction (Ku et al., Submitted).
Indeed, older adults failed to suppress neural activity to the distracting stimulus (Clapp
and Gazzaley, 2012). Importantly, older adults who exhibited more youth-like neural
activity (i.e., increased pre-distractor inhibition and increased distractor suppression)
showed better working memory performance. This is in line with research indicating
that older adults with high working memory performance exhibit comparable neural
activity to younger adults by attentionally modulating neural activity during early
visual processing stages (Werkle-Bergner et al., 2012).
Following an interruption (stimulus requiring attention), older adults do not
effectively disengage from the interruption, which may result from deficient
221
Early (0–200 ms) Mid (200–500 ms) Late (500–650 ms)
ec ec ec
1s 1s 1s
ec
2s 5 se
c ec
0–. .2–. 65 s
.5–.
P1 Amplitude P300 Amplitude

6 3
Amplitude (µV)
Amplitude (µV)
5 * 2.5
2 Alpha power
4 0
Mean ERSP (dB)

3 1.5
–0.5
2 1 –1
1 0.5 –1.5
0 0
Younger Older Younger Older –2
–2.5
Younger Older
N1 Latency Gamma power
200 2.5
Mean ERSP (dB)
*
Latency (ms)
190 2 Relevant
180 1.5 Passive
170 1
Irrelevant
160 0.5
150 0
Younger Older Younger Older
Figure 8.2 EEG data revealing an age-related deficit in top-down suppression in the earliest measures: P1 amplitude and N1 latency. All within-group t tests
are designated as significant by brackets (P < 0.05). The asterisk denotes that only P1 amplitude and N1 latency revealed a significant age x task interaction
plus a significant across-group suppression deficit. Error bars indicate standard error of the mean. Previously published (Gazzaley et al., 2008), copyright
2008 National Academy of Sciences, USA.
222
(A)
(B)
P < 0.05 P < 0.001
Figure 8.3 Gray-matter volume. Age-related declines in gray-matter volume are associated

with (a) increased distractibility and, (b) deficient neural suppression of sensory cortex activity
in older participants. Previously published (Chadick et al., 2014). (See color plate also)
frontal-posterior communication (Ku et al., Submitted). In support of this, fMRI func-

tional connectivity data between the PFC and sensory selective cortex show that mem-
ory declines in aging due to interruption stem from a deficient ability to dynamically
allocate attention and switch between functional brain networks (Clapp et al., 2011).
The flexible manner in which frontoparietal networks adapt based on attention alloca-
tion is thought to be a fundamental feature of cognitive control (Zanto and Gazzaley,
2013) and may therefore underlie multiple age-related decrements in cognition.
(A) Anterior (B)
Genu of corpus
callosum
Superior longitudinal
fasciculus
Uncinate
Right fasciculus Left
Splenium of corpus
callosum
Posterior
z = 20 z = 40
P = 0.05 P = 0.01
Figure 8.4 White-matter integrity. Age-related declines in white-matter integrity (fractional

anisotropy) are associated with (a) increased distractibility and, (b) deficient neural suppression
of sensory cortex activity in older participants. Previously published (Chadick et al., 2014).
â•‡ 223
Similar to the age-â•‰related declines in expectation processes observed during

sensory detection tasks, older adults do not utilize predictive information to over-
come deficits in suppressing irrelevant distractions (Zanto et al., 2010c; Clapp and
Gazzaley, 2012; Ku et al., Submitted). Whereas younger adults may use predictive
cues to enhance performance and modulate sensory cortical activity for an expected
stimulus (Bollinger et al., 2010), older adults do not exhibit performance benefits
from such predictive information nor do they modulate sensory cortex activity prior
to stimulus onset (Bollinger et al., 2011). However, older adults with mild cognitive
impairment may improve working memory performance via increasing cholinergic
transmission with pharmaceuticals, which results in enhanced attentional modula-
tion of sensory cortex as well as increased functional connectivity with attention and
memory networks (i.e., between sensory selective cortex, PFC, and hippocampus) (Pa
et al., 2013). Despite the functional and structural changes associated with mild cog-
nitive impairment, it is interesting to note that age-â•‰related declines in attentional mod-
ulation of sensory cortex and object selective neural networks during change detection
tasks can be reliably elicited in the same participants across multiple fMRI sessions
separated by months (Zanto et al., 2014). Overall, age-â•‰based declines in object change
detection stem from deficient structure and function in PFC to attentionally modulate
sensory cortex activity during early perceptual processing stages as well as in antici-
pation of a stimulus.
Discrimination
Magnetoencephalography (MEG) data during object discrimination has indicated that
repetition suppression effects are reduced in older adults between 50 and 800 ms post
stimulus onset when attending to repeated auditory words (Aine et al., 2005). This
age-â•‰related deficit in attenuation arises in sensory cortex and is thought to reflect
changes that span pre-â•‰attentive to attentive stages of processing. Similar age-â•‰related
declines in repetition suppression have been observed in the visual domain as well
using fMRI (Miyakoshi et al., 2012). However, with increased repetitions (> 7), older
adults may show comparable effects, indicating a delayed repetition suppression
effect in aging. Interestingly, older adults exhibit neural attenuation to repeated visual
scenes that are unattended, whereas younger adults do not exhibit such repetition
suppression (Schmitz et al., 2010). Because older adults remembered the irrelevant
information better than younger adults, this suggests that older adults do not ignore
distraction, but rather, co-â•‰encode relevant and irrelevant information. Indeed, older
adults are more negatively affected by distracting faces concomitant with a decline
in inhibiting neural activity related to distractor processing (de Fockert et al., 2009),
which may subsequently affect later processing stages of attentional control (Daniel
and Bentin, 2012).
During a word-matching task with object distractors, younger adults are able to uti-
lize long ISIs to benefit from irrelevant, yet semantically related, objects (Wilkinson
et al., 2013). This was interpreted as strategic differences such that younger adults
semantically primed neural resources by incorporating “irrelevant” information,
whereas older adults did not utilize this strategy presumably due to limited resources.
224
However, it is unclear whether this reflects a different strategy or an inability in older

adults to benefit from predictive information. Notably, in two different object discrim-
ination tasks, older adults did not utilize predictive cues to enhance performance, nor
did the older adults exhibit modulation of neural activity in anticipation of the cued
target (Zanto et al., 2011b). Together, object discrimination paradigms corroborate
other object, feature and spatial paradigms to show age-â•‰related declines in perfor-
mance result from a deficient ability to inhibit irrelevant information, which may stem
from declines in expectation processes prior to the onset of the irrelevant information.
Overall, results from object-â•‰based selective attention studies corroborate research
from spatial-â•‰and feature-â•‰based selective attention paradigms to show older adults
utilize frontoparietal attentional networks to compensate for visual processing defi-
ciencies. Such bottom-â•‰up visual processing deficits include a decrease in neural spe-
cialization, or dedifferentiation, which indicates that neural activity in sensory cortex
in older adults respond to a variety of objects, whereas the same region may only
respond to one type of object in younger adults. It is interesting to speculate whether
age-â•‰related declines in top-â•‰down modulation of sensory cortex activity may be related
to dedifferentiation in aging insofar as modulating neural activity in specific sensory
regions could have unintended perceptual consequences due to decreased specializa-
tion of the modulated region. Nonetheless, it is known that age-â•‰related declines in
top-â•‰down modulation of sensory cortex activity is marked functionally by decreased
activity in PFC as well as decreased functional connectivity between PFC and sensory
cortex. Anatomically, top-â•‰down deficits in aging are related to decreased gray mat-
ter volume in PFC and white matter degradation in PFC and frontoparietal regions
subserving attention networks. Furthermore, it should be noted that these age-â•‰related
declines in top-â•‰down modulation occur prior to stimulus onset and propagate through
various processing stages (e.g., 100–â•‰500 ms post stimulus onset). The consequence
of these functional and anatomical changes in aging is that older adults do not ignore
distraction, but rather, co-â•‰encode relevant and irrelevant information, effectively over-
loading limited cognitive resources. Thus, memory declines in aging may stem from
a deficient ability to dynamically allocate attention and switch between functional
brain networks.
Internal Selective Attention
Internal selective attention refers to the focusing of attention on representations that

are not present in the environment via refreshing working memory traces, recalling
long-â•‰term memory, or self-â•‰generation, as in mental imagery. It has been suggested
that age-â•‰related declines in mental imagery may stem from shrinkage of the prefrontal
cortex (Raz et al., 1999). Recent neuroimaging research has supported this to show an
age-â•‰related decline in modulating neural activity in stimulus-â•‰selective visual cortical
regions during imagery, but not during perception (Figure 8.5) (Kalkstein et al., 2011).
Thus, age-â•‰based declines in imagery may not be fully attributed to deficient engage-
ment of sensory cortex. Whereas imagery invokes memory processes, age-â•‰based
declines in disambiguating a relevant memory from similar irrelevant memories is
coincident with decreased activity and cortical thickness in anterior PFC (Fandakova
225
Selective Attention and Inhibitory Control in the Aging Brain 225
(A) Face Selective Motion Selective

ROI ROI
Younger adults Older adults Younger adults Older adults
*** ***
15 *** 15 15 15 ***
10 10 10 10
Beta
Beta
5 5 5 5
0 0 0 0
Perception Faces Motion Perception
(B) Younger adults Older adults Younger adults Older adults
4 4 4 * 4
**
Beta
2 2 Beta 2 2
0 0 0 0
Imagery Imagery
Figure 8.5 fMRI region of interest (ROI) analysis. Within-group comparisons of BOLD activ-
ity in face-selective and motion-selective ROIs for both younger and older adults during (a) per-
ception and (b) mental imagery. Error bars are +/−SEM. ***p < 0.0001; **p < 0.01; *p < 0.05.
Previously published (Kalkstein et al., 2011).
et al., 2014). Thus, similar to external (spatial, feature, object) selective attention, age-
related declines in internal selective attention appear to stem from deficient structure
and function within the PFC.
Similar to how older adults do not utilize predictive information to modulate neural
activity for an impending object, retrospective cues help younger, but not older, adults
enhance working memory capacity by orienting attention to internal representations
(Duarte et al., 2013). This is marked neurally by decreased parietal delay activity
only in younger adults, which suggests a reduced impact of working memory load.
Interestingly, younger, but not older, adults also utilize retrospective cues to intention-
ally forget no longer relevant information (Rizio and Dennis, 2014). Whereas inten-
tional and incidental forgetting in younger adults is associated with distinct frontal and
parietal regions, respectively, older adults do not exhibit differential neural activity to
these cognitive functions. These results indicate older adults have difficulty recruiting
inhibitory mechanisms to remove irrelevant items from memory, similar to reports
(detailed above) indicating older adults do not disengage from irrelevant items in the
external environment (Cashdollar et al., 2013; Cona et al., 2013; Ku et al., Submitted).
The consequence of failing to release irrelevant information from memory may over-
load limited memory stores and hinder memory for relevant information. Indeed, the
negative impact of distraction on memory performance is more pronounced in aging
and has been attributed to disrupted neural networks subserving attention and visual
226
imagery, which is independent of the stimulus modality (i.e., auditory or visual; Wais
and Gazzaley, 2014).
It could be argued that many mental imagery tasks involve assessing memory and
other executive functions in addition to internal attention. Another means to evaluate
internal attention is by measuring activity of brain regions in a “default network”
that is preferentially active during self-â•‰referential thoughts and mind wandering and
is suppressed during tasks that require attention to the external environment. This
default mode network (DMN) consists largely of cortical midline structures such as
the medial prefrontal cortex, anterior cingulate, and posterior cingulate cortex (e.g.,
Gusnard et al., 2001). Of interest, older adults display declines in activity (Koch et al.,
2010) and connectivity (Grady et al., 2010) within the DMN. Moreover, decreased
default mode activity during rest predicts declines in attention, processing speed, and
executive function in ageing (Damoiseaux et al., 2008) as well as decreased memory
performance (Sambataro et al., 2010) and decreased mindfulness disposition (Prakash
et al., 2013).
During cognitive tasks, older adults exhibit a decline in suppressing neural activity
in the DMN (i.e., failing to disengage; Lustig et al., 2003), which has been related to
declines in ignoring distraction and deficient working memory performance (Chadick
et al., 2014). These age-â•‰based differences become more pronounced with increased
task difficulty and occur concomitant with greater suppression of DMN activity in
younger, but not older, adults (Persson et al., 2007). Interestingly, older adults fail
to disengage DMN activity during visuospatial planning, a task that normally uti-
lizes only frontoposterior attentional networks (Spreng et al., 2013). This indicates a
decline in flexible network interactivity and reduced dynamic range of network mod-
ulation to changing task demands. This also shows that older adults exhibit reduced
specialization of neural networks during both external and internal attention. Such
reduced neural specialization, or dedifferentiation, has recently been shown to be
more pronounced during mental imagery, compared to perception, of previously
viewed videos (St-â•‰Laurent et al., 2014). Together, research addressing internal atten-
tion shows age-â•‰related declines in disambiguating relevant from irrelevant memories,
intentionally forgetting irrelevant memories, and mental imagery. These declines stem
from changes in the structure and function of PFC, which result in a loss of neural spe-
cialization in utilizing either frontoposterior networks for external attention or DMN
for internal attention related processes.
Conclusion
Throughout this chapter, several themes emerged to show that age-â•‰related declines in
selective attention are less contingent on the information to be attended/â•‰ignored (i.e.,
spatial, feature, object, internal), but rather, sensitive to level of task demands. This
suggests that disparate findings in aging research may in some cases be attributed to
differences in task difficulty. Importantly, when age-â•‰related declines in selective atten-
tion are observed, it often interacts with other cognitive domains and other known age-â•‰
related changes as discussed throughout this book. For example, response slowing is
one of the most well documented changes in aging (see Chapter 10) and may arise
â•‡ 227
from multiple sources such as bottom-â•‰up perceptual changes and motoric decline.
Yet, once perceptual and motoric slowing is accounted for, older adults often respond
slower during selective attention tasks, especially with increased task difficulty such
as a concurrent working memory load or the presence of distractors. These perfor-
mance declines in aging are often marked by structural and functional changes in the
brain (see Chapter 7), which may begin in mid-â•‰adulthood (see Chapter 14). Although
dedifferentiation in sensory cortex is typically associated with performance declines,
functional changes in the prefrontal cortex often serve as a compensatory mechanism
to retain performance abilities. In other cases, typically under increased task demands,
compensatory mechanisms are not sufficient to uphold task performance. Additional
research will be required to ascertain what processes are in need of compensation and
whether failed compensatory mechanisms reflect declines in attention and inhibitory
control or if it represents limitations in top-â•‰down control to overcome perceptual defi-
cits. Nonetheless, age-â•‰related declines in frontoparietal attentional networks, as well
as the ability to dynamically switch between networks, are often associated with defi-
cient memory performance (see Chapters 9, 12, 13).
References
Aine CJ, Adair JC, Knoefel JE, Hudson D, Qualls C, Kovacevic S, Woodruff CC, Cobb W,
Padilla D, Lee RR, Stephen JM (2005) Temporal dynamics of age-â•‰related differences in
auditory incidental verbal learning. Cognitive Brain Research 24:1–â•‰18.
Allen HA, Payne H (2012) Similar behaviour, different brain patterns: Age-â•‰related changes in
neural signatures of ignoring. Neuroimage 59:4113–â•‰4125.
Alperin BR, Mott KK, Rentz DM, Holcomb PJ, Daffner KR (2014) Investigating the age-â•‰
related “anterior shift” in the scalp distribution of the P3b component using principal com-
ponent analysis. Psychophysiology 51:620–â•‰633.
Alperin BR, Haring AE, Zhuravleva TY, Holcomb PJ, Rentz DM, Daffner KR (2013) The
Dissociation between Early and Late Selection in Older Adults. J Cogn Neurosci
25:2189–â•‰2206.
Amenedo E, Lorenzo-â•‰Lopez L, Pazo-â•‰Alvarez P (2012) Response processing during visual
search in normal aging: The need for more time to prevent cross talk between spatial
attention and manual response selection. Biol Psychol 91:201–â•‰211.
Ansado J, Monchi O, Ennabil N, Faure S, Joanette Y (2012) Load-â•‰dependent posterior-â•‰anterior
shift in aging in complex visual selective attention situations. Brain Res 1454:14–â•‰22.
Ansado J, Monchi O, Ennabil N, Deslauriers J, Jubault T, Faure S, Joanette Y (2013) Coping
with task demand in aging using neural compensation and neural reserve triggers pri-
marily intra-â•‰hemispheric-â•‰based neurofunctional reorganization. Neuroscience Research
75:295–â•‰304.
Bennett IJ, Motes MA, Rao NK, Rypma B (2012) White matter tract integrity predicts
visual search performance in young and older adults. Neurobiology of Aging 33:433.
e21–433e.31.
Berry AS, Zanto TP, Rutman AM, Clapp WC, Gazzaley A (2009) Practice-â•‰Related Improvement
in Working Memory is Modulated by Changes in Processing External Interference.
J NeuroÂ�physiol 102:1779–â•‰1789.
Berry AS, Zanto TP, Clapp WC, Hardy J, Delahunt P, Mahncke H, Gazzaley A (2010) The
influence of perceptual training on working memory in older adults. Public Library of
Science One 5(10): e11537. doi:10.1371/journal.pone.0011537
228
Bollinger J, Rubens MT, Zanto TP, Gazzaley A (2010) Expectation-driven changes in cortical
functional connectivity influence working-memory and long-term memory performance.
J Neurosci 30:14399–14410.
Bollinger J, Rubens MT, Masangkay E, Gazzaley A (2011) An expectation-based memory defi-
cit in aging. Neuropsychologia 49:1466–1475.
Burianova H, Lee Y, Grady CL, Moscovitch M (2013) Age-related dedifferentiation and com-
pensatory changes in the functional network underlying face processing. Neurobiology of
Aging 34:2759–2767.
Campbell KL, Grady CL, Ng C, Hasher L (2012) Age differences in the frontoparietal cognitive
control network: Implications for distractibility. Neuropsychologia 50:2212–2223.
Cashdollar N, Fukuda K, Bocklage A, Aurtenetxe S, Vogel EK, Gazzaley A (2013) Prolonged
Disengagement From Attentional Capture in Normal Aging. Psychology and Aging
28:77–86.
Chadick JZ, Zanto TP, Gazzaley A (2014) Structural and functional differences in medial prefron-
tal cortex underlie distractibility and suppression deficits in ageing. Nature Communications
5:4223. doi:10.1038/ncomms5223
Chao LL, Knight RT (1997) Prefrontal deficits in attention and inhibitory control with aging.
Clapp W, Gazzaley A (2012) Distinct mechanisms for the impact of distraction and interruption
on working memory in aging. Neurobiology of Aging 33:134–148.
Clapp WC, Rubens MT, Sabharwal J, Gazzaley A (2011) Deficit in switching between func-
tional brain networks underlies the impact of multitasking on working memory in older
adults. Proc Natl Acad Sci U S A 108(17):7212–7217.
Cona G, Bisiacchi PS, Amodio P, Schiff S (2013) Age-related decline in attentional shift-
ing: Evidence from ERPs. Neuroscience Letters 556:129–134.
Cowan N (1995) Attention and memory: An integrated framework. New York: Oxford
University Press.
Curran T, Hills A, Patterson MB, Strauss ME (2001) Effects of aging on visuospatial atten-
tion: an ERP study. Neuropsychologia 39:288–301.
Daffner KR, Alperin BR, Mott KK, Holcomb PJ (2014) Age-related differences in the automatic
processing of single letters: implications for selective attention. Neuroreport 25:77–82.
Damoiseaux JS, Beckmann CF, Arigita EJS, Barkhof F, Scheltens P, Stam CJ, Smith SM,
Rombouts S (2008) Reduced resting-state brain activity in the “default network” in normal
aging. Cerebral Cortex 18:1856–1864.
Daniel S, Bentin S (2012) Age-related changes in processing faces from detection to identifica-
tion: ERP evidence. Neurobiology of Aging 33(1): 206.e1–206.e28.
de Fockert JW, Ramchurn A, van Velzen J, Bergstrom Z, Bunce D (2009) Behavioral and ERP
evidence of greater distractor processing in old age. Brain Res 1282:67–73.
Deiber MP, Rodriguez C, Jaques D, Missonnier P, Emch J, Millet P, Gold G, Giannakopoulos
P, Ibanez V (2010) Aging Effects on Selective Attention-Related Electroencephalographic
Patterns during Face Encoding. Neuroscience 171:173–186.
Dempster FN (1992) The rise and fall of the inhibitory mechanism: Toward a unified theory of
cognitive devlopment and aging. Developmental Review 12:45–75.
Dew ITZ, Buchler N, Dobbins IG, Cabeza R (2012) Where Is ELSA? The Early to Late Shift
in Aging. Cerebral Cortex 22:2542–2553.
229
Drag LL, Bieliauskas LA (2010) Contemporary Review 2009: Cognitive Aging. J Geriatr
Psychiatry Neurol 23:75–93.
Duarte A, Hearons P, Jiang YS, Delvin MC, Newsome RN, Verhaeghen P (2013) Retrospective
attention enhances visual working memory in the young but not the old: An ERP study.
Psychophysiology 50:465–476.
Fandakova Y, Lindenberger U, Shing YL (2014) Deficits in Process-Specific Prefrontal and
Hippocampal Activations Contribute to Adult Age Differences in Episodic Memory
Interference. Cerebral Cortex 24:1832–1844.
Finnigan S, O’Connell RG, Cummins TDR, Broughton M, Robertson IH (2011) ERP mea-
sures indicate both attention and working memory encoding decrements in aging.
Psychophysiology 48:601–611.
Gazzaley A (2012) Top-down modulation deficit in the aging brain: An emerging theory of
cogntiive aging. In: Principles of frontal lobe function, 2nd Edition (Stuss DT, Knight RT,
eds), pp 593–608. New York: Oxford University Press.
Gazzaley A, Nobre AC (2012) Top-down modulation: Bridging selective attention and working
memory. Trends Cogn Sci 16:129–135.
Gazzaley A, Cooney JW, Rissman J, D’Esposito M (2005) Top-down suppression deficit under-
lies working memory impairment in normal aging. Nat Neurosci 8:1298–1300.
Gazzaley A, Clapp W, Kelley J, McEvoy K, Knight R, D’Esposito M (2008) Age-related
top-down suppression deficit in the early stages of cortical visual memory processing.
Proceedings of the National Academy of Science USA 105:13122–13126.
Geerligs L, Saliasi E, Maurits NM, Renken RJ, Lorist MM (2014) Brain mechanisms underly-
ing the effects of aging on different aspects of selective attention. Neuroimage 91:52–62.
Gottlob LR, Madden DJ (1998) Time course of allocation of visual attention after equating for
sensory differences: An age-related perspective. Psychology and Aging 13:138–149.
Grady CL, Yu H, Alain C (2008) Age-related differences in brain activity underlying working
memory for spatial and nonspatial auditory information. Cerebral Cortex 18:189–199.
Grady CL, Haxby JV, Horwitz B, Schapiro MB, Rapoport SI, Ungerleider LG, Mishkin M, Carson
RE, Herscovitch P (1992) Dissociation of Object and Spatial Vision in Human Extrastriate
Cortex -Age-Related-Changes in Activation of Regional Cerebral Blood-Flow Measured with
[O-15]Water and Positron Emission Tomography. J Cogn Neurosci 4:23–34.
Grady CL, Maisog JM, Horwitz B, Ungerleider LG, Mentis MJ, Salerno JA, Pietrini P, Wagner
E, Haxby JV (1994) Age-related changes in cortical blood flow activation during visual
processing of faces and location. J Neurosci 14:1450–1462.
Grady CL, Protzner AB, Kovacevic N, Strother SC, Afshin-Pour B, Wojtowicz M, Anderson JAE,
Churchill N, McIntosh AR (2010) A Multivariate Analysis of Age-Related Differences in
Default Mode and Task-Positive Networks across Multiple Cognitive Domains. Cerebral
Cortex 20:1432–1447.
Gusnard DA, Akbudak E, Shulman GL, Raichle ME (2001) Medial prefrontal cortex and self-
referential mental activity: Relation to a default mode of brain function. Proc Natl Acad
Sci U S A 98:4259–4264.
Haring AE, Zhuravleva TY, Alperin BR, Rentz DM, Holcomb PJ, Daffner KR (2013) Age-
related differences in enhancement and suppression of neural activity underlying selective
attention in matched young and old adults. Brain Res 1499:69–79.
Hartley AA, Kieley JM, Slabach EH (1990) Age differences and similarities in the effects of cues and
prompts. Journal of Experimental Psychology-Human Perception and Performance 16:523–537.
Hasher L, Zacks RT (1988) Working Memory, comprehension and aging: A review and a new
view. In: The Psychology of Learning and Motivation (Bower GH, ed), pp 193–225.
New York, NY: Academic Press.
230
Hasher L, Lustig C, Zacks JM (2007) Inhibitory mechanisms and the control of attention.
In: Variation in working memory (Conway A, Jarrold C, Kane M, Miyake A, Towse J,
eds), pp 227–249. New York: Oxford University Press.
Hedden T, Van Dijk KRA, Shire EH, Sperling RA, Johnson KA, Buckner RL (2012) Failure
to Modulate Attentional Control in Advanced Aging Linked to White Matter Pathology.
Huang CM, Polk TA, Goh JO, Park DC (2012) Both left and right posterior parietal activations
contribute to compensatory processes in normal aging. Neuropsychologia 50:55–66.
Kalkstein J, Checksfield K, Bollinger J, Gazzaley A (2011) Diminished top-down control
underlies a visual imagery deficit in normal aging. J Neurosci 31:15768–15774.
Koch W, Teipel S, Mueller S, Buerger K, Bokde ALW, Hampel H, Coates U, Reiser M, Meindl
T (2010) Effects of aging on default mode network activity in resting state fMRI: Does the
method of analysis matter? Neuroimage 51:280–287.
Kramer AF, Madden DJ (2008) Attention. In: The Handbook of Aging and Cognition, Third
Edition (Craik FI, Salthouse TA, eds). New York: Psychology Press.
Ku Y, Zanto TP, Clapp WC, Gazzaley A (Submitted) Age-related changes in neural oscillations
underlying distraction and interruption.
Langenecker SA, Nielson KA, Rao SM (2004) fMRI of healthy older adults during Stroop
interference. Neuroimage 21:192–200.
Lee Y, Grady CL, Habak C, Wilson HR, Moscovitch M (2011) Face Processing Changes in
Normal Aging Revealed by fMRI Adaptation. J Cogn Neurosci 23:3433–3447.
Leung AWS, He Y, Grady CL, Alain C (2013) Age Differences in the Neuroelectric Adaptation
to Meaningful Sounds. PLoS One 8(7):e68892. doi:10.1371/journal.pone.0068892
Li L, Gratton C, Fabiani M, Knight RT (2013) Age-related frontoparietal changes during the
control of bottom-up and top-down attention: an ERP study. Neurobiology of Aging
34:477–488.
Lorenzo-Lopez L, Amenedo E, Cadaveira F (2008a) Feature processing during visual
search in normal aging: Electrophysiological evidence. Neurobiology of Aging
29:1101–1110.
Lorenzo-Lopez L, Amenedo E, Pascual-Marqui RD, Cadaveira F (2008b) Neural correlates of
age-related visual search decline: A combined ERP and sLORETA study. Neuroimage
41:511–524.
Lorenzo-Lopez L, Doallo S, Vizoso C, Amenedo E, Rodriguez Holguin S, Cadaveira F (2002)
Covert orienting of visuospatial attention in the early stages of aging. Neuroreport
13:1459–1462.
Lorenzo-Lopez L, Gutierrez R, Moratti S, Maestu F, Cadaveira F, Amenedo E (2011) Age-
related occipito-temporal hypoactivation during visual search: Relationships between
mN2pc sources and performance. Neuropsychologia 49:858–865.
RL (2003) Functional deactivations: Change with age and dementia of the Alzheimer type.
Macpherson HN, White DJ, Ellis KA, Stough C, Camfield D, Silberstein R, Pipingas
A (2014) Age-related changes to the neural correlates of working memory which
emerge after midlife. Frontiers in Aging Neuroscience 6(70):1–10. doi:10.3389/
fnagi.2014.00070
Madden DJ (2007) Aging and visual attention. Current Directions in Psychological Science
16:70–74.
231
Madden DJ, Hoffman JM (1997) Application of positron emission tomography to age-related

cognitive changes. In: Brain Imaging in clinical psychiatry (Krishmann KRR, Doraiswamy
PM, eds). New York: M. Dekker.
Madden DJ, Turkington TG, Provenzale JM, Hawk TC, Hoffman JM (1997) Selective and
divided visual attention: Age-related changes in regional cerebral blood flow measured by
(H2O)-O-15 PET. Hum Brain Mapp 5:389–409.
Madden DJ, Spaniol J, Whiting WL, Bucur B, Provenzale JM, Cabeza R, White LE, Huettel
SA (2007) Adult age differences in the functional neuroanatomy of visual attention: a
combined fMRI and DTI study. Neurobiol Aging 28:459–476.
Milham MP, Erickson KI, Banich MT, Kramer AF, Webb A, Wszalek T, Cohen NJ (2002)
Attentional control in the aging brain: insights from an fMRI study of the stroop task.
Brain Cogn 49:277–296.
Miyakoshi M, Chen SHA, Matsuo K, Wu CY, Suzuki A, Nakai T (2012) Extensive stimulus
repetition leads older adults to show delayed functional magnetic resonance imaging adap-
tation. Brain Imaging and Behavior 6:357–365.
Mohtasib RS, Lumley G, Goodwin JA, Emsley HCA, Sluming V, Parkes LM (2012) Calibrated
fMRI during a cognitive Stroop task reveals reduced metabolic response with increasing
age. Neuroimage 59:1143–1151.
Muller-Oehring EM, Schulte T, Rohlfing T, Pfefferbaum A, Sullivan EV (2013) Visual Search
and the Aging Brain: Discerning the Effects of Age-Related Brain Volume Shrinkage on
Alertness, Feature Binding, and Attentional Control. Neuropsychology 27:48–59.
Nagel IE, Preuschhof C, Li SC, Nyberg L, Backman L, Lindenberger U, Heekeren HR (2009)
Performance level modulates adult age differences in brain activation during spatial work-
ing memory. Proc Natl Acad Sci U S A 106:22552–22557.
Nielson KA, Langenecker SA, Garavan H (2002) Differences in the functional neuroanatomy
of inhibitory control across the adult life span. Psychol Aging 17:56–71.
Nielson KA, Langenecker SA, Ross TJ, Garavan H, Rao SM, Stein EA (2004) Comparability
of functional MRI response in young and old during inhibition. Neuroreport 15:129–133.
Nissen MJ, Corkin S (1985) Effectiveness of attentional cueing in older and younger adults.
Journals of Gerontology 40:185–191.
O’connell RG, Balsters JH, Kilcullen SM, Campbell W, Bokde AW, Lai R, Upton N, Robertson
IH (2012) A simultaneous ERP/fMRI investigation of the P300 aging effect. Neurobiology
of Aging 33:2448–2461.
Pa J, Berry AS, Compagnone M, Boccanfuso J, Greenhouse I, Rubens MT, Johnson JK,
Gazzaley A (2013) Cholinergic Enhancement of Functional Networks in Older Adults
with Mild Cognitive Impairment. Annals of Neurology 73:762–773.
Parasuraman R (1998) The Attentive Brain: Issues and Prospects. In: The Attentive Brain
(Parasuraman R, ed), pp 3–15. Cambridge, MA: MIT Press.
Park DC, Reuter-Lorenz P (2009) The Adaptive Brain: Aging and Neurocognitive Scaffolding.
Annual Review of Psychology 60:173–196.
Passow S, Westerhausen R, Hugdahl K, Wartenburger I, Heekeren HR, Lindenberger U, Li SC
(2014) Electrophysiological Correlates of Adult Age Differences in Attentional Control of
Auditory Processing. Cerebral Cortex 24:249–260.
Paxton JL, Barch DM, Racine CA, Braver TS (2008) Cognitive control, goal maintenance, and
prefrontal function in healthy aging. Cerebral Cortex 18:1010–1028.
Peltz CB, Gratton G, Fabiani M (2011) Age-related changes in electrophysiological and neu-
ropsychological indices of working memory, attention control, and cognitive flexibility.
Frontiers in Psychology 2(190):1–12. doi:10.3389/fpsyg.2011.00190
Persson J, Lustig C, Nelson JK, Reuter-Lorenz PA (2007) Age differences in deactivation: A link
to cognitive control? J Cogn Neurosci 19:1021–1032.
232
Piefke M, Onur OA, Fink GR (2012) Aging-related changes of neural mechanisms underlying
visual-spatial working memory. Neurobiology of Aging 33:1284–1297.
Prakash RS, Heo S, Voss MW, Patterson B, Kramer AF (2012) Age-related differences in cor-
tical recruitment and suppression: Implications for cognitive performance. Behavioural
Brain Research 230:192–200.
Prakash RS, De Leon AA, Klatt M, Malarkey W, Patterson B (2013) Mindfulness disposition
and default-mode network connectivity in older adults. Social Cognitive and Affective
Neuroscience 8:112–117.
Prakash RS, Erickson KI, Colcombe SJ, Kim JS, Voss MW, Kramer AF (2009) Age-related
differences in the involvement of the prefrontal cortex in attentional control. Brain and
Cognition 71:328–335.
Quigley C, Muller MM (2014) Feature-Selective Attention in Healthy Old Age: A Selective
Decline in Selective Attention? J Neurosci 34:2471–2476.
Quigley C, Andersen SK, Schulze L, Grunwald M, Muller MM (2010) Feature-selective atten-
tion: Evidence for a decline in old age. Neuroscience Letters 474:5–8.
Rabbitt P, Mogapi O, Scott M, Thacker N, Lowe C, Horan M, Pendleton N, Jackson A, Lunn
D (2007) Effects of global atrophy, white matter lesions, and cerebral blood flow on
age-related changes in speed, memory, intelligence, vocabulary, and frontal function.
Raz N, Briggs SD, Marks W, Acker JD (1999) Age-related deficits in generation and manip-
ulation of mental images: II. The role of dorsolateral prefrontal cortex. Psychology and
Aging 14:436–444.
Reuter EM, Voelcker-Rehage C, Vieluf S, Winneke AH, Godde B (2013) A parietal-to-frontal
shift in the P300 is associated with compensation of tactile discrimination deficits in late
middle-aged adults. Psychophysiology 50:583–593.
Reuter-Lorenz P, Lustig C (2005) Brain aging: reorganizing discoveries about the aging mind.
Current Opinion in Neurobiology 15:245–251.
Current Directions in Psychological Science 17:177–182.
Rizio AA, Dennis NA (2014) The Cognitive Control of Memory: Age Differences in the Neural
Correlates of Successful Remembering and Intentional Forgetting. PLoS One 9(1):e87010.
doi:10.1371/journal.pone.0087010
Salthouse TA (1996) The processing-speed theory of adult age differences in cognition. Psychol
Rev 103:403–428.
Sambataro F, Murty VP, Callicott JH, Tan HY, Das S, Weinberger DR, Mattay VS (2010) Age-
related alterations in default mode network: Impact on working memory performance.
Neurobiology of Aging 31:839–852.
Sander MC, Werkle-Bergner M, Lindenberger U (2012) Amplitude modulations and inter-trial
phase stability of alpha-oscillations differentially reflect working memory constraints
across the lifespan. Neuroimage 59:646–654.
Schmitz TW, Cheng FHT, De Rosa E (2010) Failing to Ignore: Paradoxical Neural Effects
of Perceptual Load on Early Attentional Selection in Normal Aging. J Neurosci
30:14750–14758.
Schulte T, Muller-Oehring EM, Chanraud S, Rosenbloom MJ, Pfefferbaum A, Sullivan EV
(2011) Age-related reorganization of functional networks for successful conflict res-
olution: A combined functional and structural MRI study. Neurobiology of Aging
32:2075–2090.
Solbakk AK, Alpert GF, Furst AJ, Hale LA, Oga T, Chetty S, Pickard N, Knight RT (2008)
Altered prefrontal function with aging: Insights into age-associated performance decline.
Brain Res 1232:30–47.
233
Spreng RN, Sepulcre J, Turner GR, Stevens WD, Schacter DL (2013) Intrinsic Architecture
Underlying the Relations among the Default, Dorsal Attention, and Frontoparietal Control
Networks of the Human Brain. J Cogn Neurosci 25:74–86.
St-Laurent M, Abdi H, Bondad A, Buchsbaum BR (2014) Memory Reactivation in Healthy
Aging: Evidence of Stimulus-Specific Dedifferentiation. J Neurosci 34:4175–4186.
cept. Journal of the International Neuropsychological Society 8:448–460.
Stormer V, Eppinger B, Li SC (2014) Reward speeds up and increases consistency of visual
selective attention: a lifespan comparison. Cognitive Affective & Behavioral Neuroscience
14:659–671.
Stormer VS, Li SC, Heekeren HR, Lindenberger U (2013) Normative shifts of cortical mech-
anisms of encoding contribute to adult age differences in visual-spatial working memory.
Sullivan EV, Adalsteinsson E, Pfefferbaum A (2006) Selective age-related degradation of
anterior callosal fiber bundles quantified in vivo with fiber tracking. Cerebral Cortex
16:1030–1039.
Talsma D, Kok A, Ridderinkhof KR (2006) Selective attention to spatial and non-spatial visual
stimuli is affected differentially by age: Effects on event-related brain potentials and per-
formance data. Int J Psychophysiol 62:249–261.
Toepper M, Gebhardt H, Bauer E, Haberkamp A, Beblo T, Gallhofer B, Driessen M, Sammer
G (2014) The impact of age on load-related dorsolateral prefrontal cortex activation.
Frontiers in Aging Neuroscience 6(9):1–10. doi:10.3389/fnagi.2014.00009
Tse CY, Gordon BA, Fabiani M, Gratton G (2010) Frequency analysis of the visual steady-state
response measured with the fast optical signal in younger and older adults. Biol Psychol
85:79–89.
Turken AU, Whitfield-Gabrieli S, Bammer R, Baldo JV, Dronkers NF, Gabrieli JDE (2008)
Cognitive processing speed and the structure of white matter pathways: Convergent evi-
dence from normal variation and lesion studies. Neuroimage 42:1032–1044.
Vaden RJ, Hutcheson NL, McCollum LA, Kentros J, Visscher KM (2012) Older adults,
unlike younger adults, do not modulate alpha power to suppress irrelevant information.
Wais PE, Gazzaley A (2014) Distractibility during retrieval of long-term memory: domain-gen-
eral interference, neural networks and increased susceptibility in normal aging. Frontiers
in Psychology 5(280):1–12. doi:10.3389/fpsyg.2014.00280
Wang Y, Fu SM, Greenwood P, Luo YJ, Parasuraman R (2012) Perceptual load, voluntary atten-
tion, and aging: An event-related potential study. Int J Psychophysiol 84:17–25.
Werkle-Bergner M, Freunberger R, Sander MC, Lindenberger U, Klimesch W (2012) Inter-
individual performance differences in younger and older adults differentially relate to
amplitude modulations and phase stability of oscillations controlling working memory
contents. Neuroimage 60:71–82.
West RL (1996) An application of prefrontal cortex function theory to cognitive aging. Psychol
Bull 120:272–292.
Wilkinson AJ, Yang LX, Dyson BJ (2013) Modulating younger and older adults’ performance
in ignoring pictorial information during a word matching task. Brain and Cognition
83:351–359.
Wolf D, Zschutschke L, Scheurich A, Schmitz F, Lieb K, Tuscher O, Fellgiebel A (2014) Age-
related increases in stroop interference: Delineation of general slowing based on behav-
ioral and white matter analyses. Hum Brain Mapp 35:2448–2458.
Yamaguchi S, Tsuchiya H, Kobayashi S (1995) Electrophysiologic correlates of age effects on
visuospatial attention shift. Cognitive Brain Research 3:41–49.
234
Zanto TP, Gazzaley A (2013) Fronto-parietal network: flexible hub of cognitive control. Trends
Cogn Sci 17:602–603.
Zanto TP, Gazzaley A (2014) Attention and aging. In: Handbook of Attention (Nobre AC,
Kastner S, eds), pp 927–971. Oxford: Oxford University Press.
Zanto TP, Toy B, Gazzaley A (2010a) Delays in neural processing during working memory
encoding in normal aging. Neuropsychologia 48:13–25.
Zanto TP, Pa J, Gazzaley A (2014) Reliability measures of functional magnetic resonance
imaging in a longitudinal evaluation of mild cognitive impairment. Neuroimage
84:443–452.
Zanto TP, Rubens MT, Bollinger J, Gazzaley A (2010b) Top-down modulation of visual feature
processing: The role of the inferior frontal junction. Neuroimage 53:736–745.
Zanto TP, Rubens MT, Thangavel A, Gazzaley A (2011a) Causal role of the prefrontal cortex
in top-down modulation of visual processing and working memory. Nature Neuroscience
14:656–661.
Zanto TP, Sekuler R, Dube C, Gazzaley A (2013) Age-related changes in expectation based
modulation of motion detectability. PLoS One 8:e69766.
Zanto TP, Hennigan K, Ostberg M, Clapp WC, Gazzaley A (2010c) Predictive knowledge of
stimulus relevance does not influence top-down suppression of irrelevant information in
older adults. Cortex 46:561–574.
Zanto TP, Pan P, Liu H, Bollinger J, Nobre AC, Gazzaley A (2011b) Age-Related Changes in
Orienting Attention in Time. J Neurosci 31:12461–12470.
Zhuravleva TY, Alperin BR, Haring AE, Rentz DM, Holcomb PJ, Daffner KR (2014) Age-
Related Decline in Bottom-up Processing and Selective Attention in the Very Old. Journal
of Clinical Neurophysiology 31:261–271.
Zysset S, Schroeter ML, Neumann J, von Cramon DY (2007) Stroop interference, hemo-
dynamic response and aging: An event-related fMRI study. Neurobiology of Aging
28:937–946.
235
Working Memory and Executive

Functions in the Aging Brain
Patricia A. Reuter-Lorenz
Cindy Lustig
C ognition occurs by virtue of neural processes, and likewise, the aging of

cognition is rooted in the effects of age on neural structures and func-
tions. The methods of cognitive neuroscience provide the tools for relating neural and
cognitive declines. As new methods to characterize brain structure and function are
introduced, they can be applied to the analysis of cognitive abilities to elucidate how
their nature and implementation are affected by age. Likewise, as new cognitive para-
digms and behavioral analytics are developed, they can be applied in conjunction with
neural measures to elucidate how their nature and implementation are affected by age.
This dynamic scientific exchange between different levels of analysis, encompassed
by cognitive neuroscience, has been fundamental to advancing our knowledge of why
and how cognition changes as we age. In this chapter we focus on how this exchange
has been applied to understanding the aging of working memory and the executive
processes that comprise it.
Working memory refers to the active and on-line maintenance of information that
is used to guide ongoing behavior, intended actions and longer-term plans. Acting in a
dynamic world requires continuous updating, deleting, and manipulating of this infor-
mation, operations that constitute the “executive processing” components of working
memory. Traditionally, the study of aging, like the field of cognition more generally,
has been guided by the highly influential model of working memory proposed by
Baddeley and Hitch (1974), in which working memory is composed of domain-spe-
cific storage systems such as the phonological loop and the visuo-spatial sketchpad,
and the central executive that operates on the contents of these buffers. In recent years
235
236
there has been a shift away from the idea of working memory as a separate, special-
ized system and toward understanding it as an emergent phenomenon representing
a complex interplay between attention and cognitive control, primarily mediated by
frontoparietal systems, and perceptual/memorial representations, primarily mediated
by posterior cortices (e.g., Cowan, 2001; D’Esposito and Postle, 2015; Postle, 2006;
Jonides et al., 2008; Lustig et al., 2009). The term working memory still serves as
important reference point for scientific communication about this likely multidimen-
sional construct (e.g., a Web of Science search reveals over 13,000 publications on the
topic in 2014 alone). However, as we will describe below, this new understanding of
attention–representation interactions, combined with neuroimaging data, has played
an important role in understanding patterns of age deficits, preserved performance,
and compensation.
The first edition of this book (Cabeza et al., 2005), published approximately 12
years ago, included a chapter on the cognitive neuroscience of working memory and
aging (Reuter-Lorenz and Sylvester, 2005). That chapter evaluated three major issues
that arose primarily from the preceding decades of behavioral research in light of
the newly emerging body of brain imaging evidence: (1) the claim that aging spared
working memory tasks that emphasized pure maintenance relative to those requiring
maintenance plus processing; (2) the role of inhibition, attention, and interference
control deficits in age-related working memory decline; (3) the relative and poten-
tially differential decline of verbal versus visual-spatial memory with age. At that
time, there were less than 30 neuroimaging studies published on the topic, many of
which used positron emission tomography (PET) to measure task-related activity in
relatively small cross-sectional samples of extreme age groups.
The aim of the present chapter is to more o r-less pick up where the previous chap-
ter left off. This will be accomplished in four major sections. First, we provide a brief
overview of new evidence that has emerged over the past 10 years about the structural
aging of brain regions known to be involved in working memory and executive con-
trol, and the life time course of their decline. Second, we consider evidence that has
emerged from functional neuroimaging pertaining to the decline and compensatory
contributions of executive functions and their role in working memory. The third sec-
tion focuses primarily on visual working memory because of the development and
widespread use of the visual change detection paradigm and its application to aging
working memory. In the fourth section we focus on new developments in our under-
standing of age effects on inhibitory and attentional functions, and how these affect
working memory. In a final section we consider life span influences and the potential
for life long plasticity of working memory and executive functions, and factors that
may modify their effectiveness in later life.
Measures of Structural Age-Related Decline

and the Working Memory Network
Thanks to the pioneering work of Goldman- Rakic and her collaborators (e.g.,
Goldman-Rakic, 1987) it has been known for several decades that dorsolateral pre-
frontal cortex (DLPFC; Brodmann’s areas 46/9) and regions of the parietal lobes
237
Working Memory and Executive Functions in the Aging Brain 237
(superior parietal lobule BA 7 and inferior parietal lobule BA 40) are central to the
circuitry underlying working memory, although as described above these circuits
are now thought to mediate attention to and control of working-memory representa-
tions, rather than their maintenance per se (see discussions by e.g., D’Esposito and
Postle, 2015; Jonides et al., 2008; Lustig et al., 2009; Postle, 2015). Human brain
imaging studies have also identified premotor cortex (BA 6) and the inferior frontal
gyrus (BA 44/45) as part of the verbal working memory circuitry (Fegen et al., 2015).
Additionally, supplementary motor cortex including the frontal eye fields (BA 8) as
well as inferior frontal gyrus region BA 47 are known to play critical roles in spatial
working memory (Awh and Jonides, 2001; Jonides et al., 2008).
It has also been known for some time that brain regions age at different rates (e.g.,
Raz and Rodrigue, 2006; Raz et al., 2010) with subregions of the prefrontal cortices
being among the most vulnerable to shrinkage and cortical thinning as a function of
age. More recent analyses of age effects on brain structure corroborate these conclu-
sions by applying uniform analysis methods to combined data from 6 different large
scale studies that include participants (N = 883) ranging from 18–94 years of age (Fjell
et al, 2009). The strongest effects of age were evident in superior, middle and inferior
frontal gyri, which include the major PFC regions known to be important for working
memory and corresponding executive functions. The temporal-parietal junction, which
includes inferior portions of parietal cortex, was also noteworthy for its pronounced
age-related shrinkage. Medial temporal lobe regions, including the hippocampus and
entorhinal cortex are also known to undergo substantial age related atrophy, which is
a pronounced feature of the neuropathology that accompanies Alzheimer’s disease,
but recently shown to be present in older adults with preserved memory who are at
very low risk for dementia (Fjell et al., 2014). There are indications that these regions
contribute to binding processes in working memory (for a review see e.g., Sander et al,
2012), a topic that will be discussed in more detail subsequently in this chapter.
Despite the clear evidence for PFC declines with age, and the compelling evidence
from lesion studies, neuroimaging and unit recordings in behaving animals that PFC is
critical to working memory, the relationship between working memory performance
and PFC structure in older adults is relatively weak (Yuan and Raz, 2014). In a recent
meta-analysis, Yuan and Raz (2014) tested the “bigger is better” hypothesis using
33 studies that measured executive functions and PFC structure (volume or cortical
thickness), the majority published since 2005. They demonstrated a moderate associ-
ation between the size of lateral PFC and Wisconsin Card Sorting (Cohen’s d = 0.3),
but weaker associations for working memory (0.19), as well as trail-making and flu-
ency, both of which were less than 0.1. One additional relationship they observed
was between lateral PFC and interference (e.g., as measured by Stroop), which also
showed an effect size of 0.3, however the authors urged a cautious interpretation
because only five studies contributed to this analysis. The authors attribute the weak
correlation between working memory and lateral PFC structure to the heterogeneous
assortment of tests used to measure working memory, and the potential importance of
network interactions (including regions beyond PFC), as opposed to highly localized
circuitry, underlying working memory performance.
To the extent that working memory and executive functions more generally require
interactions among multiple brain regions, we would expect that the integrity of white
238
matter pathways connecting regions of PFC inter-â•‰hemispherically and with more

posterior brain regions would be related to performance on tasks that measure these
abilities. While such studies are less common than investigations of gray matter, sig-
nificant associations have been observed (see Charlton and Morris, 2015 for a review).
In particular, Kennedy and Raz (2009) used diffusion tensor imaging (DTI) to measure
the integrity of white matter microstructure in a range of regions of interest (ROIs)
including white matter in each of the major lobes, the corpus callosum, and long asso-
ciation fibers. Using verbal and nonverbal measures that required both storage and
processing components of working memory (e.g., n-â•‰back), they found that age-â•‰related
deterioration of frontal white matter, anterior corpus callosum and portions of the
internal capsule was associated with poorer working memory performance. Critically,
apart from processing speed, cognitive tasks measuring episodic memory and interfer-
ence showed different patterns of association with white matter integrity, suggesting
specificity in brain-â•‰behavior relations as opposed to global decline. As the review by
Charlton and Morris (2015) makes clear, however, there is considerable inconsistency
in the age-â•‰related associations between white matter and working memory among the
studies published to date, so more research is clearly needed in this domain.
Age Effects on Neuroâ•‰functional Measures of Working Memory
Early evidence linking aging, PFC function and working memory performance came
from animal studies that used variants of the delayed response task (see Bizon et al.,
2012 for a review) in which a to-â•‰be-â•‰remembered target is presented briefly, hidden
from view for a short delay, and then followed by a command signal that requires a
response whose accuracy is contingent upon the precise memory for the target stim-
ulus. Age-â•‰related performance deficits affecting variants of this task have been docu-
mented in rhesus monkeys (e.g., Bartus et al., 1978) together with volume reductions,
synaptic changes and decreased adrenergic receptor binding all affecting dorsolateral
PFC (Hara, et al., 2012). Moreover Wang, et al. (2011) found age-â•‰related decreases in
the firing rates of dorsolateral PFC neurons during the delay period of a spatial work-
ing memory task when active maintenance of the memorandum was required. This
deficit was partially reversed by drug-â•‰induced inhibition of cyclic AMP signaling,
which is thought to restore aspects of PFC network physiology.
Variants of the delayed response task, referred to in the human literature as “item-â•‰
recognition,” continue to be the work horse of working memory studies that focus
on maintenance of information with minimal explicit demands placed on the addi-
tional executive processes of manipulation, and rapid updating required by N-â•‰back
tasks, for example. As Reuter-â•‰Lorenz and Sylvester (2005) argued, despite the fact
that human age differences in item recognition performance are less pronounced
than on maintenance plus processing tasks, the neural substrates underlying item
recognition in younger and older adults differ in important respects. Specifically,
older adults often show more widespread activation across regions, especially in
prefrontal cortex, that do not reach threshold levels of activation in younger adults
(e.g., Reuter-â•‰Lorenz et al., 2000; Cabeza et al., 2004; Schneider-â•‰Garces et al.,
2010). When overactivation in older adults is more bilateral, it fits the hemispheric
239
asymmetry reduction pattern (Cabeza, 2002; see also Reuter-Lorenz et al., 1999).
In addition, age-related increases are not always found in the non-specialized hemi-
sphere and may extend intra-hemispherically as well (e.g., Spreng, et al., 2010).
Neuroimaging evidence for task-related overactivity in older adults is now well
established based on cross-sectional comparisons in a number of cognitive domains,
but especially for working memory and executive control tasks (Spreng et al., 2010:
Vallesi et al., 2011).
An important discovery over the last ten years comes from studies that paramet-
rically varied working memory load to establish that, compared to younger adults,
older adults tend to overactivate PFC regions at lower loads, and to underactivate
at higher loads, where performance also drops off (Cappell et al., 2010; Schneider-
Garces et al., 2009; Mattay et al., 2006; see also Nagel et al., 2009). Younger adults
also show increased activity with increasing load, suggesting that overactivity is a
typical response to demand, presumably reflecting the recruitment of additional com-
putational resources, and cognitive strategies. Older adults tend to engage these addi-
tional resources at lower levels of demand (e.g., Reuter-Lorenz et al., 1999), which
we have interpreted as a form of compensation for age-related declines in neural effi-
ciency (e.g., Reuter-Lorenz and Lustig, 2005). We have referred to this idea as the
compensation-related utilization of neural circuits hypothesis or CRUNCH (see also
Reuter-Lorenz and Cappell, 2008). By taking into account the memory loads used to
test older and younger adults, CRUNCH can explain some of the conflicting reports
in the working memory literature where older adults were reported to underactivate
(Rypma and D’Esposito, 2000), or overactivate relative to the younger adult samples
(Cabeza et al., 2004). According to CRUNCH, age-related overactivation is associ-
ated with compensation for inefficiency at lower loads, whereas underactivation is
associated with incomplete or inadequate engagement of the task-relevant network
given excessive demand. Furthermore, CRUNCH posits that while overactivation may
provide some benefits, relying on additional resources at low levels of task demands
means that a resource ceiling will be reached more readily, thereby restricting the
range of task demands that a person can respond to effectively. The dynamic range of
activity levels that the brain can assume in response to varying task demands may be a
general and important property of brain function that is indicative of neural efficiency
and brain health. As recent evidence indicates, demand-related activity modulation
appears to decline steadily across the lifespan (Kennedy et al., 2015; see also Park
and Festini, this volume), and excessive overactivity in midlife may be a sign of risk.
Along these lines, Nagel et al. (2011) demonstrated that in a verbal N-back task,
the ability to modulate components of the working memory circuitry (ie. the activation
difference between 1 back and 3 back) was related to task performance: Across age
groups, better modulators were also better performers. Furthermore, better functional
connectivity between prefrontal and parietal regions was also associated with better
task performance for younger and older adults (see also Nagel and Lindenberger,
2015). These findings are generally consistent with CRUNCH, while also calling
attention to an important principle of neurocognitive aging: While compensatory
neural responses may be fruitful in the face of decline, the optimal aging trajectory
is likely to be one in which youth-like activity patterns are maintained into older
age (Nagel and Lindenberger, 2015; Nyberg et al., 2012) especially for relatively
240
low loads, thus preserving the range of task demand to which the brain can respond
(Reuter-Lorenz and Park, 2014).
Nevertheless, there is ongoing debate about how to interpret regions of additional
activity (e.g., Cabeza and Dennis, 2012; Fabiani, 2012; Nyberg, et al., 2010; Barulli
and Stern, 2013) and whether or not the functions served when these regions are
brought on line benefit performance (Cabeza, this volume). As we argued previously
(Reuter-Lorenz and Lustig, 2005), demonstrating that performance is compromised
by temporarily deactivating the overactive regions, using transcranial magnetic stim-
ulation for example, can provide valuable causal evidence for their beneficial role.
Such effects have been documented for long-term memory tasks (e.g., Sole-Padulles
et al., 2006; see Reuter-Lorenz and Cappell (2008) for a review), but to our knowledge
have yet to be established for working memory.
If overactivation of PFC provides beneficial and compensatory support (i.e., scaf-
folding; Park and Reuter-Lorenz, 2009), what might this activity be compensating
for? We, and others (Cabeza and Dennis, 2012; Fabiani et al., 2015; Reuter-Lorenz
and Lustig, 2005) have offered several possibilities, including neural inefficiency
within the task network itself or noisy (dedifferentiated) perceptual representations
(Li et al., 2001). To our knowledge, at the time of this writing, new network-based
fMRI analyses being developed to quantify neural efficiency (Meunier et al., 2014)
have not yet been applied to data from working memory tasks. However, there is evi-
dence to indicate that the representations being maintained in working memory are
less distinctive in older than younger adults (e.g., Carp et al., 2010; Noack et al., 2012;
Payer et al., 2006). Using multivoxel pattern analyses Carp et al. (2010) compared the
distinctiveness of brain activity associated with item-recognition memory for letters
(verbal) or locations (spatial) under varying loads (see Figure 9.1). In accord with
0.25
Young
Old
0.20
Neural distinctiveness
0.15
0.10
0.05
0.00
Low Med High
Memory load
Figure 9.1 Main effect of age during working memory encoding. Older adults showed
decreased distinctiveness between verbal and visuospatial WM tasks in sensory cortex. Right
inferior occipital gyrus (x = 38). is highlighted. Bar graph displays decreased neural distinctive-
ness in older adults in right inferior occipital gyrus. From Carp et al, 2010.
241
other evidence for dedifferentiation with age (e.g., Park et al., 2004) during the encod-
ing and retrieval epochs younger adults showed more distinctive representations than
older adults, especially in perceptual and memory encoding regions, and regardless of
memory load. During the maintenance epoch, however, older adults showed greater
distinctiveness at low loads than young adults, and distinctiveness either decreased
or remained constant with increasing load, whereas activity patterns for verbal
and spatial tasks became more distinguishable with load for younger adults (See
Figures 9.2 and 9.3). This interaction between age and load was most evident in pre-
frontal regions. The authors suggested that these load dependent age differences were
consistent with CRUNCH, and likely to reflect strategy differences, whereby older
adults rely heavily on domain general resources at higher loads, whereas younger
adults engage processes that are more domain-specific (Carp et al., 2010).
What has become more evident in the past ten years is that a one-size-fits-all inter-
pretation will be inadequate to explain the results from the wide variety of tasks,
approaches to neuroimaging analyses, and cross-sectional versus longitudinal inves-
tigations of working memory and executive functions. Different age-related patterns
may emerge depending on which epoch of the working memory task is examined, the
performance level of the participants, the domain of the memoranda, and the specific
processing demands of the task (e.g., Toepper et al., 2014; Logie et al., 2015; Lustig
and Jantz, in press). Moreover, most fMRI studies of aging to date have not incorpo-
rated measures that can adjust for known age differences in vascular reactivity that can
affect blood oxygenation level dependent (BOLD) responses. Without measures of
cerebrovascular reactivity, the extent to which age differences in task-evoked BOLD
reflect differences in neural activity can be under or over estimated (Lu et al., 2011;
Tsvetanov et al., 2015), and as more efficient methods are developed they can be
more widely incorporated into future imaging studies of neurocognitive aging. The
0.25
Young
Old
0.20
0.15
0.10
0.05
0.00
−0.05
Low Mid High

Memory load
Figure 9.2 Age group by load interaction during working memory encoding. Neural distinc-
tiveness increased with load in younger adults but decreased with load or varied minimally in
older adults. Right middle frontal gyrus and left superior frontal gyrus are highlighted, and the
corresponding distinctiveness scores are displayed in the bar graphs. From Carp et al., 2010.
242
0.20 Young
Old
0.15
0.10
0.05
0.00
–0.05
Low Med High

Memory load
Figure 9.3â•‡ Age group by load interaction during working memory encoding. Neural distinc-
tiveness increased with load in younger adults but decreased with load or varied minimally in
older adults. Right middle frontal gyrus and left superior frontal gyrus are highlighted, and the
corresponding distinctiveness scores are displayed in the bar graphs. From Carp et al., 2010.
following sections offer a brief overview of some newer paradigms and associated
developments that have emerged over the past ten years in the cognitive neuroscience
of aging and working memory.
Visual Working Memory, Change Detection, and Neurocognitive Aging
In 2009, Logie and Maylor reported the results from a very large and impressive data-
set obtained via on-â•‰line administration of an assortment of working memory measures
including digit span, sentence span, visual pattern span, and feature binding. Their
sample of over 95,000 participants ranged in age from 18 to 90. While all indices
declined over the lifespan, the most dramatic effects were evident in visual pattern
span and feature binding (see also Brockmole and Logie, 2013). The disproportionate
age-â•‰related vulnerability of the visual working memory domain corroborates the work
of Hale and colleagues (e.g., see Hale et al., 2011 for a recent review) and underscores
the notion that working memory for different types of information entails somewhat
different mechanisms (e.g., representational buffers, maintenance/â•‰manipulation pro-
cesses, or both) that are differentially affected by age. Moreover, the decline in feature
binding extends to working memory an age-â•‰related deficit that was initially recog-
nized and now well â•‰documented in long-â•‰term memory (Chalfonte and Johnson, 1996;
Naveh-â•‰Benjamin, 2000; Sander, et al., 2011).
An increased focus on age effects in visual memory and feature binding stems in
part from the popularity of the change detection task developed by Luck and Vogel
(1997, 2013), which requires participants to judge the similarity of two briefly pre-
sented arrays of colored shapes. By varying the number of items to be retained and the
nature of the change between two consecutively presented displays, performance on
243
this task can provide an index of working memory capacity and the ability to integrate
features, such as object shape, color and location. By varying the response require-
ments, for example by use of a color wheel to indicate the hue of a to-be-remembered
display item, the precision of WM can be assessed (e.g., Bays et al., 2011). Like
other measures of working memory, change detection performance has been shown to
correlate with fluid intelligence (Luck and Vogel, 2013), and its substrates have been
linked to an electrophysiological measure referred to as contralateral delay activity,
or CDA (Vogel and Machizawa, 2004). The CDA is a sustained signal that is evident
during the delay interval during which participants must remember visual informa-
tion from one hemifield of a bilateral display. The amplitude of this potential varies
with memory load, reaching an asymptotic level that corresponds with the individu-
al’s memory capacity. Because the CDA is also modulated by memory instructions
and the requirement to ignore distracting information, and is disrupted by lateralized
prefrontal lesions, this electrophysiological signature is thought to reflect the contri-
butions of both bottom-up and top-down control aspects of visual working memory
(e.g., Sander et al., 2012; Ko et al., 2014).
Despite the relative simplicity of the change detection task, the evidence for age-
related deficits is not fully consistent. Estimates of visual working memory capac-
ity are generally lower in older than younger adults (Sander et al., 2012; Ko et al.,
2014). In contrast, specific deficits in binding are not always observed in healthy
older adults, but are robust in individuals with Alzheimer’s disease and their cogni-
tively normal at-risk relatives, suggesting that binding impairments may be a specific
and sensitive marker of the disease (Brockmole et al., 2008; Parra et al., 2009; 2010;
however see Cowan et al., 2006). Nevertheless, recent work combining psychophysi-
cal and modeling approaches (Peich, et al., 2013) has shown that for three-item loads
requiring the retention of line orientation, color and location the incidence of binding
errors increases significantly with increasing age from early to late adulthood (19–77
years of age). Moreover, the precision with which the features of the memoranda
could be reproduced using a continuous measure of recall fidelity indicated that even
with a set size of 1 item, memory resolution decreased with age, an effect that was
even more pronounced with a higher memory load (Peich et al., 2013; Noack et al.,
2012). Because the precision of working memory representations can be influenced
by priorities set in response to cues or monetary incentives, (Klyszejko et al., 2014),
age-related deficiencies in representational precision may involve attentional control
processes in addition to data-driven processing decrements due to age.
Indeed, despite the relative consistency of evidence indicating lower visual work-
ing memory capacity in older adults, the neurophysiological bases for this age effect is
not yet understood. Interestingly, the relationship between the CDA and memory load
appears to differ in younger and older adults in that age differences in behavioral per-
formance are not consistently reflected in age differences in the CDA (Sander et al.,
2011; Ko et al., 2014; see also Jost et al., 2011). One possible explanation for this
dissociation is that the CDA may reflect the quantity but not the quality of working
memory representations, and that as the study by Peich et al. (2013) discussed above
indicates, resolution appears to be compromised in older adults.
In summary, there is increasing interest in using visual change detection to under-
stand the components of visual working memory, its capacity limitations and the
244
effects of normal aging. However, limited progress has been made to date toward
understanding the neural underpinnings of age declines in capacity. Current debates
about whether visual memory is limited by slots, resources, or some other properties
(Suchow et al., 2014) may also be informed by future studies of older adults, espe-
cially when combined with brain imaging methodologies that better pinpoint the locus
of age-â•‰related change.
Inhibition, Attention, and Neurocognitive Aging of Working Memory
The change-â•‰detection studies described in the previous section, especially in their

consideration of the question of the quantity versus quality of representations, bear
on the influential though somewhat controversial inhibitory deficit hypothesis. This
hypothesis, first put forth by Hasher and Zacks in 1988, has since undergone several
updates to include multiple aspects of inhibition and the impact of mediating fac-
tors such as circadian arousal (e.g., Lustig, Hasher, and Zacks, 2007; Hasher, Zacks,
and May, 1999; see Weeks and Hasher, 2014 for a recent review). Its central idea
is the proposition that age differences in working memory performance (as well as
other areas of cognition) arise not because older adults’ working memory capacity is
reduced, but because it is overloaded. That is, older adults are thought to have defi-
cient inhibition of irrelevant information, which may then compete with target infor-
mation, resulting in slower and/â•‰or more inaccurate processing.
The inhibitory deficit view would be more compatible with the “resolution” inter-
pretation of the change-â•‰detection studies described above: Reduced inhibitory func-
tion should result in less precise/â•‰more cluttered representations, rather than directly
reducing the number of “slots.” Several recent findings using variations of the change
detection paradigm appear consistent with this perspective. For example, Sander et
al. (2011) tested children, young adults, and older adults on both the standard CDA
paradigm and a modified version that included irrelevant distractors (e.g., circles in
the visual display that might also change color, whereas the change detection task
was only to be performed on the colored squares). As expected, young adults outper-
formed older adults in the standard condition. The more interesting results were found
in the version of the task using distractors. In this case, children and older adults were
more impaired by the distractors than were young adults. Furthermore, model-â•‰derived
estimates indicated that compared to young adults, these two groups devoted twice as
much processing capacity toward the distractors.
One question that arises is whether older adults have a deficit in inhibition per se, or
whether other processing deficits such as reductions in processing speed lead to such
effects. As we will discuss further below, the answer to this question is complicated by
the fact that the term “inhibition” has been applied to quite a number of different pro-
cesses operating at different neurocognitive levels, and there is no particular reason to
believe that (for example) center-â•‰surround inhibition operating at the level of visual
columns represents or is subject to the same influences as response inhibition at the
behavioral level in a stop-â•‰signal task. At the level of item representations in perception
and working memory, a number of ERP studies have observed that older adults have
reduced indices of distractor suppression at early but not late processing stages (e.g.,
245
Fabiani and Gratton, 2005; Fabiani, Low, Wee, Sable, and Gratton, 2006; Gazzaley
et al., 2008). This might be interpreted as a result of slower processing speed on the
part of older adults (e.g., Gazzaley et al., 2008); alternatively, it may reflect something
like the shifts from proactive, early-stage control to reactive, late-stage control pro-
posed by Braver and colleagues (Braver and Barch, 2002; Braver et al., 2007).
The results of Sander et al. (2011) may be more consistent with the latter interpre-
tation, as there were no differential age effects as a result of increased exposure time
at encoding. However, while the degree of improvement across increases in exposure
times was roughly parallel for the three age groups, it was also the case that young
adults performed better overall, and only young adults had a complete elimination of
distractor effects at the longest exposure times. As the authors suggest, the reduced
overall performance of older adults might reflect impaired binding of stimulus-feature
representations, whereas the remaining distractor effects reflected impaired (rather
than merely slower) inhibitory control. In a related study, Jost et al. (2011) found that
older adults showed reduced early-stage filtering of distractors as measured by the
CDA, but did not show slowing in earlier ERP components (N1 and P1). Thus, the
age difference appeared to be specific to the filtering or inhibition of distractors, rather
than a general age-related decline in processing speed.
We recently suggested that this early filtering deficit, often combined with poten-
tially compensatory increases in processing at later stages, may help to resolve what
initially appear to be inconsistent findings across the different types of task often used
to assess potentially age differences in inhibitory function and interference proneness
(Lustig and Jantz, 2015). In particular, based on a meta-analysis of behavioral find-
ings from a set of putative inhibition tasks using response-time measures (inhibition
of return, negative priming, flanker, Stroop, and reading with distraction), Verhaeghen
(2011) found that only reading with distraction had reliable age effects after control-
ling for generalized slowing, and on this basis concluded that age deficits in inhibi-
tory processing (as well as many other aspects of executive processing) were “greatly
exaggerated.” This conclusion stands in contrast to findings that older adults show
increased interference and vulnerability to distraction on a variety of working memory
tasks including various measures of working memory span (Lustig et al., 2001; May
et al., 1999; Rowe et al., 2008), the change-detection visual working memory studies
described above, recent probes, and n-back (Loosli et al., 2014).
This discrepancy in findings might arise at least in part because the tasks included
in the Verhaeghen meta-analysis relied primarily on behavioral measures (specifically,
response time) from tasks that can be performed within the focus of attention, whereas
the working-memory measures required retrieval into the focus. Specifically, we sug-
gested that for tasks performed within the focus, age differences in inhibitory function
might be more evident in brain activation patterns than in behavior, as late-stage con-
trol processes intervene to rescue performance after failure of the early filtering mech-
anisms as described above. For tasks requiring retrieval into the focus, less distinctive
representations (Carp et al., 2010; Park et al., 2012; St.-Laurent et al., 2014) create an
increased opportunity for nontarget items to compete with targets for retrieval into the
focus, creating interference (Lustig and Jantz, 2015).
A number of behavioral and neuroimaging findings support this proposed sep-
aration. For example, in a relatively large sample of young and older adults who
246
all completed the same tasks, Pettigrew and Martin (2014) found that tasks requir-
ing retrieval (recent negatives, cued recall/directed forgetting, and release from PI)
showed consistent age differences in interference, whereas age differences in tasks
that could be completed within the focus of attention (flanker, picture-word inter-
ference, Stroop) were smaller and less consistent. Bearing even more directly on the
question of items within versus outside of the focus, Oztekin et al. (2012) used a
speed–accuracy tradeoff procedure in combination with the recent negatives task, and
found that young and older adults showed similar retrieval dynamics for the most
recently-presented item (assumed to be within the focus). In contrast, older adults
showed slower retrieval for items out of the focus, as well as slower and less efficient
engagement of interference-resolution processes needed to avoid false alarms. Similar
rates of retrieval into the focus help rule out a general processing-speed deficit; instead
older adults appear to have specific impairment in the controlled processes needed to
retrieve items into the focus and reduce interference.
Only a few neuroimaging studies of age differences in interference resolution
had been published at the time of the last iteration of this chapter. The number
has increased substantially since then, with Turner and Spreng (2012) including
13 in a recent meta-analysis comparing PET and fMRI studies of age differences
in inhibition versus those (n = 19) focusing on age differences related to working
memory load. Notably, most of the “inhibition” studies used tasks that we would
describe as being within the focus of attention (Stroop, flanker, etc.), whereas
the “working memory” tasks all required retrieval. The inhibition tasks showed
what Turner and Spreng called a “young plus” pattern—increased recruitment of
the same regions (right inferior frontal gyrus, left medial superior frontal gyrus,
and presupplementary motor area) recruited by young adults. In contrast, age dif-
ferences in the working-memory retrieval tasks were manifest as more bilateral
recruitment, as described above, as well as somewhat of a shift toward more ante-
rior regions (c.f., Davis et al., 2008).
The Turner and Spreng (2012) results suggest that for items in the focus, age dif-
ferences in interference often manifest as increased activation. Approximately half of
the studies included in their dataset had matched performance across young and old
adult groups; age differences in performance in several others might also have been
eliminated by controlling for slowing. This increased activation seems initially coun-
terintuitive in light of the general hypothesis that age-related inhibitory deficits are
related to prefrontal declines, raising the question of whether performance-processing
relations in this domain are also subject to a CRUNCH-like pattern similar to that
described for varying levels of working memory load as described above. This is a
difficult hypothesis to test, as it is not obvious that one can operationalize the param-
eterization of inhibitory-processing demand in the same straightforward way as oper-
ationalizing storage/retrieval demands (i.e., via number of items). To our knowledge,
the closest approximation to such a test is a recent study by Sebastian et al. (2013),
who tested adults age 22–77 on three tasks chosen to vary inhibitory demand: Go/No-
Go (inhibit predominant response only on rare trials, relatively low demand); Simon
(overcome habitual resonse, medium demand); stop-signal (cancel response in prog-
ress; high demand). Age showed a positive relationship with prefrontal activation in
the low-and medium-demand tasks, but a negative relationship in the high-demand
247
task. Furthermore, only in stop-signal was age associated with more errors as well as
specific slowing in high-demand trials even after controlling for baseline response
times; this increase errors suggests a breakdown in control rather than merely more
effort in its deployment. Although interpretation is complicated by other possible dif-
ferences between the tasks, at least in the broad view the pattern of results is consistent
with the CRUNCH idea of relatively preserved performance but higher activation at
low levels of demand, and deficits in both performance and (prefrontal) activation at
higher levels.
Dennis and Cabeza (2008) proposed an alternative “expanded inhibitory deficit the-
ory” of reduced activity by older adults in inhibitory control regions (possibly accom-
panied by increased activity in alternative inhibitory control regions), but increased
activity (disinhibition) in the regions that are the targets of that control. Notably, while
the results of the Turner and Spreng (2012) meta-analysis emphasizing tasks per-
formed within the focus of attention clearly do not support this hypothesis, studies
that involve working-memory retrieval are more consistent (e.g., Jonides et al., 2000;
Gazzaley et al., 2005). However, even most of these reports appear to show one of the
patterns (decreased activation in prefrontal control regions or increased activation in
regions representing irrelevant information); the presence of both patterns in the same
study is less common. Instead, the pattern predicted by Dennis and Cabeza appears to
play out most obviously in the interplay between prefrontal control regions and those
thought to be involved in the so-called “default network” typically associated with
internally directed, off-task thoughts. Of particular interest here, Stevens et al. (2008)
found that compared to young adults, older adults showed greater increases in regions
associated with auditory processing (which they suggested reflected greater attention
to irrelevant scanner noise) for subsequently forgotten items, as well as greater con-
nectivity between those auditory regions and default-network regions. These patterns,
combined with performance, suggested a failure to inhibit distraction from the scan-
ner noise and thus an impairment of encoding.
In summary, the data suggest that there are differences in inhibitory processing
that can be differentiated from age differences in related processes such as processing
speed or the storage/retrieval components of working memory. Furthermore, there
are distinctions within inhibitory processes that must also be taken into account. In
our discussion here we have highlighted the distinction between operations on items
within the focus versus operations that require retrieval into the focus (see Pettigrew
et al. for a similar conceptualization); a related but potentially important division is
that between whether inhibition acts to restrict access to the focus, to delete items
from the activated portion of long-term memory that competes for that access, or
restraint of strong and habitual responses (e.g., May et al., 1999). In behavioral per-
formance, these processes interact and deficits in one process may increase the load
on later processes attempting to compensate—e.g., in Stroop failures to proactively
prohibit access of inappropriate lexical information into the focus may then increase
the downstream load on reactive processes of restraining the response to that infor-
mation. Since the last iteration of this chapter, neuroimaging evidence has played an
important role in elucidating these distinctions, and we predict that both fMRI and
ERP evidence—the latter providing critical insights into timing—will be even more
prominent in its next version.
248
Finally, along with the distinctions (within/outside of the focus; access/deletion/

suppression) described here, it is important to keep in mind that “inhibition” is a term
used at many levels of analysis. In general, inhibitory deficit theory as proposed by
Hasher and colleagues is intended to pertain primarily to the controlled aspects of
attention. Interestingly, however, other levels of the system likely affect the operation
of higher-level processes. For example, at the neurotransmitter level, age deficits in
GABA center-surround inhibition likely contribute to the reduced distinction of sen-
sory representations. Reduced distinctiveness between target and irrelevant informa-
tion then increases interference bottom-up, making suppression of the irrelevant items
more difficult (see discussion by Fabiani et al., in press). A better understanding of
how both inter and intra-individual differences at lower levels of analysis play into
the differences seen at the behavioral and neuroimaging levels is likely to become
increasingly important over the next 10 years, as there is an increasing awareness
of how factors such as genetics and circadian arousal influence both (c.f., Anderson
et al., 2014).
Risk Factors, Lifestyle, and Modifiability of Working Memory

and Executive Functions
Understanding the factors that may influence, maintain, and even improve working
memory is of particular importance because of the strong relationships between work-
ing memory and success in many real-world tasks, including those related to work-
place productivity and maintaining an independent lifestyle. Worldwide demographic
trends make it abundantly clear that declining birthrates are now accompanied by
increased longevity (Source: United Nations, Department of Economic and Social
Affairs). This means that people are spending more of their lifetime in older age, and
the proportion of older people in the world is increasing. Consequently, the neces-
sity for high quality of life well into one’s golden years couldn’t be more pressing.
Also clear is that the ability to conduct complex everyday activities, i.e., the activi-
ties of daily living (ADL), is closely related to quality of life and the ability to live
independently. ADLs can be measured in a variety of ways including observational
sampling of behaviors, self-report, and informant/caregiver report. Such activities
often include financial management, managing a household, medical and medication
concerns, among other things. Both cross-sectional and longitudinal studies have doc-
umented that the level of executive functioning is one critical determinant of the how
effectively older adults can contend with such complex activities in their daily lives
(e.g., Cahn-Weiner et al., 2002; Farias et al., 2009; Tomaszewski et al., 2009; Vaughan
and Giovannello, 2010; see Reuter-Lorenz et al., 2015).
Longitudinal evidence has shown, for example, that a composite measure of exec-
utive function based on verbal and spatial backward span tasks, perseveration and flu-
ency measures was independently associated with 5-year change in informant-rated
ADL measures (Farias et al., 2009) in a sample of older age adults heterogeneous with
respect to cognitive status. In a cross-sectional study of cognitively healthy commu-
nity-dwelling older adults, Vaughan and Giovanello (2010) demonstrated that exper-
imental measures of task switching, more so than WM updating and inhibition, were
249
associated with performance-based ADL measures. Furthermore, there is increasing

evidence that levels of working memory and executive function can also modify the
trajectories of cognitive decline in healthy people and in those at risk for dementia.
For example, several studies indicate that the potentially adverse cognitive effects
of declining neural structure, as indicated by measures of white matter integrity and
brain volumes (e.g., hippocampus), can be mitigated by higher levels of executive
function (Brickman et al., 2006; Chang et al., 2010). Conversely, low levels of execu-
tive function can predict conversion to AD (Rapp and Reischies, 2005) and have been
associated with greater declines in other cognitive domains such as episodic memory
(Parks et al., 2011). Although the nature of these structure–function and function–
function relationships is complex and the direction of causality uncertain, findings
such as these underscore the fundamental and practical importance of maintaining
high levels of working memory and executive functions into older age.
Genetic, experiential and lifestyle factors have a significant impact on the level of
WM and executive function throughout the lifespan. The APOE ε4 allele is a well-
known genetic risk factor for memory decline associated with Alzheimer’s disease. Yet,
recent work by Bender and Raz (2012) indicates an interactive role of the ε4 allele in
executive function as well. They found that even in carriers of this allele whose blood
pressure was characterized as normotensive, age differences in working memory were
influenced by increased systolic blood pressure among other variables, indicating the
combined influence genetic and cardiovascular risk factors influence on executive
function in older age. Greater recruitment of prefrontal regions in middle-age ε4 car-
riers compared to non-carriers, has also been reported suggesting that aspects of aging
may be accelerated in people with this genetic risk (Evans et al., 2014).
Dopamine levels play an important role in executive functions and working mem-
ory (e.g., D’Esposito and Postle, 2015) and a number of genetic factors are known to
contribute to dopamine signaling. While a review of genetic influences on executive
function is beyond the score of this chapter, new evidence suggests that genetic mod-
ulation of dopaminergic effects on PFC functioning is not age-invariant (Nyberg et al.,
2014) and that some genetic effects may become more pronounced with increasing
age (Nagel et al., 2008).
Working memory ability and executive functions appear to be especially sensi-
tive to life course experiential factors, including those captured in such constructs
as allostatic load. Rodent work has demonstrated adverse effects of both acute
and chronic stress on the neurophysiology of PFC regions, and on PFC-dependent
behaviors, including measures of working memory, in developing rats that can
carry over into adulthood (see McEwan and Morrison, 2013 for a review). While
early in life, some recovery from the microstructural consequences of stress can
be seen after periods of rest, these forms of plasticity diminish with increasing
age (McEwan and Morrison, 2013). The adverse effects of high allostatic load are
evident in humans as well. A recent report based on over 1000 people from age 25
to 74, used multiple biomarker indices of stress and allostatic load to measure its
association with episodic memory and executive function across the adult lifespan
(Karlamanga et al., 2014). The authors documented a strong and inverse associa-
tion between allostatic load and cognitive functioning, which was present at both
middle age and late adulthood.
250
While genetics and some experiential factors are not (currently) modifiable, life-
style factors can be better controlled by the individual. The three factors consistently
shown to influence aging outcomes are cognitive/intellectual engagement, physical
activity and social engagement. Although the beneficial effects of cardiovascular fit-
ness and on prefrontal structure and function has been well documented (e.g., Hillman
et al., 2008; Weinstein et al., 2012), the combined influence of these factors on exec-
utive function per se has not been widely studied until recently. In a large-scale lon-
gitudinal study, de Frias and colleagues (deFrias et al., 2014) examined the effects
of an active life style (across the three above domains) on initial cognitive status and
on cognitive change over a 4.5-year period. Older adults who were classified as hav-
ing some cognitive impairments, but led actively engaged lives performed better on
measures of executive function than those who did not. Furthermore, longitudinally
both physical activity and intellectual engagement led to improved or stable executive
function, depending on initial cognitive status. These results are especially important
given the role of executive function and working memory in maintaining life-long
independence and quality of life.
The impact of experiential and lifestyle factors on working memory has also
led to an increased awareness within the scientific community of the need to dis-
tinguish between age-related effects measured between groups, which may lead
to large contributions from cohort effects, and aging effects as measured within
individuals over time in longitudinal studies. (See Rugg chapter for a more detailed
treatment of these issues.) The Seattle Longitudinal Study has consistently found
cohort differences on a number of measures related to working memory: positive
trends are found for inductive reasoning and spatial orientation, but more negative
trends in recent cohorts for word fluency (Schaie, Willis, and Caskie, 2004; see
also Gerstorf et al. 2015 for a report of large secular trends in Digit Symbol perfor-
mance). The cohort differences overall serve as an important reminder that mean
group differences may not predict the course of lifespan change, and the different
patterns for different abilities related to working memory reinforce that it is com-
plex, multidimensional construct and that close attention to operational definition
is required when comparing across studies.
The issue of cohort effects impacts behavioral as well as neuroimaging studies; neu-
roimaging (especially BOLD fMRI studies) also face the more specialized issue of age
and individual cerebrovascular differences/neurovascular coupling. These issues were
an area of concern in the early days of aging fMRI studies (e.g., Buckner et al., 2000;
D’Esposito et al., 1999; Huettel et al., 2001) but then somewhat neglected until more
recently (e.g., Fabiani et al., 2014a, 2014b; Lu et al., 2011; Tsvetanov et al., 2015).
Complicating matters, cardiovascular health has effects both on cognitive ability, espe-
cially the executive functions of working memory, and on the measurement of neu-
ral activity using BOLD, and careful analysis will be needed to disentangle these. The
use of parametric designs, rather than simple reliance on main effects of age group, in
many working memory studies may provide some insurance against these potential con-
founds. Methodological advances (e.g., corrections using hypercapnia, e.g., Tsvetanov
et al., 2015) in fMRI as well as converging evidence from other neuroimaging methods
(EEG/ERP, fNIRS, optical imaging) and neurostimulation techniques (TMS, TDCS)
that can provide more causal tests will be important for clarifying these issues.
251
Conclusion: Working Memory as a Contributor to and Beneficiary

of Scaffolding in Aging and Lifespan Development
Executive processes may be among the most flexible and adaptive of the human
cognitive repertoire (Miller and Cohen, 2001). Their flexibility, along with the
dynamic support provided by on-line representations active in working memory,
figure prominently in the Scaffolding Theory of Aging Cognition (Park and Reuter-
Lorenz, 2009), and the more recent lifespan, longitudinal version of this account,
STAC-r (Reuter-Lorenz and Park, 2014). These models propose that scaffolding is
a form of neuro-computational support provided to primary or newly established
task networks when new skills are acquired over the course of early development.
In adulthood, declining efficiency, noisy representations, and other factors can
contribute to the inability of primary task networks to meet task demands unless
additional resources—the scaffolding—are recruited as compensation. However,
this compensation comes at the cost of reducing the availability of those other
resources (especially attention/executive function) at higher levels of demand. As
Figure 9.4 illustrates, lifelong influences, including some reviewed in the previous
section, may enrich or deplete neural resources, affecting brain structure and func-
tion, as well as the potential for compensatory scaffolding (see Reuter-Lorenz and
Park, 2014 for more details). The potential for the brain to respond effectively to
Scaffolding Theory of Aging and Cognition-Revised (STAC-r)

Biological
Aging
Neural Brain Level of

Enrich Structure Cognition
Life Compensatory
Course Scaffolding
Rate of
Neural Brain
Cognitive
Deplete Function
Change
Figure 9.4 A simplified schematic of the conceptual model of the Scaffolding Theory of Aging
and Cognition-Revised (STAC-r). The model displays the life course–longitudinal influences of
neural enriching and neural depleting factors on brain structure and brain function which ulti-
mately determine the level of cognitive function and rate of change over time. These factors also
influence the potential for compensatory scaffolding, which Reuter-Lorenz and Park (2014)
attribute in part to adaptive recruitment of executive functions, although other neural circuitry
can provide computational support also, depending on cognitive demands. For a full description
of the model, enriching/depleting factors, etc., see Reuter-Lorenz and Park (2014).
252
the biological challenges of aging is determined in part by lifelong factors that also
influence executive processes. Identifying the life style, educational, and health
factors that can optimize executive function and working memory, along with
interventions that can remediate their decline, are crucial items on the agenda for
future research on the cognitive neuroscience of aging.
References
Anderson JA, Campbell KL, Amer T, Grady CL, Hasher L (2014). Timing is everything: Age
differences in the cognitive control network are modulated by time of day. Psychology
and Aging.
Awh E, Jonides J (2001) Overlapping mechanisms of attention and spatial working memory.
Trends in Cognitive Sciences 5:119–â•‰126.
Baddeley AD, Hitch G (1974). Working memory. In G.H. Bower (Ed.), The psychol-
ogy of learning and motivation: Advances in research and theory (Vol. 8, pp. 47–â•‰89).
New York: Academic Press.
Bartus RT, Fleming D, Johnson HR (1978) Aging in the rhesus monkey: debilitating effects on
short-â•‰term memory. J Gerontol 33:858–871.
Barulli D, Stern Y (2013) Efficiency, capacity, compensation, maintenance, plasticity: emerg-
ing concepts in cognitive reserve. Trends in Cognitive Sciences 17:502–â•‰509.
Bays PM, Wu EY, Husain M (2011) Storage and binding of object features in visual working
memory. Neuropsychologia 49:1622–â•‰1631.
Bender AR, Raz N (2012) Age-â•‰related differences in memory and executive functions in healthy
APOE epsilon 4 carriers: The contribution of individual differences in prefrontal volumes
and systolic blood pressure. Neuropsychologia 50:704–â•‰714.
Bizon JL, Foster TC, Alexander GE, Glisky EL (2012) Characterizing cognitive aging of work-
ing memory and executive functionin animal models. Frontiers in Aging Neuroscience
September 12, 4:19.
Braver TS, Barch DM (2002). A theory of cognitive control, aging cognition, and neuromodu-
lation. Neuroscience and Biobehavioral Reviews, 26(7): 809–â•‰817.
Braver TS, Gray JR, Burgess GC (2007) Explaining the Many Varieties of Working Memory
Variation: Dual Mechanisms of Cognitive Control (Chapter 4). In Variation in Working
Memory. Editors: Conway, A.R.A., Jarrold, C., Kane, M.J., Miyake, A., Towse, J. N.
Oxford Univerity Press
Brickman AM, Zimmerman ME, Paul RH, Grieve SM, Tate DF, Cohen RA, et al. (2006)
Regional white matter and neuropsychological functioning across the adult lifespan.
Biological Psychiatry 60:444–â•‰453.
Brockmole JR, Logie RH (2013) Age-â•‰related change in visual working memory: a study of
55,753 participants aged 8–â•‰75. Frontiers in Psychology 4:12.
Brockmole JR, Parra MA, Della Sala S, Logie RH (2008) Do binding deficits account
for age-â•‰related decline in visual working memory? Psychonomic Bulletin & Review
15:543–â•‰547.
Buckner RL, Snyder AZ, Sanders AL, Raichle ME, Morris JC (2000) Functional brain imaging
of young, nondemented, and demented older adults. Journal of Cognitive Neurscience 12:
24–â•‰34.
Cabeza R (2002) Hemispheric asymmetry reduction in older adults: The HAROLD model.
Psychology and Aging 17:85–â•‰100.
Cabeza R (this volume). Interpreting age-â•‰related activation differences.
253
Cabeza R, Daselaar, SM., Dolcos, F.,Prince, SE, Budde, M., Nyberg, L. (2004) Task-independent
and task-specific age effects on brain activity during working memory, Visual Attention
and Episodic Retrieval. Cerebral Cortex 14:364–375.
Cabeza R, Dennis NA (2012) Frontal lobes and aging: deterioration and compensation.
In: Frontal Lobes (Stuss DT, Knight RT, eds), pp. 628–655. New York: Oxford.
Cabeza R, Nyberg L, Park D (2005) The cognitive neuroscience of aging: linking cognitive and
cerebral aging. New York: Oxford University Press.
Cahn-Weiner DA, Boyle PA, Malloy PF (2002) Tests of executive function predict instru-
mental activities of daily living in community- dwelling older individuals. Applied
Cappell KA, Gmeindl L, Reuter-Lorenz PA (2010) Age differences in prefontal recruitment
during verbal working memory maintenance depend on memory load. Cortex 46:462–473.
Carp J, Gmeindl L, Reuter- Lorenz PA (2010) Age differences in the neural representa-
tion of working memory revealed by multi-voxel pattern analysis. Frontiers in Human
Neuroscience 4:217.
Chalfonte BL, Johnson MK (1996) Feature memory and binding in young and older adults.
Mem Cognit 24:403–416.
Chang Y-L, Jacobson MW, Fennema-Notestine C, Hagler DJ, Jr., Jennings RG, Dale AM,
McEvoy LK, Alzheimer’s Dis Neuroimaging I (2010) Level of executive function influ-
ences verbal memory in amnestic mild cognitive impairment and predicts prefrontal and
posterior cingulate thickness. Cerebral Cortex 20:1305–1313.
Charlton RA, Morris RG (2015) Associations between working memory and white matter
integrity in normal ageing. In Working Memory and Ageing (Logie, RH and Morris, RG,
eds.) pp. 97–126. New York: Psychology Press:97–128.
Cowan N (2001) The magical number 4 in short-term memory: A reconsideration of mental
storage capacity. Behavioral and Brain Sciences 24: 87–114.
Cowan N, Naveh-Benjamin M Kilk, A, Saults JS (2006) Lifespan development of working
memory: When is feature binding difficult? Developmental Psychology 42:1089–1102.
D’Esposito M, Postle BR (2015) The cognitive neuroscience of working memory. Annual
Review of Psychology 66:115–142.
D’Esposito M, Zarahn E, Aguirre GK, Rypma B (1999) The effect of normal aging on the
coupling of neural activity to the bold hemodynamic response. Neuroimage 10(1):6–14.
Davis S, Dennis NA, Fleck MS, Cabeza R (2008) Que PASA? The posterior-anterior shift in
Dennis NA, Cabeza R (2008) Neuroimaging of healthy cognitive aging. The handbook of aging
and cognition, 3:1–54.
de Frias CM, Dixon RA (2014) Lifestyle engagement affects cognitive status differences and
trajectories on executive functions in older adults. Archives of Clinical Neuropsychology
29:16–25.
Evans S, Dowell NG, Tabet N, Tofts PS, King SL, Rusted JM (2014) Cognitive and neural sig-
natures of the APOE E4 allele in mid-aged adults. Neurobiology of Aging 35:1615–1623.
Fabiani M (2012) It was the best of times, it was the worst of times: A psychophysiologist’s
view of cognitive aging. Psychophysiology 49:283–304.
Fabiani M, Gratton G (2005) Electrophysiological and optical measures of cognitive aging.
In: Cognitive neuroscience of aging: Linking cognitive and cerebral aging. (Cabeza R,
Nyberg L, Park D, eds), pp. 85–106. New York, NY US: Oxford University Press.
Fabiani M, Low KA, Wee E, Sable JJ, Gratton G (2006) Reduced suppression or labile mem-
ory? Mechanisms of inefficient filtering of irrelevant information in older adults. Journal
of Cognitive Neuroscience 18(4): 637–650.
254
Fabiani M, Zimmerman B, Gratton G (2015). Working memory and aging: A review. In P.

Joliceur and C. Lefebvre (Eds) Attention & Performance XXV: Mechanism of sensory
working memory. (pp. 131–148) Elsevier.
Fabiani Monica, Gordon, Brian A, Maclin, Edward L, Pearson, Melanie A, Brumback-Peltz,
Carrie R, Low, Kathy A, et al. (2014a) Neurovascular coupling in normal aging: A com-
bined optical, ERP and fMRI study. Neuroimage 85:592–607.
Fabiani Monica, Low, Kathy A, Tan, Chin-Hong, Zimmerman, Benjamin, Fletcher, Mark
A, Schneider-Garces, Nils, et al. (2014b) Taking the pulse of aging: Mapping pulse
pressure and elasticity in cerebral arteries with optical methods. Psychophysiology
51(11):1072–1088.
Farias ST, Murgas D, Reed BR, Harvey D, and DeCarli C (2009) Progression of mild cognitive
impairment to dementia in clinic-vs community-based cohorts. Archives of Neurology
66:1151–1157.
Fegen D, Buchsbaum BR, D’Esposito M (2015) The effect of rehearsal rate and memory load
on verbal working memory. NeuroImage 105:120–131.
Fjell AM, McEvoy L, Holland D, Dale AM, Walhovd KB, Alzheimer’s Dis Neuroimaging I
(2014) What is normal in normal aging? Effects of aging, amyloid and Alzheimer’s disease
on the cerebral cortex and the hippocampus. Progress in Neurobiology 117:20–40.
Fischl B, Dale AM, Walhovd KB (2009) High Consistency of Regional Cortical Thinning
in Aging across Multiple Samples. Cerebral Cortex 19:2001–2012.
Gazzaley A, Clapp W, Kelley J, McEvoy K, Knight RT, D’Esposito M (2008) Age-related
top-down suppression deficit in the early stages of cortical visual memory processing.
Proceedings of the National Academy of Sciences 105(35):13122–13126.
Gazzaley, A, Cooney, JW, Rissman, J, D’Esposito, M, (2005). Top-down suppression deficit under-
lies working memory impairment in normal aging. Nature Neuroscience, 8: 1298–1300.
Gerstorf D, Hueluer G, Drewelies J, Eibich P, Duezel S, Denmuth I, et al. (2015) Secular
changes in late-life cognition and well-being: Towards a long bright future with a short
brisk ending. Psychology and Aging. http://dx.doi.org/10.1037/pag0000016.
Goldman-Rakic PS (1987) Circuitry of primate prefrontal cortex and regulation of behavior by
representational memory. In: Handbook of Physiology: The Nervous System (Plum F, ed),
pp. 373–417). Bethesda, MD American Physiological Society.
Hale S, Rose NS, Myerson J, Strube MJ, Sommers M, Tye-Murray N, Spehar B (2011) The struc-
ture of working memory abilities across the adult life span. Psychology and Aging 26:92–110.
Hara Y, PR R, JH M (2012) Neuronal and morphological bases of cognitive decline in aged
rhesus monkeys. Age 34:1051–1073.
Hasher L, Zacks RT (1988) Working memory, comprehension, and aging: A review and a new
view. Psychology of Learning and Motivation 22:193–225.
Hasher L, Zacks RT, May CP (1999) Inhibitory control, circadian arousal, and age. In D. Gopher
and A. Koriat (Eds.), Attention and performance XVII. Cognitive regulation of perfor-
mance: Interaction of theory and application. (pp. 653–675). Cambridge, MA: MIT Press.
Hillman C, Erickson KI, Kramer AF (2008) Be smart, exercise your heart: exercise effects on
brain and cognition. Nature Reviews Neuroscience 9:58–65.
Huettel SA, Singerman JD, McCarthy G (2001) The effects of aging upon the hemodynamic
response measured by functional MRI. Neuroimage 13(1):161–175.
Jonides J, Lewis RL, Nee DE, Lustig CA, Berman MG, Moore KS (2008) The mind and brain
of short-term memory. Annu Rev Psychol 59:193–224.
Jonides J, Marshuetz C, Smith EE, Reuter-Lorenz PA, Koeppe RA, Hartley A. (2000). Age
differences in behavior and PET activation reveal differences in interference resolution in
verbal working memory. Journal of Cognitive Neuroscience, 12:188–196.
255
Jost K, Bryck RL, Vogel EK, Mayr U (2011) Are Old Adults Just Like Low Working Memory
Young Adults? Filtering Efficiency and Age Differences in Visual Working Memory.
Karlamangla AS, Miller- Martinez D, Lachman ME, Tun PA, Koretz BK, Seeman TE
(2014) Biological correlates of adult cognition: midlife in the United States (MIDUS).
Neurobiology of Aging, 35(2):387–394.
Kennedy KM, Rodrigue KM, Bischof GN, Hebrank AC, Reuter-Lorenz PA, Park DC (2015)
Age Trajectories of Functional Activation Under Conditions of Low and High Processing
Demands: An Adult Lifespan fMRI Study of the Aging Brain. Neuroimage 104, 21–34.
Klyszejko Z, Rahmati M, Curtis CE (2014) Attentional priority determines working memory
precision. Vision Research 105:70–76.
Ko PC, Duda B, Hussey E, Mason E, Molitor RJ, Woodman GF, Ally BA (2014) Understanding
age-related reductions in visual working memory capacity: Examining the stages of
change detection. Attention Perception & Psychophysics 76:2015–2030.
Logie RH, Horne MJ, Pettit LD (2015) When cognitive performance does not decline across the
lifespan. Working Memory and Ageing:21–47.
Logie, RH, and Maylor, EA (2009). An internet study of prospective memory across adulthood.
Psychology & Aging 24:767–774.
Loosli, SV, Rahm, B, Unterrainer, JM, Weiller, C, Kaller, C (2014) Developmental change
in proactive interference across the lifespan: Evidence from two working memory tasks.
Develoipmental Psychology 50:1060–1072.
Lu, Hanzhang, Xu, Feng, Rodrigue, Karen M, Kennedy, Kristen M, Cheng, Yamei, Flicker,
Blair, et al. (2011) Alterations in Cerebral Metabolic Rate and Blood Supply across the
Adult Lifespan. Cerebral Cortex 21(6):1426–1434.
Luck SJ, Vogel EK (1997) The capacity of visual working memory for features and conjunc-
tions. Nature 390:279–281.
Luck SJ, Vogel EK (2013) Visual working memory capacity: from psychophysics and neurobi-
ology to individual differences. Trends in Cognitive Sciences 17:391–400.
Lustig C, May CP, Hasher L (2001) Working memory span and the role of proactive interfer-
ence. Journal of Experimental Psychology: General 130(2):199.
Lustig, C., Hasher, L., & Zacks, R. (2007). Inhibitory deficit theory: Recent developments in a
“new view”. In D. S. Gorfein and C. M. MacLeod (Eds.), The place of inhibition in cogni-
tion. Washington, DC: American Psychological Association.
Lustig C, Berman M, Nee DE, Lewis RL, Moore KS Jonides J (2009) Psychological and neural
mechanisms of short-term memory. In G.G. Bernston and J.T. Caccioppo (Eds). Handbook of
Neuroscience for the Behavioral Sciences. (pp. 567–585). John Wiley & Sons, Hoboken, NJ.
Lustig C, Jantz T (2015) Questions of age differences in interference control: When and how,
not if? Brain Research 1612:59–69.
Mattay VS, Fera F, Tessitore A, Hariri AR, Berman KF, Das S, Meyer-Lindenberg A, Goldberg
TE, Callicott JH, Weinberger DR (2006) Neurophysiological correlates of age-related
changes in working memory capacity. Neurosci Lett 392:32–37.
May CP, Hasher L, Kane MJ (1999) The role of interference in memory span. Memory &
Cognition, 27(5):759–767.
McEwen BS, Morrison JH (2013) The brain on stress: vulnerability and plasticity of the pre-
frontal cortex over the life course. Neuron 79:16–29.
Meunier D, Stamatakis EA, Tyler LK (2014) Age-related functional reorganization, structural
changes, and preserved cognition. Neurobiology of Aging 35:42–54.
256
Miller, EK, Cohen, JD (2001) An integrative theory of prefrontal cortex function. Annual
Review of Neuroscience, 24:167–202.
Morrison JH, Baxter MG (2012) The ageing cortical synapse: hallmarks and implications for
cognitive decline. Nature Reviews Neuroscience 13:240–250.
Nagel IE, Chicherio C, Li SC, von Oertzen T, Sander T, Villringer A, Heekeren HR, Backman
L, Lindenberger U (2008) Human aging magnifies genetic effects on executive function-
ing and working memory. Front Hum Neurosci 2:1.
Nagel IE, Lindenberger U (2015) Adult age differences in working memory Evidence from
functional neuroimaging. In: Working Memory and Ageing (Logie RH, Morris RG, eds),
pp. 129–154. Hove: Psychology Press.
Nagel IE, Preuschhof C, Li SC, Nyberg L, Backman L, Lindenberger U, Heekeren HR (2009)
ing memory. Proc Natl Acad Sci U S A106:22552–22557.
Nagel IE, Preuschhof C, Li SC, Nyberg L, Backman L, Lindenberger U, Heekeren HR (2011) Load
Modulation of BOLD Response and Connectivity Predicts Working Memory Performance
in Younger and Older Adults. Journal of Cognitive Neuroscience 23:2030–2045.
Naveh-Benjamin M (2000) Adult age differences in memory performance: Tests of an asso-
ciative deficit hypothesis. Journal of Experimental Psychology: Learning, Memory, and
Cognition 26:1170–1187.
Noack H, Lovden M, Lindenberger U (2012) Normal Aging Increases Discriminal Dispersion
in Visuospatial Short-Term Memory. Psychology and Aging 27:627–637.
Nyberg L, Andersson M, Kauppi K, Lundquist A, Persson J, Pudas S, Nilsson LG (2014)
Age-related and Genetic Modulation of Frontal Cortex Efficiency. Journal of Cognitive
maintenance. Trends in Cognitive Sciences 16:292–305.
tion in aging. Proceedings of the National Academy of Sciences of the United States of
America 107:22682–22686.
Oztekin I, Gungor NZ, Badre D (2012) Impact of aging on the dynamics of memory
retrieval: A time-course analysis. Journal of Memory and Language 67:285–294.
Park DC, Festini S (this volume). Neural structure and function in middle-aged adults.
Park DC, Polk TA, Park R, Minear M, Savage A, Smith MR (2004) Aging reduces neural spe-
cialization in ventral visual cortex. Proc Natl Acad Sci U S A 101:13091–13095.
Park DC, Reuter-Lorenz PA (2009) The Adaptive Brain: Aging and Neurocognitive Scaffolding.
In: Annual Review of Psychology, pp. 173–196.
Park J, Carp J, Kennedy KM, Rodrigue KM, Bischof GN, Huang CM, et al. (2012) Neural
broadening or neural attenuation? Investigating age-related dedifferentiation in the face
network in a large lifespan sample. The Journal of Neuroscience 32(6):2154–2158.
Parks CM, Iosif AM, Farias S, Reed B, Mungas D, DeCarli C (2011) Executive function
mediates effects of white matter hyperintensities on episodic memory. Neuropsychologia
49:2817–2824.
Parra MA, Abrahams S, Fabi K, Logie R, Luzzi S, Della Sala S (2009) Short-term memory
binding deficits in Alzheimers disease. Brain 132:1057–1066.
Parra MA, Abrahams S, Logie RH, Della Sala S (2009) Age and binding within-dimension
features in visual short-term memory. Neuroscience Letters 449:1–5.
Payer D, Marshuetz C, Sutton B, Hebrank A, Welsh RC, Park DC (2006) Decreased neural spe-
cialization in old adults on a working memory task. NeuroReport: For Rapid Communication
of Neuroscience Research 17:487–491.
257
Peich MC, Husain M, Bays PM (2013) Age-Related Decline of Precision and Binding in Visual
Working Memory. Psychology and Aging 28:729–743.
Pettigrew C, Martin RC (2014) Cognitive declines in healthy aging: Evidence from multiple
aspects of interference resolution. Psychology and Aging 29:187–204.
Postle BR (2006) Working memory as an emergent property of the mind and brain. Neuroscience
139:23–38.
Postle BR. (2015). The cognitive neuroscience of visual short-term memory. Current Opinion
in Behavioral Sciences 1:40–46.
Rapp MA, Reischies FM (2005) Attention and executive control predict Alzheimer disease in
late life -Results from the Berlin Aging Study (BASE). Am J Geriatr Psychiatr 13:134–141.
aging in middle-aged and older adults: Regional and individual differences. Neuroimage
51:501–511.
Raz N, Rodrigue KM (2006) Differential aging of the brain: Patterns, cognitive correlates and
modifiers. Neuroscience and Biobehavioral Reviews 30:730–748.
Reuter-Lorenz PA, Festini SB, Jantz T (2015) Neurocognitive aging and executive function. In: The
Handbook of Aging. (Schaie KW, Willis S, eds) New York, NY: Academic Press pp. 245–261.
Reuter-Lorenz PA, Jonides J, Smith EE, Hartley A, Miller A, Marshuetz C, Koeppe RA (2000)
Age differences in the frontal lateralization of verbal and spatial working memory revealed
by PET. Journal of Cognitive Neuroscience 12:174–187.
Reuter-Lorenz PA, and Lustig C (2005) Brain aging: Reorganizing discoveries about the aging
mind. Current Opinion in Neurobiology 15:245–251.
Reuter-Lorenz PA, Park DC (2014) How does it STAC up? Revisiting the Scaffolding Theory
of Aging and Cognition. Neuropsychology Review 24: 355–370.
Reuter-Lorenz PA, Stanczak L, Miller AC (1999) Neural recruitment and cognitive
aging: Two hemispheres are better than one, especially as you age. Psychological
Science 10:494–500.
Reuter-Lorenz PA, Sylvester C-YC (2005) The Cognitive Neuroscience of Working Memory and
Aging. In: Cognitive neuroscience of aging: Linking cognitive and cerebral aging. (Cabeza
R, Nyberg L, Park D, eds), pp. 186–217. New York, NY US: Oxford University Press.
Rowe G, Turcotte J, Hasher L (2008) Age differences in visuospatial working memory.
Rypma B, D’Esposito M (2000) Isolating the neural mechanisms of age-related changes in
human working memory. Nature Neuroscience 3:509–515.
Sander MC, Lindenberger U, Werkle-Bergner M (2012) Lifespan age differences in working
memory: A two- component framework. Neuroscience and Biobehavioral Reviews 36:
2007–2033.
Sander MC, Werkle-Bergner M, Lindenberger U (2011) Contralateral delay activity reveals
life-span age differences in top-down modulation of working memory contents. Cerebral
Cortex 21:2809–2819.
Sander MC, Werkle-Bergner M, Lindenberger U (2011) Binding and Strategic Selection in
Working Memory: A Lifespan Dissociation. Psychology and Aging 26:612–624.
Schaie KW, Willis SL, Caskie GIL (2004) The Seattle longitudinal study: relationship between
personality and cognition. Neuropsychology, development, and cognition. Section B,
Aging, neuropsychology and cognition, 11(2–3):304–324.
Schneider-Garces NJ, Gordon BA, Brumback-Peltz CR, Shin E, Lee Y, Sutton BP, Maclin EL,
Gratton G, Fabiani M (2010) Span, CRUNCH, and Beyond: Working Memory Capacity
and the Aging Brain. J Cogn Neurosci (4):655–669.
258
Sebastian A, Baldermann C, Feige B, Katzev M, Scheller E, Hellwig B, et al. (2013)

Differential effects of age on subcomponents of response inhibition. Neurobiology of
aging 34(9):2183–2193.
Sole-Padulles C, Bartres-Faz D, Junque C, Clemente IC, Molinuevo JL, Bargallo N, Sanchez-
Aldeguer J, Bosch B, Falcon C, Valls-Sole J (2006) Repetitive transcranial magnetic stim-
ulation effects on brain function and cognition among elders with memory dysfunction.
A randomized sham-controlled study. Cerebral Cortex 16:1487–1493.
Spreng RN, Wojtowicz M, Grady CL (2010) Reliable differences in brain activity between
young and old adults: A quantitative meta-analysis across multiple cognitive domains.
Neuroscience and Biobehavioral Reviews 34:1178–1194.
St-Laurent M, Abdi H, Bondad A, Buchsbaum BR (2014) Memory reactivation in healthy
aging: evidence of stimulus- specific dedifferentiation. The Journal of Neuroscience
34(12):4175–4186.
Stevens WD, Hasher L, Chiew KS, Grady CL (2008) A Neural Mechanism Underlying Memory
Failure in Older Adults. Journal of Neuroscience 28(48):12820–12824.
Suchow JW, Fougnie D, Brady TF, Alvarez GA (2014) Terms of the debate on the format and
structure of visual memory. Attention Perception & Psychophysics 76:2071–2079.
Toepper M, Markowitsch HJ, Gebhardt H, Beblo T, Bauer E, Woermann FG, Driessen M,
Sammer G (2014) The impact of age on prefrontal cortex integrity during spatial working
memory retrieval. Neuropsychologia 59:157–168.
Tomaszewski S, Cahn-Weiner DA, Harvey DJ, Reed BR, Mungas D, Kramer JH, Chui H
(2009) Longitudinal Changes in Memory and Executive Functioning are Associated with
longitudinal change in instrumental activities of daily living in older Adults. Clinical
Neuropsychologist 23:446–461.
Tsvetanov KA, Henson RN, Tyler LK, Davis SW, Shafto MA, Taylor JR, Williams N, Cam-Can
Rowe, JB (2015) The effect of aging on fMRI: Correction for the confounding effects of
vascular reactivity evaluated by joint fMRI and MEG in 335 adults. Human Brain Maping.
Turner GR, Spreng RN (2012) Executive functions and neurocognitive aging: dissociable pat-
terns of brain activity. Neurobiology of Aging 33(4):826-e1.
Vallesi A, McIntosh AR, Stuss DT (2011) Overrecruitment in the Aging Brain as a Function of
Task Demands: Evidence for a Compensatory View. J Cogn Neurosci 23:801–815.
Vaughan L, Giovanello K (2010) Executive Function in Daily Life: Age-Related Influences of Executive
Processes on Instrumental Activities of Daily Living. Psychology and Aging 25:343–355.
Verhaeghen, P. (2011). Aging and executive control: reports of a demise greatly exaggerated.
Current Directions in Psychological Science 20(3):174–180.
Vogel EK, Machizawa MG (2004) Neural activity predicts individual differences in visual
working memory capacity. Nature 428:748–751.
Wang M, Gamo NJ, Yang Y, Jin LE, Wang XJ, Laubach M, Mazer JA, Lee D, Arnsten AFT
(2011) Neuronal basis of age-related working memory decline. Nature 476:210-U110.
Weeks JC, Hasher L (2014) The disruptive –and beneficial –effects of distraction on older adults’
perforamnce. Frontiers in Psychology. Feb 18;5:133.
Weinstein AM, Voss MW, Prakash RS, Chaddock L, Szabo A, White SM, Wojcicki TR, Mailey
E, McAuley E, Kramer AF, Erickson KI (2012) The association between aerobic fitness
and executive function is mediated by prefrontal cortex volume. Brain Behavior and
Immunity 26:811–819.
Yuan P, Raz N (2014) Prefrontal cortex and executive functions in healthy adults: A meta-
analysis of structural neuroimaging studies. Neuroscience and Biobehavioral Reviews
42:180–192.
259
10
Neural Correlates of
Age-Related Slowing
Timothy A. Salthouse
S ome of the most fundamental measures of cognitive functioning are

based on the speed with which an individual can carry out very simple
search, substitution, or comparison operations. Measures of speed of processing are
particularly interesting in the context of research on aging because they are among
the behavioral measures with the greatest sensitivity to age in both cross-sectional
(Verhaeghen and Salthouse, 1997), and longitudinal (Fozard, et al., 1994; Salthouse,
2013; Schaie, 1989) comparisons, and they have been found to have strong relations
with other types of cognitive measures involving memory, reasoning, and decision
making (e.g., Salthouse, 1996). These characteristics help explain why it has been
suggested that “… slowing … is arguably the most often replicated finding across
studies of age effects on neuropsychological test performance (Lu et al., 2011).”
In light of these properties, it is not surprising that many researchers have been
interested in investigating possible neural correlates of speed of processing. A non-
exhaustive sample of relevant studies is presented in Table 10.1. Although the table
includes many categories of neural measures and many categories of speed measures,
the entries do not fully reflect the true diversity inherent in the studies because differ-
ent types of measures are included within the neural and speed categories, and there
is no consideration of the regions in the brain from which the neural measures were
obtained. For example, the white matter integrity category includes measures of frac-
tional anisotropy and mean diffusivity along axial and radial directions derived from
many different brain locations, but none of this information is reported in the table.
Moreover, to keep the review manageable, only structural and not functional charac-
teristics are represented in the table.
259
260
Table 10.1 Studies reporting relations between neuroanatomical measures

and speed measures
Neural Measure Speed Measure N Age Range Study
White matter integrity Finger tapping speed 72 23–80 Bartzokis et al. (2010)
Trail Making 120 18–83 Bendlin et al. (2010)

Composite of P&P tests 165 54–89 Borghesani et al. (2013)
RT 38 20–28, 63–78 Bucur et al. (2008)
RT and P&P tests 93 50–81 Burgmans et al. (2011)
3 timed tests 99 50–90 Charlton et al (2006)
RT and Inspection Time 40 83 Deary et al. (2006)
RT switch costs 40 19–27, 63–76 Gold et al. (2010)
Stroop and RT 100 49–80 Haasz et al. (2013)
Trail Making 128 43–87 Jacobs et al. (2013)
P&P tests 52 18–81 Kennedy and Raz (2009)
RT 131 55–87 Kerchner et al. (2012)
4 timed tests 38 57–90 Kochunov et al. (2010)
P&P tests 253 60–87 Laukka et al. (2013)
Trail Making & Digit 152 55–80 Lu et al. (2011)
Symbol
Trail Making 17 56–85 O’Sullivan et al. (2001)

Timed tests 81 77–91 Papp et al. (2014)
RT & Inspection time 132 72 Penke et al. (2010)
RT & Inspection time 420 71–73 Penke et al. (2012)
P&P test 287 25–80 Salami et al. (2012)
RT 52 25–82 Sasson et al. (2012)
3 timed tests 64 54–91 Schiavone et al. (2009)
Digit Symbol 39 18–31 Turken et al. (2008)
Trail Making & Digit 342 72–92 Wen et al. (2011)
Symbol
Stroop & Letter-Digit 860 Mean = 67 Vernooij et al. (2009)

Stroop 100 49–80 Ystad et al. (2011)
White matter lesions Trail Making & Block 235 23–73 Duering et al. (2013)
Design
2 timed tests 832 60–90 Prins et al. (2005)
White matter RT variability 415 44–48 Bunce et al. (2013)

hyperintensities Trail Making 128 43–87 Jacobs et al. (2013)
RT 38 65–89 Lockhart et al. (2014)
Trail Making 156 Mean = 68 Oosterman et al. (2010)
Timed tests 81 77–91 Papp et al. (2014)
RT, Trail Making 150 44–82 Schmidt et al. (1993)
RT, Inspection Time & PP 634 Mean = 73 Valdes Herandez et al.
(2013)
Stroop 554 70–82 van den Heuvel et al.
(2006)
261
Neural Correlates of Age-Related Slowing 261
Table 10.1 Continued
Neural Measure Speed Measure N Age Range Study
White matter volume Trail Making variant 704 Mean = 80 Brickman et al. (2011)
Trail Making variant 42 19–79 Eckert et al. (2010)
RT and Trail Making 82 40–50 Ferreira et al. (2014)
RT distributional measures 133 46–96 Jackson et al. (2012)
4 timed tests 38 57–90 Kochunov et al. (2010)
Tapping variability 34 19–49 Ullen et al. (2008)
RT variability 71 20–88 Walhovd and Fjell (2007)
Gray matter volume Trail Making variant 42 19–79 Eckert et al. (2010)
RT and Trail Making 82 40–50 Ferreira et al. (2014)
Medial temporal lobe Trail Making 156 Mean = 68 Oosterman et al. (2010)
volume Timed tests 81 77–91 Papp et al. (2014)
Cortical atrophy 2 timed tests 832 60–90 Prins et al. (2005)

Gross Brain Volume 2 timed tests 69 62–85 Rabbitt et al. (2006)
MRS indices 4 timed tests 38 57–90 Kochunov et al. (2010)
Some of the samples in the table may overlap because it is not always clear whether some of the data in a given report
have been published before.
Nevertheless, several interesting aspects can be noted about the entries in

Table 10.1. First, the entries are dominated by neural measures that have been hypoth-
esized to reflect the efficiency of interregional communication (e.g., white matter
integrity, white matter lesions and hyperintensities). Although this emphasis could be
attributable to the prominence of the disconnection hypothesis of age-related slowing
(e.g., Burgmans et al., 2011; Chen et al., 2009; Deary et al., 2006; O’Sullivan, et al.,
2001), it is not clear from the published reports whether the disproportionate represen-
tation is attributable to an a priori focus on these types of measures, or whether many
different measures were examined and these were the measures most often found to be
related to speed. However, it is interesting that a variety of neural measures not obvi-
ously linked to connection efficiency, such as brain volumes, have also been found to
be related to measures of speed. A second noteworthy point in Table 10.1 is that the
sample sizes in many of the studies have been relatively small, which limits the preci-
sion of the estimates of the relations, and increases the likelihood that the results may
have been influenced by chance. A third point is that many of the studies had a narrow
range of ages, primarily involving adults over the age of 60, which makes it difficult
to draw inferences about processes related to aging.
Even with these acknowledged limitations, it is apparent from inspection of the
entries in Table 10.1 that relations have been examined between many neural mea-
sures and many speed measures. Although it might seem straightforward to identify
neural correlates of age-related slowing on the basis of the existing literature, the
thesis of the current chapter is that this task is deceptively complex. Specifically, it is
suggested that two critical issues need to be considered when interpreting correlations
between neural measures and speed measures: level of analysis, and influences on age
relations.
262
Level of Analysis
Level of analysis in the current context refers to the degree of aggregation or abstrac-
tion across measures in a given set of data. The level of analysis issue can be elabo-
rated by referring to Figure 10.1. The left side of the top panel in the figure indicates
that there are many possible neural measures, including information about what in
terms of the assessment of either structure or function, and information about where
in terms of measurement in different brain regions. The right side of the top panel of
Figure 10.1 indicates that the number of possible measures of cognitive functioning
is also very large, as multiple cognitive domains could be represented with multiple
measures within each domain. For example, in the domain of speed, the assessments
have ranged from scores on paper-â•‰and-â•‰pencil tests, to measures of reaction time (RT)
Neurobiological Cognitive
Structure (volume, thickness, white matter integrity, etc.) Cognitive Domain (speed, memory, reasoning, etc.)
Function (activation, functional connectivity, etc.) Measure (number correct, reaction time, etc.)
Region (frontal temporal, etc.) Property (central tendency, distributional measure, etc.)
2nd-order 2nd-order
Factor Factor
1st-order 1st-order
Factors Factors
Observed Observed
Measures Measures
FigureÂ€10.1â•‡Top panel—â•‰Illustration of many interrelated measures of brain structure and

function and many interrelated measures of cognitive functioning. Bottom panel—â•‰Illustration
of relations between neural structures and cognitive structures. The dashed line at the top rep-
resents a relation between a specific neural measure and a specific cognitive measure, the dotted
line represents a relation between first-â•‰order factors, and the solid line represents a relation
between second-â•‰order factors. As noted in the text, the first two relations may not be interpreta-
ble unless considered in the context of the third relation.
263
and inspection time (see Deary et al., 2010; Salthouse, 2000). In addition, numerous
measures have been derived from RT tasks, ranging from indices of central tendency
and within-person variability (e.g., Bunce, et al., 2013; Jackson et al., 2012; Nilsson,
et al., 2014), to measures postulated to represent specific theoretical processes (e.g.,
Balota and Yap, 2011; Jackson et al., 2012).
The double-headed arrow in the top of Figure 10.1 indicates that relations have fre-
quently been examined between a single neural measure and a single speed measure.
Bivariate relations of this type are often theoretically motivated, and it is sometimes
assumed that they are more straightforward to interpret than more complex multivariate
relations. It is therefore somewhat counter-intuitive that interpretations may actually be
less ambiguous if relations are examined in a multivariate context. That is, if interrela-
tions among the variables are not considered when attempting to isolate specific rela-
tions of neural measures with cognitive measures, the researcher may be inadvertently
studying influences on multiple cognitive abilities instead of just the measure of pri-
mary interest. Indeed, the top panel of Figure 10.1 contains lines connecting the same
types of measures with one another to represent the possibility that the measures may
not be independent of one another, but instead may be interrelated to varying degrees.
There is a long history of research with cognitive abilities indicating that individ-
uals with high values in one cognitive measure typically have higher-than-average
values in other cognitive measures. Furthermore, relations among measures vary in
strength, which allows organizational structures to be postulated based on the strength
of the relations between cognitive measures. A variety of different correlation-based
structures have been proposed to account for the organization of cognitive abilities,
with hierarchical (e.g., Salthouse, 2009; Salthouse and Ferrer-Caja, 2003), and bi-
factor (Booth et al., 2013; Salthouse and Ferrer-Caja, 2003), models among the most
popular. Structures such as these have proven valuable in indicating the level at which
variables such as age or anxiety are related to measures of cognitive functioning. For
example, recent studies have found that large proportions of the influences on cogni-
tive measures associated with age (e.g., Salthouse, 2009; Salthouse and Ferrer-Caja,
2003), and with anxiety (Salthouse, 2012), operate at a relatively broad level, with
only small contributions unique to individual measures.
Although the relevant evidence is less extensive, there are also reports of moderate
to strong relations among various neural measures (e.g., Booth et al., 2013; Ecker et al.,
2009; Kievit et al., 2012; Laukka et al., 2013; Li et al., 2012; Lovden et al., 2013; Penke
et al., 2010; 2012; Wahl et al., 2010; Wu et al., 2013). Many questions still remain about
the most meaningful dimensions in the structures, such as by cortical regions, tracts con-
necting regions, or type of neural measure. However, it is clear from recent studies that
neural measures are not independent of one another, which implies that they are likely to
share influences with one another even when they are analyzed separately.
These considerations lead to the question of level of analysis, which concerns the
extent to which the neural-speed relation is specific to a particular combination of
measures, or may, at least in part, reflect influences operating at broader levels in
the respective neural and cognitive structures. Note that this issue is not the same
as distinguishing which of several measures, such as radial versus axial diffusivity
measures of white matter integrity, or central tendency versus variability measures
of reaction time performance, are primarily responsible for a neural-speed relation.
264
Efforts such as these are valuable as an initial step in specifying the relevant relations,
but only by simultaneously examining several neural measures and several cognitive
measures is it possible to distinguish general (i.e., shared with other measures) and
specific (i.e., unique to one measure) influences on the relations.
One approach relevant to the level-of-analysis issue involves considering multiple
neural and speed measures, and then comparing the magnitude of relevant relations
across different combinations of measures. For example, some researchers have reported
stronger relations of neural measures with measures of speed than with other cognitive
measures (e.g., Bendlin et al., 2010; Booth et al., 2012; Deary et al., 2006; Farias et
al., 2012; Kerchner et al., 2012; Laukka et al., 2013; Penke et al., 2010; Prins et al.,
2005; Rabbitt et al., 2006; Salami et al., 2012; Schiavone et al., 2009; Schmidt et
al., 1993; Ystad et al., 2011). However, this is not always found (e.g., Borgehesani
et al., 2013; Brickman et al., 2011; Haasz et al., 2013; Jacobs et al., 2013; Kennedy and
Raz, 2009; Wen et al., 2011), and several of the same neural measures found to be
related to measures of speed have also been found to be related to other types of cog-
nitive measures (e.g., Bennett et al., 2012; He et al., 2012; Kloppenborg et al., 2014;
Maillard et al., 2012; Samanez-Larkin et al., 2012; Smith et al., 2011; Zahr et al., 2009).
A more informative approach to investigating neural-speed relations involves
examining the structure relating various neural measures with one another, the struc-
ture relating various cognitive measures to one another, and then investigating rela-
tions between the two structures (see Lovden et al., 2013; Salthouse, 2011). Only a
few studies of this type have been reported, but the available results suggest that many
brain-behavior relations operate at a relatively general level. For example, large pro-
portions of the relations between measures of white matter integrity and measures of
cognitive ability have been found to be shared, and are not specific to particular neural
regions or cognitive domains (e.g., Booth et al., 2013; Penke et al., 2012).
Multivariate approaches such as these are desirable in investigating brain-behavior
relations because the interrelations among neural measures and among cognitive mea-
sures suggest that merely because a researcher is analyzing the relation between a
single neural measure and a single cognitive measure does not mean that he or she is
studying relations unique to those measures. That is, when only a bivariate relation
between a measure of white matter integrity in a specific brain region and a measure
of mean RT is examined, it cannot be determined how much of that relation is unique
to those measures, and how much of it reflects influences operating at more general
levels in the neural and cognitive structures.
This issue can be elaborated by reference to the bottom panel of Figure 10.1, which
portrays relations among neural measures and cognitive measures in the context of
organizational structures representing the interrelations of each type of measure.
Neural correlates of a cognitive measure such as speed could be examined by con-
sidering the relation between a single neural measure and a single cognitive measure,
as portrayed in the dotted bi-directional arrow at the top of the panel. Although it
might be tempting to assume that this relation reflects the specific aspects that the
two measures have in common, inspection of the figure indicates that each measure
is related to higher-order factors which could also have relations to one another. This
implies that some of the relation between the specific measures may be attributable to
relations at the level of first-order or second-order factors, and only by including those
relations can the aspects unique to the target measures be estimated.
â•‡ 265
Neural Correlates of Age-Related Slowingâ•… 265
Influences on Age Relations
A second important issue to consider when trying to identify neural correlates of age-â•‰
related slowing concerns the type of evidence relevant to determining that the neural
measures are associated with the relations between age and speed, and are not simply
reflections of relations with speed independent of age. That is, if one is interested
in correlates of age-â•‰related slowing, the relevant outcome variable is the relation of
speed to age, and not the relation of a neural measure to a speed measure at a single
age, or at a single point in time. Even a discovery of parallel cross-â•‰sectional trajec-
tories of a neural measure and a speed measure (e.g., Bartzokis et al., 2010), is not
sufficient to infer that the neural measure is a correlate of age-â•‰related slowing without
evidence that the age relations in the two types of measures are linked to one another,
and are not simply independent correlates of aging.
Because longitudinal change in measures of speed within the same individuals
directly reflects age-â•‰related slowing, the relation of longitudinal change in mea-
sures of speed with neural measures can be considered the most relevant type of
information regarding neural correlates of age-â•‰related slowing. Unfortunately, only
a few studies with longitudinal data on speed measures have reported relations
between the longitudinal change in speed and the longitudinal change in neu-
ral measures (but see Prins et al., 2005; van den Heuvel, et al., 2006), and thus
there is limited direct information about correlates of age-â•‰related slowing at the
current time.
However, indirect information about correlates of age-â•‰related slowing can be
derived from cross-â•‰sectional comparisons of the relation between age and speed at
different levels of the neural measure. Consider Figure 10.2 in which age-â•‰related
slowing is represented as the slope of the function relating speed to age. At least two
types of analyses can be carried out with cross-â•‰sectional data such as that portrayed
in Figure 10.2.
Neural
Measure
Speed
Age
FigureÂ€10.2â•‡ Illustration of a moderation pattern in which the relation between speed and age
varies as a function of the value of a neural measure.
266
One type of analysis focuses on the possibility of moderation, in which the age-
speed relation varies according to the level of the neural measure. The pattern in Figure
10.2 is an example of moderation because the age-speed relation is weaker at higher
levels of the neural measure, or equivalently, the neural-speed relation is weaker at
younger ages. A moderation outcome such as this, particularly when confirmed by
a significant interaction of age and the neural measure in prediction of the speed
measure, would be consistent with the neural measure functioning as a correlate of
age-related slowing. That is, when the age-speed relation is assumed to represent age-
related slowing, a finding that the relation is moderated by a neural measure implies
that the neural measure is a correlate of age-related slowing. Although potentially
informative about neural correlates of age-related slowing, moderation analyses have
seldom been reported in studies investigating correlates of cognitive aging.
The most popular type of multivariate analysis of cross-sectional data has been
some version of mediation. The rationale is that if a neural variable is involved in
mediating the relation between age and speed, then the strength of the age-speed
relation should be reduced when variability in the mediator is statistically con-
trolled. That is, to the extent that much of the age-speed relation is attributable
to age-related reductions in the neural measure, the age-speed relation would be
expected to be small or nonexistent when there is little variation in the neural mea-
sure. Most mediation analyses rely on statistical control procedures to effectively
hold the value of the neural measure constant at the average level in the sample
when examining the relation between age and speed, but if the sample is large
enough, analyses could be conducted in subgroups stratified according to the level
of the neural measure.
A relatively large number of studies have found evidence of this type of statis-
tical mediation with different combinations of neural measures, speed measures,
and ranges of participant ages (e.g., Borghesani et al., 2013; Burgmans et al., 2011;
Eckert et al., 2010; Ferreriera et al., 2014; Gold et al., 2010; Jacobs, et al., 2013;
Kerchner et al., 2012; Laukka et al., 2013; Lu et al., 2013; Madden et al., 2014;
Rabbitt et al., 2006; Salami et al., 2012; see reviews in Bennett and Madden, 2014
and Nilsson et al., 2014). Many of the authors of these studies have concluded that
the reduction in the age-speed relation when the variation in the neural measure was
statistically eliminated implied that the neural measure causally mediates the relation
of age on speed. However, these conclusions may be premature because of a failure
to distinguish statistical mediation from causal mediation (e.g., Lindenberger, et al.,
2011; Mackinnon et al., 2007). Moreover, it is not always recognized that the results
of mediation analyses could also be consistent with a variety of causal hypotheses,
and it is therefore important to consider alternative theoretically-motivated models
of the relations among the measures before reaching conclusions from mediation
analyses (e.g., Bennett and Madden, 2014; Nilsson et al., 2014; Salthouse, 2011).
For example, plausible models for the relations among age, neural measures and
speed measures are that age-related decreases in neural measures could contribute to
age-related decreases in cognitive measures, relations between neural measures and
cognitive measures could be attributable to the influence of age on both types of mea-
sures, or age-related differences in experience with cognitive activities could con-
tribute to age-related differences in neural measures. If predictions from alternative
â•‡ 267
Neural Correlates of Age-Related Slowingâ•… 267
models are not examined, it may be as plausible to conclude that the relation between
a neural measure and a speed measure is attributable to their common relations with
age, instead of to age-â•‰related decreases in the neural measure mediating age-â•‰related
decreases in the speed measure.
It is also worth noting that, although seldom formally investigated by researchers
interested in age-â•‰cognition relations, the existence of a moderation pattern violates
the assumption of simple mediation that control of the neural measure will have a
similar effect on the speed measure at different ages. Strictly speaking, therefore,
simple mediation analyses may not be interpretable if the relation between two of
the variables varies as a function of the value of a third variable. When a moderation
pattern is detected more complex mediational models, such as moderated mediation
or mediated moderation (e.g., MacKinnon, et al., 2007), could be used, but it should
be recognized that conclusions based on simple mediation analyses may be incorrect
when the relation of the hypothesized mediator varies as a function of one of the other
measures. (See Salthouse, 2011; Salthouse, 2010, pp. 99–â•‰105; and Lovden et al., 2013
for additional discussion of mediation).
Conclusion
Measures of speed have been reported to have some of the strongest relations with age
of any behavioral measure. It therefore seems plausible to assume that identification
of neural correlates of age-â•‰related slowing will be informative about causes of age-â•‰
related cognitive decline. Although there are numerous reports of relations between
neural measures and speed measures, only limited conclusions about the neural cor-
relates of age-â•‰related slowing are possible at the current time because of weaknesses
of prior research. The final section of the chapter therefore describes recommenda-
tions designed to lead to stronger conclusions in the future.
First, because the patterns of neural-â•‰speed relations could differ at different peri-
ods in adulthood, and because powerful tests of moderation require a wide span of
ages, it is desirable to include research participants across a broad age range, and not
just individuals from a narrow period in late adulthood. Without variation in age the
results are only relevant to correlates of speed, and not to correlates of age-â•‰related
differences in speed. Second, the samples of participants should be moderately large
to have adequate power to detect theoretically relevant interactions such as those
postulated in moderation, and to obtain precise estimates of the relations. Although
neuroimaging measures are expensive to obtain, correlational patterns in research
involving individual differences may not be interpretable with small sample sizes.
Third, multiple neural measures should be examined, including measures of different
aspects of structure and function across different brain regions, and analyses should
be conducted to determine how the neural measures are related to one another prior
to considering relations with speed measures. Fourth, multiple cognitive measures
should be examined in addition to speed measures, and analyses should be conducted
to determine how the speed and other cognitive measures are related to one another
prior to considering relations with neural measures. And finally, powerful analyt-
ical procedures should be used to distinguish general and specific influences, and
268
to investigate influences on age–â•‰speed relations, and not just relations with speed
independent of age.
Implementation of these recommendations will clearly increase the complexity
(and cost) of research, and there is no guarantee that adoption of the recommenda-
tions will inevitably lead to better insight into the neural bases of age-â•‰related slowing.
However, without changes such as these it will continue to be difficult to identify
which particular neural measures are uniquely related to particular cognitive mea-
sures, and whether those relations vary as a function of age.
References
Balota DA, Yap MJ (2011) Moving beyond the mean in studies of mental chronometry: The
power of response time distributional analyses. Curr Dir Psych Sci 20:160–â•‰166.
Bartzokis G, Lu PH, Tingus K, Mendez MF, Richard A, Peters DG, Oluwadara B, Barrall KA,
Finn JP, Villblanca P, Thompson PM, Mintz J (2010) Neurobio Aging 31: 1554–â•‰1562.
Bendlin BB, Fitzgerald ME, Ries ML, Xu G, Kastman EK, Thiel BW, Rowley HA, Lazar M,
Alexander AL, Johnson SC (2010) White matter in aging and cognition: A cross-â•‰sectional
study of microstructure in adults aged eighteen to eighty-â•‰three. Dev Neuropsych 35: 257–â•‰277.
Bennett, IJ, Madden DJ, (2014). Disconnected aging: Cerebral white matter integrity and
age-related differences in cognition. Neuroscience 276: 187–205.
Bennett IJ, Motes MA, Rao NK, Rypma B (2012) White matter tract integrity predicts visual
search performance in young and older adults. Neurobiol Aging 33:433.e421–â•‰e433.
Booth T, Bastin ME, Penke L, Munoz Maniega S, Murray C, Royle NA, Gow AJ, Corley J,
Henerson RD, Valdes Hernandez MC, Starr JM, Wardlaw JM, Deary IJ (2013) Brain white
matter tract integrity and cognitive abilities in community-â•‰dwelling older people: The
Lothian Birth Cohort, 1936. Neuropsych 27: 595–â•‰607.
Borghesani PR, Madhyastha TM, Aylward EH, Reiter MA, Swarny BR, Warner Schaie K,
Willis SL (2013) The association between higher order abilities, processing speed, and age
are variably mediated by white matter integrity during typical aging. Neuropsychologia
51:1435–â•‰1444.
Brickman AM, Siedlecki KL, Muraskin J, Manly JJ, Luchsinger JA, Yeung L-â•‰K, Brown TR,
DeCarli C, Stern Y (2011) White matter hyperintensities and cognition: Testing the reserve
hypothesis. Neurobio Aging 32:1588–â•‰1598.
Bucur B, Madden DJ, Spaniol J, Provenzale JM, Cabeza R, White LE, Huettel SA (2008). Age-â•‰
related slowing of memory retrieval: contributions of perceptual speed and cerebral white
matter integrity. Neurobiol Aging 29:1070–â•‰1079.
Bunce D, Bielak AAM, Cherbuin N, Batterham PJ, Wen W, Sachdev P, Anstey KJ (2013)
Utility of intraindividual reaction time variability to predict white matter hyperintensi-
ties: A potential tool for clinical contexts? J Int Neurosych Soc 19: 971–â•‰976.
by differences in axonal integrity. Neuroimage 55:1287–â•‰1297.
Charlton RA, Barrick TR, McIntyre DJ, Shen Y, O’Sullivan M, Howe FA, Clark CA, Morris
RG, Markus HS (2006) White matter damage on diffusion tensor imaging correlates with
age-â•‰related cognitive decline. Neurology 66:217–â•‰222.
Chen NK, Chou YH, Song AW, Madden DJ (2009) Measurement of spontaneous signal fluc-
tuations in fMRI: adult age differences in intrinsic functional connectivity. Brain Struct
Funct 213:571–â•‰585.
269
Deary IJ, Bastin ME, Pattie A, Clayden JD, Whalley LJ, Starr JM, Wardlaw JM (2006) White
matter integrity and cognition in childhood and old age. Neurol 66: 505–512.
Deary IJ, Johnson W, Starr JM (2010) Are processing speed tasks biomarkers of cognitive
aging? Psych Aging 25:219–228.
Duering M, Gonik M, Malik R, Zieren N, Reyes S, Jouvent E, Herve D, Gchwendtner A, Opherk
C, Chabriat H, Dichgans M (2013) Identification of a strategic brain network underlying
processing speed deficits in vascular cognitive impairment. NeuroImage 66: 177–183.
Ecker C, Stahl D, Daly E, Johnston P, Thomson A, Murphy DGM (2009). Is there a com-
mon underlying mechanism for age-related decline in cortical thickness? NeuroRepor
20:1155–1160.
Eckert MA, Keren NI, Roberts DR, Calhoun VD, Harris KC (2010) Age-related changes in
processing speed: Unique contributions of cerebellar and prefrontal cortex. Front Hum
Neurosci, 4: Article 10.
Farias ST, Mungas D, Reed B, Carmichael O, Beckett L, Harvey D, Olichney J, Simmons A,
DeCarli C (2012) Maximal brain size remains an important predictor of cognition in old
age, independent of current brain pathology. Neurobio Aging 33:1758–1768.
Ferreira D. Molina Y, Machado A, Westman E, Wahlund L-O, Nieto A, Correia R, Junque C,
Diaz-Flores L, Barroso J (2014) Cognitive decline is mediated by gray matter changes
during middle age. Neurobio Aging 35:1086–1094.
Fozard JL, Vercruyssen M, Reynolds SL, Hancock PA, Quilter RE (1994) Age differences
and changes in reaction time: the Baltimore Longitudinal Study. J Geronto Psychol Sci
49: P179-P189.
Gold BT, Powell DK, Xuan L, Jicha GA, Smith CD (2010) Age-related slowing of task switch-
ing is associated with decreased integrity of frontoparietal white matter. Neurobiol Aging
31:512–522.
Haasz J, Westlye ET, Fjaer S, Espeseth T, Lundervold A, Lundervold AJ (2013) General fluid-
type intelligence is related to indices of white matter structure in middle-aged and old
adults. Neuroimage 83C:372–383.
He J, Wong VS, Fletcher E, Maillard P, Lee DY, Iosif AM, Singh B, Martinez O, Roach AE,
Lockhart SN, Beckett L, Mungas D, Farias ST, Carmichael O, DeCarli C (2012) The con-
tributions of MRI-based measures of gray matter, white matter hyperintensity, and white
matter integrity to late-life cognition. AJNR Am J Neuroradiol 33:1797–1803.
Jackson JD, Balota DA, Duchek JM, Head D (2012) White matter integrity and reaction
time intraindividual variability in healthy aging and early- stage Alzheimer disease.
Jacobs HI, Leritz EC, Williams VJ, Van Boxtel MP, van der Elst W, Jolles J, Verhey FR,
McGlinchey RE, Milberg WP, Salat DH (2013) Association between white matter micro-
structure, executive functions, and processing speed in older adults: the impact of vascular
health. Hum Brain Mapp 34:77–95.
Kerchner GA, Racine CA, Hale S, Wilheim R, Laluz V, Miller BL, Kramer JH (2012) Cognitive
processing speed in older adults: Relationship with white matter integrity. PLoS One
7(11):e50425.
Kievit RA, van Rooijen H, Wicherts JM, Waldorp LJ, Kan KJ, Scholte HS, Borsboom D (2012)
Intelligence and the brain: A model-based approach. Cog Neurosci 3:89–97.
Kloppenborg RP, Nederkoorn PJ, Geerlings MI, van der Berg, E (2014) Presence and progres-
sion of white matter hyperintensities and cognition. Neurology 82:2127–2138.
270
Kochunov P, Coyle T, Lancaster J, Robin DA, Hardies J, Kochunov V, Bartzokis G, Stanley

J, Royall D, Schlosser AE, Null M, Fox PT (2010) Processing speed is correlated with
cerebral health markers in the frontal lobes as quantified by neuroimaging. NeuroImage
49:1190–1199.
Laukka EJ, Lovden M, Kalpouzos G, Li T-Q, Jonsson T, Wahlund L-O, Fratiglioni L, Backman
L (2013) Associations between white matter microstructure and cognitive performance in
old and very old age. PLoS One 8:e81419.
Li YO, Yang FG, Nguyen CT, Cooper SR, LaHue SC, Venugopal S, Mukherjee P (2012)
Independent component analysis of DTI reveals multivariate microstructural correlations
of white matter in the human brain. Hum Brain Mapp 33:1431–1451.
extraction: What’s change got to do with it? Psychology and Aging 26:34–47.
Lockhart SN, Roach AE, Luck SJ, Geng J, Beckett L, Carmichael O, DeCarli C (2014) White
matter hyperintensities are associated with visual search behavior independent of general-
ized slowing in aging. Neuropsychologia 52:93–101.
Lovden M, Laukka EJ, Rieckmann A, Kalpouzos G, Li TQ, Jonsson T, Wahlund LO, Fratiglioni
L, Backman L (2013) The dimensionality of between-person differences in white matter
microstructure in old age. Hum Brain Mapp 34:1386–1398.
Lu PH, Lee GJ, Raven EP, Tingus K, Khoo T., Thompson PM, Bartzokis G (2011) Age-related
slowing in cognitive processing speed is associated with myelin integrity in a very healthy
elderly sample. J Clin Exp Neuro 33:1059–1068.
MacKinnon DP, Fairchild AJ, Fritz MS (2007) Mediation analysis. Ann Rev Psychol
58: 593–614.
Madden DJ, Parks EL, Davis SW, Diaz MT, Potter GG, Chou Y, Chen N, Cabeza, R (2014)
Age mediation of frontoparietal activation during visual feature search. Neuroimage
102:262–274.
Maillard P, Carmichael O, Fletcher E, Reed B, Mungas D, DeCarli C (2012) Coevolution of
white matter hyperintensities and cognition in the elderly. Neurology 79:442–448.
Nilsson J, Thomas AJ, O’Brien JT, Gallagher P. (2014). White matter and cognitive decline in
aging: A focus on processing speed and variability. J Int Neuropsych Soc 20:262–267.
Oosterman JM, Vogels RLC, van Harten B, Gouw AA,Poggesi A, Scheltens P, Kessels RPC,
Scherder EJA (2010) Assessing mental flexibility: Neuroanatomical and neuropsycho-
logical correlates of the trail making test in elderly people. The Clin Neuropsychologist
24:203–219.
O’Sullivan M, Jones DK, Summers PE, Morris RG, Williams SCR, Markus HS (2001) Evidence
for cortical “disconnection” as a mechanisms of age-related cognitive decline. Neurology
57:632–638.
Papp KV, Kaplan RF, Springate B, Moscufo N, Wakefield DB, Guttmann CRG, Wolfson L
(2014) Processing speed in normal aging: Effects of white matter hyperintensities and
hippocampal volume loss. Aging Neuropsych Cog 21:197–213.
Penke L, Maniega SM, Murray C, Gow AJ, Valdes Hernandez MC, Clayden JD, Starr
JM, Wardlaw JM, Bastin ME, Deary IJ (2010) A general factor of brain white matter
integrity predicts information processing speed in healthy older people. J. Neurosci,
30:7569–7574.
Penke L, Maniega SM, Bastin ME, Valdes Hernandez MC, Murray C, Royle NA, Starr JM,
Wardlaw JM, Deary IJ (2012) Brain white matter tract integrity as a neural foundation for
general intelligence. Molecular Psych 17:1026–1030.
Prins ND, van Dijk EJ, den Heijer T, Vermeer SE, Jolles J, Koudstaal PJ, Hofman A, Breteler
MMB (2005) Cerebral small-vessel disease and decline in information processing speed,
executive function and memory. Brain 128:2034–2041.
271
Rabbitt P, Scott M, Thacker N, Lowe C, Jackson A, Horan M, Pendleton N (2006) Losses in

gross brain volume and cerebral blood flow account for age-related differences in speed
but not in fluid intelligence. Neuropsych 20:549–557.
Salami A, Eriksson J, Nilsson LG, Nyberg L (2012) Age-related white matter microstructural
differences partly mediate age-related decline in processing speed but not cognition.
Salthouse TA (2000) Aging and measures of processing speed. Biol Psych 54:35–54.
Salthouse TA (2009) Decomposing age correlations on neuropsychological and cognitive vari-
ables. J Int Neuropsych Soc 15:650–661.
Salthouse TA (2011) Neuroanatomical substrates of age-related cognitive decline. Psych Bull
137:753–784.
Salthouse, TA (2012) How general are effects of trait anxiety and depressive symptoms on cog-
nitive functioning? Emotion 12:1075–1084.
Salthouse TA (2014) Correlates of cognitive change. J Exp Psych: Gen 143:1026–1048.
Salthouse TA, Ferrer-Caja E (2003) What needs to be explained to account for age-related
effects on multiple cognitive variables? Psych Aging 18:91–110.
Samanez-Larkin GR, Levens SM, Perry LM, Dougherty RF, Knutson B (2012). Frontostriatal
white matter integrity mediates adult age differences in probabilistic reward learning. J
Neurosci 32:5333–5337.
Sasson E, Doniger GM, Pasternak O, Tarrasch R, Assaf Y (2012) Structural correlates of
cognitive domains in normal aging with diffusion tensor imaging. Brain Struct Funct
217:503–515.
Sasson E, Doniger GM, Pasternak O, Tarrasch R, Assaf Y (2013) White matter correlates of
cognitive domains in normal aging with diffusion tensor imaging. Front Neurosci 7:32.
Schaie KW (1989). Perceptual speed in adulthood: Cross-sectional and longitudinal studies.
Schiavone F, Charlton RA, Barrick TR, Morris RG, Markus HS (2009) Imaging age-related
cognitive decline: a comparison of diffusion tensor and magnetization transfer MRI. J
Magn Reson Imaging 29:23–30.
Schmidt R, Fazekas F, Offenbacher H, Dusek T, Zach E, Reinhart B, Grieshofer P, Freidl
W, Eber B, Schumacher M, Koch M, Lechner H (1993) Neuropsychologic correlates
of MRI white matter hyperintensities: A study of 150 normal volunteers. Neurology
43:2490–2494.
Smith EE, Salat DH, Jeng J, McCreary CR, Fischl B, Schmahmann JD, Dickerson BC,
Viswanathan A, Albert MS, Blacker D, Greenberg SM (2011) Correlations between MRI
white matter lesion location and executive function and episodic memory. Neurology
76:1492–1499.
Turken AU, Whitfield-Gabrieli S, Bammer R, Baldo JV, Dronkers NF, Gabrieli JDE (2008)
Cognitive processing speed and the structure of white matter pathways: Convergent evi-
dence from normal variation and lesion studies. NeuroImage 42:1032–1044.
Ullen F, Forsman L, Blom O, Karabanov A, Madison G (2008) Intelligence and variability
in a simple timing task share neural substrates in the prefrontal white matter. J Neurosci
28:4238–4243.
Valdes Hernandez MD, Booth T, Murray C, Gow AJ, Penke L, Morris Z, Maniega SM, Royle
NA, Aribisala BS, Bastin ME, Starr JM, Deary IJ, Wardlaw JM (2013) Brain white
matter damage in aging and cognitive ability in youth and older age. Neurobiol Aging
34:2740–2747.
Van den Heuvel DMJ, ten Dam VH, de Craen AJM, Admiraal-Behloul F, Olofsen H, Bollen
ELEM, Jolles J, Murray HM, Blauw GJ, Westendorp RGJ, van Buchem MA (2006)
Increase in periventricular white matter hyperintensities parallels decline in mental
272
processing speed in a non-demented elderly population. J Neurol Neurosurg Psychiatry

777:149–153.
Verhaeghen P, Salthouse TA (1997) Meta- analyses of age-
cognition relations in adult-
hood: estimates of linear and nonlinear age effects and structural models. Psychol Bull
122:231–249.
Vernooij MW, Ikram MA, Vrooman HA, Wielopolski PA, Krestin GP, Hofman A, Niessen WJ,
Van der Lugt A, Breteler MM (2009) White matter microstructural integrity and cognitive
function in a general elderly population. Arch Gen Psychiatry 66:545–553.
Wahl M, Li Y-O, Ng J, LaHue SC, Cooper SR, SHerr EH, Mukherjee P (2010) Microstructural
correlations of white matter tracts in the human brain. NeuroImage 51:531–541.
Wahlovd KB, Fjell AM (2007) White matter volume predicts reaction time instability.
Waiter GD, Fox HC, Murray AD, Starr JM, Staff RT, Bourne VJ, Whalley LJ, Deary IJ (2008)
Is retaining the youthful functional anatomy underlying speed of information processing a
signature of successful cognitive ageing? An event-related fMRI study of inspection time
performance. NeurImage 41:581–595.
Wen W, Zhu W, He Y, Kochan NA, Reppermund S, Slavin MJ, Brodaty H, Crawford J, Xia A,
Sachdev P (2011) Discrete neuroanatomical networks are associated with specific cogni-
tive abilities in old age. J Neurosci 31(4):1204–1212.
Wu K, Taki Y, Sato K, Qi H, Kawashima R, Fukuda H (2013) A longitudinal study of structural
brain network changes with normal aging. Front Hum Neurosci 7: Article 113.
Ystad M, Hodneland E, Adolfsdottir S, Haasz J, Lundervold AJ, Eichele T, Lundervold A
(2011) Cortico-strial connectivity and cognition in normal aging: A combined DTI and
resting state fMRI study. NeuroImage 55:24–31.
Zahr NM, Rohlfing T, Pfefferbaum A, Sullivan EV (2009) Problem solving, working mem-
ory, and motor correlates of association and commissural fiber bundles in normal aging: a
quantitative fiber tracking study. Neuroimage 44:1050–1062.
273
11
The Aging Hippocampus

Linking Animal and Human Research
Shauna M. Stark
Craig E. L. Stark
O ne of the key concerns of older adults is the experience of memory

loss, both in the normal course of aging and as it is associated with
neural degenerative disease. Age-related memory impairment affects various types of
memory but leaves others untouched. For example, there are impairments in episodic
memory—the knowledge of where or when information was encoded (Johnson et
al., 1993; Schacter et al., 1997). Older adults are more prone to false recollections,
miscombining features of different events that are confidently held as true (Koutstaal
et al., 2001; Lyle et al., 2006). Yet, other domains of memory remain unchanged with
age, such as procedural memory (Fleischman et al., 2004), while semantic knowledge
(e.g., vocabulary) actually improves (Verhaeghen, 2003). Many of the types of mem-
ory that demonstrate age-related decline are mediated by the hippocampus and other
medial temporal lobe structures, and the frontal lobes (Morrison and Baxter, 2012),
relying on similar underlying fundamental mnemonic processes.
While studies of human behavior and pathology have contributed greatly to our
understanding of the cognitive and neuroanatomical changes associated with aging,
there are limitations to studying humans that can be addressed by the use of animal
models of aging. First, since the structural and functional foundation of monkey and
rodent brains, particularly the medial temporal lobe, parallels that of humans (Clark
and Squire, 2013), we can meaningfully utilize techniques that are impossible, such as
directed lesions and pharmacogenetic manipulations, or rare, such as extensive elec-
trophysiological recordings, in humans. Second, many animals, rodents in particular,
age relatively quickly compared to humans, which allows us to examine the neural
273
274
and behavioral changes associated with the aging process in just a few years. Third,
rodents and nonâ•‰human primates show age-â•‰related cognitive decline, including effects
on memory (Rapp and Bachevalier, 2011). Fourth, the individual variability exhibited
by humans in brain pathology and cognition that increases with age also appears to
be present in other species (Gallagher and Rapp, 1997), allowing us to examine this
potentially critical variable. These parallels allow us to examine the neurobiological
and cognitive functions most affected by normal aging in other animals.
There are some caveats to consider in the use of animal models however. With the
sole exception of the degu (Octagon degus) (Van Groen et al., 2011), only humans
develop hallmark Alzheimer’s disease (AD) pathology (specifically, amyloid-â•‰β plaques
and tau-â•‰protein tangles) spontaneously with age. Several transgenic and non-â•‰transgenic
animal models of AD have been developed (Dodart and May, 2005), many of which
overexpress human β-â•‰amyloid precursor protein and mutant forms of tau to study the
pathogenesis of AD at the molecular, cellular, and behavioral levels. However, inas-
much as one of the goals of aging research involves identifying the differences in the
progression of normal aging from those associated with AD, the difference in the natural
occurrence of AD between humans and other animals must be considered. Importantly,
rodents do demonstrate a range of performance in a variety of memory tasks, which
parallels the individual differences in cognition observed in humans. Thus, while other
species do not naturally exhibit AD pathology, we advocate the use of animal â•‰models
in aging research as they provide an important contribution to understanding the under-
lying neural and cognitive changes associate with healthy aging.
Of course, for an animal model of aging to be informative, it must be clearly linked
with the human in as many ways as possible. By validating findings from the animal
model in humans where feasible, we gain confidence in the legitimacy of applying
other, less readily measured effects. While there are many other theories regarding
age-â•‰related changes in memory (e.g., recollection and familiarity), we focus here on
those that have been heavily explored in the animal literature. In this chapter, we will
review the neurobiological and memory changes associated with aging that have been
studied with animals. We will then review a cross-â•‰species model of neurocognitive
aging that emphasizes the underlying alterations in behavior, connectivity, and activ-
ity for both animals and humans. These age-â•‰related changes reflect a loss in the ability
to rapidly encode and retrieve specific details (particularly in the face of interference)
and a shift toward a greater reliance upon gist.
Hippocampus and Medial Temporal Lobe
Amnesic patients gave us the first insight into the role of the hippocampus and sur-
rounding medial temporal lobe cortex in the formation of long-â•‰ term memories
(Penfield and Milner, 1958). Most notably, in the 1950s the hippocampus and most of
the surrounding medial temporal cortex was surgically removed in epileptic patient
H.M. in an attempt to cure his seizures (Scoville and Milner, 1957). His resulting
profound amnesia, an inability to learn new things and recall them at a later time,
illustrated the importance of these brain regions for memory encoding and retrieval.
The following decades of research in humans, nonhuman primates, and rodents have
contributed to an understanding that the structures in the medial temporal lobe (MTL)
275
The Aging Hippocampus 275
play a vital role in certain forms of memory (Morris, 2007). Fortuitously, these struc-
tures, and the forms of memory they support, are well preserved across the species
(Allen and Fortin, 2013).
Broadly speaking, the MTL consists of the hippocampus and the adjacent ento-
rhinal, perirhinal, and parahippocampal cortices (which together form the parahip-
pocampal gyrus). These structures form a loose hierarchy such that hippocampus
receives projections from entorhinal cortex, which in turn receives projections from
the perirhinal and parahippocampal cortices (Figure 11.1). These connections are
Auto-associative
Fibers
Schaffer
Collaterals
CA3 CA1
MS Sub-
Mossy (ACh) cort
Fibers
VTA
(DA)
DG ParS
PreS Sub
Perforant Path
II III IV-VI
Amyg. RSC
ERC
PHC/POR PRC
Neocortical
Association Areas
Figure 11.1 A simplified model of the anatomy of the medial temporal lobe (after Amaral and
Lavenex, 2007). Line thickness denotes the strength of the connectivity. On dashed lines, open
circles and closed circles denote inhibitory and excitatory modulation (structures in circles).
Gray connections have been observed in animal models to be affected by age (after Wilson
et al., 2006). White background connotes being part of the classic MTL memory system.
PHC/POR: parahippocampal/postrhinal cortex, PRC: perirhinal cortex, EC: entorhinal cortex,
Amyg: amygdala, RSC: retrosplenial cortex, Sub: subiculum, ParS: parasubiculum, PreS: pre-
subiculum, DG: dentate gyrus, MS: medial septum, VTA: ventral tegmental area.
276
bidirectional; feeding information forward into the hippocampus and then returning
it back through the cortices. In addition, heavy projections run between the perirhinal
and parahippocampal cortices, which in turn receive widespread projections from uni-
modal and polymodal areas in the frontal, temporal, and parietal lobes (Amaral, 1999;
Lavenex and Amaral, 2000). The hippocampus itself is not a monolithic structure.
The “hippocampus proper” consists of the dentate gyrus (DG) and the CA1, CA2, and
CA3 subfields and the “hippocampal formation” extends this to include the subicu-
lum (Amaral and Lavenex, 2007). These regions communicate via a “direct pathway”
in which information flows from layer 2 of the entorhinal cortex into CA3, to CA1
via the Schaffer Collaterals, and to the subiculum before returning to the entorhinal
cortex in layer 3. There is also an “indirect pathway,” which leaves entorhinal cortex
via the same perforant path connections as the direct pathway but branches off to the
DG before heading again into CA3 via the Mossy Fibers. These fibers terminate in
the CA3 with exceptionally strong “detonator synapses” (McNaughton and Morris,
1987b) that can forcibly impose a pattern of activity in CA3 (Ribak et al., 1985). This
indirect pathway (EC-â•‰DG, DG-â•‰CA3, CA3-â•‰CA1) has become known as the “trisynap-
tic loop” (Andersen et al., 1971). In addition to these pathways, however, information
can leave the EC (layer 3) and bypass the both the DG and CA3, heading directly into
CA1. It can also be further processed within CA3 via the extensive recurrent connec-
tions or auto-â•‰associative fibers. As can be seen in Figure 11.1, even this level of detail
reflects a mere sketch of the anatomy (and Figure 11.1 is less complex that what is
currently known; see Van Strien et al., 2009).
Computational Role of the Hippocampus in the MTL
The division of labor in the MTL has been a heavily debated issue in the cognitive
neuroscience and the neurobiology of learning and memory (Morris, 2007; Squire
et al., 2004). Theories have suggested that the hippocampus is tuned for “associa-
tions” rather than items (Brown and Aggleton, 2001), “configural memory” (O’Reilly
and Rudy, 2001), “relational” memory (Eichenbaum et al., 1994), “binding items in
context” (Diana et al., 2007), or that its role is a less specified one that in some ways
“combines and extends” the processing carried out in the structures of the parahip-
pocampal gyrus (Squire, 1992). A common theme that runs throughout these popular
theories of hippocampal function involves the ability to rapidly store arbitrary, asso-
ciative information in such a way that it is free from significant interference by and
with any similar information. Binding in context and relational memory, for example,
may represent the outcome of the rich, detailed representation formed as a result of
separating that experience from other, similar experiences. These theories diverge in
numerous places (e.g., what information is being stored and the nature of the associa-
tion), but this principle is highly prevalent.
What allows a system to rapidly store arbitrary, associative information? There is
a long history of computational models of the hippocampus that provides at least one
solid answer to that question and can give us insight into the potential roles the various
subfields and projections may play in hippocampal function. For example, Marr (1971)
suggested that the recurrent collaterals of the CA3 region of the hippocampus enable it
277
to act as an auto-association network capable of pattern completion. This is the process

by which incomplete or degraded representations are filled in based on previously stored
representations—the retrieval of the missing portions of a memory given a partial cue
(e.g., remembering a person’s name upon seeing their face). Such a process is clearly
helpful at retrieval (as all memory cues are likely to be incomplete or degraded in some
manner) and may be helpful in determining whether something is new (and thereby
potentially worth storing), but does not lend itself to rapid, associative encoding.
While pattern completion is designed to increase the similarity between the current
item and a previously stored item, pattern separation is designed to do the converse.
In a large class of computational models (Rolls, 2013), similar representations (of
similar information or events) are pulled apart in the dentate gyrus (DG) into non-
overlapping (orthogonalized) representations (Treves and Rolls, 1994). Evidence for
this theory comes from several anatomical findings and from cellular recordings. For
example, the DG has approximately four to five times as many granule cells in it as
there are EC cells projecting to it (Boss et al., 1985; Amaral et al., 1990), consistent
with the notion that the DG is responsible for reducing the similarity of input patterns
(e.g., pattern separation; McNaughton and Morris, 1987a; McClelland and Goddard,
1996). In addition, activity in the DG is notoriously sparse with only a small propor-
tion of the cells active at any one time (Chawla et al., 2005). Finally, DG cells have
been found to alter their firing in response to small changes in input that are insuf-
ficient to alter firing properties elsewhere in the hippocampus (Leutgeb et al., 2007;
Neuneubel and Knierim, 2014) and BOLD fMRI signals have likewise shown the DG/
CA3 to be more responsive to stimulus change than CA1 (Lacy et al., 2011; Bakker et
al., 2008; see Yassa and Stark, 2011 for review).
This process of pattern separation in the hippocampus is a potential computational
mechanism that allows the system to rapidly store arbitrary associative informa-
tion, thereby giving the hippocampus a preferential role in multiple forms of mem-
ory, such as source and recollective memory (Norman and O’Reilly, 2003; Norman,
2010; Yassa and Stark, 2011). Source and recollective memory refer to components
of episodic memory that are both designed to probe the unique features of an epi-
sodic memory trace, which differentiate it from other similar experiences. According
to the Complementary Learning Systems (CLS) theory, the hippocampal system of
pattern separation is necessary to rapidly store novel, arbitrary, associative informa-
tion without interference from (or to) similar information (McClelland et al., 1995;
Norman and O’Reilly, 2003). Thus, damage to such a system, by age perhaps, would
lead to predictions of impairments on those forms of memory which all place strong
demands on pattern separation and to a greater reliance on gist rather than specific
details (Yassa and Stark, 2011).
However, it is worth noting that the computations of pattern completion and pat-
tern separation can certainly act in parallel and there are clear opportunities for them
to interact or compete as to which one sends the strongest signal forward to the next
stage. For example, in the CA3 subfield, the auto-associative recurrent collaterals are
typically hypothesized to provide a mechanism for pattern completion. If we assume
a pattern separation signal arriving from the Mossy Fibers from the dentate, we have
a clear example of the computations being performed separately, but interacting or
competing to set activity in CA3.
278
Neurobiological Changes in the MTL Associated

with Aging in Animal Models
Human imaging studies have reported that hippocampal volume is reduced in older
adults (Raz et al., 2005), at a rate of approximately 1.5% per year in adults aged
70–â•‰89 (Jack et al., 1998). This observation not only draws attention to the role of
the hippocampus in cognitive aging, but also suggests that neuronal loss may play a
critical role. After all, a reduction in volume would be a clear outcome of neuronal
loss. However, this example demonstrates the power of combining human research
with animal models. While early reports indicated that age-â•‰related brain changes were
the result of gross neuronal loss (Brizzee, 1981), there is very strong evidence that
this brain volume loss is not due to fewer neurons (Burke and Barnes, 2006). Across
species, there are consistent reports that there is not significant cell death in the hip-
pocampus or neocortex in humans (West, 1993; Pakkenberg and Gundersen, 1997),
nonâ•‰human primates (Merrill et al., 2000; Keuker et al., 2003), or rodents (Rapp and
M, 1996; Merrill et al., 2001). One exception may be area 8A of the dorsolateral
PFC, which exhibits a 30% reduction in neuronal number in aged nonhuman primates
(Smith et al., 2004). There are several of other neural changes however that may be
responsible for this volume loss, such as synaptic loss, reductions in microglia, demy-
elination, and other connectivity changes.
Connectivity Changes
There is evidence for synaptic changes in animal models, observed in select areas of
the hippocampus (Nicholson et al., 2004). In particular, there are changes in the con-
nectivity of the DG and the CA3 subfields. While the synapses in the CA3 subfield
of the hippocampus that form the recurrent collaterals in the auto-â•‰associative network
are not reduced in aging (Smith et al., 2000), the number of synapses from the ento-
rhinal cortex (EC) into the DG via the perforant path is reduced (Scheff et al., 2006).
Early observations showed a 27% reduction in synapse number in the middle molecu-
lar layer of the dentate gyrus of aged rats (Bondareff and Geinisman, 1976; Geinisman
et al., 1977), with many of these synapses becoming nonâ•‰functional or silent. This
reduction is now known to be from the layer II EC to both the DG and the CA3—â•‰pre-
sumably via the perforant path (Smith et al., 2000). Likewise, there are age-â•‰related
increases in dendritic length and branching in layer II of the EC, the origin of the per-
forant pathway to the DG (Buell and Coleman, 1979; Buell and Coleman, 1981; Flood
et al., 1987). In contrast, these changes are not observed in hippocampal subregions
CA1 (Hanks and Flood, 1991), CA3 (Flood et al., 1987), or the subiculum (Flood,
1991). Outside of the MTL, dendritic branching of pyramidal neurons decreases with
age for both apical and basal dendrites in superficial layers of the cortex (Grill and
Riddle, 2002), the anterior cingulate of the rat (Markham and Juraska, 2002), and the
medial prefrontal cortex (PFC) in the human (de Brabander et al., 1998). Finally, there
is an age-â•‰related reduction in spine density in the subiculum in nonâ•‰human primates
(Uemura, 1985), but no reduction in spine density in the DG (Curcio and Hinds, 1983)
or CA1 (Markham et al., 2005).
â•‡ 279
The Aging Hippocampusâ•… 279
Several other relevant connections have shown age-â•‰related changes as well. The
fornix connects the hippocampus bidirectionally with numerous subcortical struc-
tures. In both the rat (Naranjo and Greene, 1977) and monkey (Peters et al., 2010),
age-â•‰related axonal degeneration has been found in the fornix and has been associated
with age-â•‰related loss of myelinated fibers and alterations in the myelin sheath (Peters
et al., 2010). Similar changes have been found in the hippocampal cingulum, which
connects the hippocampus and EC to the retrosplenial cortex and other portions of the
cingulate cortex. Here, there have been reports of age-â•‰related declines in fractional
anisotropy (FA) using diffusion tensor imaging (DTI) imaging in monkeys (Makris
et al., 2007), loss of myelinated fibers (Bowley et al., 2010), and axonal degeneration
(Naranjo and Greene, 1977).
This age-â•‰related change is not endemic, however, as many connections show no
sign of age-â•‰related changes (Burke and Barnes, 2010). For example, synapses in the
recurrent collaterals forming the CA3 the auto-â•‰associative network are not reduced in
aging (Smith et al., 2000). Likewise, the Schaffer Collaterals do not show age-â•‰related
decline (Barnes et al., 2000b). Thus, large components of the hippocampal circuitry
remain structurally unaffected.
Neuromodulatory Changes
There are also age-â•‰related shifts in neuromodulatory influences in the hippocampus.
The cholinergic input from the medial septum reduced in aged rats (Perry et al., 1992),
the extent of which is correlated with the degree of memory impairment (Chouinard
et al., 1995; Sugaya et al., 1998). Furthermore, reducing cholinergic input replicates
age-â•‰related deficits (Ikonen et al., 2002) while activating the septum using a choliner-
gic agonist reverses them (Sava and Markus, 2008). Aging also leads to a reduction in
dopamine producing neurons in the ventral tegmental area (VTA) in monkeys (Siddiqi
et al., 1999) leading to a reduction in dopamine levels in the hippocampus that also
correlates with memory impairments (Stemmelin et al., 2000). Thus, aging is asso-
ciated with changes in the neuromodulatory context the hippocampus is operating
in, leading to alterations of the dynamics of information processing throughout this
complex circuit.
Neurophysiological Changes
In addition to synaptic changes associated with aging, deficits in plasticity also occur.
Electrophysiological data in aged rats show reductions in field excitatory postâ•‰synaptic
potentials recorded in the DG (Barnes, 1979; Barnes and McNaughton, 1980), as well
as presynaptic fiber potentials at the perforant path–â•‰DG synapse (Barnes et al., 2000b;
Dieguez and Barea-â•‰Rodriguez, 2004; Burke and Barnes, 2010). The DG also requires
greater stimulation to induce LTP (Barnes et al., 2000a) and potentiation decays faster
in aged rats than in young rats (Barnes, 1979). Similarly, there are smaller proportions
of neurons that express Arc in the granule cells of the dentate gyrus (Small et al.,
2004). Specifically, aged pyramidal cells in CA1 are less excitable, firing fewer action
potentials than young neurons (Moyer et al., 2011), while the firing rates of CA3
280
pyramidal neurons are higher in aged than young rats (Wilson et al., 2005). Likewise,
there are decreases in gene expression associated with aging (Berchtold et al., 2013),
such as a smaller proportion of neurons that express the plasticity-â•‰related Arc in the
granule cells of the dentate gyrus (Small et al., 2004) and reduced reelin expression in
the entorhinal cortex (Stranahan et al., 2010).
Neurogenesis and Synaptogenesis Changes

It is now well established that the DG acquires new neurons (neurogenesis) throughout
the lifespan in rodents (Ming and Song, 2005; Li et al., 2009) and humans (Spalding
et al., 2013). Manipulations that decrease neurogenesis typically lead to poorer cog-
nitive performance in older animals, while increasing neurogenesis tends to improve
learning and memory (Kempermann et al., 1998; van Praag et al., 2005; Montaron
et al., 2006). However, other studies have demonstrated that rats with the lowest lev-
els of neurogenesis in the DG have better spatial memory than those with more new
neurons (Bizon and Gallagher, 2003; Bizon et al., 2004). Given these contradictory
results, the role of neurogenesis in age-â•‰related cognitive is still a subject of investiga-
tion. There are several factors that contribute to the survival and integration of newborn
neurons into the brain circuitry, including, but not limited to, proliferation rate (which
can be modulated by exercise and environmental enrichment), availability of oxygen,
changes in vasculature, and levels of steroid hormones and growth factors (Artegiani
and Calegari, 2012). While the decrease in neurogenesis is not correlated with memory
performance on some tasks (Bizon et al., 2004), the maturation and survival of new-
born neurons may be dependent on new learning (Kempermann et al., 1997).
Synaptogenesis refers to synaptic connections formed between neurons, which dis-
plays a mild to severe decline during aging (Adams and Jones, 1982), including in the
DG (Geinisman et al., 1992). Interestingly, while there is a decrease in the density of
synapses, particularly in aged animals who display memory impariments (Nicholson
et al., 2004). There is evidence that synaptogenesis increases as a function of behav-
ioral learning in CA3 (Kumazawa-â•‰Manita et al., 2013) and CA1 (Wenzel et al., 1980).
Synaptogenesis may be play a critical role in the integration of new neurons into the
existing structure of the brain, but unfortunately, markers for newly established syn-
apses are not yet available. However, using multiple-â•‰synapse boutons (MSBs) as a
proxy measure revealed the number of MSBs is reduced in the aging monkey dentate
gyrus and is correlated with deficits in recognition memory (Hara et al., 2011).
MTL Changes Outside the Hippocampus

In addition to hippocampal dysfunction, there is evidence that age-â•‰related changes
extend into the perirhinal cortex (PRC) (Burke et al., 2010; Burke et al., 2011). Like
the hippocampus, age-â•‰related effects in the PRC are not the result of a loss of neurons
(Rapp et al., 2002). Instead, there are several cellular and molecular changes includ-
ing reductions in calcium binding proteins (Moyer et al., 2011) and glutamate (Liu
et al., 2008) along with effects on LTP and LTD induction and maintenance (Burke
and Barnes, 2006) that could certainly lead to an age-â•‰related disruption of function
â•‡ 281
in the PRC. When coupled with its putative role in object recognition, this raises the
compelling possibility that age-â•‰related changes in the PRC contribute to age-â•‰related
changes in memory (Burke et al., 2012).
Changes in the Prefrontal Cortex

While ample evidence exists to support memory encoding and retrieval dependence
on the MTL, the PFC is also critically involved in certain aspects of memory that
exhibit declines in healthy aging, such as memory for the source of acquired informa-
tion (Naveh-â•‰Benjamin et al., 2004) and for temporal order (Giovanello and Schacter,
2012; Tolentino et al., 2012; Pirogovsky et al., 2013). Aged monkeys demonstrate
deficits in delayed response tests (i.e. delayed non-â•‰match to sample), though they per-
form normally on standard tests of recognition (Rapp and Amaral, 1989; Bachevalier
et al., 1999), which may require PFC function. This standard delayed response testing
may tap memory for temporal order (Rapp and Amaral, 1989) and incorporate a spa-
tial component that is thought to engage processing of the dorsolateral PFC (Wilson
et al., 1993). Additional signs of prefrontal declines in aged monkeys include marked
perseveration and behavioral rigidity when faced with shifting task contingencies
(Gallagher and Rapp, 1997).
The rodent PFC is not as complex as the primate, but many of the neuroanatomical
and functional characteristics are preserved, allowing for meaningful cross-â•‰species
comparisons related to changes in PFC across the lifespan. The medial portion of
rodent PFC, which includes the anterior cingulate, and prelimbic and infralimbic cor-
tices, receives afferents from thalamic nuclei and limbic structures and sends efferents
to caudate-â•‰putamen and nucleus accumbens, sharing strong anatomical homology
with primate dorsolateral PFC (Uylings et al., 2003). Structural alterations in the
prefrontal cortex have been identified, including a substantial decline in the density
of prefrontal cortex synapses, striking changes in dendritic architecture, and marked
abnormalities in the myelination of axons (Peters et al., 1998; Morrison and Baxter,
2012). Similar to the synaptic changes in the hippocampus, there is evidence of a
functional synaptic imbalance toward inhibitory activity that corresponds to cognitive
impairment in aged rats, while there was an increase in excitatory synaptic tone in
aged unimpaired rats (Bories et al., 2013). In addition, there are reports of a significant
age-â•‰related decline in the number of principal neurons in the dorsal PFC in rodents
(Stranahan et al., 2012).
Implications of Neurobiological Changes

Together, these findings suggest a clear disruption of processing within the hippo-
campus associated with age, which extends to the surrounding medial temporal and
frontal lobes to create a memory system that is vulnerable to age-â•‰related changes. The
loss is not simply a reduction in overall capacity of the system (as one might assume
were there an evenly distributed loss of neurons) but rather a disruption of specific
aspects. As we will argue in more detail later, these changes observed in the animal
suggests a reduction in the capacity of the system to perform pattern separation and a
282
resulting shift in its behavior toward pattern completion. The behavioral result of this
alteration might be seen as a loss in the ability to rapidly encode and retrieve specific
details (especially in the face of interference)—â•‰hallmarks of temporal memory, source
memory, and recollection—â•‰and a shift toward a greater reliance upon gist.
Behavioral Changes
In humans, many tests of memory function rely on learning and later recall of verbal
or verbalizable material, such as word pairs, short stories, and face-â•‰name pairs. In ani-
mal models however, learning and memory must be assessed in ways that the animal
can learn and express, thus the heavy reliance on mazes or reward-â•‰based contingency
learning. It is important to understand the tasks used to probe memory in these animal
models in order to assess the parallels with human behavior, which often relies on
very different kinds of tasks.
In rodents, tasks that tap spatial memory have been heavily used, such as the Morris
water maze (Morris, 1981) and the Barnes maze (Barnes, 1979). In the water maze,
opaque water is used to conceal an escape platform submerged just below the surface
of the water. By placing the rat in the water at variable starting positions, memory for
the spatial location of the platform (relative to extra-â•‰maze spatial cues) can be mea-
sured as time to find the platform, distance swam, or time searching in the location
of the platform once it has been further submerged (D’Hooge and De Deyn, 2001).
Performance on this task is often viewed as a defining feature of age-â•‰related memory
loss in the rodent (Gallagher and Rapp, 1997). The Barnes maze was designed to take
advantage of rodent’s natural aversion to open, well-â•‰lit areas. The animal is placed in
a circular platform with many openings around the circumference, only one of which
will allow the animal to escape into a dark box beneath the platform’s surface. By
recording from neurons in the hippocampus while rodents performed this task, Barnes
(1979) first demonstrated that LTP decayed approximately twice as fast in older rats
than younger rats and correlated with the memory of the correct escape hole in the
Barnes maze.
Place fields are another way to assess neural changes underlying spatial memory
in rodents. While recording from the hippocampus of a freely roaming rat, O’Keefe
and Dostrovsky (1971) discovered that individual hippocampal cells would fire only
when the animal was in a particular location in a given environment. Recording from
these place fields can be used to reconstruct where an animal is in space (Wilson and
McNaughton, 1993), creating the concept of a cognitive map (O’Keefe and Nadel,
1978). Place cell maps of older rats are less stable than those for younger rats, with
a 30% chance of older rats retrieving an incorrect hippocampal map (Barnes et al.,
1997). Part of this between-â•‰session instability may be due to faulty LTP mechanisms
in aging, consistent with the outcomes of blocking NMDA receptors in young animals
(Kentros et al., 1998). Similarly, the size of the place fields in CA1 expands in young
rats during repeated runs around a track, but this place field expansion is significantly
decreased in aged rats (Shen et al., 1997). Again, these results are likely due to LTP
deficits, as similar results are shown in young rats given an NMDA receptor agonist
(Ekstrom et al., 2001). These cognitive maps can also “remap” when the environment
283
is changed (Muller and Kubie, 1987). This ability to rapidly store a new representa-
tion in response to a potentially small change in the environment is critical for epi-
sodic memory (Colgin et al., 2008). In aged rodents, place field maps are abnormally
“rigid” and resistant to change in the environment (Tanila et al., 1997). This rigidity—
representing a new environment as if it were an old one—is predictive of aged rats’
spatial memory performance on the Morris water maze (Wilson et al., 2003).
Another task that has been used to assess memory changes associated with aging
is the delayed non-match to sample (DNMS) task, using trial-unique stimuli (Mishkin
and Delacour, 1975). In the DNMS task, a novel sample stimulus is presented to
the subject. After a delay, the sample is presented again, along with a new stimulus
and the subject is then rewarded for selecting the new stimulus (the “nonmatching to
sample”). This task takes advantage of the natural novelty-seeking behavior in both
primates and rodents. Aged monkeys demonstrated impairments on this task with
increasing delays between the sample and the test phase, from 60 seconds to 10 min-
utes (Rapp and Amaral, 1991). These findings are consistent with animals that have
medial temporal lobe damage such as perirhinal lesions, and to a lesser degree, hip-
pocampal lesions (Baxter and Murray, 2001; Christie et al., 2008; Clark et al., 2001;
Nielsen et al., 2009). Further studies have identified that both synaptic density and
overall spine size in the dorsolateral PFC of rhesus monkeys were predictive of the
number of trials required for the acquisition of DNMS (Dumitriu et al., 2010).
One final task worth noting is the delayed response (DR) task. Like the DNMS,
there is a sample phase, a delay, and a match-nonmatch test phase, but here the reward
is hidden behind one of two spatial locations (with the location varying from trial to
trial, providing substantial interference). This task is sensitive to aging (Bachevalier
et al., 1991) and places strong demands on both the hippocampus and the dorsolateral
prefrontal cortex. For example, in aged monkeys, hippocampal volume correlates with
acquisition of and performance in the DR task (Shamy et al., 2011).
Many of these tasks have analogous human versions (Monk et al., 2002; Schoenfeld
et al., 2014), emphasizing their utility in tapping into the same underlying mnemonic
processes across species. Understanding the tasks used in animals is important for
evaluating the age-related changes associated with their behavioral performance.
These tasks, among others, are critical for testing predictions made by a cross-species
model of age-related memory decline.
A Cross-Species Model of Hippocampal Aging

and Memory Decline
Wilson et al. (2006) proposed a model that drew upon the computational models
and age-related neurobiological changes observed in animals to account for age-
related memory decline. They were particularly interested in the frequent finding
that the ability to form memories of new events and associations declines with age,
while memories of older events are often preserved (Burke and Mackay, 1997).
This model was designed to account for three key characteristics of episodic mem-
ory that occur in normal aging: 1) loss of contextual details, 2) susceptibility to
interference, and 3) an attenuated response to novelty. For example, older adults
284
are impaired on recalling contextual or source details, such as the color, case, or
font of words (Zacks et al., 2000) or location of an item (Chalfonte and Johnson,
1996). Older adults are disproportionately affected by interference from irrelevant
information, such as proactive interference when earlier learned memories inter-
fere with more recent memories (Lustig and Hasher, 2001). Finally, older adults
have a reduced novelty response, with smaller physical startle reactions to sudden
sounds (Ellwanger et al., 2003) and attenuated electrophysiological responses to
new items (Friedman, 2000).
To account for these findings, the model focuses on age-related changes within the
hippocampal circuit and the projections into the hippocampus (summarized in the gray
arrows in Figure 11.1). If the dentate gyrus is required for pattern separation and if pat-
tern separation is a computational requirement of these behavioral abilities, there should
be a clear set of relationships that could be observed. The data reviewed above are in
accord with this model, demonstrating, for example that the DG receives approximately
one-third fewer synaptic contacts from the EC than in young rats via the perforant path
(Geinisman et al., 1992), and that the extent of the synaptic reduction correlates with the
degree of spatial memory impairment in aged rats (Smith et al., 2000).
Reduced input to the DG is coupled with a decrease in modulation by inhibitory
neurons in this region (Vela et al., 2003) resulting in a hypoactive DG. As discussed
earlier, the DG granule cells are capable of performing especially strong pattern sepa-
ration on the distributed representation coming from the entorhinal neurons. In aging,
the reduction in input from the EC results and decreased interneuron activity in the
DG may result in a failure of the DG to reduce the similarity of input patterns and
project this on to the CA3, leading to a decrease in pattern separation performance.
In addition, there is a decrease in cholinergic modulation in the hippocampus in
normal aging (Perry et al., 1992). This reduction in cholinergic input creates a hyper-
activity of the aged CA3, resulting in high firing rates in animals (Hasselmo et al.,
1995). Reduced inputs from the EC and reduced cholinergic modulation might impair
processing, forcing CA3 to rely on its own internal storage of the external world.
Reduced cholinergic input releases the CA3 auto-associative fibers from inhibition,
thereby promoting completion of the familiar pattern with an existing representation
(Hasselmo, 1995).
Finally, age-related changes in CA1 are a consequence of the input from EC and
CA3, but also changes that are intrinsic to the CA1 itself. The CA1 subregion receives
dopaminergic projections that are reduced in aging (Stemmelin et al., 2000), the
extent of which correlates with memory ability in animals (Salvatore et al., 2003). In
addition, dopamine receptors are reduced in number in the aged CA1 hippocampus
(Hemby et al., 2003). This age-related modulation of dopamine in the CA1 may con-
tribute to impairments in synaptic plasticity (Lisman and Grace, 2005). This weak-
ened plasticity in aged CA1 may compromise its ability to encode new information
with sufficient information to form new memories. As a consequence, CA1 may shift
its reliance from inputs from CA3 to the direct inputs from layer III of the EC (Brun et
al., 2002). This shift in the reliance of its input may account for the shift in neurocog-
nitive aging toward pattern completion and away from pattern separation.
These data can be used to create a model of the input transformation for the young
and aged CA1 and CA3 (Figure 11.2). Here, in both young and aged individuals, the
â•‡ 285
DG
CA3
CA1
Change in output
Change in output
CA1
3
A
G/C
3
/CA
gD
DG
3
You
/CA
DG
Pattern separation
ed
Pattern completion
Ag
Change in input Change in input
similar distinct similar distinct
Figure 11.2â•‡ (Left) Model of DG, CA1 and CA3 transformations of input. The CA1 subregion
responds in linear manner to changes in input (a degree of match/â•‰mismatch) and is consistent
for young and aged individuals. In contrast, changes in the DG and CA3 shift the transforma-
tion for aged adults such that a greater change in the input is necessary to produce a change
in the output (e.g., consistent with a bias away from pattern separation processes). (Right) In
human BOLD fMRI, as the DG and CA3 are typically combined (see text), the transfer func-
tions are slightly changed but respect a young/â•‰aged difference.
CA1 subregion responds in a linear manner to changes in input with an output whose
difference is equal to the difference between the two inputs. The CA3 region responds
in a nonlinear manner to changes in the input, such that it requires a greater change in
input to affect a change in the output. In contrast to the CA1, the input transformation
in the CA3 is shifted for younger and older individuals. Specifically, the aged CA3
requires a greater change in the input to result in a change in the output (resulting in
greater pattern completion with aging).
If this model could be validated in humans, we would have a significantly more
detailed understanding of the nature and mechanism of age-â•‰related memory decline.
Several labs, including our own, have been testing the assumptions and predictions
of this model in humans. By necessity, the techniques that can be applied in humans
are primarily limited to behavior, magnetic resonance imaging (MRI) and positron
emission tomography (PET).
Alterations in Human Behavior

We developed a novel behavioral task to test the tenets of this neurobiologically based
â•‰
model. In the Mnemonic Similarity Task (MST) (Figure 11.3), participants encode
color pictures of everyday objects and are later administered a surprise recognition
test wherein they are shown repeated items, novel items, and critically, lures that are
similar to previously viewed items, but not exactly the same. Their task is to deter-
mine if each item is old (a repeated item), new (a novel item), or similar (a lure
item). Identifying these lure trials as “Old” (i.e., overgeneralization) is likely driven
by incorrect pattern completion processes. In contrast, discriminating these lure trials
286
Encoding phase Test phase

Indoor/Outdoor? Old/Similar/New?
Young: CA3/DG Aged: CA3/DG

0.02 lure activity 0.02 lure activity
Percent change in signal
Percent change in signal

0.01 0.01
0.00 0.00
–0.01 –0.01
–0.02 –0.02
L1 L2 L3 L4 L5 L1 L2 L3 L4 L5
FigureÂ€11.3â•‡ (Top) A schematic of the Mnemonic Similarity Task (Stark et al., 2013)—â•‰colored
boxes are for demonstration purposes only and were not shown to participants (gray—â•‰novel
items; black—â•‰repeated items; dashed—â•‰lure items). (Bottom) Activity in the aged DG/â•‰CA3DG/â•‰
CA3 tracks the lures as they increase in their mnemonic similarity (L1-â•‰least similar to L5—â•‰
most similar), in contrast to the young DG/â•‰CA3DG/â•‰CA3, which is not sensitive to lure similar-
ity (Yassa et al., 2011).
from the similar study item requires a distinct representation of the objects—â•‰a hall-
mark of pattern separation.
Utilizing this task, accurately identifying lures as similar items declines with age,
consistent with an age-â•‰related impairment in pattern separation (Kirwan and Stark,
2007; Toner et al., 2009; Stark et al., 2010; 2013; 2015). Similar results have been
reported utilizing variants of the task that emphasize spatial processing (Stark et al.,
2010; Reagh et al., 2013). Importantly, consistent with rodent studies demonstrating
variability in age-â•‰related memory impairments (Gallagher et al., 2006), a subset of
aging participants performed within the range of the young adults, while another sub-
set were markedly impaired. Furthermore, larger changes in the input (greater dissim-
ilarity) were required in order for older adults to encode new information as distinct
from previously learned information (Yassa et al., 2010a; Stark et al., 2013).
Alterations in Human Connectivity

The perforant pathway, which perforates the subiculum carrying input from the EC to
the hippocampal formation, is particularly vulnerable to age-â•‰related change (Witter,
â•‡ 287
2007). Diffusion tensor imaging (DTI) provides several indirect measures of white
matter integrity by measuring the diffusion properties of water molecules in fiber
tracts. We developed a high resolution DTI scan, demonstrating that the integrity of
this pathway is reduced in older adults (Yassa et al., 2010b). Furthermore, reduced
perforant path integrity in older adults was correlated with worse lure discrimination
performance (Yassa et al., 2011). Similar findings have been observed in older adults
with memory impairments, with synaptic loss in the perforant path found at autopsy
(Scheff et al., 2006) and marked atrophy in the perforant path in older adults with
dementia (Kalus et al., 2005).
Similarly, age-â•‰related axonal degeneration has been found in the fornix, which con-
nects the hippocampus bidirectionally with numerous subcortical structures (Naranjo
and Greene, 1977; Peters et al., 2010). DTI studies have demonstrated decreased
integrity of the fornix is related to impairments in recall and recognition memory
performance in individuals aged 9–â•‰93 (Lockhart et al., 2012; Sato et al., 2012). We
found reduced integrity in the fornix of healthy adults that was correlated with lure
discrimination performance on the MST (Bennett et al., 2013).
Alterations in Activity
Since extracellular recording in humans is rare, brain activity is primarily measured
using functional MRI. Historically, imaging hippocampal subfields has been chal-
lenging due to the small size of these structures. In recent years though, techniques
have been developed to examine functional activity from these subfields (Kirwan
et al., 2007; Yassa and Stark, 2009), though separating DG and CA3 is typically
not possible. Not only does the resolution we can scan at provide limitations in
separating the subfields, but the hemodynamic nature of the BOLD effect provides
challenges as well. While BOLD is related to synaptic activity (Logothetis and
Wandall, 2004), this relationship is known to be a complex one in the hippocam-
pus (Hargreaves et al., 2012). When coupled with the exceptional anatomy of the
DG (sparse activity with large amounts of inhibition) and CA3 (incredible number
of recurrent excitatory connections coupled with massive synaptic drive from the
Mossy fiber detonator synapses), meaningful interpretation of BOLD fMRI from
the two regions independently is wrought with difficulty and thus, they are often
combined (DG/â•‰CA3).
The connectivity studies demonstrated in humans that the input from EC to DG
appears to be weakened in older adults, but what about changes in activity within
DG? Using cerebral blood volume as a dependent measure in humans, Small et al.
(Small et al., 2002) showed reduced activity in the aging DG, but no differences in
CA1, subiculum, or EC. Additionally, there was a significant correlation between the
brain activity measure in the DG and memory performance, with poorer brain activity
corresponding to poorer memory scores (Small et al., 2002; Small et al., 2004). We
have reported age-â•‰related changes in the DG/â•‰CA3 during an object discrimination
task (the MST) (Yassa et al., 2010a; Yassa et al., 2011). These data demonstrate that
the failures to encode new memories in aged individuals arise in part from insufficient
pattern separation in a hypoactive DG.
The model also predicts hyperactivity in the CA3 as a result of reduced cholin-
ergic input releasing the CA3 auto-â•‰associative fibers from inhibition. Hippocampal
288
hyperactivity has been reported in older adults (Miller et al., 2008a) and in individuals
with mild cognitive impairment (Dickerson et al., 2004; Dickerson et al., 2005; Miller
et al., 2008b). Using high-resolution MRI, we found an increase in DG/CA3 activity
during a variant of the MST (Yassa et al., 2010a). These data are consistent with the
hypothesis that the aged CA3 demonstrates greater representational rigidity, requir-
ing a greater change in the input before stimuli can be orthogonalized, resulting in
decreased pattern separation (Yassa and Stark, 2011). In fact, reduction of this hyper-
activity in amnestic MCI using a low-dose antiepileptic was associated with improved
performance on the MST (Bakker et al., 2012).
It is worth noting that alternative hypotheses have been proposed to account for the
age-related hyperactivity observed in the hippocampus during human neuroimaging
(Miller et al., 2008a). For example, the compensation hypothesis proposes that over-
activation may reflect greater processing demand from a structure that is itself defi-
cient or it may be a result of extra effort required to process degraded input coming
from other structures (Reuter-Lorenz and Cappell, 2008). Hippocampal hyperactivity
is associated with worse performance on memory tasks, both in aging and in early
AD (Sperling, 2007). While the hyperactivity observed in CA3 may be the result of
reduced cholinergic input, hyperactivity in other brain regions likely results from a
variety of other causes.
In contrast to the age-related changes in DG/CA3, neuroimaging of the CA1 sub-
field of the hippocampus appears unaffected and is consistent with robust pattern
completion performance (Yassa et al., 2010a). Similarly, in young adults, activity was
found in CA1 during maintenance of items with overlapping features (Newmark et al.,
2013) and when a test item matched the studied sample item (Dudukovic et al., 2011).
Some structural studies have identified age-related changes in CA1 volume (Mueller
and Weiner, 2009), but others suggest that CA1 is spared in healthy aging and can be
used to identify changes associated with early memory impairment and AD (Frisoni
et al., 2008; Devanand et al., 2012).
Thus, this neurobiological model of aging provides a consistent framework for
evaluating neural and behavioral changes in animals and humans. We have chosen
to focus here on the hippocampus and subfield-level changes associated with aging,
largely because the animal models have focused on this system and because much of
the research in the human literature has been focused on age-related changes outside
of the hippocampus. In linking the animal-model data on age-related changes in the
hippocampus to humans, we can more strongly conclude that there are age-related
alterations in the hippocampus outside of disease or dementia. Looking forward,
while there are known age-related changes to the surrounding medial temporal lobe,
prefrontal cortex, and other cortical regions, these have been studied in less detail in
animal models of aging. Likewise, while some constructs such as recollection and
familiarity have been investigated in the rodent (Eichenbaum et al., 2010), these have
not yet been applied to the aging rodent. As the field of aging research advances, we
anticipate that investigations of temporal sequence learning in the prefrontal cortex
and functional specialization of the medial temporal cortices will expand to be incor-
porated into the animal models to make more contact with the human research in these
domains.
â•‡ 289
Conclusion
The comparison across species advances our understanding of the neurobiological

and cognitive alterations that occur over the course of healthy aging. The tools and
techniques that can be utilized in these different species complement and contrast
different aspects of these models of aging in a way that could not be achieved by
a single application. For example, lesion studies have limited utility for capturing
the performance of neural systems in the aged brain, which exhibits changes in neu-
rotransmitters, reduced synapses, and remodeling. The type of gross neuronal loss
produced by a lesion-â•‰based model cannot account for these subtle shifts and gradual
neurodegeneration. However, by incorporating a variety of methods, including lesion
techniques, we can develop complex models of age-â•‰related changes that can be rig-
orously tested across species. While differences between species certainly exist, all
of these species naturally exhibit individual differences in neurocognitive aging. This
natural variability may be important for dissociating the normal aging process from a
preclinical condition that heralds dementia.
With humans, there is always a concern that poor memory performance may
reflect early changes associated with dementia. It has been considered that
Alzheimer’s disease may be an inevitable consequence of aging if we lived long
enough. We can evaluate these hypotheses by examining aging in animals that do
not spontaneously develop AD and comparing their neural and cognitive changes
to those in humans. Using this technique, there is ample evidence that alterations
in the DG play a large role in normal aging, while changes in CA1 and EC are dra-
matically affected in AD. While the changes associated with aging and Alzheimer’s
disease may be superimposed, there does not appear to be a simple linear trajec-
tory wherein all humans ultimately contract AD. Rather, aging and AD appear to
be differentiable and have different neural signatures to at least some degree and
with great variability in the natural aging process (Wu et al., 2008; Nelson et al.,
2011). By studying this variability, we may gain a greater understanding of how to
overcome these age-â•‰related changes in memory that remain the primary cognitive
complaint as we all get older.
References
Adams, I, Jones, DG (1982) Quantitative ulstrastructural changes in rat cortical synapses during
early-â•‰, mid-â•‰and late-â•‰adulthood. Brain Research, 239:349–â•‰363.
Allen, TA, Fortin, NJ (2013) The evolution of episodic memory. Proc Natl Acad Sci U S A,
110:10379–â•‰10386.
Amaral, DG (1999) Introduction: what is where in the medial temporal lobe? Hippocampus, 9:1–â•‰6.
Amaral, DG, Ishizuka, N, Claiborne, B (1990) Neurons, numbers and the hippocampal net-
work. Prog Brain Res, 83:1–â•‰11.
Amaral, DG, Lavenex, P (2007) Hippocampal neuroanatomy. In: The Hippocampus Book
(Amaral, DG, Andersen, P, Bliss, T, Morris, RGM, O’Keefe, J, eds), pp. 37–â•‰ 114.
New York: Oxford University Press.
Andersen, P, Bliss, TVP, Skrede, KK (1971) Lamellar organization of hippocampal excitatory
pathways. Experimental Brain Research 13:222–â•‰238.
290
Artegiani, B, Calegari, F (2012) Age-related cognitive decline: can neural stem cells help us?
Aging, 4:176–186.
Bachevalier, J, Beauregard, M, Alvarado, MC (1999) Long-term effects of neonatal dam-
age to the hippocampal formation and amygdaloid complex on object discrimination
and object recognition in rhesus monkeys (Macaca mulatta). Behavioral Neuroscience,
113:1127–1151.
Bachevalier, J, Landis, LS, Walker, LC, Brickson, M, Mishkin, M, Price, DL, Cork, LC (1991)
Aged monkeys exhibit behavioral deficits indicative of widespread cerebral dysfunction.
Neurobiol Aging, 12:99–111.
Bakker, A, Kirwan, CB, Miller, M, Stark, CE (2008) Pattern separation in the human hippocam-
pal CA3 and dentate gyrus. Science, 319:1640–1642.
Bakker, A, Krauss, GL, Albert, MS, Speck, CL, Jones, LR, Stark, CE, Yassa, MA, Bassett, SS,
Shelton, AL, Gallagher, M (2012) Reduction of hippocampal hyperactivity improves cog-
nition in amnestic mild cognitive impairment. Neuron, 74:467–474.
Barnes, CA (1979) Memory deficits associated with senescence: A neurophysiological and
behavioral study in the rat. Journal of Comparative Physiological Psychology, 93:74–104.
Barnes, CA, McNaughton, BL (1980) Physiological compensation for loss of afferent synapses
in rat hippocampal granule cells during senescence. J Physiol, 309:473–485.
Barnes, CA, Rao, G, Houston, FP (2000a) LTP induction threshold change in old rats at the
perforant path--g ranule cell synapse. Neurobiol Aging, 21:613–620.
Barnes, CA, Rao, G, Orr, G (2000b) Age-related decrease in the Schaffer Collateral -evoked
EPSP in awake, freely behaving rats. Neural Plasticity, 7:167–178.
Barnes, CA, Suster, MS, Shen, J, McNaughton, BL (1997) Multistability of cognitive maps in
the hippocampus of old rats. Nature, 388:272–275.
Baxter, MG, Murray, EA (2001) Opposite relationship of hippocampal and rhinal cortex dam-
age to delayed nonmatching-to-sample deficits in monkeys. Hippocampus, 11:61–71.
Bennett, IJ, Huffman, DJ, Stark, CEL (2013) Integrity of limbic tracts contributes to epi-
sodic memory performance across the lifespan. Annual Meeting of the Society for
Neuroscience,
Berchtold, NC, Coleman, PD, Cribbs, DH, Rogers, J, Gillen, DL, Cotman, CW (2013) Synaptic
genes are extensively downregulated across multiple brain regionsin normal human aging
and Alzheimer’s disease. Neurobiol Aging, 34:1653–1661.
Bizon, JL, Gallagher, M (2003) Production of new cells in the rat dentate gyrus over the lifes-
pan: relation to cognitive decline. Eur J Neurosci, 18:215–219.
Bizon, JL, HJ, L, Gallagher, M (2004) Neurogenesis in a rat model of age-related cognitive
decline. Aging Cell, 3:227–234.
Bondareff, W, Geinisman, Y (1976) Loss of synapses in the dentate gyrus of the senescent rat.
American Journal of Anatomy, 145:129–136.
Bories, C, Husson, Z, Guitton, MJ, De Koninck, Y (2013) Differential balance of perfrontal syn-
aptic activity in successful versus unsuccessful cognitive aging. J NeuroSci, 33:1344–1356.
Boss, BD, Peterson, GM, Cowan, WM (1985) On the number of neurons in the dentate gyrus.
Brain Research, 338:144–150.
Bowley, MP, Cabral, H, Rosene, DL, Peters, A (2010) Age changes in myelinated nerve fibers
of the cingulate bundle and corpus callosum in the rhesus monkey. J Comp Neurol,
518:3046–3064.
Brizzee, KR (1981) Structural correlates of the aging process in the brain. Psychopharmacology
Bulletin, 17:43–52.
Brown, MW, Aggleton, JP (2001) Recognition memory: what are the roles of the perirhinal
cortex and hippocampus? Nat Rev Neurosci, 2:51–61.
291
Brun, VH, Otnass, MK, Molden, S, Steffenach, HA, Witter, MP, Moser, MB, Moser, EI (2002)
Place cells and place recognition maintained by direct entorhinal-hippocampal circuitry.
Science, 296:2243–2246.
Buell, SJ, Coleman, PD (1979) Dendritic growth in the aged human brain and failure of growth
in senile dementia. Science, 206:854–856.
Buell, SJ, Coleman, PD (1981) Quantitative evidence for selective dendritic growth in normal
human aging but not in senile dementia. Brain Research, 214:23–41.
Burke, DM, Mackay, DG (1997) Memory, language, and aging. Philos Trans R Soc Lond B
Biol Sci, 352:1845–1856.
Burke, SN, Barnes, CA (2006) Neural plasticity in the ageing brain. Nat Rev Neurosci, 7:30–40.
Burke, SN, Barnes, CA (2010) Senescent synapses and hippocampal circuit dynamics. Trends
Neurosci, 33:153–161.
Burke, SN, Ryan, L, Barnes, CA (2012) Characterizing cognitive aging of recognition memory
and related processes in animal models and in humans. Front Aging Neurosci, 4:15.
Burke, SN, Wallace, JL, Hartzell, AL, Nematollahi, S, Plange, K, Barnes, CA (2011) Age-
associated deficits in pattern separation functions of the perirhinal cortex: a cross-species
consensus. Behav Neurosci, 125:836–847.
Burke, SN, Wallace, JL, Nematollahi, S, Uprety, AR, Barnes, CA (2010) Pattern separation
deficits may contribute to age- associated recognition impairments. Behav Neurosci,
124:559–573.
Chalfonte, BL, Johnson, MK (1996) Feature memory and binding in young and older adults.
Memory and Cognition, 24:403–416.
Chawla, MK, Guzowski, JF, Ramirez-Amaya, V, Lipa, P, Hoffman, KL, Marriott, LK, Worley,
PF, McNaughton, BL, Barnes, CA (2005) Sparse, environmentally selective expression
of Arc RNA in the upper blade of the rodent fascia dentata by brief spatial experience.
Hippocampus, 15:579–586.
Chouinard, ML, M, G, Yasuda, RP, Wolfe, BB, McKinney, M (1995) Hippocampal muscarinic
receptor function in spatial learning-impaired aged rats. Neurobiol Aging, 16:955–963.
Christie, LA, Saunders, RC, Kowalska, DM, MacKay, WA, Head, E, Cotman, CW (2008)
Rhinal and dorsolateral prefrontal cortex lesions produce selective impairments in object
and spatial learning and memory in canines. J Comp Neurol, 511:257–270.
Clark, RE, Squire, LR (2013) Similarity in form and function of the hippocampus in rodents,
monkeys, and humans. Proc Natl Acad Sci U S A, 110:10365–10370.
Clark, RE, West, AN, Zola, SM, Squire, LR (2001) Rats with lesions of the hippocampus are
impaired on the delayed nonmatching-to-sample task. Hippocampus, 11: 176–186.
Colgin, LL, Moser, EI, Moser, MB (2008) Understanding memory through hippocampal
remapping. Trends in Neuroscience, 31:469–477.
Curcio, CA, Hinds, JW (1983) Stability of synaptic density and spine volume in dentate gyrus
of aged rats. Neurobiol Aging, 4:77–87.
D’Hooge, R, De Deyn, PP (2001) Applications of the Morris water maze in the study of learn-
ing and memory. Brain Research, 36:60–90.
de Brabander, JM, Kramers, RJ, Uylings, HB (1998) Layer- specific dendritic regres-
sion of pyramidal cells with ageing in the human prefrontal cortex. Eur J Neurosci,
10:1261–1269.
Devanand, DP, Bansal, R, Liu, J, Hao, X, Pradhaban, G, Peterson, BS (2012) MRI hippo-
campal and entorhinal cortex mapping in predicting conversion to Alzheimer’s disease.
Neuroimage, 60:1622–1629.
Diana, RA, Yonelinas, AP, Ranganath, C (2007) Imaging recollection and familiarity in the
medial temporal lobe: a three-component model. Trends Cogn Sci, 11:379–386.
292
Dickerson, BC, Salat, CE, Bates, JF, Atiya, M, Killiany, RJ, Greve, DN, Dale, AM, Stern, CE,
Blacker, D, Albert, MS, Sperling, RA (2004) Medial temporal function and structure in
mild cognitive impairment. Annals of Neurology, 56:27–35.
Dickerson, BC, Salat, CE, Greve, DN, Chua, EF, Rand-Giovannetti, E, Rentz, D, Bertram, L,
Mullin, K, Tanzi, RE, Blacker, D, Albert, MS, Sperling, R (2005) Increased hippocampal
activation in mild cognitive impairment compared to normal aging and AD. Neurology,
65:404–411.
Dieguez, DJ, Barea-Rodriguez, EJ (2004) Aging impairs the late phase of long-term potentia-
tion at the medial perforant path-CA3 synapse in awake rats. Synapse, 52:53–61.
Dodart, JC, May, P (2005) Overview on rodent models of Alzheimer’s disease. Current
Protocols in Neuroscience, 9.22.1–9.22.16.
Dudukovic, NM, Preston, AR, Archie, JJ, Glover, GH, Wagner, AD (2011) High-resolution
fMRI reveals match enhancement and attentional modulation in the human medial tempo-
ral lobe. Journal of Cognitive Neuroscience, 23:670–682.
Dumitriu, D, Hao, J, Hara, Y, Kaufmann, J, Janssen, WG, Lou, W, Rapp, PR, Morrison, JH
(2010) Selective changes in thin spine density and morphology in monkey prefrontal cor-
tex correlate with aging-related cognitive impairment. J Neurosci, 30:7507–7515.
Eichenbaum, H, Otto, T, Cohen, NJ (1994) Two functional components of the hippocampal
memory system. Behavioral and Brain Sciences, 17:449–472.
Eichenbaum, H, Fortin, N, Sauvage, M, Robitsek, RJ, Farovik, A (2010) An animal model of
amnesia that uses Receiver Operator Characteristics (ROC) analysis to distinguish recollec-
tion from familiarity deficits in recognition memory. Neuropsychologia, 48(8):2281–2289.
Ekstrom, AD, Meltzer, J, McNaughton, BL, Barnes, CA (2001) NMDA receptor antagonism
blocks experience-dependent expansion of hippocampal “place fileds.” Neuron, 31:631–638.
Ellwanger, J, Geyer, MA, Braff, DL (2003) The relationship of age to prepulse inhibition and
habituation of the acoustic startle response. Biological Psychology, 62:175–195.
Fleischman, DA, Wilson, RS, Gabrieli, JD, Bienias, JL, Bennett, DA (2004) A longitudinal
study of implicit and explicit memory in old persons. Psychol Aging, 19:617–625.
Flood, DG (1991) Region-specific stability of dendritic extent in normal human aging and
regression in Alzheimer’s disease. II. Subiculum. Brain Research, 540:83–95.
Flood, DG, Buell, SJ, Horwitz, GJ, Coleman, PD (1987) Dendritic extent in human dentate
gyrus granule cells in normal aging and senile dementia. Brain Research, 402:205–216.
Friedman, D (2000) Event- related brain potential investigations of memory and aging.
Biological Psychology, 54:175–206.
Frisoni, GB, Ganzola, R, Canu, E, Rub, U, Pizzini, FB, Alessandrini, F, Zoccatelli, G,
Beltramello, A, Caltagirone, C, Thompson, P (2008) Mapping local hippocampal changes
in Alzheimer’s disease and normal ageing with MRI at 3 Tesla. Brain, 131:3266–3276.
Gallagher, M, Colantuoni, C, Eichenbaum, H, Haberman, RP, Rapp, PR, Tanila, H, Wilson, IA
(2006) Individual differences in neurocognitive aging of the medial temporal lobe. Age
(Dordr), 28:221–233.
Gallagher, M, Rapp, PR (1997) The use of animal models to study the effects of aging on cog-
nition. Annu Rev Psychol, 48:339–370.
Geinisman, Y, Bondareff, W, Dodge, JT (1977) Partial deafferentation of neurons in the dentate
gyrus of the senescent rat. Brain Research, 134:541–545.
Geinisman, Y, deToledo-Morrell, L, Morrell, F, Persina, IS, Rossi, M (1992) Age-related loss of
axospinous synapses formed by two afferent systems in the rat dentate gyrus as revealed
by the unbiased stereological dissector technique. Hippocampus, 2:437–444.
Giovanello, KS, Schacter, DL (2012) Reduced specificity of hippocampal and posterior ven-
trolateral prefrontal acticity during relational retrieval in normal aging. J Cogn Neurosci,
24:159–170.
293
Grill, JD, Riddle, DR (2002) Age-related and laminar-specific dendritic changes in the medial
frontal cortex of the rat. Brain Research, 937:8–21.
Hanks, SD, Flood, DG (1991) Region-specificity stability of dendritic extent in normal human
aging and regression in Alzheimer’s diesase. I. CA1 of hippocampus. Brain Research,
540:63–82.
Hara, Y, Parks, CS, Janssen, WG, Punsoni, M, Rapp, PR, Morrison, JH (2011) Synaptic charac-
teristics of dentate gyrus axonal boutons and their relationship with aging, menopause, and
memory in female rhesus monkeys. Journal of Neuroscience, 31:7737–7744.
Hargreaves, E, Mattfeld, AT, Stark, CE, Suzuki, WA (2012) Conserved fMRI and LFP signals
during new associative learning in the human and macaque monkey medial temporal lobe.
Neuron, 74:743–752.
Hasselmo, ME (1995) Neuromodulation and cortical function: modeling the physiological
basis of behavior. Behav Brain Res, 67:1–27.
Hasselmo, ME, Schnell, E, Barkai, E (1995) Dynamics of learning and recall at excitatory
recurrent synapses and cholinergic modulation in rat hippocampal region CA3. J Neurosci,
15:5249–5262.
Hemby, SE, Tojanowski, JQ, Ginsberg, SD (2003) Neuron-specific age-related decreasesin
dopamine receptor subtype mRNAs. Journal of Comparative Neurology, 456:176–183.
Ikonen, S, McMahan, RW, M, G, Tanila, H (2002) Cholinergic system regulation of spatial
representation by the hippocampus. Hippocampus, 12:386–397.
Jack, CR, Petersen, RC, Xu, Y, O’Brien, PC, Smith, GE, Ivnik, RJ, Tangalos, EG, Kokmen, R
(1998) The rate of medial temporal lobe atrophy in typical aging and Alzheimer’s disease.
Neurology, 51:993–999.
Johnson, MK, Hashtroudi, S, Lindsay, DS (1993) Source monitoring. Psychological Bulletin,
114:3–28.
Kalus, P, Slotboom, J, Gallinat, J, Mahlberg, R, Cattapan-Ludewig, K, Wiest, R, Nyffeler, T,
Buri, C, Federspiel, A, Kunz, D, Schroth, G, Kiefer, C (2005) Examining the gateway
to the limbic system with diffusion tensor imaging: the perforant pathway in dementia.
Neuroimage, 30:713–720.
Kempermann, G, Kuhn, HG, Gage, FH (1997) More hippocampal neurons in adult mice living
in an enriched environment. Nature, 386:493–495.
Kempermann, G, Kuhn, HG, Gage, FH (1998) Experience-induced neurogenesis in the senes-
cent dentate gyrus. J Neurosci, 18:3206–3212.
Kentros, C, Hargreaves, E, Hawkins, RD, Kandel, ER, Shapiro, M, Muller, RV (1998) Abolition
of long-term stability of new hippocampal place cell maps by NMDA receptor blockade.
Science, 280:2121–2126.
Keuker, JI, Luiten, PG, Fuchs, E (2003) Preservation of hippocampal neuron numbers in aged
rhesus monkeys. Neurobiol Aging, 24:157–165.
Kirwan, CB, Jones, CK, Miller, MI, Stark, CE (2007) High-resolution fMRI investigation of the
medial temporal lobe. Hum Brain Mapp, 28:959–966.
Kirwan, CB, Stark, CE (2007) Overcoming interference: an fMRI investigation of pattern sep-
aration in the medial temporal lobe. Learn Mem, 14:625–633.
Koutstaal, W, Schacter, DL, Brenner, C (2001) Dual task demands and gist-based false recogni-
tion of pictures in younger and older adults. Journal of Memory and Language, 44:399–426.
Kumazawa-Manita, N, Hama, H, Miyawaki, A, Iriki, A (2013) Tool use specific adult neu-
rogenesis and synaptogenesis in rodent (Octodon degus) hippocampus. PLOS One,
DOI:10.1371/journal.pone.0058649.
Lacy, JW, Yassa, MA, Stark, SM, Muftuler, LT, Stark, CE (2011) Distinct pattern separation
related transfer functions in human CA3/dentate and CA1 revealed using high-resolution
fMRI and variable mnemonic similarity. Learn Mem, 18:15–18.
294
Lavenex, P, Amaral, DG (2000) Hippocampal-neocortical interaction: a hierarchy of associativ-

ity. Hippocampus, 10:420–430.
Leutgeb, JK, Leutgeb, S, Moser, MB, Moser, EI (2007) Pattern separation in the dentate gyrus
and CA3 of the hippocampus. Science, 315:961–966.
Li, Y, Mu, Y, Gage, FH (2009) Development of neural circuits in the adult hippocampus. Current
Topics in Developmental Biology, 87:149–174.
Lisman, JE, Grace, AA (2005) The hippocampal-VTA loop controlling the entry of information
into long-term memory. Neuron, 46:703–713.
Liu, P, Gupta, N, Jing, Y, Zhang, H (2008) Age-related changes in polyamines in memory-
associated brain structures in rats. Neuroscience, 155:789–796.
Lockhart, SN, Mayda, AB, Roach, AE, Fletcher, E, Carmichael, O, Maillard, P, Schwarz, CG,
Yonelinas, AP, Ranganath, C, Decarli, C (2012) Episodic memory function is associated
with multiple measures of white matter integrity in cognitive aging. Frontiers in Human
Neuroscience, 6:56.
Logothetis, NK, Wandall, BA (2004) Interpreting the BOLD signal. Annual Review of
Physiology, 66:735–769.
Lustig, C, Hasher, L (2001) Interference. In: The Encyclopedia of Aging (Maddox, GL, ed), pp
553–555. Springer.
Lyle, KB, Bloise, SM, Johnson, MK (2006) Age-related binding deficits and the content of false
memories. Psychol Aging, 21:86–95.
Makris, N, Papadimitriou, GM, van der Kouwe, A, Kennedy, DN, Hodge, SM, Dale, AM,
Benner, T, Wald, LL, Wu, O, Tuch, DS, Caviness, VS, Moore, TL, Killiany, RJ, Moss,
MB, Rosene, DL (2007) Frontal connections and cognitive changes in normal aging rhe-
sus monkeys: a DTI study. Neurobiol Aging, 28:1556–1567.
Markham, JA, Juraska, JM (2002) Aging and sex influence the anatomy of the rat anterior cin-
gulate cortex. Neurobiol Aging, 23:579–588.
Markham, JA, McKian, KP, Stroup, TS, Juraska, JM (2005) Sexually dimorphic aging of den-
dritic morphology in CA1 of hippocampus. Hippocampus, 15:97–103.
Marr, D (1971) Simple memory: a theory for archicortex. Philos Trans R Soc Lond B Biol Sci,
262:23–81.
McClelland, JL, Goddard, NH (1996) Considerations arising from a complementary learning
systems perspective on hippocampus and neocortex. Hippocampus, 6:654–665.
McClelland, JL, McNaughton, BL, O’Reilly, RC (1995) Why there are complementary learn-
ing systems in the hippocampus and neocortex: insights from the successes and failures of
connectionist models of learning and memory. Psychol Rev, 102:419–457.
McNaughton, BL, Morris, RG (1987a) Hippocampal synaptic enhancement and information
storage within a distributed memory system. Trends Neurosci, 10:408–415.
McNaughton, BL, Morris, RG (1987b) Hippocampal synaptic enhancement and information
storage withing a distributed memory system. Trends in Neuroscience, 10:408–415.
Merrill, DA, Chiba, AA, Tuszynski, MH (2001) Conservation of neuronal number and
size in the entorhinal cortex of behaviorally characterized aged rats. J Comp Neurol,
438:445–456.
Merrill, DA, Roberts, JA, Tuszynski, MH (2000) Conservation of neuron number and
size in entorhinal cortex layers II, III, and V/VI of aged primates. J Comp Neurol,
422:396–401.
Miller, SL, Celone, K, DePeau, K, Diamond, E, Dickerson, BC, Rentz, D, Pihlajamaki, M,
Sperling, RA (2008a) Age- related memory- impairment associated with the loss of
parietal deactivation but preserved hippocampal activation. Proc Natl Acad Sci U S A,
105:2181–2186.
295
Miller, SL, Fenstermacher, E, Bates, J, Blacker, D, Sperling, RA, Dickerson, BC (2008b)

Hippocampal activation in adults with mild cognitive impairment predicts subsequent cog-
nitive decline. Journal of Neurology, Neurosurgery, and Psychiatry, 79:630–635.
Ming, GL, Song, H (2005) Adult neurogenesis in the mammalian central nervous system.
Annual Review of Neuroscience, 28:223–250.
Mishkin, M, Delacour, J (1975) An analysis of short-term visual memory in the monkey. Journal
of Experimental Psychology: Animal Behavior Processes, 1:326–334.
Monk, CS, Zhuang, J, Curtis, WJ, Ofenloch, IT, Tottenham, N, Nelson, CA, Hu, X (2002) Human
hippocampal activation in the delayed matching-and nonmatching-to-sample memory
tasks: an event-related functional MRI approach. Behavioral Neuroscience, 116:716–721.
Montaron, MF, Drapeau, E, Dupret, D, Kitchener, P, Aurousseau, C, Le Moal, M, Piazza, PV,
Abrous, DN (2006) Lifelong corticosterone level determines age-related decline in neuro-
genesis and memory. Neurobiol Aging, 27:645–654.
Morris, R (2007) Theories of hippocampal function. In: The Hippocampus Book (Amaral, DG,
Andersen, P, Bliss, T, Morris, RGM, O’Keefe, J, eds), pp 581–714. New York: Oxford
University Press.
Morris, RG (1981) Spatial localization does not require the presence of local cues. Learning
and Motivation, 12:239–260.
Morrison, JH, Baxter, MG (2012) The ageing cortical synapse: hallmarks and implications for
cognitive decline. Nature Reviews Neuroscience, 13:240–250.
Moyer, JRJ, Furtak, SC, McGann, JP, Brown, TH (2011) Aging-related changes in calcium-
binding proteins in rat perirhinal cortex. Neurobiol Aging, 32:1693–1706.
Mueller, SG, Weiner, MW (2009) Selective effect of age, Apo e4, and Alzheimer’s disease on
hippocampal subfields. Hippocampus, 19:558–564.
Muller, RU, Kubie, JL (1987) The effects of changes in the environment on the spatial firing of
hippocampal complex-spike cells. Journal of Neuroscience, 7:1951–1968.
Naranjo, N, Greene, E (1977) Use of reduced silver staining to show loss of connections in aged
rat brain. Brain Research Bulletin, 2:71–74.
Naveh-Benjamin, M, Guez, J, Shulman, S (2004) Older adults’ associative deficits in episodic
memory: assessing the role of decline in attentional resources. Psychonomic Bulletin &
Review, 11:1067–1073.
Nelson, PT, Head, E, Schmitt, FA, Davis, PR, Neltner, JH, Jicha, GA, Abner, EL, Smith, CD,
Van Eldik, LJ, Kryscio, RJ, Scheff, SW (2011) Alzheimer’s disease is not “brain aging”:
Neuropathological, genetic, and epidemiological human studies. Acta Neuropathology,
121:571–587.
Neuneubel, JP, Knierim, JJ (2014) CA3 retrieves coherent representations from degraded
input: direct evidence for CA3 pattern completion and dentate gyrus pattern separation.
Neuron, 81:416–427.
Newmark, RE, Schon, K, Ross, RS, Starn, CE (2013) Contributions of the hippocampal sub-
fields and entorhinal cortex to disambiguation during working memory. Hippocampus,
23:467–475.
Nicholson, DA, Yoshida, R, Berry, RW, Gallagher, M, Geinisman, Y (2004) Reduction in size
of perforated postsynaptic densities in hippocampal axospinous synapses and age-related
spatial learning impairments. J Neurosci, 24:7648–7653.
Nielsen, TR, Kornum, BR, Moustgaard, A, Gade, A, Lind, NM, Knudsen, GM (2009) A novel
spatial delayed non-match to sample (DNMS) task in the Gottingen minipig. Behav Brain
Res, 196:93–98.
Norman, KA (2010) How hippocampus and cortex contribute to recognition memory: revisit-
ing the complementary learning systems model. Hippocampus, 20:1217–1227.
296
Norman, KA, O’Reilly, RC (2003) Modeling hippocampal and neocortical contributions

to recognition memory: a complementary- learning-
systems approach. Psychol Rev,
110:611–646.
O’Keefe, J, Dostrovsky, J (1971) The hippocampus as a spatial map: Preliminary evidence from
unit activity in the freely-moving rat. Brain Research, 34:171–175.
O’Keefe, J, Nadel, L (1978) The hippocampus as a cognitive map. Clarendon Press: Oxford.
O’Reilly, RC, Rudy, JW (2001) Conjunctive representations in learning and memory: princi-
ples of cortical and hippocampal function. Psychology Reviews, 108:311–345.
Pakkenberg, B, Gundersen, HJ (1997) Neocortical neuron number in humans: effect of sex and
age. J Comp Neurol, 384:312–320.
Penfield, W, Milner, B (1958) Memory deficit produced by bilateral lesions in the hippocampal
zone. AMA Archives of Neurology & Psychiatry, 79:475–497.
Perry, EK, Johnson, M, Kerwin, JM, Piggott, MA, Court, JA, Shaw, PJ, Ince, PG, Brown,
A, Perry, RH (1992) Convergent cholinergic activities in aging and Alzheimer’s disease.
Peters, A, Morrison, JH, Rosene, DL, Hyman, BT (1998) Are neurons lost from primate cere-
bral cortex during normal aging? Cereb Cortex, 8:295–300.
Peters, A, Sethares, C, Moss, MB (2010) How the primate fornix is affected by age. Journal of
Comparative Neurology, 18:3962–3980.
Pirogovsky, E, Holden, HM, Jenkins, C, Peavy, GM, Salmon, DP, Galasko, DR, Gilbert, P
(2013) Temporal sequence learning in healthy aging and amnestic mild cognitive impair-
ment. Experimental Aging Research, 39:doi:10.1080/0361073X.2013.808122.
Rapp, PR, Amaral, DG (1989) Evidence for task-dependent memory dysfunction in the aged
monkey. Journal of Neuroscience, 9:3568–3576.
Rapp, PR, Amaral, DG (1991) Recognition memory deficits in a subpopulation of aged mon-
keys resemble the effects of medial temporal lobe damage. Neurobiol Aging, 12:481–486.
Rapp, PR, Bachevalier, J (2011) Cognitive development and aging. In: Fundamental
Neuroscience (Squire, LR, Berg, D, Bloom, FE, du Lac, S, Ghosh, A, Spitzer, NC, eds),
pp 1039–1066. San Diego, CA: Academic Press.
Rapp, PR, Deroche, PS, Mao, Y, Burwell, RD (2002) Neuron number in the parahippocampal
region is preserved in aged rats with spatial learning deficits. Cereb Cortex, 12:1171–1179.
Rapp, PR, M, G (1996) Preserved neuron numbers in the hippocampus of aged rats with spatial
learning deficits. Proc Natl Acad Sci U S A, 93:9926–9930.
Raz, N, Lindenberger, U, Rodrigue, KM, Kennedy, KM, Head, D, Williamson, A, Dahle, C,
Gerstorf, D, Acker, JD (2005) Regional brain changes in aging healthy adults: general
trends, individual differences and modifiers. Cereb Cortex, 15:1676–1689.
Reagh, ZM, Roberts, JM, Ly, M, DiProspero, N, Murray, E, Yassa, MA (2013) Spatial discrim-
ination deficits as a function of mnemonic interference in aged adults with and without
memory impairment. Hippocampus, 24:303–314.
Reuter-Lorenz, PA, Cappell, KA (2008) Neurocognitive aging and the compensation hypthesis.
Current Directions in Psychological Science, 17: 177–182.
Ribak, CE, Seress, L, Amaral, DG (1985) The development, ultrastructure and synaptic connec-
tions of the mossy cells of the dentate gyrus. J Neurocytol, 14:835–857.
Rolls, ET (2013) The mechanisms for pattern completion and pattern separation in the hippo-
campus. Frontiers in Systems Neuroscience, 7:doi: 10.3389/fnsys.2013.00074.
Salvatore, ME, Apparsundaram, S, Gerhardt, GA (2003) Decreased plasma membrane expres-
sion of striatal dopamine transporter in aging. Neurobiol Aging, 24:1147–1154.
Sato, T, Maruyama, N, Hoshida, T, Minato, K (2012) Correlation between uncinate fasciculus
and memory tasks in healthy individuals using diffusion tenory tractography. Conference
Proceedings IEEE End Med Biol Soc, 424–427.
297
Sava, S, Markus, EJ (2008) Activation of medial septum reverses age-related hippocampal

encoding deficits. Journal of Neuroscience, 28:1841–1853.
Schacter, DL, Koutstaal, W, Johnson, MK, Gross, MS, Angell, KE (1997) False recollec-
tion induced by photographs: comparison of older and younger adults. Psychol Aging,
12:203–215.
Scheff, SW, Price, DA, Schmitt, FA, Mufson, EJ (2006) Hippocampal synaptic loss in early
Alzheimer’s disease and mild cognitive impairment. Neurobiol Aging, 10:137–184.
Schoenfeld, R, Foreman, N, Leplow, B (2014) Ageing and spatial reversal learning in
humans: findings from a virtual water maze task. Behav Brain Res, 270:47–55.
Scoville, WB, Milner, B (1957) Loss of recent memory after bilateral hippocampal lesions. J
Neurol Neurosurg Psychiatry, 20:11–21.
Shamy, JL, Habeck, C, Hof, PR, Amaral, DG, Fong, SG, Buonocore, MH, Stern, Y, Barnes,
CA, Rapp, PR (2011) Volumetric correlates of spatiotemporal working and recognition
memory impairment in aged rhesus monkeys. Cereb Cortex, 21:1559–1573.
Shen, J, Barnes, CA, McNaughton, BL, Skaggs, WE, Weaver, KL (1997) The effect of
aging on experience- dependent plasicity of hippocampal place cells. J Neurosci,
17:6769–6782.
Siddiqi, Z, Kemper, TL, Killiany, RJ (1999) Age-related neuronal loss from the substana-
tial nigra-pars compacta and ventral tegmental are of the rhesus monkey. Journal of
Neuropathology and Experimental Neurology, 58:959–971.
Small, SA, Chawla, MK, Buonocore, M, Rapp, PR, Barnes, CA (2004) Imaging correlates of
brain function in monkeys and rats isolates a hippocampal subregion differentially vulner-
able to aging. Proc Natl Acad Sci U S A, 101:7181–7186.
Small, SA, Tsai, WY, DeLaPaz, R, Mayeux, R, Stern, Y (2002) Imaging hippocampal
function across the human life span: is memory decline normal or not? Ann Neurol,
51:290–295.
Smith, DE, Rapp, PR, McKay, HM, Roberts, JA, Tuszynski, MH (2004) Memory impairment in
aged primates is associated with focal death of cortical neurons and atrophy of subcortical
neurons. J Neurosci, 24:4373–4381.
Smith, TD, Adams, MM, Gallagher, M, Morrison, JH, Rapp, PR (2000) Circuit-specific alter-
ations in hippocampal synaptophysin immunoreactivity predict spatial learning impair-
ment in aged rats. J Neurosci, 20:6587–6593.
Spalding, KL, Bergmann, O, Alkass, A, Bernard, S, Salehpour, M, Huttner, HB, Bostrom, E,
Westerlund, I, Vial, C, Buchholz, BA, Possnert, G, Mash, DC, Druid, H, Frisen, J (2013)
Dynamics of hippocampal neurogenesis in adult humans. Cell, 153:1219–1227.
Sperling, R (2007) Functional MRI studies of associative encoding in normal aging, mild cogni-
tive impairment, and Alzheimer’s disease. Annals of the New York Academy of Sciences,
1097: 146–155.
Squire, LR (1992) Memory and the hippocampus: a synthesis of findings with rats, monkeys,
and humans. Psychological Review, 99:195–231.
Squire, LR, Stark, CE, Clark, RE (2004) The medial temporal lobe. Annu Rev Neurosci,
27:279–306.
Stark, SM, Yassa, MA, Lacy, JW, Stark, CE (2013) A task to assess behavioral pattern sep-
aration (BPS) in humans: Data from healthy aging and mild cognitive impairment.
Neuropsychologia, 51:2442–2449.
Stark, SM, Yassa, MA, Stark, CE (2010) Individual differences in spatial pattern separation
performance associated with healthy aging in humans. Learn Mem, 17:284–288.
Stark, SM, Stevenson, R, Wu, C, Rutledge, S, Stark, CE (2015) Stability of age-related defi-
cits in the mnemonic similarity task across task variations. Behavioral Neuroscience,
129(3):257–268.
298
Stemmelin, J, Lazarus, C, Cassel, S, Kelche, C, Cassel, JC (2000) Immunohistochemical and

neurochemical correlates of learning deficits in aged rats. Neuroscience, 24:275–289.
Stranahan, AM, Haberman, RP, Gallagher, M (2010) Cognitive Decline Is Associated with
Reduced Reelin Expression in the Entorhinal Cortex of Aged Rats. Cereb Cortex,
21(2):392–400.
Stranahan, AM, Jiam, NT, Spiegel, AM, M, G (2012) Aging reduces total neuron number in the
dorsal component of the rodent prefrontal cortex. J Comp Neurol, 520:1318–1326.
Sugaya, K, R, G, Personett, D, Robbins, M, Kent, C, Bryan, D, Skiba, E, Gallagher, M,
McKinney, M (1998) Septo-hippocampal cholinergic and neurotrophin markers in age-
induced cognitive decline. Neurobiol Aging, 19:351–361.
Tanila, H, Sipila, M, Shapiro, H, Eichenbaum, H (1997) Brain aging: impaired coding of novel
environmental cues. Journal of Neuroscience, 17:5167–5174.
Tolentino, JC, Pirogovsky, E, Luu, T, Toner, CK, Gilbert, P (2012) The effect of interference
on temporal order memory for random and fixed sequences in nondemented older adults.
Learn Mem, 19:doi:10.1101/lm.026062.112.
Toner, CK, Pirogovsky, E, Kirwan, CB, Gilbert, PE (2009) Visual object pattern separation
deficits in nondemented older adults. Learn Mem, 16:338–342.
Treves, A, Rolls, ET (1994) Computational analysis of the role of the hippocampus in memory.
Hippocampus, 4:374–391.
Uemura, E (1985) Age-related changes in the subiculum of teh Macaca mulatta: dendritic
branching pattern. Experimental Neurology, 87:412–427.
Uylings, HB, Groenewegen, HJ, Kolb, B (2003) Do rats have a prefrontal cortex? Behav Brain
Res, 146:3–17.
Van Groen, T, Kadish, I, Popovic, N, Caballero-Bieda, M, Bano-Otalora, B, Vivanco, P, Rol,
MA, Madrid, JA (2011) Age-related brain pathology in Octogon degu: blood vessel, white
matter, and Alzheimer-like pathology. Neurobiol Aging, 32:1651–1661.
van Praag, H, Shubert, T, Zhao, C, Gage, FH (2005) Exercise enhances learning and hippocam-
pal neurogenesis in aged mice. J Neurosci, 25:8680–8685.
Van Strien, NM, Cappaert, NL, Witter, MP (2009) The anatomy of memory: an interactive
overview of the parahippocampal-hippocampal network. Nat Rev Neurosci, 10:272–282.
Vela, J, Gutierrez, A, Vitorica, J, Ruano, S (2003) Rat hippocampal GABAergic molecular
markers are differentially affected by ageing. Journal of Neurochemistry, 85:368–377.
Verhaeghen, P (2003) Aging and vocabulary scores: a meta-analysis. Psychol Aging, 18:332–339.
Wenzel, J, Kammerer, E, Kirsche, W, Matthies, H, Wenzel, M (1980) Electron microscopic and
morphometric studies on synaptic plasticity in the hiipocampus of the rat following condi-
tioning. Journal for Hirnforschung, 21:647–654.
West, MJ (1993) Regionally specific loss of neurons in the aging human hippocampus.
Wilson, FA, Scalaidhe, SP, Goldman-Rakic, PS (1993) Dissociation of object and spatial pro-
cessing domains in primate prefrontal cortex. Science, 260:1955–1958.
Wilson, IA, Gallagher, M, Eichenbaum, H, Tanila, H (2006) Neurocognitive aging: prior mem-
ories hinder new hippocampal encoding. Trends Neurosci, 29:662–670.
Wilson, IA, Ikonen, S, Gurevicius, K, McMahan, RW, Gallagher, M, Eichenbaum, H, Tanila, H
(2005) Place cells of aged rats in two visually identical compartments. Neurobiol Aging,
26:1099–1106.
Wilson, IA, Ikonen, S, McMahan, RW, Gallagher, M, Eichenbaum, H, Tanila, H (2003) Place
cell rigidity correlates with impaired spatial learning in aged rats. Neurobiol Aging,
24:297–305.
Wilson, MA, McNaughton, BL (1993) Dynamics of the hippocampal ensemble code for space.
Science, 261:1055–1058.
299
Witter, MP (2007) The perforant path: projections from the entorhinal cortex to the dentate
gyrus. Prog Brain Res, 163:43–61.
Wu, WE, Brickman, A, Luchsinger, J, Farrazano, P, Pichiule, P, Brown, TH, DeCarli, C, Barnes,
CA, Mayeux, R, Vannucci, S, Small, SA (2008) The brain in the age of old: The hippo-
campal formation is targeted differentially by diseases of late-life. Annals of Neurology,
64:698–706.
Yassa, MA, Lacy, JW, Stark, SM, Albert, MS, Gallagher, M, Stark, CE (2010a) Pattern sepa-
ration deficits associated with increased hippocampal CA3 and dentate gyrus activity in
nondemented older adults. Hippocampus, Epub ahead of print doi: 10.1002/hipo.20808
Yassa, MA, Mattfeld, AT, Stark, SM, Stark, CE (2011) Age- related memory deficits
linked to circuit-specific disruptions in the hippocampus. Proc Natl Acad Sci U S A,
108:8873–8878.
Yassa, MA, Muftuler, LT, Stark, CE (2010b) Ultrahigh-resolution microstructural diffusion ten-
sor imaging reveals perforant path degradation in aged humans in vivo. Proc Natl Acad Sci
U S A, 107:12687–12691.
Yassa, MA, Stark, CE (2009) A quantitative evaluation of cross-participant registration tech-
niques for MRI studies of the medial temporal lobe. Neuroimage, 44:319–327.
Yassa, MA, Stark, CE (2011) Pattern separation in the hippocampus. Trends Neurosci,
34:515–525.
Zacks, RT, Hasher, L, Li, KZH (2000) Human Memory. In: Handbook of Aging and Cognition
(Salthouse, TA, Craik, FIM, eds), pp 293–357. Mahwah, NJ: Lawrence Erlbaum.
300
â•‡ 301
12
Episodic Memory Encoding

and Retrieval inÂ€the AgingÂ€Brain
Wei-â•‰Chun Wang
RobertoÂ€Cabeza
W ith healthy aging comes a host of cognitive changes. Compared to

young adults (YAs), older adults (OAs) are impaired in many cogni-
tive abilities, including episodic memory, executive function, attention, and process-
ing speed, whereas abilities such as language and semantic knowledge are relatively
preserved (Light, 1991; Salthouse, 2004). Functional neuroimaging methods allow
researchers in the new domain of cognitive neuroscience of aging to associate age-â•‰
related differences in cognition and brain function. In particular, techniques such as
event-â•‰related fMRI can identify age effects on brain activity associated with success-
ful vs. failed cognitive performance.
The current chapter reviews event-â•‰related fMRI studies of episodic memory encod-
ing and retrieval, as well as fMRI studies of episodic-â•‰memory-â•‰related functional con-
nectivity. Before turning to these studies, we first provide a brief summary of relevant
background areas: (1) episodic memory and its neural mechanisms; (2) age-â•‰related
episodic memory decline and its neural mechanisms; (3) task-â•‰independent age effects
on brain activity and the concept of compensation.
Episodic Memory and its Neural Mechanisms
Episodic memory refers to conscious memory for personally experienced past events
(Tulving, 1984). This includes memory for what events happened (item memory), as
well as memory for where, when, and how they happened (context memory). The term
301
302
associative memory is broader than context memory, as it includes both item–context

and item–item associations, and for this reason, we prefer it in this review. We use
“associative memory” and “item memory” to refer to forms of memory while the terms
“associative tasks” and “item tasks” refer to memory tests that explicitly test for asso-
ciative or item memory. Associative tasks include context memory (spatial, temporal,
etc.) and associative recognition (e.g., word pairs) tests, and item tasks, old/new item
recognition tests. The distinction between forms of memory and tasks is important
because tasks are not process pure. For example, even if participants are only asked
to decide if an item is old or new, if they can retrieve associations about the item, they
can use it infer or confirm the item decision (e.g., remembering where it happened
confirms that it happened). It is also important to note that the associative and item
distinction largely overlaps with the recollection and familiarity distinction (Yonelinas,
2002). Recollection can be generally defined as associative memory plus item mem-
ory, whereas familiarity as item memory without associative memory. The Remember/
Know recognition paradigm allows us to isolate the associative component in recogni-
tion tests by contrasting “Remember” (i.e., associative plus item) to “Know” (i.e., item
only) responses. Similarly, because high-confidence “old” recognition responses are
thought to reflect recollection (Yonelinas, 2001; Koen and Yonelinas, 2010), they can
sometimes be used as an associative measure when compared to low-confidence “old”
recognition responses (item only) or all other responses.
Episodic memory is primarily dependent on the medial temporal lobe (MTL),
prefrontal cortex (PFC), and posterior cortical regions, as well as on their inter-
actions (Simons and Spiers, 2003; Eichenbaum et al., 2007; Cabeza et al., 2008;
Uncapher and Wagner, 2009; Rugg and Vilberg, 2013). During encoding, MTL—
and particularly the hippocampus (HPC)—is assumed to store mnemonic rep-
resentations as well as pointers to posterior cortical memory traces (e.g., visual
cortex for visual memory representations), and during retrieval, MTL mediates
the reactivation of posterior cortices and access to stored memory traces (Alvarez
and Squire, 1994; McClelland et al., 1995; Nadel et al., 2000; Sutherland and
McNaughton, 2000; Norman and O’Reilly, 2003; Danker and Anderson, 2010;
Ritchey et al., 2013). According to a three-process MTL model (Davachi, 2006;
Eichenbaum et al., 2007), perirhinal cortex primarily mediates memory for items,
parahippocampal cortex, memory for contexts, and HPC, memory for associa-
tions between items and contexts (i.e., associative binding). During encoding,
ventrolateral PFC is assumed to mediate controlled semantic processing (Prince et
al., 2007) while dorsolateral PFC is thought to regulate organizational processes
(Blumenfeld and Ranganath, 2007). During retrieval, dorsolateral PFC is pre-
sumed to mediate memory search and monitoring processes (Achim and Lepage,
2005; Hayama and Rugg, 2009). In posterior cortices, modality-specific regions
(visual, auditory, etc.) are assumed to hold perceptual memory representations,
and lateral/ventral temporal cortices, semantic memory representations. Dorsal
parietal cortex (DPC; including medial precuneus) is hypothesized to mediate
top-down attention during encoding and retrieval (Cabeza, 2008; Uncapher and
Wagner, 2009), whereas ventral parietal cortex (VPC) is presumed to mediate
bottom-up attention to (Cabeza et al., 2008) or maintenance of (Vilberg and Rugg,
2008) recovered memories during retrieval.
â•‡ 303
Episodic Memory Encoding and Retrieval in the AgingÂ€Brainâ•… 303
Age-â•‰Related Episodic Memory Decline and its Neural Mechanisms

Episodic memory is arguably the cognitive function most affected by healthy aging
and early Alzheimer’s disease. However, not all forms of episodic memory are equally
affected: whereas associative deficits in healthy OAs are substantial, item deficits are
small and usually nonsignificant (for reviews, see Kausler, 1994; Spencer and Raz,
1995; Yonelinas, 2002; Koen and Yonelinas, 2016). The dissociation between associa-
tive and item memory is one of the main topics in this review. According to a resource
deficit hypothesis (Craik, 1986), age-â•‰related associative deficits reflect a reduction
in executive control and attentional resources, which are more critical for associa-
tive than item memory. Consistent with this hypothesis, when control resources are
reduced in YAs using a divided attention manipulation, YAs’ memory performance
resembles that of OAs (Craik and Byrd, 1982; Rabinowitz et al., 1982). According to
the associative deficit hypothesis (Naveh-â•‰Benjamin, 2000), associative impairments
in OAs reflect a deficit in memory binding, which is essential for associative but not
item memory. There is a growing agreement that both executive control and binding
deficits are required to explain age-â•‰related associative impairments (Castel and Craik,
2003; Naveh-â•‰Benjamin et al., 2004). In keeping with this idea, factor analyses of
OAs’ associative deficits have identified two quasi-â•‰orthogonal factors: an “executive”
factor that correlates with tests primarily dependent on PFC function, and a “memory”
factor that correlates with tests primarily dependent on MTL function (Glisky et al.,
1995; Glisky et al., 2001; Prull et al., 2006).
This two-â•‰factor model fits well with evidence that PFC and MTL are among the
brain regions most affected by healthy aging. PFC displays the largest age-â•‰related
atrophy in the brain, (Raz et al., 1997; Raz, 2000; Raz et al., 2005) (Raz et al., 1997;
Raz, 2000; Raz et al., 2005) and its function is also impaired by age-â•‰related deficits
in dopamine function (Bäckman and Farde, 2004). Within MTL, age-â•‰related atrophy
is substantial in HPC but minimal in perirhinal cortex (Raz et al., 2005). Given that
HPC has been linked to associative memory and perirhinal cortex to item memory
(Davachi, 2006; Eichenbaum et al., 2007), this atrophy pattern fits well with the afore-
mentioned associative-â•‰item dissociation. However, brain regions cannot operate inde-
pendently, and hence episodic memory also depends on the integrity of white-â•‰matter
tracts linking regions such as PFC and MTL to each other and to the rest of the brain.
Thus, it is not surprising that age-â•‰related episodic memory deficits are significantly
correlated with age-â•‰related white â•‰matter decline (Davis et al., 2009; Kennedy and Raz,
2009; Charlton et al., 2010; Daselaar et al., 2015).
Task-â•‰Independent Age Effects on Brain Activity

and the Concept of Compensation
Whereas reduced activity in OAs compared to YAs is usually attributed to some cog-
nitive deficit, increased activity in OAs is typically attributed to “compensation” (for
reviews, see Dennis and Cabeza, 2008; Cabeza and Dennis, 2013). Some age-â•‰related
decreases and increases are specific to an individual cognitive domain, such as asso-
ciative memory, but others have been found to be reliable across several cognitive
domains. One of these task-â•‰independent effects is the posterior–â•‰anterior shift in aging
304
or PASA (Davis et al., 2008), which refers to an age-related decrease in occipito-

temporal activity coupled with an age-related increase in PFC activity. PASA was
first found in a visual matching task by Grady et al. (1994), who suggested that PFC
increases compensate for occipito-temporal decreases. Consistent with this hypothe-
sis, there is evidence that PFC activations are negatively correlated with occipito-tem-
poral activations (Cabeza et al., 2004; Davis et al., 2008), and are positively correlated
with task performance (Davis et al., 2008).
Another task-independent effect is the Hemispheric Asymmetry Reduction in
OLDer Adults or HAROLD, which refers to more bilateral PFC activations in OAs
than YAs (Cabeza, 2002). HAROLD was first reported by Cabeza et al. (1997),
who suggested that OAs compensate for age-related decline by recruiting both
hemispheres instead of one. We, and others, have suggested that the compensation
account of HAROLD is supported by evidence that bilateral recruitment is posi-
tively correlated with cognitive performance across participants (Reuter-Lorenz et
al., 2000; Cabeza, 2002; Rosen et al., 2002). However, we no longer believe that
correlations across participants provide good evidence in favor of compensation.
Similarly, we believe that negative across-subject correlations—such as between
greater PFC activity and performance, which have been couched as evidence for
PFC dysfunction (e.g., Rypma and D’Esposito, 2000; Rypma et al., 2007; Stern et
al., 2012; McDonough et al., 2013)—should also not be attributed to compensation.
Given accumulated evidence that both high-and low-performing OAs can display
compensatory activity, we now believe that positive correlations with performance
support compensation only when they are established within-subjects, as is now pos-
sible using event-related fMRI paradigms.
It is also, however, important to note alternative viewpoints of age-related increases
in activity. One such viewpoint is dedifferentiation, which posits that older adults have
difficulty in recruiting specialized neural mechanisms (Li et al., 2001; Park et al.,
2004). This explanation is most commonly supported by evidence from studies of
different visual stimuli (Park et al., 2004; Goh et al., 2010; Park et al., 2012) but is also
reported in studies of different memory processes (Dennis and Cabeza, 2011; Wang
et al., 2015). Alternatively, it has also been proposed that age-related increases may
reflect neural inefficiency (Rypma and D’Esposito, 2000; Reuter-Lorenz and Lustig,
2005; Morcom et al., 2007; Nyberg et al., 2014).
Current data do not allow for a clear consensus as to which of these explanations
account for age-related increases, although we also note that the concept of compen-
sation is not necessarily incompatible with other explanations of age-related increases
in activity. Specifically, dedifferentiation and neural inefficiency can both occur con-
currently with compensation. For example, compensatory activity may occur in cohort
with dedifferentiation (e.g., Burianová et al., 2013). Additionally, compensatory activ-
ity need not be efficient in the same way that using a cane to walk is compensatory but
not necessarily efficient.
In the case of episodic memory, event-related fMRI paradigms—as opposed to
blocked designs—allow researchers to identify activity associated with successful
memory processes both during encoding and during retrieval. During encoding,
one can identify Subsequent Memory Effects (SMEs) by comparing activity for
subsequently remembered vs. forgotten items (Paller and Wagner, 2002). During
â•‡ 305
retrieval, one can identify Retrieval Success Effects (RSEs) by comparing hits to
correct rejections or misses. SMEs and RSEs are assumed to reflect successful
encoding and retrieval processes respectively. We interpret larger SMEs/â•‰RSEs for
OAs than YAs as evidence for compensation—â•‰regardless of whether these effects
are displayed by high-â•‰or low-â•‰performing OAs. We do note, however, that these
effects may also be interpreted as neural inefficiency (regardless of whether it is
compensatory in nature). Thus, the current review focuses on age effects on epi-
sodic SMEs and RSEs.
Event-â•‰Related fMRI Studies of Episodic Memory and Aging
The fact that OAs are more impaired in associative than item memory (Spencer and
Raz, 1995) may reflect difficulties with encoding, retrieval, or both. We review the
effects of aging on SMEs and RSEs in separate sections, and then turn to the effects of
aging on “negative” SMEs (nSMEs) and RSEs (nRSEs), which occur when success-
ful memory is associated with reduced rather than increased activity. Throughout the
review, we interpret age-â•‰related increases and decreases relative to YAs. That is, we
consider increases to be compensatory given that they indicate an effect is larger in
OAs than YAs; if a region typically shows SMEs in YAs, then increased SMEs in OAs
could be interpreted as compensation, and if a region typically shows nSMEs in YAs,
then increased nSMEs in OAs could also be interpreted as compensation. Finally, we
review evidence regarding age effects on connectivity during episodic memory tasks.
Studies were selected via a pubmed and google scholar search with the following
search query: “(fmri OR ‘functional mri’ OR ‘functional magnetic resonance imag-
ing’) AND ‘event related’ AND (aging OR age OR older) AND (memory OR ‘recog-
nition memory’).” Studies were limited to visual stimuli and excluded from the tables
reported below if whole brain results or if “memory success” contrasts (e.g., hit versus
miss) were not reported.
Subsequent Memory Effects (SMEs)

Table 12.1 displays the results of studies investigating age effects on SMEs. Most
studies compared high-â•‰confidence hits vs. misses but some exceptions are noted
below. To simplify the description of the findings, brain regions are classified very
coarsely. Parietal cortex is divided into VPC (Brodmann Areas –â•‰BAs 39 and 40) and
DPC (BA7 and precuneus). The occipito-â•‰temporal pathway (OTP) includes medial
and lateral occipital cortex, ventral and lateral temporal cortex, as well as two MTL
regions, the parahippocampal gyrus (PHG)—â•‰including perirhinal and parahippocam-
pal cortex and HPC. The table also includes a posterior midline region (PMR: retro-
splenial and posterior cingulate cortices), the insula, and the anterior cingulate cortex
(ACC). Finally, PFC is divided into four large areas: posterior (i.e., premotor, BAs 6
and 8), ventrolateral (BAs 44, 45, and 47), dorsolateral (BAs 9 and 46), and anterior
(BAs 10 and 11, including anteromedial) regions. For each of these regions, there are
separate columns for the left and right hemisphere.
306
Table 12.1 Subsequent Memory Effects (SMEs)
â•‡ 307
As for the symbols in Table 12.1, small black diamonds indicate a significant SME
in both YAs and OAs with no difference between these groups. Squares indicate that
there was a significant SME in only one group (black squares for YAs, white squares
for OAs) but no SME x age interaction. Lastly, circles indicate that the interaction was
significant either because the SME was smaller in OAs than YAs (black circles) or
because it was larger in OAs than YAs (white circles). If the interaction was driven by a
large negative SME in one group and a nonsignificant SME in the other, the finding is
included in Table 12.3 and discussed in the section on negative SME/â•‰RSE effects. For
the sake of brevity, black squares/â•‰circles are referred to as “age-â•‰related decreases” or just
“decreases,” and white squares/â•‰circles, as “age-â•‰related increases” or simply “increases.”
Depending on the encoding and retrieval tasks employed, SMEs may predict item
performance or associative performance. Item SMEs and associative SMEs are dis-
played at the top and bottom panels of Table 12.1, and at the bottom of each panel, a
greyed row identifies “recurrent” findings. “Recurrent” is defined as an effect (e.g.,
increase) reported (1) by at least 25% of the studies in the category, and (2) twice as
frequently as the opposite effect (e.g., decrease). Given that there were no notable
between-â•‰age effects in the left or the right hemisphere (e.g., age-â•‰related increases
or decreases were not more apparent in one hemisphere than the other), recurrent
findings were collapsed across hemispheres. Additionally, we also note that the tables
do not display hemispheric asymmetry differences (i.e., HAROLD) and studies that
report it are mentioned in the text. For Table 12.1 and all subsequent tables, we cat-
egorized imaging contrasts as item memory or associative memory primarily based
on the task, unless it specifically tested recollection (which approximates associative
memory) or familiarity (which approximates item memory). Given the caveat that
memory tasks are not process pure, we consider whether associative contamination
in item tasks—â•‰or vice versa—â•‰may account for some of the observed patterns across
studies in the General Discussion.
Item Memory SMEs (Item SMEs)

As illustrated by the bottom row of the item SME panel, these studies yielded four
recurrent findings; compared to YAs, SMEs in OAs often showed (1) decreases in OTP
(occipital cortex, PHG, and/â•‰or HPC), (2) increases in dorsolateral and anterior (except
for scene studies) PFC, (3) decreases in DPC, and (4) increases in VPC (except in face/â•‰
object studies). Taken together, the first two effects constitute the PASA pattern. We
will consider the PASA pattern and DPC decreases in the General Discussion. Here, we
focus on interesting item memory findings. We group these findings according to the
stimuli used (as in Table 12.1: scenes, words, and faces/â•‰objects).
Age-â•‰related SME increases in VPC are noteworthy given that in YAs, VPC tends
to exhibit negative SMEs (Cabeza, 2008; Cabeza et al., 2008; Uncapher and Wagner,
2009). Like other default mode network regions, VPC tends to be deactivated during
demanding cognitive tasks (Buckner et al., 2008). There is evidence that healthy OAs
and AD patients (Lustig et al., 2003; Persson et al., 2007) may not display deactiva-
tions in some default mode network regions. We will return to this issue in the section
on negative SMEs and RSEs.
308
Word studies showed prominent SME increases in PFC regions. Interestingly, sev-
eral of these studies found HAROLD (these findings cannot be seen in Table 12.1,
which does not display lateralization differences). For example, in Morcom et al.
(2003), SMEs in ventrolateral and dorsolateral PFC were left-â•‰lateralized in YAs but
bilateral in OAs. This HAROLD finding was replicated by Duverne et al. (2009) and
Dennis et al. (2007a) in the true memory condition. Another finding that is not dis-
played in Table 12.1 is from one study (Dennis et al., 2007b) that found PFC increases
when SMEs were measured at the trial level (transient activity), but decreases when
they were measured at the block level (sustained activity). The authors attributed
the decrease in block-â•‰level SMEs to a sustained attention deficit. Turning to OTP,
one study (Dennis et al., 2007a) found that increased left temporal activity in OAs
predicted subsequent false memory for associated words, possibly reflecting OAs’
greater reliance on semantic gist.
Finally, studies using faces/â•‰objects showed the PASA pattern but also two differ-
ent patterns than studies using scenes and words: they did not show any effect in
parietal cortex and showed increases in insula and ACC. The lack of parietal effects
might reflect the fact that faces/â•‰objects are primarily processed by the ventral pathway
(OTP). Both studies using faces reported reduced SMEs in the amygdala (Dennis et
al., 2008b; Fischer et al., 2010), which is a region frequently activated by face stimuli
(Breiter et al., 1996; Vuilleumier et al., 2001; Fusar-â•‰Poli et al., 2009). Insula increases
were found in studies using emotional stimuli (Kensinger and Schacter, 2008; Fischer
et al., 2010), consistent with its role in emotion. The ACC increases are consistent
with similar increases in associative SME studies.
Associative Memory SMEs (Associative SMEs)

In addition to DPC decreases, associative SME studies also report recurrent SME
increases in ACC and decreases in OTP and dorsolateral PFC. Despite the variability of
associative SME findings, it is worth noting a few interesting findings. First, evidence
of dorsolateral PFC decreases (Dennis et al., 2008b; Miller et al., 2008; Morcom et al.,
2010; Dulas and Duarte, 2011; Kim and Giovanello, 2011; Bangen et al., 2012) likely
reflects deficits in executive functions related to associative encoding. Two studies
investigated age effects on face-â•‰name associative memory (Miller et al., 2008; Bangen
et al., 2012), which is an important topic given OAs’ frequent complaints about diffi-
culties remembering people’s names (Zelinski and Gilewski, 1988; Leirer et al., 1990).
Similar to item SME studies, both showed a PASA pattern (i.e., occipital decreases
coupled with anterior PFC increases). Bangen et al. (2012), also found a HAROLD
pattern as PFC SMEs were left lateralized in YAs but bilateral in OAs.
The two studies investigating word-â•‰pair associative memory yielded inconsistent
findings: Kim and Giovanello (2011) found mostly SME decreases in OTP and PFC,
whereas de Chastelaine et al. (2011) found mostly SME increases in these regions,
largely driven by negative SMEs in YAs. Interestingly, in the Kim and Giovanello
(2011) study, there was an SME increase in perirhinal cortex (included in PHG col-
umn) that is consistent with evidence of age-â•‰related perirhinal compensation during
retrieval (Daselaar et al., 2006b), as discussed later in that section.
Like Bangen et al. (2012), the scene recollection study by Duzel et al. (2011) and
the context recognition study by Dulas and Duarte (2011) found SME increases in
â•‡ 309
HPC. This finding is difficult to explain because it is generally assumed that one of
the causes of associative deficits in OAs is a binding impairment during encoding
due to HPC dysfunction (e.g., Mitchell et al., 2000). However, HPC increases during
associative encoding are consistent with similar findings during associative retrieval,
as discussed later in that section. One possibility is that it reflects neural inefficiency
or compensation.
Lastly, it is important to note a recurrent null finding—â•‰that is to say, three studies
reported one or fewer age differences across the entire brain. Interestingly, two of these
studies utilized objects (Kukolja et al., 2009; Dulas and Duarte, 2014) while the third
utilized object-â•‰scene pairs (Leshikar and Duarte, 2014). Although object SME studies
did not show a similar pattern, these results indicate that associative SME differences
may be minimized when examining the context under which objects are encoded.
Age Effects on SMEs: Summary and Discussion

Similarities and differences between item and associative SME studies can be identi-
fied by comparing the bottom rows in the two panels in Table 12.1. Regarding simi-
larities, both item and associative SME studies frequently showed occipital decreases
and recurrent DPC decreases. Additionally, item SME studies exhibited a PASA
pattern (i.e., OTP decreases and dorsolateral and anterior PFC increases) that was
also observed in face-â•‰name associative SME studies. We will return to PASA and
DPC findings in the General Discussion. Regarding differences in recurrent findings,
one disparity is between dorsolateral PFC increases in item studies but decreases in
associative studies. This is consistent with evidence that associative memory is more
impaired than item memory in OAs. Another is between MTL decreases (both HPC
and PHG) in item studies but not in associative studies. Given that SME decreases
suggest impaired encoding processes, this difference is surprising because OAs tend
to be more impaired in associative than item memory. However, the MTL discrepancy
could reflect differences in the stimuli and/â•‰or methods used by item and associative
studies rather than true differences between item and associative memory.
In fact, the study by Dennis et al. (2008b), which directly compared age effects on
SMEs for associative vs. item memory within-â•‰participants, found the opposite pattern.
The associative and item conditions used comparable stimuli (face-â•‰scene pairs vs. indi-
vidual faces or scenes) and retrieval tests (associative recognition vs. old/â•‰new recogni-
tion). The results showed significant age-â•‰related SME reductions in HPC for associative
but not item memory (see Figure 12.1). This was the only associative encoding study to
report a significant SME x age interaction (all other studies reported SMEs in only one
group but not a significant interaction). Thus, it is possible that the MTL difference in
item vs. associative studies in Table 12.1 is more apparent than real and will disappear
with the addition of more studies. The same might be true for PFC differences.
Retrieval Success Effects (RSEs)

The top panel of Table 12.2 lists item studies while the middle and bottom panels list
associative studies. The middle panel includes studies that used recognition tests sen-
sitive to associative memory (associative recognition), such as Remember/â•‰Know and
310
12
10
YAs
8 *
HPC SME (Rem-For)
OAs
6
–2
–4
Faces Scenes Pairs
FigureÂ€12.1â•‡ OAs showed reduced SMEs in HPC for face-â•‰scene pairs but not for individual
faces or individual scenes, consistent with the idea that OAs are more impaired in associative
than item memory. Adapted with permission from Dennis NA, Hayes SM, Prince SE, Madden
DJ, Huettel SA, Cabeza R (2008) Effects of aging on the neural correlates of successful item
and source memory encoding. J Exp Psychol Learn 34: 791–â•‰808.
associative recognition tasks, whereas the bottom panel includes studies that used con-
text memory tests (context memory). Even though both associative recognition and
context memory tests are assumed to measure associative memory, the latter are more
dependent on the generation and monitoring of contextual information. Similarities and
differences among the three groups of studies are discussed in the RSE summary and
discussion section; the sections on each study only mention a few noteworthy findings.
Item Memory RSEs (Item RSEs)

As indicated by the bottom row of the item RSE panel, OAs exhibited recurrent RSE
decreases in DPC, temporal cortex, insula, ACC, and PFC regions, and increases in
occipital cortex. Interestingly, in Daselaar et al. (2003) study, RSE decreases in the
insula were observed in both high-â•‰and low-â•‰OAs, suggesting that decreases generalize
across individual differences. Dennis et al. (2008a) linked false memories for words
in OAs to left temporal increases, which they attribute to greater OAs’ reliance on
semantic gist in keeping with the role of left temporal cortex in semantic processing.
In contrast, a study using objects (Duarte et al., 2010) found mostly decreased false
memory SMEs in OAs, particularly in PFC, fusiform, and DPC. The authors suggest
that these decreases affect the ability to discriminate between studied and unstudied
stimuli, and may in part explain memory deficits in OAs.
Associative Memory: Recognition Tasks (Associative Recognition RSEs)

All recurrent findings in associative recognition studies were RSE reductions, includ-
ing decreases in parietal, OTP, midline, and PFC regions. The most consistent neu-
ral correlate for reduced associative retrieval in OAs was an RSE decrease in VPC
311
Table 12.2 Retrieval Success Effects (RSEs)
312
(Daselaar et al., 2006b; Duarte et al., 2008; Duarte et al., 2010; Giovanello et al., 2010;
Tsukiura et al., 2011; Dennis et al., 2014). Tsukiura et al. (2011) reported RSE reduc-
tions in several regions associated with associative memory, including parietal cortex,
dorsolateral PFC, HPC, and PMR when retrieving face-â•‰name/â•‰job pairs. In fact, HPC
decreases were reported by other studies as well (Daselaar et al., 2006b; Giovanello et
al., 2010), consistent with associative deficits in aging. Interestingly, studies reported
PHG decreases for during both true (Angel et al., 2013; McDonough et al., 2014) and
false (Dennis et al., 2014) recollection. Although Dennis et al. (2014) associated false
recollection in OAs with decreases in multiple regions including PHG, Giovanello
et al. (2010) associated it with a PHG increase. Another inconsistency emerges from
Wang et al. (2016), which did not report any age differences, although this lack of
differences is similar to associative studies examining object SMEs.
Associative Memory: Context Memory Tasks (Context RSEs)

In stark contrast to associative recognition studies, context studies reported recur-
rent RSE increases in most brain regions, including parietal, OTP, ACC, and PFC
areas. Five studies (Morcom et al., 2007; Dew et al., 2012; Dulas and Duarte, 2012;
Dulas and Duarte, 2014; Leshikar and Duarte, 2014) examined context memory for
encoding condition (e.g., in which encoding task was this item studied?). Within this
group, RSE increases in parietal and most PFC regions disappeared when context
memory misses were used as controls instead of correct rejections, and for two of
them—â•‰one using objects (Dulas and Duarte, 2014) and the other using object-â•‰scene
pairs (Leshikar and Duarte, 2014)—â•‰no age differences were reported. The difference
is particularly clear in the two contrasts from Dulas and Duarte (2012), in which
the context memory hits were the same and only the control condition was different.
AM (Recollection) IM (Familiarity)
1.8 0.7
1.6 0.6
HPC exponential rate
1.4
Rhinal linear slope
0.5
1.2
1 0.4
0.8 0.3
0.6
0.2
0.4
0.2 0.1
0 0
YAs OAs YAs OAs
Figure 12.2â•‡ OAs showed reduced associative-â•‰related HPC activity, but increased item-â•‰related
rhinal activity. These results suggest that OAs compensate for impaired associative processes
in HPC by over-recruiting item processes in rhinal cortex. Adapted with permission from
Daselaar SM, Fleck MS, Dobbins IG, Madden DJ, Cabeza R (2006) Effects of healthy aging
on hippocampal and rhinal memory functions: An event-â•‰related fMRI study. Cereb Cortex 16:
1771–â•‰1782.
â•‡ 313
These findings suggest that parietal and PFC increases for context memory could
be partly driven by age-â•‰related decreases in control correct rejections. Three studies
investigated context memory for background, spatial location, and temporal order.
The background memory study by Duverne et al. (2008), which used correct rejec-
tions as a control, found not only parietal increases but also decreases in PFC. The
object spatial location study by Kukolja et al. (2009) only found HPC decreases (and
was the only study to report HPC decreases), the finding most memory models would
predict for age effects on associative memory.
Age Effects on RSEs: Summary and Discussion

In this section, we summarize and discuss similarities and differences between item
memory (top panel of Table 12.2), associative recognition (middle panel), and asso-
ciative context (bottom panel) studies. Beginning with the comparison between item
and associative recognition, both types of studies found RSE decreases in DPC, tem-
poral cortex, ACC, and PFC regions. Both temporal and PFC decreases are consistent
with the PASA pattern. Turning to differences, several regions showed reduced RSEs
in associative but not item recognition studies, consistent with behavioral evidence
that OAs are more impaired in associative than item memory. One of the regions that
showed RSE reductions for associative but not item recognition was VPC, which fits
well with evidence linking this region to associative memory and recollection (Cabeza
et al., 2008; Vilberg and Rugg, 2008). Occipital cortex and PMR also showed more
reductions for associative than item recognition, perhaps reflecting the difficulty of
associative recognition tasks. The finding of HPC decreases in associative but not
item recognition studies is consistent with evidence that HPC is significantly impaired
by aging (Raz et al., 2005) and is more critical for associative than item memory
(Davachi, 2006; Eichenbaum et al., 2007). PHG decreases were also observed, which
may reflect a deficit in recollecting contextual information.
Given that it relates to nRSE effects reviewed in the next section, it is worth
noting here that although HPC is associated with associative memory and recollec-
tion, perirhinal cortex, another MTL region, is associated with item memory and
familiarity (Davachi, 2006; Eichenbaum et al., 2007). Consistent with the idea that
OAs are impaired in associative but not item memory, aging has different effects on
retrieval activity in these two regions. Whereas Table 12.2 shows that HPC effects
related to associative memory were reduced in OAs, Table 12.3 below shows that
perirhinal effects related to familiarity were actually increased in OAs (i.e., the
nRSEs were increased). For example, Daselaar et al. (2006b) compared the effects
of aging on the responses of HPC and perirhinal cortex to associative memory in
the form of recollection and item memory in the form of familiarity. Recollection-â•‰
related activity was quantified by the exponential rate, and familiarity by the slope
of the linear decreases (Brown and Aggleton, 2001; Henson et al., 2003; Daselaar
et al., 2006a; Montaldi et al., 2006; Wang et al., 2014). As illustrated by Figure
12.2, this contrast yielded a clear cross-â•‰over dissociation: OAs showed a reduction
in an associative-â•‰related effect in HPC, but an increase in an item-â•‰related effect
in perirhinal cortex. The authors suggested that OAs compensated for deficits in
HPC-â•‰mediated recollection by relying more on perirhinal-â•‰mediated familiarity.
314
Table 12.3 Negative SMEs and RSEs
â•‡ 315
This interpretation was supported by functional connectivity data and by a logistic

regression analysis showing that perirhinal cortex had a stronger effect on recogni-
tion accuracy in OAs than in YAs.
Finally, the most dramatic difference effect in Table 12.2 is that associative rec-
ognition showed mainly RSE decreases whereas context studies displayed primarily
RSE increases. On the basis of preexistent hypotheses, one would have expected that
the main difference between the three panels of Table 12.2 would be between associ-
ative and item memory, not between two types of associative memory. As mentioned
above, one speculative explanation for some RSE increases in context studies is the
use of correct rejections as the control condition, as several of the increases seem
to disappear when context misses are used as controls instead of correct rejections
(e.g., Dulas and Duarte, 2012). RSE increases could partly reflect a reduced nov-
elty response for correct rejections in OAs. In a study specifically examining novelty
responses, YAs but not OAs exhibited greater activity for correct rejections in OTP
and ACC; moreover, the region showing the strongest novelty effects in OAs was
ventrolateral PFC (Bowman and Dennis, 2015). Consistent with this, OTP and ACC
exhibit RSE increases in many context studies that compare context hits to correct
rejections, while ventrolateral PFC generally does not exhibit RSE increases (Table
12.2). However, this account cannot explain why the studies that used context misses
also yielded mostly increases, rather than mostly decreases as in associative recogni-
tion studies. We will return to the marked difference between associative and context
recognition findings in the General Discussion.
Negative SMEs (nSMEs) and Negative RSEs (nRSEs)

Previous sections focused on positive SMEs and RSEs, which occur when mem-
ory success is associated with increased activity. However, there are also negative
SMEs (nSMEs) and negative RSEs (nRSEs), which happen when memory success
is associated with reduced activity. Table 12.3 summarizes age effects on nSMEs and
nRSEs. The symbols are similar to those used in Tables 12.1 and 12.2, but in the
negative direction (e.g., a black square indicate a significant nSME/â•‰nRSE in YAs but
not in OAs, and a black circle means that YAs had a larger nSME/â•‰nRSE than OAs).
Although we mention a few recurrent findings, given the small number of studies,
these trends should be interpreted carefully as many could disappear with the addition
of new studies.
Starting with nSMEs, OAs showed reduced nSMEs in VPC, DPC, and dorsolat-
eralgt and anterior PFC. The nSME reduction in VPC is interesting because VPC
typically shows nSMEs in YAs (Otten and Rugg, 2001; Wagner and Davachi, 2001;
Cabeza, 2008; Uncapher and Wagner, 2009). The typical explanation for nSMEs in
VPC, PMR, and other regions in YAs is that encoding activity in these regions pre-
dict subsequent forgetting because these regions are involved in the processing of,
or bottom-â•‰up attention to, distracting information (Otten and Rugg, 2001; Wagner
and Davachi, 2001; Cabeza, 2008; Uncapher and Wagner, 2009). Given that VPC
and PMR are part of the default mode network, which has been linked to day-
dreaming and task-â•‰independent thoughts (Buckner et al., 2008), reduced nSMEs
316
in these regions can be described as a failure to suppress the default mode network
(Daselaar et al., 2004). Reduced nSMEs in OAs were first explained in the context
of OAs having difficulty suppressing the default mode network (Lustig et al., 2003;
Persson et al., 2007).
The nSME reduction in VPC is consistent with evidence that nSMEs in default
mode network regions tend to be attenuated in OAs. As illustrated by Figure 12.3,
for example, a large cross-sectional fMRI study including YAs, middle-aged adults,
and OAs found that nSMEs in typical default mode network regions, including VPC,
PMR, and ACC, decreased as function of age (Park et al., 2013). Also, a recent meta-
analysis of SMEs and aging found that within default mode network regions, neg-
ative SMEs in YAs partially overlapped with positive SMEs in OAs (Maillet and
Rajah, 2014).
Moreover, studies examining connectivity within default mode network regions
support the idea that OAs may differentially utilize default mode network regions
during memory encoding (Stevens et al., 2008; Rizio and Dennis, 2014). However, it
is important to emphasize that in the present review, only VPC showed evidence of
nSME reductions. Other core default mode network components, including PMR and
ACC, did not show these effects. Even in VPC, only two item studies and two associa-
tive studies showed the effect (see Table 12.3).
Figure 12.3 Increase in nSMEs as a function of age in several default mode network regions.
Reprinted with permission from Park H, Kennedy KM, Rodrigue KM, Hebrank A, Park DC
(2013) An fMRI study of episodic encoding across the lifespan: Changes in subsequent mem-
ory effects are evident by middle-age. Neuropsychologia 51: 448–56. (See color plate also)
â•‡ 317
Turning to nRSEs, there are two main accounts of these effects in YAs. First, when
nRSEs reflect more activity for correct rejections (new) than hits (old), one com-
mon explanation is “repetition suppression,” which refers to less activity for old than
novel items. Repetition suppression is assumed to mediate not only implicit memory
(priming) effects but also familiarity-â•‰related item memory during episodic recogni-
tion tasks (Brown and Aggleton, 2001; Henson et al., 2003). Second, nRSEs could
also reflect greater activity in regions mediating control processes when correct
rejections or misses are more demanding than hits. As illustrated by Table 12.3, OAs
showed increased nRSEs in PHG (item memory), and nRSE decreases in ACC (item
memory) and anterior PFC (associative memory). The age-â•‰related increases in PHG
occurred within perirhinal cortex (Daselaar et al., 2003; Daselaar et al., 2006b) and
could reflect greater reliance on familiarity, as previously discussed (see Figure 12.2).
In contrast, decreases in ACC and anterior PFC could reflect a difficulty in recruiting
control processes for correct rejections. Clearly, these interpretations are speculative
and further research is required.
Task-â•‰Based Functional Connectivity Studies of Recognition Memory

Previous sections reviewed age-â•‰related differences in the contribution of individual
brain regions to successful episodic encoding (SMEs) and retrieval (RSEs). However,
successful encoding and retrieval cannot be achieved by the operation of individual
brain regions; it also requires close interactions among these regions. FMRI studies
can investigate these complex interactions by measuring whether activity in different
regions co-â•‰varies over time, a measure typically known as functional connectivity
(here, just connectivity). Few studies have investigated connectivity associated with
encoding and retrieval tasks, and even fewer have examined how this task-â•‰based con-
nectivity is affected by aging. The review of these studies suggests at least four differ-
ent patterns of age-â•‰related episodic-â•‰related connectivity differences.
First, the most reliable task-â•‰based connectivity finding is an age-â•‰related increase
in PFC connectivity. In several studies, this PFC connectivity increase was coupled
with a decrease in connectivity with posterior brain regions, resembling the PASA
pattern often found in regional activity. Like PASA, the PFC connectivity increase has
usually been attributed to functional compensation (Daselaar et al., 2006b; Dennis et
al., 2008b; Murty et al., 2009; Oh and Jagust, 2013; Waring et al., 2013; Oedekoven
et al., 2015). In several studies, age-â•‰related increases in PFC connectivity were found
using neutral stimuli. For example, in the aforementioned word recognition study
by Daselaar et al. (2006b), OAs showed reduced MTL connectivity with posterior
cortical regions (VPC and retrosplenial cortex), but increased MTL connectivity with
bilateral dorsolateral PFC regions (Figure 12.4A), resembling a PASA pattern. This
pattern was also found in a previously reviewed study in which OAs showed SME
reductions in HPC during associative but not item encoding (Dennis et al., 2008b). In
this study, HPC connectivity in OAs was reduced with posterior temporal regions, but
was increased with PFC (Figure 12.4B). Moreover, two other studies reported greater
PFC connectivity with MTL in OAs, one during face-â•‰name recognition (Oedekoven
et al., 2015), and another during successful scene encoding (Oh and Jagust, 2013).
318
(A) L DLPFC R DLPFC (B) (C) L DLPFC
L IPL RSC HPC
MTL conn. HPC conn. Amygdala conn.
YAs > OAs OAs > YAs
Figure 12.4 PASA-like effect in functional connectivity. (A) OAs exhibited reduced MTL
connectivity with parietal and retrosplenial cortices, but increased connectivity with dorsolat-
eral PFC. Adapted with permission from Daselaar SM, Fleck MS, Dobbins IG, Madden DJ,
Cabeza R (2006) Effects of healthy aging on hippocampal and rhinal memory functions: An
event-related fMRI study. Cereb Cortex 16: 1771–1782. (B) OAs displayed reduced HPC con-
nectivity with OTP regions, but increased HPC connectivity with PFC regions. Adapted with
permission from Dennis NA, Hayes SM, Prince SE, Madden DJ, Huettel SA, Cabeza R (2008)
Effects of aging on the neural correlates of successful item and source memory encoding. J
Exp Psychol Learn 34: 791–808. (C) OAs showed reduced amygdala connectivity with HPC,
but increased connectivity with dorsolateral PFC. Adapted with permission from St. Jacques
PL, Dolcos F, Cabeza R (2009) Effects of aging on functional connectivity of the amygdala
during subsequent memory for negative pictures: A network analysis of fMRI data. Psychol Sci
20: 74–84. (See color plate also)
In another group of studies, greater PFC connectivity in OAs was found during
tasks involving emotional stimuli. In one study (St. Jacques et al., 2009), amygdala
activity during the encoding of emotional pictures was similar in YAs and OAs while
amygdala connectivity differed across groups: it was stronger with HPC in YAs and
with bilateral dorsolateral PFC regions in OAs (Figure 12.4C). Murty et al. (2009)
reported similar findings: OAs displayed reduced amygdala connectivity with HPC
but increased amygdala connectivity with dorsolateral PFC, both during encoding and
retrieval of emotional pictures. A third study (Addis et al., 2010) found age effects in
HPC connectivity during successful encoding (i.e., SMEs) of positive stimuli: HPC
connectivity was stronger with the thalamus in YAs but with PFC in OAs. Finally, in
a study in which participants viewed neutral scenes paired with positive or negative
items (Waring et al., 2013), parahippocampal connectivity with fusiform cortex was
reduced in OAs, whereas parahippocampal connectivity with PFC was increased in
OAs, particularly for subsequently remembered pairs (i.e., SMEs).
Second, another type of age-related effect on connectivity is reductions in OAs
among components supporting memory in the specific task employed. For exam-
ple, Tsukiura et al. (2011) found an age-related reduction in connectivity between
HPC and anterior temporal cortex for faces paired with names and jobs (i.e., asso-
ciative memory). Given that anterior temporal cortex was associated with memory
for names in this study, as well as in other studies, the authors suggested the reduced
319
Episodic Memory Encoding and Retrieval in the Aging Brain 319
temporal connectivity could contribute to OAs’ difficulty with memory for names.
Another example comes from a study of picture associative encoding (Leshikar et al.,
2010) where OAs exhibited reduced connectivity between HPC and occipital cortex.
Furthermore, when restricting the analyses to unrelated pictures, OAs were found to
have reduced connectivity between HPC and several OTP and PFC regions. A study
examining face recognition reported parietal connectivity with a DPC seed in both
age groups, but decreased connectivity with fusiform cortex (Oedekoven et al., 2013),
a region critical for face representations. A graph theory analysis showed significant
connectivity reductions in “long-range connections” between fronto-temporal, fronto-
occipital, fronto-parietal, and tempero-parietal regions during both memory encoding
and retrieval (Wang et al., 2010). The authors propose that this reduced connectivity
supports the idea that white matter declines across cortical regions may at least par-
tially explain cognitive deficits found in aging (O’Sullivan et al., 2001; Davis et al.,
2009). Lastly, a study from Gutchess et al. (2005) reported positive connectivity in
YAs but negative connectivity in OAs between parahippocampal cortex—a critical
scene processing region—and PFC during scene encoding. The authors suggest that
because parahippocampal cortex exhibited reduced SMEs, this negative correlation
reflects PFC compensation for impaired MTL encoding processes.
Third, some age-related increases in connectivity suggest a failure of inhibitory
control processes. For example, a face encoding study (Stevens et al., 2008) found an
age-related increase in nSMEs in auditory cortex (not shown in Table 12.1 because
only nSMEs for predicted ROIs were reported), possibly because OAs were distracted
by scanner noise. Consistent with this interpretation, auditory cortex in OAs displayed
stronger connectivity with temporal and parietal cortex and PFC regions, including sev-
eral default mode network regions typically suppressed during successful encoding. The
authors suggest that auditory distraction from the scanner environment may explain the
nSMEs in auditory cortex as well as its connectivity with default mode network regions.
A second study also found increased connectivity in OAs related to impaired inhibitory
control, but this increase was attributed to compensatory processes (Rizio and Dennis,
2014). This study investigated a directed forgetting manipulation in which participants
were instructed to forget some stimuli. This intentional forgetting was compared to nat-
urally occurring, incidental forgetting. Interestingly, VPC exhibited negative connectiv-
ity with HPC and parahippocampal cortex in OAs during intentional but not incidental
forgetting. The authors suggested that negative VPC connectivity mediates inhibitory
processes in OAs, who displayed a failure in PFC-mediated inhibition.
Finally, some age-related connectivity differences during episodic memory tasks
have been associated with a dedifferentiation of memory systems and subsystems.
For example, one study comparing episodic memory encoding and implicit learning
(Dennis and Cabeza, 2011) found that YAs showed a clear dissociation between mem-
ory systems, recruiting MTL during episodic encoding and the striatum during implicit
learning, whereas OAs showed no preferential recruit for either task. Consistent with
the idea of dedifferentiation, YAs displayed negative connectivity between MTL and
the striatum while OAs did not. Thus, the opposing relationship between episodic and
implicit learning systems (Poldrack et al., 2001) seems to disappear in OAs, with an
associated attenuation of negative connectivity. A similar age-related dedifferentiation
finding was recently reported by our laboratory between two subsystems of episodic
320
memory: associative and item memory (Wang et al., 2015). In a word recognition task,
each word was preceded by a masked word that was either conceptually related to the
target word or unrelated. The presence of the conceptual prime increased familiarity-â•‰
based false alarms and perirhinal activity reductions (repetition suppression) in both
age groups. OAs also exhibited activity increases in HPC, possibly due to dedifferen-
tiation (i.e., fluency response in both HPC and perirhinal cortex in aging). Increased
connectivity between HPC and perirhinal cortex was also observed in OAs during false
alarms that were primed by a conceptual cue, further supporting the possibility that
HPC may be recruited during familiarity-â•‰based decisions in healthy aging.
In sum, four different age effects on connectivity have been reported. First, the
most reliable pattern, reported in eight different studies, was stronger PFC connec-
tivity in OAs than YAs. This age-â•‰related increase in PFC connectivity was observed
during both encoding and retrieval using a variety of different tasks and stimuli. In
several cases, this effect was coupled with an age-â•‰related reduction in connectivity
with posterior regions, yielding a connectivity effect that resembles the PASA pat-
tern. Second, another age effect on connectivity is a reduction among components of
the network supporting task performance, such as reduced connectivity with anterior
temporal cortex during a face-â•‰name association task. Third, other studies reported
an age-â•‰related increase in connectivity consistent with a deficit in inhibitory control.
Finally, a fourth age-â•‰related difference in connectivity is consistent with a hypothe-
sized dedifferentiation between different memory systems.
General Discussion
FMRI studies have made an important contribution to our understanding of the effects
of aging on the neural correlates of episodic memory. Event-â•‰related fMRI studies have
been particularly useful because they can be used to directly compare successful and
unsuccessful memory trials during encoding (i.e., SMEs) and during retrieval (i.e.,
RSEs), allowing the identification of compensatory activity within-â•‰participants (i.e.,
increased SMEs or increased RSEs). Although there is substantial variability across
studies, our review has revealed several recurrent findings. In this section, we first
discuss the most consistent recurrent findings, and then consider factors that could
explain inconsistencies across studies.
Main Recurrent Findings

We focus here on the most frequent age-â•‰related findings on SMEs and RSEs: (1) the
PASA pattern in SMEs, (2) DPC reductions in SMEs and RSEs, (3) widespread RSE
increases during context recognition, and (4) increases in PFC connectivity.
Posterior–â•‰Anterior Shift with Aging (PASA) in SMEs

As shown in Table 12.1, item and face-â•‰name associative encoding studies showed
occipital SME decreases coupled with anterior PFC SME increases. These effects fit
â•‡ 321
with the PASA pattern, which was originally attributed to a visual processing deficit in
OTP compensated for by the recruitment of higher-â•‰order cognitive processes in PFC
(Grady et al., 1994). Given that all surveyed SME studies employed visual stimuli,
this interpretation could be also applied to the SME findings. Encoding fMRI stud-
ies that use visual stimuli frequently report significant SMEs in visual cortex (Kim,
2011), which makes intuitive sense given that better visual processing is likely to lead
to better subsequent memory for visual stimuli.
One potential criticism of the idea that OTP reductions reflect visual process-
ing deficits in OAs is that gross occipital volume is well preserved in OAs, particu-
larly when compared with regions such as PFC or HPC (Raz et al., 2005). However,
occipital decreases could reflect a reduction of the visual input due to deficits in the
peripheral visual system, including impaired crystalline lens transparency (Sekuler
and Sekuler, 2000) and reduced axons and myelin in the optic nerve (Peters, 2002).
Also, even if gross occipital volume is preserved, there is evidence of significant age-â•‰
related reductions in synapses (Peters et al., 2001b) and myelin (Peters et al., 2001a)
in primary visual cortex. The idea that age-â•‰related sensory deficits have a negative
impact on higher-â•‰order cognitive functions such as episodic memory is consistent
with evidence of significant correlations between sensory and cognitive measures in
OAs (Baltes and Lindenberger, 1997; Li and Lindenberger, 2002). One explanation
for these correlations is that sensory deficits cascade through the processing system,
impacting higher-â•‰order cognitive abilities. Consistent with this idea, degrading sen-
sory stimuli by adding noise yields cognitive deficits in YAs that resemble the ones
associated with aging (Pichora-â•‰Fuller et al., 1995; Murphy et al., 2000; Gilmore et
al., 2006).
Attributing SME increases in anterior PFC to compensation fits with our conser-
vative application of the term compensation only to cases in which activity is linked
to successful performance within-â•‰participants. The reason that PFC increases were
most frequent in anterior PFC is unclear, but the fact that this region is associated
with higher-â•‰order relational and abstract cognitive operations (Badre, 2008) suggests
that OAs may be relying on their spared conceptual knowledge to compensate for
impaired visual processes.
Although item SME and face-â•‰name associative SME studies yielded a PASA pat-
tern, other associative SME studies and RSE studies did not. A speculative expla-
nation of the difference in RSE studies is that memory could be less dependent on
executive control than encoding. For example, there is evidence that memory perfor-
mance suffers when attention is divided during encoding but not when it is divided
during recognition (Craik et al., 1996; Anderson et al., 1998). In general, however, it
seems that the PASA pattern is not reliably found across episodic memory studies.
Age-â•‰Related DPC Reductions in SMEs and RSEs

DPC shows recurrent age-â•‰related reductions in item SMEs and subthresholded in
associative SMEs (Table 12.1) as well as in RSEs during item and associative rec-
ognition tasks (Table 12.2). In YAs, DPC frequently shows SMEs (Uncapher and
Wagner, 2009) and RSEs (Cabeza et al., 2008), and both effects have been attributed
to top-â•‰down attention (Cabeza, 2008; Cabeza et al., 2008; Uncapher and Wagner,
322
2009). During encoding, top-â•‰down attention is necessary for focusing processing on

relevant new information, and during retrieval, for performing demanding memory
search and monitoring processes. Accordingly, age-â•‰related SME and RSE reduc-
tions in DPC could be attributed to a deficit in top-â•‰down attention. As mentioned
before, the resource deficit hypothesis postulates that insufficient attentional control
resources is one of the main causes of memory deficits in OAs, a hypothesis that
is consistent with evidence that divided-â•‰attention manipulations can yield memory
deficits in YAs that resemble those in OAs (Craik and Byrd, 1982; Rabinowitz et al.,
1982). A challenge for the resource deficit hypothesis is explaining why OAs show
reductions in top-â•‰down attention processes mediated by DPC but not in higher-â•‰order
control processes mediated by anterior PFC. This is an important question for future
research.
Age-â•‰Related RSE Increases in Context Studies

One of the most interesting patterns that emerged from our review is evidence of wide-
spread age-â•‰related RSEs increases in studies assessing associative memory with con-
text memory tasks. These increases are in stark contrast to age-â•‰related RSE decreases
in studies investigating recognition memory (see Table 12.2). Opposite age effects in
context vs. recognition memory RSE studies occurred in most brain regions. These
included regions associated with demanding retrieval search, monitoring, and top-â•‰
down attention processes (“retrieval effort”), such as PFC and DPC, as well as regions
associated with memory recovery (“retrieval success”), such as OTP and VPC. Given
that the RSE increases were more pronounced when context hits were compared to
correct rejections than when context hits were compared to context misses, we spec-
ulated that the increases could be partly driven by decreases in control correct rejec-
tions, perhaps related to a reduced novelty response in OAs. Here we consider two
alternative explanations.
In the case of regions associated with “retrieval effort,” one possible explanation
is that context memory tasks have greater executive control and top-â•‰down attentional
demands than recognition tests, such as the Remember/â•‰Know paradigm. Context
memory tasks require a demanding memory search for a specific target (e.g., was
this item encoded in context A or B?), as well as an effortful monitoring process (the
response can be either correct or incorrect), whereas the Remember/â•‰Know paradigm
requires only a subjective assessment of the conscious quality of the memories trig-
gered by a stimulus (e.g., are you “remembering?”) that is not constrained to one
specific criterion (“non-â•‰criterial recollection,” Yonelinas and Jacoby, 1996), and in
principle, responses are not correct or incorrect (the response is about a private con-
scious state). Thus, context memory tasks are more likely to feel demanding and trig-
ger compensatory “retrieval effort” processes than subjective recollection tests such
as the Remember/â•‰Know paradigm.
A related explanation is that compared to associative recognition tests, context
memory tasks are more dependent on late strategic retrieval processes, such as post-â•‰
retrieval monitoring, which are the ones that could be driving RSE increases in OAs.
There is evidence that OAs display a shift from a proactive to a retroactive decision
323
strategy (Paxton et al., 2008) and may compensate for deficits in early retrieval
processes by extending processing to later retrieval stages (Velanova et al., 2007).
Consistent with such a Early-to-Late Shift with Aging (ELSA), in one of the stud-
ies listed in the context panel (Dew et al., 2012), we found that OAs showed less
activity in PFC (and HPC) than YAs during an early, retrieval preparation phase, but
more activity than YAs during a late, retrieval completion phase (see Figure 12.5). As
illustrated by Figure 12.5, the ELSA account may apply not only to “retrieval effort”
regions such as PFC but also to “retrieval success” regions such as HPC. While it
is intuitive that OAs over-recruit control processes to compensate for impaired epi-
sodic memory, it is less clear how OAs can over-recruit retrieval success regions
given that their episodic memory is impaired. One possible explanation is that top-
down modulation from fronto-parietal regions increases activity in posterior regions
during demanding context memory tasks. At any rate, all three explanations of the
widespread age-related RSE increases during context tasks (correct rejection novelty,
control demands of context memory tasks, and ELSA) are highly speculative, and
additional evidence is clearly required. Ideally, researchers should directly compare
age effects on Remember/Know and context memory tasks within participants, while
manipulating variables relevant to the three hypotheses.
2
YAs
1.5
OAs
HPC RSE (Hit-Miss)
0.5
–0.5
–1
2
DLPFC RSE (Hit-Miss)
1.5
1
0.5
0
–0.5
–1
Preparation Retrieval
Figure 12.5 Consistent with an early-to-late shift in aging, in both left HPC and dorsolateral
PFC, OAs showed smaller RSEs than YAs during an early retrieval preparation phase but larger
RSEs than YAs during a late retrieval completion phase. Adapted with permission from Dew
ITZ, Buchler N, Dobbins IG, Cabeza R (2012) Where is ELSA? The early to late shift in aging.
Cereb Cortex 22: 2542–2553.
324
Age-â•‰Related Increases in PFC Connectivity

At least eight different studies reported age-â•‰related increases in PFC connectivity. In
some studies, the increases in PFC connectivity were coupled with decreases in pos-
terior connectivity (Daselaar et al., 2006b; Dennis et al., 2008b; Murty et al., 2009;
St. Jacques et al., 2009; Waring et al., 2013), whereas in other studies they were not
(Addis et al., 2010; Oh and Jagust, 2013; Oedekoven et al., 2015). The former fits
with the PASA pattern discussed above, whereas the latter seems to be a PFC-â•‰specific
phenomenon. In both cases, increases in PFC connectivity have usually been inter-
preted as evidence of compensation in OAs. This interpretation consistently points to
a role for PFC in age-â•‰related compensation as PFC regions not only show increased
SMEs and RSEs, but are also recruited for task-â•‰related functional networks related
to episodic memory. One important open question is how the increased connectivity
with PFC in OAs relates to disruptions in their white matter tracts (e.g., does increased
functional connectivity compensate for reduced structural connectivity?).
Potential Factors Explaining Inconsistent Findings Across Studies

While we observed several recurrent patterns across studies, there are also many
inconsistencies. For example, although several studies reported the HAROLD pattern,
or reduced PFC lateralization in OAs, in SMEs (Morcom et al., 2003; Dennis et al.,
2007a; Duverne et al., 2009; Bangen et al., 2012), many others failed to find this effect
when examining SMEs (Daselaar et al., 2003; Gutchess et al., 2005; de Chastelaine
et al., 2011) or context RSEs (Morcom et al., 2007; Dulas and Duarte, 2012). As
another example, there was no consistent evidence of associative SME reductions
in HPC—â•‰which would be one of the strongest a priori predictions given the role of
HPC in associative memory (Davachi, 2006; Eichenbaum et al., 2007) and evidence of
binding deficits in OAs (Naveh-â•‰Benjamin, 2000)—â•‰with some studies reporting reduc-
tions (Dennis et al., 2008b; Kim and Giovanello, 2011), and others reporting increases
(Duzel et al., 2011; Bangen et al., 2012). Even the findings we identified as recurrent
were not necessarily reliable (our definition of “recurrent” was only 25% of the stud-
ies). One possibility is that these effects (e.g., HAROLD or HPC-â•‰related deficits) are
not always present in OAs, but there are also several factors differing between studies
that would make it difficult to confirm this possibility with confidence. Within this
section we elaborate on several variables that may either separately or in cohort explain
these differences.
First, there are many ways to measure age-â•‰differences in fMRI studies of episodic
memory. As shown in Tables 12.1–â•‰12.3, one can be particularly stringent in defining
an age difference as an interaction (e.g., SME greater in one group than the other), or
simply as the presence of an effect in one group but not the other. Moreover, while we
did our best to separate studies reporting nSMEs and nRSEs from those that did not,
most studies did not report the direction of the interaction. That is, a positive SME
interaction (i.e., white circle) could be due to (1) a larger significant SME in OAs than
YAs, (2) a significant SME in OAs but a nonsignificant SME in YAs, (3) a significant
SME in OAs and a significant nSME in YAs, (4) a nonsignificant SME in OAs and a
significant nSME in YAs, or (5) a larger significant nSME in YAs than OAs. Studies
â•‡ 325
reporting these latter two patterns would have only been included in Table 12.3 if the
authors noted that such a pattern existed.
Relatedly, studies defined SMEs and RSEs differently. For item SMEs, successful
encoding can be defined as either subsequent high-â•‰confidence hits contrasted with
subsequent low-â•‰confidence hits or subsequent misses, subsequent hits contrasted
with subsequent misses, or some other possibility. For associative RSEs, this is fur-
ther complicated by the addition of correct rejection trials and also the utilization of
different retrieval paradigms (e.g., Remember/â•‰Know or context retrieval). Relatedly,
as mentioned in the introduction, item and associative tasks are not process pure.
Particularly, many item tasks that ostensibly tap item memory may also recruit asso-
ciative processes. Thus, some similarities seen between item and associative memory
studies may be due to contamination of these two forms of memory across tasks.
The third issue relates to individual differences in memory. While YAs had better
memory performance than OAs in many studies, they had balanced performance in
other studies. Although event-â•‰related studies allow us to examine memory success
rather than performance differences per se, differences in memory may still lead to an
interaction between memory differences and univariate activity. For example, a more
difficult task (in OAs) may lead to greater activity for hits than misses (while increased
activity for both hits and misses would not affect the memory success comparisons).
Lastly, studies utilized vastly different stimuli, which likely also contributed to the
inconsistencies across studies. Different cortical areas are critical for different stimuli
types (Epstein et al., 1999; Downing et al., 2001; Haxby et al., 2001), which may
weaken evidence of any recurrent findings. Regarding stimuli, one interesting null
result emerges across associative SME and RSE studies utilizing objects: all studies
that reported no age differences (or only one age difference) utilized object stimuli
or object stimuli in combination with another stimulus type. However, not all object
studies reported no age differences, therefore this remains pattern that requires further
study. For example, it may be the case that neural correlates of associative memory are
less unaffected by aging than previously thought (e.g., Wang et al., 2016). Together,
differences in how contrasts are reported, how memory success is defined, whether
memory performance is matched, and what stimuli are used may all contribute to the
inconsistent findings found across item and associative encoding and retrieval. Thus,
it remains important for future research to develop a more standardized approach to
studying age effects on the neural correlates of episodic memory.
Additional Methodological Considerations

We note in this section several methodological concerns that must be addressed in
future work in order to better advance our understanding of age-â•‰related effects on
episodic memory. One methodological issue revolves around the necessity of control-
ling for vascular changes. Given that fMRI signal is dependent on blood flow, cere-
brovascular reactivity changes in aging has significant implications for interpretations
of age-â•‰related differences in fMRI activity (Ito et al., 2002; D’Esposito et al., 2003;
Lu et al., 2011; Tsvetanov et al., 2015). For example, one study recently compared
uncorrected fMRI activity compared to cerebrovascular-â•‰reactivity-â•‰corrected fMRI
326
activity during memory encoding (Liu et al., 2013). The uncorrected activity for task
vs. fixation yielded results consistent with many studies in the aging literature, namely
a PASA pattern. However, the occipital and MTL decreases were no longer present
when accounting for cerebrovascular reactivity differences, while the PFC increases
became larger. When contrasting subsequent hits and misses, a similar pattern was
observed in PFC. Thus it is particularly important for future work to account for cere-
brovascular reactivity differences in aging, especially given that posterior cortices
most often exhibit age-â•‰related decreases (see Tables 12.1 and 12.2).
It is also critical to acknowledge that many of the studies reviewed in the current
chapter are cross-â•‰sectional rather than longitudinal. In brief, age-â•‰related differences
in brain and behavior observed in cross-â•‰sectional studies are not necessarily due to
aging (for a detailed discussion, see Chapters 6 and 7). Rather, they could be due to
differences between the YA and OA samples (i.e., selection bias or cohort effects),
or could be simply an age-â•‰invariant effect (e.g., worse memory or activity decreases
that were present throughout adulthood). For example, a comparison of cross-â•‰sec-
tional and longitudinal methods indicates that age-â•‰related over-recruitment of PFC
in a cross-â•‰sectional analysis actually manifest as underrecruitment in a longitudinal
analysis (Nyberg et al., 2010). A follow-â•‰up study from a subset of the same sample
reported that longitudinal increases in PHG activity—â•‰approximately in the perirhi-
nal cortex—â•‰correlated with longitudinal decreases in memory performance whereas
decreases in HPC activity (and volume) correlated with decreases in memory perfor-
mance (Persson et al., 2012). The former finding is inconsistent with a compensation
account, but could reflect inefficient processing. The latter finding is, however, con-
sistent with both cross-â•‰sectional studies (see Table 12.1) and theoretical proposals
(e.g., Naveh-â•‰Benjamin, 2000) that HPC dysfunction in aging contributes to episodic
memory deficits. Although these aforementioned studies examine gross activity (i.e.,
block design) rather than memory success differences, they offer promising insight
into age-â•‰related changes in episodic memory.
Lastly, the studies reviewed in the current chapter varied vastly in the statis-
tical power afforded by the sample sizes. The sample sizes of studies listed in
Tables 12.1–12.3 ranged from 19 (9 YAs, 10 OAs) to 192 (64 YAs, 64 middle-â•‰aged
adults, 64 OAs). A quantitative meta-â•‰analysis that fully accounts for differences in
power is beyond the scope of the current study (for meta-â•‰analyses, see Spreng et al.,
2010; Maillet and Rajah, 2014). Thus, it is possible that the age-â•‰related differences
found in studies with smaller sample sizes are less reliable than those with larger sam-
ple sizes and may explain some inconsistencies seen across studies.
Conclusion
In the current chapter, we reviewed event-â•‰related studies of item and associative epi-
sodic encoding and retrieval, as well as episodic memory studies of task-â•‰based func-
tional connectivity. While the field has advanced greatly over the past decade with the
use of these methods, many unanswered questions remain. Future work must seek
to adopt more standardized methods for assessing age differences in event-â•‰related
studies to better understand whether and how compensation—â•‰and other explanatory
â•‡ 327
models such as dedifferentiation and neural inefficiency—â•‰in the aging brain relate to
changes in episodic memory.
References
Achim AM, Lepage M (2005) Dorsolateral prefrontal cortex involvement in memory post-â•‰
retrieval monitoring revealed in both item and associative recognition tests. NeuroImage
24:1113–â•‰1121.
Addis DR, Leclerc CM, Muscatell KA, Kensinger EA (2010) There are age-â•‰related changes
in neural connectivity during the encoding of positive, but not negative, information.
46:425–â•‰433.
Alvarez P, Squire LR (1994) Memory consolidation and the medial temporal lobe: a simple
network model. Proceedings of the National Academy of Sciences 91:7041–â•‰7045.
Anderson ND, Craik FIM, Naveh-â•‰Benjamin M (1998) The attentional demands of encoding
and retrieval in younger and older adults: I. Evidence from divided attention costs. Psychol
Aging 13:405–â•‰423.
Angel L, Bastin C, Genon S, Balteau E, Phillips C, Luxen A, Maquet P, Salmon E, Collette F
(2013) Differential effects of aging on the neural correlates of recollection and familiarity.
49:1585–â•‰1597.
Bäckman L, Farde L (2004) The role of dopamine systems in cognitive aging. In: Cognitive
Neuroscience of Aging: Linking Cognitive and Cerebral Aging (Cabeza R, Nyberg L, Park
DC, eds), pp 58–â•‰84. New York: Oxford University Press.
Badre D (2008) Cognitive control, hierarchy, and the rostro-â•‰caudal organization of the frontal
lobes. Trends Cogn Sci 12:193–â•‰200.
Baltes PB, Lindenberger U (1997) Emergence of a powerful connection between sensory and
cognitive functions across the adult life span: a new window to the study of cognitive
aging? Psychology and aging 12:12–â•‰21.
Bangen KJ, Kaup AR, Mirzakhanian H, Wierenga CE, Jeste DV, Eyler LT (2012) Compensatory
brain activity during encoding among older adults with better recognition memory for
face-â•‰name pairs: an integrative functional, structural, and perfusion imaging study. Journal
of the International Neuropsychological Society 18:402–â•‰413.
Blumenfeld RS, Ranganath C (2007) Prefrontal cortex and long-â•‰term memory encoding: an inte-
grative review of findings from neuropsychology and neuroimaging. The Neuroscientist
13:280–â•‰291.
Bowman CR, Dennis NA (2015) Age differences in the neural correlates of novelty processing:
The effects of item-â•‰relatedness. Brain research 1612:2– 15.
Breiter HC, Etcoff NL, Whalen PJ, Kennedy WA, Rauch SL, Buckner RL, Strauss MM, Hyman
SE, Rosen BR (1996) Response and habituation of the human amygdala during visual
processing of facial expression. Neuron 17:875–â•‰887.
Brown MW, Aggleton JP (2001) Recognition memory: What are the roles of the perirhinal
cortex and hippocampus? Nat Rev Neurosci 2:51–â•‰61.
Buckner RL, Andrews-â•‰Hanna JR, Schacter DL (2008) The brain’s default network: anatomy,
function, and relevance to disease. Ann N Y Acad Sci 1124:1–â•‰38.
Burianová H, Lee Y, Grady CL, Moscovitch M (2013) Age-â•‰related dedifferentiation and com-
pensatory changes in the functional network underlying face processing. Neurobiology of
aging 34:2759–â•‰2767.
Cabeza R (2002) Hemispheric asymmetry reduction in old adults: The HAROLD Model.
Psychol Aging 17:85–â•‰100.
Cabeza R (2008) Role of posterior parietal regions in episodic memory retrieval: The dual
attentional processes hypothesis. Neuropsychologia 46:1813–â•‰1827.
328
Cabeza R, Dennis NA (2013) Frontal lobes and aging: Deterioration and Compensation. In:
Principles of frontal lobe function, Second Edition (Stuss DT, Knight RT, eds), pp. 628–652.
New York: Oxford University Press.
Cabeza R, Ciaramelli E, Olson IR, Moscovitch M (2008) The parietal cortex and episodic mem-
ory: an attentional account. Nat Rev Neurosci 9:613–625.
Cabeza R, Daselaar SM, Dolcos F, Prince S, Budde M, Nyberg L (2004) Age-related changes in
brain activity during working memory, attention, and episodic memory: Task-independent
vs. task-specific effects. Cereb Cortex 14:365–375.
Cabeza R, Grady CL, Nyberg L, McIntosh AR, Tulving E, Kapur S, Jennings JM, Houle S,
Craik FIM (1997) Age-related differences in neural activity during memory encoding and
retrieval: A positron emission tomography study. J Neurosci 17:391–400.
Castel AD, Craik FIM (2003) The effects of aging and divided attention on memory for item
and associative information. Psychol Aging 18:873–885.
Charlton R, Barrick T, Markus H, Morris R (2010) The relationship between episodic long-term
memory and white matter integrity in normal aging. Neuropsychologia 48:114–122.
Craik FI, Govoni R, Naveh-Benjamin M, Anderson ND (1996) The effects of divided atten-
tion on encoding and retrieval processes in human memory. Journal of Experimental
Psychology: General 125:159–180.
Craik FIM (1986) A functional account of age differences in memory. In: Human memory and
cognitive capabilities. (Klix F, Hagendorf H, eds), pp 409–422. New York: Elsevier Sceince
Publishers.
Craik FIM, Byrd M (1982) Aging and cognitive deficits: The role of attentional resources.
In: Aging and cognitive processes (Craik FIM, Trehub S, eds), pp 191–211. New York:
Plenum.
D’Esposito M, Deouell LY, Gazzaley A (2003) Alterations in the BOLD fMRI signal with age-
ing and disease: a challenge for neuroimaging. Nature Reviews Neuroscience 4:863–872.
Danker JF, Anderson JR (2010) The Ghosts of Brain States Past: Remembering Reactivates the
Brain Regions Engaged During Encoding. Psychol Bull 136:87–102.
Daselaar SM, Prince SE, Cabeza R (2004) When less means more: deactivations during encod-
ing that predict subsequent memory. Neuroimage 23:921–927.
Daselaar SM, Fleck MS, Cabeza RE (2006a) Triple Dissociation in the Medial Temporal
Lobes: Recollection, Familiarity, and Novelty. J Neurophysiol 96:1902–1911.
Daselaar SM, Veltman DJ, Rombouts SA, Raaijmakers JG, Jonker C (2003) Neuroanatomical
correlates of episodic encoding and retrieval in young and elderly subjects. Brain
126:43–56.
Daselaar SM, Fleck MS, Dobbins IG, Madden DJ, Cabeza R (2006b) Effects of healthy aging
on hippocampal and rhinal memory functions: an event-related fMRI study. Cerebral cortex
16:1771–1782.
Daselaar SM, Iyengar V, Davis SW, Eklund K, Hayes SM, Cabeza RE (2015) Less wiring, more
firing: low-performing older adults compensate for impaired white matter with greater
neural activity. Cerebral cortex 25:983–990.
Davachi L (2006) Item, context and relational episodic encoding in humans. Curr Opin
Neurobiol 16:693–700.
anterior shift in aging. Cereb Cortex 18:1201–1209.
Davis SW, Dennis NA, Buchler NG, White LE, Madden DJ, Cabeza R (2009) Assessing the effects
of age on long white matter tracts using diffusion tensor tractography. Neuroimage 46:530–541.
de Chastelaine M, Wang TH, Minton B, Muftuler LT, Rugg MD (2011) The effects of age,
memory performance, and callosal integrity on the neural correlates of successful associa-
tive encoding. Cerebral Cortex 21:2166–2176.
329
Dennis NA, Cabeza R (2008) Neuroimaging of healthy cognitive aging. In: The Handbook of
Aging and Cognition, Third Edition (Craik FIM, Salthouse TA, eds), pp. 1–54. New York,
NY: Psychology Press.
Dennis NA, Cabeza R (2011) Age-related dedifferentiation of learning systems: an fMRI study
of implicit and explicit learning. Neurobiol Aging 32:2318.e17–2318.e30.
Dennis NA, Kim H, Cabeza R (2007a) Effects of aging on true and false memory formation: an
fMRI study. Neuropsychologia 45:3157–3166.
Dennis NA, Daselaar SM, Cabeza R (2007b) Effects of aging on transient and sustained suc-
cessful memory encoding activity. Neurobiology of aging 28:1749–1758.
Dennis NA, Kim H, Cabeza R (2008a) Age-related differences in brain activity during true and
false memory retrieval. Journal of cognitive neuroscience 20:1390–1402.
Dennis NA, Bowman CR, Peterson KM (2014) Age-related differences in the neural correlates
mediating false recollection. Neurobiology of aging 35:395–407.
Dennis NA, Hayes SM, Prince SE, Madden DJ, Huettel SA, Cabeza R (2008b) Effects of aging on
the neural correlates of successful item and source memory encoding. J Exp Psychol Learn
Mem Cogn 34:791–808.
Dew ITZ, Buchler N, Dobbins IG, Cabeza R (2012) Where is ELSA? The early to late shift in
Downing PE, Jiang Y, Shuman M, Kanwisher N (2001) A cortical area selective for visual pro-
cessing of the human body. Science 293:2470–2473.
Duarte A, Henson RN, Graham KS (2008) The effects of aging on the neural correlates of sub-
jective and objective recollection. Cereb Cortex 18:2169–2180.
Duarte A, Graham KS, Henson RNA (2010) Age-related changes in neural activity asso-
ciated with familiarity, recollection and false recognition. Neurobiology of aging
31:1814–1830.
Dulas MR, Duarte A (2011) The effects of aging on material-independent and material-
dependent neural correlates of contextual binding. NeuroImage 57:1192–1204.
Dulas MR, Duarte A (2012) The effects of aging on material-independent and material-
dependent neural correlates of source memory retrieval. Cerebral cortex 22:37–50.
Dulas MR, Duarte A (2014) Aging affects the interaction between attentional control and source
memory: an fMRI study. Journal of cognitive neuroscience 26:2653–2669.
Duverne S, Habibi A, Rugg MD (2008) Regional specificity of age effects on the neural cor-
relates of episodic retrieval. Neurobiology of Aging 29:1902–1916.
Duverne S, Motamedinia S, Rugg MD (2009) The relationship between aging, performance,
and the neural correlates of successful memory encoding. Cerebral cortex 19:733–744.
Duzel E, Schutze H, Yonelinas AP, Heinze HJ (2011) Functional phenotyping of successful
aging in long-term memory: Preserved performance in the absence of neural compensation.
Hippocampus 21:803–814.
Eichenbaum H, Yonelinas AP, Ranganath C (2007) The Medial Temporal Lobe and Recognition
Memory. Annu Rev Neurosci 30:123–152.
Epstein R, Harris A, Stanley D, Kanwisher N (1999) The parahippocampal place
area: Recognition, navigation, or encoding? Neuron 23:115–125.
Fischer H, Nyberg L, Backman L (2010) Age-related differences in brain regions supporting
successful encoding of emotional faces. Cortex 46:490–497.
Fusar-Poli P, Placentino A, Carletti F, Landi P, Allen P, Surguladze S, Benedetti F, Abbamonte
M, Gasparotti R, Barale F (2009) Functional atlas of emotional faces processing: a voxel-
based meta-analysis of 105 functional magnetic resonance imaging studies. Journal of
Psychiatry & Neuroscience 34:418.
Gilmore GC, Spinks RA, Thomas CW (2006) Age effects in coding tasks: Componential anal-
ysis and test of the sensory deficit hypothesis. Psychol Aging 21:7–18.
330
Giovanello KS, Kensinger EA, Wong AT, Schacter DL (2010) Age-related neural changes
during memory conjunction errors. Journal of cognitive neuroscience 22:1348–1361.
Glisky EL, Polster MR, Routhieaux BC (1995) Double dissociation between item and source
memory. Neuropsychology 9:229–235.
Glisky EL, Rubin SR, Davidson PSR (2001) Source memory in oder adults: An encoding or
retrieval problem? Journal of Experimental Psychology: Learning, Memory, and Cognition
27:1131–1146.
Goh JO, Suzuki A, Park DC (2010) Reduced neural selectivity increases fMRI adaptation with
age during face discrimination. NeuroImage 51:336–344.
Grady CL, Maisog JM, Horwitz B, Ungerleider LG, Mentis MJ, Salerno JA, Pietrini P, Wagner
E, Haxby JV (1994) Age-related changes in cortical blood flow activation during visual
processing of faces and location. J Neurosci 14:1450–1462.
Gutchess AH, Welsh RC, Hedden T, Bangert A, Minear M, Liu LL, Park DC (2005) Aging and
the neural correlates of successful picture encoding: frontal activations compensate for
decreased medial-temporal activity. J Cogn Neurosci 17:84–96.
Haxby JV, Gobbini MI, Furey ML, Ishai A, Schouten JL, Pietrini P (2001) Distributed and overlap-
ping representations of faces and objects in ventral temporal cortex. Science 293:2425–2430.
Hayama HR, Rugg MD (2009) Right dorsolateral prefrontal cortex is engaged during post-retrieval
processing of both episodic and semantic information. Neuropsychologia 47:2409–2416.
Henson RN, Cansino S, Herron JE, Robb WGK, Rugg MD (2003) A familiarity signal in
human anterior medial temporal cortex? Hippocampus 13:259–262.
Ito H, Kanno I, Ibaraki M, Hatazawa J (2002) Effect of aging on cerebral vascular response
to Paco2 changes in humans as measured by positron emission tomography. Journal of
Cerebral Blood Flow & Metabolism 22:997–1003.
Kausler DH (1994) Learning and memory in normal aging. San Diego, CA: Academic Press.
Kennedy KM, Raz N (2009) Aging white matter and cognition: differential effects of
Kensinger EA, Schacter DL (2008) Neural processes supporting young and older adults’ emo-
tional memories. Journal of cognitive neuroscience 20:1161–1173.
Kim H (2011) Neural activity that predicts subsequent memory and forgetting: A meta-analysis
of 74 fMRI studies. NeuroImage 54:2446–2461.
Kim SY, Giovanello KS (2011) The effects of attention on age-related relational memory
deficits: fMRI evidence from a novel attentional manipulation. Journal of Cognitive
Koen JD, Yonelinas AP (2010) Memory variability is due to the contribution of recollection and
familiarity, not to encoding variability. Journal of Experimental Psychology: Learning,
Memory, and Cognition 36:1536–1542.
Koen JD, Yonelinas AP (2016) Recollection, not familiarity, decreases in healthy ageing:
Converging evidence from four estimation methods. Memory 24:75–88.
Kukolja J, Thiel CM, Wilms M, Mirzazade S, Fink GR (2009) Ageing-related changes of neural
activity associated with spatial contextual memory. Neurobiology of Aging 30:630–645.
Leirer VO, Morrow DG, Sheikh JI, Pariante GM (1990) Memory skills elders want to improve.
Experimental Aging Research 16:155–158.
Leshikar ED, Duarte A (2014) Medial prefrontal cortex supports source memory for self-refer-
enced materials in young and older adults. Cognitive, Affective & Behavioral Neuroscience
14:236–252.
Leshikar ED, Gutchess AH, Hebrank AC, Sutton BP, Park DC (2010) The impact of increased
relational encoding demands on frontal and hippocampal function in older adults. Cortex
46:507–521.
331
Li KZ, Lindenberger U (2002) Relations between aging sensory/sensorimotor and cognitive

functions. Neuroscience and Biobehavioral Reviews 26:777–783.
Li S-C, Lindenberger U, Sikström S (2001) Aging cognition: from neuromodulation to repre-
sentation. Trends in cognitive sciences 5:479–486.
Light LL (1991) Memory and aging: Four hypotheses in search of data. Annual Review of
Psychology 42:333–376.
Liu P, Hebrank AC, Rodrigue KM, Kennedy KM, Section J, Park DC, Lu H (2013) Age-related
differences in memory-encoding fMRI responses after accounting for decline in vascular
reactivity. NeuroImage 78:415–425.
Lu H, Xu F, Rodrigue KM, Kennedy KM, Cheng Y, Flicker B, Hebrank AC, Uh J, Park DC
(2011) Alterations in cerebral metabolic rate and blood supply across the adult lifespan.
RL (2003) Functional deactivations: change with age and dementia of the Alzheimer type.
Maillet D, Rajah MN (2014) Age-related differences in brain activity in the subsequent memory
paradigm: a meta-analysis. Neuroscience and Biobehavioral Reviews 45:246–257.
McClelland JL, McNaughton BL, O’Reilly RC (1995) Why there are complementary learning
systems in the hippocampus and neocortex: insights from the successes and failures of
connectionist models of learning and memory. Psychological Review 102:419.
McDonough IM, Wong JT, Gallo DA (2013) Age-related differences in prefrontal cortex
activity during retrieval monitoring: Testing the compensation and dysfunction accounts.
McDonough IM, Cervantes SN, Gray SJ, Gallo DA (2014) Memory’s aging echo: Age-related
decline in neural reactivation of perceptual details during recollection. NeuroImage 98:346–358.
Miller SL, Celone K, DePeau K, Diamond E, Dickerson BC, Rentz D, Pihlajamaki M, Sperling
RA (2008) Age-related memory impairment associated with loss of parietal deactivation
but preserved hippocampal activation. Proceedings of the National Academy of Sciences
Mitchell KJ, Johnson MK, Raye CL, D’Esposito M (2000) fMRI evidence of age-related hip-
pocampal dysfunction in feature binding in working memory. Cognitive Brain Research
10:197–206.
Montaldi D, Spencer TJ, Roberts N, Mayes AR (2006) The neural system that mediates famil-
iarity memory. Hippocampus 16:504–520.
Morcom AM, Li J, Rugg MD (2007) Age effects on the neural correlates of episodic
retrieval: increased cortical recruitment with matched performance. Cerebral Cortex
17:2491–2506.
Morcom AM, Good CD, Frackowiak RS, Rugg MD (2003) Age effects on the neural correlates
of successful memory encoding. Brain 126:213–229.
Morcom AM, Bullmore ET, Huppert FA, Lennox B, Praseedom A, Linnington H, Fletcher
PC (2010) Memory encoding and dopamine in the aging brain: a psychopharmacological
neuroimaging study. Cerebral Cortex 20:743–757.
Murphy DR, Craik FI, Li KZ, Schneider BA (2000) Comparing the effects of aging and back-
ground noise on short-term memory performance. Psychol Aging 15:323–334.
Murty VP, Sambataro F, Das S, Tan HY, Callicott JH, Goldberg TE, Meyer-Lindenberg A,
Weinberger DR, Mattay VS (2009) Age-related alterations in simple declarative memory and
the effect of negative stimulus valence. Journal of Cognitive Neuroscience 21:1920–1933.
Nadel L, Samsonovich A, Ryan L, Moscovitch M (2000) Multiple trace theory of human
memory: computational, neuroimaging, and neuropsychological results. Hippocampus
10:352–368.
332
Naveh-Benjamin M (2000) Adult age differences in memory performance: tests of an asso-

ciative deficit hypothesis. Journal of Experimental Psychology Learning, Memory, and
Cognition 26:1170–1187.
Naveh-Benjamin M, Guez J, Shulman S (2004) Older adults’ associative deficit in episodic
memory: Assessing the role of decline in attentional resources. Psychon Bull Rev
11:1067–1073.
Norman KA, O’Reilly RC (2003) Modeling hippocampal and neocortical contributions to rec-
ognition memory: a complementary-learning-systems approach. Psychological Review
110:611.
Nyberg L, Andersson M, Kauppi K, Lundquist A, Persson J, Pudas S, Nilsson L-G (2014)
Age-related and genetic modulation of frontal cortex efficiency. Journal of Cognitive
tion in aging. Proceedings of the National Academy of Sciences 107:22682–22686.
O’Sullivan M, Jones DK, Summers PE, Morris RG, Williams SC, Markus HS (2001) Evidence
57:632–638.
Oedekoven CS, Jansen A, Kircher TT, Leube DT (2013) Age-related changes in parietal lobe
activation during an episodic memory retrieval task. Journal of Neural Transmission
120:799–806.
Oedekoven CS, Jansen A, Keidel JL, Kircher T, Leube D (2015) The influence of age and mild
cognitive impairment on associative memory performance and underlying brain networks.
Brain Imaging and Behavior 9:776–789.
Oh H, Jagust WJ (2013) Frontotemporal network connectivity during memory encoding
is increased with aging and disrupted by beta-amyloid. The Journal of Neuroscience
33:18425–18437.
Otten LJ, Rugg MD (2001) When more means less: neural activity related to unsuccessful
memory encoding. Current Biology 11:1528–1530.
Paller KA, Wagner AD (2002) Observing the transformation of experience into memory. Trends
in Cognitive Sciences 6:93–102.
Park DC, Polk TA, Park R, Minear M, Savage A, Smith MR (2004) Aging reduces neural spe-
cialization in ventral visual cortex. Proceedings of the National Academy of Sciences of
the United States of America 101:13091–13095.
Park H, Kennedy KM, Rodrigue KM, Hebrank A, Park DC (2013) An fMRI study of epi-
sodic encoding across the lifespan: changes in subsequent memory effects are evident by
middle-age. Neuropsychologia 51:448–456.
Park J, Carp J, Kennedy KM, Rodrigue KM, Bischof GN, Huang C-M, Rieck JR, Polk TA,
Park DC (2012) Neural broadening or neural attenuation? Investigating age-related dedif-
ferentiation in the face network in a large lifespan sample. The Journal of Neuroscience
32:2154–2158.
Paxton JL, Barch DM, Racine CA, Braver TS (2008) Cognitive control, goal maintenance, and
prefrontal function in healthy aging. Cereb Cortex 18:1010–1028.
Persson J, Lustig C, Nelson JK, Reuter-Lorenz PA (2007) Age differences in deactivation: a
link to cognitive control? Journal of Cognitive Neuroscience 19:1021–1032.
Persson J, Pudas S, Lind J, Kauppi K, Nilsson L-G, Nyberg L (2012) Longitudinal structure–
function correlates in elderly reveal MTL dysfunction with cognitive decline. Cerebral
Cortex 22:2297–2304.
333
Peters A (2002) The effects of normal aging on myelin and nerve fibers: A review. Journal of
Neurocytology 31:581–593.
Peters A, Sethares C, Killiany RJ (2001a) Effects of age on the thickness of myelin sheaths in
monkey primary visual cortex. Journal of Comparative Neurology 435:241–248.
Peters A, Moss MB, Sethares C (2001b) The effects of aging on layer 1 of primary visual cortex
in the rhesus monkey. Cereb Cortex 11:93–103.
Pichora-Fuller MK, Schneider BA, Daneman M (1995) How Young and Old Adults Listen
to and Remember Speech in Noise. Journal of the Acoustical Society of America
97:593–608.
Poldrack RA, Clark J, Pare-Blagoev EJ, Shohamy D, Creso Moyano J, Myers C, Gluck MA
(2001) Interactive memory systems in the human brain. Nature 414:546–550.
Prince SE, Tsukiura T, Daselaar SM, Cabeza R (2007) Distinguishing the neural correlates of
episodic memory encoding and semantic memory retrieval. Psychol Sci 18:144–151.
Prull MW, Dawes LLC, Martin AM, Rosenberg HF, Light LL (2006) Recollection and familiar-
ity in recognition memory: Adult age differences and neuropsychological test correlates.
Rabinowitz JC, Craik FIM, Ackerman BP (1982) A processing resource account of age differ-
ences in recall. Canadian Journal of Psychology 36:325–344.
Raz N (2000) Aging of the brain and its impact on cognitive performance: Integration of struc-
tural and functional findings. In: The Handbook of Aging and Cognition, Second Edition
(Craik FIM, Salthouse TA, eds), pp. 1–90. Mahwah, NJ: Erlbaum.
Raz N, Gunning FM, Head D, Dupuis JH, McQuain J, Briggs SD, Loken WJ, Thornton AE,
Acker JD (1997) Selective aging of the human cerebral cortex observed in vivo: differen-
tial vulnerability of the prefrontal gray matter. Cerebral Cortex 7:268–282.
Reuter-Lorenz P, Jonides J, Smith ES, Hartley A, Miller A, Marshuetz C, Koeppe RA (2000)
Age differences in the frontal lateralization of verbal and spatial working memory revealed
by PET. J Cogn Neurosci 12:174–187.
Reuter-Lorenz PA, Lustig C (2005) Brain aging: reorganizing discoveries about the aging mind.
Curr Opin Neurobiol 15:245–251.
Ritchey M, Wing EA, LaBar KS, Cabeza R (2013) Neural similarity between encoding and
retrieval is related to memory via hippocampal interactions. Cerebral Cortex 23:2818–2828.
Rizio AA, Dennis NA (2014) The cognitive control of memory: age differences in the neural
correlates of successful remembering and intentional forgetting. PloS One 9:e87010.
Rosen AC, Prull MW, O’Hara R, Race EA, Desmond JE, Glover GH, Gabrieli JDE (2002) Variable
effects of aging on frontal lobe contributions to memory. Neuroreport 13:2425–2428.
Rugg MD, Vilberg KL (2013) Brain networks underlying episodic memory retrieval. Curr Opin
Neurobiol 23:255–260.
Rypma B, D’Esposito M (2000) Isolating the neural mechanisms of age-related changes in
human working memory. Nature Neuroscience 3:509–515.
Rypma B, Eldreth DA, Rebbechi D (2007) Age-related differences in activation-performance
relations in delayed-response tasks: a multiple component analysis. Cortex 43:65–76.
Salthouse TA (2004) What and when of cognitive aging. Current Directions in Psychological
Science 13:140–144.
Sekuler R, Sekuler AB (2000) Vision and aging. In: Encyclopedia of Psychology (Kazdin A,
ed), pp 180–183: American Psychological Association and Cambridge University Press.
334
Simons JS, Spiers HJ (2003) Prefrontal and medial temporal lobe interactions in long-term
memory. Nat Rev Neurosci 4:637–648.
Spencer WD, Raz N (1995) Differential effects of aging on memory for content and context: a
meta-analysis. Psychol Aging 10:527–539.
Spreng RN, Wojtowicz M, Grady CL (2010) Reliable differences in brain activity between
young and old adults: a quantitative meta-analysis across multiple cognitive domains.
Neuroscience & Biobehavioral Reviews 34:1178–1194.
St. Jacques PL, Dolcos F, Cabeza R (2009) Effects of aging on functional connectivity of the
amygdala during subsequent memory for negative pictures: A network analysis of fMRI
data. Psychological Science 20:74–84.
Stern Y, Rakitin BC, Habeck C, Gazes Y, Steffener J, Kumar A, Reuben A (2012) Task difficulty
modulates young–old differences in network expression. Brain Research 1435:130–145.
Stevens WD, Hasher L, Chiew KS, Grady CL (2008) A neural mechanism underlying memory
failure in older adults. The Journal of Neuroscience 28:12820–12824.
Sutherland GR, McNaughton B (2000) Memory trace reactivation in hippocampal and neocor-
tical neuronal ensembles. Curr Opin Neurobiol 10:180–186.
Tsukiura T, Sekiguchi A, Yomogida Y, Nakagawa S, Shigemune Y, Kambara T, Akitsuki Y, Taki
Y, Kawashima R (2011) Effects of aging on hippocampal and anterior temporal activations
during successful retrieval of memory for face-name associations. Journal of Cognitive
Tsvetanov KA, Henson RN, Tyler LK, Davis SW, Shafto MA, Taylor JR, Williams N, Rowe
JB (2015) The effect of ageing on fMRI: Correction for the confounding effects of vas-
cular reactivity evaluated by joint fMRI and MEG in 335 adults. Human Brain Mapping
36:2248–2269.
Tulving E (1984) Precis of elements of episodic memory. Behavioral and Brain Sciences
7:223–238.
Uncapher MR, Wagner AD (2009) Posterior parietal cortex and episodic encoding: insights
from fMRI subsequent memory effects and dual-attention theory. Neurobiol Learn Mem
91:139–154.
Velanova K, Lustig C, Jacoby LL, Buckner RL (2007) Evidence for frontally mediated con-
trolled processing differences in older adults. Cerebral Cortex 17:1033–1046.
Vilberg KL, Rugg MD (2008) Memory retrieval and the parietal cortex: A review of evidence
from a dual-process perspective. Neuropsychologia 46:1787–1799.
Vuilleumier P, Armony JL, Driver J, Dolan RJ (2001) Effects of attention and emotion on face
processing in the human brain: an event-related fMRI study. Neuron 30:829–841.
Wagner AD, Davachi L (2001) Cognitive neuroscience: forgetting of things past. Curr Biol
11:R964–R967.
Wang L, Li Y, Metzak P, He Y, Woodward TS (2010) Age-related changes in topological pat-
terns of large-scale brain functional networks during memory encoding and recognition.
Wang TH, Johnson JD, de Chastelaine M, Donley BE, Rugg MD (2016) The Effects of
Age on the Neural Correlates of Recollection Success, Recollection-Related Cortical
Reinstatement, and Post-Retrieval Monitoring. Cerebral Cortex 26:1698–1714.
Wang WC, Ranganath C, Yonelinas AP (2014) Activity reductions in perirhinal cortex predict
conceptual priming and familiarity-based recognition. Neuropsychologia 52:19–26.
Wang WC, Dew ITZ, Cabeza R (2015) Age-related differences in medial temporal lobe involve-
ment during conceptual fluency. Brain Research 1612:48–58.
Waring JD, Addis DR, Kensinger EA (2013) Effects of aging on neural connectivity underlying
selective memory for emotional scenes. Neurobiology of Aging 34:451–467.
335
Yonelinas AP (2001) Consciousness, control, and confidence: the 3 Cs of recognition memory.

Journal of Experimental Psychology: General 130:361–379.
Yonelinas AP (2002) The nature of recollection and familiarity: A review of 30 years of
research. J Mem Lang 46:441–517.
Yonelinas AP, Jacoby LL (1996) Noncriterial recollection: familiarity as automatic, irrelevant
recollection. Consciousness and Cognition 5:131–141.
Zelinski EM, Gilewski MJ (1988) Assessment of memory complaints by rating scales and ques-
tionnaires. Psychopharmacology Bulletin 24:523–529.
336
â•‡ 337
13
Emotion and EmotionalÂ€Memory
Elizabeth A. Kensinger
Jaclyn H.Â€Ford
A ging is associated with changes across mental and physical domains.

Yet despite losses in cognitive flexibility (see Chapters 8, 9, and 10) and
physical health (Deary et al., 2006), older adults’ emotional well-â•‰being is often as good
as that of younger adults, or even better (reviewed by Scheibe and Carstensen, 2010).
Moreover, as we describe in this chapter, older adults appear to retain the ability to
process and to remember emotional events, despite deficits in many related cognitive
processes. Cognitive neuroscience research has elucidated some of the reasons why
emotional processing and emotional memory do not seem as affected by age as other
cognitive processes. In this chapter, we first review how emotional events are pro-
cessed across the adult lifespan. Because how we process events is part-â•‰and-â•‰parcel of
how we encode them into memory (see Rugg, Otten, and Henson, 2002; Blumenfeld
and Ranganath, 2007), we interweave discussions of the effects of age on emotion
processing with discussions of emotional memory encoding. We then examine what is
known about how age affects emotional memory retrieval (see Figure 13.1).
Emotion Processing and Emotional Memory Encoding
Events that elicit emotion tend to be those that are surprising or distinctive, and of
some personal consequence (see Brown and Kulik, 1977; Talmi, 2013). These events
draw our attention and often become the target of elaborative processes and of addi-
tional rehearsal (see Mather and Sutherland, 2011; Talmi, 2013 for recent discussion).
As a result, these emotional experiences usually are retained in memory with a greater
likelihood than non-â•‰emotional events. Even decades later, people retain memory of
337
338
Encoding Consolidation Retrieval
Sleep-dependent Cue Content

Processing Orienting, Sustained Elaboration, Retrieval goals, processing Validity elaboration,
reactivation,
goals detection attention regulation expectations search, decision regulation
rehearsal
monitoring
Figure 13.1 Memory Timeline. The large boxes represent the three predominant phases of memory. The internal boxes represent
the sub-processes that take place within each phase and that have been shown to be influenced by emotion. The bolded processes
are those that are influenced by the interaction between emotion and age and that are focused upon in the chapter. The consolida-
tion box is in dotted lines because it is not the focus of the chapter due to the paucity of data investigating age-related changes in
these processes.
â•‡ 339
Emotion and EmotionalÂ€Memoryâ•… 339
highly emotional events such as assassinations, terrorist attacks, or natural disasters

(Brown and Kulik, 1977; Conway, 1995), even if not all the details of those memories
are accurate (Neisser and Harsch, 1992). As we describe next, this memory benefit
for emotional events is relatively preserved across the adult lifespan, as demonstrated
not only by studies of public events but also by studies assessing memory for stimuli
presented within a controlled, laboratory setting.
Many of the first studies examining the effects of age on emotional memory
focused on memory for events thought to elicit a “flashbulb memory” (Brown and
Kulik, 1977), a subjectively vivid memory for a surprising and consequential event.
Although some age differences have been revealed, as a whole, these studies suggest
that older adults can form vivid memories that meet the criteria for “flashbulb mem-
ory,” (reviewed by Kensinger, Allard, Krendl, 2014). Thus, although older adults often
show impairments in the ability to vividly remember prior events (e.g., Levine et al.,
2002; Spencer and Raz, 1995), they seem to retain the ability to remember emotional
experiences with a subjective vividness.
Because the public events assessed usually elicited negative emotions (but see
Holland and Kensinger, 2012; Petrican et al., 2008) and high-â•‰arousal states, these
studies left unanswered questions about the effect of age on memory for different
types of emotional events. Although there are many models of emotion (see Scherer,
2000 for review), one prevalent model (Russell, 1979) describes emotion in a two-â•‰
dimensional space consisting of valence (describing its pleasantness) and arousal
(describing its elicited excitement or agitation). Over the past decade, a major focus
of research on age-â•‰related changes in emotion processing and emotional memory has
been targeted at understanding how arousal and valence influence older adults’ event
processing and retention.
Effects of Arousal
The majority of research has demonstrated that younger and older adults’ attention is
directed toward (Hahn, Carlson, Singer, and Gronlund, 2006; Leclerc and Kensinger,
2008; Mather and Knight, 2006) and sustained by (Wurm et al., 2004) arousing stim-
uli. Thus, regardless of age, arousing information appears to be prioritized during
encoding. Older adults also show enhanced memory for arousing events as compared
to nonarousing ones, even when participants are not directly instructed to remember
the stimuli. They are more likely to recall (Charles et al., 2003; Kensinger et al., 2002;
Kensinger, 2008) or to recognize (Denburg et al., 2003; May et al., 2005; Rahhal,
May, and Hasher, 2002) arousing events and to remember some details associated
with arousing events (Nashiro and Mather, 2011; Kvavilashvili et al., 2010; Davidson,
Cook, and Glisky, 2006). These effects generally parallel those seen in younger adults
(see Kensinger et al., 2002 and Murray and Kensinger, 2013 for exceptions), and few
studies have revealed interactions between age and arousal when examining the pro-
cessing or retention of emotional events. This age similarity in the effects of arousal is
particularly interesting given some evidence that older adults may show reductions in
autonomic responses to arousing stimuli (Levenson et al., 1991; Labouvie-â•‰Vief et al.,
2003; but see Kunzmann and Gruhn, 2005).
340
Importantly, arousal does not enhance the processing of, or the memory for, all
event details. For instance, the presence of an arousing stimulus can impair the ability
to detect or to remember preceding or simultaneously presented stimuli. Although
the exact nature of these effects of arousal remains a point of discussion and research
(e.g., Bennion et al., 2013; Mather and Sutherland, 2011), in a general sense, the
effects of arousal on memory appear to be best described as selective enhancements.
This selectivity appears to be consistent across the adult lifespan (Denburg et al.,
2003; Kensinger et al., 2005; Kensinger, Gutchess, and Schacter, 2007; Nashiro and
Mather, 2011).
Many of these selective enhancements have been attributed to amygdala engage-
ment. Individuals with damage to the amygdala show neither the enhancing nor the
impairing effects of arousal on memory (e.g., Adolphs, Tranel, and Buchanan, 2005;
Strange, Hurlemann, and Dolan, 2003). A plethora of neuroimaging studies conducted
in younger adults have confirmed a link between amygdala engagement and success-
ful encoding of emotional events. The amygdala is activated as attention is drawn
toward salient, novel stimuli, and amygdala activity is greater during the processing of
subsequently remembered emotional items than during the processing of subsequently
forgotten items, even when the need to remember these stimuli is not emphasized to
participants (reviewed by Hamann, 2001; LaBar and Cabeza, 2006). Amygdala activ-
ity does not correspond with subsequent memory for all details of an arousing event,
however, consistent with the idea of selective enhancements (Kensinger and Schacter,
2006; Kensinger, Addis, and Atapattu, 2011).
With regard to successful encoding of arousing events, the amygdala is thought
to exert its effects through interactions with other medial temporal lobe regions
(reviewed by Hamann, 2001; LaBar and Cabeza, 2006). There is an increased like-
lihood of encoding information when there is co-activation of the amygdala and the
hippocampal complex, and the interaction may be reciprocal (Binder et al., 2012;
Richardson et al., 2004). This interaction is consistent with the purported role of the
amygdala in the processing of salient information (e.g., Liberzon et al., 2003; Sander
et al., 2003) and of the hippocampus in binding elements of an experience into a mem-
ory trace (see Ranganath, 2010).
In contrast to other medial temporal lobe regions, which often show relatively large
effects of age on brain structure and function (see Chapter 13), the effects of age on
volumetric measurements of the amygdala usually are found to be no greater than
what can be accounted for by the age differences in whole-brain volume (reviewed by
Nashiro, Sakaki, and Mather, 2012). The neural activity levels within the amygdala,
as measured via functional magnetic resonance imaging (fMRI), also seem to be rel-
atively preserved with aging. Older adults engage the amygdala during the process-
ing of emotionally arousing information (e.g., Krendl et al., 2009; Moriguchi et al.,
2011) and generally show comparable levels of engagement to younger adults when
age differences in the emotional response to stimuli are taken into account (see St.
Jacques et al., 2009a for discussion). Although older adults sometimes show reduced
functional connectivity between the amygdala and hippocampus during the encoding
of emotional information (Murty et al., 2009; St. Jacques et al., 2009b), older adults’
amygdalar engagement while encoding emotional experiences corresponds with the
likelihood that they will later remember the emotional events (Kensinger and Schacter,
â•‡ 341
2008; Fischer et al., 2010; Ritchey et al., 2011; St. Jacques et al., 2009b), suggesting
a relatively preserved link between the amygdala and emotional memory across the
adult lifespan (see discussion by Nashiro, Sakaki, and Mather, 2011).
The amygdala and hippocampal formation are not the only nodes within the emo-
tional network (Murty et al., 2010; see Figure 13.2), and age differences appear to be
more prevalent in other nodes of this network. Most notably, as compared to younger
adults, older adults often over-â•‰activate both lateral (e.g., Gunning-â•‰Dixon et al., 2003;
Murty et al., 2009; Roalf et al., 2011; Tessitore et al., 2005) and medial (e.g., Leclerc
and Kensinger, 2008a; Leclerc and Kensinger, 2011; Tessitore et al., 2005; Williams
et al., 2006) prefrontal regions when processing emotional material (reviewed by
Kensinger and Leclerc, 2009; St. Jacques, Bessette-â•‰Symons, and Cabeza, 2009).
Emotion processing is not the only domain in which age-â•‰related shifts toward pre-
frontal processing have been noted. A posterior-â•‰to-â•‰anterior shift with aging (PASA;
Davis et al., 2008; see Chapter 7) has been revealed in studies of attention and percep-
tion, working memory, and long-â•‰term memory. In these domains, the over-â•‰recruitment
of prefrontal regions has been proposed to be a compensatory mechanism, related to
older adults’ tendency to approach tasks in a more global, top-â•‰down, or controlled
fashion than younger adults (see also Grady, McIntosh, and Craik, 2003). Somewhat
similarly, the shift to prefrontal processing of emotional information has been inter-
preted as evidence that older adults process emotional information in a more con-
trolled fashion than younger adults (Ray and Zald, 2012), perhaps because they are
motivated to regulate their emotional reactions (Reed and Carstensen, 2012) or to
process the conceptual or self-â•‰referential meaning of emotional information (Ritchey
et al., 2011; Kensinger and Leclerc, 2009). This controlled-â•‰processing explanation
is consistent with the extant data examining older adults’ emotional memory perfor-
mance, insofar as their prefrontal engagement has been linked to more effective subse-
quent memory for emotional information (Kensinger and Schacter, 2008), particularly
when deep processing of information is encouraged (Ritchey et al., 2011; Murray and
Kensinger, 2013; Waring et al., 2013).
An important caveat to these findings, which we will elaborate upon in the follow-
ing section, is that the age-â•‰related shift toward prefrontal processing may not occur
equally for all emotional stimuli or for all individuals. Next, we review evidence that
the effects of age on prefrontal recruitment may depend on the valence of the infor-
mation being processed and may be most pronounced in older adults who show a
tendency to remember positive information disproportionately. More generally, we
address the possibility that—â•‰in contrast to the relative preservation of arousal effects
on emotion processing and emotional memory across the adult lifespan—â•‰larger
changes in valence-â•‰based processing may occur as adults age.
Effects of Valence
As noted earlier, emotional experiences can be described in a two-â•‰dimensional space
of valence and arousal. This yields two approaches for isolating valence effects on
age-â•‰related changes in emotion processing or emotional memory. The first, and most
often used, approach is to compare the processing and retention of stimuli that are
342
Medial PFC Amygdala Parahippocampus
Lateral PFC Hippocampus Ventral visual processing areas
Figure 13.2 Neuroanatomy of Emotional Memory. The key regions implicated in the formation and retrieval of emotional memories. (See color plate
also)
â•‡ 343
Arousal effects
selective memory enhancements
enhanced amygdala activity
amygdala-hippocampal connectivity
may be weakened with age;
PFC activity may compensate
Valence-specific effects
memory consistency & specificity reliance on gist
increased sensory recruitment increased PCF recruitment
especially or older adults may read of
preferential memory or positive items
neg pos
Valence-only Effects Valence-only Effects

younger adults show larger controlled processing older adults show larger
memory enhancement for memory enhancement for
strong age x valance
positive low-arousal items positive low-arousal
interaction in memory items
FigureÂ€13.3â•‡ Summary of Encoding Effects. Overview of the effects of arousal and the effects
of valence on memory encoding, shown on a circumplex model of emotion (Russell, 1980).
Bolded statements refer to consistent age-â•‰related differences seen across multiple studies.
of comparably high arousal but that differ in whether they are positive or negative.
We refer to these as valence-â•‰specific effects. The second approach is to compare the
processing and memory of stimuli that have emotional valence (i.e., are pleasant or
unpleasant) but do not elicit arousal. We refer to these as valence-â•‰only effects (see
Figure 13.3).
Valence-â•‰Specific Effects
Although positive and negative events both tend to be better remembered than neutral
events, valence can affect the quality of the memories. In most of these studies, as in
everyday life, participants’ encoding task did not direct them specifically to memorize
the materials. With this incidental encoding, negative events often are remembered
with a greater subjective vividness, resolution of detail, or consistency than positive
events (reviewed by Kensinger, 2009; but see Chipchase and Chapman, 2013; Waters
et al., 2013). Positive emotion, on the other hand, can lead to a broader scope of atten-
tion (Fredrickson, 2004; Clore et al., 2001) and to memory for the heuristics or gist of
an event (Kensinger, Garoff-â•‰Eaton, and Schacter, 2007; Yegiyan and Yonelinas, 2011).
344
Relatively few studies have examined how age affects the types of details, or
the quantity of details, that are retrieved for negative and positive events. That is,
given that an event is remembered, does age affect the types of details that will be
recalled about positive and negative events? The extant data suggest that the effects of
valence on the likelihood of remembering event details are fairly constant across the
adult lifespan. Younger and older adults who feel negative about the outcome of an
event remember its details more consistently or accurately over time than those who
feel positive about the outcome (Holland and Kensinger, 2012; see also Bohn and
Berntsen, 2007). Younger and older adults also remember more specific visual details
of negative items than of positive ones (Kensinger, Garoff-Eaton, and Schacter, 2007)
and can make more accurate distinctions between imagined and actual negative events
than positive ones (Kensinger, O’Brien et al., 2007).
These valence differences likely relate to the fact that, even when items are equated
for arousal, valence affects the regions that are recruited during event processing and
encoding. Although converging evidence suggests that the amygdala is engaged for
all salient stimuli, regardless of valence (e.g., Blackford et al., 2010; Liberzon et al.,
2003; Sander et al., 2003), negative valence may lead to greater recruitment of sensory
regions, while positive valence may lead to greater recruitment of prefrontal regions
(reviewed by Kensinger, 2009) and midline regions (Ritchey et al., 2011). More spe-
cifically, the encoding of negative images tends to be associated with greater fusiform
activity, and with greater functional connectivity between the fusiform and amyg-
dala, than does the encoding of neutral or positive images (reviewed by Kensinger,
2009). Conversely, the processing of positive items often is associated with greater
recruitment of medial and lateral prefrontal and parietal regions than negative items
(reviewed by Kensinger, 2009; see also Ritchey et al., 2011).
These differences in the regions engaged during the processing of positive and neg-
ative information appear to be fairly consistent across the adult lifespan (Kensinger
and Schacter, 2008; Addis et al., 2010). As noted in the prior section, older adults
tend to recruit the prefrontal cortex more than younger adults during the processing of
emotional information, and they more generally show a shift away from recruitment
of posterior regions and toward the recruitment of frontal regions during all encod-
ing sessions. Yet, despite their general shifts away from the recruitment of posterior
regions (Ritchey et al., 2011; Tessitore et al., 2005), older adults do still recruit visual
sensory regions more during the encoding of negative information compared to pos-
itive information (Kensinger and Schacter, 2008), and those sensory regions appear
to be connected to the amygdala (Addis et al., 2010; but see St. Jacques et al., 2009b)
or to other nodes of the emotional memory network (Waring et al., 2013) in older
adults. Also consistent with the younger adults’ pattern of neural engagement, older
adults’ successful encoding of positive information tends to be associated with greater
activity in a number of medial and lateral prefrontal regions (Kensinger and Schacter,
2008; Ritchey et al., 2011) as compared to their encoding of negative information.
Although younger and older adults may engage sensory regions more for negative
than for positive events, and may remember some details of negative events more accu-
rately or consistently, on the whole, older adults may have a propensity to remem-
ber the positive (Mather and Carstensen, 2005). Thus, age differences may be larger
when studies examine how valence affects the likelihood of remembering that an event
345
Emotion and Emotional Memory 345
occurred rather than how it affects the types of details associated with a remembered
event. In assessments of autobiographical memory, long-term memory for visual or
verbal stimuli, working memory, and decision making, older adults’ performance has
been shown to benefit disproportionately from positive valence (reviewed by Mather
and Carstensen, 2005; Reed and Carstensen, 2012). Thus, across a number of domains,
an interaction between age and valence has been revealed to affect task performance.
This “positivity effect” in older adults’ processing and retention of emotional infor-
mation has been associated with their engagement of cognitive control, a presumably
resource-limited system that directs complex information processing to be in line with
internal or external task goals (reviewed by Reed and Carstensen, 2012). Thus, the
positivity effect is strongest when older adults have resources available to devote to
processing positive information (Mather and Knight, 2005; Knight et al., 2007) and
when deep processing of the information is encouraged (Kryla-Lighthall and Mather,
2009; Mather and Knight, 2005; see also Ritchey et al., 2011).
Support for the controlled- processing hypothesis has come from event- related
potential (ERP) and fMRI studies. In ERP studies of emotion processing, an age by
valence interaction has been revealed in the amplitude of the Late Positive Potential
(LPP; Kisley et al., 2007; Wood and Kisley, 2006) and this neurophysiological marker
of a positivity effect co-occurs with a behavioral positivity effect in memory (Langeslag
and van Strien, 2009). This potential has been linked to motivated attentional process-
ing (Matsuda and Nittono, 2014), and so the age differences in this electrophysiologi-
cal signature are consistent with the presence of motivational differences between the
age groups. Because the engagement of prefrontal regions is frequently tied to the
implementation of cognitive control processes during emotion processing (reviewed
by Ray and Zald, 2012), older adults’ over-recruitment of prefrontal regions during
the processing of emotional information also is consistent with a controlled-processing
explanation for the age differences. Moreover, the older adults most likely to show
enhanced functional connectivity between the prefrontal cortex and amygdala at rest
(i.e., when no task is instructed) also are those who are most likely to disproportion-
ately remember positive information (Sakaki, Nga, and Mather, 2013).
When emotion processing is instructed, there is even greater evidence that age-
related enhancements in prefrontal engagement may not occur equally during the
processing of positive and negative stimuli. Older adults often activate the medial
prefrontal cortex more for positive than negative stimuli (reviewed by Kensinger and
Leclerc, 2008a; see also Cassidy et al., 2013; Ritchey et al., 2011), and older adults’
prefrontal engagement habituates more quickly for negative stimuli than for posi-
tive stimuli (Roalf et al., 2011). This pattern suggests a more sustained salience for
the positive stimuli among older adults. As compared to younger adults, older adults
also show increased activity (Kensinger and Schacter, 2008; Ritchey et al., 2011) and
increased functional connectivity (Addis et al., 2010; Waring et al., 2013; Waldinger et
al., 2011) among prefrontal regions when successfully encoding positive information.
One explanation for the resource-demanding processes that enable this positivity
effect is that older adults are utilizing controlled processes to regulate their emotional
responses (see Nashiro, Sakaki, and Mather, 2011; Reed and Carstensen, 2012). This
hypothesis stems from evidence that older adults are motivated to feel good in the
here-and-now, and this motivation may cause them to prioritize the processing of
346
positive rather than negative information (Mather and Carstensen, 2005). Supporting
this motivational hypothesis is evidence that task instructions can influence whether
age-related positivity effects emerge. For instance, when older adults are encouraged
to focus on information-gathering rather than regulatory goals, older adults no longer
show a positivity effect (Löckenhoff and Carstensen, 2007).
Although this regulatory account may explain the findings, the extant data can-
not clearly disambiguate this hypothesis from others that suggest age differences in
the engagement of other forms of top-down processing of emotional information.
The ventromedial prefrontal regions that often show interactions between age and
valence—showing stronger activity while younger adults process negative infor-
mation but while older adults process positive information (Leclerc and Kensinger,
2008a, 2010, 2011; Cassidy et al., 2013)—are implicated in processes other than emo-
tion regulation. For instance, these medial prefrontal regions have been proposed to
extract conceptual and self-referential information about affective stimuli, drawing on
information from long-term memory stores (Roy, Shohamy, and Wager, 2012). Thus,
age may affect the way this conceptual and self-referential information is assembled
(see Kensinger and Leclerc, 2009) or the way it is integrated with autobiographi-
cal information (see Ritchey et al., 2011). For example, there may be age-related
increases in the self-referential processing of positive information, if older adults are
more likely than younger adults to consider how positively valenced stimuli relate to
their self concept or connect to their autobiographical memories (see Kensinger and
Leclerc, 2009; Gutchess et al., 2007).
If a regulatory explanation for the valence-specific effects in prefrontal cortex is
correct, then the prefrontal activity should have an impact on amygdala engagement
and on experienced affect. When prefrontal processes are engaged for emotion reg-
ulation, there is not only increased frontal activity but also a corresponding change
in amygdala activity, and a decrease in experienced negative affect (reviewed by
Ochsner and Gross, 2005). Although some evidence suggests age disruptions in emo-
tion regulation ability (Opitz et al., 2012; Tucker et al., 2012; Winecoff et al., 2011),
older adults can engage prefrontal regions in the service of emotion regulation (Allard
and Kensinger, 2014), resulting in a down-regulation of amygdala activity (Winecoff
et al., 2011) and affective response (Krendl et al., 2009).
When age differences in amygdala engagement are present, they have been in the
direction predicted by a regulatory account: less amygdala activity for negative than
for positive stimuli (reviewed by Reed and Carstensen, 2012), and an inverse relation
between prefrontal activity and amygdala activity during the processing of negative
information (e.g., St. Jacques et al., 2009b). Two studies also have revealed individual
differences consistent with a regulatory account. In one study (Erk et al., 2008), older
adults who endorsed using reappraisal strategies more frequently showed reduced
amygdala activity to negative stimuli. In another study (Waldinger et al., 2011), only
older adults who rated themselves to have high life satisfaction showed stronger func-
tional connectivity between the amygdala and other nodes of an emotional memory
network when successfully encoding positive compared to negative images. It is
important to remember, however, that age differences in amygdala engagement are
not always found; in fact, such differences are infrequently revealed when stimuli
are equated for arousal in younger and older adults (see St. Jacques et al., 2009a).
â•‡ 347
Although these null effects of age could occur for technical as well as theoretical
reasons, they suggest the need for future research to elucidate whether older adults’
prefrontal engagement is tied to regulatory actions (e.g., changes in amygdala activity
and in experienced affect) only in some circumstances.
Although most studies have examined effects of age on valence-â•‰specific effects
on high-â•‰arousal stimuli, which activate the amygdala, it is also possible to examine
valence-â•‰specific effects at lower levels of arousal as well. There is some evidence
to suggest that the interactions between age and valence may be even more appar-
ent in neural recruitment (Leclerc and Kensinger, 2011) and in behavioral outcomes
(Kensinger, 2008; Mickley and Kensinger, 2009) for those low-â•‰arousal stimuli. As
we discuss in the next section, this increased effect of age may.be because processing
low-â•‰arousal stimuli relies on controlled processing to a greater degree than processing
high-â•‰arousal stimuli; thus, low-â•‰arousal stimuli may recruit processes that are most
likely to be differentially deployed by younger and older adults.
Valence-â•‰only Effects
Stimuli can be pleasant or unpleasant and yet evoke no arousal response. For instance,
people understand the affective meaning of “fatigued” or “relaxed” but often have no
subjective feeling of arousal nor a physiological response to the words. Despite this
lack of an arousal response, memory for these words often is better than for words
without affective meaning (Kensinger and Corkin, 2003; Kensinger and Corkin,
2004).
Although memory is enhanced for these valenced stimuli, the mnemonic processes
appear to differ from those that guide memory for high-â•‰arousal stimuli. The enhanced
encoding of these valenced words appears to be accounted for by increases in the
same processes that enable the encoding of other deeply-â•‰processed neutral words.
At a neural level, activity in the left ventrolateral prefrontal cortex and hippocampus
leads to subsequent memory for both valenced-â•‰only and neutral stimuli, but valenced-â•‰
only stimuli engage the lateral prefrontal processes to a greater degree (Kensinger and
Corkin, 2004). This finding is consistent with the proposal that valenced-â•‰only stimuli
are remembered well because they are deeply processed, insofar as the lateral pre-
frontal cortex has been connected to the elaborative processing of verbal stimuli (e.g.,
Blumenfeld and Ranganath, 2007). Further evidence for an overlap in the types of
processes that support the encoding of valenced-â•‰only and neutral stimuli comes from
studies that have divided the attention of younger adults as they encode the stimuli:
The memory enhancement for these valence-â•‰only stimuli disappears as soon as atten-
tion is divided (Kensinger and Corkin, 2004). Without the ability to engage controlled,
elaborative processing, the memory advantage for these stimuli no longer exists. This
result directly counters the often-â•‰found results for high-â•‰arousal stimuli, which retain
their memory advantage even under conditions of divided attention (e.g., Kensinger
and Corkin, 2004; Kern et al., 2005; Kang et al., 2014).
In the one study to compare valence-â•‰only memory enhancement effects in younger
and older adults, younger adults showed a memory enhancement for both positive and
negative valence-â•‰only words as compared to neutral words, although the enhancement
was significantly higher for the negative words (Kensinger, 2008). Older adults, by
348
contrast, showed a large memory enhancement for the positive words but not for the
negative words (Kensinger, 2008). A related finding was reported in Mickley and
Kensinger (2009), with older adults reporting greater subjective vividness for mem-
ories of positive low-â•‰arousal events but younger adults reporting greater subjective
vividness for memories of negative low-â•‰arousal events. In both of these studies, the
age-â•‰related positivity effect was more pronounced for low-â•‰arousal stimuli than for
high-â•‰arousal stimuli.
Because the valence-â•‰only memory enhancement is thought to rely on controlled
processing, this pattern is consistent with the proposal that older adults’ positivity
effect relates to the way in which they engage controlled processing resources. While
these findings support the role of controlled processing in the effects of age on emo-
tional memory, it is unclear whether they are consistent with the regulatory account
in particular. One recent fMRI study suggests they may be: Dolcos and colleagues
(2014) revealed a neural pattern that they interpreted as suggestive evidence that older
adults chronically activate regulatory goals while processing low-â•‰arousal stimuli but
not high-â•‰arousal stimuli. In this study, age differences were greater during the pro-
cessing of low-â•‰arousal images than high-â•‰arousal images. For the low-â•‰arousal stimuli,
older adults showed greater activity than younger adults in the ventral ACC, and they
also showed lesser activity than younger adults in the amygdala. Moreover, the greater
the older adults’ ventral ACC activity, the lower their arousal ratings for the low-â•‰
arousal negative images. Thus, although participants were not instructed to regulate
their emotions, the older adults’ pattern of neural engagement for the low-â•‰arousal
stimuli suggests they may have done so spontaneously. It is therefore possible that
older adults’ greater positivity effect for low-â•‰arousal than for high-â•‰arousal stimuli
could be explained by their greater tendency to evoke regulatory strategies for the
low-â•‰arousal stimuli. There is not yet direct support for this hypothesis, however, as no
research has examined how arousal affects the links between age, emotion regulation,
and memory performance.
Emotional Memory Retrieval
Although age effects on neural recruitment during encoding of emotional memories

have been well-â•‰characterized, effects during retrieval have received considerably less
attention. Nevertheless, there are a number of behavioral lines of evidence that point
to age differences in emotional retrieval processes. First, even when younger and older
adults’ emotional responses to events do not differ initially, older adults’ subsequent
retellings are more likely to become increasingly positive (Kennedy, Mather, and
Carstensen, 2004; Levine and Bluck 1997). Such a change over time is unlikely to
be accounted for by age differences in the original experience of the event and more
likely reflects retrieval differences. Moreover, such increases in positivity can occur
even when older adults recall events from when they were younger adults, suggesting
that the effects must relate to their older age at retrieval. Second, older adults some-
times adopt a more liberal responses bias with positive stimuli compared to younger
adults, such that they have disproportionately more false alarms to positive relative to
negative items (Kapucu et al., 2008; Fernandes et al., 2008), particularly for related
349
lures (Piguet et al., 2008). Although response bias can be influenced by the depth or
specificity with which information is encoded, it also critically depends upon retrieval
monitoring and retrieval expectations (e.g., Gallo, 2013). Finally, instructions pre-
sented at the time of retrieval can shift emotional memory effects so that younger adults
perform more like older adults (Kennedy et al., 2004; Mather and Johnson, 2000) or
older adults perform more like younger adults (Kennedy et al., 2004; Hashtroudi et
al.,1994). These lines of evidence suggest that age differences in emotional memory
performance are not solely a product of age-related changes at encoding.
Neuroimaging methods provide an important tool for understanding the mecha-
nisms active during retrieval, because they enable the study of the processes invoked
at the moment a past event is recognized or recalled. In younger adults, fMRI stud-
ies have highlighted an emotional memory retrieval network similar to the network
engaged during encoding (see Buchanan, 2007 for review). Specifically, emotional
memory retrieval is associated with a bilateral network that includes prefrontal,
medial-temporal, medial-parieto-occipital, lateral parietal, anterior cingulate, occipi-
tal, and cerebellar regions (Cabeza and Nyberg 2000; Spaniol et al., 2009). The over-
lap between the encoding and retrieval networks suggests that recalling an emotional
event elicits some degree of recapitulation of both the features of the original event as
well as the emotional experience associated with it (see Chapter 12).
Despite the behavioral evidence suggesting that aging may significantly alter neu-
ral recruitment during emotional memory retrieval, to date, there are only a handful
of fMRI and ERP studies specifically addressing these changes (see Figure 13.4).
Notably, study methodology deviates drastically across studies, making it difficult
to draw firm conclusions regarding the processes being recruited or to disambiguate
arousal from valence effects, as the literature has done at encoding. Nevertheless,
the studies do begin to identify a pattern whereby younger and older adults recruit
similar neural networks during retrieval of emotional events, but with significant age-
related increases in prefrontal activity. Interestingly, these age-related increases in
prefrontal activation can be greater during negative event retrieval than neutral (Murty
et al., 2009) event retrieval. Although these age differences can be seen even when
memory is tested after relatively short delays, the differences may become greater
after longer delays. One study revealed that older adults engaged the prefrontal cortex
more when negative event retrieval occurred after a three week delay compared to a
one week delay, while younger adults showed the opposite pattern (Kalpouzos et al.,
2012). In addition, there is evidence that age-related increases in prefrontal recruit-
ment are greater for negative relative to positive event retrieval (Ford et al., 2015;
Ford and Kensinger, 2014). It is notable that such patterns are the opposite of the age-
by-valence interactions seen during encoding, where older adults recruit prefrontal
regions more during retrieval of positive compared to negative events (e.g., Leclerc
and Kensinger, 2008a). Such reversals may reflect the fact that emotion influences
different mechanisms at encoding and retrieval.
In the sections below, we present alternate accounts to explain these age-related
changes. We first discuss the possibility that age-related increases in prefrontal activ-
ity during negative event retrieval may be a compensatory response, offsetting other
age-related disruptions present during negative event retrieval. Although this account
cannot be firmly ruled out, the extant data suggest that this account may not provide
350
Methodology fMRI fMRI/DTI ERP
Continuous Face Source

Task Recognition Recognition
Recognition Recognition Memory
Positive v. Positive v. Positive v.

Emotion Negative v. Positive v. Negative v.
Negative v. Negative v. Negative v.
Contrast Neutral Negative Neutral
Neutral Neutral Neutral
Intentional*
Encoding Incidental Intentional Intentional Intentional (*Continuous Incidental Incidental
Recognition)
1 week and 3
Delay 2 Min 30 Min 30 Min N/A 1 Hour Immediate
weeks
Ford & Langeslag &

Murty et al., Kalpouzos Ford et al., Schefter Newsome
Study Kensinger, Van Strien,
2009 et al., 2012 2015 et al., 2012 et al., 2012
2014 2008
Figure 13.4 Procedures Used in Studies Examining Emotional Memory Retrieval.

â•‡ 351
the best explanation for age-â•‰by-â•‰valence interactions. We then discuss an emotion reg-
ulation account, whereby the age-â•‰related increases in prefrontal activity may serve to
dampen the negative emotions experienced during retrieval. The extant data are more
consistent with this account, yet we also note some empirical findings that suggest
the need to look beyond compensatory or regulatory accounts to fully understand the
reasons for the enhanced prefrontal recruitment.
Compensatory Accounts
In non-â•‰emotional domains, age-â•‰related increases in prefrontal activation during episodic
memory retrieval often have been proposed to serve a compensatory function, support-
ing successful performance despite decreased function in more posterior neural regions
(Davis et al., 2008). Therefore, one potential explanation for age-â•‰by-â•‰valence interac-
tions in prefrontal recruitment is that they, too, reflect a compensatory shift in neural
activity from posterior to anterior regions for negative relative to positive and neutral
event retrieval. In other words, increased prefrontal recruitment could compensate for
other declines during negative event retrieval and aid older adults’ memory for negative
events.
Indeed, Murty and colleagues (2009) found that young and older adults both exhib-
ited increased performance during negative relative to neutral memory retrieval (they
did not assess memory for positive), but the associated neural processes differed in
the two age groups. The authors reported age-â•‰related decreases in amygdalar activity
and age-â•‰related increases in dorsolateral PFC activity during negative event retrieval;
based on this pattern, the authors suggested that older adults may be recruiting prefron-
tal regions to compensate for decreased limbic activity during negative event retrieval.
Although such a pattern generally is consistent with a compensatory account, recogni-
tion accuracy was not correlated with prefrontal recruitment during retrieval. Notably,
accuracy was correlated with prefrontal recruitment during encoding, particularly in
older adults, suggesting that age-â•‰related increases in prefrontal activity may serve a
greater compensatory function at encoding than at retrieval.
The two studies that have compared retrieval of negative and positive events have
revealed greater age-â•‰related increases in prefrontal recruitment for negative relative
to positive event retrieval (Ford et al., 2015; Ford and Kensinger, 2014). Because
age-â•‰by-â•‰valence interactions during encoding (such as differences in self-â•‰referential
processing, see Kensinger and Leclerc, 2009) may cause positive information to be
processed more deeply than negative in older adults, it is possible that increased
mnemonic support may be necessary for negative event retrieval in older adults.
Enhanced difficulty of negative event retrieval may explain greater age-â•‰ related
increases in prefrontal activation for negative relative to positive events to equate
performance in negative and positive events (Ford et al., 2015). In other words, the
enhanced prefrontal activity during retrieval could be compensating for the weaker
strength with which that negative information initially was encoded. However, stron-
ger evidence for this view would be revealed by a correlation between prefrontal
engagement and hit rates for the negative items—â•‰a correlation not revealed by the
extant data.
352
Regulatory Accounts
An alternative explanation for increased prefrontal activity during negative event
retrieval is that older adults are motivated to regulate their emotional state to a greater
extent than younger adults. As reviewed above, there is an extensive literature suggest-
ing that older adults are more motivated to maintain a positive outlook (Carstensen,
1995) and have superior regulation abilities (Blanchard-â•‰Fields, 2007) compared to
younger adults. Therefore, although regulation strategies require prefrontally-â•‰medi-
ated cognitive control processes (Mather and Carstensen, 2005), and healthy aging has
been associated with significant reductions in these processes (Dennis and Cabeza,
2008), older adults may be motivated to allocate more of their limited resources to
recruiting regulation processes (see Hess, 2014 for relevant discussion). Such age-â•‰
related shifts in motivation have been suggested by behavioral studies examining
divided attention during encoding (Mather and Knight, 2005), but regulation may also
alter the way older adults interact with to-â•‰be-â•‰remembered stimuli, both in the time
between encoding and retrieval and at the time of retrieval. Specifically, prefrontal
regions may regulate emotions by guiding retrieval of emotional details and either
increasing or decreasing neural activity in limbic regions associated with emotional
responses and memory retrieval (Ochsner and Gross, 2005). Such an explanation
could account for increased recruitment of prefrontal regions during negative relative
to positive or neutral event retrieval in older adults.
The paucity of data examining age differences in neural recruitment during posi-
tive and negative event retrieval makes it difficult to draw firm conclusions regarding
the role of prefrontal recruitment in older adults’ in emotion regulation. However,
the current data are in line with a regulatory account of prefrontal involvement. Ford
and colleagues found that individuals who were more likely to recruit ventrolateral
prefrontal regions during negative event retrieval were also those who exhibited a
greater positivity bias in hit rate, as measured by a greater difference between their hit
rate for positive and negative items (Ford and Kensinger, 2015b). Further, older adults
who recruited the ventrolateral prefrontal cortex more during negative event retrieval
provided lower overall ratings of arousal for those items; this same relation was not
seen in young adults. Parametric modulation analyses also revealed that older adults
were more likely to recruit dmPFC during retrieval of negative events that they later
rated as being less vivid and less arousing (Ford and Kensinger, 2015b). By contrast,
healthy aging was associated with an increased relation between ventromedial and
anterior prefrontal recruitment during positive event retrieval and higher ratings of
positive event vividness (Ford and Kensinger, 2015b). These findings suggest that,
within older adults, prefrontal regions may be preferentially recruited to enhance the
quality of positive event retrieval and to diminish the vividness and arousal of negative
memories.
Beyond Compensatory and Regulatory Accounts

Together, the studies reviewed above suggest that age-â•‰by-â•‰emotion interactions may
be partially explained by shifts in regulation motivation in older adults. However,
there is reason to believe that these interactions are not sustained throughout the
â•‡ 353
entire retrieval trial. Recent data from our laboratory suggest that age-â•‰by-â•‰valence
interactions in prefrontal recruitment are greater during the initial search phase
compared to a period of subsequent elaboration (Ford and Kensinger, 2015a). In
addition, a recent ERP study using positive, negative, and neutral images as retrieval
cues suggests that differences in positive and negative event retrieval in younger and
older adults may begin well before regulation strategies are implemented (Newsome
et al., 2012). Specifically, the authors present greater early old/â•‰new effects (150–
250 ms) for positive events in older adults and for negative events in younger adults,
likely reflecting age-â•‰specific valence biases in relatively automatic alerting and
attention processes. This finding suggests that multiple mechanisms may be con-
tributing to age-â•‰by-â•‰valence interactions during retrieval, with some of the age dif-
ferences arising from relatively early differences in the processing of emotional
retrieval cues.
The existence of age-â•‰related neural differences in cue processing could suggest that
all age effects at retrieval are driven by how young and older adults interact with emo-
tional cues. However, Newsome and colleagues (2012) only identified this effect in
the earliest analysis window (150–â•‰250 ms), suggesting that age differences in cue pro-
cessing do not extend into subsequent retrieval processes. In addition, recent evidence
suggests that the effects of age on cue processing and retrieval may be separable. One
ERP study utilized a facial emotional switch paradigm in which authors were able to
examine age effects on the influence of the encoded emotion (i.e., whether the initial
memory representation was negative or neutral) and of the retrieval cue (i.e., whether
the test face was negative or neutral) 380–â•‰700 ms following stimulus presentation
(Schefter et al., 2012). In these later analysis windows, age differences were identified
as a function of encoding emotion (specifically, greater emotion enhancement of old/â•‰
new effects for young relative to older adults) but not as a function or retrieval cue
emotion. Finally, two of the recent studies reviewed above (Ford and Kensinger, 2014;
Ford et al., 2015) identified significant age-â•‰by-â•‰valence effects in neural recruitment
when neutral titles were used as retrieval cues for memories with different encoded
valences. Together, these studies suggest that age-â•‰by-â•‰valence effects at retrieval rep-
resent differences in how young and older adults search and elaborate on emotional
memory representations, in addition to very early differences in how they may interact
with emotional retrieval cues.
Conclusion
Cognitive neuroscience research can not only elucidate the neural mechanisms sup-
porting cognitive processes, but also refine theories of how and why those processes
are implemented. As we hope this chapter has demonstrated, cognitive neuroscience
methods have been critical in the development and improvement of theories regarding
age differences in emotional memory. Neuroimaging studies have provided clear evi-
dence that, even when there are not age differences in behavioral outcomes, there are
often age differences in the mechanisms underlying the processing of emotional infor-
mation. Most notably, older adults engage with arousing information in a more con-
trolled or top-â•‰down fashion, recruiting more prefrontal regions than younger adults.
354
These arousal differences are present at encoding and retrieval, but their valence-â•‰
specificity may differ across these phases of memory.
While these insights have been gained, many open questions remain. Most notably:
• Although the prioritization of emotional goals appears to shift across the entire
adult lifespan (e.g., Carstensen, 1995), few studies of encoding (Williams et
al., 2006) or retrieval (Ford et al., 2015; Ford and Kensinger, 2014) have used
a continuous age design rather than a comparison of extreme age groups. Thus,
little is known about the lifespan trajectory of the neural effects described (see
Chapter 14 for more discussion of middle-â•‰aged adults).
• Distinguishing events by their valence and arousal has proved a powerful way to
investigate age differences in emotional memory. However, many studies have
assessed memory for high-â•‰arousal stimuli; this leaves many open questions about
age differences at lower levels of arousal, including why valence-â•‰specific age
differences may become exaggerated at these lower arousal levels. Moreover, no
neuroimaging studies have examined how age affects the neural processes asso-
ciated with memories that trigger particular emotions (e.g., fear, anger), moral
emotions (e.g., shame, contempt), or that are retrieved for regulatory purposes.
• It is well known that the way information is encoded will influence how it is
retrieved (and see Chapter 12); however, little research speaks to how age and
emotion affect the overlap between encoding and retrieval processes. As noted
earlier, age differences at encoding do not always parallel differences at retrieval,
suggesting that downstream consequences of encoding on retrieval do not nec-
essarily lead to reinstatement of the same age effects. Moreover, these differ-
ences may reflect the different processing required during the two phases, with
encoding reflecting the processing of an externally â•‰presented event and retrieval
including the re-â•‰experience of the internal representation of that event.
• The way information is retrieved will be influenced not only by how it is encoded
but also by how it is rehearsed and consolidated over time. Although age dif-
ferences in emotional memory can become exaggerated over time (Kennedy,
Mather, and Carstensen, 2004; Levine and Bluck 1997; Waring and Kensinger,
2009), with one exception (Kalpouzos et al., 2012), the neuroimaging studies of
emotional memory retrieval have assessed memory after a delay of 1 hr. or less,
leaving open the question of how age differences unfold over longer delays.
• Most research examining the effects of age on emotional memory has focused
on age differences in the processing and retention of emotional stimuli rather
than age differences in the effects of emotional states on the ability to remember
information. There is some behavioral evidence for age differences in mood
congruency effects (Knight et al., 2002), yet to our knowledge, there have not
been studies examining whether age and mood interact to influence the neural
mechanisms recruited at the time of encoding or retrieval.
References
Adolphs R, Tranel D, Buchanan TW (2005) Amygdala damage impairs emotional memory for
gist but not details of complex stimuli. Nat Neurosci 8:512–â•‰518.
355
Addis DR, Leclerc CM, Muscatell K, Kensinger EA (2010) There are age-related changes in
neural connectivity during the successful encoding of positive, but not negative, informa-
tion. Cortex 46:425–433.
Allard ES, Kensinger EA (2014). Age-related differences in neural recruitment during the use
of cognitive reappraisal and selective attention as emotion regulation strategies. Front
Psychol 5:296.
Bennion KA, Ford JH, Murray BD, Kensinger EA (2013). Oversimplification in the study of
emotional memory. J Int Neuropsychol Soc 19(9):953–961.
Binder J, De Quervain DJ, Friese M, Luechinger R, Boesiger P, Rasch B (2012) Emotion sup-
pression reduces hippocampal activity during successful memory encoding. Neuroimage
63(1):525–532.
Blackford JU1, Buckholtz JW, Avery SN, Zald DH (2010) A unique role for the human amyg-
dala in novelty detection. Neuroimage. 15;50(3):1188–1193.
Blanchard-Fields F (2007) Everyday problem solving and emotion: An adult developmental
perspective. Curr Dir Psychol Sci 16:26–31.
Blumenfeld RS, Ranganath C (2007) Prefrontal cortex and long-term memory encoding: an
integrative review of findings from neuropsychology and neuroimaging. Neuroscientist
13(3):280–291.
Bohn A, Berntsen D (2007) Pleasantness bias in flashbulb memories: positive and negative
flashbulb memories of the fall of the Berlin Wall among East and West Germans. Mem
Cognit 35(3):565–577.
Brown R, Kulik J (1977). Flashbulb memories. Cognition 5:73–99.
Buchanan TW, Denburg N, Tranel D, Adolphs R (2001) Verbal and nonverbal emotional mem-
ory following unilateral amygdala damage. Learn Mem 8:326–335.
Cabeza R, Nyberg L (2000) Imaging cognition: II. An empirical review of 275 PET and fMRI
studies. J Cogn Neurosci 12:1–47.
Carstensen LL (1995) Evidence for a life-span theory of socioemotional selectivity. Curr Dir
Psychol Sci 4:151–156.
Cassidy BS, Leshikar ED, Shih JY, Aizenman A, Gutchess AH (2013) Valence-based age differ-
ences in medial prefrontal activity during impression formation. Soc Neurosci 8(5):462–473.
Charles S, Mather M, Carstensen LL (2003). Aging and emotional memory: The forgettable
nature of negative images for older adults. JEP:Gen 132:310–324.
Chipchase SY, Chapman P (2013) Trade-offs in visual attention and the enhancement of mem-
ory specificity for positive and negative emotional stimuli. Q J Exp Psychol 66(2):277–298.
Clore GL, Gasper K, Garvin E (2001) Affect as information. In Forgas Jp (Ed.). Handbook of
Affect and Social Cognition (pp. 121–144). Mahwah, NJ: Lawrence Erlbaum Associates.
Conway MA (1995) Flashbulb memories. Hillsdale, NJ: Lawrence Erlbaum.
Davidson PS, Cook SP, Glisky EL (2006) Flashbulb memories for September 11th can
be preserved in older adults. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn
13(2):196–206.
Davis SW, Dennis NA, Daselaar SM, Fleck MS, Cabeza R (2008). Que PASA? The posterior-
Deary IJ, Whalley LJ, Batty GD, Starr JM (2006) Physical fitness and lifetime cognitive change.
Neurology: 67(7):1195–1200.
Denburg NL, Buchanan TW, Tranel D, Adolphs R (2003) Evidence for preserved emotional
memory in normal older persons. Emotion 3:239–253.
Dennis NA, Cabeza R (2008) Neuroimaging of healthy cognitive aging. In Craik Fim, Salthouse
Ta (Eds.), The handbook of aging and cognition (3rd ed.) (pp. 1–54). New York, NY, US:
Psychology Press.
356
Dolcos S, Katsumi Y, Dixon RA (2014) The role of arousal in the spontaneous regulation of
emotions in healthy aging: a fMRI investigation. Front Psychol 5:681.
Erk S, Walter H, Abler B (2008) Age-related physiological responses to emotion anticipation
and exposure. Neuroreport 19:447–452.
Fernandes M, Ross M, Wiegand M, Schryer E (2008) Are the memories of older adults posi-
tively biased? Psychol Aging 23:297–306.
Fischer H, Nyberg L, Bäckman L (2010) Age-related differences in brain regions supporting
successful encoding of emotional faces. Cortex 46:490–497.
Ford JH, Kensinger EA (2014) The relation between structural and functional connectivity
depends on age and on task goals. Front Hum Neurosci 8:307.
Ford JH, Kensinger EA (2015a) The posterior-to-anterior shift in aging is reversed during
episodic memory elaboration. Poster presented at the annual Rotman Research Institute
Conference, Toronto, CA.
Ford JH, Kensinger EA (2015b) Effects of subjective vividness on the neural correlates of
memory retrieval. Poster presented at the annual meeting of the Cognitive Neuroscience
Society, San Francisco, CA.
Ford Jh, Morris Ja, Kensinger Ea (2015). Neural recruitment during emotional memory retrieval
across the adult lifespan. Neurobiol Aging 35(12): 2770–2784.
Fredrickson BL (2004) The broaden-and-build theory of positive emotions. Philos Trans R Soc
Lond B Biol Sci 359(1449):1367–1378.
Gallo DA (2013). Retrieval Expectations Affect False Recollection: Insights from a Criterial
Recollection Task. Curr Dir Psychol Sci 22(4):316–323.
Gallo DA, Korthauer LE, Mcdonough IM, Teshale S, Johnson EL (2011) Age-related positivity
effects and autobiographical memory detail: evidence from a past/future source memory
task. Memory 19(6):641–652.
Grady CL, Mcintosh AR, Craik FI (2003) Age-related differences in the functional connectivity
of the hippocampus during memory encoding. Hippocampus 13:572–586.
Gunning-Dixon FM, Gur RC, Perkins AC, Schroder L, Turner T, Turetsky BI, et al. (2003) Age-
related differences in brain activation during emotional face processing. Neurobiology of
Aging 24:285–295.
Gutchess AH, Kensinger EA, Schacter DL (2007) Aging, Self-referencing and Medial Prefrontal
Cortex. Social Neuroscience 2(2):117–133.
Hahn S, Carlson C, Singer S, Gronlund SD (2006) Aging and visual search: Automatic and
controlled attentional bias to threat faces. Acta Psychologica 123:312–336.
Hamann S (2001) Cognitive and neural mechanisms of emotional memory. Trends in Cognitive
Sciences 5:394–400.
Hashtroudi S, Johnson MK, Vnek N, Ferguson SA (1994) Aging and the effects of affective and
factual focus on source monitoring and recall. Psychol Aging 9:160–170.
Hess Tm (2014) Selective engagement of cognitive resources: Motivational Influences on older
adults’ cognitive functioning. Perspectives on Psychological Science 9(4):388–407.
Holland AC, Kensinger EA (2012) Younger, middle-aged, and older adults’ memories for the
2008 U.S. Presidential Election. J Appl Res Mem Cogn 1(3):163–170.
Kalpouzos G, Fischer H, Rieckmann A, MacDonald SWS, Backman L (2012) Impact of nega-
tive emotion on the neural correlates of long-term recognition in younger and older adults.
Front Int Neurosci 6:74.
Kang C, Wang Z, Surina A, Lü W (2014) Immediate emotion-enhanced memory dependent
on arousal and valence: the role of automatic and controlled processing. Acta Psychol
150:153–160.
Kapucu A, Rotello CM, Ready RE, Seidl KN (2008) Response bias in “remembering” emo-
tional stimuli: A new perspective on age differences. J Exp Psychol: Learn Mem Cogn
34:703–711.
357
Kennedy Q, Mather M, Carstensen L (2004) The Role of Motivation in the Age-Related

Positivity Effect in Autobiographical Memory. Psych Sci 15:208–214.
Kensinger EA (2008) Age differences in memory for arousing and nonarousing emotional
words. J Gerontol B Psychol Sci Soc Sci 63(1):P13–P18.
Kensinger EA (2009) Remembering the Details: Effects of Emotion. Emot Rev 1(2):99–113.
Kensinger EA, Corkin S (2003) Memory enhancement for emotional words: are emotional
words more vividly remembered than neutral words? Mem Cognit 31(8):1169–1180.
Kensinger EA, Corkin S (2004) Two routes to emotional memory: Distinct neural processes for
valence and arousal. PNAS 101:3310–3315.
Kensinger EA, Leclerc CM (2009) Age-related changes in the neural mechanisms supporting
emotion processing and emotional memory. Eur J Cogn Psych 21:192–215.
Kensinger EA, Schacter DL (2008) Neural processes supporting young and older adults’ emo-
tional memories. J Cogn Neurosc 20:1161–1173.
Kensinger EA, Schacter DL (2006) Amygdala activity is associated with the successful encod-
ing of item, but not source, information for positive and negative stimuli. J Neurosci. 2006
Mar 1;26(9):2564–2570.
Kensinger EA, Brierley B, Medford N, Growdon JH, Corkin S (2002) Effects of normal aging
and Alzheimer’s disease on emotional memory. Emotion 2:118–134.
Kensinger EA, O’brien J, Swanberg K, Garoff-Eaton RJ, Schacter DL (2007) The effects of
emotional content on reality-monitoring performance in young and older adults. Psychol
Aging 22:752–764.
Kensinger EA, Gutchess AH, Schacter DL (2007). Effects of aging and encoding instructions
on emotion-induced memory trade-offs. Psychol Aging 22:781–795.
Kensinger EA, Garoff-Eaton RJ, Schacter DL (2007). Effects of emotion on memory specificity
in young and older adults. J Gerontol B Psychol Sci Soc Sci 62:208–215.
Kensinger EA, Addis DR, Atapattu RK (2011) Amygdala activity at encoding corresponds with
memory vividness and with memory for select episodic details. Neuropsychologia. 2011
Mar;49(4):663–673.
Kensinger EA, Allard E, Krendl AC (2014). The Effects of Age on Memory for Socioemotional
Material: An Affective Neuroscience Perspective. In: The Oxford Handbook of Emotion,
Social Cognition, and Problem Solving in Adulthood (Verhaeghen P, Hertzog C, Eds.)
New York: Oxford UP.
Kensinger EA, Piguet O, Krendl AC, Corkin S (2005) Memory for contextual details: effects of
emotion and aging. Psychol Aging 20(2):241–250.
Kern RP, Libkuman TM, Otani H, Holmes K (2005) Emotional stimuli, divided attention, and
memory. Emotion 5(4):408–417.
Kisley MA, Wood S, Burrows CL (2007) Looking at the sunny side of life: age-related
change in an event-related potential measure of the negativity bias. Psychol Sci
18(9):838–843.
Knight M, Seymour TL, Gaunt Jt, Baker C, Nesmith K, Mather M (2007) Aging and
goal-directed emotional attention: distraction reverses emotional biases. Emotion
7(4):705–714.
Knight BG, Maines ML, Robinson GS (2002) The effects of sad mood on memory in older
adults: a test of the mood congruence effect. Psychol Aging 17(4):653–661.
Krendl AC, Heatherton TF, Kensinger EA (2009) Aging minds and twisting attitudes: an fMRI
investigation of age differences in inhibiting prejudice. Psychol Aging 24(3):530–541.
Kryla-Lighthall N, Mather M (2009) The role of cognitive control in older adults’ emotional
well-being. In Berngtson, V., Gans, D., Putney, N., & Silverstein, M. (Eds.), Handbook of
Theories of Aging, 2nd Edition, Springer Publishing, 323–344.
Kunzmann U, Grühn D (2005) Age differences in emotional reactivity: the sample case of sad-
ness. Psychol Aging 20(1): 47–59.
358
Kvavilashvili L, Mirani J, Schlagman S, Erskine JA, Kornbrot DE (2010) Effects of age on phe-
nomenology and consistency of flashbulb memories of September 11 and a staged control
event. Psychol Aging 25(2):391–404.
Labar KS, Cabeza R (2006) Cognitive neuroscience of emotional memory. Nat Rev Neurosci
7:54–56.
Labouvie-Vief G, Lumley MA, Jain E, Heinze H (2003) Age and gender differences in car-
diac reactivity and subjective emotion responses to emotional autobiographical memories.
Emotion 3(2):115–126.
Langeslag SJE, Van Strien JW (2009). Aging and emotional memory: The co-occurrence of
neurophysiological and behavioral positivity effects. Emotion 9:369–377.
Leclerc CM, Kensinger EA (2008). Effects of age on detection of emotional information.
Leclerc CM, Kensinger EA (2008a) Age-related differences in medial prefrontal activation in
response to emotional images. Cogn Affect Behav Neurosci 8:153–164.
Leclerc CM, Kensinger EA (2010). Age-related valence-based reversal in recruitment of medial
prefrontal cortex on a visual search task. Soc Neurosci 5:560–576.
Leclerc CM, Kensinger EA (2011) Neural processing of emotional pictures and words: a com-
parison of young and older adults. Devel Neuropsychol 36:519–538.
Levenson RW, Carstensen LL, Friesen WV, Ekman P (1991) Emotion, physiology, and expres-
sion in old age. Psychol Aging 6(1):28–35.
Levine LJ, Bluck S (1997) Experienced and remembered emotional intensity in older adults.
Levine B, Svoboda E, Hay JF, Winocur G, Moscovitch M (2002) Aging and autobiographical
memory: dissociating episodic from semantic retrieval. Psychol Aging 17(4):677–689.
Liberzon I, Phan KL, Decker LR, Taylor SF. (2003) Extended amygdala and emotional salience:
a PET activation study of positive and negative affect. Neuropsychopharmacology
28(4):726–733.
Löckenhoff CE, Carstensen LL (2007) Aging, emotion, and health-related decision strategies:
Motivational manipulations can reduce age differences. Psychol Aging 22:134–146.
Mather M, Carstensen LL (2005) Aging and motivated cognition: The positivity effect in atten-
tion and memory. Trends Cogn Sci 9:496–502.
Mather M, Johnson MK (2000) Choice-supportive source monitoring: do our decisions seem
better to us as we age? Psychol Aging 15:596–606.
Mather M, Knight M (2005) Goal-directed memory: The role of cognitive control in older
adults’ emotional memory. Psychol Aging 20:554–570.
Mather M, Knight MR (2006) Angry faces get noticed quickly: Threat detection is not impaired
among older adults. J Gerontol B Psychol Sci Soc Sci 61:54–57.
Mather M, Sutherland MR (2011) Arousal-biased competition in perception and memory.
Perspect Psychol Sci 6(2):114–133.
Matsuda I, Nittono H (2014) Motivational significance and cognitive effort elicit different late
positive potentials. Clin Neurophysiol.
May CP, Rahhal T, Berry EM, Leighton EA (2005). Aging, source memory, and emotion.
Mickley KR, Kensinger EA (2009) Phenomenological characteristics of emotional memories in
younger and older adults. Memory 17(5):528–543.
Moriguchi Y, Negreira A, Weierich M, Dautoff R, Dickerson BC, Wright CI, Barrett LF (2011)
Differential hemodynamic response in affective circuitry with aging: an FMRI study of
novelty, valence, and arousal. J Cogn Neurosci 23(5):1027–1041.
Murray BD, Kensinger EA (2013) Age-related changes in associative memory for emotional
and nonemotional integrative representations. Psychol Aging 28(4):969–983.
359
Murty VP, Sambataro F, Das S, Tan HY, Callicott JH, Goldberg TE, Meyer-Lindenberg A,
Weinberger DR, Mattay VS (2009) Age-related alterations in simple declarative memory
and the effect of negative stimulus valence. J Cogn Neurosci 21(10):1920–1933
Murty Vp, Ritchey M, Adcock Ra, Labar Ks (2010) fMRI studies of successful emotional
memory encoding: A quantitative meta-analysis. Neuropsychologia 48(12):3459–3469.
Nashiro K, Mather M (2011) Effects of emotional arousal on memory binding in normal aging
and Alzheimer’s disease. Am J Psychol 124(3):301–312.
Nashiro K, Sakaki M, Mather M. (2012). Age differences in brain activity during emotion pro-
cessing: Reflections of age-related decline or increased emotion regulation? Gerontology
58:156–163.
Neisser U, Harsch N (1992) Phantom flashbulbs: False recollections of hearing the news
about Challenger. In Affect and accuracy in recall: Studies of ‘flashbulb’ memories (E.
Winograd, U. Neisser, Eds., pp. 9–31). New York, NY US: Cambridge University Press.
Newsome RN, Dulas MR, Duarte A (2012) The effects of aging on emotion-induced modula-
tions of source retrieval ERPs: Evidence for valence biases. Neuropsychol 50:3370–3384.
Ochsner KN, Gross JJ (2005) The cognitive control of emotion. Trends Cogn Sci 9:292–249.
Opitz PC, Rauch LC, Terry DP, Urry HL (2012) Prefrontal mediation of age differences in
cognitive reappraisal. Neurobiol Aging 33(4):645–655.
Petrican R, Moscovitch M, Schimmack U (2008) Cognitive resources, valence, and memory
retrieval of emotional events in older adults. Psychol Aging 23(3):585–594.
Piguet O, Connally E, Krendl AC, Huot JR, Corkin S (2008) False memory in aging: Effects of
emotional valence on word recognition accuracy. Psychol Aging 23:307–314.
Rahhal T, May CP, Hasher L (2002). Truth and character: Sources that older adults can remem-
ber. Psychol Sci 13:101–105.
Ranganath C (2010). A unified framework for the functional organization of the medial tem-
poral lobes and the phenomenology of episodic memory. Hippocampus 20:1263–1290.
Ray RD, Zald DH (2012) Anatomical insights into the interaction of emotion and cognition in
the prefrontal cortex. Neurosci Biobehav Rev 36(1):479–501.
Reed AE, Carstensen LL (2012) The theory behind the age-related positivity effect. Front
Psychol 3:339.
Richardson MP, Strange B, Dolan RJ (2004) Encoding of emotional memories depends on the
amygdala and hippocampus and their interactions. Nat Neurosci 7:278–285.
Ritchey M, Bessette-Symons B, Hayes SM, Cabeza R (2011) Emotion processing in the aging
brain is modulated by semantic elaboration. Neuropsychologia 49(4):640–650.
Roalf DR, Pruis TA, Stevens AA, Janowsky JS (2011) More is less: Emotion induced prefrontal
cortex activity habituates in aging. Neurobiol Aging 32:1634–1650.
Roy M, Shohamy D, Wager TD (2012) Ventromedial prefrontal-subcortical systems and the
generation of affective meaning. Trends Cogn Sci 16(3):147–156.
Rugg MD, Otten LJ, Henson RN (2002) The neural basis of episodic memory: evidence from
functional neuroimaging. Philos Trans R Soc Lond B Biol Sci 357(1424):1097–1110.
Russell JA (1979) Affective space is bipolar. J Pers Soc Psychol 37(3):345–356.
Sakaki M, Nga L, Mather M (2013) Amygdala functional connectivity with medial prefron-
tal cortex at rest predicts the positivity effect in older adults’ memory. J Cogn Neurosci
25(8):1206–1224.
Sander D, Grafman J, Zalla T (2003) The human amygdala: an evolved system for relevance
detection. Rev Neurosci. 14(4):303–316.
Scheibe S, Carstensen LL (2010) Emotional aging: recent findings and future trends. J Gerontol
B Psychol Sci Soc Sci 65B(2):135–144.
Schefter M, Knorr S, Kathmann N, Werheid K (2012) Age differences on ERP old/new effects
for emotional and neutral faces. Intl J Psychophysiol 85:257–269.
360
Scherer KR (2000) Psychological models of emotion. In The neuropsychology of emotion.

(Borod J., Ed.). New York: Oxford UP, pgs. 1337–1662.
Spaniol J, Davidson PSR, Kim ASN, Han H, Moscovitch M, Grady CL (2009) Event-related
fMRI studies of episodic encoding and retrieval: Meta-analysis using activation likelihood
estimation. Neuropsychol 47:1765–1779.
Spencer WD, Raz N (1995) Differential effects of aging on memory for content and context: A
meta-analysis. Psychol Aging 10:527–539.
St Jacques PL, Bessette-Symons B, Cabeza R (2009) Functional neuroimaging studies of aging
and emotion: Fronto-amygdalar difference during emotional perception and episodic
memory. J Int Neuropsychol Soc 15:819–825.
St Jacques PL, Dolcos F, Cabeza R (2009) Effects of aging on functional connectivity of the
amygdala for subsequent memory of negative pictures: a network analysis of functional
magnetic resonance imaging data. Psychol Sci 20(1):74–84.
Strange BA, Hurlemann R, Dolan RJ (2003) An emotion-induced retrograde amnesia in humans
is amygdala-and beta-adrenergic-dependent. PNAS 100(23):13626–13631.
Szpunar KK, Addis DR, Schacter DL (2012) Memory for emotional simulations: remembering
a rosy future. Psychol Sci 23(1):24–29.
Talmi D (2013) Enhanced Emotional Memory: Cognitive and Neural Mechanisms. Curr Dir in
Psychol Sci 22(6):430–436
Tessitore A, Hariri AR, Fera F, Smith WG, Das S, Weinberger DR, Mattay VS (2005) Functional
changes in the activity of brain regions underlying emotion processing in the elderly.
Psychiatry Res 139(1):9–18.
Tucker AM, Feuerstein R, Mende-Siedlecki P, Ochsner KN, Stern Y (2012) Double dissocia-
tion: circadian off-peak times increase emotional reactivity; aging impairs emotion regu-
lation via reappraisal. Emotion 12(5):869–874.
Waldinger RJ, Kensinger EA, Schultz MS (2011) Neural activity, neural connectivity, and the
processing of emotionally valenced information in older adults: Links with life satisfac-
tion. Cogn Aff Behav Neurosci 11:426–236.
Waring JD, Addis DR, Kensinger EA (2013) Effects of aging on neural connectivity underlying
selective memory for emotional scenes. Neurobiol Aging 34(2):451–467.
Waring JD, Kensinger EA (2009) Effects of emotional valence and arousal upon memory trade-
offs with aging. Psychol Aging 24(2):412–422.
Waters TE, Bohanek JG, Marin K, Fivush R (2013) Null’s the word: a comparison of memory
quality for intensely negative and positive events. Memory 21(6):633–645.
Williams LM, Brown KJ, Palmer D, Liddell BJ, Kemp AH, Olivieri G, Peduto A, Gordon E
(2006) The mellow years? Neural basis of improving emotional stability over age. Journal
of Neuroscience 26:6422–6430.
Winecoff A, Labar KS, Madden DJ, Cabeza R, Huettel SA (2011) Cognitive and neu-
ral contributors to emotion regulation in aging. Soc Cogn Affect Neurosci. 2011
Apr;6(2):165–176.
Wood S, Kisley MA (2006) The negativity bias is eliminated in older adults: age-related reduc-
tion in event-related brain potentials associated with evaluative categorization. Psychol
Aging 21(4):815–820.
Wurm LH, Labouvie-Vief G, Aycock J, Rebucal KA, Koch HE (2004) Performance in auditory
and visual emotional Stroop tasks: a comparison of older and younger adults. Psychol
Aging 19(3):523–535.
Yegiyan NS, Yonelinas AP (2011) Encoding details: positive emotion leads to memory broad-
ening. Cogn Emot 25(7):1255–1262.
361
III
HEALTH AND DISEASE

362
363
14
The Middle-Aged Brain
A Cognitive Neuroscience Perspective
Denise C. Park
Sara B. Festini
M iddle age is a unique component of the lifespan that has received rela-
tively little focused study in the vast literature on cognitive aging. As
more has been learned about the brain and behavior, it has become clear that some
aspects of neuropathology may have detectable neural signatures many years before
symptoms appear. As cognitive neuroscience advances, it seems likely that it will
become increasingly possible to predict one’s neurocognitive aging trajectory earlier
in life. Behavioral and pharmaceutical interventions may become available to treat
neurological disease or, more importantly, to prevent diseases such as Alzheimer’s
entirely. Middle-aged adults are increasingly likely to become the targets of such
interventions.
One obstacle to realizing the above goals is the limited information that is availa-
ble about what healthy neurocognitive function looks like in middle age. The present
chapter integrates existing data on brain structure, function, and cognition in mid-
dle-aged adults, using the STAC model (Scaffolding Theory of Aging and Cognition;
Park and Reuter-Lorenz, 2009) as a theoretical guide. The original STAC model
provided a theoretical framework that explained how age-related decline in cogni-
tion could be accounted for by brain structure and function. The model proposed that
varying levels of brain structure and brain function influence compensatory scaffold-
ing resources, all of which impact cognitive performance. Recently, Reuter-Lorenz
and Park (2014) updated STAC into the revised STAC-r model to include life-course
experiences, which enrich or deplete brain structure and function and affect the ability
to develop compensatory neural scaffolding. The revised model proposes that all of
363
364
364 Health and Disease
Scaffolding theory of aging and cognition-revised (STAC-r)
Biological
aging
Neural Brain Level of

enrich structure cognition
Life Compensatory
course scaffolding
Rate of
Neural Brain
cognitive
deplete function
change
Figure 14.1 A schematic diagram of the Scaffolding Theory of Aging and Cognition-Revised

(STAC-r). The model outlines general factors that influence adult cognition and cognitive
change. Various life course events (e.g., education, physical fitness, vascular health) promote
neural enrichment or neural depletion, which, along with biological aging, affect brain struc-
ture, brain function, and compensatory scaffolding. The confluence of these factors is theorized
to influence one’s current cognitive ability and one’s rate of cognitive change. Adapted from
Reuter-Lorenz and Park (2014), Neuropsychology Review.
these factors collectively affect current cognition as well as the rate of change in cog-
nition over time. See Figure 14.1.
As we learn more about individual trajectories of neurocognition across the lifespan
(e.g., Josefsson et al., 2012), it is clear that rates of change in both brain and behavior
vary considerably across individuals. Because the field of cognitive neuroscience is
young, longitudinal studies measuring concomitant changes of brain and behavior
across the lifespan have often focused on later stages of life, excluding middle-aged
adults, or the studies are so new that there are only a few years of brain/behavior data
that include middle age. In the present chapter, we integrate the data that do exist, both
cross-sectional and longitudinal, summarize what is known about the middle-aged
brain and cognition, and suggest future directions for research.
Before we address the neural and cognitive data, we must first define middle age.
Currently, life expectancy in the United States is approximately age 79 (CDC, 2015),
collapsed across gender. If we take the term “middle” literally, and split the 79 years
up into thirds, the middle years of the entire lifespan would be from age 27 to 53.
Childhood and adolescence, however, have a special status and set of associated devel-
opmental issues related to growth and learning, so it is perhaps more appropriate to
â•‡ 365
The Middle-AgedÂ€Brainâ•… 365
think about middle age in the context of the adult lifespan. If we consider that there
appears to be continuing development of the brain, particularly neocortical regions,
until about age 25 (Giedd, 2004), and conceptualize middle age to be the middle years
of the adult lifespan, then middle age would fall from age 43 to 60, with ages 61+ repre-
senting old age, and ages 26–â•‰42 representing young adulthood. Taking life expectancy,
neural maturation, and current research practices into account, in the present chapter
we refer to adult stages of the lifespan as follows: Young Adulthood: Age 18–39 (taking
into account that many 18–â•‰21-â•‰year-â•‰old college students are treated as young adults,
even though brain maturity may not have been fully reached); Middle Age: Age 40–â•‰59;
Older Adults: Age 60–â•‰80; and Very Old (exceeding life expectancy): Age 81+.
The Psychosocial Context of Middle Age
Adults of all ages perceive middle age to be a time of high responsibility and demands
(Lachman et al., 1994). Most adults reach their peak work success during midlife,
so middle age is often characterized by high work demands (Helson et al., 2006;
Lachman et al., 2015). Additionally, many middle-â•‰aged adults are managing the
responsibilities associated with rearing children (Lachman, 2004), and it is also not
uncommon to be a grandparent toward the end of midlife (Census, 2014). Coupled
with these commitments, the parents of middle-â•‰aged adults are beginning to approach
very old age, and there is frequently a need to provide assistance to these aging parents
(Lachman et al., 2015). Thus, middle age is characterized by a significant amount of
contextual pressure from many different directions, which results in high allostatic
load (i.e., physiological repercussions of frequent stress).
This contextual pressure likely plays a role in the relatively sparse data associated
with the middle-â•‰aged segment of the lifespan due to a smaller pool of middle-â•‰aged
volunteers available for research studies. Middle-â•‰aged adults who do volunteer may
be different in their characteristics from their younger and older counterparts. This is
particularly problematic for cross-â•‰sectional designs, as middle-â•‰aged birth cohorts are
likely to differ from older and younger participants both in terms of individual differ-
ences, such as health, education, and socioeconomic status (SES), as well as in terms of
experiences associated with birth years. There is no easy way around this issue, but one
possibility is to enroll young, middle-â•‰aged, and old samples that are matched in terms of
education and SES, and then to follow each group over a time interval (i.e., ten years).
Such a design has the advantage of providing longitudinal change data in neural and
cognitive function in well-â•‰matched age groups in a span of time that would both allow
for significant change in cognition, but also be readily manageable within the career
span of an individual scientist. This design does not, of course, correct for cohort effects,
but it does realistically address many problems with a traditional cross-â•‰sectional design.
Goals of Chapter
The primary goal of this chapter is to provide a roadmap for the study of the cognitive
neuroscience of middle age. We initially review the literature on cognition in midlife,
366
366â•… Health and Disease
followed by structural and functional findings on the middle-â•‰aged brain. Using the
STAC-â•‰r model as a guide, we then explore neural enrichment and depletion factors that
may alter the course of neurocognitive aging. We believe a careful understanding and
study of the middle-â•‰aged brain will allow us to chart the cognitive aging trajectory of an
individual into old age, revealing markers of both future resilience and pathology. The
maintenance of a healthy mind for life and a cure for Alzheimer’s Disease may very well
depend on our ability to accurately piece together the puzzle of the middle-â•‰aged brain.
Cognition, Brain Structure, and Brain Function in Midlife
What Differences are Present in Core Cognitive Abilities in Midlife?
Middle age is generally characterized by modest declines in fluid processing, with

growth in crystallized knowledge and experience (e.g., see Lachman et al., 2015).
Middle-â•‰aged adults occupy the transition period between young adulthood and old
age, and may arguably be at the height of their cognitive ability. Although they have
experienced some declines in fluid abilities, these declines are relatively modest, and
middle-â•‰aged adults have experienced significant growth in their knowledge base,
which likely provides significant support to everyday cognitive performance. Indeed,
evidence from the workplace suggests that individuals tend to reach their career peak
in their 50s (Helson and Soto, 2005), potentially because they have optimal levels of
both cognitive ability and world knowledge and experience.
In the following section, we provide an overview of what is known about mid-
life cognition (see Table 14.1 for a summary), noting when differences are observed
between longitudinal and cross-â•‰ sectional studies, each of which have different
strengths and weaknesses. Longitudinal studies likely underestimate decline because
of re-â•‰test and dropout effects (e.g., Salthouse, 2009), and cross-â•‰sectional studies may
overestimate age-â•‰related differences because of cohort effects and the inability to
measure within-â•‰person change. This, in combination with a considerable range of
individual differences (e.g., Willis and Schaie, 2005; Gunstad et al., 2006; Salthouse,
2014), makes the picture of midlife cognition complex. Moreover, a recent study by
Hartshorne and Germine (2015) highlights that different cognitive abilities peak at
different ages, with evidence for certain peaks in midlife. Thus, changes in cognition
are not uniform across the lifespan.
Processing Speed
Much longitudinal and cross-â•‰sectional evidence indicates that processing speed
decreases from young adulthood to middle adulthood to old age (Willis and Schaie,
1999; Park et al., 2002; Salthouse, 2009; but see Martin and Zimprich, 2005). Data
from the Midlife in the United States (MIDUS) study revealed smaller differences
in processing speed between middle-â•‰aged and younger adults, whereas accelerated
declines after age 65 were observed (Soederberg, Miller, and Lachman, 2000). Thus,
while processing speed tends to decline with age, midlife declines are not as extensive
as those in later life.
367
Table 14.1 Overview of midlife cognition as revealed by cross-sectional

and longitudinal studies.
Cognitive Function Study Type Conclusion References
Processing Speed Cross-Sectional Decline (Park et al., 2002;

Salthouse, 2009)
Small decline (Soederberg Miller and
Lachman, 2000)
Longitudinal Decline (Willis and Schaie, 1999)
Small increase (Martin and Zimprich, 2005)
Executive Cross-Sectional Decline (Garden et al., 2001;

Functioning MacPherson et al., 2002)
Longitudinal 20% show decline (Willis and Schaie, 2005)
Working Memory Cross-Sectional Decline (MacPherson et al., 2002; Park

et al., 2002)
Little or no decline (Soederberg Miller and
Lachman, 2000; Park et al.,
2013)
Longitudinal Decline (Hultsch et al., 1992)
Decline, if vascular risk (Raz et al., 2007)
Episodic Memory Cross-Sectional Decline (Park et al., 2002; Mitchell and

Bruss, 2003)
Longitudinal Peaks at midlife (Willis and Schaie, 1999;
Martin and Zimprich, 2005)
Little or no decline (Zelinski and Burnight, 1997;
Ronnlund et al., 2005)
31% show decline (Willis and Schaie, 2005)
Implicit Memory Cross-Sectional Stability (Hamberger and Friedman,

1992; Mitchell and Bruss,
2003)
Longitudinal Stability (Hultsch et al., 1992)
Spatial Processing Cross-Sectional Decline (Salthouse, 2009, 2013)

Longitudinal Peaks at midlife (Zelinski and Burnight, 1997;
Willis and Schaie, 1999)
Reasoning Cross-Sectional Stability (Soederberg Miller and

Lachman, 2000)
Decline (Denney and Heidrich, 1990;
Salthouse, 2009)
Longitudinal Peaks at midlife (Zelinski and Burnight, 1997;
Willis and Schaie, 1999)
Decline (Raz et al., 2007)
Social & Emotional Cross-Sectional Stability (MacPherson et al., 2002)

Processing Increase (Grossmann et al., 2010)
Decline (Heidrich and Denney, 1994)
Longitudinal — —
Crystallized Cross-Sectional Increase (Soederberg Miller and

Knowledge Lachman, 2000; Park et al.,
2002; Park et al., 2013;
Hartshorne and Germine, 2015)
Longitudinal Stability or Increase (Finkel et al., 2003; Ronnlund
et al., 2005)
Everyday Functioning Cross-Sectional Stability (Baltes et al., 1995; Garden

et al., 2001)
Decline (Denney and Pearce, 1989)
Longitudinal — —
Note. This table is not intended to be a comprehensive list.

368
Executive Functioning
Deficits in executive functioning have also been observed in midlife. Compared to
their younger counterparts, middle-â•‰aged adults tend to have impaired executive per-
formance on novel experimental tasks (e.g., Garden et al., 2001; MacPherson et al.,
2002). The limited longitudinal data available suggests that only a subset of individuals
exhibits observable midlife executive declines. Specifically, the Seattle Longitudinal
Study measured verbal fluency at ages 46, 53, and 60, and only 20% of participants
exhibited declines over this interval; 69% were stable; and 11% improved (Willis
and Schaie, 2005). Note, however, that repeated testing may mask some decline in
longitudinal designs. Thus, while cross-â•‰sectional evidence suggests that middle-â•‰aged
adults have more difficulty rapidly and efficiently performing higher-â•‰level mental
manipulations, such as inhibiting, updating, and shifting, longitudinal data suggest
differences to be more modest.
Working Memory
Cross-â•‰sectional data yield mixed findings regarding working memory performance
at midlife. Park et al. (2002) documented reductions in Line Span, Letter Rotation,
Reading Span, and Computation Span from young to middle age (see also MacPherson
et al., 2002). Nevertheless, in a later study, when compared to younger adults,
middle-â•‰aged adults had similar performance on Operation Span and Letter-â•‰Number
sequencing (Park et al., 2013; see also Soederberg Miller and Lachman, 2000). The
discrepancy in these findings may be due to differences in task difficulty, such that
simple span tasks require fewer processing demands than complex working memory
tasks, or due to differences in the sample being tested. Longitudinal assessments of
working memory in midlife are less frequent. Hultsch et al. (1992) report 3-â•‰year lon-
gitudinal decline in complex span tasks in a sample that included middle-â•‰aged adults.
Another longitudinal study documented midlife declines in listening span, although,
interestingly, this decline was only observed in individuals with poor vascular health
(Raz et al., 2007).
Episodic Memory
Both deficits and improvements in episodic memory have been documented in
middle-â•‰aged adults relative to young adults. Longitudinally, Willis and Schaie (1999)
noted peak levels of verbal memory at midlife, and Zelinski and Burnight (1997)
report no longitudinal change in episodic list recall (see also Martin and Zimprich,
2005; Ronnlund et al., 2005). In contrast, cross-â•‰sectional assessments have consist-
ently revealed declines in episodic memory with age, even in midlife (Park et al.,
2002; Mitchell and Bruss, 2003). Both marked individual differences in maintenance
of memory ability and pronounced retest effects on memory tasks may be responsible
for the discrepancy between cross-â•‰sectional and longitudinal findings. First, when
looking for inter-â•‰individual variability, Willis and Schaie (2005) found that 31% of
their participants declined, 53% were stable, and 16% improved on tests of delayed
â•‡ 369
recall from age 46 to 53 to 60. Second, retest effects are commonly observed on
memory tasks (see Salthouse, 2009). Nevertheless, even when adjusting for practice
effects, some studies find no longitudinal declines in midlife episodic memory (e.g.,
Ronnlund et al., 2005).
Implicit Memory
Implicit memory shows stability in midlife. In cross-â•‰sectional implicit tests of word-â•‰
fragment completion, word-â•‰stem completion, and picture-â•‰fragment identification, there
were no age differences between younger, middle-â•‰aged, or older adults (Mitchell and
Bruss, 2003). Moreover, both behavioral and electrophysiological correlates of repetition
priming have revealed similar patterns in young, middle, and older adults (Hamberger
and Friedman, 1992). Longitudinal evidence appears consistent with these findings, as
Hultsch et al. (1992) found no decline in word-â•‰stem completion after 3 years.
Spatial Processing
Peak levels of spatial processing on mental rotation tasks were observed in mid-
life in the Seattle Longitudinal Study (Willis and Schaie, 1999). In accord with
this finding, 30–â•‰36-â•‰year-â•‰olds had improved spatial ability on Figure and Object
Rotation 16 years later (Zelinski and Burnight, 1997). Cross-â•‰sectionally, however,
middle-â•‰aged adults have exhibited inferior spatial processing than younger adults
(Salthouse, 2009, 2013).
Reasoning
There are mixed results concerning reasoning performance in midlife. Cross-â•‰sectional
data has provided evidence for both stability (Soederberg Miller and Lachman,
2000) and decline (Denney and Heidrich, 1990; Salthouse, 2009). Longitudinally,
Willis and Schaie (1999) found peak inductive reasoning performance at midlife,
and Zelinski and Burnight (1997) found stability in reasoning in people aged 30–â•‰36
after 16 years. Nevertheless, fluid reasoning has also exhibited longitudinal midlife
declines (Raz et al., 2007).
Social and Emotional Processing

Some cross-â•‰sectional evidence suggests that the ability to detect emotions and to
make moral decisions is preserved in midlife (MacPherson et al., 2002), and that
social reasoning improves with age (Grossmann et al., 2010). However, Heidrich and
Denney (1994) report that social problem solving ability increased from 20-â•‰year-â•‰olds
to 40-â•‰year-â•‰olds, but declined thereafter. These differences may be driven by the extent
to which the social and emotional tasks relied on fluid skills. Additional work should
investigate the source of these discrepancies, as well as examine longitudinal midlife
trajectories of social and emotional processing.
370
Crystallized Knowledge
Both longitudinal and cross-â•‰sectional findings indicate that crystallized intelligence
tends to remain stable or increase in midlife (Finkel et al., 2003; Ronnlund et al.,
2005; Park et al., 2013; Hartshorne and Germine, 2015), with middle-â•‰aged adults
having high levels of semantic memory and general world knowledge. For instance,
middle-â•‰aged adults have been repeatedly shown to have superior vocabulary than
young adults (Soederberg Miller and Lachman, 2000; Park et al., 2002). Continuing
this trajectory, however, middle-â•‰aged adults sometimes exhibit inferior vocabulary
than older adults (Park et al., 2002). Nevertheless, middle-â•‰aged adults maintain higher
levels of fluid processing than older adults, which may outweigh this difference in
crystallized knowledge.
Everyday Functioning
Middle-â•‰
aged adults appear to have few deficits on real-â•‰ world activities. For
instance, Baltes et al. (1995) documented good performance on existential life
planning and life management tasks up to age 80. Moreover, Garden et al. (2001)
found that middle-â•‰aged adults successfully scheduled and planned real-â•‰life tasks
and errands better than younger adults, breaking fewer task rules. This superior
everyday functioning is also reflected in their peak levels of work performance
(Helson and Soto, 2005). Nevertheless, tasks that rely more strongly on fluid pro-
cessing can show deficits, as Denney and Pearce (1989) found that everyday prob-
lem solving declined after age 40. Overall, when they are able to draw upon their
knowledge and experiences, middle-â•‰aged adults tend to perform everyday tasks
well. We note, however, that additional longitudinal assessments of everyday func-
tioning in midlife are necessary.
Does Middle-â•‰Aged Cognition Predict Future Cognitive Function?

Several studies have begun to address the impact of midlife cognition on future cog-
nitive performance. Midlife performance on delayed recall and word fluency have
been shown to predict cognitive impairment at old age (Willis and Schaie, 2005).
Moreover, early deficits in midlife cognition may be an indicator of a preclinical
progression to dementia, as LaRue and Jarvik (1987) found that midlife cognitive
deficits predicted dementia 20 years later. Beginning even earlier in the lifespan,
data from the Lothian Birth Cohort indicated that intelligence at age 11 was asso-
ciated with better cognition at age 79, although childhood mental ability was not
related to the rate of late-life cognitive decline from 79–â•‰83 (Gow et al., 2008).
Additionally, within-â•‰individual cognitive variability at initial testing may serve as
a useful predictor of future cognition. Older adults with the largest 3–â•‰4 year longi-
tudinal decline in cognition exhibited the most variability in their cognitive perfor-
mance at initial testing (Salthouse and Soubelet, 2014), although the authors note
that within-â•‰individual variability may be a predictor of pathological change rather
than normal aging.
â•‡ 371
What Structural Brain Differences are Present in Midlife?
On the whole, middle-â•‰aged adults exhibit declines in structural brain integrity, exhib-
iting reductions in gray matter and white matter and slight increases in tau tangles and
amyloid plaques. However, much of the extant work on structural brain changes in
midlife comes from cross-â•‰sectional studies. The field would benefit from additional
longitudinal studies to document trajectories of brain change within-â•‰individuals,
including middle-â•‰aged adults.
Gray Matter Declines

Numerous studies find that gray matter volume declines with age (e.g., Ge et al.,
2002; Fjell et al., 2009), and that cortical thinning is observable by middle age (Salat
et al., 2004). The biggest reductions in cortical gray matter have been observed in
the prefrontal cortex (Salat et al., 2004; Raz et al., 2007; Fjell et al., 2009) and the
hippocampus (Raz et al., 2005; Raz et al., 2007), with the hippocampus showing an
increasing rate of decline with age (e.g., Raz et al., 2004; Raz et al., 2010). Shrinkage
in the association cortices, cerebellum, caudate, and inferior temporal cortex also has
been observed (Raz et al., 2005). Nevertheless, careful examination of individual
trajectories of change has revealed considerable individual differences in shrinkage
rates, especially in the lateral prefrontal cortex and cerebellum (Raz et al., 2010).
Thus, on average, declines are consistently observed, but certain individuals may not
exhibit this pattern of gray matter reduction in midlife. These individual differences
may be driven by variations in the enrichment and depletion factors that people have
experienced, which we discuss later.
White Matter Declines

Three methodologies have been used to assess white matter integrity in middle-â•‰aged
adults. First, volumetric decreases in white matter have been observed in midlife,
with accelerated, nonâ•‰linear declines starting in the 40s (Ge et al., 2002). Paralleling
the gray matter findings, prefrontal white matter tends to show the largest change
relative to other regions (Raz et al., 2005). Second, white matter hyperintensities
(WMH), which reflect damage to white matter and appear as bright spots on MRI
scans, have been observed to increase from young adulthood through middle age to
old age (DeCarli et al., 1995). Third, Diffusion Tensor Imaging (DTI) evidence sim-
ilarly documents midlife change, such that white matter water diffusivity increases
with age (Pfefferbaum and Sullivan, 2003; Salat et al., 2005; Sullivan et al., 2010).
Thus, converging evidence indicates some white matter degradation even in midlife.
β-â•‰Amyloid Plaques and Neurofibrillary Tangles

Widespread accumulation of beta-â•‰amyloid in the brain is a hallmark of Alzheimer’s
Disease (AD) (Hardy and Selkoe, 2002), although research suggests that extensive
372
amyloid deposition is necessary but not sufficient to develop AD (e.g., Sperling et al.,
2011). Autopsy data has shown that a subset of cognitively healthy individuals also
exhibit amyloid plaques, and postmortem amyloid deposition even has been observed
in some healthy middle-â•‰aged adults (Braak and Braak, 1991, 1997). More recently,
the ability to image amyloid deposition in vivo has become available (e.g., Klunk et
al., 2004), and, at present, the limited data on middle-â•‰aged adults suggests that amy-
loid deposition increases significantly from ages 30–â•‰89 in cognitively normal people
(Rodrigue et al., 2012). Among older adults, greater levels of β-â•‰amyloid have been
associated with worse episodic memory (Hedden et al., 2013), altered neural activation
during memory encoding (Sperling et al., 2009; Kennedy et al., 2012), and progression
to dementia (Villemagne et al., 2011). Moreover, it is thought that amyloid accumu-
lates over decades, implying that initial amyloid deposition may begin in midlife (Jack
et al., 2013), as is consistent with autopsy data. Additional research is needed to deter-
mine if and when midlife amyloid is predictive of cognitive decline or dementia.
Another defining feature of AD is the accumulation of tau neurofibrillary tan-
gles throughout the neocortex (Braak and Braak, 1991, 1997). At autopsy, many tau
tangles are observed in those with Mild Cognitive Impairment (MCI) and AD (e.g.,
Guillozet et al., 2003), and the extent of tau deposition is highly correlated with cogni-
tive decline and disease severity (Mitchell et al., 2002). Moreover, autopsy data indi-
cates that the earliest focal deposits of neurofibrillary tangles can be detected as early
as age 25, with about 50% of individuals exhibiting these focal neurofibrillary tangles
by age 55 (Braak et al., 2011). It is theorized that amyloid induces the spread of neu-
rofibrillary tangles beyond these focal deposits (Bolmont et al., 2007). However, little
is known about the effects of lower levels of tau on cognition throughout the lifespan.
The recent development of new ligands that permit in vivo tau PET imaging will
allow scientists to more critically examine the relationship between midlife tau levels
and current and future cognitive health. The new tau PET imaging will enable scien-
tists to examine the interrelationships between amyloid, tau, and dementia, including
whether these biomarkers are informative in midlife.
Relative Preservation in Specific Brain Structures

Despite the widespread documentation of age-â•‰related structural brain degeneration,
not all neural structures are equally affected. Whereas the hippocampus shows large
declines even at midlife (Raz et al., 2004; Raz et al., 2005), primary sensory cortices
like the primary visual cortex tend to display considerable age-â•‰related stability (Raz
et al., 2005; but see Salat et al., 2004). Variations in the rate at which brain structures
deteriorate have been proposed to occur for a myriad of reasons, including that brain
deterioration may mirror the developmental order of brain maturation (i.e., first in,
last out; Ribot’s law; Ribot, 1891).
Patterns of Neural Activity During Middle Age
One of the most dramatic findings from the functional imaging literature on aging is
that, when presented with verbal materials, young adults tend to show left-â•‰lateralized
373
The Middle-Aged Brain 373
activation, whereas older adults tend to show bilateral patterns of neural activity in
dorsolateral prefrontal cortex (e.g., see Cabeza, 2002; Reuter-Lorenz and Cappell,
2008). This pattern is pervasive across many tasks and is also manifested in parietal
cortices (Huang et al., 2012). The STAC model suggests that the recruitment of the
additional hemisphere by older adults is a compensatory mechanism to maintain cog-
nitive function in the face of structural degradation.
An interesting and important question about which little is known is at what stage
in the lifespan the shift from left-lateralized activity to bilateral activity occurs during
verbal encoding and working memory tasks. The STAC model suggests that the age
at which individuals manifest a bilateral pattern may be a predictor of neural frailty,
with those manifesting earlier bilateral recruitment potentially exhibiting faster rates
of future cognitive decline. Results from the Betula Aging Study are in line with this
hypothesis. Pudas et al. (2014) used both midlife memory performance and change
in memory performance from middle to old age to predict activation at old age in
the bilateral hippocampus and left inferior frontal gyrus during a verbal encoding
task. The results showed that memory performance in middle age predicted twice as
much variance in the functional data as the memory change data predicted. The results
provide compelling evidence that middle-aged cognition and brain function may be
reliable harbingers of future brain health. The Betula dataset, due to its longitudinal
nature and relatively early use of neuroimaging, is likely to be a rich resource in
addressing these issues. Presently, no longitudinal studies categorize individual differ-
ences in middle-aged recruitment patterns and relate them to future cognitive aging.
Limited cross-sectional data on this topic do exist, however, and suggest that the age
at which neural activity differences are observed varies as a function of both the task
type and the engaged brain networks. We used the first wave of data from the Dallas
Lifespan Brain Study to examine cross-sectional activation patterns on four different
functional imaging tasks. These included (a) a passive face-viewing and house-viewing
task that loaded on ventral visual cortex, particularly the fusiform gyrus (Park et al.,
2012), (b) a picture encoding task that yielded robust activation in mediotemporal and
frontal-parietal regions (Park et al., 2013), (c) a semantic word judgment task that had
strong frontal-parietal loading (Kennedy et al., 2015), and (d) a resting state task where
formal network analyses using graph theory were conducted (Chan et al., 2014). All
four studies included participants from ages 20–89, with the number of subjects in each
study ranging from 192 to 318. Of particular importance, the same subjects performed
all four tasks in a single imaging session. Hence, any observed differences in middle
age across studies are unlikely to be due to sampling differences.
The ventral visual study was designed to detect age differences in neural selectivity
in highly specialized brain regions, particularly the core fusiform face area (FFA),
as well as additional regions associated with the extended face network (Park et al.,
2012). The study’s main finding was that age-related dedifferentiation in the core face
network was driven by increased responses to houses with age, supporting the broad-
ening hypothesis of dedifferentiation; in contrast, age-related dedifferentiation in the
extended face network was driven by decreased activation to faces, suggesting that
attenuation of domain-specific activation was the mechanism underlying the effect.
These results suggest that age-related dedifferentiation may not be a unitary phenom-
enon. Rather, age-related losses of neural selectivity reflect different processes in dif-
ferent regions. With respect to midlife performance, all reported lifespan effects were
374
(A) ** (B) * **
0.0
1.0
Parameter Estimates
0.8
(Hc–hits–Miss)
0.6 –0.4
0.4
0.2 –0.8
0.0
Young Middle Older Young Middle Older
Figure 14.2 Average parameter estimates for activation and deactivation in the positive sub-
sequent memory contrast (A) and the negative subsequent memory contrast (B), as a function
of age group. Large differences in positive subsequent memory effects were observed from
young to middle age, whereas large differences in negative subsequent memory effects were
most apparent from middle to older age. Reproduced with permission from Park et al. (2013),
Neuropsychologia.
linear in nature, suggestive of continuous decline in neural selectivity from young

adulthood to middle age to old age.
The findings from this ventral visual study contrast with results from a subsequent
memory task, where 192 participants encoded complex outdoor scenes and judged
whether or not the scene included water (Park et al., 2013). Recognition of targets
and lures was performed outside the scanner. Results showed that positive subsequent
memory effects (contrast of subsequently remembered and forgotten items) occurred
largely in object-processing and higher order visual regions (including fusiform,
precuneus, inferior temporal, and superior parietal sites), and importantly, that age
differences were most pronounced in these regions between young and middle-aged
adults (see Figure 14.2A). In contrast, negative subsequent memory effects (revealed
by the contrast between subsequently forgotten and remembered items), occurred
primarily in default network sites and exhibited the largest differences from middle
to old age (see Figure 14.2B). The important finding from this study, with implica-
tions for understanding the middle-aged brain, is as follows: Brain regions are not
equally influenced by age, as lifespan discontinuities differed here for task-positive
and default regions.
In a third task, Kennedy et al. (2015) examined semantic encoding, where 316
participants judged whether words were living or non-living. Some items were of
a concrete nature (e.g., bear, truck, desk) and easy to judge, while other items were
ambiguous and thus more difficult (e.g., ghost, virus, bat). The difference in neural
activity between the easy versus ambiguous items reflected the modulatory capacity
of the brain to respond to increased challenge. Of particular interest, was the finding
that frontal and subcortical regions exhibited the strongest age differences between
middle-aged (40–59) and older (60–79) adults. However, dopaminergic pathways
in midbrain/brainstem regions displayed the largest age differences between older
(60–79) and very old participants (80–89). Thus, specific regions associated with
â•‡ 375
(A) Sensory-motor system segregation vs. age (B) Association system segregation vs. age
0.90 0.80
Mean sensory-motor system
Mean association system

segregation
segregation
0.70 0.60
0.50 0.40
0.30 0.20
20 30 40 50 60 70 80 90 20 30 40 50 60 70 80 90
FigureÂ€14.3â•‡ Neural network segregation as a function of age in sensory-â•‰motor systems (A) and

association systems (B). Sensory-â•‰motor systems showed linear decreases in segregation with
age. Association systems showed accelerated segregation decline after age 50. Reproduced with
permission from Chan et al. (2014), Proceedings of the National Academy of Sciences.
controlled-â•‰processing declined from middle to old age, whereas midbrain/â•‰brainstem

regions showed a larger decline from old to very old age.
Finally, a fourth study that included 210 participants, focused on resting state con-
nectivity (Chan et al., 2014). Network analyses guided by graph theory were used
to describe age differences in 10 major intrinsic brain systems. Results showed that
younger adults’ brain systems were highly segregated, with strong connectivity
between nodes within a given system. In contrast, increasing age was associated with
less segregation of systems. Of particular interest is evidence shown in Figure 14.3,
where segregation linearly decreased with age for sensory and motor systems (A), but
showed accelerated decline beginning at age 50 for systems involved in associative
and controlled processing (B).
To summarize the findings from these four studies, we note that both Park et al.
(2012) and Chan et al. (2014) show evidence for linear decline in brain regions and
systems specialized for visual and sensory processing. In contrast, tasks that involved
controlled associative processing exhibited nonlinearity (Park et al., 2013; Kennedy
et al., 2015). Together, these findings suggest that, with respect to the brain, age is
far from a universal construct. Different brain regions and systems manifest critical
discontinuities at different points in the lifespan.
Enrichment and Depletion Factors at Midlife
The STAC-â•‰r model (Reuter-â•‰Lorenz and Park, 2014), portrayed in Figure 14.1, sug-
gests that both enrichment and depletion factors influence brain structure, brain
function, and cognition. Enrichment and depletion factors are individual differences
variables that can either enhance brain structure and function (enrichment factors) or
have negative consequences (depletion factors). Enrichment and depletion variables
can be introduced at any point in the life course and can often be represented as “two
sides of the same coin.” For example, good vascular health would be viewed as an
376
enrichment factor whereas poor vascular health would be depleting. Hence, we have
chosen to focus our discussion of these important individual differences variables that
modify brain health by distinguishing between biologically â•‰based factors and contex-
tually based lifestyle factors.
Biological Factors that Affect Brain Health
There are many biologically â•‰based variables that can affect brain structure and func-
tion. Below we summarize key findings, focusing in particular on biological factors in
midlife that influence current and subsequent cognitive and neural functioning (see also
Bendlin et al., 2010 for a review of midlife predictors of AD). It is important to note,
however, that some factors are collinear with one another (i.e., individuals with poor
vascular health may be at higher risk for diabetes, obesity, and hypercholesterolemia).
One important challenge is to understand the difference in the predictive value between
clusters of interâ•‰related variables compared to one core biomarker. Much of the extant
work comes from cross-â•‰sectional data; we note longitudinal studies when applicable.
Genes
The apolipoprotein E type 4 (APOE ε4) allele has been associated with poorer cog-
nition and higher risk for development of AD (e.g., Coon et al., 2007). Presence of
the APOE ε4 allele was related to inferior spatial working memory and visuospatial
attention at middle age (Greenwood et al., 2005; but see Greenwood et al., 2014), and
it predicted midlife longitudinal cognitive decline (Blair et al., 2005; Caselli et al.,
2009). Structural reductions in white matter volume were also observed in carriers of
APOE ε4 who were also homozygous TT for the cholinergic receptor gene CHRNA
(Espeseth et al., 2006). Moreover, middle-â•‰aged adults with these genetic characteris-
tics, on average, had differentially slower response times for invalid cues on a Posner
attention task compared to older adults, suggesting that these genetic variations had
the biggest impact on attention during midlife. Finally, functional neuroimaging
measures have revealed altered neural activation in APOE ε4 carriers relative to non-â•‰
carriers, with evidence for both increases and decreases in activation. Filippini et al.
(2009) found that young APOE ε4 carriers had greater brain activity during memory
encoding and abnormally increased default mode activity at rest than non-â•‰carriers,
with implications for altered neural activity continuing to midlife. Moreover, in a
sample of middle-â•‰aged adults, APOE ε4 carriers exhibited reduced neural activation
during a semantic categorization task (Lind et al., 2006). Overall, alterations in cog-
nition, neural structure, and neural function have been observed in individuals with
certain genetic predispositions, even in midlife.
Vascular Health
Poor vascular health, including hypertension, high blood pressure, atherosclero-
sis, and arterial stiffness, is a common risk factor for impaired cognition and brain
â•‡ 377
health. Midlife hypertension has been associated with declines in processing speed
(Knopman et al., 2001), impaired block design (Elias et al., 2004), and faster pro-
gression of WMH (Raz et al., 2007; Debette et al., 2011). Similarly, high midlife
systolic blood pressure predicted larger WMHs 10 â•‰years later (Swan et al., 1998),
as well as decreased cognitive function 26 â•‰years later (Launer et al., 1995). Finally,
neuroimaging assessments during a 2-â•‰back task revealed that middle-â•‰aged adults
with poor vascular health, metabolic syndrome, or increased arterial stiffness had
decreased neural activation of parietal and frontal regions (see Haley, 2014).
Glucose Tolerance
Impaired glucose tolerance at midlife has been related to poorer cognitive and neural
health. Midlife diabetes was associated with a decline in verbal fluency and pro-
cessing speed (Knopman et al., 2001; see also Blair et al., 2005), as well as a faster
increase in temporal horn volume, indicating hippocampal atrophy (Debette et al.,
2011). Moreover, middle-â•‰aged adults with impaired glucose tolerance exhibited a
greater decline in regional blood flow to frontal, parietal, and temporal cortices, than
people with normal glucose tolerance (Thambisetty et al., 2013).
Weight
Midlife obesity has also been affiliated with reduced executive function, visuomotor
ability, and memory (e.g., Wolf et al., 2007; Debette et al., 2011; Bischof and Park,
2015). Singh-â•‰Manoux et al. (2012) found that those with midlife obesity and meta-
bolic abnormality had the fastest decline in cognition after 10 years. Furthermore,
higher midlife BMI was associated with worse episodic memory, after controlling
for age, sex, education, blood pressure, and diabetes (Cournot et al., 2006). Notably,
Sabia et al. (2009) report that being obese or underweight during early adulthood,
early midlife, or late midlife was associated with lower cognition in late midlife. After
adjusting for cardiovascular risk factors, obesity around age 40 was related to later life
cognition, but obesity around age 60 was not, indicating that midlife obesity played
a more critical role. Weight-â•‰related structural changes are also present. For instance,
BMI was found to be independently associated with decreased brain volume in a mid-
dle-â•‰aged sample (Ward et al., 2005).
Cholesterol
Similar to the other health factors, high cholesterol in middle age has been linked
to cognitive decline (Blair et al., 2005). Kivipelto et al. (2001) found that ele-
vated midlife cholesterol was a significant risk factor for MCI. Functional changes
have also been observed, such that high cholesterol was associated with reduced
activation of the left inferior parietal lobe, right superior frontal gyrus, and right
middle frontal gyrus during a 2-â•‰back task in middle-â•‰aged participants (Gonzales
et al., 2011).
378
Depression
Persistent depressive symptoms during midlife seem to compromise cognitive and
neural function. Specifically, extensive midlife depressive symptoms were associ-
ated with poorer memory, reasoning, vocabulary, and fluency at late midlife (Singh-â•‰
Manoux et al., 2010). Moreover, those with major depressive disorder at midlife
activated the hippocampus and anterior cingulate less during episodic encoding than
controls (Bremner et al., 2004). Thus, psychobiological mental health conditions like
depression appear to influence midlife and later-â•‰life cognition and neural function.
However, the directionality of the effects remains unknown, as altered neural and cog-
nitive function may contribute to the development of various mental health conditions.
Lifestyle Factors
Like biological factors, both controllable and uncontrollable lifestyle factors influence
cognitive and neural health. In contrast to biological variables, these lifestyle factors
correspond to personal inclinations and societal influences on daily living. Whereas
large longitudinal studies on lifestyle factors have not solely examined middle-â•‰aged
individuals, in samples of both middle-â•‰aged and older adults, decreases in lifestyle
engagement in physical, social, and cognitive activities predicted decreases in verbal
speed, episodic memory, and semantic memory (Small et al., 2012; see also Sabia
et al., 2012; Lachman et al., 2015). Thus, everyday engagement has been shown to
play a critical role in cognitive aging, although Small et al. importantly acknowledge
that decreases in cognitive function also contribute to decreased lifestyle engagement.
Here, we briefly summarize research on midlife lifestyle factors and their influence
on current and future brain health.
Socioeconomic Status (SES)
People with low SES do not have equal access to certain enrichment factors (i.e., higher
education), can experience greater everyday stress, and tend to have less time for leisure
(e.g., Baum et al., 1999). Thus, low SES may minimize the effects of protective factors
and exacerbate the effects of risk factors. For example, Singh-â•‰Manoux et al. (2008) found
that atherosclerosis was correlated with inferior cognition in those with low SES only,
indicating that SES may moderate the influence of risk factors on midlife cognition.
Education and Intellectual Engagement

High levels of education and challenging intellectual engagement have been proposed
to protect against cognitive decline (e.g., see Tucker and Stern, 2011; Reuter-â•‰Lorenz
and Park, 2014). Interestingly, Mortensen et al. (2014) report that the strongest rela-
tionship between education and cognitive ability occurred in middle age. Thus, high
levels of education and cognitive engagement may serve as a protective factor, whereas
low education may predict greater cognitive decline (see Bosma et al., 2003; Lachman
et al., 2010).
â•‡ 379
Beliefs
Varying personal beliefs have also been linked to cognitive and functional differences.
Middle-â•‰aged adults who believed that they had better control over their cognition
had better reasoning performance (Soederberg Miller and Lachman, 2000; see also
Agrigoroaei and Lachman, 2011). Additionally, greater perceived cognitive dysfunc-
tion was related to decreased frontal activity during a working memory task (Haley
et al., 2011). While interesting, the direction of causality is ambiguous, as individual
belief styles may be based on self-â•‰observed cognitive performance.
Social Engagement
As mentioned in the section introduction, increased social engagement is associated
with better cognition; yet, poorer cognition may contribute to social withdrawal (e.g.,
Small et al., 2012). Specifically, in a sample of adults aged 35–â•‰85, Seeman et al.
(2011) documented that those with greater social engagement had better executive
function and episodic memory (see also Agrigoroaei and Lachman, 2011). Even
during midlife, maintaining social connections appears to benefit cognitive health.
Stress
Humans show an inverted-â•‰U relationship between stress and cognition, such that
moderate stress is associated with better cognition, but high stress is associated with
poorer cognition (Lupien et al., 2009). Consistently across the lifespan, greater levels
of self-â•‰reported daily stress were related to lower performance on a complex epi-
sodic memory task that required executive functioning (VonDras et al., 2005). Those
middle-â•‰aged adults who reported high levels of stress had performance similar to
individuals 40 years older who reported low stress.
Exercise
Physical activity has consistently been associated with better cognition and neural
health (see Kramer and Erickson, 2007). For instance, Hillman et al. (2006) report
that higher physical activity was related to better executive control on a Flanker task
in people between 40–â•‰71, and exercise has been shown to improve executive func-
tioning, attention, processing speed, and memory (see Chang et al., 2010; Randolph
and Randolph, 2013). Structurally, greater midlife exercise predicted larger total brain
volume and increased gray matter, primarily in the frontal lobes (Rovio et al., 2010).
Thus, increasing midlife exercise appears to improve overall cognition and to help
maintain brain structure.
Diet
Variations in diet have also been linked with cognitive health. High saturated fat
intake has been related to decreased cognition, whereas high polyunsaturated fatty
380
acid intake has been related to better semantic memory (Eskelinen et al., 2008). Lower
levels of Vitamin D have also been associated with slower processing speed in both
middle-â•‰aged and older adults (Lee et al., 2009). Correspondingly, a vitamin-â•‰rich, low-â•‰
fat diet may be beneficial for cognitive performance.
Sleep
Nightly sleep of between 6–â•‰8 hours seems to benefit cognition. Middle-â•‰aged and
older individuals receiving greater than 8 or fewer than 6 hours of sleep a night have
exhibited impaired cognition, including deficits in reasoning, semantic fluency, and
vocabulary (Randolph and Randolph, 2013). Compensatory strategies of extended
sleep periods after sleep withdrawal and napping have been shown to have benefi-
ciary effects, although sleep recovery periods do not increase cognitive performance
to that of healthy sleep patterns (Randolph and Randolph, 2013).
Smoking
Finally, midlife smoking has been related to decreased brain volume and increased
risk for WMH (Debette et al., 2011). Smokers also had worse cognitive performance
in middle age and a steeper rate of decline 5 â•‰years later (Nooyens et al., 2008; see also
Randolph and Randolph, 2013). Thus, midlife smoking may accelerate cognitive and
neural decline.
Conclusion
The present review reveals many gaps in our present knowledge of the middle-â•‰aged
brain. Nonetheless, some general conclusions can be drawn from the relatively scant
data that exist. We summarize these overarching themes below, also noting related
directions for future research.
Contextual Pressure During Middle Age

Middle age is characterized by high allostatic load, as middle-â•‰aged adults have high
demands in work and family life. However, the consequences of such intensity are not
well understood. It would be worthwhile to assess whether certain cognitive functions
are strengthened by contextual pressure, such that adults with the highest level of
environmental demand are facilitated. Additionally, it is possible that high contextual
pressure could be accompanied by sleep deprivation and low levels of health mainte-
nance. Work focusing on whether this is the case and the consequences for late-â•‰life
cognition would be worthwhile.
Middle Age and Peak Performance

With relatively modest declines in cognitive abilities combined with extensive life and
work experiences, middle-â•‰aged adults may possess an optimal blend of processing
â•‡ 381
capacity and knowledge. Studies that attempt to understand the relative contributions
of fluid processing and crystallized knowledge at different points in the life cycle as
relevant to complex behaviors are needed.
Individual Differences and Longitudinal Designs

Relatively little work focuses specifically on the middle-â•‰aged brain, with a particular
dearth of longitudinal studies. An investment in cross-â•‰sequential designs that include
10-â•‰to-â•‰15-â•‰year longitudinal assessments of young, middle-â•‰aged, and old cohorts, or
that even exclusively focus on middle-â•‰aged adults, could yield significant information
about the middle-â•‰aged brain and cognition within a manageable time period for indi-
vidual researchers. Such studies could fill many gaps in our incomplete knowledge of
the cognitive neuroscience of midlife.
Brain Health and Neural Plasticity in Middle Age

Much remains to be learned about the rate of change in neural health and function
during middle age. In particular, almost nothing is known to date about the plasticity
and potential for change based on life experiences or interventions exclusively during
middle age. For instance, midlife plasticity may predict a favorable future aging tra-
jectory, whereas elevated levels of amyloid relative to peers may be predictive of a
later diagnosis of Alzheimer’s disease far into the future.
Nonuniform Functional Changes in Midlife

More pronounced functional changes have been observed in brain regions involved in
effortful controlled processing in middle age compared to visual and other sensory regions.
The data are scant and it is premature to conclude that frontal, parietal, and mediotempo-
ral structures change earlier and more rapidly than sensory structures, but both regional
changes and mechanisms of change should be specifically investigated in middle age.
Foreseeing Future Cognition

It is important to understand the role that individual differences in brain health at mid-
dle age may play in predicting future cognitive aging trajectories or the diagnosis of
neuropathology. Continued study of individual differences in enrichment and deple-
tion factors at midlife will allow us to understand how early different factors influence
cognition and how deterministic they are of brain function.
Overall, it is clear that the cognitive neuroscience of midlife is nuanced. Some
cognitive abilities peak in midlife, whereas others begin to show measurable declines.
Some neural networks exhibit differences between young adulthood and middle age,
and some remain stable until old age. Moreover, ample individual differences exist.
As brain health and function at midlife may prove to be a useful harbinger of future
cognition and pathology, more research is needed to continue uncovering the intricate
details of the complex and varied middle-â•‰aged brain.
382
References
Agrigoroaei S, Lachman ME (2011) Cognitive functioning in midlife and old age: combined
effects of psychosocial and behavioral factors. The journals of Gerontology Series B,
Psychological Sciences and Social Sciences 66 Suppl 1:i130–â•‰i140.
Baltes PB, Staudinger UM, Maercker A, Smith J (1995) People nominated as wise: a compara-
tive study of wisdom-â•‰related knowledge. Psychology and Aging 10:155–â•‰166.
Baum A, Garofalo JP, Yali AM (1999) Socioeconomic status and chronic stress: Does stress
account for SES effects on health? Annals of the New York Academy of Sciences
896:131–â•‰144.
Bendlin BB, Carlsson CM, Gleason CE, Johnson SC, Sodhi A, Gallagher CL, Puglielli L,
Engelman CD, Ries ML, Xu G, Wharton W, Asthana S (2010) Midlife predictors of
Alzheimer’s disease. Maturitas 65:131–â•‰137.
Bischof Gn, Park Dc (2015) Obesity and aging: Consequences for cognition, brain structure,
and brain function. Journal of Psychosomatic Medicine 77:697–709.
Blair CK, Folsom AR, Knopman DS, Bray MS, Mosley TH, Boerwinkle E (2005) APOE geno-
type and cognitive decline in a middle-â•‰aged cohort. Neurology 64:268–â•‰276.
Bolmont T, Clavaguera F, Meyer-â•‰Luehmann M, Herzig MC, Radde R, Staufenbiel M, Lewis J,
Hutton M, Tolnay M, Jucker M (2007) Induction of tau pathology by intracerebral infusion
of amyloid-â•‰beta -â•‰containing brain extract and by amyloid-â•‰beta deposition in APP x Tau
transgenic mice. The American Journal of Pathology 171:2012–â•‰2020.
Bosma H, Van Boxtel MPJ, Ponds RWHM, Houx PJH, Jolles J (2003) Education and age-â•‰
related cognitive decline: The contribution of mental workload. Educational Gerontology
29:165–â•‰173.
Braak H, Braak E (1991) Neuropathological stageing of Alzheimer-â•‰related changes. Acta
Neuropathol 82:239–â•‰259.
Braak H, Braak E (1997) Frequency of stages of Alzheimer-â•‰related lesions in different age
categories. Neurobiology of Aging 18:351–â•‰357.
Braak H, Thal DR, Ghebremedhin E, Del Tredici K (2011) Stages of Pathologic Process in Alzheimer
Disease: Age Categories From 1 to 100 Years. J Neuropathol Exp Neurol 70:960–â•‰969.
Bremner JD, Vythilingam M, Vermetten E, Vaccarino V, Charney DS (2004) Deficits in hippo-
campal and anterior cingulate functioning during verbal declarative memory encoding in
midlife major depression. American Journal of Psychiatry 161:637–â•‰645.
Cabeza R (2002) Hemispheric Asymmetry Reduction in Older Adults: The HAROLD Model.
Psychology and Aging 17:85–â•‰100.
Caselli RJ, Dueck AC, Osborne D, Sabbagh MN, Connor DJ, Ahern GL, Baxter LC, Rapcsak
SZ, Shi J, Woodruff BK, Locke DEC, Snyder CH, Alexander GE, Rademakers R, Reiman
EM (2009) Longitudinal modeling of age-â•‰related memory decline and the APOE e4 effect.
New England Journal of Medicine 361:255–â•‰263.
Cdc (2015) Life Expectancy. In: http://â•‰www.cdc.gov/â•‰nchs/â•‰fastats/â•‰life-â•‰expectancy.htm.
Census (2014) Grandparents and Grandchildren. In: http://â•‰www.census.gov/â•‰population/â•‰www/â•‰
socdemo/â•‰grandparents.html.
Chan MY, Park DC, Savalia NK, Petersen SE, Wig GS (2014) Decreased segregation of
brain systems across the healthy adult lifespan. Proceedings of the National Academy of
Sciences of the United States of America 111:E4997–â•‰E5006.
Chang M, Jonsson PV, Snaedal J, Bjornsson S, Saczynski JS, Aspelund T, Eiriksdottir G,
Jonsdottir MK, Lopez OL, Harris TB, Gudnason V, Launer LJ (2010) The effect of
midlife physical activity on cognitive function among older adults: AGES-â•‰-â•‰Reykjavik
Study. The journals of gerontology Series A, Biological Sciences and Medical Sciences
65:1369–â•‰1374.
383
Coon KD et al. (2007) A High-Density Whole-Genome Association Study Reveals That APOE
Is the Major Susceptibility Gene for Sporadic Late-Onset Alzheimer’s Disease. Journal of
Clinical Psychiatry 68:613–619.
Cournot M, Marquie JC, Ansiau D, Martinaud C, Fonds H, Ferrieres J, Ruidavets JB (2006)
Relation between body mass index and cognitive function in healthy middle-aged men and
women. Neurology 67:1208–1214.
Debette S, Seshadri S, Beiser A, Au R, Himali JJ, Palumbo C, Wolf PA, Decarli C (2011)
Midlife vascular risk factor exposure accelerates structural brain aging and cognitive
Decarli C, Murphy DGM, Tranh M, Grady CL, Haxby JV, Gillette JA, Salerno JA, Gonzales-
Aviles A, Horwitz B, Rapoport SI, Schapiro MB (1995) The effect of white matter hyper-
intensity volume on brain structure, cognitive performance, and cerebral metabolism of
glucose in 51 healthy adults. Neurology 45:2077–2084.
Denney NW, Pearce KA (1989) A Developmental Study of Practical Problem Solving in Adults.
Denney NW, Heidrich SM (1990) Training Effects on Raven’s Progressive Matrices in Young,
Middle-Aged, and Elderly Adults. Psychology and Aging 5:144–145.
Elias PK, Elias MF, Robbins MA, Budge MM (2004) Blood pressure-related cognitive decline:
does age make a difference? Hypertension 44:631–636.
Eskelinen MH, Ngandu T, Helkala E-L, Tuomilehto J, Nissinen A, Soininen H, Kivipelto M
(2008) Fat intake at midlife and cognitive impairment later in life: a population-based
CAIDE study. International Journal of Geriatric Psychiatry 23:741–747.
Espeseth T, Greenwood PM, Reinvang I, Fjell AM, Walhovd KB, Westlye LT, Wehling
E, Lundervold A, Rootwelt H, Parasuraman R (2006) Interactive effects of APOE and
CHRNA4 on attention and white matter volume in healthy middle-aged and older adults.
Cognitive, Affective & Behavioral Neuroscience 6:31–43.
Filippini N, Macintosh BJ, Hough MG, Goodwin GM, Frisoni GB, Smith SM, Matthews PM,
Beckmann CF, Mackay CE (2009) Distinct patterns of brain activity in young carriers of
the APOE-epsilon4 allele. Proceedings of the National Academy of Sciences of the United
States of America 106:7209–7214.
Finkel D, Reynolds CA, Mcardle JJ, Gatz M, Pedersen NL (2003) Latent growth curve analyses
of accelerating decline in cognitive abilities in late adulthood. Developmental Psychology
39:535–550.
Fischl B, Dale AM, Walhovd KB (2009) High consistency of regional cortical thinning in
aging across multiple samples. Cerebral Cortex 19:2001–2012.
Garden SE, Phillips LH, Macpherson SE (2001) Midlife aging, open-ended planning, and lab-
oratory measures of executive function. Neuropsychology 15:472–482.
Ge Y, Grossman RI, Babb JS, Rabin ML, Mannon LJ, Kolson DL (2002) Age-related total gray
matter and white matter changes in normal adult brain. Part I: volumetric MR imaging
analysis. American Journal of Neuroradiology 23:1327–1333.
Giedd JN (2004) Structural magnetic resonance imaging of the adolescent brain. Annals of the
Gonzales MM, Tarumi T, Eagan DE, Tanaka H, Biney FO, Haley AP (2011) Current serum
lipoprotein levels and FMRI response to working memory in midlife. Dementia and
Geriatric Cognitive Disorders 31:259–267.
Gow AJ, Johnson W, Pattie A, Whiteman MC, Starr J, Deary IJ (2008) Mental ability in child-
hood and cognitive aging. Gerontology 54:177–186.
Greenwood PM, Lambert C, Sunderland T, Parasuraman R (2005) Effects of apolipopro-
tein E genotype on spatial attention, working memory, and their interaction in healthy,
384
middle-aged adults: results From the National Institute of Mental Health’s BIOCARD
study. Neuropsychology 19:199–211.
Greenwood PM, Espeseth T, Lin M-K, Reinvang I, Parasuraman R (2014) Longitudinal change
in working memory as a function of APOE genotype in midlife and old age. Scandinavian
Journal of Psychology 55:268–277.
Grossmann I, Na J, Varnum ME, Park Dc, Kitayama S, Nisbett RE (2010) Reasoning about
social conflicts improves into old age. Proceedings of the National Academy of Sciences
Guillozet AL, Weintraub S, Mash DC, Mesulam MM (2003) Neurofibrillary tangles, amyloid,
and memory in aging and mild cognitive impairment. Archives of Neurology 60:729–736.
Gunstad J, Paul RH, Brickman AM, Cohen RA, Arns M, Roe D, Lawrence JJ, Gordon E (2006)
Patterns of cognitive performance in middle-aged and older adults: A cluster analytic
examination. Journal of Geriatric Psychiatry and Neurology 19:59–64.
Haley AP (2014) Vascular Functions and Brain Integrity in Midlife: Effects of Obesity and
Metabolic Syndrome. Advances in Vascular Medicine 2014:1–7.
Haley AP, Eagan DE, Gonzales MM, Biney FO, Cooper RA (2011) Functional magnetic res-
onance imaging of working memory reveals frontal hypoactivation in middle-aged adults
with cognitive complaints. Journal of the International Neuropsychological Society
17:915–924.
Hamberger M, Friedman D (1992) Event-related potential correlates of repetition priming and
stimulus classification in young, middle-aged, and older adults. Journal of Gerontology
47:P395–P405.
Hardy J, Selkoe DJ (2002) The amyloid hypothesis of Alzheimer’s disease: progress and prob-
lems on the road to therapeutics. Science 297:353–356.
Hartshorne JK, Germine LT (2015) When Does Cognitive Functioning Peak? The Asynchronous
Rise and Fall of Different Cognitive Abilities Across the Life Span. Psychol Sci 26:433–443.
Hedden T, Oh H, Younger AP, Patel TA (2013) Meta-analysis of amyloid-cognition relations in
cognitively normal older adults. Neurology 80:1341–1348.
Heidrich SM, Denney NW (1994) Does social problem solving differ from other types of prob-
lem solving during the adult years? Experimental aging research 20:105–126.
Helson R, Soto CJ (2005) Up and down in middle age: monotonic and nonmonotonic
changes in roles, status, and personality. Journal of Personality and Social Psychology
89:194–204.
Helson R, Soto CJ, Cate RA (2006) From Young Adulthood Through the Middle Ages. In:
(Mroczek DK, Little TD, eds), pp. 337–352. New York: Psychology Press.
Hillman CH, Motl RW, Pontifex MB, Posthuma D, Stubbe JH, Boomsma DI, De Geus EJC
(2006) Physical activity and cognitive function in a cross-section of younger and older
community-dwelling individuals. Health Psychology 25:678–687.
Huang CM, Polk TA, Goh JO, Park DC (2012) Both left and right posterior parietal activations
contribute to compensatory processes in normal aging. Neuropsychologia 50:55–66.
Hultsch DF, Hertzog C, Small BJ, Mcdonald-Miszczak L, Dixon RA (1992) Short-term longi-
tudinal change in cognitive performance in later life. Psychology and Aging 7:571–584.
Jack CR, Knopman DS, Jagust WJ, Petersen RC, Weiner MW, Aisen PS, Shaw LM, Vemuri P,
Wiste HJ, Weigand SD, Lesnick TG, Pankratz VS, Donohue MC, Trojanowski JQ (2013)
Tracking pathophysiological processes in Alzheimer’s disease: an updated hypothetical
model of dynamic biomarkers. The Lancet Neurology 12:207–216.
Josefsson M, De Luna X, Pudas S, Nilsson LG, Nyberg L (2012) Genetic and lifestyle pre-
dictors of 15-year longitudinal change in episodic memory. Journal of the American
Geriatrics Society 60:2308–2312.
385
Kennedy KM, Rodrigue KM, Devous MD, Sr., Hebrank AC, Bischof GN, Park DC (2012)
Effects of beta-amyloid accumulation on neural function during encoding across the adult
lifespan. NeuroImage 62:1–8.
Kennedy KM, Rodrigue KM, Bischof GN, Hebrank AC, Reuter-Lorenz PA, Park DC (2015)
Age trajectories of functional activation under conditions of low and high processing
demands: an adult lifespan fMRI study of the aging brain. NeuroImage 104:21–34.
Kivipelto M, Helkala E-L, Hanninen T, Laakso MP, Hallikainen M, Alhainen K, Soininen H,
Tuomilehto J, Nissinen A (2001) Midlife vascular risk factors and late-life mild cognitive
impairment: A population-based study. Neurology 56:1683–1689.
Klunk WE et al. (2004) Imaging Brain Amyloid in Alzheimer’s Disease with Pittsburgh
Compound-B. Ann Neurol 55:306–319.
Knopman D, Boland LL, Mosley T, Howard G, Liao D, Szklo M, Mcgovern P, Folsom AR
(2001) Cardiovascular risk factors and cognitive decline in middle-aged adults. Neurology
56:42–48.
Kramer AF, Erickson KI (2007) Capitalizing on cortical plasticity: influence of physical activ-
ity on cognition and brain function. Trends in Cognitive Sciences 11:342–348.
Lachman ME (2004) Development in midlife. Annual Review of Psychology 55:305–331.
Lachman ME, Teshale S, Agrigoroaei S (2015) Midlife as a Pivotal Period in the Life Course:
Balancing Growth and Decline at the Crossroads of Youth and Old Age. International
Journal of Behavioral Development 39:20–31.
Lachman ME, Lewkowicz C, Marcus A, Peng Y (1994) Images of midlife development among
young, middle-aged, and older adults. Journal of Adult Development 1:201–211.
Lachman ME, Agrigoroaei S, Murphy C, Tun PA (2010) Frequent cognitive activity compen-
sates for education differences in episodic memory. Am J Geriatr Psychiatry 18:4–10.
Larue A, Jarvik LF (1987) Cognitive function and prediction of dementia in old age. International
Journal of Aging and Human Development 25:79–89.
Launer LJ, Masaki K, Petrovitch H, Foley D, Havlik RJ (1995) The association between midlife
blood pressure levels and late-life cognitive function. The Honolulu-Asia Aging Study.
Journal of the American Medical Association 274:1846–1851.
Lee DM et al. (2009) Association between 25-hydroxyvitamin D levels and cognitive perfor-
mance in middle-aged and older European men. Journal of Neurology, Neurosurgery, and
Lind J, Persson J, Ingvar M, Larsson A, Cruts M, Van Broeckhoven C, Adolfsson R, Backman L,
Nilsson LG, Petersson KM, Nyberg L (2006) Reduced functional brain activity response
in cognitively intact apolipoprotein E epsilon4 carriers. Brain 129:1240–1248.
Lupien SJ, Mcewen BS, Gunnar MR, Heim C (2009) Effects of stress throughout the lifespan
on the brain, behaviour and cognition. Nature Reviews Neuroscience 10:434–445.
Macpherson SE, Phillips LH, Della Sala S (2002) Age, Executive Function, and Social Decision
Making: A Dorsolateral Prefrontal Theory of Cognitive Aging. Psychology and Aging
17:598–609.
Martin M, Zimprich D (2005) Cognitive development in midlife. In: Middle Adulthood:
A Lifespan Perspective (Willis Sl, Martin M, eds), pp. 179–206. Thousand Oaks: Sage
Publications, Inc.
Mitchell DB, Bruss PJ (2003) Age differences in implicit memory: conceptual, perceptual, or
methodological? Psychology and ging 18:807–822.
Mitchell TW, Mufson EJ, Schneider JA, Cochran EJ, Nissanov J, Han LY, Bienias JL, Lee
VM, Trojanowski JQ, Bennett DA, Arnold SE (2002) Parahippocampal tau pathology
in healthy aging, mild cognitive impairment, and early Alzheimer’s disease. Ann Neurol
51:182–189.
386
Mortensen EL, Flensborg-Madsen T, Molbo D, Fagerlund B, Christensen U, Lund R, Osler M,

Avlund K (2014) The relationship between cognitive ability and demographic factors in
late midlife. Journal of Aging and Health 26:37–53.
Nooyens AC, Van Gelder BM, Verschuren WM (2008) Smoking and cognitive decline among
middle-aged men and women: the Doetinchem Cohort Study. American Journal of Public
Health 98:2244–2250.
Park DC, Reuter-Lorenz PA (2009) The adaptive brain: aging and neurocognitive scaffolding.
Annual Review of Psychology 60:173–196.
Park DC, Lautenschlager G, Hedden T, Davidson NS, Smith AD, Smith PK (2002) Models
of Visuospatial and Verbal Memory Across the Adult Life Span. Psychology and Aging
17:299–320.
Park H, Kennedy KM, Rodrigue KM, Hebrank A, Park DC (2013) An fMRI study of episodic
encoding across the lifespan: changes in subsequent memory effects are evident by mid-
dle-age. Neuropsychologia 51:448–456.
Park J, Carp J, Kennedy KM, Rodrigue KM, Bischof GN, Huang CM, Rieck JR, Polk TA,
Park DC (2012) Neural broadening or neural attenuation? Investigating age- related
dedifferentiation in the face network in a large lifespan sample. Journal of Neuroscience
32:2154–2158.
Pfefferbaum A, Sullivan EV (2003) Increased brain white matter diffusivity in normal adult
aging: relationship to anisotropy and partial voluming. Magnetic Resonance in Medicine
49:953–961.
Pudas S, Persson J, Nilsson L-G, Nyberg L (2014) Midlife memory ability accounts for brain
activity differences in healthy aging. Neurobiology of Aging 35:2495–2503.
Randolph JS, Randolph JJ (2013) Modifiable lifestyle factors and cognition through midlife. In:
Positive Neuropsychology: Evidence-Based Perspectives on Promoting Cognitive Health
(Randolph JJ, ed), pp. 25–55. New York, NY: Springer New York.
Raz N, Rodrigue KM, Head D, Kennedy KM, Acker JD (2004) Differential aging of the medial
temporal lobe: A study of a five-year change. Neurology 62:433–438.
aging in middle-aged and older adults: regional and individual differences. NeuroImage
51:501–511.
Reuter-Lorenz PA, Park DC (2014) How Does it STAC Up? Revisiting the Scaffolding Theory
of Aging and Cognition. Neuropsychol Rev 24:355–370.
Ribot T (1891) Les Maladies de la Mémoire. Paris: Germer-Baillière.
Rodrigue KM, Kennedy KM, Devous MD, Sr., Rieck JR, Hebrank AC, Diaz-Arrastia R,
Mathews D, Park DC (2012) beta-Amyloid burden in healthy aging: regional distribution
and cognitive consequences. Neurology 78:387–395.
Ronnlund M, Nyberg L, Backman L, Nilsson LG (2005) Stability, growth, and decline in adult
population-based study. Psychology and Aging 20:3–18.
Rovio S, Spulber G, Nieminen LJ, Niskanen E, Winblad B, Tuomilehto J, Nissinen A, Soininen
H, Kivipelto M (2010) The effect of midlife physical activity on structural brain changes
in the elderly. Neurobiology of Aging 31:1927–1936.
387
Sabia S, Kivimaki M, Shipley MJ, Marmot MG, Singh-Manoux A (2009) Body mass index
over the adult life course and cognition in late midlife: the Whitehall II Cohort Study.
American Journal of Clinical Nutrition 89:601–607.
Sabia S, Singh-Manoux A, Hagger-Johnson G, Cambois E, Brunner EJ, Kivimaki M (2012)
Influence of individual and combined healthy behaviours on successful aging. Canadian
Medical Association Journal 184:1985–1992.
Salat DH, Tuch DS, Hevelone ND, Fischl B, Corkin S, Rosas HD, Dale AM (2005) Age-related
changes in prefrontal white matter measured by diffusion tensor imaging. Annals of the
Salat DH, Buckner RL, Snyder AZ, Greve DN, Desikan RS, Busa E, Morris JC, Dale AM,
Fischl B (2004) Thinning of the cerebral cortex in aging. Cerebral Cortex 14:721–730.
Salthouse TA (2009) When does age-related cognitive decline begin? Neurobiology of Aging
30:507–514.
Salthouse TA (2013) Effects of age and ability on components of cognitive change. Intelligence
41:501–511.
Salthouse TA (2014) Correlates of cognitive change. Journal of Experimental Psychology
General 143:1026–1048.
Salthouse TA, Soubelet A (2014) Heterogeneous ability profiles may be a unique indicator of
impending cognitive decline. Neuropsychology 28:812–818.
Seeman TE, Miller-Martinez DM, Stein Merkin S, Lachman ME, Tun PA, Karlamangla AS
(2011) Histories of social engagement and adult cognition: midlife in the U.S. study. The
Journals of Gerontology Series B, Psychological Sciences and Social Sciences 66 Suppl
1:i141–i152.
Singh-Manoux A, Britton A, Kivimaki M, Gueguen A, Halcox J, Marmot M (2008)
Socioeconomic status moderates the association between carotid intima-media thick-
ness and cognition in midlife: evidence from the Whitehall II study. Atherosclerosis
197:541–548.
Singh-Manoux A, Akbaraly TN, Marmot M, Melchior M, Ankri J, Sabia S, Ferrie JE (2010)
Persistent depressive symptoms and cognitive function in late midlife: the Whitehall II
study. Journal of Clinical Psychiatry 71:1379–1385.
Singh-Manoux A, Czernichow S, Elbaz A, Dugravot A, Sabia S, Hagger- Johnson G,
Kaffashian S, Zins M, Brunner Ej, Nabi H, Kivimaki M (2012) Obesity phenotypes
in midlife and cognition in early old age: the Whitehall II cohort study. Neurology
79:755–762.
Small BJ, Dixon RA, Mcardle JJ, Grimm KJ (2012) Do changes in lifestyle engagement mod-
erate cognitive decline in normal aging? Evidence from the Victoria Longitudinal Study.
Soederberg Miller LM, Lachman ME (2000) Cognitive Performance and the Role of Control
Beliefs in Midlife. Aging, Neuropsychology, and Cognition 7:69–85.
Sperling RA, Laviolette PS, O’keefe K, O’brien J, Rentz DM, Pihlajamaki M, Marshall G,
Sperling RA et al. (2011) Toward defining the preclinical stages of Alzheimer’s disease:
recommendations from the National Institute on Aging-Alzheimer’s Association work-
groups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & Dementia
7:280–292.
Sullivan EV, Rohlfing T, Pfefferbaum A (2010) Quantitative fiber tracking of lateral and
interhemispheric white matter systems in normal aging: relations to timed performance.
Neurobiology of Aging 31:464–481.
388
Swan GE, Decarli C, Miller BL, Reed T, Wolf PA, Jack LM, Carmelli D (1998) Association
of midlife blood pressure to late-life cognitive decline and brain morphology. Neurology
51:986–993.
Thambisetty M, Beason-Held LL, An Y, Kraut M, Metter J, Egan J, Ferrucci L, O’brien R,
Resnick SM (2013) Impaired glucose tolerance in midlife and longitudinal changes in
brain function during aging. Neurobiology of Aging 34:2271–2276.
Tucker AM, Stern Y (2011) Cognitive Reserve in Aging. Current Alzheimer Research 8:354–360.
Villemagne VL, Pike KE, Chetelat G, Ellis KA, Mulligan RS, Bourgeat P, Ackermann U,
Jones G, Szoeke C, Salvado O, Martins R, O’keefe G, Mathis CA, Klunk WE, Ames D,
Masters Cl, Rowe CC (2011) Longitudinal assessment of Abeta and cognition in aging and
Alzheimer disease. Ann Neurol 69:181–192.
Vondras DD, Powless MR, Olson AK, Wheeler D, Snudden AL (2005) Differential effects of
everyday stress on the episodic memory test performances of young, mid-life, and older
adults. Aging & Mental Health 9:60–70.
Ward MA, Carlsson CM, Trivedi MA, Sager MA, Johnson SC (2005) The effect of body
mass index on global brain volume in middle-aged adults: a cross sectional study. BMC
Neurology 5:23.
Willis SL, Schaie KW (1999) Intellectual Functioning in Midlife. In: Life in the Middle:
Psychological and Social Development in Middle Age (Willis SL, Reid JD, eds). San
Diego: Academic Press.
Willis SL, Schaie KW (2005) Cognitive Trajectories in Midlife and Cognitive Functioning in
Old Age. In: Middle Adulthood: A Lifespan Perspective (Willis SL, Martin M, eds), pp.
243–275. Thousand Oaks, California: Sage Publications, Inc.
Wolf PA, Beiser A, Elias MF, Au R, Vasan RS, Seshadri S (2007) Relation of obesity to cog-
nitive function: importance of central obesity and synergistic influence of concomitant
hypertension. The Framingham Heart Study. Current Alzheimer Research 4:111–116.
Zelinski EM, Burnight KP (1997) Sixteen-year longitudinal and time lag changes in memory
and cognition in older adults. Psychology and Aging 12:503–513.
389
15
The Modifying Role of Hypertension

in Cognitive and Brain Aging
Karen M. Rodrigue
Gérard N. Bischof
E ssential hypertension, defined by a chronic elevation in blood pressure for

which there is no identifiable cause, is a disease of considerable propor-
tion affecting an estimated 33% of adults over the age of 20 years in the United States,
with approximately half of those diagnosed lacking adequate medical control to target
blood pressure levels of systolic pressure ≤ 140 mm HG (millimeters of mercury) and
diastolic pressure ≤ 90 mm HG (Go et al., 2013). Prevalence of the disease increases
sharply with age, as approximately 67% of adults over age 60 in the U.S. are hyper-
tensive (Nwankwo et al., 2013). Critically, hypertension, particularly if untreated, is
a major risk factor for cardiovascular disease including myocardial infarction, stroke,
and congestive heart failure (Stamler, Stamler, and Neaton, 1993).
Chronic elevation of blood pressure is associated with a cascade of pathophysi-
ological changes to the vasculature including increased cardiac output and greater
perfusion of tissue (Folkow et al., 1984), which leads to a vasoconstrictive response
and remodeling of the vascular wall including a decrease in the diameter of blood
vessels (Boegehold, 2007) and arterial thickening (Touyz, 2005). In addition to pro-
moting cardiovascular disease, these hypertension-related changes are associated with
cognitive performance decrements and changes in brain structure and function. In
fact, it has been proposed that the brain is an early and critical target of initial disease
mechanisms (see Jennings and Zanstra, 2009 for review). Maintaining optimal brain
and cognitive function is in part dependent upon a healthy vascular system to deliver
a steady supply of essential nutrients to neural tissue. Thus, the increase in the preva-
lence of hypertension with age combined with its detrimental effects on the vascular
389
390
system makes it an important factor to consider in the pursuit of understanding brain

and cognitive aging.
The primary goal of this chapter is to provide an overview of the literature exam-
ining the modifying effects of hypertension on non-â•‰pathological brain and cognitive
aging. We briefly review the overall pattern of findings from the hypertension and
cognitive aging literature and discuss the effects of hypertension on normal brain
aging, including studies of brain structure, resting cerebral blood flow and the limited
task-â•‰related functional activation literature. Finally, the role of hypertension and over-
all vascular health as a risk factor for pathological aging and Alzheimer’s disease will
be discussed. We conclude with a summary of key findings and discuss the implica-
tions of this research for understanding neurocognitive aging.
Hypertension Effects on Normal Cognitive Aging
A considerable body of research has documented the association of blood pressure

elevations with poorer cognitive performance, with experiments dating back to as
early as the 1950s reporting significant decline in psychomotor function in schizo-
phrenic patients with hypertension compared to patients diagnosed with hypoten-
sion (King, 1956). Since this time, numerous studies utilizing both cross-â•‰sectional
and longitudinal designs further delineated the specific effects of hypertension on
cognitive performance in both normal and pathological aging. Here we focus on
the effects of hypertension in non-â•‰pathological aging. Overall, two interesting pat-
terns emerge from the literature: (1) Across multiple studies, hypertension appears
to affect cognitive domains that are most age-â•‰sensitive (i.e., processing speed,
executive and fluid abilities and memory performance, but not crystallized or ver-
bal abilities); (2) the relationship of blood pressure to cognition is complex and
age-â•‰dependent.
Effect of Hypertension on Specific Cognitive Domains

A plethora of studies illustrate the adverse effects of hypertension on a wide array of
cognitive tasks sensitive to aging (Waldstein et al., 1991 for an early review) including
executive functions (Bucur and Madden, 2010; Madden and Blumenthal, 1998), fluid
intelligence (Raz, Rodrigue, et al., 2007), working memory (M. F. Elias, Dore, et al.,
2010; Raz et al., 2007), processing speed (Hannesdottir et al., 2009), and episodic
memory performance (Brady, Spiro, and Gaziano, 2005; M. F. Elias, Wolf, et al.,
1993; Swan, Carmelli, and Larue, 1998). Recent meta-â•‰analyses (Gifford et al., 2013;
van den Berg et al., 2009) examining the cognitive correlates of hypertension reported
that the strength of effect sizes across cognitive domains varied, with the strongest
effects observed in memory (d = −â•‰0.4) and executive functions/â•‰attention (d = −â•‰0.4),
with weaker effects for processing speed (d = −â•‰0.2), general intelligence (d = −â•‰0.1),
and cognitive flexibility (d = −â•‰0.1), and no reliable effects on language ability (van
den Berg et al., 2009). The most recent meta-â•‰analysis published to date adjusted for
â•‡ 391
The Modifying Role of Hypertension in Cognitive and BrainÂ€Agingâ•… 391
demographic (age, education) and other cardiovascular factors (diabetes, cholesterol,

cardiac disease) and showed a similar, but more modest pattern reporting significant
associations for episodic memory (r = -â•‰.20), attention (r = .14), and global cognition (r
= -â•‰.11), with only trend level effects for language (Gifford et al., 2013).
Longitudinal studies, which offer insight into the long-â•‰term effects of hypertension
over time, demonstrate an accelerated effect of the disease on cognition. Specifically,
hypertension has been associated with decline in both executive function (Wolf et al.,
2007) and episodic memory (Swan, Carmelli, et al., 1998) in middle-â•‰aged individu-
als across longer (i.e., 25 to 30 years) and shorter (i.e., 4 to 12 years) time intervals.
Further, chronicity of hypertension over a ten-â•‰year period was associated with subse-
quent cognitive decline, beyond the effects of aging, on measures of attention, mem-
ory and a composite of neuropsychological tests (M. F. Elias et al., 1993). Increases in
mid-â•‰life blood pressure measured over two years predicted poorer executive function
performance and visuomotor skills eight and twelve years later (Wolf et al., 2007).
Thus, longitudinal studies illustrate the role of hypertension as one important predic-
tor or modifier of individual differences in cognitive decline over time.
Overall, hypertension is associated with poorer cognitive performance across
several age-â•‰sensitive cognitive domains and chronic hypertension over long peri-
ods appears to accelerate cognitive aging, even within healthy adults. The general
pattern of results suggests that hypertension may exert its strongest effects on age-â•‰
sensitive domains of cognition (i.e., memory and executive function) whereas cog-
nitive function known to be preserved with advanced aging such as verbal abilities
(Baltes, Staudinger, and Lindenberger, 1999; Park et al., 2002) appears to be spared
from the detrimental effects of hypertension. There are however, some discrepancies
in the pattern of findings across the literature, including null results that cannot be
reviewed in depth in a chapter of this scope (Desmond, Tatemichi, Paik, and Stern,
1993; Harrington et al., 2000; Scherr et al., 1991). The specificity and sensitivity of
selected cognitive functions to hypertension may be dependent upon several factors
which can vary within and across studies including type of anti-â•‰hypertensive medi-
cation used, duration of hypertension, the extent of elevation in actual blood pressure
levels, the presence of other vascular risk factors, in addition to differences in the
selection criteria for normal aging samples.
Age-â•‰Dependent Associations of Blood Pressure and Cognition

Hypertension effects on cognitive aging may vary depending upon when in the
lifespan blood pressure is assessed. There is mounting evidence that optimal blood
pressure levels may differ across the lifespan, with both low and high blood pres-
sure linked to poorer cognition. Additionally, low blood pressure or hypotension has
been associated with decreased global cognition (i.e., MMSE or Cognitive Ability
Screening Instrument CASI) in older adults (Guo et al., 1997; Kähönen-â•‰Väre et al.,
2004; Launer et al., 1995; Pandav et al., 2003). A characteristic among the major-
ity of studies reporting hypotension effects is that participants are usually aged 75
years or older, suggesting that the association between elevated blood pressure and
392
cognition is age-dependent. Other studies including cohorts older than 65 years of age
reported curvilinear associations of blood pressure and cognition (Morris et al., 2002;
Waldstein et al., 2005). Poorer cognitive performance was associated with lower sys-
tolic pressure of 100 mm Hg as well as higher levels of 180 mm Hg. This curvilinear
association was also observed for diastolic blood pressure, with lower neuropsycho-
logical test performance linked to both low (60 mm Hg) and high (100 mm Hg) dia-
stolic pressures (Morris et al., 2002). This complex pattern of results has been also
reported for executive function and memory performance (Waldstein et al., 2005).
Hypotension in old age is associated with an increased risk for mortality, cognitive
impairment (e.g., MMSE < 24) and heart failure (Guo et al., 1997), possibly reflective
of pathological aging (Swan, Carmelli, et al., 1998). Therefore, both hypotension and
hypertension may predict cognitive decrements, suggesting that an optimal range of
blood pressure is desirable for the maintenance of normal cognitive function with
advanced age.
Interestingly, the majority of longitudinal studies examining the influence of
mid-life hypertension report adverse effects on cognitive function in later life (P.
K. Elias, Elias, et al., 2004; Kilander, Nyman, Boberg, and Lithell, 2000; Reijmer
et al., 2012; Swan, Carmelli, et al., 1998). In contrast, longitudinal studies of
late-life hypertension and cognition produced mixed results. On the one hand, a
population-based study in older adults (≥ 65 years) over a six-year interval did
not report an association of baseline hypertension with cognitive decline (Hebert
et al., 2004). On the other hand, a study with a similar re-test interval (7 years in
individuals ≥ 58 years of age) reported greater decline on processing speed for
individuals with hypertension at baseline (Knopman et al., 2001), while others
only found significant association of hypertension and cognition in older men, but
not women (M. F. Elias, Elias, et al., 2003). In addition, global cognitive decline
(i.e., MMSE) over nine years was associated with both elevated and decreased
systolic blood pressure in a population sample of 3657 adults aged 65 and older
(Glynn et al., 1999). Over a nine-year follow-up interval, elevated pulse pressure
(i.e., systolic minus diastolic pressure) in participants aged 70 to 75 significantly
predicted declines in verbal learning, whereas for participants aged 76 to 80 years,
elevated systolic blood pressure was associated with decline in executive function
(Yasar, Ko, et al., 2011). Therefore, whereas adverse effects of non-optimal blood
pressure on cognitive decline are relatively consistent in studies of middle-aged
adults, the association between late-life blood pressure and late-life cognition is
less clear.
An important factor to consider in evaluating apparent inconsistencies across the
literature on aging, hypertension and cognition is the presence and degree of blood
pressure control. Although some experiments nicely control for this variability, study
samples may include a mixture of untreated, successfully medicated and inadequately
controlled blood pressure. Evidence suggests that individuals whose blood pressure is
medically controlled are at lesser risk for cognitive declines than those whose hyperten-
sion remains untreated (Dufouil et al., 2001; Fukuda and Kitani, 1995; Richards et al.,
2000). However, even treated hypertension may be associated with selected decreases
in cognitive performance (e.g., Raz, Rodrigue, and Acker, 2003). Interestingly, the
â•‡ 393
inclusion of otherwise healthy, medically treated hypertensive participants in a sample

can change the pattern of correlations among age, cognitive performance and regional
brain volume (Head, Raz, et al., 2002; Raz et al., 2005). For example, in one study
(Head et al., 2002), the significant effect of prefrontal cortical volume on the executive
performance component of the Tower of Hanoi task in an aging sample was elimi-
nated with exclusion of medically treated hypertensive participants.
Effects of Hypertension on Non-â•‰pathological

Structural Brain Aging
The impact of years of treated hypertension or uncontrolled blood pressure may

exert its effect on cognition at least in part, through changes in brain integrity and
function (Jennings and Heim, 2012). Many studies have shown that hypertension
is associated with differences in brain structure beyond the changes expected in
the normal brain aging process. Hypertension effects have been reported on the
volumetric level in both in vivo MRI studies and in postmortem brain studies.
Decreased total cerebral cortical volume (Hatazawa et al., 1984; Launer et al.,
2000), ventricular expansion and sulcal widening (Knopman et al., 2011; Salerno
et al., 1992) have been reported in hypertension persons relative to the brains of
individuals without hypertension.
Effects of Hypertension on Grey Matter

Studies examining the effect of hypertension on grey matter show differential regional
vulnerability. For example, volume differences in cortical regions have been linked
to hypertension, with a particular vulnerability of frontal regions (Chen et al., 2006;
Jennings et al., 2012; Leritz et al., 2011; Raz et al., 2005; Raz et al., 2007; Taki et al.,
2004) such as the superior frontal gyrus or dorsolateral prefrontal cortex (Gianaros
et al., 2006; Raz et al., 2003). Hypertension has also been associated with decreased
volume of more posterior regions such as inferior temporal, inferior parietal and
occipital cortex (Leritz et al., 2011; Raz et al., 2005; Taki et al., 2004). In addition
to cortical brain volume differences, subcortical effects of hypertension have been
reported in regions such as the thalamus (Jennings and Heim, 2012; Strassburger et
al., 1997). The most notable subcortical effects have been found in the hippocampus,
which appears to be especially vulnerable to hypertension (den Heijer et al., 2005;
Korf, White, Scheltens, and Launer, 2004; Raz et al., 2005; Raz et al., 2007). A
recent meta-â•‰analysis accounting for important co-â•‰variables (i.e., age, sex, education,
vascular risk factors) reported significantly smaller hippocampal volume in individ-
uals with hypertension compared to controls (Beauchet et al., 2013). A high-â•‰resolu-
tion hippocampal subfield measurement study recently showed that CA1-â•‰2 subfield
volumes specifically may be reduced in hypertensive adults (Shing et al., 2011), in
accord with animal models showing decreased CA1 neurons in hypertensive rats
(Sabbatini et al., 2000).
394
The majority of studies to date used cross-â•‰sectional designs to examine the effect
of hypertension on grey matter volume. However, longitudinal designs are required
to assess within-â•‰person change and variability. Interestingly, longitudinal data illus-
trate a nonâ•‰linear, accelerated effect of hypertension on hippocampal shrinkage in both
medicated and unmedicated individuals (Korf et al., 2004). Duration of elevated blood
pressure is also an important factor, as it has been associated with accelerated whole
brain atrophy and atrophy of frontal areas including orbitofrontal and prefrontal cor-
tex (DeCarli et al., 1999; den Heijer et al., 2005; Raz et al., 2003; Swan, Carmelli,
et al., 1998). Thus, time spent with hypertension is an important variable to consider
when interpreting study results (Muller et al., 2014; Raz et al., 2005). Longitudinal
evidence suggests an increased susceptibility over time to brain regions such as the
frontal lobes, which show substantial age-â•‰related shrinkage in normal aging (Raz et
al., 2005). Additionally, shrinkage in regions usually affected to a lesser extent by the
aging process such as such as primary visual and inferior parietal cortex have been
observed in persons carrying multiple vascular risk factors, including hypertension
(Raz et al., 2007).
One hypothesized mechanism underlying the effect of hypertension on regional
brain volume is a decrease in cerebral perfusion (Veglio et al., 2009). Hippocampus
and frontal regions (Beason-â•‰Held et al., 2007; Cohen, 2007; Dai et al., 2008), as well
as occipital and parietal areas (Bangen et al., 2014; Dai et al., 2008) exhibit reduced
blood flow with hypertension. These areas of hypoperfusion partially correspond with
regions of hypertension-â•‰related grey matter atrophy, suggesting a potential mechanis-
tic link, although studies directly testing this hypothesis are needed. Atherosclerotic
changes in an animal model of hypertension supports this hypothesis, as spontaneous
hypertension in the rat is associated with decreased cerebral blood flow (Al-â•‰Sarraf
and Philip, 2003) and leakage of the blood-â•‰brain barrier, which leads to small vessel
injuries (Schreiber et al., 2013) and possibly contributes to brain atrophy (Tayebati,
Tomassoni, and Amenta, 2012).
Studies cumulatively suggest that hypertension not only exacerbates the known
pattern of normal structural brain aging, but also affects more posterior grey matter
volumes that are generally invariant in healthy aging. It is evident that hypertension is
associated with focal grey matter differences, most notably in frontal and hippocam-
pal regions, and to a lesser extent in posterior regions including parietal and occipital
cortex. The link between hypertension and structural brain changes may be indirect,
and additional research is needed to understand the underlying mechanisms, such
as changes in cerebral perfusion, driving these effects. Finally, longitudinal studies
examining brain volume loss are needed to understand the long-â•‰term effects of treated
vs. untreated hypertension on brain structure.
Effects of Hypertension on White Matter

Volume of the different regions of the white matter compartment are reportedly
smaller in individuals with hypertension, especially in white matter underlying pre-
frontal cortex (Raz et al., 2003; Scullin et al., 2013) and in the parietal lobes (Hajjar
395
The Modifying Role of Hypertension in Cognitive and Brain Aging 395
et al., 2010). However, not all studies find significant effects of hypertension on white
matter volume (Burgmans et al., 2010). White matter volume is a relatively gross
measure of white matter structure and may be less sensitive to structural alterations
during aging compared to finer-grained MRI measurements of white matter integrity
(Burgmans et al., 2011). One such frequently reported indicator of white matter health
is the presence of white matter hyperintensities (WMH) (Raz and Kennedy, 2009).
The pathology underlying these WMH is leukoaraiosis (Hachinski, Potter, and
Merskey, 1987), which is radiologically manifested as areas of increased signal
intensities on T2/Flair sequences on MRI scans. WMH may reflect multiple vascular
pathologies including arteriosclerotic and ischemic lesions, expansion of perivascu-
lar spaces and axonal loss and patches of demyelination (de Leeuw, De Groot, and
Breteler, 2001; Pantoni and Garcia, 1997). Indeed, one of the most well documented
predictors of WMH burden, besides age, is hypertension (DeCarli et al., 1999; Dufouil
et al., 2001; Gunning-Dixon et al., 2009; Longstreth et al., 1996). Longitudinal evi-
dence of WMH progression over five years showed a doubling of WMH in the pari-
etal lobe in a vascular risk group in contrast to very mild increases of WMH seen in
risk-free individuals (Raz et al., 2007). Research from the northern Manhattan study
showed that high blood pressure at baseline and increases over seven years were both
independently associated with greater volume of WMH in a population of 1281 par-
ticipants ranging in age from 40 to 94 years (Marcus et al., 2011). Interestingly, exam-
ination of the progression of WMH with shorter re-test intervals of only 15 months
indicated that elevated blood pressure and history of hypertension are associated with
WMH expansion in the deep white matter of the frontal cortex (Jennings et al., 2012).
These studies suggest that hypertension manifests itself in increased WMH burden
over short and longer periods of time, showing a specific pattern of regional vulnera-
bility, with frontal regions affected in the short run and parietal regions affected in the
longer run. Overall, these findings suggest an anterior to posterior spread of vascular
related insult to white matter, with earlier vulnerability seen in anterior brain regions
followed by later involvement of posterior areas (Artero et al., 2004).
More recently, studies examining microstructural differences in white matter using
diffusion tensor imaging (DTI) in hypertensive individuals have been conducted.
Early findings reported significant effects of hypertension on white matter integrity in
the occipital and parietal lobes in normal aging, which was modified further by a pro-
longed history of hypertension (Kennedy and Raz, 2009). Increases in arterial blood
pressure have been associated with decreases in white matter integrity selectively in
frontal and parietal lobes (Salat et al., 2012) and anterior portions of the corpus cal-
losum, but not the posterior splenium in hypertensive adults (Leritz et al., 2010). One
of the largest samples of middle-aged and older adults (N = 4537) to date observed
that severe hypertension (>160/100 mm Hg) was associated with decreased white
matter integrity in multiple association tracts including the inferior-frontal occipital
fasciculus, inferior longitudinal fasciculus, posterior-thalamic radiation, superior lon-
gitudinal fasciculus and the anterior cingulum (de Groot et al., 2015).
Studies examining the impact of hypertension on both indices of white matter
health (i.e., integrity as assessed with DTI and WMH) simultaneously are mixed in
their findings. Thus, the nature of the relationship between white matter degradation
396
processes as measured by DTI and WMH remains unclear. Two studies found that
effects of hypertension on white matter integrity persist, although may be somewhat
attenuated when controlling for global WMH burden (de Groot et al., 2015; Salat et
al., 2012). One study reported stronger age-â•‰related effects on white matter integrity
than on volume of WMH in individuals with hypertension or elevated blood pres-
sure (Burgmans et al., 2010), whereas another study found the opposite, suggesting a
stronger link of blood pressure to WMH but minimal relationship with DTI measures
(Jacobs et al., 2013). Overall, it appears that microstructural properties of white matter
as quantified by DTI and WMH measurement are more sensitive to negative effects of
hypertension compared to volumetric measurements. It is possible that the appearance
of WMH and degradation of white matter fiber integrity (measured via DTI) may
track earlier detrimental changes with aging and/â•‰or hypertension before overt white
matter volume losses emerge. However, the precise temporal order of these white mat-
ter changes remains to be established by longitudinal investigations of hypertension
and brain aging.
One underlying mechanism that may partly explain the vulnerability of white mat-
ter to hypertension is altered cerebral blood flow. Hypertension may be associated
with an imbalance in the blood supply and a restriction of blood flow to specific
watershed regions of the cerebral cortex that likely promotes the accumulation of
WMH (Jennings et al., 2012). Indeed, reductions in blood flow have been documented
within areas of WMH lesions suggesting a vascular mechanism underlying this white
matter pathology (Brickman et al., 2009; Nobili et al., 1993). Decreases in cerebral
vasoreactivity, or the ability of blood vessels to adequately dilate to meet the meta-
bolic demands of the hemodynamic response, (Austin et al., 2011; Zhang, Huang, and
Shi, 2011) may also contribute to white matter decline, as decreased vasoreactivity
underlies hypoperfusion (Mandell et al., 2008). A recent study linked hypertension
with decreased vasoreactivity, particularly in frontal and parietal lobes (Hajjar et al.,
2010), regions that show considerable WMH expansion in hypertensive individuals.
Of course, further research directly examining cerebral perfusion, vasoreactivity and
white matter integrity are needed to support this supposition.
Hypertension Effects on Cognition via Structural Brain Alterations

Links between hypertension and brain structure and hypertension and cognition have
been separately studied to great extent. Studies incorporating both brain and cognitive
measures are critical to examining the hypothesis that structural brain changes are
one potential mechanism through which hypertension may promote cognitive decline.
Existing studies of this nature are surprisingly sparse, but do show evidence that that
the impact of hypertension on cognition is partly exerted through effects on brain
structure.
Overall, longitudinal studies tracking changes in brain structure and cognitive
performance over time in hypertensive relative to normotensive adults reveal that
hypertension in older adulthood, and in mid-â•‰life particularly, is associated with dele-
terious downstream effects on both grey and white matter integrity, which in turn sig-
nificantly impact cognitive performance. For example, hypertension in middle-â•‰aged
â•‡ 397
adults has been associated with ten-â•‰year longitudinal grey matter atrophy, which
predicted declines in processing speed and global cognitive function. In the same
cohort, WMH expansion over time predicted both slower processing speed and
poorer memory in hypertensive participants (Swan, DeCarli, et al., 1998). There is
also evidence that elevated blood pressure itself rather than merely a diagnosis of
hypertension per se, exerts effects on cognition via alterations to brain structure.
Older adults (mean age = 82 years) with higher ambulatory blood pressure (24-hour
average measurement), showed a significant increase of WMH over two years, which
was associated with declines in executive function performance (Burgmans et al.,
2011). Further evidence for the link of hypertension to cognition via structural brain
changes comes from mediation analyses. Findings in older individuals (68 to 78
years) from the Aberdeen birth cohorts showed that the relationship of hypertension
and late-â•‰life general intelligence was significantly mediated by the accumulation of
WMH (Murray et al., 2012).
A diagnosis of hypertension is often coâ•‰morbid with other vascular and metabolic
disorders such as diabetes and more significant expressions of cerebrovascular disease
such as stroke and transient ischemic attack for example, even in normal aging studies.
Longitudinal aging studies that classify individuals into groups that remain free of vas-
cular risk and those who develop hypertension and other associated risk factors have
also illustrated the impact of vascular heath on brain structure and cognitive change.
For instance, significant progression of WMH over five â•‰years, as well as atrophy of the
fusiform gyrus, was observed in a vascular risk group, whereas the middle-â•‰aged and
older adults without such risk factors showed little change over time (Raz et al., 2007).
These brain changes were associated with longitudinal declines in working memory.
In sum, the findings of the few studies that directly assessed the relationship of
brain aging and cognition among hypertensive individuals relative to controls sug-
gests that changes in both grey and white matter beyond the extent of decline expected
in normal brain aging are apparent and contribute to accelerated cognitive perfor-
mance decrements in middle-â•‰aged and older adults.
Hypertension and Brain Function
Resting Cerebral Blood Flow

In addition to impacting the brain at a structural level, hypertension has also been
linked to differences in resting cerebral blood flow (CBF). The human brain requires
high metabolic maintenance, as it utilizes approximately 20% of all cardiac output
(Edvinsson and Krause, 2002). Steady and sufficient blood flow supply is critical for
the delivery of oxygen and nutrients to support healthy brain function. Interruption
of this process may lead to chronic suppression of blood flow (i.e., hypoperfusion),
which could result in ischemia, neuronal dysfunction, and eventually degeneration
(Iadecola, 2004). A large body of literature dating back to the 1950s shows that aging
is accompanied by a decrease in resting CBF (Kety, 1956). However, individuals with
irregular blood pressure and/â•‰or a diagnosis of hypertension may be additionally sus-
ceptible to blood flow changes with aging.
398
On the whole, evidence suggests that hypertension is associated with a modest

decrease in resting CBF (Jennings and Heim, 2012; Jennings et al., 2005). A recent
study (Dai et al., 2008) utilizing arterial spin-labeled MRI showed that a small group
of cognitively normal hypertensive subjects showed relatively lower regional CBF in
select cortical regions including left inferior, superior and orbital frontal gyri, inferior
parietal, left superior temporal cortex, and left anterior cingulate gyrus. CBF decreases
were additionally observed in subcortical regions including left hippocampal forma-
tion and paleostriatum. This pattern of decreased blood flow associated with hyper-
tension is in accord with results from an earlier PET study (Fujishima et al., 1995).
Longitudinal studies examining changes in regional CBF within individuals over
time confirmed that hypertensive middle-aged and older adults demonstrate greater
declines than normotensive adults (Beason-Held et al., 2007; Muller et al., 2012).
Hypertension-related CBF decreases were observed in the prefrontal cortex, anterior
cingulate, and occipital areas over a six-year period, and longer duration of hyperten-
sion was linked to the frontal decreases in CBF (Beason-Held et al., 2007). Declines
in CBF have also been correlated (Beason-Held et al., 2012) with global assessment
of vascular risk as measured by the Framingham Cardiovascular Risk Profile (FCRP).
Higher FCRP at baseline was associated with accelerated longitudinal decline in CBF
in the orbitofrontal, medial frontal/anterior cingulate, insular, precuneus, and brain
stem regions. However, among the various factors that contribute to the FCRP, higher
diastolic blood pressure and diabetes each emerged independently as associated with
larger decreases in CBF in the frontal and insular areas.
An important methodological issue in the literature is disentangling effects that are
attributable to uncontrolled elevations in blood pressure vs. medically treated hyper-
tension. Pre-post test antihypertensive treatment study designs, as well as careful clas-
sification of hypertensive groups (with blood pressure measurements), help clarify
these issues. Interestingly, a single photon emission computed tomography (SPECT)
study (Efimova et al., 2008) showed that anti-hypertensive treatment over a six-month
period resulted in a 7%–11% increase in total CBF. This increase was accompanied
by modest improvements in some aspects of cognitive functioning, such as atten-
tion and psychomotor speed. A recent longitudinal study examined the association
of blood pressure and anti-hypertensive medications on changes in CBF over four
years in 575 middle-aged patients with atherosclerotic disease (Muller et al., 2012).
Their results clearly demonstrated that untreated and poorly controlled hypertension
predicted greater decline in CBF compared to both controls and successfully medi-
cated hypertensive individuals, who showed little to no decrease over the interval (see
Figure 15.1).
A dose-response effect has been documented where increasing systolic and dia-
stolic blood pressure predicted greater declines in blood flow. Supporting this idea
of an incremental and graded effect of blood pressure on cerebral perfusion, a recent
intervention study (Tryambake et al., 2013) showed that intensive antihypertensive
treatment which reduced blood pressure to the prehypertension range of <130/80 mm
Hg in older adults was associated with increases in total grey matter CBF, compared to
no reliable increase in the more moderate anti-hypertensive group reduced to a target
blood pressure at or slightly beneath the criteria for stage 1 hypertension (i.e., <140/
85 mm Hg). However, on the flip side, it must also be noted that a hypotensive state,
399
(A) N = 101 200 274 (B) N = 83 371 121 (C) N = 106 186 166 117
1.0 1.0 1.0
0.0 0.0 0.0

<120 120–140 >140 <70 70–90 >90 no controlled poorly untreated
controlled
Change in pCBF (mL/min/100mL)

–1.0 –1.0 –1.0
–2.0 –2.0 –2.0
–3.0 –3.0 –3.0
* *
**
–4.0 –4.0 –4.0
**
* **
** *
–5.0 –5.0 –5.0
Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Hypertension
Figure 15.1 Association of baseline systolic blood pressure (A), diastolic blood pressure (B), and categories of hypertension (C) with change
in parenchymal cerebral blood flow (pCBF) in the total population (N = 575). Bars represent change (standard error) in pCBF after 3.9 years
of follow-up, adjusted for age, sex, follow-up time, and vascular risk factors. *p < 0.05, **p < 0.01. Adapted from Muller et al., (2012), Annals
of Neurology, with permission.
400
subsequent to medical treatment, could emerge and increase the risk for cerebrovas-
cular disease. Decreases in arterial pressure associated with blood pressure lowering
can result in a susceptibility of brain tissue to ischemia, increasing the risk for white
matter disease and brain atrophy (Cruickshank, Thorp, and Zacharias, 1987; Fagan,
Payne, and Houtekier, 1989).
Overall, carrying a diagnosis of hypertension is consistently associated with at
least a mild reduction in CBF, with most studies overlapping to report effects in pre-
frontal brain regions, although decreases in several posterior brain regions have been
less consistently reported. A dose-â•‰response relationship has been illustrated where
longer duration of the disease may be associated with greater CBF decline. Finally,
preliminary evidence suggests that sufficient blood pressure lowering may be effec-
tive in preventing or even reversing these declines in CBF.
Hypertension and Functional Brain Activation

There is a paucity of literature focused on the effects of hypertension on functional
brain activation in response to a stimulus or cognitive task. However, the studies that
do exist reveal differences in the pattern of activation between individuals with and
without hypertension. Both increases and reductions in task relevant brain activation
have been reported in hypertensive groups, as well as utilization of additional brain
regions not observed in normotensive participants during cognitive processing.
An early study quantifying relative cerebral blood flow with [15O] water positron
emission tomography (PET) in response to a verbal working memory task demon-
strated in an uncontrolled hypertensive group (nine participants, age 60–â•‰67 years old)
a lower amplitude of blood flow response in some cortical regions, but also utiliza-
tion of a greater number of cortical areas compared to control participants (Jennings
et al., 1998). Specifically, the hypertension group showed decreased CBF response in
the right hemisphere of dorsolateral prefrontal and posterior parietal areas on a verbal
working memory span task with two levels of difficulty, in combination with addi-
tional activation observed in the left medial temporal lobe. A follow-â•‰up PET activation
study with 37 hypertensive and 59 normotensive older adults of similar age, confirmed
reduced blood flow in the posterior parietal area during verbal working memory per-
formance in the hypertensive group (Jennings et al., 2005). Interestingly, the findings
also revealed increased hippocampal/â•‰amygdala area activation only in the hypertensive
adults who performed well on the task, whereas hypertensive subjects who performed
poorly showed less prefrontal activation. Blood flow in that medial temporal area was
significantly correlated with the task-â•‰related activation in the prefrontal cortex. Thus,
there is preliminary evidence that hypertension may be associated with a potentially
compensatory CBF response that is linked to better working memory performance.
The interpretation of compensatory brain activity in response to cognitive chal-
lenge is a central issue in the cognitive neuroscience of aging literature, with many
studies showing evidence of increased brain activation. This phenomenon has been
interpreted both as working to support cognitive performance in older adults (Cabeza,
2002; Grady, 2008; Reuter-â•‰Lorenz and Cappell, 2008) and also as evidence to the
contrary showing that increased activation is a result of inefficient neural processes
with aging that are not facilitative of cognitive performance (Rypma, Berger, and
401
D’esposito, 2002). It is plausible that hypertension is one (of likely several) individ-
ual difference factors that contribute to the need for additional neural and/or vascu-
lar resources to support cognitive performance in older age. Much more prospective
research focused on the modifying role of blood pressure in the relationship between
functional brain activity and cognitive performance is needed to clarify this issue.
Only two fMRI studies focused on the effects of blood pressure related differ-
ences in functional activation patterns to cognitive task have been published to date,
with both studies utilizing a multiple vascular risk approach. One study of 30 cog-
nitively healthy adults (aged 42–77) without vascular disease, examined the effects
of systolic blood pressure, body mass index (BMI) and cholesterol levels on acti-
vation during a verbal associative memory task (Braskie, Small, and Bookheimer,
2010). Higher systolic pressure and BMI, but not cholesterol, were correlated with
increased activation in frontal, temporal and parietal lobes, including posterior cin-
gulate. This association was not dependent on a clinical diagnosis of hypertension as
the effect was observed within the recommended ranges of blood pressure and BMI.
Another fMRI study (Chuang et al., 2014) examining cardiovascular risks and brain
activity to an executive function task reported a significant association between
increased FCRP and greater task-relevant activation in the left inferior parietal
(see Figure 15.2 below). Importantly, this increased activation was associated with
(A) SMALL CUE LARGE CUE
CONGRUENT
INCONGRUENT
(B) A IncLg > ConLg B 2.3 3.2
R Z = 22 mm X = –52 mm
Figure 15.2 (A) The flanker task used to measure attention and inhibition. Incongruent flanker
symbols are more difficult to process than congruent. (B) Increased activation to incongruent
trials vs. congruent trials in the left middle temporal, left inferior parietal, and left supramarginal
gyri was seen as vascular risk score increased. (C) Scatterplot of the association of increased
activation with increased vascular risk score from the inferior parietal cluster. Modified from
Chuang et al. (2014), Neurobology of Aging, with permission.
402
Figure 15.2 Continued
â•‡ 403
slower reaction time on the flanker attention task, suggesting that it stemmed from
less efficient neural function. Thus, preliminary fMRI evidence suggests that subtle
elevations in blood pressure and vascular risk may be correlated with increased acti-
vation to a cognitive task.
It is difficult to reconcile the earlier PET activation findings with the more recent
fMRI studies as the population studied differed significantly, diverse cognitive par-
adigms were used, and the approach to quantifying vascular risk and estimating its
effects varied. However, the initial evidence suggests that vascular risk does appear to
be associated with differences in brain activity in both task relevant regions and in addi-
tional brain areas. It is premature to draw conclusions about the role of hypertension
specifically (versus more generalized vascular risk) from the pattern of the few studies
published thus far. Future research with sufficiently large sample sizes is needed to
carefully quantify and isolate specific vascular risks to alterations in neural activation
patterns. Linking the differences in activation (whether increases or decreases) to cog-
nitive performance is critical in understanding the nature of brain activation changes
in an aging context. An important methodological issue for future work to address is
disentangling neuronal and vascular function tracked in fMRI, as activation differences
linked to hypertension could be reflective of vascular system dysfunction, rather than
neuronal differences. A recent study (Liu et al., 2013) using a CO2-â•‰inhalation task to
measure cerebrovascular reactivity (CVR), a proxy for vessels’ ability to expand and
contract, showed differences in activation patterns in a memory encoding task after
the BOLD signal was adjusted for age-â•‰related decreases in CVR. The CVR-â•‰adjusted
results revealed additional increases in bilateral inferior frontal gyri. It remains unclear
if or to what extent age-â•‰related differences in cerebrovascular function measured in
the BOLD signal are affected by individual differences in vascular risk factors such as
hypertension.
Role of Hypertension in Pathological Aging
Thus far, this review discussed the effects of hypertension on non-â•‰pathological cog-
nitive and brain aging. However, a large body of epidemiological literature has linked
risk factors for cardiovascular disease, such as diabetes and hypertension, to increased
risk of pathological aging including Alzheimer’s disease and related dementias
(Breteler, 2000; Decarli, 2004; Kivipelto et al., 2002; Launer et al., 1995; Shah et al.,
2012). Late-â•‰onset AD is a neurodegenerative disorder characterized by progressive
cognitive impairment and memory loss. The defining neuropathology of AD, neuro-
fibrillary tangles composed of abnormally phosphorylated tau and plaques composed
of beta-â•‰amyloid protein (Aβ), precedes clinical impairment by 10–â•‰20 years (Reiman
et al., 2012). Initial triggers of the disease and the precise pathological processes that
transpire across these decades are unclear. However, vascular diseases such as hyper-
tension have been identified as risk factors for progression to AD. For example, a
large, population based, case-â•‰control study of AD patients identified solely hyperten-
sion, among 20 examined factors, as an independent risk factor for AD (Kokmen et
al., 1991).
404
The high level of comorbidities often observed among vascular-related neuro-

degenerative diseases such as AD and vascular dementia complicates understand-
ing the unique effects of single risk factors on pathological brain aging. However,
a recent review (Barnes and Yaffe, 2011) that estimated the approximate impact
of individual risk factor reduction on AD prevalence by calculating population
attributable risks, estimated that up to half of AD cases might be attributable to
the combined effect of seven modifiable risk factors examined. Mid-life hyper-
tension emerged as the third strongest of these vascular risk factors (after ciga-
rette smoking and physical inactivity). The authors estimated that a 25% reduction
in hypertension cases worldwide could result in 400,000 fewer cases of AD. In
fact, the most robust finding observed across multiple large-scale longitudinal
studies is that carrying a diagnosis of hypertension specifically during middle age
is associated with the development of dementia later in the lifespan (Kennelly,
Lawlor, and Kenny, 2009; Kilander et al., 1998; Kivipelto et al., 2001; Launer et
al., 2000; Polidori, Pientka, and Mecocci, 2012; Purnell et al., 2009). However, the
relationship between hypertension and dementia is complex as once an individ-
ual progresses to AD elevated blood pressure is less predictive of symptomatol-
ogy (Feldstein, 2012), and as AD progresses to later stages, lower blood pressure
is sometimes associated with declines in cognitive performance (Moretti et al.,
2008). It is unclear what effects are attributable to years of carrying the disease
itself versus the effects of peripheral blood pressure fluctuations and variability in
treatment effectiveness. However, the links between hypertension diagnosis, blood
pressure, and dementia are correlational and bidirectional (Jennings and Zanstra,
2009), and there are several mechanisms which could explain the association.
These underlying mechanisms of the increased risks for the development for
Alzheimer’s dementia in persons with hypertension are incompletely understood.
However, work from animal models of hypertension, as well as more recent in vivo
neuroimaging studies examining biomarkers for AD, offer clues about this associa-
tion. Results from rodent models of induced hypertension suggest that elevated blood
pressure may lead to increased amyloid-beta peptides (Diaz-Ruiz et al., 2009), an
early event in the neuropathological cascade of AD. Chronic hypertension appears to
increase blood-brain barrier permeability, which may lead to Aβ deposition (Gentile et
al., 2009). Additionally, hypoperfusion and neuroinflammation in the cortex and hip-
pocampus have been shown to precede the development of beta-amyloid deposition
in a hypertensive mouse model (Carnevale et al., 2012). Indeed, it has been suggested
that cerebral hypoperfusion, a consequence of hypertension (Daulatzai, 2012) and a
risk factor for both AD (Austin et al., 2011) and vascular dementia (Moretti et al.,
2008), is linked to oxidative stress, which may trigger Aβ plaque formation (Dickstein
et al., 2010; Lanari et al., 2007).
Recent human data from the Honolulu Asia Aging study (Shah et al., 2012)
reported that mid-life diastolic blood pressure modulated the link between lower
levels of circulating plasma Aβ, which typically begins decreasing about 15 years
before AD onset and is reflective of increases in beta-amyloid plaque in the brain,
and a subsequent diagnosis of probable AD. Low plasma Aβ was also associated with
cerebral amyloid angiopathy (CAA), and the authors suggest that elevated BP may
405
compromise vascular integrity of the brain, leading to CAA and impaired Aβ clear-
ance. Hypertension has been additionally linked to increased beta-amyloid burden as
measured in vivo. One study using the FCRP to examine the influence of combined
vascular risk factors (Reed et al., 2012), found greater aggregate vascular risk (but not
individual risk factors) was associated with greater amyloid burden. A more recent
beta-amyloid PET imaging study (Rodrigue et al., 2013) focused on amyloid depo-
sition in a normal aging sample, reported that cognitively normal older adults with
both a genetic risk for AD (heterozygosity or homozygosity for apolipoprotein ε4)
and elevated blood pressure showed elevated Aβ deposition. Neither the normotensive
group with APOE ε4, nor the hypertensive group without APOE ε4 showed elevated
beta-amyloid in comparison to a group with neither a genetic risk nor hypertension.
Specifically, the individuals with both untreated elevations in blood pressure and
APOE ε4 showed the greatest beta-amyloid burden (see Figure 15.3 below), suggest-
ing that blood pressure may work synergistically with genetic risk factors to impact
the accumulation of AD biomarkers in preclinical aging, as well as to affect cognitive
performance (Caselli et al., 2011).
It remains to be elucidated if the hypertension-related accumulation of AD bio-
markers in normal aging samples is ultimately reflective of impending pathological
decline. The precise nature of the relationship between vascular risk factors and the
subsequent progression to disease requires further longitudinal study to disentangle,
which will require studies of normal aging that carefully measure cognitive perfor-
mance and key brain changes over time, while tracking the development of known
risk factors such as hypertension and estimating their influence on neurocognitive
decline.
*
Normotensive
Mean cortical amyloid (SUVR)
1.4 Medicated hypertensive

Unmedicated hypertensive
1.3
1.2
1.1
1.0
APOE ε4– APOE ε4+
Figure 15.3 Mean β-amyloid deposition by genetic risk and hypertensive subgroup. Individuals
with uncontrolled hypertension and at least 1 apolipoprotein E (APOE) ϵ4 allele show greater
β-amyloid deposition. SUVR indicates standardized uptake value ratio. *Significant interaction
(P = .02) between vascular risk group and genetic status, with unmedicated ϵ4+ participants with
hypertension showing the greatest amounts of β-amyloid deposition. Adapted from Rodrigue,
et al., (2013), JAMA Neurology, with permission.
406
Conclusion
In sum, we have presented an overview of the evidence showing that a diagnosis of

hypertension is linked to poorer cognitive performance in otherwise healthy samples
of cognitive aging. The cognitive domains most likely to be affected are the areas of
cognition most susceptible to cognitive aging effects, such as memory ability, rea-
soning, and executive function. Hypertension-â•‰related changes to brain structure may
in part underlie the relationship between blood pressure and cognitive performance.
Age-â•‰related shrinkage of brain structures vulnerable to aging, such as the frontal lobes
and hippocampus, appear to be additionally sensitive to hypertension. Across a vari-
ety of cognitive and brain modalities there appears to be evidence that vascular risk
factors such as hypertension may accelerate the aging process and may be associated
with a spread in the pattern of structural brain aging, resulting in changes to more
posterior regions of the brain which are less affected in samples of healthy aging
without vascular risk factors. Hypertension-â•‰related decreases in cerebral perfusion
are frequently reported and may also partly explain the detrimental effects of hyper-
tension on cognitive performance. More recently, there is accumulating evidence that
hypertension is associated with alterations in brain activation patterns in response to
cognitive challenge. Although additional work is needed, hypertension may represent
one individual difference factor that underlies the observed age-â•‰related shift in func-
tional activation patterns in fMRI studies of aging.
Overall there is evidence that medically controlling blood pressure offers a significant
degree of protection to brain and cognitive decline compared to aging effects observed in
uncontrolled hypertension. However, not all intervention studies attempting to use anti-
hypertensive treatment to reverse or slow brain and cognitive aging have been effective
(Jennings and Zanstra, 2009). A key issue to address in future treatment studies is the
selection of dependent variables that are not too far downstream from the earlier changes
associated with blood pressure increases, such as CBF as compared to measuring changes
in structural brain volume, which may occur further into the disease process.
The demonstrated effects of hypertension on brain and cognition in even highly
screened samples of healthy aging has important implication for the larger cognitive
neuroscience of aging literature, which for the most part does not routinely report
percentage of hypertension or other vascular risks in study demographics. On the one
hand, given the high prevalence of hypertension in aging populations it can be argued
that carrying a diagnosis of hypertension is indeed reflective of the “normal” aging
process. Yet, a failure to control for or measure variance due to variability in blood
pressure may result in an overestimation of aging effects or result in a different pattern
of results altogether compared to assessments of brain and cognitive aging without
inclusion of vascular risk factors.
Finally, hypertension and related vascular factors are linked to increased risk for
Alzheimer’s disease and related dementias. The effect may be especially pronounced
in individuals who carry a genetic risk for AD. It is currently unclear if chronic eleva-
tions in blood pressure play a causative role in the development of pathological aging.
However, given that hypertension is a significantly modifiable risk factor, whether
through pharmacological or lifestyle interventions, increased awareness of the risks
â•‡ 407
of hypertension for neurocognitive aging and more aggressive and earlier treatment
of the disease may bring about a critical reduction in pathological brain aging and
cognitive decline.
Acknowledgmentsâ•‡
This work was supported in part by grant K99/â•‰R00-â•‰AG-â•‰36848-â•‰05 to KR. The authors
thank Dr. Kristen M. Kennedy for helpful comments on an earlier version of the
manuscript.
References
Al-â•‰Sarraf H, and Philip L (2003) Effect of hypertension on the integrity of blood brain and blood
CSF barriers, cerebral blood flow and CSF secretion in the rat. Brain Res 975:179–â•‰188.
Artero S, Tiemeier H, Prins ND, Sabatier R, Breteler MM, and Ritchie K (2004) Neuroanatomical
localisation and clinical correlates of white matter lesions in the elderly. J Neurol Neurosurg
Psychiatry 75:1304–â•‰1308.
Austin BP, Nair VA, Meier TB, Xu G, Rowley HA, Carlsson CM, et al. (2011) Effects of
hypoperfusion in Alzheimer’s disease. J Alzheimers Dis, 26 Suppl 3:123–â•‰133.
Baltes PB, Staudinger UM, and Lindenberger U (1999) Lifespan psychology: Theory and
application to intellectual functioning. Annual Review of Psychology 50:471–â•‰507.
Bangen KJ, Nation DA, Clark LR, Harmell AL, Wierenga CE, Dev SI, et al. (2014) Interactive
effects of vascular risk burden and advanced age on cerebral blood flow. Front Aging
Neurosci 6:159.
Barnes DE, and Yaffe K (2011) The projected effect of risk factor reduction on Alzheimer’s
disease prevalence. Lancet Neurol 10:819–â•‰828.
Beason-â•‰Held LL, Moghekar A, Zonderman AB, Kraut MA, and Resnick SM (2007) Longitudinal
changes in cerebral blood flow in the older hypertensive brain. Stroke 38:1766–â•‰1773.
Beason-â•‰Held LL, Thambisetty M, Deib G, Sojkova J, Landman BA, Zonderman AB, et al.
(2012) Baseline cardiovascular risk predicts subsequent changes in resting brain function.
Stroke 43:1542–â•‰1547.
Beauchet O, Celle S, Roche F, Bartha R, Montero-â•‰Odasso M, Allali G, and Annweiler C (2013)
Blood pressure levels and brain volume reduction: a systematic review and meta-â•‰analysis.
J Hypertens 31:1502–â•‰1516.
Boegehold MA (2007) Vascular remodeling and rarefaction in hypertension. In G. Y. H. Lip and J.
E. Hall (Eds.), Comprehensive Hypertension (pp. 157–â•‰166). Philadelphia, PA: Mosby/â•‰Elsevier.
Brady CB, Spiro A, and Gaziano JM (2005) Effects of age and hypertension status on cogni-
tion: the Veterans Affairs Normative Aging Study. Neuropsychology 19:770–â•‰777.
Braskie MN, Small GW, and Bookheimer SY (2010) Vascular health risks and fMRI activation
during a memory task in older adults. Neurobiol Aging 31:1532–â•‰1542.
Breteler MM (2000) Vascular risk factors for Alzheimer’s disease: an epidemiologic perspec-
tive. Neurobiol Aging 21:153–â•‰160.
Brickman AM, Zahra A, Muraskin J, Steffener J, Holland CM, Habeck C, et al. (2009)
Reduction in cerebral blood flow in areas appearing as white matter hyperintensities on
magnetic resonance imaging. Psychiatry Res 172:117–â•‰120.
Bucur B, and Madden DJ (2010) Effects of adult age and blood pressure on executive function
and speed of processing. Exp Aging Res 36:153–â•‰168.
408
Burgmans S, Gronenschild EH, Fandakova Y, Shing YL, Van Boxtel MP, Vuurman EF, et al.
(2011) Age differences in speed of processing are partially mediated by differences in
axonal integrity. NeuroImage 55:1287–1297.
Burgmans, S., Van Boxtel, M. P., Gronenschild, E. H., Vuurman, E. F., Hofman, P. H., Uylings,
H. B., … Raz, N. (2010). Multiple indicators of age-related differences in cerebral white
matter and the modifying effects of hypertension. NeuroImage, 49, 2083–2093.
Carnevale D, Mascio G, Ajmone-Cat MA, D’Andrea I, Cifelli G, Madonna M, et al. (2012)
Role of neuroinflammation in hypertension-induced brain amyloid pathology. Neurobiol
Aging 33:205.e219–e229.
Caselli RJ, Dueck AC, Locke DE, Sabbagh MN, Ahern GL, Rapcsak SZ, et al. (2011)
Cerebrovascular risk factors and preclinical memory decline in healthy APOE ε4 homozy-
gotes. Neurology, 76, 1078–1084.
Chen X, Wen W, Anstey KJ, and Sachdev PS (2006) Effects of cerebrovascular risk factors
on gray matter volume in adults aged 60–64 years: a voxel-based morphometric study.
Psychiatry Res 147:105–114.
Chuang YF, Eldreth D, Erickson KI, Varma V, Harris G, Fried LP, et al. (2014) Cardiovascular
risks and brain function: a functional magnetic resonance imaging study of executive func-
tion in older adults. Neurobiol Aging 35:1396–1403.
Cohen RA (2007) Hypertension and cerebral blood flow: implications for the development of
vascular cognitive impairment in the elderly. Stroke 38:1715–1717.
Cruickshank JM, Thorp JM, and Zacharias FJ (1987) Benefits and potential harm of lowering
high blood pressure. Lancet 1:581–584.
Dai W, Lopez OL, Carmichael OT, Becker JT, Kuller LH, and Gach HM (2008) Abnormal
regional cerebral blood flow in cognitively normal elderly subjects with hypertension.
Stroke 39:349–354.
Daulatzai MA (2012) Quintessential risk factors: their role in promoting cognitive dysfunction
and Alzheimer’s disease. Neurochem Res 37:2627–2658.
de Groot M, Ikram MA, Akoudad S, Krestin GP, Hofman A, van der Lugt A, et al. (2015). Tract-
specific white matter degeneration in aging. The Rotterdam Study. Alzheimers Dement
11, 321–330.
De Leeuw FE, De Groot JC, and Breteler MMB (2001) White matter changes: Frequency and
risk factors. In L. Pantoni, D. Inzitari & A. Wallin (Eds.), The Matter of White Matter:
Clinical and Pathophysiological Aspects of White Matter Disease Related to Cognitive
Decline and Vascular Dementia (pp. 19– 33). Utrecht, The Netherlands: Academic
Pharmaceutical Productions.
Decarli, C. (2004). Vascular factors in dementia: an overview. J Neurol Sci, 226, 19–23.
Decarli C, Miller BL, Swan GE, Reed T, Wolf PA, Garner J, et al. (1999) Predictors of brain
morphology for the men of the NHLBI twin study. Stroke 30:529–536.
Den Heijer T, Launer LJ, Prins ND, Van Dijk EJ, Vermeer SE, Hofman A, et al. (2005)
Association between blood pressure, white matter lesions, and atrophy of the medial tem-
poral lobe. Neurology 64:263–267.
Desmond DW, Tatemichi TK, Paik M, and Stern Y (1993) Risk factors for cerebrovascu-
lar disease as correlates of cognitive function in a stroke-free cohort. Arch Neurol
50(2):162–166.
Diaz-Ruiz C, Wang J, Ksiezak-Reding H, Ho L, Qian X, Humala N, et al. (2009) Role of
Hypertension in Aggravating Aβ Neuropathology of AD Type and Tau-Mediated Motor
Impairment. Cardiovasc Psychiatry Neurol: 107286.
409
Dickstein DL, Walsh J, Brautigam H, Stockton SD, Gandy S, and Hof PR (2010) Role of
vascular risk factors and vascular dysfunction in Alzheimer’s disease. Mt Sinai J Med
77:82–102.
Dufouil C, De Kersaint-Gilly A, Besançon V, Levy C, Auffray E, Brunnereau L, et al. (2001)
Longitudinal study of blood pressure and white matter hyperintensities: the EVA MRI
Cohort. Neurology 56:921–926.
Edvinsson L, and Krause DN (2002) Cerebral Blood Flow and Metabolism (2, illustrated ed.):
Lippincott Williams & Wilkins.
Efimova IY, Efimova NY, Triss SV, and Lishmanov YB (2008) Brain perfusion and cognitive
function changes in hypertensive patients. Hypertens Res 31:673–678.
Elias MF, Dore GA, Davey A, Robbins MA, and Elias PK (2010) From blood pressure to phys-
ical disability: the role of cognition. Hypertension 55:1360–1365.
Elias MF, Elias PK, Sullivan LM, Wolf PA, and D’agostino RB (2003) Lower cognitive func-
tion in the presence of obesity and hypertension: the Framingham heart study. Int J Obes
Relat Metab Disord 27:260–268.
Elias MF, Wolf PA, D’agostino RB, Cobb J, and White LR (1993) Untreated blood pressure
level is inversely related to cognitive functioning: the Framingham Study. Am J Epidemiol
138:353–364.
Elias PK, Elias MF, Robbins MA, and Budge MM (2004) Blood pressure-related cognitive
decline: does age make a difference? Hypertension 44:631–636.
Fagan SC, Payne LW, and Houtekier SC (1989) Risk of cerebral hypoperfusion with antihyper-
tensive therapy. DICP 23:957–962.
Feldstein CA (2012) Association between chronic blood pressure changes and development of
Alzheimer’s disease. J Alzheimers Dis 32:753–763.
Folkow B, Aurell M, Friberg P, Göthberg G, Hallbäck-Nordlander M, Karlström G, et al. (1984)
Cardiovascular studies in rats with respect to some functional and structural relationships
of relevance in hypertension and ordinary aging. Clin Exp Hypertens A 6:587–598.
Fujishima M, Ibayashi S, Fujii K, and Mori S (1995) Cerebral blood flow and brain function in
hypertension. Hypertens Res 18:111–117.
Fukuda H, and Kitani M (1995) Differences between treated and untreated hypertensive sub-
jects in the extent of periventricular hyperintensities observed on brain MRI. Stroke
26:1593–1597.
Gentile MT, Poulet R, Di Pardo A, Cifelli G, Maffei A, Vecchione C, et al. (2009) Beta-amyloid
deposition in brain is enhanced in mouse models of arterial hypertension. Neurobiol Aging
30:222–228.
Gianaros, P. J., Greer, P. J., Ryan, C. M., & Jennings, J. R. (2006). Higher blood pressure
predicts lower regional grey matter volume: Consequences on short-term information pro-
cessing. NeuroImage, 31, 754–765.
Gifford, K. A., Badaracco, M., Liu, D., Tripodis, Y., Gentile, A., Lu, Z., … Jefferson, A. L.
(2013). Blood pressure and cognition among older adults: a meta-analysis. Arch Clin
Neuropsychol, 28, 649–664.
Glynn, R. J., Beckett, L. A., Hebert, L. E., Morris, M. C., Scherr, P. A., & Evans, D. A. (1999).
Current and remote blood pressure and cognitive decline. JAMA, 281, 438–445.
Go, A. S., Mozaffarian, D., Roger, V. L., Benjamin, E. J., Berry, J. D., Borden, W. B., … Turner,
M. B. (2013). Executive summary: heart disease and stroke statistics--2 013 update: a
report from the American Heart Association. Circulation, 127, 143–152.
Grady, C. L. (2008). Cognitive neuroscience of aging. Ann N Y Acad Sci, 1124, 127–144.
Gunning-Dixon, F. M., Brickman, A. M., Cheng, J. C., & Alexopoulos, G. S. (2009). Aging of
cerebral white matter: a review of MRI findings. Int J Geriatr Psychiatry, 24, 109–117.
410
Guo, Z., Fratiglioni, L., Winblad, B., & Viitanen, M. (1997). Blood pressure and performance
on the Mini-Mental State Examination in the very old. Cross-sectional and longitudinal
data from the Kungsholmen Project. Am J Epidemiol, 145, 1106–1113.
Hachinski, V. C., Potter, P., & Merskey, H. (1987). Leuko-araiosis. Arch Neurol, 44, 21–23.
Hajjar, I., Zhao, P., Alsop, D., Abduljalil, A., Selim, M., Novak, P., & Novak, V. (2010).
Association of blood pressure elevation and nocturnal dipping with brain atrophy, per-
fusion and functional measures in stroke and nonstroke individuals. Am J Hypertens,
23, 17–23.
Hannesdottir, K., Nitkunan, A., Charlton, R. A., Barrick, T. R., MacGregor, G. A., & Markus,
H. S. (2009). Cognitive impairment and white matter damage in hypertension: a pilot
study. Acta Neurol Scand, 119, 261–268.
Harrington, F., Saxby, B. K., Mckeith, I. G., Wesnes, K., & Ford, G. A. (2000). Cognitive perfor-
mance in hypertensive and normotensive older subjects. Hypertension, 36(6), 1079–1082.
Hatazawa, J., Yamaguchi, T., Ito, M., Yamaura, H., & Matsuzawa, T. (1984). Association of
hypertension with increased atrophy of brain matter in the elderly. J Am Geriatr Soc, 32,
370–374.
Head, D., Raz, N., Gunning-Dixon, F., Williamson, A., & Acker, J. D. (2002). Age-related dif-
ferences in the course of cognitive skill acquisition: the role of regional cortical shrinkage
and cognitive resources. Psychol Aging, 17, 72–84.
Hebert, L. E., Scherr, P. A., Bennett, D. A., Bienias, J. L., Wilson, R. S., Morris, M. C., & Evans,
D. A. (2004). Blood pressure and late-life cognitive function change: a biracial longitudi-
nal population study. Neurology, 62, 2021–2024.
Iadecola, C. (2004). Neurovascular regulation in the normal brain and in Alzheimer’s disease.
Nat Rev Neurosci, 5, 347–360.
Jacobs, H. I., Leritz, E. C., Williams, V. J., Van Boxtel, M. P., Van Der Elst, W., Jolles, J., …
Salat, D. H. (2013). Association between white matter microstructure, executive functions,
and processing speed in older adults: the impact of vascular health. Hum Brain Mapp,
34, 77–95.
Jennings, J. R., & Heim, A. F. (2012). From Brain to Behavior: Hypertension’s Modulation of
Cognition and Affect. Int J Hypertens, 2012, 701385.
Jennings, J. R., Mendelson, D. N., Muldoon, M. F., Ryan, C. M., Gianaros, P. J., Raz, N., &
Aizenstein, H. (2012). Regional grey matter shrinks in hypertensive individuals despite
successful lowering of blood pressure. J Hum Hypertens, 26, 295–305.
Jennings, J. R., Muldoon, M. F., Ryan, C., Price, J. C., Greer, P., Sutton-Tyrrell, K., … Meltzer,
C. C. (2005). Reduced cerebral blood flow response and compensation among patients
with untreated hypertension. Neurology, 64, 1358–1365.
Jennings, J. R., Muldoon, M. F., Ryan, C. M., Mintun, M. A., Meltzer, C. C., Townsend, D. W.,
… Manuck, S. B. (1998). Cerebral blood flow in hypertensive patients: an initial report
of reduced and compensatory blood flow responses during performance of two cognitive
tasks. Hypertension, 31, 1216–1222.
Jennings, J. R., & Zanstra, Y. (2009). Is the brain the essential in hypertension? NeuroImage,
47, 914–921.
Kähönen-Väre, M., Brunni-Hakala, S., Lindroos, M., Pitkala, K., Strandberg, T., & Tilvis, R.
(2004). Left ventricular hypertrophy and blood pressure as predictors of cognitive decline
in old age. Aging Clin Exp Res, 16, 147–152.
Kennedy, K. M., & Raz, N. (2009). Pattern of normal age-related regional differences in white
matter microstructure is modified by vascular risk. Brain Res, 1297, 41–56.
Kennelly, S. P., Lawlor, B. A., & Kenny, R. A. (2009). Blood pressure and dementia -a compre-
hensive review. Ther Adv Neurol Disord, 2, 241–260.
Kety, S. S. (1956). Human cerebral blood flow and oxygen consumption as related to aging. Res
Publ Assoc Res Nerv Ment Dis, 35, 31–45.
411
Kilander, L., Nyman, H., Boberg, M., Hansson, L., & Lithell, H. (1998). Hypertension is related
to cognitive impairment: a 20-year follow-up of 999 men. Hypertension, 31, 780–786.
Kilander, L., Nyman, H., Boberg, M., & Lithell, H. (2000). The association between low dia-
stolic blood pressure in middle age and cognitive function in old age. A population-based
study. Age Ageing, 29, 243–248.
King, H. E. (1956). Comparison of fine psychomotor movement by hypertensive and hypoten-
sive subjects. Perceptual and Motor Skills, 6, 199–204.
Kivipelto, M., Helkala, E. L., Laakso, M. P., Hänninen, T., Hallikainen, M., Alhainen, K., …
Nissinen, A. (2001). Midlife vascular risk factors and Alzheimer’s disease in later life: lon-
gitudinal, population based study. BMJ, 322, 1447–1451.
Kivipelto, M., Laakso, M. P., Tuomilehto, J., Nissinen, A., & Soininen, H. (2002). Hypertension
and hypercholesterolaemia as risk factors for Alzheimer’s disease: potential for pharmaco-
logical intervention. CNS Drugs, 16, 435–444.
Knopman, D., Boland, L. L., Mosley, T., Howard, G., Liao, D., Szklo, M., … Investigators, A.
R. I. C. S. (2001). Cardiovascular risk factors and cognitive decline in middle-aged adults.
Neurology, 56, 42–48.
Knopman, D. S., Penman, A. D., Catellier, D. J., Coker, L. H., Shibata, D. K., Sharrett, A. R.,
& Mosley, T. H. (2011). Vascular risk factors and longitudinal changes on brain MRI: the
ARIC study. Neurology, 76, 1879–1885.
Kokmen, E., Beard, C. M., Chandra, V., Offord, K. P., Schoenberg, B. S., & Ballard, D. J.
(1991). Clinical risk factors for Alzheimer’s disease: a population-based case-control
study. Neurology, 41, 1393–1397.
Korf, E. S., White, L. R., Scheltens, P., & Launer, L. J. (2004). Midlife blood pressure and the
risk of hippocampal atrophy: the Honolulu Asia Aging Study. Hypertension, 44, 29–34.
Lanari, A., Silvestrelli, G., De Dominicis, P., Tomassoni, D., Amenta, F., & Parnetti, L. (2007).
Arterial hypertension and cognitive dysfunction in physiologic and pathologic aging of the
brain. Am J Geriatr Cardiol, 16, 158–164.
Launer, L. J., Masaki, K., Petrovitch, H., Foley, D., & Havlik, R. J. (1995). The association
between midlife blood pressure levels and late-life cognitive function. The Honolulu-Asia
Aging Study. JAMA, 274, 1846–1851.
Launer, L. J., Ross, G. W., Petrovitch, H., Masaki, K., Foley, D., White, L. R., & Havlik, R. J.
(2000). Midlife blood pressure and dementia: the Honolulu-Asia aging study. Neurobiol
Aging, 21, 49–55.
Leritz, E. C., Salat, D. H., Milberg, W. P., Williams, V. J., Chapman, C. E., Grande, L. J., …
McGlinchey, R. E. (2010). Variation in blood pressure is associated with white matter
microstructure but not cognition in African Americans. Neuropsychology, 24, 199–208.
Leritz, E. C., Salat, D. H., Williams, V. J., Schnyer, D. M., Rudolph, J. L., Lipsitz, L., …
Milberg, W. P. (2011). Thickness of the human cerebral cortex is associated with met-
rics of cerebrovascular health in a normative sample of community dwelling older adults.
NeuroImage, 54, 2659–2671.
Liu, P., Hebrank, A. C., Rodrigue, K. M., Kennedy, K. M., Section, J., Park, D. C., & Lu, H.
(2013). Age-related differences in memory-encoding fMRI responses after accounting for
decline in vascular reactivity. NeuroImage, 78, 415–425.
Longstreth, W. T., Manolio, T. A., Arnold, A., Burke, G. L., Bryan, N., Jungreis, C. A., …
Fried, L. (1996). Clinical correlates of white matter findings on cranial magnetic resonance
imaging of 3301 elderly people. The Cardiovascular Health Study. Stroke, 27, 1274–1282.
Madden, D. J., & Blumenthal, J. A. (1998). Interaction of hypertension and age in visual selec-
tive attention performance. Health Psychol, 17, 76–83.
Mandell, D. M., Han, J. S., Poublanc, J., Crawley, A. P., Kassner, A., Fisher, J. A., & Mikulis,
D. J. (2008). Selective reduction of blood flow to white matter during hypercapnia corre-
sponds with leukoaraiosis. Stroke, 39(7), 1993–1998.
412
Marcus, J., Gardener, H., Rundek, T., Elkind, M. S., Sacco, R. L., Decarli, C., & Wright, C.
B. (2011). Baseline and longitudinal increases in diastolic blood pressure are associated
with greater white matter hyperintensity volume: the northern Manhattan study. Stroke,
42, 2639–2641.
Moretti, R., Torre, P., Antonello, R. M., Manganaro, D., Vilotti, C., & Pizzolato, G. (2008).
Risk factors for vascular dementia: hypotension as a key point. Vasc Health Risk Manag,
4, 395–402.
Morris, M. C., Scherr, P. A., Hebert, L. E., Bennett, D. A., Wilson, R. S., Glynn, R. J., & Evans,
D. A. (2002). Association between blood pressure and cognitive function in a biracial
community population of older persons. Neuroepidemiology, 21, 123–130.
Muller, M., Sigurdsson, S., Kjartansson, O., Aspelund, T., Lopez, O. L., Jonnson, P. V., … Age,
G. N. E. S.-R. S. I. (2014). Joint effect of mid-and late-life blood pressure on the brain: the
AGES-Reykjavik study. Neurology, 82, 2187–2195.
Muller, M., Van Der Graaf, Y., Visseren, F. L., Mali, W. P., Geerlings, M. I., & Group, S. S.
(2012). Hypertension and longitudinal changes in cerebral blood flow: the SMART-MR
study. Ann Neurol, 71, 825–833.
Murray, A. D., Staff, R. T., McNeil, C. J., Salarirad, S., Starr, J. M., Deary, I. J., & Whalley, L.
J. (2012). Brain lesions, hypertension and cognitive ageing in the 1921 and 1936 Aberdeen
birth cohorts. Age (Dordr), 34, 451–459.
Nobili, F., Rodriguez, G., Marenco, S., De Carli, F., Gambaro, M., Castello, C., … Rosadini, G.
(1993). Regional cerebral blood flow in chronic hypertension. A correlative study. Stroke,
24, 1148–1153.
Nwankwo, T., Yoon, S. S., Burt, V., & Gu, Q. (2013). Hypertension among adults in the United
States: National Health and Nutrition Examination Survey, 2011–2012. NCHS Data
Brief, 1–8.
Pandav, R., Dodge, H. H., DeKosky, S. T., & Ganguli, M. (2003). Blood pressure and cognitive
impairment in India and the United States: a cross-national epidemiological study. Arch
Neurol, 60, 1123–1128.
Pantoni, L., & Garcia, J. H. (1997). Pathogenesis of leukoaraiosis: a review. Stroke, 28,
652–659.
Park, D. C., Lautenschlager, G., Hedden, T., Davidson, N. S., Smith, A. D., & Smith, P. K.
(2002). Models of visuospatial and verbal memory across the adult life span. Psychology
and Aging, 17(2), 299–320.
Polidori, M. C., Pientka, L., & Mecocci, P. (2012). A review of the major vascular risk factors
related to Alzheimer’s disease. J Alzheimers Dis, 32, 521–530.
Purnell, C., Gao, S., Callahan, C. M., & Hendrie, H. C. (2009). Cardiovascular risk factors and
incident Alzheimer disease: a systematic review of the literature. Alzheimer Dis Assoc
Disord, 23, 1–10.
Raz, N., & Kennedy, K. M. (2009). A systems approach to the aging brain: Neuroanatomic
changes, their modifiers, and cognitive correlates. In W. Jagust & M. D’esposito (Eds.),
Imaging the Aging Brain (pp. 43–70). New York, NY: Oxford University Press.
Raz, N., Lindenberger, U., Rodrigue, K. M., Kennedy, K. M., Head, D., Williamson, A., …
Acker, J. D. (2005). Regional brain changes in aging healthy adults: general trends, indi-
vidual differences and modifiers. Cereb Cortex, 15, 1676–1689.
Raz, N., Rodrigue, K. M., & Acker, J. D. (2003). Hypertension and the brain: vulnerability of
the prefrontal regions and executive functions. Behav Neurosci, 117, 1169–1180.
Raz, N., Rodrigue, K. M., Kennedy, K. M., & Acker, J. D. (2007). Vascular health and longitu-
dinal changes in brain and cognition in middle-aged and older adults. Neuropsychology,
21, 149–157.
413
Reed, B. R., Marchant, N. L., Jagust, W. J., Decarli, C. C., Mack, W., & Chui, H. C. (2012).
Coronary risk correlates with cerebral amyloid deposition. Neurobiol Aging, 33,
1979–1987.
Reijmer, Y. D., Van Den Berg, E., Dekker, J. M., Nijpels, G., Stehouwer, C. D., Kappelle, L. J.,
& Biessels, G. J. (2012). Development of vascular risk factors over 15 years in relation to
cognition: the Hoorn Study. J Am Geriatr Soc, 60, 1426–1433.
Reiman, E. M., Quiroz, Y. T., Fleisher, A. S., Chen, K., Velez-Pardo, C., Jimenez-Del-Rio,
M., … Lopera, F. (2012). Brain imaging and fluid biomarker analysis in young adults at
genetic risk for autosomal dominant Alzheimer’s disease in the presenilin 1 E280A kin-
dred: a case-control study. Lancet Neurol, 11, 1048–1056.
Reuter-Lorenz, P. A., & Cappell, K. A. (2008). Neurocognitive Aging and the Compensation
Hypothesis. Current Directions in Psychological Science, 17, 177–182.
Richards, S. S., Emsley, C. L., Roberts, J., Murray, M. D., Hall, K., Gao, S., & Hendrie, H. C.
(2000). The association between vascular risk factor-mediating medications and cogni-
tion and dementia diagnosis in a community-based sample of African-Americans. J Am
Geriatr Soc, 48(9), 1035–1041.
Rodrigue, K. M., Rieck, J. R., Kennedy, K. M., Devous, M. D., Diaz-Arrastia, R., & Park, D.
C. (2013). Risk factors for β-amyloid deposition in healthy aging: vascular and genetic
effects. JAMA Neurol, 70, 600–606.
Rypma, B., Berger, J. S., & D’Esposito, M. (2002). The influence of working-memory demand
and subject performance on prefrontal cortical activity. J Cogn Neurosci, 14, 721–731.
Sabbatini, M., Strocchi, P., Vitaioli, L., & Amenta, F. (2000). The hippocampus in spontaneously
hypertensive rats: a quantitative microanatomical study. Neuroscience, 100, 251–258.
Salat, D. H., Williams, V. J., Leritz, E. C., Schnyer, D. M., Rudolph, J. L., Lipsitz, L. A., …
Milberg, W. P. (2012). Inter-individual variation in blood pressure is associated with
regional white matter integrity in generally healthy older adults. NeuroImage, 59, 181–192.
Salerno, J. A., Murphy, D. G., Horwitz, B., Decarli, C., Haxby, J. V., Rapoport, S. I., & Schapiro,
M. B. (1992). Brain atrophy in hypertension. A volumetric magnetic resonance imaging
study. Hypertension, 20, 340–348.
Scherr, P. A., Hebert, L. E., Smith, L. A., & Evans, D. A. (1991). Relation of blood pressure to
cognitive function in the elderly. American Journal of Epidemiology, 134(11), 1303–1315.
Schreiber, S., Bueche, C. Z., Garz, C., & Braun, H. (2013). Blood brain barrier breakdown as
the starting point of cerebral small vessel disease? -New insights from a rat model. Exp
Transl Stroke Med, 5, 4.
Scullin, M. K., Gordon, B. A., Shelton, J. T., Lee, J. H., Head, D., & McDaniel, M. A. (2013).
Evidence for a detrimental relationship between hypertension history, prospective mem-
ory, and prefrontal cortex white matter in cognitively normal older adults. Cogn Affect
Behav Neurosci, 13, 405–416.
Shah, N. S., Vidal, J. S., Masaki, K., Petrovitch, H., Ross, G. W., Tilley, C., … Launer, L. J.
(2012). Midlife blood pressure, plasma β-amyloid, and the risk for Alzheimer disease: the
Honolulu Asia Aging Study. Hypertension, 59, 780–786.
Shing, Y. L., Rodrigue, K. M., Kennedy, K. M., Fandakova, Y., Bodammer, N., Werkle-Bergner,
M., … Raz, N. (2011). Hippocampal subfield volumes: age, vascular risk, and correlation
with associative memory. Front Aging Neurosci, 3, 2.
Stamler, J., Stamler, R., & Neaton, J. D. (1993). Blood pressure, systolic and diastolic, and
cardiovascular risks. US population data. Arch Intern Med, 153, 598–615.
Strassburger, T. L., Lee, H. C., Daly, E. M., Szczepanik, J., Krasuski, J. S., Mentis, M. J., …
Alexander, G. E. (1997). Interactive effects of age and hypertension on volumes of brain
structures. Stroke, 28, 1410–1417.
414
Swan, G. E., Carmelli, D., & Larue, A. (1998). Systolic blood pressure tracking over 25 to
30 years and cognitive performance in older adults. Stroke, 29, 2334–2340.
Swan, G. E., Decarli, C., Miller, B. L., Reed, T., Wolf, P. A., Jack, L. M., & Carmelli, D. (1998).
Association of midlife blood pressure to late-life cognitive decline and brain morphology.
Neurology, 51, 986–993.
Taki, Y., Goto, R., Evans, A., Zijdenbos, A., Neelin, P., Lerch, J., … Fukuda, H. (2004). Voxel-
based morphometry of human brain with age and cerebrovascular risk factors. Neurobiol
Aging, 25(4), 455–463.
Tayebati, S. K., Tomassoni, D., & Amenta, F. (2012). Spontaneously hypertensive rat as a
model of vascular brain disorder: microanatomy, neurochemistry and behavior. J Neurol
Sci, 322, 241–249.
Touyz, R. M. (2005). Intracellular mechanisms involved in vascular remodelling of resistance
arteries in hypertension: role of angiotensin II. Exp Physiol, 90, 449–455.
Tryambake, D., He, J., Firbank, M. J., O’brien, J. T., Blamire, A. M., & Ford, G. A. (2013).
Intensive blood pressure lowering increases cerebral blood flow in older subjects with
hypertension. Hypertension, 61, 1309–1315.
Van Den Berg, E., Kloppenborg, R. P., Kessels, R. P., Kappelle, L. J., & Biessels, G. J. (2009).
Type 2 diabetes mellitus, hypertension, dyslipidemia and obesity: A systematic compari-
son of their impact on cognition. Biochim Biophys Acta, 1792, 470–481.
Veglio, F., Paglieri, C., Rabbia, F., Bisbocci, D., Bergui, M., & Cerrato, P. (2009). Hypertension
and cerebrovascular damage. Atherosclerosis, 205, 331–341.
Waldstein, S. R., Giggey, P. P., Thayer, J. F., & Zonderman, A. B. (2005). Nonlinear relations
of blood pressure to cognitive function: the Baltimore Longitudinal Study of Aging.
Hypertension, 45, 374–379.
Waldstein, S. R., Manuck, S. B., Ryan, C. M., & Muldoon, M. F. (1991). Neuropsychological
correlates of hypertension: review and methodologic considerations. Psychol Bull, 110,
451–468.
Wolf, P. A., Beiser, A., Elias, M. F., Au, R., Vasan, R. S., & Seshadri, S. (2007). Relation of
obesity to cognitive function: importance of central obesity and synergistic influence
of concomitant hypertension. The Framingham Heart Study. Curr Alzheimer Res, 4,
111–116.
Yasar, S., Ko, J. Y., Nothelle, S., Mielke, M. M., & Carlson, M. C. (2011). Evaluation of the
effect of systolic blood pressure and pulse pressure on cognitive function: the Women’s
Health and Aging Study II. PLoS One, 6, e27976.
Zhang, P., Huang, G., & Shi, X. (2011). Cerebral vasoreactivity during hypercapnia is reset by
augmented sympathetic influence. J Appl Physiol 110, 352–358.
415
16
Genetics and the Cognitive

Neuroscience of Aging
Goran Papenberg
Ulman Lindenberger
Lars Bäckman
N ormative aging is associated with decline in different cognitive domains

(e.g., Schaie et al., 1998; Rönnlund et al., 2005). Independent of age,
there are large inter-individual differences in various cognitive functions, which may
increase from early to late adulthood (e.g., de Frias et al., 2007; Tucker-Drob et al.,
2014). These individual differences are sometimes paralleled by greater between-per-
son differences in neural processing with advancing age, as revealed by functional
imaging techniques (Lindenberger et al., 2013). Conceivably, multiple factors contrib-
ute to between-person differences at neural and behavioral levels, including genetic
predispositions and lifestyle factors (Nyberg et al., 2012).
In this chapter, we review studies that investigate the effects of candidate genes
on behavior, as well as on brain structure and functioning in relatively healthy older
adults. The chapter is organized as follows: First, we introduce the resource-modu-
lation notion proffered by Lindenberger and colleagues (2008). This view posits that
effects of common genetic variations on brain and behavior may become stronger
in old age. Our goal is to give a broad overview of the evidence from behavioral,
structural, and functional imaging studies supporting this hypothesis. Next, we briefly
introduce the candidate-gene approach and review the extent to which the available
data support the resource-modulation hypothesis. We focus on four candidate genes
that have attracted considerable attention in cognitive neuroscience of aging: apolipo-
protein E gene (APOE, rs429358, rs7412), brain-derived neurotrophic factor (BDNF,
rs6265) Val66Met polymorphism, catechol- O-methyltransferase (COMT, rs4680)
415
416
Val66Met polymorphism, and the kidney and brain–expressed protein or WW and C2

domain containing 1 (KIBRA, WWC1, rs17070145) polymorphism. Of these candi-
date genes, the number of studies investigating the effects of APOE on cognitive and
brain integrity is considerably larger compared to the other candidate genes, which is
due to the fact that APOE e4 is the strongest genetic risk factor for dementia (Ferencz
and Gerritsen, 2015). Thus, APOE may have also stronger effects on cognition in
healthy elderly than the other candidate genes (e.g., Laukka et al., 2013). Given mixed
results regarding the replication of effects of candidate genes and the fact that most
genes account for only a small fraction of the variance in cognition (Deary et al.,
2010), we will delineate potential factors that may limit or enhance the likelihood to
observe magnified genetic effects in aging.
The Resource-Modulation Hypothesis: Aging-Related

Magnification of Genetic Effects
The resource-modulation hypothesis assumes that losses in neurochemical (e.g.,

Bäckman et al., 2006) and anatomical (e.g., Raz et al., 2005) brain resources in nor-
mal aging modulate the effects of common genetic variations on cognitive function-
ing (see Figure 16.1). Whereas genetic effects in younger adults may be small or
Genetic effects on cognition for two hypothetical

subjects across the adult lifespan
High Young
adulthood
Cognitive performance
Old age
Dementia,
Non-beneficial genotype terminal
Low Beneficial genotype
decline
High Low
Brain resources
Figure 16.1 The resource-modulation hypothesis assumes that the function relating brain
resources to cognition is nonlinear and predicts magnified genetic effects on cognitive perfor-
mance in old age. In healthy aging, associated with decline in anatomical and chemical brain
resources, constant amounts of genetic variation translate into increasingly larger performance
differences. With resources further depleted, genetic effects are expected to diminish. The col-
ored circles represent two hypothetical individuals with different genetic predispositions as
they move from early adulthood through old age to dementia or terminal decline. Adapted from
Papenberg, Lindenberger, & Bäckman (2015) with permission from Elsevier.
417
Genetics and the Cognitive Neuroscience of Aging 417
not detectable, these effects become stronger as people age. The idea is based on
the assumption that the function relating brain resources to cognition is nonlinear,
so that genetic differences exert increasingly larger effects on cognition as resources
recede from high to medium levels. Put differently, older adults may benefit more
from beneficial genetic predispositions relative to younger adults and be able to main-
tain higher brain and cognitive functioning, as a consequence of their more benefi-
cial genetic make-up (Nyberg et al., 2012). Indeed, evidence from heritability studies
suggests that acceleration of cognitive decline in old age is strongly associated with
genetic factors (e.g., Finkel et al., 2005; Deary et al., 2006; Tucker-Drob et al., 2014).
For instance, Tucker-Drob and colleagues (2014) found that one third of individual
differences in global cognition changes in late adulthood are attributable to genetic
influences. The model depicted in Figure 16.1 also predicts that genetic effects are
expected to diminish once individuals have reached very low resource levels, as in
terminal decline or dementia. In line with this, heritability for cognition seems to
decrease once individuals reach dementia, suggesting that genes account less for indi-
vidual differences in Alzheimer’s disease (AD) patients. More specifically, heritabil-
ity estimates for different forms of memory are smaller in samples of individuals with
AD and their unaffected family members than for unaffected family members alone
(Wilson et al., 2011).
As with cognition, there is considerable heritability for brain structure and func-
tion (Toga and Thompson, 2005). Furthermore, given that brain integrity is an inter-
mediate phenotype for behavior (Harris and Deary, 2011), it has been argued that
neural indicators may be more sensitive to genetic effects than behavioral measures
(e.g., Rasch et al., 2010). Older age is associated with reduced grey-matter volumes
(e.g., Raz et al., 2005) and lower white-matter integrity (for a review, see Madden et
al., 2012), as indicated by different measures of diffusion tensor imaging (DTI). For
instance, higher fractional anisotropy (FA) and lower mean diffusivity (MD) reflect
more preserved white-matter integrity. Unfortunately, the interpretation of functional
magnetic resonance imaging (fMRI) signals is not straightforward, as higher blood-
oxygen-level-dependent (BOLD) signal has been suggested to reflect better as well as
worse neural functioning in the absence of behavioral differences. On the one hand,
higher brain activation may reflect less efficient neural processing to achieve the same
behavioral outcomes, a result typically observed during less demanding cognitive
tasks (e.g., Papassotiropoulos et al., 2006). That said, the inefficiency argument may
not always hold, in light of additional information, such as performance differences
between allelic groupings in more demanding offline tasks that may be more suit-
able to disclose genetic differences (e.g., Muse et al., 2014). For example, Muse and
colleagues documented higher hippocampal activation during a recognition memory
task and better immediate and delayed recall performance outside the scanner for car-
riers of the advantageous KIBRA T-allele. In these cases, genetic effects at the brain
level may simply reflect that neural measures are more sensitive to genetic effects
than behavioral measures. Indeed, it has been shown that genetic differences in brain
activation remained the same both in the presence of performance differences and in
performance-matched groups (Kauppi et al., 2011).
Furthermore, increased activation in additional brain regions are commonly
observed in older adults, a phenomenon suggested to reflect decreased selectivity or
418
dedifferentiation of neural processing, but also compensation for deficits elsewhere in

the brain or selection of alternative processing strategies (Grady, 2012; Lindenberger
et al., 2013). Finally, greater brain activation may also reflect pathological aging,
which may in some cases overshadow genetic effects. In line with this, longitudinal
imaging studies have found that older adults with increased and more diffuse brain
activation patterns declined more in their clinical and neurological status, although
these individuals did not differ in baseline memory performance from those with more
distinct activation patterns (Bookheimer et al., 2000; O’Brien et al., 2010). Taken
together, higher brain activations may be interpreted to reflect better as well as worse
neural processing depending on task demands, behavioral outcomes, and sample
characteristics.
Candidate Gene Approach
In the past decade, there has been an increasing interest in investigating the impact of
genetic predispositions on behavioral phenotypes in cognitive neuroscience (Rasch
et al., 2010), often with a focus on candidate genes. The general assumption is that
individual differences in genetic predispositions result in differences in protein expres-
sion in the brain, affecting neural processing and consequently behavior.
Candidate genes are typically selected based on knowledge from previous studies,
linking specific genes to a specific protein and the protein’s involvement in brain and
cognitive functions. In the following, we will focus on four extensively investigated
candidate genes and review their associations with behavioral outcomes, brain struc-
ture and function in adulthood and aging.
APOE Polymorphism
Apolipoprotein E (APOE) is a low-â•‰density lipoprotein that modulates cholesterol

transport (Mahley et al., 2009) and is implicated in several key mechanisms in the
central nervous system, such as neuronal development, brain plasticity, and neural
repair (Bu, 2009). There are three isoforms of APOE, as defined by the two poly-
morphisms, rs429358 and rs7412: the e2, e3, and e4, of which e3 and e2 are the most
and least common, respectively. The e4 allele is the strongest genetic risk factor for
Alzheimer´s disease (AD), such that the presence of one or two ε4 alleles is associated
with higher risk of developing AD in a dose-â•‰dependent manner (e.g., Okuizumi et al.,
1995; Farrer et al., 1997). The e4 allele is also related to increased risk for mild cog-
nitive impairment (MCI; Boyle et al., 2010) and accelerated cognitive decline during
normal aging (Smith, 2002; Bretsky et al., 2003). In line with earlier meta-â•‰analytic
findings (Small et al., 2004), a recent meta-â•‰analysis showed that the e4 allele is associ-
ated with lower performance on measures of episodic memory, executive functioning,
general cognitive ability, and perceptual speed (Wisdom et al., 2011). In addition,
genetic effects were more pronounced in older than younger individuals with respect
to episodic memory and global cognitive ability. In line with this pattern, longitu-
dinal studies have documented interactions between age and APOE on the memory
419
and learning subdomains of the Adult Verbal Learning Test, with stronger negative
effects of e4 in persons older than 50 years than in those below 50 years (Liu et al.,
2010; Figure 16.2A). Similarly, in older adults aged 79 years, the e4 allele was associ-
ated with more rapid decline in verbal memory and abstract reasoning across 8 years
(Schiepers et al., 2012). However, one study showed that the relationship of APOE
to episodic memory and global cognition was attenuated, after exclusion of future
(A) 50
45 no e4
one e4
40 e4/e4
35
Learning
30
25
20
15
10
10 20 30 40 50 60 70 80 90
Age (in years)
(B) 300
DRD2 C/C
DRD2 Any T
250
SSRT in ms
200
150
100
Younger adults Older adults
Figure 16.2 (A) Effects of APOE on learning, with increased negative dose-response effects
of the e4 allele across adult age. Learning reflects number of correctly recalled words in the Rey
Auditory Verbal Learning Test. Adapted from Liu et al. 2010. (B) Age-magnification of DRD2
effects on inhibitory control (in ms) measured by the stop-signal reaction time task (SSRT),
with older any T carriers (fewer dopamine D2 receptors) showing disproportionate slowing.
Reprinted from Papenberg et al. (2015) with permission from Elsevier.
420
dementia cases (Laukka et al., 2013). This suggests that dementia-related processes
may contribute to the link between APOE and cognition in old age.
Functional imaging data on APOE are less consistent than the available behavioral
data. Here we focus on fMRI studies in persons from middle age through old age that
investigate neural correlates of episodic memory, the cognitive domain most severely
affected by AD. Several fMRI studies have documented that elderly e4 carriers have
increased brain activation relative to non-carriers during different episodic memory
tasks, with no performance differences between genotypes (e.g., Bookheimer et al.,
2000; Bondi et al., 2005; Fleisher et al., 2005; Han et al., 2007). For example, increased
BOLD activation have been documented in left MTL, as well as in parietal and prefron-
tal regions for e4 carriers during episodic encoding and recall (Bookheimer et al., 2000;
Fleisher et al., 2005). Using longitudinal assessment over two years, Bookheimer and
colleagues further showed that the degree of baseline brain activation was linked to
the degree of memory decline. The authors suggest a compensatory mechanism such
that individuals at genetic risk for AD (i.e., e4 carriers) have to invest greater cogni-
tive effort to achieve the same performance level as persons with lower genetic risk.
However, efficiency of the elevated BOLD response may decline over time as reported
in a longitudinal imaging study in individuals with prodromal AD (O’Brien et al.,
2010). More specifically, individuals with more rapid cognitive decline were charac-
terized by the highest hippocampal activation at baseline and the greatest reduction of
hippocampal activation over two years during an associative memory task.
On the other hand, there is a bulk of evidence documenting decreased brain activa-
tion for older e4 carriers during memory tasks (e.g., Lind et al., 2006; Fleisher et al.,
2009b; Xu et al., 2009; Kauppi et al., 2014). Employing incidental word encoding and
an associative memory task, two studies reported decreased activity in right MTL for
e4 carriers relative to non-carriers despite similar performance levels (Kauppi et al.,
2014; Lind et al., 2006). Lower MTL activity was interpreted as an early sign of
hippocampal pathology for individuals at risk for AD. Another study demonstrated
equivalent face recognition performance accompanied by decreased BOLD activity
in left posterior cingulate cortex (PCC) and precuneus in the high-risk group (i.e.,
first-degree family history of dementia and presence of the APOE e4 allele) compared
to the low-risk group (Xu et al., 2009). The authors speculated that decreased BOLD
activity in those regions might disrupt the functional architecture of the default mode
network (DMN), which is severely affected in MCI and AD. Related to this argu-
ment, genetic differences have been observed with respect to decreased deactivation
of the DMN (e.g., Fleisher et al., 2009a; Pihlajamaki et al., 2010). Using an identical
memory task, Fleisher and colleagues also reported no BOLD differences between e4
carriers and non-carriers during encoding (Fleisher et al., 2009a). Instead, encoding-
associated deactivation was greater in magnitude among non-e4 carriers in medial
and right lateral parietal cortex. This finding is consistent with results from Xu and
colleagues (2009), demonstrating functional alterations in critical nodes of the DMN
for e4 carriers compared to non-carriers.
However, studies with larger numbers of participants (n > 100) failed to find
BOLD differences between e4 carriers and non-carriers during memory tasks (Bassett
et al., 2006; Johnson et al., 2006). Considering first-degree family history of demen-
tia, Johnson and colleagues showed that e4 carriers with and without family history
421
exhibited the least and the greatest signal change in MTL, respectively. This suggests
that other genetic factors and possibly non-genetic risk factors associated with AD
may exacerbate the negative effects of APOE e4 on functional brain measures (e.g.,
Donix et al., 2012).
So far, only two studies investigated the effect of the APOE genotype on brain
function across the adult lifespan. Using a memory encoding task in younger (aged
20 to 35) and older individuals (aged 50 to 78), Filippini and colleagues (2011)
showed a significant interaction between age and e4 status in MTL, frontal pole,
subcortical nuclei, middle temporal gyrus, and cerebellum, such that aging was
associated with decreased activity in e4 carriers and increased activity in non-
carriers. Further, cerebral blood flow was reduced for older e4 carriers relative
to non-carriers, which could partly account for decreased activity in individuals
at genetic risk for AD. However, data from another lifespan sample (aged 19–77
years) suggested an opposite pattern for hippocampal activation in older adults
(Nichols et al., 2012). In this study, there was decreased hippocampal activity
during encoding and retrieval of neutral pictures with increasing age, and these
decreases were weaker for e4 carriers than for non-carriers. These two studies
provide evidence in line with the resource-modulation hypothesis, although they
document opposing genetic effects on neural functioning in old age. One source of
variation that may account for this discrepancy between studies is task difficulty.
Older adults at higher risk for cognitive decline typically show more brain activity
during relatively simple tasks than individuals at lower risk. During more difficult
tasks, the pattern may be reversed (Grady, 2012). In line with this notion, partici-
pants were instructed to remember images in the study where older e4 carriers had
lower brain activity at encoding (Filippini et al., 2011). This task is clearly more
cognitively challenging than judging the contents of images during study, a task
for which greater brain activity in older e4 carriers was observed (Nichols et al.,
2012). With respect to markers of brain structure, interactive effects between age
and APOE on white m atter microstructure have not been reported (e.g., Heise et
al., 2011; Westlye et al., 2012; Nyberg and Salami, 2014). However, longitudinal
studies have documented more hippocampal atrophy for e4 carriers than for non-
carriers (Cohen et al., 2001; Crivello et al., 2010) among healthy elderly persons.
Taken together, elevated BOLD activation in e4 carriers relative to non-carriers
were interpreted as a compensatory mechanism to circumvent initial AD-related
damage using more cognitive effort (Bookheimer et al., 2000; Han et al., 2007).
Some studies, however, have reported decreased BOLD activation in e4 carriers,
assumed to reflect pathological disruption (Kauppi et al., 2014; Lind et al., 2006;
Pihlajamaki et al., 2010; Xu et al., 2009). To some extent, differences in task dif-
ficulty may account for inconsistent findings across studies. Older adults at higher
risk for cognitive decline may show higher brain activation at lower cognitive
loads than individuals at less risk (Grady, 2012). With increasing load, however,
the pattern may be reversed. Further, as noted above, longitudinal imaging evi-
dence suggests decreases in hippocampal activation for individuals with greater
cognitive decline, despite over-recruitment at baseline (O’Brien et al., 2010).
Thus, higher task-related brain activation for e4 carriers may reflect initial com-
pensatory attempts, which with advancing age and further age-related decline in
422
brain resources become less efficient, resulting in lower brain activation. Although
behavioral and structural imaging data suggest that negative effects of the e4 allele
may increase with age, corresponding longitudinal fMRI data are lacking and age-
comparative cross-sectional research is scarce.
BDNF Polymorphism
The brain-derived neurotrophic factor (BDNF) promotes activity-dependent synaptic

plasticity and is critical for learning and memory (e.g., Binder and Scharfman, 2004).
The Val66Met polymorphism (rs6265) of the BDNF gene is associated with individ-
ual differences in secretion of BDNF, which is higher in Val homozygotes than in
Met carriers (Egan et al., 2003). At the behavioral level, older Val homozygotes have
exhibited superior episodic memory, but also processing speed and general intelli-
gence compared to any Met carriers (Miyajima et al., 2008). Indeed, meta-analytic
evidence confirms adverse effects of the BDNF Met allele on human episodic mem-
ory (Kambeitz et al., 2012). Adult age-comparative studies have reported magnified
effects of BDNF in old age for episodic memory (Li et al., 2010), with older Val homo-
zygotes performing better on backward serial recall. In line with the resource-mod-
ulation hypothesis, longitudinal data demonstrate exacerbated decline in perceptual
speed across 13 years among older BDNF Met carriers (Ghisletta et al., 2014; Figure
16.3A), an effect that remained after excluding prodromal dementia cases. Similarly,
Sanchez et al. (2011) reported that pilots carrying the Met allele (aged 40–69 years)
declined disproportionately in flight-simulator performance, presumably reflecting
executive functioning.
In young adults, BDNF Met carriers have shown lower activity in hippocampus
during encoding and retrieval of episodic memories (e.g., Egan et al., 2003; for meta-
analysis, see Kambeitz et al., 2012; Hariri et al., 2003) and during performance of a
virtual navigation task (Banner et al., 2011). Whereas Hariri and colleagues (2003)
found BDNF-related differences in episodic memory, accompanied by differences in
hippocampal activation during encoding and retrieval, other studies have failed to
find behavioral genotype differences, although Met carriers had lower hippocampal
activity during encoding of episodic memories (Hashimoto et al., 2008). Of particular
interest, one lifespan study (n = 125; aged 19–85 years) showed more pronounced
decrease in hippocampal activity with advancing age for Met carriers than for Val
homozygotes (Sambataro et al., 2010). This pattern was evident during both episodic
encoding and retrieval (Figure 16.3B). Importantly, after adjusting for inter-individ-
ual differences in hippocampal volume, the differential activation pattern remained.
Again, however, this study did not find any behavioral genotype differences, support-
ing the notion that neural measures may be more sensitive to genetic differences than
behavioral measures (e.g., Rasch et al., 2010).
Age magnification of the effects of BDNF has also been documented for measures
of brain integrity, emphasizing the role of BDNF in modulating myelin expression
(Ikeda et al., 2002) and survival of neurons in the adult brain (Morse et al., 1993).
Cross-sectional imaging studies have shown age magnification of the effects of BDNF
for grey matter volumes and white matter microstructure. Specifically, Met carriers
423
(A) (B) 0.8

70
Cognitive performance 0.6
Hippocampal activity
60 0.4
50 0.2
0
40
–0.2
30
Val/Val –0.4 Val/Val
20 Any Met Any Met
–0.6
70 75 80 85 90 95 100 105 20 40 60 80
Age (in years) Age (in years)
(C) (D)
0.90
6.2
Fractional anisotropy
5.8 0.80
5.4
5.0 0.70
4.6
Val/Val
0.60
4.2 Any Met
3.8 0.00
< 65 ≥ 65 < 65 ≥ 65
0 40 50 60 70 80 90
Val/Val Any Met Age (in years)
Age (in years)
Figure 16.3 Effects of BDNF on (A) longitudinal decline in perceptual speed across 13 years,
with steeper decline for BDNF Met carriers. Perceptual speed is measured using the digit-letter
task, which required participants to name letters associated with a digit, according to a template.
The y-axis indicates total number of correct responses after 3 min. Adapted from Ghisletta et al.
(2014). Interaction between age and BDNF, reflecting (B) lower hippocampal activity during
retrieval of episodic memories, (C) smaller hippocampal volumes, and (d) lower white m atter
integrity in the splenium for older BDNF Met carriers. Hippocampal activity in (b) indicates
parameter estimates of the BOLD response measured in arbitrary units in left hippocampus,
which is greater during retrieval relative to a baseline condition. White matter integrity is indi-
cated by fractional anisotropy. Reprinted from Papenberg et al. (2015) with permission from
Elsevier. (See color plate also)
had lower hippocampal volumes after age 65 than Val homozygotes, whereas no such
differences were apparent before 65 (Sanchez et al., 2011; Figure 16.3C). Critically,
age was uncorrelated with hippocampal volume in Val homozygotes, supporting the
idea that brain maintenance in old age may be partly due to genetic factors (Nyberg
et al., 2012). Another study in individuals with prodromal AD reported that the Met
allele was associated with increased memory decline, paralleled by more hippocampal
atrophy, across three years (Lim et al., 2014). Similarly, age-related decline in white
matter microstructure (i.e., lower fractional anisotropy) was found for Met carriers in
the splenium of the corpus callosum, although no age-related decline was evident for
Val homozygotes (Kennedy et al., 2009; Figure 16.3D).
Taken together, behavioral evidence as well as data from functional and structural
imaging studies suggest magnified effects of BDNF on brain and cognition in aging,
with greater decline in functioning for older Met carriers.
424
COMT Polymorphism
The COMT Val158Met polymorphism (rs4680) is the most studied dopamine-related

polymorphism. The COMT enzyme is involved in extracellular degradation of syn-
aptically released dopamine in the prefrontal (PFC) cortex (e.g., Matsumoto et al.,
2003). Dopamine concentration increases neuronal signal-to-noise ratio in PFC, crit-
ical to efficient cognitive processing (Egan et al., 2001). COMT Val homozygotes
have three to four times higher turnover rates of this enzyme than Met homozygotes
(Lotta et al., 1995), resulting in lower prefrontal DA availability and presumably less
efficient processing.
Studies with young adults have demonstrated a COMT Met advantage in executive
functions and memory performance compared to Val homozygotes (for review, see
Witte and Flöel, 2011). Such cognitive benefits have been found to correlate with
lower brain activation, indicating more efficient information processing (Egan et al.,
2001; Sambataro et al., 2009). However, several studies have failed to replicate these
findings (Bolton et al., 2010; de Frias et al., 2010; Blanchard et al., 2011; Stuart et al.,
2014), and two meta-analyses only documented a COMT Met superiority on measures
of general cognition, such as IQ (Barnett et al., 2007; Barnett et al., 2008).
Although the effect of COMT on cognition in healthy aging has received less atten-
tion, there is evidence that COMT-related differences become more apparent with
increasing adult age (de Frias et al., 2005; Nagel et al., 2008; Papenberg et al., 2013b).
For instance, evidence for such age interactions comes from cross-sectional studies
demonstrating faster response times for Met homozygotes during spatial working
memory in older, but not in younger individuals (Nagel et al., 2008). Most impor-
tantly, longitudinal data reveal less decline of executive function over a five-year inter-
val (de Frias et al., 2005) and less episodic memory decline across 15 years (Josefsson
et al., 2012) for middle-aged and older COMT Met carriers than for Val homozygotes.
Both older age and COMT Val status have been associated with altered fronto-striatal
dopamine functioning (e.g., Slifstein et al., 2008; Klostermann et al., 2012) that might
translate into reduced neural efficiency and lower cognitive performance. Although
only a few studies have tested whether aging modulates the link between COMT and
brain functioning, there is some evidence suggesting the existence of such interac-
tions. In a recent fMRI study, Nyberg and colleagues (2014) investigated the inde-
pendent effects of aging and COMT on working memory performance and patterns of
brain activation in a large population-based sample. Given previous observations in
older adults (e.g., Nagel et al., 2009), they predicted weaker PFC response during high
working-memory load (manipulation), along with increased BOLD response during
low working-memory load (maintenance) in older adults and COMT Val carriers com-
pared to younger adults and COMT Met carriers. In line with the predictions, older
individuals had weaker BOLD modulation in PFC during working-memory manipu-
lation. Also, the weakest PFC activation during manipulation was observed in COMT
Val carriers. Conversely, older adults and Val carriers had elevated BOLD response in
PFC during the less cognitively taxing maintenance condition.
In another fMRI study, Sambataro and colleagues (2009) took a multivariate
approach to investigate the effects of COMT on brain-network connectivity using a
low-load working memory task. They found that a network including left PFC and
425
parietal cortex was modulated by COMT, with Val homozygotes showing increased
connectivity from dorsolateral PFC to other components in this network compared to
Met homozygotes. Notably, the largest COMT-related difference was seen in older,
relative to younger, individuals. This suggests additive effects of age-related dopa-
mine deterioration and COMT. The findings from these two studies of increased PFC
activation and functional connectivity in older individuals and Val carriers is in agree-
ment with the efficiency hypothesis of COMT influences on brain functioning (Mier
et al., 2010).
In conclusion, there is preliminary evidence to support the notion that the COMT
Val158Met polymorphism and age jointly modulate dopamine and PFC efficiency
such that older Val carriers display limited PFC upregulation in response to increased
task demands. Moreover, initial evidence suggests that there is an interaction between
age and COMT on measures of PFC functioning. Relatedly, a structural-imaging
study in a population-based sample found that COMT Val status was associated with
reduced white m atter integrity, reflected by lower fractional anisotropy and higher
mean diffusivity, of several prefrontal white-matter tracts in the oldest age group (81–
87 years), although there were no reliable associations between COMT and white
matter microstructure in two younger age groups (60–66 and 72–78 years; Papenberg
et al., 2014a). This finding is particularly relevant, as BOLD responsivity does not
predict working-memory performance after controlling for individual differences in
white matter integrity (Burzynska et al., 2011). This raises the question of whether the
magnified effects of COMT on white m atter microstructure may be related to those on
functional brain activity.
KIBRA Polymorphism
A genetic variation (rs17070145) in the WWC1 gene, which encodes the KIBRA
protein, has been associated with episodic memory in humans through genome-wide
screening, with T-allele carriers exhibiting better performance (Papassotiropoulos et
al., 2006). In the human brain, KIBRA is mainly expressed in hippocampus and has
been linked to long-term potentiation and synaptic plasticity (Schneider et al., 2010). A
meta-analysis reported a reliable association between rs17070145 and episodic mem-
ory as well as working memory, explaining 0.5% and 0.1% of variance, respectively
(Milnik et al., 2012; for review, see Schwab et al., 2014). In line with the resource-mod-
ulation hypothesis, a recent behavioral study reported that older adults carrying the
T-allele showed better spatial learning compared to C homozygotes, whereas no geno-
type effects were found in younger adults (Schuck et al., 2013). Interestingly, Almeida
and colleagues (2008) showed better episodic memory for older KIBRA T-carriers, but
there was no effect of this polymorphism in a sample of older adults with mild cogni-
tive impairment. This pattern is supportive of the lower end of the distribution portray-
ing the resource-modulation hypothesis (see Figure 16.1), predicting genetic effects to
diminish once individuals approach dementia or death.
fMRI studies have documented lower as well as higher brain activation for car-
riers of the beneficial T-allele. Papassoptiropolous and colleagues (2006) reported
lower hippocampal activity during episodic retrieval for young T-allele carriers, in
426
the absence of behavioral differences in the scanner task. The authors interpreted this
pattern in terms of more efficient processing for T-â•‰allele carriers. A different pattern
was reported by a more recent fMRI study with elderly persons (Kauppi et al., 2011).
Specifically, the behavioral data in a sample of 2230 participants (aged 35–â•‰85 years)
suggested an advantageous effect of the T-â•‰allele on immediate free recall, which
was magnified with increasing age. Further, in a subsample (n = 83), Kauppi and
colleagues demonstrated increased hippocampal activity in T-â•‰allele carriers, which
was evident when the genotype groups differed in memory performance, but also
when the groups were matched for in-â•‰scanner task performance. However, KIBRA
modulated episodic memory and hippocampal activation only in middle-â•‰aged adults
(aged 55 to 60 years). Despite age magnification of KIBRA effects on behavior in
the larger sample, there was no genetic modulation of brain activity and memory in
the scanner task in the older age group (aged 65–â•‰75 years). The authors speculated
that older adults with the disadvantageous genotype might have increased hippocam-
pal activation associated with pathological aging that overshadow the genetic effects.
Another lifespan study reported further evidence in favor of the resource-â•‰modulation
hypothesis (Muse et al., 2014). First, increasing age was associated with stronger
effects of the disadvantageous C-â•‰allele on immediate and delayed (30 minute) free
recall (Figure 16.4A, B). In contrast, there were no behavioral differences between
genotypes or interactions between age and genotype for the recognition memory task
during scanning. However, older, but not younger, C homozygotes had lower hippo-
campal activation during encoding and retrieval, suggesting stronger genetic effects
in advanced age (Figure 16.4C, D). The latter study provides additional support for
the resource-â•‰modulation hypothesis and the notion that brain measures are more sen-
sitive to genetic effects; genetic effects on behavior became evident only during more
demanding episodic-â•‰recall tasks, performed outside the scanner.
Miscellenous Genes and Gene–Gene Interaction
Taken together, the data reviewed above suggest that effects of common genetic varia-
tions on behavior, brain structure and functioning may become stronger with increas-
ing adult age, supporting the resource-â•‰modulation hypothesis. We focused on a few
of the most extensively investigated polymorphisms. However, several other studies
investigating relationships of common genetic variations to brain and behavior have
also found stronger genetic effects in old age. Notably, studies have also documented
gene–â•‰gene interactions and additive effects of different polymorphisms on brain and
behavior in older adults, emphasizing the importance of investigating effects of mul-
tiple genes.
Examining the association between a dopamine D2 receptor (DRD2, C957T,
rs6277) polymorphism and inhibitory control, Colzato and colleagues (2013) reported
that genetic predisposition for higher density of extrastriatal D2 receptors (DRD2
CC) was associated with better inhibition of unwanted action tendencies (Figure
16.2B), an effect that was most pronounced in older adults. In another study, DRD2
interacted with the dopamine transporter (DAT) gene on backward serial recall (Li
et al., 2012): Homozygotes for the DRD2 C and DAT 9-â•‰repeat alleles (associated
427
(A) (B) Recall after 30 minutes

KIBRA genotype KIBRA genotype
T carriers T carriers
(C) (D)
KIBRA genotype KIBRA genotype
T carriers T carriers
Figure 16.4 KIBRA genotype groups show different correlations between increasing age and per-
formance on (A) immediate and (B) 30-minute delayed recall of a story, as measured with the
Wechsler Memory Scale. (c, d) KIBRA genotype group differences in the correlation between age
and brain activation during an episodic memory task. (C) The KIBRA CC group (red) exhibits a
negative correlation between age and activity in left hippocampus during encoding, which is not
observed for T allele carriers (blue). (D) The KIBRA CC group (red) exhibits a negative correlation
between age and activity in right hippocampus during retrieval, which is not observed for T allele
carriers (blue). Hippocampal activity indicates parameter estimates of the BOLD response mea-
sured in arbitrary units, which is greater during encoding and retrieval relative to a baseline condi-
tion. Reprinted from Papenberg et al. (2015) with permission from Elsevier. (See color plate also)
with higher synaptic DA levels) showed overall higher recall accuracy. The genetic
main effects and the gene–gene interaction were again larger in older than in younger
adults. Similarly, age magnification of the effects of two other dopamine-related
polymorphisms, namely COMT and the dopamine betahydroxylase (DBH; C-1021T;
rs1611115), were observed on working memory (Greenwood et al., 2014), with older
adults with lowest synaptic dopamine performing worst. As both the dopaminer-
gic and glutamatergic systems modulate episodic memory consolidation, one study
investigated whether DRD2 and a variation of the N-methyl-D-aspartate 3A (NR3A;
rs10989591) gene, coding for the NR3A subunit of the glutamate N-methyl-D-aspar-
tate (NMDA) receptor, interactively modulate episodic memory (Papenberg et al.,
2014b). The gene–gene interaction was observed in older adults only, with individuals
428
carrying genotypes associated with greater D2 and NMDA receptor efficacy showing
the highest episodic memory performance. In another study, genetic predispositions
for DA-â•‰relevant genes affecting DAT expression and D2-â•‰like receptors (i.e., D2 and
D3) were aggregated into a composite gene score (Papenberg et al., 2013a). Older
adults carrying more beneficial alleles showed an episodic memory advantage, this
time in terms of less forgetting after 1 week. No genetic effects were observed in
younger adults. Recently, additive adverse effects of COMT Val/â•‰Val, BDNF Any Met,
and age have been reported on executive functioning in middle-â•‰aged and older adults
(53–â•‰95 years), indicating that older adults with a high-â•‰risk combination performed
worse. The effects were further intensified by the presence of the APOE e4 allele
(Sapkota et al., 2014). In addition, a genome-â•‰wide association study demonstrated
effects of APOE (rs769449) on rate of cognitive decline (Zhang and Pierce, 2014),
thereby supporting the magnification notion.
With respect to fMRI studies, the TaqIA polymorphism of the dopamine D2 recep-
tor (DRD2)/â•‰ANKK1 gene (rs1800497) has been related to striatal dopamine recep-
tors, with A1 allele carriers having reduced density of D2 receptors (Jonsson et al.,
1999). Persson and colleagues (2014) showed lower performance in long-â•‰term mem-
ory updating among older A1 carriers compared to non-â•‰carriers. In addition, older A1-â•‰
carriers had less BOLD activation in left caudate nucleus, a region critical to updating.
None of these effects were present in younger adults. Ebner and colleagues (2013)
investigated the association of the oxytocin receptor (OXTR, rs237887) polymorphism,
previously associated with susceptibility to prosocial behavior, to face recognition and
BOLD activity in younger and older adults. Results showed that OXTR modulated
activity in anterior cingulate cortex of older adults only. Specifically, higher brain
activity, indicating more affective processing of happy compared to angry faces, were
observed for older A homozygotes compared to Any G carriers. Behaviorally, this was
reflected in faster response times in identifying happy faces for older A homozygotes.
Further, an electroencephalography study examined the association between the AD-â•‰
related clusterin (CLU; rs11136000) polymorphism and resting-â•‰state alpha-â•‰rhythm
activity in healthy non-â•‰carriers of the APOE e4 allele (Ponomareva et al., 2013). CLU
modulated alpha activity only in older adults (50–â•‰80 years), with no genetic effects
in younger adults (20–â•‰50 years). Imaging studies investigating additive or interactive
genetic effects in old age are rare. However, data suggest that considering more than
one polymorphism may help explain more variance in brain activity. For instance,
Kauppi and colleagues (2014) demonstrated that MTL activity during episodic encod-
ing decreased as a function of number of APOE e4 and BDNF Met alleles (none, one,
or both), yielding stronger effects than those of the individual genes.
Factors Affecting Age Magnification of Genetic Effects
Despite increasing evidence in favor of the resource-magnification model, the avail-

able evidence is not unequivocal. Other than the gene–â•‰gene interactions reported
above, there are several reasons for this fact. Whereas most age-â•‰comparative studies
include carefully selected convenience samples, use of population-â•‰based samples
likely introduces many lifestyle and individual-â•‰difference factors that may wash
429
out genetic effects. For example, in a large-scale population-based study, Laukka

and colleagues (2013) reported relationships of APOE to episodic memory, percep-
tual speed, and global cognition. However, no associations were found for COMT,
BDNF, or KIBRA. Similarly, some authors have argued that higher BOLD activity
may reflect pathological processes associated with a disadvantageous genotype,
which may overshadow genetic effects in older age (e.g., Kauppi et al., 2011), or
lead to inconsistent results across studies. Indeed, most available genetic studies on
brain and cognition did not control for incident dementia. This is a serious omis-
sion, given that the prodromal phase of dementia might start more than 10 years
before clinical diagnosis (Thorvaldsson et al., 2011). The resource-modulation
hypothesis holds that genetic effects become weaker when individuals approach
dementia or terminal decline. For instance, APOE genotype does not modify rate
of decline in AD after the clinical diagnosis has been made (e.g., Corder et al.,
1995), and even progression from the preclinical stage to clinically verified AD is
indistinguishable for carriers and non-carriers of the e4 allele (e.g., Bunce et al.,
2004). Thus, including preclinical dementia cases may make it difficult to observe
already small effects of various genetic polymorphisms (Figure 16.1). However,
the prediction of the resource-modulation hypothesis that genetic effects become
weaker once individuals approach the preclinical stage of dementia and eventually
dementia diagnosis has still not been extensively tested.
Research further suggests that once additional factors are taken into account, it
may be easier to disclose genetic effects in old age. Lifestyle factors, such as physical
activity, may modulate genetic effects on brain and cognition. It has been suggested
that the evolution of physical activity approximately 2 million years ago resulted
in the reduction of amyloid plaques and vascular burden associated with the APOE
e4 genotype, relaxing genetic constraints on aging (Raichlen and Alexander, 2014).
Indeed, increased physical exercise changes levels of DNA methylation and gene
expression in human adipose tissue, supporting metabolic changes through epigene-
tic modifications (Rönn et al., 2013). So far, imaging studies investigating interactive
effects between genes and environmental factors in older age are scarce. However,
one study reported that older APOE e4 carriers who are more physically active had
higher activity in posterior temporal and parietal regions during an episodic memory
task than non-carriers or those with lower physical activity levels. These data suggest
that physical activity may circumvent the negative effects of carrying a disadvan-
tageous genotype on brain functioning (see also Erickson et al., 2013; Ferencz et
al., 2014). Similar to lifestyle factors, behavior-genetic studies indicate that differ-
ent diseases may lower resources and make it easier to disclose genetic effects. For
instance, interactive effects of KIBRA and the calsyntenin 2 (CLSTN2; rs6439886)
polymorphism have been observed for episodic memory in older adults with depres-
sion, with individuals carrying both risk alleles (KIBRA CC and CLSTN2 TT) per-
forming the worst (Pantzar et al., 2014). However, no genetic effects were observed
in non-depressed individuals, suggesting that genetic effects are most easily detected
at suboptimal levels of brain integrity. Similarly, COMT Any Val carriers treated
with chemotherapy performed worse on tests of attention than healthy controls with
the same genotype, but no history of breast cancer (Small et al., 2011). Relatedly, an
imaging study reported stronger effects of COMT in populations with reduced brain
430
resources (Ceaser et al., 2013) compared to healthy controls. Specifically, patients

with schizophrenia and their siblings, but not healthy controls, who were Val homo-
zygotes displayed greater activity in frontal regions, striatum, and cerebellum during
a working memory task, presumably reflecting inefficient processing.
Thus, taking into account different lifestyle and disease-â•‰related factors may further
elucidate the heterogeneity in brain and cognitive aging. Importantly, most studies
reviewed in this chapter do not report data from middle-â•‰aged adults. This calls for large-â•‰
scale longitudinal studies covering the whole lifespan to better understand the temporal
dynamics of genetic influences on brain functioning and behavior, including their inter-
actions with lifestyle factors and different diseases. Longitudinal studies that seek to
address epigenetic mechanisms (e.g., Sweatt, 2013) may help to close the gap between
heritability estimates derived from behavior-â•‰genetic data and the small amounts of
variance predicted by common genetic variation related to individual polymorphisms
(Turkheimer, 2011). Given that common genetic variations may affect both brain struc-
ture and function, multimodal imaging studies are imperative to uncover genetic and
epigenetic effects on individual differences in the aging of brain and behavior.
Conclusion
Increasing evidence at behavioral and neural levels of analysis suggests that effects
of common genetic variations on behavior and brain become stronger in late life,
supporting the resource-â•‰modulation hypothesis. Similar patterns have been reported
in other populations characterized by reduced brain resources, by contrasting samples
with different diseases and healthy controls. So far, the bulk of studies are cross-â•‰sec-
tional. In particular, longitudinal structural and functional imaging studies are needed
to confirm the patterns reported in the cross-â•‰sectional imaging data. Furthermore,
some of the inconsistent patterns likely stem from gene–â•‰gene interactions, and from
environmental and lifestyle factors that result in epigenetic differences. Behavioral,
structural, and functional imaging studies are needed that consider the operation of
these factors during the transition from early to late adulthood.
Acknowledgments
Preparation of this chapter was supported by grants from the Swedish Research
Council, the Swedish Research Council for Health, Working Life, and Welfare,
Swedish Brain Power, an Alexander von Humboldt Research Award, and a donation
from the AF Jochnick Foundation to Lars Bäckman.
References
Almeida OP, Schwab SG, Lautenschlager NT, Morar B, Greenop KR, Flicker L, Wildenauer D
(2008) KIBRA genetic polymorphism influences episodic memory in later life, but does
not increase the risk of mild cognitive impairment. J Cell Mol Med 12:1672–â•‰1676.
431
Bäckman L, Nyberg L, Lindenberger U, Li S-C, Farde L (2006) The correlative triad among
aging, dopamine, and cognition: Current status and future prospects. Neurosci Biobehav
Rev 30:791–807.
Banner H, Bhat V, Etchamendy N, Joober R, Bohbot Vd (2011) The brain-derived neurotrophic
factor Val66Met polymorphism is associated with reduced functional magnetic resonance
imaging activity in the hippocampus and increased use of caudate nucleus-dependent
strategies in a human virtual navigation task. Eur J Neurosci 33:968–977.
Barnett Jh, Scoriels L, Munafo Mr (2008) Meta-analysis of the cognitive effects of the catechol-
O-methyltransferase gene Val158/108Met polymorphism. Biol Psychiatry 64:137–144.
Barnett Jh, Jones Pb, Robbins Tw, Müller U (2007) Effects of the catechol-O-methyltransferase
Val158Met polymorphism on executive function: a meta-analysis of the Wisconsin Card
Sort Test in schizophrenia and healthy controls. Mol Psychiatry 12:502–509.
Bassett Ss, Yousem Dm, Cristinzio C, Kusevic I, Yassa Ma, Caffo Bs, Zeger Sl (2006) Familial
risk for Alzheimer’s disease alters fMRI activation patterns. Brain 129:1229–1239.
Binder Dk, Scharfman He (2004) Brain-derived neurotrophic factor. Growth factors 22:123–131.
Blanchard Mm, Chamberlain Sr, Roiser J, Robbins Tw, Muller U (2011) Effects of two
dopamine-modulating genes (DAT1 9/10 and COMT Val/Met) on n-back working mem-
ory performance in healthy volunteers. Psychol Med 41:611–618.
Bolton Jl, Marioni Re, Deary Ij, Harris Se, Stewart Mc, Murray Gd, Fowkes Fgr, Price Jf (2010)
Association between polymorphisms of the dopamine receptor D2 and catechol-o-methyl
transferase genes and cognitive function. Behav Genet 40:630–638.
Bondi Mw, Houston Ws, Eyler Lt, Brown Gg (2005) fMRI evidence of compensatory mecha-
nisms in older adults at genetic risk for Alzheimer disease. Neurology 64:501–508.
Bookheimer Sy, Strojwas Mh, Cohen Ms, Saunders Am, Pericak-Vance Ma, Mazziotta Jc,
Small Gw (2000) Patterns of brain activation in people at risk for Alzheimer’s disease. N
Engl J Med 343:450–456.
Boyle Pa, Buchman As, Wilson Rs, Kelly Jf, Bennett Da (2010) The APOE epsilon4 allele
is associated with incident mild cognitive impairment among community-dwelling older
persons. Neuroepidemiology 34:43–49.
Bretsky P, Guralnik Jm, Launer L, Albert M, Seeman Te, Macarthur Studies Of Successful A
(2003) The role of APOE-epsilon4 in longitudinal cognitive decline: MacArthur Studies
of Successful Aging. Neurology 60:1077–1081.
Bunce D, Fratiglioni L, Small Bj, Winblad B, Bäckman L (2004) APOE and cognitive decline
in preclinical Alzheimer’s disease and non-demented aging. Neurology 63:816–821.
Bu G (2009) Apolipoprotein E and its receptors in Alzheimer’s disease: pathways, pathogenesis
and therapy. Nat Rev Neurosci 10:333–344.
Burzynska Az, Nagel Ie, Preuschhof C, Li Sc, Lindenberger U, Backman L, Heekeren Hr
(2011) Microstructure of frontoparietal connections predicts cortical responsivity and
working memory performance. Cereb Cortex 21:2261–2271.
Ceaser A, Csernansky Jg, Barch Dm (2013) COMT influences on prefrontal and striatal blood
oxygenation level-dependent responses during working memory among individuals with
schizophrenia, their siblings, and healthy controls. Cogn neuropsychiatry 18:257–283.
Cohen Rm, Small C, Lalonde F, Friz J, Sunderland T (2001) Effect of apolipoprotein E gen-
otype on hippocampal volume loss in aging healthy women. Neurology 57:2223–2228.
Colzato Ls, Van Den Wildenberg Wp, Hommel B (2013) The genetic impact (C957T-DRD2) on
inhibitory control is magnified by aging. Neuropsychologia 51:1377–1381.
Corder Eh, Saunders Am, Strittmatter Wj, Schmechel De, Gaskell Pc, Rimmler Jb, Locke Pa,
Conneally Pm, Schmader Ke, Tanzi Re, Gusella Jf, Small Gw, Roses Ad, Pericakvance
Ma, Haines Jl (1995) Apolipoprotein E, survival in Alzheimer’s disease patients, and the
competing risks of death and Alzheimer’s disease. Neurology 45:1323–1328.
432
Crivello F, Lemaitre H, Dufouil C, Grassiot B, Delcroix N, Tzourio-Mazoyer N, Tzourio C,

Mazoyer B (2010) Effects of ApoE-epsilon4 allele load and age on the rates of grey matter
and hippocampal volumes loss in a longitudinal cohort of 1186 healthy elderly persons.
De Frias Cm, Lövdén M, Lindenberger U, Nilsson L-G (2007) Revisiting the dedifferentiation
hypothesis with longitudinal multi-cohort data. Intelligence 35:381–392.
De Frias Cm, Annerbrink K, Westberg L, Eriksson E, Adolfsson R, Nilsson Lg (2005) Catechol
O-methyltransferase Val158Met polymorphism is associated with cognitive performance
in nondemented adults. J Cogn Neurosci 17:1018–1025.
De Frias Cm, Marklund P, Eriksson E, Larsson A, Oman L, Annerbrink K, Backman L, Nilsson
Lg, Nyberg L (2010) Influence of COMT gene polymorphism on fMRI-assessed sustained
and transient activity during a working memory task. Journal of cognitive neuroscience
22:1614–1622.
Deary Ij, Spinath Fm, Bates Tc (2006) Genetics of intelligence. Eur J Hum Genet 14:690–700.
Deary Ij, Penke L, Johnson W (2010) The neuroscience of human intelligence differences. Nat
Rev Neurosci 11:201–211.
Donix M, Small Gw, Bookheimer Sy (2012) Family history and APOE-4 genetic risk in
Alzheimer’s disease. Neuropsychol Rev 22:298–309.
Ebner Nc, Maura Gm, Macdonald K, Westberg L, Fischer H (2013) Oxytocin and socioemo-
tional aging: Current knowledge and future trends. Front Hum Neurosci 7:487.
Egan Mf, Goldberg Te, Kolachana Bs, Callicott Jh, Mazzanti Cm, Straub Re, Goldman D,
Weinberger Dr (2001) Effect of COMT Val108/158 Met genotype on frontal lobe function
and risk for schizophrenia. Proc Natl Acad Sci USA 98:6917–6922.
Egan Mf, Kojima M, Callicott Jh, Goldberg Te, Kolachana Bs, Bertolino A, Zaitsev E, Gold B,
Goldman D, Dean M, Lu B, Weinberger Dr (2003) The BDNF val66met polymorphism
affects activity-dependent secretion of BDNF and human memory and hippocampal func-
tion. Cell 112:257–269.
Erickson Ki, Banducci Se, Weinstein Am, Macdonald Aw, Ferrell Re, Halder I, Flory Jd,
Manuck Sb (2013) The brain-derived neurotrophic factor Val66Met polymorphism mod-
erates an effect of physical activity on working memory performance. Psychological
Science, 24:1770–1779
Farrer La, Cupples La, Haines Jl, Hyman B, Kukull Wa, Mayeux R, Myers Rh, Pericak-Vance
Ma, Risch N, Van Duijn Cm (1997) Effects of age, sex, and ethnicity on the association
between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and
Alzheimer Disease Meta Analysis Consortium. JAMA 278:1349–1356.
Ferencz B, Gerritsen L (2015) Genetics and underlying pathology of dementia. Neuropsychol
Rev 25:113–124.
Ferencz B, Laukka Ej, Welmer Ak, Kalpouzos G, Angleman S, Keller L, Graff C, Lövdén M,
Bäckman L (2014) The benefits of staying active in old age: physical activity counteracts
the negative influence of PICALM, BIN1, and CLU risk alleles on episodic memory func-
tioning. Psychol. Aging 29:440–449.
Filippini N, Ebmeier Kp, Macintosh Bj, Trachtenberg Aj, Frisoni Gb, Wilcock Gk, Beckmann
Cf, Smith Sm, Matthews Pm, Mackay Ce (2011) Differential effects of the APOE geno-
type on brain function across the lifespan. NeuroImage 54:602–610.
Finkel D, Reynolds Ca, Mcardle Jj, Pedersen Nl (2005) The longitudinal relationship between
processing speed and cognitive ability: genetic and environmental influences. Behav
Genet 35:535–549.
Fleisher As, Sherzai A, Taylor C, Langbaum Jb, Chen K, Buxton RB (2009a) Resting-state
BOLD networks versus task-associated functional MRI for distinguishing Alzheimer’s
disease risk groups. NeuroImage 47:1678–1690.
433
Fleisher As, Houston Ws, Eyler Lt, Frye S, Jenkins C, Thal Lj, Bondi Mw (2005) Identification
of Alzheimer disease risk by functional magnetic resonance imaging. Arch Neurol
62:1881–1888.
Fleisher As, Podraza Km, Bangen Kj, Taylor C, Sherzai A, Sidhar K, Liu Tt, Dale Am, Buxton
RB (2009b) Cerebral perfusion and oxygenation differences in Alzheimer’s disease risk.
Ghisletta P, Bäckman L, Bertram L, Brandmaier Am, Gerstorf D, Liu T, Lindenberger U (2014)
The Val/Met polymorphism of the brain-derived neurotrophic factor (BDNF) gene pre-
dicts decline in perceptual speed in older adults. Psychol Aging 29:384–392.
Grady C (2012) The cognitive neuroscience of ageing. Nat Rev Neurosci 13:491–505.
Greenwood Pm, Lin Mk, Sundararajan R, Fryxell Kj, Parasuraman R (2014) Healthy aging
increases the cognitive effects of two genes that influence extracellular dopamine. Psychol
Aging 29:363–373.
Han Sd, Houston Ws, Jak Aj, Eyler Lt, Nagel Bj, Fleisher As, Brown Gg, Corey-Bloom J,
Salmon Dp, Thal Lj, Bondi Mw (2007) Verbal paired-associate learning by APOE geno-
type in non-demented older adults: fMRI evidence of a right hemispheric compensatory
response. Neurobiol Aging 28:238–247.
Hariri Ar, Goldberg Te, Mattay Vs, Kolachana Bs, Callicott Jh, Egan Mf, Weinberger Dr (2003)
Brain-derived neurotrophic factor val66met polymorphism affects human memory-related
hippocampal activity and predicts memory performance. J Neurosci 23:6690–6694.
Harris Se, Deary Ij (2011) The genetics of cognitive ability and cognitive ageing in healthy
older people. Trends Cogn Sci 15:388–394.
Hashimoto R, Moriguchi Y, Yamashita F, Mori T, Nemoto K, Okada T, Hori H, Noguchi H,
Kunugi H, Ohnishi T (2008) Dose-dependent effect of the Val66Met polymorphism of the
brain-derived neurotrophic factor gene on memory-related hippocampal activity. Neurosci
Res 61:360–367.
Heise V, Filippini N, Ebmeier Kp, Mackay Ce (2011) The APOE varepsilon4 allele modulates
brain white matter integrity in healthy adults. Mol Psychiatry 16:908–916.
Ikeda O, Murakami M, Ino H, Yamazaki M, Koda M, Nakayama C, Moriya H (2002) Effects
of brain- derived neurotrophic factor (BDNF) on compression- induced spinal cord
injury: BDNF attenuates down-regulation of superoxide dismutase expression and pro-
motes up-regulation of myelin basic protein expression. Journal Neuropath Exp Neur
61:142–153.
Johnson Sc, Schmitz Tw, Trivedi Ma, Ries Ml, Torgerson Bm, Carlsson Cm, Asthana S,
Hermann Bp, Sager Ma (2006) The influence of Alzheimer disease family history and
apolipoprotein E epsilon4 on mesial temporal lobe activation. J Neurosci 26:6069–6076.
Jonsson Eg, Nothen Mm, Grunhage F, Farde L, Nakashima Y, Propping P, Sedvall Gc (1999)
Polymorphisms in the dopamine D2 receptor gene and their relationships to striatal dopa-
mine receptor density of healthy volunteers. Mol Psychiatry 4:290–296.
Josefsson M, De Luna X, Pudas S, Nilsson Lg, Nyberg L (2012) Genetic and lifestyle predic-
tors of 15-year longitudinal change in episodic memory. J Am Geriatr Soc 60:2308–2312.
Kambeitz Jp, Bhattacharyya S, Kambeitz-Ilankovic Lm, Valli I, Collier Da, Mcguire P (2012)
Effect of BDNF val(66)met polymorphism on declarative memory and its neural sub-
strate: a meta-analysis. Neurosci Biobehav Rev 36:2165–2177.
Kauppi K, Nilsson Lg, Persson J, Nyberg L (2014) Additive genetic effect of APOE and BDNF
on hippocampus activity. NeuroImage 89:306–313.
Kauppi K, Nilsson Lg, Adolfsson R, Eriksson E, Nyberg L (2011) KIBRA polymorphism
is related to enhanced memory and elevated hippocampal processing. J Neurosci
31:14218–14222.
434
Kennedy Km, Rodrigue Km, Land Sj, Raz N (2009) BDNF Val66Met polymorphism influences
age differences in microstructure of the Corpus Callosum. Front Hum Neurosci 3:19.
Klostermann Ec, Braskie Mn, Landau Sm, O’neil Jp, Jagust Wj (2012) Dopamine and fronto-
striatal networks in cognitive aging. Neurobiol Aging 33:623 e615–e624.
Laukka Ej, Lövdén M, Herlitz A, Karlsson S, Ferencz B, Pantzar A, Keller L, Graff C, Fratiglioni
L, Bäckman L (2013) Genetic effects on old-age cognitive functioning: a population-based
study. Psychol Aging 28:262–274.
Li S-C, Chicherio C, Nyberg L, Von Oertzen T, Nagel Ie, Papenberg G, Sander T, Heekeren
Hr, Lindenberger U, Backman L (2010) Ebbinghaus revisited: influences of the BDNF
Val66Met polymorphism on backward serial recall are modulated by human aging. J Cogn
Neurosci 22:2164–2173.
Li S-C, Papenberg G, Nagel Ie, Preuschhof C, Schroder J, Nietfeld W, Bertram L, Heekeren Hr,
Lindenberger U, Backman L (2012) Aging magnifies the effects of dopamine transporter
and D2 receptor genes on backward serial memory. Neurobiol Aging 34: 358.e1–e10.
Lim Yy, Villemagne Vl, Laws Sm, Ames D, Pietrzak Rh, Ellis Ka, Harrington K, Bourgeat P,
Bush Ai, Martins Rn, Masters Cl, Rowe Cc, Maruff P, Group Ar (2014) Effect of BDNF
Val66Met on memory decline and hippocampal atrophy in prodromal Alzheimer’s dis-
ease: a preliminary study. PloS One 9:e86498.
Lind J, Persson J, Ingvar M, Larsson A, Cruts M, Van Broeckhoven C, Adolfsson R, Backman
L, Nilsson Lg, Petersson Km, Nyberg L (2006) Reduced functional brain activity response
in cognitively intact apolipoprotein E epsilon4 carriers. Brain 129:1240–1248.
Lindenberger U, Burzynska Az, Nagel Ie (2013) Heterogeneity in frontal lobe aging. In: Principles
of frontal lobe functions (Stuss Dt, Knight Rt, eds). New York: New York: Oxford
University Press.
Lindenberger U, Nagel Ie, Chicherio C, Li Sc, Heekeren Hr, Bäckman L (2008) Age-related
decline in brain resources modulates genetic effects on cognitive functioning. Front
Neurosci 2:234–244.
Liu F, Pardo Lm, Schuur M, Sanchez-Juan P, Isaacs A, Sleegers K, De Koning I, Zorkoltseva
Iv, Axenovich Ti, Witteman Jc, Janssens Ac, Van Swieten Jc, Aulchenko Ys, Oostra Ba,
Van Duijn Cm (2010) The apolipoprotein E gene and its age-specific effects on cognitive
function. Neurobiol Aging 31:1831–1833.
Lotta T, Vidgren J, Tilgmann C, Ulmanen I, Melen K, Julkunen I, Taskinen J (1995) Kinetics
of human soluble and membrane- bound catechol O- methyltransferase: a revised
mechanism and description of the thermolabile variant of the enzyme. Biochemistry
34:4202–4210.
Madden Dj, Bennett Ij, Burzynska A, Potter Gg, Chen Nk, Song Aw (2012) Diffusion ten-
1822:386–400.
Mahley Rw, Weisgraber Kh, Huang Y (2009) Apolipoprotein E: structure determines function,
from atherosclerosis to Alzheimer’s disease to AIDS. J Lipid Res 50 Suppl:S183–S188.
Matsumoto M, Weickert Cs, Akil M, Lipska Bk, Hyde Tm, Herman Mm, Kleinman Je,
Weinberger Dr (2003) Catechol O-methyltransferase mRNA expression in human and rat
brain: Evidence for a role in cortical neuronal function. Neurosci 116:127–137.
Mier D, Kirsch P, Meyer-Lindenberg A (2010) Neural substrates of pleiotropic action of genetic
variation in COMT: a meta-analysis. Mol Psychiatry 15:918–927.
Milnik A, Heck A, Vogler C, Heinze Hj, De Quervain Dj, Papassotiropoulos A (2012)
Association of KIBRA with episodic and working memory: a meta-analysis. Am J Med
Genet B Neuropsychiatr Genet 159B:958–969.
435
Miyajima F, Ollier W, Mayes A, Jackson A, Thacker N, Rabbitt P, Pendleton N, Horan M,

Payton A (2008) Brain-derived neurotrophic factor polymorphism Val66Met influences
cognitive abilities in the elderly. Genes Brain Behav 7:411–417.
Morse Jk, Wiegand Sj, Anderson K, You Y, Cai N, Carnahan J, Miller J, Distefano Ps, Altar Ca,
Lindsay Rm, et al. (1993) Brain-derived neurotrophic factor (BDNF) prevents the degen-
eration of medial septal cholinergic neurons following fimbria transection. J Neurosci
13:4146–4156.
Muse J, Emery M, Sambataro F, Lemaitre H, Tan Hy, Chen Q, Kolachana Bs, Das S, Callicott
Jh, Weinberger Dr, Mattay Vs (2014) WWC1 genotype modulates age-related decline in
episodic memory function across the adult life span. Biol Psychiatry 75:693–700.
Nagel Ie, Preuschhof C, Li Sc, Nyberg L, Backman L, Lindenberger U, Heekeren Hr (2009)
ing memory. Proc Natl Acad Sci USA 106:22552–22557.
Nagel Ie, Chicherio C, Li S-C, Von Oertzen T, Sander T, Villringer A, Heekeren Hr, Bäckman L,
Lindenberger U (2008) Human aging magnifies genetic effects on executive functioning
and working memory. Front Hum Neurosci 2:1.
Nichols Lm, Masdeu Jc, Mattay Vs, Kohn P, Emery M, Sambataro F, Kolachana B, Elvevåg
B, Kippenhan S, Weinberger Dr, Berman Kf (2012) Interactive effect of apolipoprotein e
genotype and age on hippocampal activation during memory processing in healthy adults.
Arch Gen Psychiatry 69: 804–813.
Nyberg L, Salami A (2014) The APOE epsilon4 allele in relation to brain white-matter micro-
structure in adulthood and aging. Scand J Psychol 55:263–267.
Nyberg L, Lövdén M, Riklund K, Lindenberger U, Bäckman L (2012) Memory aging and brain
maintenance. Trends in cognitive sciences 16:292–305.
Nyberg L, Andersson M, Kauppi K, Lundquist A, Persson J, Pudas S, Nilsson Lg (2014) Age-
related and genetic modulation of frontal cortex efficiency. J Cogn Neurosci 26:746–754.
O’brien Jl, O’keefe Km, Laviolette Ps, Deluca An, Blacker D, Dickerson Bc, Sperling Ra
(2010) Longitudinal fMRI in elderly reveals loss of hippocampal activation with clinical
Okuizumi K, Onodera O, Namba Y, Ikeda K, Yamamoto T, Seki K, Ueki A, Nanko S, Tanaka H,
Takahashi H, Oyanagi K, Mizusawa H, Kanazawa I, Tsuji S (1995) Genetic association of
the very low density lipoprotein (VLDL) receptor gene with sporadic Alzheimer’s disease.
Nat Genetics 11:207–209.
Pantzar A, Laukka Ej, Atti Ar, Papenberg G, Keller L, Graff C, Fratiglioni L, Backman L (2014)
Interactive effects of KIBRA and CLSTN2 polymorphisms on episodic memory in old-
age unipolar depression. Neuropsychologia 62:137–142.
Papassotiropoulos A, Stephan Da, Huentelman Mj, Hoerndli Fj, Craig Dw, Pearson Jv, Huynh
Kd, Brunner F, Corneveaux J, Osborne D, Wollmer Ma, Aerni A, Coluccia D, Hanggi
J, Mondadori Cr, Buchmann A, Reiman Em, Caselli Rj, Henke K, De Quervain Dj
(2006) Common Kibra alleles are associated with human memory performance. Science
314:475–478.
Papenberg G, Bäckman L, Nagel Ie, Nietfeld W, Schroder J, Bertram L, Heekeren Hr,
Lindenberger U, Li Sc (2013a) Dopaminergic Gene Polymorphisms Affect Long-term
Forgetting in Old Age: Further Support for the Magnification Hypothesis. J Cogn Neurosci
25:571–579.
Papenberg G, Bäckman L, Nagel Ie, Nietfeld W, Schroder J, Bertram L, Heekeren Hr,
Lindenberger U, Li Sc (2013b) COMT Polymorphism and Memory Dedifferentiation in
Old Age. Psychol Aging 29:374–383.
436
Papenberg G, Lövdén M, Laukka Ej, Kalpouzos G, Keller L, Graff C, Köhncke Y, Li Tq,

Fratiglioni L, Bäckman L (2014a) Magnified effects of the COMT gene on white-matter
microstructure in very old age. Brain Struct Funct 5:2927–2938.
Papenberg G, Lindenberger U, Bäckman L (2015) Aging-related magnification of genetic
effects on cognitive and brain integrity. Trends Cogn Sci 19:506–514.
Papenberg G, Li Sc, Nagel Ie, Nietfeld W, Schjeide Bm, Schroder J, Bertram L, Heekeren Hr,
Lindenberger U, Bäckman L (2014b) Dopamine and glutamate receptor genes interac-
tively influence episodic memory in old age. Neurobiol Aging 35:1213 e1213–e1218.
Persson J, Rieckmann A, Kalpouzos G, Fischer H, Bäckman L (2014) Influences of a DRD2
polymorphism on updating of long-term memory representations and caudate BOLD
activity: Magnification in aging. Hum Brain Mapp 36:1325–1334.
Pihlajamaki M, K Ok, Bertram L, Tanzi Re, Dickerson Bc, Blacker D, Albert Ms, Sperling Ra
(2010) Evidence of altered posteromedial cortical FMRI activity in subjects at risk for
Alzheimer disease. Alzheimer disease and associated disorders 24:28–36.
Ponomareva N, Andreeva T, Protasova M, Shagam L, Malina D, Goltsov A, Fokin V, Mitrofanov
A, Rogaev E (2013) Age-dependent effect of Alzheimer’s risk variant of CLU on EEG
alpha rhythm in non-demented adults. Front Aging Neurosci 5:86.
Raichlen Da, Alexander Ge (2014) Exercise, APOE genotype, and the evolution of the human
lifespan. Trends in Neurosci 37: 247–255.
Rasch B, Papassotiropoulos A, De Quervain Df (2010) Imaging genetics of cognitive func-
tions: Focus on episodic memory. NeuroImage 53:870–877.
Raz N, Lindenberger U, Rodrigue Km, Kennedy Km, Head D, Williamson A, Dahle C, Gerstorf
D, Acker Jd (2005) Regional brain changes in aging healthy adults: general trends, individ-
ual differences and modifiers. Cereb Cortex 15:1676–1689.
Rönn T, Volkov P, Davegardh C, Dayeh T, Hall E, Olsson Ah, Nilsson E, Tornberg A, Nitert Md,
Eriksson Kf, Jones Ha, Groop L, Ling C (2013) A six months exercise intervention influ-
ences the genome-wide DNA methylation pattern in human adipose tissue. PLoS Genet
9:e1003572.
Rönnlund M, Nyberg L, Bäckman L, Nilsson Lg (2005) Stability, growth, and decline in adult
population-based study. Psychol Aging 20:3–18.
Sambataro F, Reed Jd, Murty Vp, Das S, Tan Hy, Callicott Jh, Weinberger Dr, Mattay Vs (2009)
Catechol-O-methyltransferase valine(158)methionine polymorphism modulates brain net-
works underlying working memory across adulthood. Biol Psychiatry 66:540–548.
Sambataro F, Murty Vp, Lemaitre Hs, Reed Jd, Das S, Goldberg Te, Callicott Jh, Weinberger
Dr, Mattay Vs (2010) BDNF modulates normal human hippocampal ageing [corrected].
Mol Psychiatry 15:116–118.
Sanchez Mm, Das D, Taylor Jl, Noda A, Yesavage Ja, Salehi A (2011) BDNF polymorphism
predicts the rate of decline in skilled task performance and hippocampal volume in healthy
individuals. Transl Psychiatry 1:e51.
Sapkota S, Vergote D, Westaway D, Jhamandas J, Dixon Ra (2014) Synergistic associations of
COMT and BDNF with executive function in aging are selective and modified by APOE.
Schaie Kw, Maitland Sb, Willis Sl, Intrieri Rc (1998) Longitudinal invariance of adult psycho-
metric ability factor structures across 7 years. Psychol Aging 13:8–20.
Schiepers Oj, Harris Se, Gow Aj, Pattie A, Brett Ce, Starr Jm, Deary Ij (2012) APOE E4 status
predicts age-related cognitive decline in the ninth decade: longitudinal follow-up of the
Lothian Birth Cohort 1921. Mol Psychiatry 17:315–324.
Schneider A, Huentelman Mj, Kremerskothen J, Duning K, Spoelgen R, Nikolich K (2010)
KIBRA: A New Gateway to Learning and Memory? Front Aging Neurosci 2:4.
437
Schuck Nw, Doeller Cf, Schjeide Bm, Schroder J, Frensch Pa, Bertram L, Li Sc (2013) Aging
and KIBRA/WWC1 genotype affect spatial memory processes in a virtual navigation task.
Hippocampus 23:919–930.
Schwab Lc, Luo V, Clarke Cl, Nathan Pj (2014) Effects of the KIBRA Single Nucleotide
Polymorphism on Synaptic Plasticity and Memory: A Review of the Literature. Curr
Neuropharmacol 12:281–288.
Sweatt Dj (2013) The emerging field of neuroepigenetics. Neuron 80:624–632.
Slifstein M, Kolachana B, Simpson Eh, Tabares P, Cheng B, Duvall M, Frankle Wg, Weinberger
Dr, Laruelle M, Abi-Dargham A (2008) COMT genotype predicts cortical-limbic D1
receptor availability measured with [11C]NNC112 and PET. Mol Psychiatry 13:821–827.
Small Bj, Rosnick Cb, Fratiglioni L, Backman L (2004) Apolipoprotein E and cognitive perfor-
mance: a meta-analysis. Psychol Aging 19:592–600.
Small Bj, Rawson Ks, Walsh E, Jim Hs, Hughes Tf, Iser L, Andrykowski Ma, Jacobsen Pb
(2011) Catechol-O-methyltransferase genotype modulates cancer treatment-related cogni-
tive deficits in breast cancer survivors. Cancer 117:1369–1376.
Smith Jd (2002) Apolipoproteins and aging: emerging mechanisms. Ageing Res Rev
1:345–365.
Stuart K, Summers Mj, Valenzuela Mj, Vickers Jc (2014) BDNF and COMT polymorphisms
have a limited association with episodic memory performance or engagement in complex
cognitive activity in healthy older adults. Neurobiol Learn Mem 110:1–7.
Thorvaldsson V, Macdonald Sw, Fratiglioni L, Winblad B, Kivipelto M, Laukka Ej, Skoog I,
Sacuiu S, Guo X, Ostling S, Borjesson-Hanson A, Gustafson D, Johansson B, Backman
L (2011) Onset and rate of cognitive change before dementia diagnosis: findings from
two Swedish population-based longitudinal studies. J Int Neuropsychol Soc 17:154–162.
Toga Aw, Thompson Pm (2005) Genetics of brain structure and intelligence. Annu Rev Neurosci
28:1–23.
Turkheimer E (2011) Still Missing. Res Hum Dev 8:227–241.
Tucker-Drob Em, Reynolds Ca, Finkel D, Pedersen Nl (2014) Shared and unique genetic and
environmental influences on aging-related changes in multiple cognitive abilities. Dev
Psychol 50:152–166.
Westlye Lt, Reinvang I, Rootwelt H, Espeseth T (2012) Effects of APOE on brain white matter
microstructure in healthy adults. Neurology 79:1961–1969.
Wilson Rs, Barral S, Lee Jh, Leurgans Se, Foroud Tm, Sweet Ra, Graff-Radford N, Bird Td,
Mayeux R, Bennett Da (2011) Heritability of different forms of memory in the Late Onset
Alzheimer’s Disease Family Study. J Alzheimers Dis 23:249–255.
Wisdom Nm, Callahan Jl, Hawkins Ka (2011) The effects of apolipoprotein E on non-impaired
cognitive functioning: a meta-analysis. Neurobiol Aging 32:63–74.
Witte Av, Flöel A (2011) Effects of COMT polymorphisms on brain function and behavior in
health and disease. Brain Res Bull 88:418–428.
Xu G, Mclaren Dg, Ries Ml, Fitzgerald Me, Bendlin Bb, Rowley Ha, Sager Ma, Atwood C,
Asthana S, Johnson Sc (2009) The influence of parental history of Alzheimer’s disease
and apolipoprotein E epsilon4 on the BOLD signal during recognition memory. Brain
132:383–391.
Zhang C, Pierce Bl (2014) Genetic susceptibility to accelerated cognitive decline in the US
Health and Retirement Study. Neurobiol Aging 35:1512 e1511–e1518.
438
439
17
Effects of Exercise on Cognition,

Brain Structure, and Brain Function in
Older Adults
Kirk I. Erickson*
Lauren E. Oberlin
T he brain is inherently plastic and is altered by stimulation from the

environment (Kolb and Muhammad, 2014). The capability for brain
plasticity is thought to vary across the lifespan with a greater propensity for plas-
ticity earlier in life than in later years. Nonetheless, there has been great interest in
examining factors that may influence the potential for brain plasticity in late adult-
hood, especially given the lack of successful medications to treat or prevent dementia
and normal age-related cognitive decline (Lovden et al., 2013). This has become an
increasingly relevant and challenging issue given the expected increase in the number
of adults over the age of 65 in many countries around the world and the anticipation
that this increase will result in a concomitant increase in age-related diseases, includ-
ing dementias (Erickson et al., 2013a).
Fortunately physical activity is one of the most promising methods for positively
influencing brain structure and function in late adulthood. As will be described in
this chapter, there is now overwhelming evidence that physical activity is effective at
influencing the brain throughout the lifespan. Yet, the majority of evidence for effects
of physical activity on brain function focuses on late adulthood when risk for both car-
diovascular disease and cognitive decline escalates. As the close affiliation between
cardiovascular diseases and brain disorders becomes more and more recognized, it is
clear that reducing cardiovascular risks through increased physical activity might also
lead to reducing risks for cognitive losses. Yet, despite the growing awareness that
increased physical activity is useful for both the heart and the brain, there are many
439
440
remaining questions that have yet to be answered. We will address these and related
issues throughout the chapter.
Experimental Designs and Exercise Effects on Cognitive Outcomes
Spirduso and Clifford (1978) conducted some of the earliest research examining the
association among physical activity, cardiovascular fitness, and cognitive outcomes
in older adults. In a series of studies, they found that higher fit and more active older
adults tended to outperform their more sedentary peers on a variety of cognitive and
psychomotor tasks (Spirduso, 1975). This seminal finding was critical for launch-
ing exercise and cognition research, as it was not only highly consistent with animal
research on the beneficial effects of environmental enrichment on brain and cognitive
measures, but also was one of the first studies to suggest that maintaining a physi-
cally active lifestyle might be an important factor for influencing cognitive function
in late adulthood. In fact, this finding has been frequently replicated in other cross-â•‰
sectional studies. In this chapter, the term cross-â•‰sectional means a study design in
which a single snapshot of physical activity habits or fitness levels is measured on
individuals and correlated with cognitive performance or brain outcomes (Etnier et
al., 2006). This type of design can be contrasted with prospective epidemiological
studies, which often follow individuals for an extended period and examine how phys-
ical activity habits or fitness levels are related to changes in cognitive function or risk
for dementia over the measured interval. For example, in a meta-â•‰analysis of 15 pro-
spective epidemiological studies that followed 33,816 adults without dementia, Sofi
et al. (2011) found that greater amounts of physical activity was associated with a
~38% reduced risk of experiencing cognitive decline as measured by changes in
performÂ�ance on neuropsychological tests. This and other similar studies and meta-â•‰
analyses make a strong case for an association between physical activity and cogni-
tive function (Barnes and Yaffe, 2011). However, causality cannot be conclusively
determined from either cross-â•‰sectional or prospective epidemiological studies. For
example, it is possible that older adults that show less cognitive decline are more
physically capable of engaging in physical activity, and that continued participation in
physical activity in late adulthood is simply a sign of overall better physical and cog-
nitive health rather than a causal factor that leads to better cognitive function. It is with
this in mind that randomized controlled trials of exercise have been developed to more
closely examine the causal role between physical activity and cognitive function.
In the context of physical activity, the term randomized controlled trial refers to
an experimental design in which participants are randomly assigned to either receive
physical activity or to a less physically active (or non–â•‰physically active) control con-
dition for some period of time. The overarching goal of a randomized controlled trial
is to manipulate the amount, frequency, or intensity of physical activity in order to test
whether such a manipulation influences the dependent variable, which in this case is
either cognitive or brain outcomes. Similar to any experimental design, investigators
designing a randomized controlled trial face many choices in the properties of the
study that could lead to lack of experimental control, poor data quality, and erroneous
interpretations. For example, in designing a randomized controlled trial of physical
441
Effects of Exercise on Cognition, Brain Structure and Brain Function 441
activity, investigators must choose the type of activity, how and to what extent the
activity will be monitored, the intensity and frequency of the activity, and on what
factors the control group will be matched. Given the number of variables to choose
from, it should be no surprise that randomized controlled trials of physical activity
often differ in how they are conducted, and, as such, the quality of the results and
the validity of the interpretations can also vary dramatically. Nonetheless, random-
ized controlled trials of exercise are considered the “gold s tandard” of experimental
designs that attempt to examine whether physical activity plays an important causal
role in cognitive performance or brain health.
Unfortunately, randomized controlled trials of exercise are generally more costly
and time-consuming than cross-sectional or observational studies (i.e., studies not
manipulating physical activity levels). As such, there are fewer randomized trials
examining the effects of physical activity on cognitive outcomes as compared to the
number of cross-sectional and observational studies. Randomized trials of physical
activity have been conducted using several approaches, each with its own set of lim-
itations. First, most of the randomized trials are single-blind designs such that partic-
ipants generally know if they have been assigned to the physical activity treatment
group or to the control group, but the experimenters themselves are usually blind to
which participants are in which group. Second, many studies utilize a stretching-and-
toning control group, which usually tries to equate the amount of social interaction
in treatment and control groups but differs in the level and type of physical activity.
Other studies, however, have utilized “usual care” control groups or education control
groups, which purposefully avoid having participants engage in any physical activity.
The choice of the most appropriate control group depends on the questions being
asked by the researchers—and great thought should be given to what type of control
should be included, or even whether multiple control groups might be most effective.
Similar challenges regarding the length or amount of physical activity treatment
experimental subjects receive is also an area of concern. Specifically, when designing
a randomized controlled trial of exercise, researchers need to consider the intensity of
the physical activity, the frequency of the intervention, how to maintain adherence and
compliance, ways to avoid frustration and fatigue among participants, the duration of
the intervention, and how to best monitor and test the fidelity of the intervention. Most
of the studies discussed in this chapter have not adhered to any hard and fast rules of
exercise, but have generally utilized a moderate-intensity intervention based on close
monitoring of heart rate. Such interventions are often delivered for a period of 3–6
months and generally occur for 2–4 days per week.
Results from randomized trials on cognitive outcomes in late adulthood have been
largely positive, but slightly more variable than results from observational studies.
Again, by “observational” we mean here a study design that does not attempt to
manipulate or alter physical activity levels in participants, but just examines either
single snapshots of fitness levels or longitudinally follows individuals for some period
of time. In this context, some randomized controlled studies have failed to find signif-
icant effects of exercise on cognitive outcomes while others have reported more pos-
itive findings. Such variability might be due to the types of cognitive tasks employed
or the design of the exercise intervention (e.g., intensity of exercise)—as we will see,
both play a critically important role. In a seminal paper, Kramer et al. (1999) reported
442
that a 6-month randomized physical activity intervention in older adults significantly

improved performance on tasks that measured executive function including response
compatibility and switching, but not for tasks that were low in executive function.
Here, executive function is used as a broad term reflecting cognitive processes that
demand attentional or inhibitory control, working memory, task coordination, or rapid
switching between tasks. Executive function is thought to be primarily supported by
the prefrontal cortex along with white matter tracts that provide communicating paths
to medial temporal and parietal cortices.
The results from Kramer et al. (1999) led to a “selective improvement hypothe-
sis” that predicted that executive functions would be more influenced by exercise than
other “non-executive” cognitive domains. This hypothesis has been largely supported,
at least for older adults, across a number of studies and meta-analyses. For example,
Colcombe and Kramer (2003) conducted a meta-analysis of 18 randomized controlled
trials of physical activity and found that physical activity was effective at improving
performance across all cognitive domains, but that larger effect sizes were found for
tests of executive function than for other cognitive domains. This specificity of effects
on executive function is interesting from several perspectives (Hindin and Zelinski,
2012). First, the results suggest that the same cognitive domains that show early and
precipitous losses in function with increasing age are the same ones that may be most
amenable to an exercise intervention. For example, age-related losses in executive
function may precede changes in episodic memory, language processes, or visuo-spa-
tial processing, and it is executive functions that appear the most receptive to change
from an exercise intervention. Second, the results provide an answer to why some
studies fail to find significant effects. That is, some studies may fail to find significant
effects because they do not include adequate testing of executive function. For exam-
ple, Bakken et al. (2001) conducted an 8-week randomized exercise trial in 15 older
adults, but focused on verbal fluency, verbal discrimination, and perceptual paradigms
rather than executive function. They reported nonsignificant effects of the intervention
on cognitive outcomes except for small improvements on a perceptual finger-tracking
test. The lack of significant effects could be related to the lack of executive function
measures in the cognitive battery. Third, the Kramer et al. (1999) results led to the pre-
diction that the brain areas that support executive function (i.e., frontal cortex) might
be more strongly influenced by physical activity than other brain areas (i.e., parietal
cortex) and that physical activity may not have a uniform effect across all brain areas.
The results from the meta-analysis by Colcombe and Kramer (2003) have been
replicated in more recent meta-analyses, but across the lifespan (Smith et al., 2010).
However, the results from these meta-analyses also indicate a few additional issues
relevant for both the prescription of physical activity and for interpreting the results
from randomized trials. First, Colcombe and Kramer (2003) found that effect sizes
tended to increase when studies with more women were included. This suggests a
possible moderating effect of gender, which could be contributing to heterogeneity
across studies. Second, meta-analyses have also reported that studies 6 months or
longer show larger effect sizes than interventions of shorter durations (Colcombe and
Kramer, 2003). This might seem intuitive, but studies with shorter durations may be
too short to significantly elicit changes in brain health and cognitive function, which
may also help explain some of the heterogeneity across studies.
â•‡ 443
Effects of Exercise on Cognition, Brain Structure and Brain Functionâ•… 443
In sum, we can conclude that results from randomized trials largely support the
results from cross-â•‰sectional and prospective epidemiological studies and indicate
that moderate-intensity physical activity several days per week over a several-month
period is an effective approach to improve cognitive function, especially executive
functioning, in late adulthood.
Exercise and Gray Matter Volume Outcomes
Evidence that executive functions are more affected by physical activity than other
cognitive domains led to the hypothesis that brain areas that help support executive
functions (i.e., prefrontal cortex, medial temporal lobe, and connecting white matter
paths) would be more affected by exercise than other brain areas that are less associ-
ated with executive function (i.e., thalamus, occipital cortex, lateral temporal regions).
The first study to test this hypothesis was cross-â•‰sectional and used voxel-â•‰based mor-
phometry (VBM) to segment and examine T1-â•‰weighted magnetic resonance imag-
ing (MRI) structural gray matter and white matter data from 55 older adults that
had cardiorespiratory fitness data (Colcombe et al., 2003). Cardiorespiratory fitness
is an objective measure of fitness and assesses aerobic capacity. Cardiorespiratory
fitness is modifiable by participation in regular exercise and is often used in two
ways: (a) as a cross-â•‰sectional measure of aerobic capacity and regular engagement in
physical activity, and (b) as a way to assess the efficacy of an exercise intervention.
In this study (Colcombe et al., 2003), the authors conducted a linear regression anal-
ysis between the VBM results and cardiorespiratory fitness levels of the participants
and found that higher fitness levels offset an age-â•‰related decline in both gray and
white matter volume. Furthermore, consistent with the hypothesis that the associa-
tion would not be uniform throughout the entire brain, they found that prefrontal and
temporal brain regions, including along the frontal medial wall and anterior cingulate
cortex (ACC), were more strongly associated with fitness than other brain areas (e.g.,
visual cortex).
As described previously, cross-â•‰sectional studies that take single snapshots of fit-
ness and exercise habits are typically much easier to conduct and less expensive than
randomized controlled trials. As such, there have now been many other cross-â•‰sec-
tional studies that take a single snapshot of physical activity habits or fitness levels,
but fewer randomized trials of exercise, demonstrating associations with gray and/â•‰or
white matter volumes (see Erickson et al., 2014 for a more comprehensive review).
For example, Bugg and Head (2011) used FreeSurfer to segment gray matter volume
in a sample of 52 adults with a mean age of 69 and found that higher fitness levels
were associated with greater gray matter volume in the frontal lobe and that greater
engagement in exercise also offset an age-â•‰related loss of volume in the medial tempo-
ral lobe. Similarly, Floel et al. (2010) found in 75 adults with a mean age of 60.5 that
greater amounts of self-â•‰reported physical activity were associated with greater gray
matter volume in the frontal cortex. These and other cross-â•‰sectional studies of phys-
ical activity and fitness suggest that the prefrontal cortex may be especially sensitive
to fitness and exercise, which is highly consistent with the hypotheses generated from
research on the effects of exercise on cognitive function. However, only longitudinal
444
or randomized controlled trials are capable of testing whether the prefrontal cortex is
modifiable by engaging in regular exercise.
To test whether a randomized exercise intervention would influence the volume
of the prefrontal cortex and other brain areas, Colcombe et al. (2006) used VBM to
assess gray matter volume in 59 adults ranging between 60 and 79 years of age that
had been randomized to either a stretching-and-toning control condition or to a mod-
erate-intensity exercise intervention 3 days per week for 6 months. Consistent with
their hypothesis, the exercise intervention increased the volume of the frontal cortex,
along the medial wall and ACC, and in the lateral temporal lobes. The results from this
study were important as they demonstrated that older adult brains remain relatively
plastic and that only 6 months of exercise may be capable of taking advantage of this
brain property (also see Ruscheweyh et al., 2011).
In addition to the effects of fitness and exercise on the frontal cortex, research
has now demonstrated consistent effects on the size of the hippocampus, another
region that is susceptible to age-related atrophy, involved in memory formation and
Alzheimer’s disease (AD). In the first study testing this association, Erickson et al.
(2009) found in 165 adults with a mean age of 66.5 that higher cardiorespiratory
fitness levels were associated with larger left and right hippocampal volumes and
that larger hippocampal volumes translated to better spatial memory performance.
This finding has now been replicated in other cross-sectional studies in older adults
(Bugg et al., 2012; Head et al., 2012). Again, however, associations reported in cross-
sectional studies that take a single snapshot of fitness rather than manipulating fitness
or exercise in an intervention, are limited in drawing causal conclusions about the
effects of exercise on brain health.
The hippocampus, located in the medial temporal lobe, is an important brain struc-
ture with respect to aging. Longitudinal studies have reported that in adults over the
age of 60 the hippocampus deteriorates between 1%–2% per year and that this rate
increases in adults with AD (Raz et al., 2005; Raz et al., 2010; Suzuki et al., 2013;
Petersen et al., 2014). We note that animal studies have unequivocally found benefi-
cial effects of exercise on the structure and function of the hippocampus including
the production of brain-derived neurotrophic factor (BDNF) (van Praag et al., 2005).
However, until recently, human data were lacking. To test whether exercise is capa-
ble of modifying the size of the hippocampus, Erickson et al. (2011) conducted a
randomized intervention trial of 120 adults with 60 in a stretching-and-toning con-
trol condition and 60 in a moderate-intensity walking condition for 12 months. Both
groups began the intervention with equivalent sizes of the hippocampus, but after
12 months of brisk walking the treatment group showed an increase in the size of
the hippocampus while the stretching-and-toning control group showed a reduction
(see Figure 17.1). Furthermore, increase in hippocampal volume was correlated with
improvements on a spatial memory task. These findings indicate that the hippocampus
retains its capacity for plasticity and that exercise is capable of modifying the size of
the structure, even in late adulthood. Importantly, another 12-month study of exercise
and motor coordination training replicated these findings (Niemann et al., 2014), but a
study of 3 m onths of exercise in adults with depression found no significant effects on
the size of the hippocampus (Krogh et al., 2014). These patterns suggest that it might
445
LEFT HIPPOCAMPUS RIGHT HIPPOCAMPUS
(A) Hippocampus 5.2 5.2
5.1 5.1
5 5
4.9 4.9
4.8 4.8
4.7 4.7
4.6 4.6
(B) Caudate nucleus

LEFT CAUDATE NUCLEUS RIGHT CAUDATE NUCLEUS
4.9 5.3
4.8 5.2
4.7 5.1
4.6 5
4.5 4.9
4.4 4.8
(C) Thalamus THALAMUS
15.00 Exercise
Stretching
14.50
14.00
13.50
13.00
Baseline 6 months 1 year
Figure 17.1 Results from a 12-month exercise intervention in which 120 older adults were randomized to either
a walking exercise condition or to a stretching control condition. The results demonstrated that there were no
significant changes in the size of either the thalamus or caudate nucleus with the intervention, but there were significant
increases in the size of the hippocampus for the walking exercise group. Adapted from Erickson et al. (2011).
446
take longer than 3 months to be able to observe significant changes in the size of the
hippocampus—â•‰at least in older adults.
There is mounting evidence that exercise is capable of modifying both the pre-
frontal cortex and hippocampus in late adulthood (randomized trials) and that indi-
viduals that report greater amounts of physical activity or have higher fitness levels
have larger volumes than their lesser fit or inactive peers (cross-â•‰sectional studies). In
addition to this body of work, several longitudinal observational studies have reported
that engaging in physical activity reduces the loss of gray matter volume over a sev-
eral-year span. For example, Erickson et al. (2010) examined self-â•‰reported physi-
cal activity in 299 individuals with a mean age of 78 without cognitive impairment
and then examined gray matter volume 9 â•‰years after the assessment. They found that
greater amounts of self-â•‰reported physical activity were associated with greater pre-
frontal cortex and hippocampal volume 9 â•‰years after the assessment. Furthermore,
greater volume of these structures was associated with a two â•‰fold reduced risk of
experiencing cognitive impairment 4 â•‰years after the imaging assessment (also see
Rovio et al., 2010b; Gow et al., 2012b). These results suggest that engaging in phys-
ical activity may have long-â•‰term consequences for preserving the hippocampus and
other brain areas from the development of dementia.
In summary, there are now numerous publications from both cross-â•‰sectional, lon-
gitudinal observational, and randomized trials that demonstrate associations between
exercise, fitness, and gray matter volume. These studies suggest that the prefrontal
cortex and hippocampus are consistently more affected relative to other brain areas.
Although these results are consistent with predictions made from research on cog-
nitive function, the reason for the regional specificity remains a matter of debate.
Nonetheless, the studies described in this section, including others on AD and other
at-â•‰risk populations (Honea et al., 2009; Makizako et al., 2014; Ten Brinke et al., 2014),
suggests that exercise may be a relatively low-â•‰cost, highly accessible, and effective
method of influencing gray matter volume and function in late adulthood.
The Effects of Exercise on White Matter Integrity
Prior research demonstrates that exercise may be capable of increasing gray matter
volume, particularly in the hippocampus and prefrontal cortex. This increase, or spar-
ing, of gray matter volume may account partially for the improvements in executive
functioning. Yet, executive processes are not localizable to one specific brain region,
but rather require complex communication among temporal, parietal, and subcortical
structures (Madden et al., 2009). Therefore, white matter, which consists primarily of
glia and myelinated axons, is necessary to facilitate communication within and across
neural networks and to support cognitive function.
Along with changes in gray matter, normal aging is also accompanied by alterations
in white matter (Westlye et al., 2009). White matter degeneration occurs in regions
that support higher-â•‰level cognitive processes, with tract disruption following an ante-
rior to posterior gradient (Jernigan et al., 2001; Head et al., 2004). Notably, compro-
mised white matter circuitry predicts impairments in cognitive functioning in normal
aging (Charlton et al., 2006; Gold et al., 2010) and is part of the neuropathology
447
of age-related neurodegenerative diseases, including AD, vascular dementia, and

Parkinson’s disease (Barber et al., 1999; Bartzokis et al., 2004; Bohnen and Albin,
2011).
Physical activity may modify the age-related decline in white matter tissue. While
the existing body of literature assessing the relationship between physical activity
and white matter structure is relatively sparse, particularly in relation to similar stud-
ies involving gray matter volume, cumulative evidence suggests that there is greater
white matter tissue structure with higher levels of physical activity.
There are multiple ways to assess white matter in vivo, with each modality pro-
viding fundamentally distinct information about white matter architecture. One tech-
nique involves the quantification of white matter hyperintensities (WMH). These
white matter lesions indicate focal areas of white matter damage. Enhanced presence
and/or severity of WMH predicts impairments in cognition as well as dementia risk
(Barber et al., 1999; Debette and Markus, 2010). At present, few studies have assessed
the relation between WMH and physical activity, with existing literature citing mixed
results. Using a prospective epidemiological design as described in preceding sec-
tions, Rovio et al. (2010a) assessed the relationship between self-reported leisure time
physical activity in midlife and development of WMH 21 years later in 75 adults. At
follow-up, the sedentary group (physical activity < 2 times/wk) had a greater risk of
developing WMH than their more active peers, but this association did not survive
adjustments for demographic and vascular factors. Another prospective epidemiolog-
ical study (Podewils et al., 2007) was in agreement, citing no association between
self-reported physical activity and rate of white matter lesion progression 5 years later
(also see Rosano et al., 2012).
Yet, results from several other studies have provided evidence for an association
between physical activity and WMH. Tseng et al. (2013) assessed the relationship
between physical activity and WMH in a sample of 10 older master’s athletes (endur-
ance training > 15 years) and 10 older sedentary subjects. Relative to their seden-
tary counterparts, master’s athletes showed an 83% reduction in deep WMH volume.
Further support for this association was established in a cross-sectional physical activ-
ity study (Gow et al., 2012a), which found that higher levels of self-reported physical
activity were associated with reduced white matter lesion load among 691 cognitively
healthy older adults. Taken together, the literature on WMH and physical activity
is mixed, with recent evidence demonstrating a link between physical activity and
WMH. The studies highlighted above are limited by small sample sizes (Rosano et
al., 2010; Tseng et al., 2013) and subjective self-reports of physical activity (Podewils
et al., 2007; Rovio et al., 2010b; Gow et al., 2012a; Rosano et al., 2012), which may
partially contribute to incongruous findings. Also, of note, lesion analysis provides
gross estimates of focal white matter damage and is not a measure of diffuse or micro-
structural white matter changes.
White matter can also be measured volumetrically, with some evidence demonstrat-
ing a link between white matter volume and physical activity or fitness (Colcombe et
al., 2003; Colcombe et al., 2006; Burns et al., 2008; Ho et al., 2011; Benedict et al.,
2013). For instance, adults aged 60–79 showed an increase in white matter volume
following 6 months of a moderate-intensity aerobic exercise intervention. In this sam-
ple, volumetric changes were observed in the genu and the anterior body of the corpus
448
callosum, areas which facilitate interhemispheric communication between anterior

brain regions (Colcombe et al., 2006). Fitness-related associations have also been
established in samples with early AD. For example, Burns et al. (2008) found that
higher fitness levels were predictive of greater whole brain white matter volume in a
sample with early AD. However, this association was not found among similarly aged
cognitively healthy adults (also see Ho et al., 2011). Other studies have also failed
to demonstrate a relationship between fitness or physical activity and white matter
volume. For instance, in a study by Bugg and Head (2011) exercise engagement did
not account for a significant amount of variance in cortical white matter volume,
although only global white matter volume was explored in this sample. Similarly,
using a voxel-wise approach, Gordon et al. (2008) did not observe an association
between fitness and white matter atrophy in a sample of healthy older adults. Thus, the
report by Colcombe et al. (2006) provides a causal link between white matter volume
and exercise, an association which is further supported by some cross-sectional work
that only measures activity or fitness at one point in time (Ho et al., 2011). But, the
implications of these findings are tempered by the lack of other studies showing an
association between physical activity and white matter volume. Methodological dif-
ferences across studies in the quantification of white matter volume may be partially
responsible for disparate findings, as studies varied by whether they used a region of
interest approach, voxel-wise assessment, or simply a global estimate of white matter
volume.
Taken together, evidence suggests that greater levels of physical activity may be
associated with fewer focal lesions and reduced white matter atrophy, although the
scope of interpretation is constrained by discordant results. Unlike the techniques
described above, diffusion tensor imaging (DTI) may be more sensitive to subtle alter-
ations or abnormalities in white matter tissue properties and may be able to detect
variations in white matter microstructure. DTI measures white matter structure by
estimating the rate and directionality of water diffusion along white matter fibers.
Water diffusion will be anisotropic, or directionally homogenous, when restricted by
barriers such as axons, neurofilaments, and myelin (Hagmann et al., 2006). In con-
trast, when the microstructure is compromised white matter becomes less constrained
and therefore more isotropic. Fractional anisotropy (FA) is the most common esti-
mate of white matter structure, and is an overall measure of anisotropy, or directional
dependence of diffusion, in selected tissue (Pfefferbaum and Sullivan, 2002; Vernooij
et al., 2009).
Although limited in number, studies using DTI have provided converging evidence
of favorable effects of physical activity on white matter microstructure (Marks et al.,
2010; Gow et al., 2012a; Gons et al., 2013). Gons et al. (2013) examined the asso-
ciation between leisure-time physical activity and white matter structure among 440
adults aged 50–85 with cerebral small vessel disease. They found an increase in FA
with higher levels of physical activity in almost all voxels of the white matter skel-
eton, even after adjusting for age, sex, and cardiovascular risk factors. Additionally,
Gow et al. (2012a) assessed self-reported physical activity from 691 adults aged 70,
and followed up with MRI scans three years later. Higher levels of physical activ-
ity at baseline predicted greater white matter structure three years later. While these
large-scale studies suggest an inverse relationship between physical activity and white
â•‡ 449
matter degeneration, they are limited by the use of subjective assessments of physical
activity, which are prone to reporter bias.
To test whether objective fitness estimates were associated with microstructural
white matter measures, Marks et al. (2010) found that higher cardiorespiratory fit-
ness was associated with greater FA in the middle cingulum segment and explained
28.5% of the total variance of FA in this region. These findings are in agreement
with other research, which has demonstrated positive associations between fitness and
white matter structure in the corpus callosum (Johnson et al., 2012), the superior lon-
gitudinal fasciculus, the inferior fronto-â•‰occipital fasciculus, and the superior corona
radiata (Tseng et al., 2013). These cross-â•‰sectional results suggest that higher fitness
levels may be able to offset age-â•‰related degeneration in white matter tissue, but these
studies include sample sizes of 26 or fewer, restricting generalizability. This limitation
was recently addressed (Tian et al., 2014), in a study with 164 adults over the age
of 80. Their findings indicated that higher cardiorespiratory fitness was associated
with greater FA in the cingulum, a medial tract with extensive cortical and subcorti-
cal connections, which facilitates signal transmission between networks involved in
attention, memory, and emotion. Thus, increased objective levels of fitness predict
greater anisotropic diffusion in white matter tissue, indicating enhanced structural
soundness of axons.
To date, we are aware of only one randomized controlled trial that has examined
changes in white matter microstructure following aerobic exercise training. Briefly,
70 cognitively healthy older adults completed either 12 â•‰months of structured stretch-
ing and toning or aerobic exercise training. DTI data was collected pre and post-â•‰inter-
vention to examine changes in white matter as a function of the intervention. Voss et
al. (2013) found that gains in aerobic fitness within the exercise group were associated
with enhanced memory performance, as well as significant increases in prefrontal and
temporal FA. This result was not found in the stretching and toning control group,
suggesting that enhanced cardiorespiratory fitness following one â•‰year of moderate-in-
tensity aerobic exercise is predictive of greater focal white matter microstructure.
The results highlighted above suggest that physical activity may preserve white
matter structure in older adulthood, thus promoting efficient and effective signal
transmission within neural systems. Unfortunately, other measures of white mat-
ter have yielded conflicting results, suggesting that physical activity may induce
microstructural changes in white matter tissue not detectable by volumetric or lesion
assessments.
Exercise on Task-â•‰Evoked fMRI Patterns
We have now outlined the evidence for the effects and associations between exer-
cise and cardiorespiratory fitness on gray matter volume, white matter structure, and
cognitive function. In addition, several studies have utilized task-â•‰evoked functional
MRI (fMRI) patterns to examine how exercise influences the functioning of particular
brain systems under a cognitive challenge.
In one of the first studies, Colcombe et al. (2004) conducted a randomized inter-
vention on 41 older adults that were assigned to receive moderate-intensity exercise
450
or to a stretching-â•‰and-â•‰toning control group for 6 â•‰months. They used a version of the

response compatibility (flanker) task in which participants were asked to respond to
the direction of a center arrow in a series of five arrows. Half of the trials were “incon-
gruent” such that the center arrow pointed in the opposite direction of the four flank-
ing arrows, while the other half of the trials were “congruent” such that the center
arrow and flanking arrows were all pointing in the same direction. This paradigm is
thought to rely on attentional control, or the active allocation of attentional resources
to focus attention, in this case on the direction of the center arrow, and is often used
as an exemplar of executive function. The results indicated that exercise increased
brain activity during the incongruent task condition in the lateral prefrontal cortex and
parietal cortex, while reducing conflict-â•‰related activity in the ACC. Importantly, these
changes corresponded to improved performance on the task. In another study, Smith
et al. (2013) enrolled older adults that were classified as either cognitively normal or
cognitively impaired in a 12-â•‰week exercise intervention. They found that 11 different
regions showed consistent changes in activation (decreases) during a semantic mem-
ory task regardless of the degree of cognitive impairment. Since performance on the
task increased they interpreted the results as being consistent with a model of neural
efficiency (also see Chapman et al., 2013). These randomized trials are in line with
results from cross-â•‰sectional studies. For example, Smith et al. (2011) found that self-â•‰
reported physical activity was associated with elevated levels of activation during a
semantic memory paradigm in the caudate nucleus in individuals with amnestic mild
cognitive impairment (also see Prakash et al., 2011).
In sum, although few studies have examined the impact of exercise on task-â•‰evoked
fMRI, the results are promising in that detectable changes in brain activity have been
observed that correspond to improvements in cognitive performance. However, most
of these studies have not controlled for individual variations in blood flow or structure
that might change as a result of an exercise intervention. Another related issue is that
there are both increases and decreases in activation after exercise that could be inter-
preted in several different ways (i.e, does a decrease represent increased efficiency
or a decline in function?). Hence, this field will require more research with clearer
hypotheses about the direction of the effects, the regions affected, and the tasks that
might elicit the most consistent changes in activation.
The Impact of Exercise on Intrinsic Brain Connectivity
The corpus of the work discussed so far emphasizes the favorable effects of exercise on
cognition and brain structure, with preliminary evidence also supporting differences
in task-â•‰evoked brain activity as a function of aerobic exercise training. The evidence
presented up to this point has largely been devoted to the structure and function of
individual brain regions. A particularly intriguing and important question is whether
exercise is related to coordinated activity across multiple brain regions that define
neural networks or systems within the brain. Very recently, researchers have utilized
resting-â•‰state functional connectivity to address this question. Resting state functional
connectivity provides information on functional collaboration between spatially distal
regions through the quantification of intercorrelations among brain regions while at
â•‡ 451
rest. Although quite limited in number, the studies conducted to date have demon-
strated consistent positive effects of exercise on task-â•‰independent functional connec-
tivity within brain networks that are particularly vulnerable to age-â•‰related changes.
Burdette et al. (2010) assessed exercise-â•‰induced changes in hippocampal connectiv-
ity following a 4-â•‰month intervention among a group of older adults, aged 70–â•‰85 years.
Following the intervention, they found enhanced connectivity between the hippocam-
pus and ACC (also see Siette et al., 2013). Fitness-â•‰related variation in functional connec-
tivity may have cognitive implications. In a cross-â•‰sectional study by Voss et al. (2010a),
120 older adults aged 55–â•‰80 were administered an extensive neuropsychological bat-
tery as well as an objective cardiorespiratory fitness assessment. The authors found that
greater fitness levels were associated with increased functional connectivity within the
default mode network. The default mode network is engaged in young adults when the
mind is relaxed and suppressed when a cognitive challenge is presented. Disruption of
the default network is characteristic of both healthy and pathological aging.
As a follow-â•‰up, Voss et al. (2010b) conducted a randomized exercise intervention to
examine whether the resting state networks that they identified in their study where level
of fitness was treated as a cross-â•‰sectional variable could also be modified in an interven-
tion study where subjects were randomly assigned to a fitness or control condition. The
intervention involved 12 â•‰months of exercise training, in which older adult participants
were randomized to 1) a walking-â•‰based aerobic exercise condition or 2) a non-â•‰aero-
bic stretching and toning condition. Following one â•‰year of moderate-intensity walking
exercise, results showed an increase in functional connectivity within the default mode
network, as well as the fronto-â•‰executive network. Notably, exercise-â•‰induced increases in
functional connectivity within the default mode network correlated with improvements
in executive function post-â•‰intervention. Thus, 12 â•‰months of aerobic exercise increased
task-â•‰independent functional connectivity within two specific neural networks, changes
which were further linked to improvements in executive processes.
These results underscore the remarkable plasticity of brain networks even in older
adulthood, a time when the brain is largely susceptible to structural and functional
deterioration. Importantly, exercise-â•‰induced increases in resting-â•‰state functional con-
nectivity have been correlated with improvements in higher-â•‰order cognitive processes,
particularly executive function. Yet, whether the findings described above would also
be observed in clinical populations is, at present, unknown. Thus, while the current
body of literature demonstrates variation in the plasticity of neural networks as a func-
tion of fitness and aerobic exercise training, there remain many important questions
yet to be answered.
Modifiers of Exercise Effects on Cognition and Brain Function
Not all studies report significant effects of exercise on cognition or brain outcomes.
Similarly, not all individuals benefit at the same rate or to the same degree from exer-
cise. What factors could be contributing to this variability? One possibility is that
there are factors that are interacting with physical activity that modify the strength or
direction of the association. For example, engaging in both cognitively stimulating
activities and physical activity may exert a greater effect than either factor by itself.
452
In contrast, benefits of physical activity may be attenuated if it is accompanied by an

unhealthy diet (Leckie et al., 2012).
Several studies have reported that the beneficial effects of exercise or fitness are
magnified in later life (Colcombe et al., 2003; Erickson et al., 2011; Head et al., 2012).
Similarly, studies with more women tend to have larger effect sizes than studies with
fewer women (Colcombe and Kramer, 2003). This gender effect is also apparent in
mood outcomes, with women showing greater effects of physical activity on depressive
symptoms compared to men (Gujral et al., 2014). In sum, the demographic character-
istics of the sample are important factors to consider when inconsistent findings are
observed, since they may account for some of the variability across and within studies.
Similarly, dietary factors may also play an important role in modifying the effects of
physical activity. In a cross-sectional study of 344 adults, Leckie et al. (2014) found that
the level of omega-3 fatty acids in the blood modified the association between physical
activity and working memory and executive function. This physical activity x omega-3
interaction revealed that greater amounts of omega-3 levels offset the effects of lower lev-
els of physical activity on cognitive performance. There were no benefits of having both
higher levels of physical activity and higher levels of omega-3. This intriguing finding
suggests that a nutritious diet may be able to mitigate some of the negative consequences
of a more sedentary lifestyle. In addition, it also may provide an explanation for heteroge-
neity of findings in both the physical activity literature and the nutrition literature.
Genetic factors also play a role in risk for cognitive decline and likely interact
with physical activity to influence cognitive or brain outcomes. For example, Smith
et al. (2014) conducted a longitudinal study over 18 months and found that physically
active older adults showed less hippocampal atrophy than their less active peers, but
this effect was moderated by genetic risk for AD, such that less active APOE ε4 car-
riers showed more precipitous losses in volume than APOE ε4 carriers that remained
more physically active. In line with this, Erickson et al. (2013b) recently reported that
a common genetic polymorphism on the gene that codes for BDNF modifies the asso-
ciation between physical activity and working memory performance. In 1032 adults,
Erickson and colleagues found that a greater amount of self-reported physical activity
was associated with better working memory performance, but that this association
was magnified for those carrying the risk allele. More specifically, higher physical
activity levels eliminated the genetic differences on the working memory paradigm.
In sum, there is now considerable evidence that numerous demographic (e.g., age),
genetic (e.g., APOE), and lifestyle (e.g., omega-3) factors moderate the effects of
physical activity on cognitive and brain outcomes. These factors often go unstudied
and could help explain some of the individual differences within studies as well as
heterogeneity across studies.
Possible Mediators of the Effects of Exercise on Cognitive

and Brain Health
The above sections argue that exercise and fitness are associated with better brain
health. But, what are the mechanisms by which exercise modifies brain outcomes?
An accumulation of evidence in animal models demonstrates that exercise-induced
upregulation of growth factors, including BDNF, mediate the effect of exercise on
453
the brain (Gómez-Pinilla et al., 2002; Vaynman et al., 2003; Vaynman et al., 2004).
BDNF facilitates neural repair (Yang et al., 2014), induces long-term potentiation,
enhances learning and memory (Pang and Lu, 2004), and promotes synaptic plasticity
and neurogenesis (Vaynman et al., 2003). Vaynman and colleagues (2004) found in
rats that exercise training improved learning acquisition and recall during a spatial
learning task, but blocking BDNF in the hippocampus abrogated these improvements.
In addition to BDNF, mouse models have also demonstrated favorable alterations in
brain structure with exercise-induced increases in 1) insulin-like growth factor-1,
and 2) vascular endothelial growth factor, (Trejo et al., 2001; Cotman et al., 2007).
Unfortunately the contribution of these factors in humans is not well understood, in
part due to limitations in measuring central levels of these in vivo.
Another route by which exercise may be influencing the brain is via physiological
pathways linked to metabolism. Exercise reduces peripheral metabolic risk factors
including hypertension, adiposity, hypercholesterolemia, and insulin insensitivity,
which converge to cause cognitive impairment and neural degeneration. Prospective
research has demonstrated that the risks for cognitive decline and dementia are
increased considerably by the presence of cardiometabolic risk factors (Elias et al.,
2003; Kivipelto et al., 2005; Whitmer et al., 2008). Additionally, cardiometabolic risk
factors have been linked to global and regional gray matter atrophy, as well as degen-
eration of white matter integrity (Debette et al., 2011; Stanek et al., 2011; Maillard
et al., 2012).
Physical activity reduces the risk for development or progression of metabolic
disturbances, and may also mitigate the neural and cognitive consequences of these
conditions. Intervention and cross-sectional studies have found that increased physi-
cal activity favorably affects blood pressure (Dunn et al., 1999; Carroll and Dudfield,
2004), insulin resistance (Carroll and Dudfield, 2004), visceral fat (Ross and Janssen,
2001), and cholesterol and triglyceride levels (Healy et al., 2008). By attenuating
these risk factors, exercise may mitigate their adverse effects on the brain, but this is
currently unknown.
A common feature of many of the conditions mentioned above is chronic low-
grade systemic inflammation, which stimulates a host of neural and cognitive con-
sequences but may be regulated by exercise. Increases in pro-inflammatory factors
within the central nervous system interferes with multiple neuromolecular processes
including long-term potentiation, synaptic plasticity, and neurogenesis, with such
changes facilitating brain atrophy and impairments in cognitive function (Rosano et
al., 2012). For example, higher levels of circulating pro-inflammatory cytokines have
been associated with reductions in learning, memory, and executive function, and
have been prospectively linked to future cognitive decline (Yaffe et al., 2003; Lim et
al., 2013). Several cross-sectional studies have demonstrated an inverse relationship
between physical activity and markers of inflammatory load (Beavers et al., 2010)
suggesting that changes in inflammation may underlie some of the cognitive benefits.
At present, we have only a circumscribed understanding of the mechanisms by
which exercise influences brain outcomes. Research suggests that the propitious effects
of exercise may be orchestrated by exercise-dependent regulation of growth factor
expression and reductions in peripheral vascular risk factors. Importantly, it is likely
that multiple mechanistic pathways are at play. Additionally, the specific mechanisms
involved may vary as a function of the moderators discussed in the previous section.
454
Conclusion
In this chapter we have reviewed the effects of physical activity and exercise on cog-
nitive and brain outcomes. Based on this review, we can be confident that (a) physical
activity, especially in the form of moderate-intensity exercise, is capable of improving
cognitive function in late adulthood. These effects seem to have some specificity to
executive function, but there is evidence that several different cognitive domains are
affected. Further, there is a biological basis for these improvements to cognitive func-
tion. That is, (b) physical activity appears to influence several neuroimaging metrics
of brain health that associate with cognitive improvements.
Despite these promising effects of physical activity on cognitive and brain health
in late adulthood, there are many remaining questions that limit both the widespread
prescription of exercise for prevention and treatment of cognitive and brain disorders
as well as the scientific acceptance of physical activity as an important and effective
means of intervention. For example, we still know very little about the dose-â•‰response
nature of physical activity on the brain. Most of the studies described above have
used moderate-intensity exercise several days a week, but whether more vigorous
exercise fewer times per week or light exercise more days per week would have
similar effects remains largely unknown. We also have a poor understanding of the
extent to which these effects translate to cognitively impaired populations or those
with psychiatric or neurologic disorders. A few studies have shown promising trends
in at-â•‰risk populations (Smith et al., 2014) or in those with early dementia (Honea
et al., 2009; Erickson et al., 2013c; Ten Brinke et al., 2014), but more studies are
needed before making more widespread claims about the potential for physical activ-
ity to ameliorate symptoms. Finally, despite a wealth of animal research outlining
the possible molecular mechanisms of exercise on the brain, we still have a poor
understanding of these effects in humans. For physical activity interventions to hold
more weight as valid approaches to manipulate brain health and function among
scientists, we need more research that determines the mechanisms by which exercise
influences the brain.
In sum, the brain is a surprisingly plastic organ and at least some level of plasticity
is retained well into late adulthood. Fortunately, there is ample evidence that even a
modest amount of physical activity for several months will tap into inherent brain
plasticity and improve cognitive function. Nevertheless, much remains to be learned
about the mediators, moderators, and applications of physical activity to enhance
brain health, we can be certain that maintaining a healthy heart is critical for main-
taining a healthy brain.
References
Bakken Rc, Carey Jr, Di Fabio Rp, Erlandson Tj, Hake Jl, Intihar Tw (2001) Effect of aero-
bic exercise on tracking performance in elderly people: a pilot study. Physical therapy
81:1870–â•‰1879.
Barber R, Scheltens P, Gholkar A, Ballard C, Mckeith I, Ince P, Perry R, O’brien J (1999) White
matter lesions on magnetic resonance imaging in dementia with Lewy bodies, Alzheimer’s
disease, vascular dementia, and normal aging. Journal of Neurology, Neurosurgery &
Psychiatry 67:66–â•‰72.
455
Barnes De, Yaffe K (2011) The projected effect of risk factor reduction on Alzheimer’s disease
prevalence. Lancet Neurol 10:819–828.
Bartzokis G, Sultzer D, Lu Ph, Nuechterlein Kh, Mintz J, Cummings Jl (2004) Heterogeneous
age-related breakdown of white matter structural integrity: implications for cortical “dis-
connection” in aging and Alzheimer’s disease. Neurobiology of aging 25:843–851.
Beavers Km, Brinkley Te, Nicklas Bj (2010) Effect of exercise training on chronic inflamma-
tion. Clinica Chimica Acta 411:785–793.
Benedict C, Brooks Sj, Kullberg J, Nordenskjöld R, Burgos J, Le Grevès M, Kilander L,
Larsson E-M, Johansson L, Ahlström H (2013) Association between physical activity and
brain health in older adults. Neurobiology of aging 34:83–90.
Bohnen Ni, Albin Rl (2011) White matter lesions in Parkinson disease. Nature Reviews
Bugg Jm, Head D (2011) Exercise moderates age-related atrophy of the medial temporal lobe.
Neurobiology of aging 32:506–514.
Bugg Jm, Shah K, Villareal Dt, Head D (2012) Cognitive and neural correlates of aerobic fit-
ness in obese older adults. Exp Aging Res 38:131–145.
Burdette Jh, Laurienti Pj, Espeland Ma, Morgan A, Telesford Q, Vechlekar Cd, Hayasaka S,
Jennings Jm, Katula Ja, Kraft Ra, Rejeski Wj (2010) Using network science to evaluate
exercise-associated brain changes in older adults. Frontiers in aging neuroscience 2:23.
Burns Jm, Cronk Bb, Anderson Hs, Donnelly Je, Thomas Gp, Harsha A, Brooks Wm,
Swerdlow Rh (2008) Cardiorespiratory fitness and brain atrophy in early Alzheimer dis-
ease. Neurology 71:210–216.
Carroll S, Dudfield M (2004) What is the relationship between exercise and metabolic abnor-
malities? Sports Medicine 34:371–418.
Chapman Sb, Aslan S, Spence Js, Defina Lf, Keebler Mw, Didehbani N, Lu H (2013) Shorter
term aerobic exercise improves brain, cognition, and cardiovascular fitness in aging.
Frontiers in aging neuroscience 5:75.
Charlton R, Barrick T, Mcintyre D, Shen Y, O’sullivan M, Howe F, Clark C, Morris R, Markus
H (2006) White matter damage on diffusion tensor imaging correlates with age-related
cognitive decline. Neurology 66:217–222.
Colcombe S, Kramer Af (2003) Fitness effects on the cognitive function of older adults: a meta-
analytic study. Psychol Sci 14:125–130.
Colcombe Sj, Erickson Ki, Raz N, Webb Ag, Cohen Nj, Mcauley E, Kramer Af (2003) Aerobic
fitness reduces brain tissue loss in aging humans. J Gerontol A Biol Sci Med Sci 58:176–180.
Colcombe Sj, Kramer Af, Erickson Ki, Scalf P, Mcauley E, Cohen Nj, Webb A, Jerome Gj,
Marquez Dx, Elavsky S (2004) Cardiovascular fitness, cortical plasticity, and aging. Proc
Natl Acad Sci U S A 101:3316–3321.
Colcombe Sj, Erickson Ki, Scalf Pe, Kim Js, Prakash R, Mcauley E, Elavsky S, Marquez
Dx, Hu L, Kramer Af (2006) Aerobic exercise training increases brain volume in aging
humans. J Gerontol A Biol Sci Med Sci 61:1166–1170.
Cotman Cw, Berchtold Nc, Christie L-A (2007) Exercise builds brain health: key roles of
growth factor cascades and inflammation. Trends in neurosciences 30:464–472.
Debette S, Markus H (2010) The clinical importance of white matter hyperintensities on brain
magnetic resonance imaging: systematic review and meta-analysis. Bmj 341:c3666.
Debette S, Seshadri S, Beiser A, Au R, Himali J, Palumbo C, Wolf P, Decarli C (2011) Midlife
vascular risk factor exposure accelerates structural brain aging and cognitive decline.
Dunn Al, Marcus Bh, Kampert Jb, Garcia Me, Kohl Iii Hw, Blair Sn (1999) Comparison of
lifestyle and structured interventions to increase physical activity and cardiorespiratory
fitness: a randomized trial. Jama 281:327–334.
456
Elias M, Elias P, Sullivan L, Wolf P, D’agostino R (2003) Lower cognitive function in the
presence of obesity and hypertension: the Framingham heart study. International journal
of obesity 27:260–268.
Erickson Ki, Gildengers Ag, Butters MA (2013a) Physical activity and brain plasticity in late
adulthood. Dialogues Clin Neurosci 15:99–108.
Erickson Ki, Leckie Rl, Weinstein Am (2014) Physical activity, fitness, and gray matter vol-
ume. Neurobiology of aging 35 Suppl 2:S20–S28.
Erickson Ki, Banducci Se, Weinstein Am, Macdonald Aw, 3rd, Ferrell Re, Halder I, Flory
Jd, Manuck Sb (2013b) The brain-derived neurotrophic factor Val66Met polymorphism
moderates an effect of physical activity on working memory performance. Psychol Sci
24:1770–1779.
Erickson Ki, Prakash Rs, Voss Mw, Chaddock L, Hu L, Morris Ks, White Sm, Wojcicki Tr,
Mcauley E, Kramer Af (2009) Aerobic fitness is associated with hippocampal volume in
elderly humans. Hippocampus 19:1030–1039.
Erickson Ki, Raji Ca, Lopez Ol, Becker Jt, Rosano C, Newman Ab, Gach Hm, Thompson
Pm, Ho Aj, Kuller Lh (2010) Physical activity predicts gray matter volume in late adult-
hood: the Cardiovascular Health Study. Neurology 75:1415–1422.
Erickson Ki, Barr Ll, Weinstein Am, Banducci Se, Akl Sl, Santo Nm, Leckie Rl, Oakley M,
Saxton J, Aizenstein Hj, Becker Jt, Lopez Ol (2013c) Measuring physical activity using
accelerometry in a community sample with dementia. Journal of the American Geriatrics
Society 61:158–159.
Erickson Ki, Voss Mw, Prakash Rs, Basak C, Szabo A, Chaddock L, Kim Js, Heo S, Alves H,
White Sm, Wojcicki Tr, Mailey E, Vieira Vj, Martin Sa, Pence Bd, Woods Ja, Mcauley E,
Kramer Af (2011) Exercise training increases size of hippocampus and improves memory.
Etnier Jl, Nowell Pm, Landers Dm, Sibley Ba (2006) A meta-regression to examine the rela-
tionship between aerobic fitness and cognitive performance. Brain research reviews
52:119–130.
Floel A, Ruscheweyh R, Kruger K, Willemer C, Winter B, Volker K, Lohmann H, Zitzmann
M, Mooren F, Breitenstein C, Knecht S (2010) Physical activity and memory func-
tions: are neurotrophins and cerebral gray matter volume the missing link? Neuroimage
49:2756–2763.
Gold Bt, Powell Dk, Xuan L, Jicha Ga, Smith Cd (2010) Age-related slowing of task switch-
ing is associated with decreased integrity of frontoparietal white matter. Neurobiology of
aging 31:512–522.
Gómez-Pinilla F, Ying Z, Roy Rr, Molteni R, Edgerton Vr (2002) Voluntary exercise induces
a BDNF-mediated mechanism that promotes neuroplasticity. Journal of neurophysiology
88:2187–2195.
Gons Ra, Tuladhar Am, De Laat Kf, Van Norden Ag, Van Dijk Ej, Norris Dg, Zwiers Mp, De
Leeuw F-E (2013) Physical activity is related to the structural integrity of cerebral white
matter. Neurology 81:971–976.
Gordon Ba, Rykhlevskaia Ei, Brumback Cr, Lee Y, Elavsky S, Konopack Jf, Mcauley E, Kramer
Af, Colcombe S, Gratton G (2008) Neuroanatomical correlates of aging, cardiopulmonary
fitness level, and education. Psychophysiology 45:825–838.
Gow Aj, Bastin Me, Maniega Sm, Hernández Mcv, Morris Z, Murray C, Royle Na, Starr Jm,
Deary Ij, Wardlaw Jm (2012a) Neuroprotective lifestyles and the aging brain Activity,
atrophy, and white matter integrity. Neurology 79:1802–1808.
Gow Aj, Bastin Me, Munoz Maniega S, Valdes Hernandez Mc, Morris Z, Murray C, Royle
Na, Starr Jm, Deary Ij, Wardlaw Jm (2012b) Neuroprotective lifestyles and the aging
brain: Activity, atrophy, and white matter integrity. Neurology 79:1802–1808.
457
Gujral S, Manuck Sb, Ferrell Re, Flory Jd, Erickson Ki (2014) The BDNF Val66Met poly-
morphism does not moderate the effect of self-reported physical activity on depressive
symptoms in midlife. Psychiatry research 218:93–97.
Hagmann P, Jonasson L, Maeder P, Thiran J-P, Wedeen Vj, Meuli R (2006) Understanding
Diffusion MR Imaging Techniques: From Scalar Diffusion-weighted Imaging to Diffusion
Tensor Imaging and Beyond 1. Radiographics 26:S205-S223.
Head D, Singh T, Bugg Jm (2012) The moderating role of exercise on stress-related effects on
the hippocampus and memory in later adulthood. Neuropsychology 26:133–143.
Head D, Buckner Rl, Shimony Js, Williams Le, Akbudak E, Conturo Te, Mcavoy M, Morris Jc,
Snyder Az (2004) Differential vulnerability of anterior white matter in nondemented aging
with minimal acceleration in dementia of the Alzheimer type: evidence from diffusion
tensor imaging. Cerebral Cortex 14:410–423.
Healy Gn, Wijndaele K, Dunstan Dw, Shaw Je, Salmon J, Zimmet Pz, Owen N (2008)
Objectively measured sedentary time, physical activity, and metabolic risk the Australian
Diabetes, Obesity and Lifestyle Study (AusDiab). Diabetes care 31:369–371.
Hindin Sb, Zelinski Em (2012) Extended practice and aerobic exercise interventions benefit
untrained cognitive outcomes in older adults: a meta-analysis. Journal of the American
Geriatrics Society 60:136–141.
Ho Aj, Raji Ca, Becker Jt, Lopez Ol, Kuller Lh, Hua X, Dinov Id, Stein Jl, Rosano C, Toga Aw
(2011) The effects of physical activity, education, and body mass index on the aging brain.
Human brain mapping 32:1371–1382.
Honea Ra, Thomas Gp, Harsha A, Anderson Hs, Donnelly Je, Brooks Wm, Burns Jm (2009)
Cardiorespiratory fitness and preserved medial temporal lobe volume in Alzheimer dis-
ease. Alzheimer Dis Assoc Disord 23:188–197.
Jernigan Tl, Archibald Sl, Fennema-Notestine C, Gamst Ac, Stout Jc, Bonner J, Hesselink Jr
(2001) Effects of age on tissues and regions of the cerebrum and cerebellum. Neurobiology
of aging 22:581–594.
Johnson Nf, Kim C, Clasey Jl, Bailey A, Gold Bt (2012) Cardiorespiratory fitness is posi-
tively correlated with cerebral white matter integrity in healthy seniors. Neuroimage
59:1514–1523.
Kivipelto M, Ngandu T, Fratiglioni L, Viitanen M, Kåreholt I, Winblad B, Helkala E-L,
Tuomilehto J, Soininen H, Nissinen A (2005) Obesity and vascular risk factors at midlife
and the risk of dementia and Alzheimer disease. Archives of neurology 62:1556–1560.
Kolb B, Muhammad A (2014) Harnessing the power of neuroplasticity for intervention.
Frontiers in human neuroscience 8:377.
Kramer Af, Hahn S, Cohen Nj, Banich Mt, Mcauley E, Harrison Cr, Chason J, Vakil E, Bardell
L, Boileau Ra, Colcombe A (1999) Ageing, fitness and neurocognitive function. Nature
400:418–419.
Krogh J, Rostrup E, Thomsen C, Elfving B, Videbech P, Nordentoft M (2014) The effect of
exercise on hippocampal volume and neurotrophines in patients with major depression--a
randomized clinical trial. Journal of affective disorders 165:24–30.
Leckie Rl, Weinstein Am, Hodzic Jc, Erickson Ki (2012) Potential moderators of physical
activity on brain health. Journal of aging research 2012:948981.
Leckie Rl, Manuck Sb, Bhattacharjee N, Muldoon Mf, Flory Jm, Erickson Ki (2014) Omega-3
fatty acids moderate effects of physical activity on cognitive function. Neuropsychologia
59:103–111.
Lim A, Krajina K, Marsland AL (2013) Peripheral inflammation and cognitive aging.
Lovden M, Wenger E, Martensson J, Lindenberger U, Backman L (2013) Structural brain
plasticity in adult learning and development. Neuroscience and biobehavioral reviews
37:2296–2310.
458
Madden Dj, Bennett Ij, Song Aw (2009) Cerebral white matter integrity and cogni-
tive aging: contributions from diffusion tensor imaging. Neuropsychology review
19:415–435.
Maillard P, Seshadri S, Beiser A, Himali Jj, Au R, Fletcher E, Carmichael O, Wolf Pa,
Decarli C (2012) Effects of systolic blood pressure on white-matter integrity in young
adults in the Framingham Heart Study: a cross-sectional study. The Lancet Neurology
11:1039–1047.
Makizako H, Liu-Ambrose T, Shimada H, Doi T, Park H, Tsutsumimoto K, Uemura K, Suzuki T
(2014) Moderate-Intensity Physical Activity, Hippocampal Volume, and Memory in Older
Adults With Mild Cognitive Impairment. J Gerontol A Biol Sci Med Sci 70:480–486.
Marks B, Katz L, Styner M, Smith J (2010) Aerobic fitness and obesity: relationship to cerebral
white matter integrity in the brain of active and sedentary older adults. British journal of
sports medicine 45:1208–1215.
Niemann C, Godde B, Voelcker-Rehage C (2014) Not only cardiovascular, but also coor-
dinative exercise increases hippocampal volume in older adults. Frontiers in aging
neuroscience 6:170.
Pang Pt, Lu B (2004) Regulation of late-phase LTP and long-term memory in normal and
aging hippocampus: role of secreted proteins tPA and BDNF. Ageing research reviews
3:407–430.
Petersen Rc, Caracciolo B, Brayne C, Gauthier S, Jelic V, Fratiglioni L (2014) Mild cognitive
impairment: a concept in evolution. Journal of internal medicine 275:214–228.
Pfefferbaum A, Sullivan Ev (2002) Microstructural but not macrostructural disruption of white
matter in women with chronic alcoholism. Neuroimage 15:708–718.
Podewils L, Guallar E, Beauchamp N, Lyketsos C, Kuller L, Scheltens P (2007) Physical activ-
ity and white matter lesion progression Assessment using MRI. Neurology 68:1223–1226.
Prakash Rs, Voss Mw, Erickson Ki, Lewis Jm, Chaddock L, Malkowski E, Alves H, Kim J, Szabo
A, White Sm, Wojcicki Tr, Klamm El, Mcauley E, Kramer Af (2011) Cardiorespiratory
fitness and attentional control in the aging brain. Frontiers in human neuroscience 4:229.
Raz N, Ghisletta P, Rodrigue Km, Kennedy Km, Lindenberger U (2010) Trajectories of brain
aging in middle-aged and older adults: regional and individual differences. NeuroImage
51:501–511.
ual differences and modifiers. Cerebral cortex 15:1676–1689.
Rosano C, Marsland Al, Gianaros Pj (2012) Maintaining brain health by monitoring inflamma-
tory processes: a mechanism to promote successful aging. Aging Dis 3:16–33.
Rosano C, Venkatraman Vk, Guralnik J, Newman Ab, Glynn Nw, Launer L, Taylor Ca,
Williamson J, Studenski S, Pahor M (2010) Psychomotor speed and functional brain MRI
2 years after completing a physical activity treatment. The Journals of Gerontology Series
A: Biological Sciences and Medical Sciences 65:639–647.
Ross R, Janssen I (2001) Physical activity, total and regional obesity: dose-response consider-
ations. Medicine and science in sports and exercise 33:S521–S527.
Rovio S, Spulber G, Nieminen Lj, Niskanen E, Winblad B, Tuomilehto J, Nissinen A, Soininen
H, Kivipelto M (2010a) The effect of midlife physical activity on structural brain changes
in the elderly. Neurobiology of aging 31:1927–1936.
Rovio S, Spulber G, Nieminen Lj, Niskanen E, Winblad B, Tuomilehto J, Nissinen A, Soininen
H, Kivipelto M (2010b) The effect of midlife physical activity on structural brain changes
in the elderly. Neurobiology of aging 31:1927–1936.
459
Ruscheweyh R, Willemer C, Kruger K, Duning T, Warnecke T, Sommer J, Volker K, Ho Hv,

Mooren F, Knecht S, Floel A (2011) Physical activity and memory functions: an interven-
tional study. Neurobiology of aging 32:1304–1319.
Siette J, Westbrook Rf, Cotman C, Sidhu K, Zhu W, Sachdev P, Valenzuela Mj (2013) Age-
specific effects of voluntary exercise on memory and the older brain. Biological psychiatry
73:435–442.
Smith Jc, Nielson Ka, Antuono P, Lyons Ja, Hanson Rj, Butts Am, Hantke Nc, Verber Md
(2013) Semantic memory functional MRI and cognitive function after exercise interven-
tion in mild cognitive impairment. Journal of Alzheimer’s disease: JAD 37:197–215.
Smith Jc, Nielson Ka, Woodard Jl, Seidenberg M, Verber Md, Durgerian S, Antuono P, Butts
Am, Hantke Nc, Lancaster Ma, Rao Sm (2011) Does physical activity influence semantic
memory activation in amnestic mild cognitive impairment? Psychiatry research 193:60–62.
Smith Jc, Nielson Ka, Woodard Jl, Seidenberg M, Durgerian S, Hazlett Ke, Figueroa Cm, Kandah
Cc, Kay Cd, Matthews Ma, Rao Sm (2014) Physical activity reduces hippocampal atrophy
in elders at genetic risk for Alzheimer’s disease. Frontiers in aging neuroscience 6:61.
Smith Pj, Blumenthal Ja, Hoffman Bm, Cooper H, Strauman Ta, Welsh-Bohmer K, Browndyke
Jn, Sherwood A (2010) Aerobic exercise and neurocognitive performance: a meta-analytic
review of randomized controlled trials. Psychosom Med 72:239–252.
Sofi F, Valecchi D, Bacci D, Abbate R, Gensini Gf, Casini A, Macchi C (2011) Physical activ-
ity and risk of cognitive decline: a meta-analysis of prospective studies. J Intern Med
269:107–117.
Spirduso Ww (1975) Reaction and movement time as a function of age and physical activity
level. Journal of gerontology 30:435–440.
Spirduso Ww, Clifford P (1978) Replication of age and physical activity effects on reaction and
movement time. Journal of gerontology 33:26–30.
Stanek Km, Grieve Sm, Brickman Am, Korgaonkar Ms, Paul Rh, Cohen Ra, Gunstad Jj (2011)
Obesity Is Associated With Reduced White Matter Integrity in Otherwise Healthy Adults*.
Obesity 19:500–504.
Suzuki T, Shimada H, Makizako H, Doi T, Yoshida D, Ito K, Shimokata H, Washimi Y, Endo H,
Kato T (2013) A randomized controlled trial of multicomponent exercise in older adults
with mild cognitive impairment. PloS one 8:e61483.
Ten Brinke Lf, Bolandzadeh N, Nagamatsu Ls, Hsu Cl, Davis Jc, Miran-Khan K, Liu-Ambrose
T (2014) Aerobic exercise increases hippocampal volume in older women with proba-
ble mild cognitive impairment: a 6-month randomised controlled trial. British journal of
sports medicine 49:248–254.
Tian Q, Erickson Ki, Simonsick Em, Aizenstein Hj, Glynn Nw, Boudreau Rm, Newman Ab,
Kritchevsky Sb, Yaffe K, Harris Tb (2014) Physical Activity Predicts Microstructural
Integrity in Memory-Related Networks in Very Old Adults. The Journals of Gerontology
Series A: Biological Sciences and Medical Sciences 69:1284–1290.
Trejo Jl, Carro E, Torres-Alemán I (2001) Circulating insulin-like growth factor I mediates
exercise-induced increases in the number of new neurons in the adult hippocampus. The
Journal of neuroscience 21:1628–1634.
Tseng B, Gundapuneedi T, Khan M, Diaz-Arrastia R, Levine B, Lu H, Huang H, Zhang R
(2013) White matter integrity in physically fit older adults. Neuroimage 82:510–516.
Van Praag H, Shubert T, Zhao C, Gage Fh (2005) Exercise enhances learning and hippocampal
neurogenesis in aged mice. The Journal of Neuroscience 25:8680–8685.
Vaynman S, Ying Z, Gomez-Pinilla F (2003) Interplay between brain-derived neurotrophic
factor and signal transduction modulators in the regulation of the effects of exercise on
synaptic-plasticity. Neuroscience 122:647–657.
460
Vaynman S, Ying Z, Gomez‐Pinilla F (2004) Hippocampal BDNF mediates the efficacy

of exercise on synaptic plasticity and cognition. European Journal of Neuroscience
20:2580–2590.
Vernooij Mw, Ikram Ma, Vrooman Ha, Wielopolski Pa, Krestin Gp, Hofman A, Niessen
Wj, Van Der Lugt A, Breteler Mm (2009) White matter microstructural integrity and
cognitive function in a general elderly population. Archives of General Psychiatry
66:545–553.
Voss Mw, Heo S, Prakash Rs, Erickson Ki, Alves H, Chaddock L, Szabo An, Mailey El,
Wojcicki Tr, White Sm (2013) The influence of aerobic fitness on cerebral white matter
integrity and cognitive function in older adults: Results of a one‐year exercise intervention.
Human brain mapping 34:2972–2985.
Voss Mw, Erickson Ki, Prakash Rs, Chaddock L, Malkowski E, Alves H, Kim Js, Morris
Ks, White Sm, Wojcicki Tr, Hu L, Szabo A, Klamm E, Mcauley E, Kramer Af (2010a)
Functional connectivity: a source of variance in the association between cardiorespiratory
fitness and cognition? Neuropsychologia 48:1394–1406.
Voss Mw, Prakash Rs, Erickson Ki, Basak C, Chaddock L, Kim Js, Alves H, Heo S, Szabo An,
White Sm, Wojcicki Tr, Mailey El, Gothe N, Olson Ea, Mcauley E, Kramer AF (2010b)
Plasticity of brain networks in a randomized intervention trial of exercise training in older
adults. Frontiers in aging neuroscience 2.
Westlye Lt, Walhovd Kb, Dale Am, Bjørnerud A, Due-Tønnessen P, Engvig A, Grydeland H,
Tamnes Ck, Østby Y, Fjell Am (2009) Life-span changes of the human brain white matter:
diffusion tensor imaging (DTI) and volumetry. Cerebral cortex 20:2055–2068.
Whitmer R, Gustafson D, Barrett-Connor E, Haan M, Gunderson E, Yaffe K (2008) Central
obesity and increased risk of dementia more than three decades later. Neurology
71:1057–1064.
Yaffe K, Lindquist K, Penninx B, Simonsick E, Pahor M, Kritchevsky S, Launer L, Kuller
L, Rubin S, Harris T (2003) Inflammatory markers and cognition in well-functioning
African-American and white elders. Neurology 61:76–80.
Yang J-L, Lin Y-T, Chuang P-C, Bohr Va, Mattson Mp (2014) BDNF and exercise enhance
neuronal DNA repair by stimulating CREB-Mediated production of apurinic/apyrimidinic
endonuclease 1. Neuromolecular medicine 16:161–174.
461
18
The Link of Intellectual Engagement

to Cognitive and Brain Aging
Martin Lövdén
Lars Bäckman
Ulman Lindenberger
I s human cognitive ability static and is the brain fixed? Certainly not. An
individual’s cognitive performance varies systematically from day- to-
day, minute-to-minute, and second-to-second (Rabbitt et al., 2001; Schmiedek et al.,
2013). Flux at the level of the brain is not in any way an exception, but rather the
normal modus operandi of brain structure and function (Faisal et al., 2008). Cognitive
performance also changes over longer time periods: it improves considerably during
child development (Jones and Conrad, 1933) and declines in aging (Schaie, 1994).
The volume and integrity of the brain change in a similar way over the lifespan (Giedd
et al., 1999; Raz et al., 2005).
Can humans improve their cognitive performance? Certainly. Eating breakfast is
a good idea (Hoyland et al., 2009). Drinking a cup of coffee may sometimes enhance
performance (Nehlig, 2010). Though long-term side effects prevail, the right dos-
age of nicotine improves attention and working memory (Heishman et al., 2010).
Acquiring knowledge and strategies in a domain affects memory for domain-rele-
vant information (Bartlett, 1932; Chase and Simon, 1973). Education in childhood
and early adulthood improves performance on tests of intelligence (Cliffordson and
Gustafsson, 2008; Brinch and Galloway, 2012). During the 20th century, each new
generation performed better than their parents on tests measuring a variety of cogni-
tive functions (Flynn, 1984). No doubt, then, human behavior can modify cognitive
performance, but of course, not every aspect of human behavior does the trick. In
this chapter, we evaluate the evidence for and against the hypothesis that engagement
461
462
in cognitively demanding activities positively influences cognitive performance in

healthy aging, and we review which brain mechanisms can be linked to such effects.
We focus this review on whether cognitive activity influences processing efficiency
(Lövdén et al., 2010a); that cognitively healthy older adults can acquire new skills and
enhance their task-relevant cognitive strategies is undisputed.
Between-Person Differences in Intellectual Engagement

and Cognitive Performance
One approach to investigate whether engagement in cognitively demanding activities

improves performance in old age is to sample, typically with questionnaires, individ-
uals’ involvement in various types of cognitively stimulating activities (e.g., leisure
activities of various kinds, such as reading books, solving cross-words, and playing
board games). Between-person differences in engagement (e.g., frequency, duration)
in these activities are then related to differences in cognitive performance in various
ways. Obviously, longitudinal within-person data provide the most powerful founda-
tion for estimating these various associations (e.g., Hertzog et al., 2009).
Results from the Victoria Longitudinal Study (VLS) reveal that decline in cognitive
leisure activities over a 6-year period in old age was associated with decline in cogni-
tive performance, including fact recall (Hultsch et al., 1999) and aspects of processing
speed (Bielak et al., 2007; see also Bielak et al., 2014). The magnitude of these asso-
ciations was, however, relatively small, with the median correlation between change
in cognitive activity and change in the various measures of cognitive performance
in the VLS studies being only 0.10. However, a comprehensive analysis of four dif-
ferent longitudinal studies with up to 21 years of follow-up data support these initial
findings, with changes in participation in cognitively stimulating leisure activities
being consistently associated with changes in reasoning, verbal fluency, memory, and
knowledge (rs = .23–.50; Mitchell et al., 2012).
Note that correlations between cognitive performance and lifestyle are generally
not informative of the causal direction of influence (Hultsch et al., 1999; Lövdén et
al., 2005; Gow et al., 2012b). That is, the cognitive engagement hypothesis would pre-
dict such associations, but reverse causation, with decline in cognitive performance
leading to a less active life, is equally tenable, given that people are not randomly
assigned to lifestyles that differ in the degree of cognitive challenge. In addition,
changes in some third variable (e.g., health) may drive both activity and cognitive
changes. Investigating the association between level of engagement in cognitive stim-
ulating activities at one point in time and subsequent change in cognitive performance
has been one way to try to approach these issues of causality. Several findings of a
positive association between engagement levels and change in performance have been
reported from longitudinal studies (e.g., Hultsch et al., 1999; Schooler and Mulatu,
2001; Bosma et al., 2002; Wilson et al., 2003; Wang et al., 2013). However, other
studies have failed to find such associations (e.g., Aartsen et al., 2002; Gow et al.,
2012b; Gow et al., 2012a; Mitchell et al., 2012). Some have found that cognitive
performance predicts subsequent change in activity engagement (e.g., Schooler and
Mulatu, 2001; Aartsen et al., 2002; Bosma et al., 2002). A comprehensive way to
463
The Link of Intellectual Engagement to Cognitive and Brain Aging 463
approach these level-change associations is to apply a statistical model for longitudi-

nal data termed the dual-change score model (McArdle and Hamagami, 2001). In this
model, both hypothetical level-change influences can be simultaneously estimated,
while accounting for the effect of previous level of a variable on subsequent change
in that variable, as well as for the overall linear change. With respect to intellectual
activities, Ghisletta et al. (2006) found that media consumption (e.g., radio, TV, news-
papers) and intellectual leisure activities (games and cross-words) were related to
subsequent changes in perceptual speed, but not vice versa (see also Lövdén et al.,
2005). Supporting this pattern, Small et al. (2012) observed that cognitive activities
were associated with change in episodic and semantic memory, without the reverse
influence. However, Small et al. (2012) also reported reciprocal associations between
cognitive activities and speed of lexical access.
The mixed nature of the findings likely reflects the presence of several sources
of influence (Hertzog et al., 2009; Bielak, 2010). For example, the sampled activ-
ities and the variables that they form vary substantially across studies, and differ-
ent activities may mean different effects for different individuals (Salthouse et al.,
2002). A related problem is that individuals’ engagement in select leisure activities
is only a very small portion of the cognitive demands in their lives, and the impor-
tance of this portion may vary among older adults and certainly across the lifespan.
For example, during extensive periods of life we spend more time in work than in
leisure activities. This balance shifts across the lifespan, with retirement perhaps
being the most dramatic change. Indeed, support for positive associations between
intellectually challenging occupations and level and change trajectories of cognitive
performance in old age is available (Jorm et al., 1998; Schooler et al., 1999; Bosma
et al., 2003; Andel et al., 2007; Finkel et al., 2009; Marquie et al., 2010; Van der
Elst et al., 2012). Also here reverse causation might, however, operate. Individuals
higher in cognitive ability are more likely to make their way into more demanding
jobs, and the cognitive ability differences, rather than intellectual engagement, may
drive old-age differences in level and change of cognitive performance (Salthouse,
2006). A few attempts to account for early differences in cognitive ability have been
reported. In one study, Gow et al. (2014) observed a seemingly counterintuitive neg-
ative association between occupational demands and cognitive performance after
statistically controlling for early cognitive ability. This finding implies that, of two
persons with the same cognitive ability at age 50, the individual with the more intel-
lectually demanding job had lower cognitive ability in old age (60–80 years) than
the one with the less demanding job. Even if we take these findings at face value
and disregard potential methodological problems with controlling for earlier perfor-
mance at the observed level (Glymour et al., 2005), we note that such findings are
not incompatible with the hypothesis that cognitive engagement is beneficial for per-
formance. This becomes clear when, for example, factoring in evidence indicating
that retirement can have a detrimental effect on cognitive performance (Schaie, 2005;
Finkel et al., 2009; Rohwedder and Willis, 2010; Roberts et al., 2011; Bonsang et
al., 2012; Mazzonna and Peracchi, 2012). For example, Finkel et al. (2009) reported
that individuals in occupations characterized by high complexity of work with people
(e.g., jobs with mentoring and negotiations demands) displayed steeper decline after
retirement than individuals with jobs scoring low on this dimension (but see Fisher
464
et al., 2014). For individuals in occupations with higher complexity, retirement may
constitute a greater change in life conditions, and cognitive performance may thus
be differentially affected. With such a background, the findings reported by Gow et
al. (2014) could also make sense from the point of view of the engagement hypoth-
esis: Of two persons with the same cognitive ability at age 50, the individual with
the more intellectually demanding job may have the same cognitive ability as the
person in the job with lower demands just because job conditions positively affect
cognitive ability. When this influence is not there anymore, the individual in the more
intellectually demanding job may have lower cognitive performance. In this sense,
failure to find that level of engagement positively affects subsequent change (i.e.,
differential preservation of cognitive abilities; Salthouse, 2006) is not inconsistent
with the engagement hypothesis. Such findings could just be reflecting that individu-
als’ current engagement in cognitively demanding activities is what matters for per-
formance, which would play out in observations of level–level and change–change
associations. Level-change associations, such as whether initial differences are pre-
served or whether they are differentially preserved, are not necessarily informative,
and may differ depending on the time frame they capture. In line with this view,
studies with shorter longitudinal time spans are also the ones that more often find sig-
nificant level-change associations (Lövdén et al., 2005; Ghisletta et al., 2006; Small
et al., 2012). Such a pattern fits theoretical models of adult plasticity that pinpoint
ongoing adaptations to a mismatch between experiential demands and functional
capacity as partially determining performance (Lövdén et al., 2010a). That is, what
you do, rather than what you did, could be the key player. However, we also note that
retirement may affect cognitive performance via mechanisms other than reductions
of cognitive stimulation, so that effects of retirement on performance do not provide
strong evidence for the engagement hypothesis.
Overall, this line of inquiry may thus benefit from a more systematic approach
to sampling intellectual activities, including the entire life space of intellectual
demands (leisure activities, work, family life, and so forth) and from more attention
to individual differences in the balance of these aspects of life and how they change
in importance over time. As things stand, it seems safe to conclude that there is an
association between engagement in cognitively demanding activities and cogni-
tive performance in aging. Studies of the association between level of engagement
and subsequent change that focus on a shorter longitudinal time span (around 2–3
years) support the notion that cognitive activity influences subsequent change in
performance. The causal nature of the association is, however, likely to be complex
and scientific consensus on this issue is unlikely to be reached based on studies
of naturally occurring between-person differences alone. Progress in this field is
likely to come from abandoning simplistic attempts to pit the hypothesis that activ-
ity affects cognitive ability against the prediction that ability affects selection of
activities. The path dependency of the life course needs more careful conceptual
consideration and longitudinal study: People with higher cognitive abilities may be
more likely to select or be selected into more challenging environments, and these
environments, in turn, may further improve their abilities, whereas the reverse
may be true for individual with lower cognitive abilities (Schooler et al., 1999;
Schooler and Mulatu, 2001). Individuals with higher abilities may experience the
â•‡ 465
The Link of Intellectual Engagement to Cognitive and BrainÂ€Agingâ•… 465
same environment differently from individuals with lower abilities. The resulting
scenario would be one in which the two directions of influence—â•‰abilities affecting
lifestyles, and lifestyles affecting abilities—â•‰are positively correlated over time.
This would be consistent with theoretical and empirical claims pointing to the
importance of gene–â•‰environment correlations in understanding individual differ-
ences in development (e.g., Beam and Turkheimer, 2013).
Effects of Cognitive Training on Performance in Old Age
Intervention studies could provide less ambiguous support for the hypothesis that
intellectual engagement affects performance. A few such studies have used multiâ•‰
modal engagement interventions (e.g., group-â•‰based diverse problem solving tasks,
computer and photo-â•‰editing courses), with encouraging results for memory (Park et
al., 2014) and reasoning (Stine-â•‰Morrow et al., 2008; Tranter and Koutstaal, 2008).
Studies focusing on cognitive training promise to more specifically localize such
effects to the impact of cognitive activity. The first generation of such studies generally
included teaching individuals to use efficient cognitive strategies (e.g., method-â•‰of-â•‰loci
for memorizing words). Results were disappointing in the sense that improvements
on the trained tasks did not transfer to related but nontrained tasks (Verhaeghen et
al., 1992; Ball et al., 2002; Hertzog et al., 2009). Probing transfer of improvements is
important because this provides a tool for examining whether processing efficiency
has been improved (Lövdén et al., 2010a). That is, if improvements can be observed
on tasks where training-â•‰related acquisition of knowledge (e.g., better strategies and
improved response mapping) can be reasonably well excluded as a factor behind any
improvements, then it can be assumed that training has affected processing efficiency.
A more recent generation of studies has examined the effects of practice on various
types of cognitively challenging tasks, such as off-â•‰the-â•‰shelf video games (Basak et
al., 2008), working memory tasks (Dahlin et al., 2008b), and a mix of cognitive tasks
(Schmiedek et al., 2010). Karbach and Verhaeghen (2014) recently summarized this
literature in a meta-â•‰analysis that focused on training of working memory and exec-
utive tasks. This analysis yielded net training effects (gains for training group minus
gains for controls) of 0.5 SD for near transfer (measuring the trained ability using
untrained tasks) and 0.2 SD for far transfer (measuring any nonâ•‰trained ability, such as
reasoning, episodic memory, and speed). Younger and older adults displayed similar
effects sizes.
With near-transfer effects, it is, without a detailed model or task analysis, diffi-
cult to exclude that acquisition of knowledge (e.g., strategies) is responsible for the
observed gains. One may therefore argue that this type of outcome measure is prob-
lematic in a meta-â•‰analysis. A skeptical reader may also argue that far-â•‰transfer effects
suffer from the same problem, especially when selected far-â•‰transfer tasks are hetero-
geneous, so that the tasks in each individual study need to be carefully analyzed. We
therefore conducted our own meta-â•‰analysis, focusing exclusively on reasoning as an
outcome measure. To gain power and generality we included all types of process-
ing-â•‰based cognitive training (e.g., working memory, inhibition, episodic memory, and
computer game training, but no combinations with other activities, such a physical
466
training, and no strategy training, general enrichment interventions, and meditation

studies). Further, we only included studies if the method sections were sufficiently
detailed to make sure that training procedures did not directly include reasoning or
strategy training. Further inclusion criteria were (1) a pretest–posttest design includ-
ing a control group; (2) a healthy older sample (mean age > 60); and (3) publication
of the study between January 2000 and July 2014.
Our initial screening of search results from Pubmed, Web of Science, and avail-
able meta-analyses and reviews (Kueider et al., 2012; Reijnders et al., 2013; Karr et
al., 2014; Kelly et al., 2014; Noack et al., 2014; Brehmer et al., 2014; Karbach and
Verhaeghen, 2014) resulted in 73 candidate studies, of which 20 were eligible accord-
ing to the above criteria. These studies reported results for samples with a mean age of
69 years (range 61–79 years) and a mean total sample size of 51 (29–139) individuals.
The training groups trained on average 993 minutes (range = 180–6000 minutes).
Eleven studies had an active control group. Thirteen studies focused on working-
memory training, one on task switching, one on computer game training, one on inhi-
bition, and four on several cognitive domains. Ten studies had Raven’s matrices as
a single reasoning outcome, four had Cattell’s culture fair test, one had reasoning
tasks from WAIS III, one had a letter series task only, and four had several reasoning
tasks. As main measures for the meta-analyses we computed one Standardized Mean
Difference (SMD; Hedges’ g) for each study of the difference between the training
and control groups at pretest and one at posttest. An average SMD was computed
across tasks for the studies reporting multiple reasoning tasks.
Results of a random effects analysis (maximum likelihood in Open Meta Analyst;
Wallace et al., 2012) of the post-test difference in reasoning performance between
training and control groups showed a significant weighted mean group differences
favoring the training group (g = 0.192, SE = .097, p = .049). There was no such dif-
ference at pretest (g = −0.012, SE = .081, p = .878). The standardized mean increase
for the training groups was 0.342 (SE = .076, p < .001). The increase for the control
groups was .160 (SE = .067, p = .018). The difference between these effects (i.e., the
net training effect) is 0.182, which corresponds well with the observed post-test dif-
ference between the groups.
Publication bias was addressed by first computing standardized net effect sizes
for each study, which arguably is the effect that may drive a publication bias. We are
not aware of a way to compute the standard errors for these effects, so we related
this effect size to total sample size (rather than SE, which is otherwise the preferred
measure; Sterne and Egger, 2001). The scatterplot (Figure 18.1) of this association
showed no indication of publication bias, and sample size was not significantly
related to effect size (p = .61). However, the plot shows that two of the small-sized
(and thus likely low-powered) studies (Borella et al., 2010; Carretti et al., 2013)
report somewhat deviant net effect sizes. We therefore excluded these two studies
in a sensitivity analysis, which showed a lower and nonsignificant posttest differ-
ence between the groups (g = 0.136, SE = .094, p = .148). The pretest difference
was essentially zero (g = −0.007, SE = .091, p = .939). The increase for the training
groups included in this analysis was 0.255 (SE = .062, p < .001). The increase for
the control groups was .158 (SE = .071, p = .025). The difference between these
effects (i.e., the net training effect) is 0.097. Figure 18.2 shows a forest plot of the
467
160
140
120
Total sample size

100
80
60
40
20
0
–1.5 –1 –0.5 0 0,5 1 1,5
Net SMD
Figure 18.1 Funnel plot relating total sample size to net SMD (Standardized Mean Difference)
of cognitive training on reasoning (SMDpost-pre for the training group—SMDpost-pre for the control
group). The dashed line indicates the mean net SMD.
individual effect sizes and the weighted mean effect sizes across all studies and
when excluding the two deviant studies. Figure 18.3 shows the corresponding infor-
mation at pretest.
The effect sizes were significantly heterogeneous at posttest (I2 = 52.59, p = 0.002),
but not at pretest (I2 = 32.84, p = 0.058). Increases were significantly heterogeneous
for the training groups (I2 = 38.33, p = 0.029), but not for the control groups (I2 = 0.00,
p = 0.959). These results suggest that the training regimens used in the various stud-
ies may differ in efficiency (e.g., due to differences in sample composition, outcome
variables, and training paradigms). We therefore explored associations with a few
potentially moderating factors.
Mean age of the sample was unrelated to posttest differences (p = .911) and to
gains in the training groups (p = .37). Training length was also unrelated to posttest
differences (p = .174) and not related to gains in the training groups (p = .419).
When mutually adjusted, neither training length nor age had a significant effect (both
ps > .217). Excluding the two studies with outlying positive effects and two studies
with outlying length of training did not change these findings. Studies with Raven’s
matrices as an outcome did not demonstrate larger posttest differences than other
studies (p = .361), and gains were not larger in these studies either (p = .146). Studies
with working memory training tended to report larger posttest differences than other
studies (p = .079), but gains were not larger in these studies (p = .827). Excluding
the two studies with outlying positive effects did not substantially alter these results.
Studies with an active control group did not report smaller posttest differences than
other studies (p = .844), and neither gains in the training group (p = .387) nor gains
in the control group differed between active and passive control groups (p = .897).
Thus, we conclude that the observed heterogeneity remains unexplained.
In summary, studies of the effects of practicing cognitive tasks on reasoning per-
formance in old age report a significant but small average effect size (roughly 0.2
SD). This estimate is virtually identical to the far-transfer effect reported by Karbach
468
Studies Estimate (95% C.I.)
Ackerman et al. 2010 0.247 (–0.198, 0.693)

Basak et al. 2008 0.534 (–0.113, 1.181)
Boot et al. 2013 –0.772 (–1.440, –0.104)
Borella et al. 2014 young–old 0.426 (–0.201, 1.052)
Borella et al. 2013 0.119 (–0.535, 0.773)
Brehmer et al. 2012 –0.100 (–0.692, 0.492)
Burki et al. 2014 –0.333 (–0.942, 0.277)
Dahlin et al. 2008 –0.044 (–0.776, 0.688)
Heinzel et al. 2014 0.186 (–0.531, 0.904)
Karbach et al. 2009 0.014 (–0.614, 0.642)
Richmond et al. 2011 0.127 (–0.494, 0.749)
Schmiedek et al. 2010 0.016 (–0.354, 0.386)
Stepankova et al. 2014 20 session group 0.942 (0.323, 1.562)
Theill et al. 2013 0.217 (–0.494, 0.928)
van Muijden et al. 2012 0.236 (–0.290, 0.761)
von Bastian et al. 2013 0.096 (–0.424, 0.617)
Wilkinsson et al. 2012 –0.526 (–1.139, 0.087)
Zinke et al. 2014 0.791 (0.336, 1.246)
Subgroup 0 (I^2 = 48% , P = 0.013) 0.136 (–0.048, 0.320)
Borella et al. 2010 1.195 (0.522, 1.867)

Carretti et al. 2013 0.386 (–0.274, 1.047)
Subgroup 1 (I^2 = 65% , P = 0.093) 0.785 (0.225, 1.345)
Overall (I^2 = 54% , P = 0.002) 0.192 (0.001, 0.383)
–1 –0.5 0 0.5 1 1.5

Standardized mean difference
Figure 18.2 Forest plot of the effect sizes (and 95% confidence intervals) for the difference between training and control groups (SMD; Hedges g) at
posttest in studies investigating effects of cognitive training on reasoning in older adults. Overall weighted effect size is displayed at the bottom of the
figure. The overall weighted effects for subgroup 0 (upper triangle) does not include the two studies in subgroup 1 that report outlying net effect sizes
(see Figure 18.1).
469
Studies Estimate (95% C.I.)
Ackerman et al. 2010 0.466 (0.016, 0.916)

Basak et al. 2008 –0.133 (–0.770, 0.503)
Boot et al. 2013 –0.785 (–1.454, –0.116)
Borella et al. 2014 young–old 0.170 (–0.451, 0.791)
Borella et al. 2013 0.000 (–0.653, 0.653)
Brehmer et al. 2012 0.245 (–0.349, 0.839)
Burki et al. 2014 –0.441 (–1.054, 0.172)
Dahlin et al. 2008 –0.108 (–0.840, 0.624)
Heinzel et al. 2014 –0.097 (–0.813, 0.620)
Karbach et al. 2009 –0.322 (–0.953, 0.310)
Richmond et al. 2011 0.556 (–0.077, 1.188)
Schmiedek et al. 2010 –0.051 (–0.421, 0.319)
Stepankova et al. 2014 20 session group 0.362 (–0.231, 0.955)
Theill et al. 2013 0.122 (–0.588, 0.832)
van Muijden et al. 2012 –0.336 (–0.863, 0.191)
von Bastian et al. 2013 0.474 (–0.053, 1.001)
Wilkinsson et al. 2012 –0.748 (–1.368, –0.127)
Zinke et al. 2014 0.127 (–0.312, 0.565)
Subgroup 0 (I^2 = 41% , P = 0.036) –0.006 (–0.179, 0.167)
Borella et al. 2010 0.066 (–0.554, 0.686)

Carretti et al. 2013 –0.281 (–0.938, 0.377)
Subgroup 1 (I^2 = 0% , P = 0.453) –0.097 (–0.549, 0.354)
Overall (I^2 = 36% , P = 0.058) –0.012 (–0.172, 0.147)
–1 –0.5 0 0.5 1
Standardized mean difference
Figure 18.3 Forest plot of the effect sizes (and 95% confidence intervals) for the difference between training and control groups (SMD; Hedges
g) at pretest in studies investigating effects of cognitive training on reasoning in older adults. Overall weighted effect size is displayed at the bottom
of the figure. The overall weighted effects for subgroup 0 (upper triangle) does not include the two studies in subgroup 1 that report outlying net
effect sizes (see Figure 18.1).
470
and Verhaeghen (2014). The effect is not very robust. In particular, the exclusion of
two extreme effect sizes reduced the overall effect to a nonâ•‰significant point estimate
of around 0.13 SD. Similar positive outliers were also included in the analysis by
Karbach and Verhaeghen (2014). The effect size tends to be lower than a report of a
significant average effect of working memory training on reasoning performance in
younger adults of 0.24 SD (Au et al., 2015). Note however that the effect size in these
studies of younger adults that had an active control group was only 0.06. In our anal-
ysis of studies on older adults, the difference between an active and passive control
group was, however, minimal. Though power is low with only 20 studies and with,
at best, a small overall effect, it is worrisome that theoretically-â•‰predicted moderators
of the effects, such as for example length of training (Lövdén et al., 2010a), are not
associated with effect sizes (see also Au et al., 2015; Karbach and Verhaeghen, 2014).
We thus conclude that cognitive training, as currently implemented, at best has a very
small effect on cognitive processing efficiency, as indexed by transfer to reasoning
tasks, in old age. A firm conclusion of the trustworthiness of this effect must await
accumulation of more studies. We also note that almost all of the published studies
are seriously underpowered (a total sample of around 200 subjects is needed to detect
a net effect of 0.2 SD with a power of 0.8; the power for the typical study with a total
sample size of 40 subjects is only around 0.20 (Faul et al., 2009)). This fact substan-
tially limits the value of reviewing results from individual studies in this field. At
the same time, we note that meta-â•‰analyses are no methodological remedy for flawed
studies. Several factors, such as publication bias, which we tested for, but also other
confounds that are harder to detect, such as selective reporting of only significant
findings, may positively bias the average effect size. Finally, we note that, with only
observed indicators of reasoning ability, it is difficult to entirely exclude that ability-â•‰
extraneous changes, such as strategy improvements, influence the measures of rea-
soning. To reduce this problem, future studies should obtain several measures of the
target ability and form a factor of the common variance of these measures (Noack et
al., 2009; Lövdén et al., 2010a; Schmiedek et al., 2010; Noack et al., 2014).
Effects of Intellectual Engagement on the Brain in Old Age
Understanding how the brain responds structurally to cognitive activity and how such
changes relate to cognitive performance may provide much credibility to the engage-
ment hypothesis by providing feasible mechanistic pathways. Reports from cross-â•‰
sectional studies have shown that individual differences in participation in leisure
activities, including cognitively demanding activities, in old age are related to individ-
ual differences in a variety of brain measures, including total brain volume, grey matter
volume, white matter volume, white matter lesions (Hafsteinsdottir et al., 2012), and
beta-â•‰amyloid deposition (Landau et al., 2012). Interestingly, the associations between
activity levels (both physical and cognitive) and cognition, as well as with beta-â•‰amy-
loid deposition, have been reported to be stronger for individuals who have greater
genetic risk regarding cognitive impairment and dementia (most notably APOE ε4
carriers; (Kivipelto et al., 2008; Head et al., 2012; Ferencz et al., 2014; Wirth et al.,
2014). Under the assumption that a sedentary lifestyle is a phylogenetically recent
471
phenomenon, these results suggest that this variation in genetic risk has remained in the
population because it is relatively inconsequential under non-sedentary living condi-
tions and now becomes effective for some individuals (Raichlen and Alexander, 2014).
It may even have been advantageous for offspring not to carry the burden of an older
individual that for some reason (e.g., injury) cannot remain active and contribute to
bringing food on the table. This line of reasoning is consistent with the notion that so-
called “vulnerability alleles” may also serve as plasticity alleles (Belsky et al., 2009).
The association between educational attainment, as a proxy for early exposure to
cognitively demanding activities, and brain variables has also been studied quite inten-
sively. For cognitive performance in healthy aging, educational exposure is related to
levels of cognitive performance, but not to trajectories of change, in old age (Lövdén
et al., 2004; Zahodne et al., 2011). Causal pathways are likely to be complex here,
but there is evidence from natural quasi-experiments that education may partly serve
to improve cognitive performance (Ceci, 1991; Cliffordson and Gustafsson, 2008;
Brinch and Galloway, 2012). When triangulating individual differences in education
and cognition with measures of brain integrity, interesting patterns have emerged. For
example, Bennett et al. (2003) reported a study of older Catholic priests, nuns, and
monks who underwent annual clinical evaluations and brain autopsy at death. A com-
posite index of amyloid plaques and neurofibrillary tangles (i.e., classic Alzheimer’s
disease pathology) was associated with lower level of cognitive function in close
proximity to death. This association was, however, smaller in participants with higher
levels of education. This finding suggests that cognitive activities may result in long-
term advantages (e.g., availability of alternative cognitive strategies, better integrity
of other aspects of the brain that matter for cognitive performance) that may offset
the effect of this type of pathology on cognitive performance. In general, this notion
has been supported in studies of both education (Wilson et al., 2004) and other types
of cognitive activity (Scarmeas et al., 2003; Helzner et al., 2007; Hall et al., 2009).
The relative absence of long-term longitudinal brain data in this domain of research
is, however, a major shortcoming. To our knowledge, the few published longitudinal
studies with extensive brain measurements and activity measures have only longitudi-
nal data of activities and not of brain structure (e.g., Gow et al., 2012c; Vaughan et al.,
2014), which limit the conclusions that can be drawn.
Again, effects of training studies offer a more direct test of whether and how cog-
nitive engagement affects the brain. In humans, a large body of literature has reported
that regional grey-matter volume and cortical thickness, probed with T1-weighted
magnetic resonance (MR) imaging, changes in response to motor (Draganski et al.,
2004), cognitive (Draganski et al., 2006), and physical (Erickson et al., 2011) activity
in younger adults. The biological nature, behavioral correlates, and time-course of
these changes are, however, largely unknown (for reviews, see May, 2011; Zatorre
et al., 2012; Lövdén et al., 2013). Evidence on effects of cognitive activity on brain
volume in older adults is also scarce. In one of the few available studies, Engvig et
al. (2010) studied middle-aged and older adults taking part in an 8-week training
regimen in a mnemonic (the Method of Loci) aimed at improving episodic mem-
ory. Compared to controls, the trained persons showed a regional increase of cor-
tical thickness in right insula, left lateral orbitofrontal cortex, and fusiform cortex.
Increases in right fusiform and lateral orbitofrontal cortex were related to larger
472
improvement in memory performance. Lövdén et al. (2012) investigated the effects of

spatial navigation training on hippocampal volume and integrity in younger and older
men. The training group navigated in a virtual world while walking on a treadmill
for 45 minutes every other day over a period of four months. A walking-only control
group was also included. Results showed navigation-related performance gains and
stable hippocampal volume that were also maintained four months after termination
of training. In contrast, control groups showed the typical age-related hippocampal
decrease in volume. Follow-up analyses revealed training-related cortical thickening
in precuneus and paracentral lobule in younger, but not in older, participants (Wenger
et al., 2012). In the COGITO study (Schmiedek et al., 2010), younger and older adults
trained for a total of 101 1-hour sessions on a set of working memory, episodic mem-
ory, and perceptual speed tasks. Using data from this study, Raz et al. (2013) showed
that cognitive training was associated with less decrease of cerebellar volumes, but
that training did not modify cortical volume changes. Thus, also older adults may
display experience-dependent changes in grey matter structure, but the magnitude of
these changes may sometimes be reduced. This pattern is consistent with correspond-
ing animal work (for a review, see Lövdén et al., 2013).
Animal research shows that also the brain’s white matter can be shaped by expe-
rience (Fields, 2008). In humans, amount of piano practicing in childhood and early
adulthood relates to white-matter microstructure, as assessed with diffusion-tensor
imaging (DTI; Bengtsson et al., 2005). Practicing juggling (Scholz et al., 2009), med-
itation (Tang et al., 2010), and reasoning (Mackey et al., 2012) in younger adulthood
also results in microstructural changes in regional white matter of the brain. Lövdén
et al. (2010b) reported that such experience-dependent plasticity extends into old age.
In this study, younger and older adults trained for a total of 101 1-hour sessions on a
set of working memory, episodic memory, and perceptual speed tasks. As compared
with a control group, training affected several DTI metrics and increased the area of
the anterior, but not the posterior, part of corpus callosum. These brain changes were
of similar magnitude in both age groups. Effects on white-matter microstructure in old
age have also been observed after other types of cognitive interventions (Strenziok et
al., 2014), including training that has been mainly strategy based (Engvig et al., 2012;
Chapman et al., 2015). For example, Engvig et al. (2012) reported training-related
changes in DTI metrics from their study of method-of-loci training. Participants in
the training group showed a relative increase in fractional anisotropy (FA), a measure
of density and coherence of the white matter tissue, in a frontal region compared with
a decrease in controls. Increases in memory performance correlated with changes in
FA. Though biological interpretations of changes in DTI metrics are difficult, the pat-
tern of changes (i.e., primarily decreases of radial diffusivity) in some of these studies
suggests a role for myelin-related processes in plasticity of white matter (Lövdén et
al., 2010b; Engvig et al., 2012).
Considering the theoretical importance of efficient large-scale connectivity for
higher-order cognition in general (Fields, 2008) and for cognitive performance in
aging (O’Sullivan et al., 2001; Andrews-Hanna et al., 2007; Bartzokis, 2011; but see
Lövdén et al., 2014), as well as effects of cognitive activity on white-matter structure,
attention to effects of training on functional connectivity is warranted. In line with
this notion, Anguera et al. (2013) reported that older adults training in dual tasking
473
increased their task-related long-range phase (theta) coherence, as assessed with EEG,
between frontal and parietal areas. In another study, Chapman et al. (2015) found
increases in cerebral blood flow, particularly in the default mode network and the
central executive network, as well as greater connectivity at rest in these networks,
as observed with functional MR imaging, after strategy-based reasoning training
(see also Li et al., 2014). Other studies have, however, also observed decrease in
connectivity in a ventral attention network at rest (Strenziok et al., 2014). Studies
of training-related changes in classic measures of functional activity in old age are
delivering even worse scattering of results. Just as in studies of younger adults (Kelly
and Garavan, 2005), practice on cognitive tasks has been associated with both corti-
cal activity decreases (e.g., Dahlin et al., 2008a; Brehmer et al., 2011) and increases
(e.g., Erickson et al., 2007; for reviews, see Lustig et al., 2009; Brehmer et al., 2014).
In a study by Erickson et al. (2007), performance increases after dual-task training
were related to activity increases during the dual-task condition in left ventro-lateral
frontal cortex and in right dorsolateral prefrontal cortex among trained older adults.
Dahlin et al. (2008b) reported training-related cortical activation decreases (in right
anterior prefrontal cortex, right somatosensory association cortex, and right supram-
arginal gyrus) in younger and older adults after five weeks of updating training. In
addition to these cortical decreases, younger adults who, in contrast to older adults,
also improved in a near-transfer task, showed striatal increases in both the trained
and the transfer task, but these changes were not observed in older adults. Increases
in striatal activity after updating training in younger adults have later been replicated,
but are also shown to decrease again after an initial increase (Kuhn et al., 2013). These
findings suggest that the time course of training-related brain changes needs to stud-
ied with better resolution to develop, test, and refine cerebral models of learning and
transfer (see also Lövdén et al., 2013).
Mechanisms behind the Influence of Intellectual

Engagement in Aging
Which are the brain mechanisms that mediate potential effects of cognitive activ-
ity on cognitive processing efficiency in aging? From a general perspective, these
mechanisms may come in three major and complementary forms (Barulli and Stern,
2013): (a) through improving brain functioning and performance in younger age with-
out altering brain aging per se (Satz, 1993; Lövdén et al., 2010a); (b) by aiding com-
pensatory reactions to primary brain aging (Baltes et al., 1999; Stern, 2002; Park and
Reuter-Lorenz, 2009; Stern, 2009); and (c) by fostering maintenance of a young-adult
like brain in old age (Nyberg et al., 2012).
According to the first form, cognitive activities during the life course (e.g., edu-
cation, occupation) may improve brain integrity in the life period during which they
operate (Satz, 1993; Stern, 2002, 2009). To the extent that these improvements are
maintained, it will take more time before a critical threshold for functional impair-
ments (e.g., compromised independence, dementia diagnosis) will be reached. This
simple but important point is sometimes called “brain reserve” (Satz, 1993; Stern,
2002, 2009).
474
Lifetime cognitive activities may also aid compensatory reactions to primary

aging, so that effects of brain aging on performance may be reduced in old age, which
implies different trajectories of performance change in old age for individuals with
the same change in brain integrity but who differ in life time cognitive activity. This
notion, often referred to as cognitive reserve (Stern, 2002, 2009) or simply flexibility
(Lövdén et al., 2010a), holds that lifetime cognitive activities (e.g., education) may
give individuals better opportunities to handle the negative effects of brain aging on
cognitive performance, perhaps by giving the individual a larger and more flexible set
of neurophysiological routes and cognitive skills (e.g., knowledge and strategies) to
handle different situations (see also Lövdén et al., 2010a). This notion has received
tentative support by studies reporting a negative association between a cognitively
engaged lifestyle and brain integrity in some groups (e.g., in dementia; Scarmeas et
al., 2003) when cognitive performance is controlled for.
In the likely absence of an active gene program that causes human aging (Kirkwood,
2005), the major goals in the cognitive neuroscience of aging must be to identify the
mechanisms causing damage to accumulate in the brain and the variety of mecha-
nisms that operate to protect us from this damage. Related to this view, individuals
that decline less in cognitive performance in aging do not have to be those individuals
who can compensate for decline in brain integrity, but rather those who can main-
tain brain integrity in the first place (Nyberg et al., 2012). According to this “brain
maintenance” view, mechanisms protecting the brain against age-graded risks differ
in number and strength across and within individuals. Certain types of behavior may
lead to neurophysiological effects that modify brain aging directly, but others may
also indirectly protect the brain by offsetting negative conditions. For example, an
engaged lifestyle in old age may confer advantages for brain aging because nega-
tive pathways, such as loneliness, depression, stress, and malnutrition, become less
likely. Cognitive activity may also have direct effects on brain integrity through a
multitude of mechanisms. For example, basic neuroscience work indicates that neural
activity can induce myelination (Fields, 2008; Wake et al., 2011). Cognitive training
may improve white matter integrity (Lövdén et al., 2010b). Myelin undergoes many
negative changes in aging (Bartzokis, 2011), and white matter integrity in general also
shows negative adult age differences (Madden et al., 2012), of which some thus could
be postponed by positive effects of cognitive activities. To the extent that aging of
white-matter integrity contributes to cognitive aging (Salthouse, 2011; Lövdén et al.,
2014), one may thus speculate that cognitive activity could serve to maintain cognitive
functioning through relatively preserved white-matter integrity. Many other similar
direct mechanisms are also available, such as activity-dependent release of growth
factors (Lövdén et al., 2011), alterations of release of neurotransmitters (Bäckman
et al., 2011), and activity-dependent structural changes of the neuron (Zatorre et al.,
2012; Lövdén et al., 2013; Lindenberger, 2014).
At first glance, the brain-maintenance concept seems unable to deal with the find-
ing that commonly measured aging-related brain changes (Salthouse, 2011) and puta-
tive markers of pathology (Boyle, 2013) leave a major portion of age-related cognitive
decline unexplained. However, such findings do not require that a cognitive reserve
account must be evoked to explain the remaining individual differences in cognition.
Rather, the presence of such residual variability may simply mean that our current
475
knowledge of the mappings between brain changes and changes in behavior is incom-
plete. These mappings may also differ across individuals if there is more than one
physiological pathway into cognitive decline. In addition, such mappings may change
within individuals, possibly reflecting selection mechanisms (Lindenberger, 2014; see
also Edelman, 1987; Lautrey, 2003).
In particular, aging-related cognitive decline and dementia are likely to have
related and multifactorial etiologies (Drachman, 2007). Seventy percent of all indi-
viduals suffering from dementia are older than 75 years (Fratiglioni and Qiu, 2011).
A majority of these persons show not only the pathology typically associated with
Alzheimer’s disease (e.g., plaques) but also vascular injuries (Viswanathan et al.,
2009). To this we must add a wide range of known (e.g., changes in neurotransmitter
functioning; Bäckman et al., 2010) and as yet unknown aging-related brain changes,
all of which may contribute to individual differences in late-life cognitive functioning.
Of course, determinants of individual differences in cognitive abilities before aging-
related changes have emerged must also be factored into the equation.
In summary, researchers are confronted with many-to-many mappings between
brain integrity and cognitive abilities in aging and dementia. For example, assume
that the amyloid burden of two individuals is identical. Assume also that one of these
individuals is diagnosed with dementia or has low cognitive performance, whereas
the other individual has no dementia diagnosis or shows high performance. Further
assume that the low-functioning individual has fewer years of education than the
high-functioning individual. Reserve concepts offer viable explanations of such a
scenario. The cognitive reserve notion describes well the possibility that the high-
performing individual may have been able to better cope with the accumulation of
amyloid, perhaps due to the higher flexibility that comes with added years of edu-
cation (Lövdén et al., 2010a). However, an alternative explanation in terms of brain
maintenance seems just as viable: The high-functioning individual may have accu-
mulated fewer vascular injuries—or any number of other unknown, imperfectly mea-
sured, or unmeasured alterations—perhaps reflecting advantageous lifestyle habits
associated with education. Thus, the presence of residual variability in functioning
after accounting for select aspects of age-related brain pathology does not discrim-
inate between the reserve and maintenance views, because several aspects of brain
integrity determine functioning and dementia diagnosis in old age, including those
that have not been observed in the particular study in question, or that have not yet
been discovered.
Thus, although these general models never have been proposed as mechanistic
theories of aging, but rather as general frameworks, further progress in the field is
likely to come from operationally defining the concepts and from the generation of
predictions that can tease the models apart. Future work needs to develop models that
can estimate the contribution of brain reserve, cognitive reserve, and maintenance to
successful cognitive aging. Conceivably, these general models in the cognitive neu-
roscience of aging are complementary, related across individuals, and differing in
importance across the adult life span. To delineate the multiple sources of individ-
ual differences in aging and the potential effects of cognitive activity on cognitive
aging, we need to intensify our efforts at discovering and measuring what matters.
In our view, what matters are the between-person differences in change that we can
476
predict and explain, rather than speculations about why there is variance that we can-
not account for.
Acknowledgments
Martin Lövdén was supported by a program grant from FORTE (2013-â•‰2277) and a
“distinguished younger researcher” grant from the Swedish Research Council (446-â•‰
2013-â•‰7189). Lars Bäckman was supported by grants from the Swedish Research
Council, the Swedish Council for Health, Working Life, and Welfare, an Alexander
von Humbolds Research Award, and a donation from the af Jochnick Foundation. The
authors thank Ylva Köhncke and Martin Bellander for valuable discussions about the
contents of this chapter.
References
Aartsen Mj, Smits Ch, Van Tilburg T, Knipscheer Kc, Deeg Dj (2002) Activity in older
adults: cause or consequence of cognitive functioning? A longitudinal study on everyday
activities and cognitive performance in older adults. J Gerontol B Psychol Sci Soc Sci
57:P153–â•‰P162.
Andel R, Kareholt I, Parker Mg, Thorslund M, Gatz M (2007) Complexity of primary lifetime
occupation and cognition in advanced old age. J Aging Health 19:397–â•‰415.
Andrews-â•‰Hanna Jr, Snyder Az, Vincent Jl, Lustig C, Head D, Raichle Me, Buckner Rl (2007)
Disruption of large-â•‰scale brain systems in advanced aging. Neuron 56:924–â•‰935.
Anguera Ja, Boccanfuso J, Rintoul Jl, Al-â•‰Hashimi O, Faraji F, Janowich J, Kong E, Larraburo Y,
Rolle C, Johnston E, Gazzaley A (2013) Video game training enhances cognitive control
in older adults. Nature 501:97–â•‰101.
Au J, Sheehan E, Tsai N, Duncan GJ, Buschkuehl M, Jaeggi SM (2015) Improving fluid intelli-
gence with training on working memory: a meta-analysis. Psychon Bull Rev 22:366–377.
Ball K, Berch Db, Helmers Kf, Jobe Jb, Leveck Md, Marsiske M, Morris Jn, Rebok Gw, Smith
Dm, Tennstedt Sl, Unverzagt Fw, Willis Sl, Advanced Cognitive Training For I, Vital
Elderly Study G (2002) Effects of cognitive training interventions with older adults: a
randomized controlled trial. JAMA 288:2271–â•‰2281.
Baltes Pb, Staudinger Um, Lindenberger U (1999) Lifespan psychology: theory and application
to intellectual functioning. Annu Rev Psychol 50:471–â•‰507.
Bartlett Fc (1932) Remembering: A study in experimental and social psychology. Cambridge,
UK: Cambridge University press.
Bartzokis G (2011) Alzheimer’s disease as homeostatic responses to age-â•‰related myelin break-
down. Neurobiol Aging 32(8):1341–1371.
Barulli D, Stern Y (2013) Efficiency, capacity, compensation, maintenance, plasticity: emerg-
ing concepts in cognitive reserve. Trends Cogn Sci 17, 502–509.
Basak C, Boot Wr, Voss Mw, Kramer Af (2008) Can training in a real-â•‰time strategy video game
attenuate cognitive decline in older adults? Psychol Aging 23:765–â•‰777.
Beam CR, Turkheimer E (2013) Phenotype-environment correlations in longitudinal twin mod-
els. Development and Psychopathology 25(1):7-16.
Belsky J, Jonassaint C, Pluess M, Stanton M, Brummett B, Williams R (2009) Vulnerability
genes or plasticity genes? Mol Psychiatry 14:746–â•‰754.
477
Bengtsson Sl, Nagy Z, Skare S, Forsman L, Forssberg H, Ullen F (2005) Extensive piano
practicing has regionally specific effects on white matter development. Nat Neurosci
8:1148–1150.
Bennett DA, Wilson RS, Schneider JA, Evans DA, Mendes De Leon CF, Arnold SE, Barnes LL,
Bienias JL (2003) Education modifies the relation of AD pathology to level of cognitive
function in older persons. Neurology 60:1909–1915.
Bielak Aa (2010) How can we not ‘lose it’ if we still don’t understand how to ‘use it’?
Unanswered questions about the influence of activity participation on cognitive perfor-
mance in older age--a mini-review. Gerontology 56:507–519.
Bielak Aa, Gerstorf D, Anstey Kj, Luszcz Ma (2014) Longitudinal associations between
activity and cognition vary by age, activity type, and cognitive domain. Psychol Aging
29:863–874.
Bielak Aam, Hughes Tf, Small Bj, Dixon R (2007) It’s never too late to engage in lifestyle
activities: Significant concurrent but not change relationshships between lifestyle activ-
ities and cognitiev speed. Journal of Gerontology: Psychological Sciences 62B:331–339.
Bonsang E, Adam S, Perelman S (2012) Does retirement affect cognitive functioning? Journal
of Health Economics 31:490–501.
Borella E, Carretti B, Riboldi F, De Beni R (2010) Working memory training in older adults: evi-
dence of transfer and maintenance effects. Psychol Aging 25:767–778.
Bosma H, Van Boxtel Mp, Ponds Rw, Houx Pj, Burdorf A, Jolles J (2003) Mental work demands
protect against cognitive impairment: MAAS prospective cohort study. Exp Aging Res
29:33–45.
Bosma H, Van Boxtel Mp, Ponds Rw, Jelicic M, Houx P, Metsemakers J, Jolles J (2002)
Engaged lifestyle and cognitive function in middle and old-aged, non-demented persons: a
reciprocal association? Z Gerontol Geriatr 35:575–581.
Boyle PA, Wilson RS, Yu L, Barr AM, Honer WG, Schneider JA, Bennett DA (2013) Much
of late life cognitive decline is not due to common neurodegenerative pathologies. Ann
Neurol 74(3):478–489.
Brehmer Y, Kalpouzos G, Wenger E, Lövdén M (2014) Plasticity of brain and cognition in
older adults. Psychol Res 78(6):790–802.
Brehmer Y, Rieckmann A, Bellander M, Westerberg H, Fischer H, Backman L (2011) Neural cor-
relates of training-related working-memory gains in old age. Neuroimage 58:1110–1120.
Brinch Cn, Galloway Ta (2012) Schooling in adolescence raises IQ scores. Proc Natl Acad Sci
U S A 109:425–430.
Bäckman L, Lindenberger U, Li Sc, Nyberg L (2010) Linking cognitive aging to alterations
in dopamine neurotransmitter functioning: Recent data and future avenues. Neurosci
Biobehav R 34:670–677.
Bäckman L, Nyberg L, Soveri A, Johansson J, Andersson M, Dahlin E, Neely As, Virta J, Laine
M, Rinne Jo (2011) Effects of working-memory training on striatal dopamine release.
Science 333:718.
Carretti B, Borella E, Zavagnin M, De Beni R (2013) Gains in language comprehension relating
to working memory training in healthy older adults. Int J Geriatr Psychiatry 28:539–546.
Ceci Sj (1991) How much does schooling influence general intelligence and its cognitive com-
ponents? A reassessment of the evidence. Developmental Psychology 27:703.
Chapman SB, Aslan S, Spence JS, Hart JJ, Jr., Bartz EK, Didehbani N et al. (2015) Neural
mechanisms of brain plasticity with complex cognitive training in healthy seniors. Cerebral
Cortex 25:396–405.
Chase Wg, Simon Ha (1973) The mind’s eye in chess. In: Visual information processing (Chase
Wg, ed), pp 215–281. New York: Academic Press.
478
Cliffordson C, Gustafsson J (2008) Effects of age and schooling on intellectual perfor-

mance: Estimates obtained from analysis of continuous variation in age and length of
schooling. Intelligence 36:143–152.
Dahlin E, Nyberg L, Bäckman L, Neely As (2008a) Plasticity of executive functioning in
young and older adults: immediate training gains, transfer, and long-term maintenance.
Dahlin E, Neely As, Larsson A, Backman L, Nyberg L (2008b) Transfer of learning after updat-
ing training mediated by the striatum. Science 320:1510–1512.
Drachman D (2007) Rethinking Alzheimer’s disease: The role of age-related changes. Curr
Neurol Neurosci 7:265–268.
Draganski B, Gaser C, Busch V, Schuierer G, Bogdahn U, May A (2004) Neuroplasticity: Changes
in grey matter induced by training. Nature 427:311–312.
Draganski B, Gaser C, Kempermann G, Kuhn Hg, Winkler J, Buchel C, May A (2006) Temporal
and spatial dynamics of brain structure changes during extensive learning. J Neurosci
26:6314–6317.
Edelman G (1987) Neural Darwinism: The Theory of Neuronal Group Selection (Basic Books,
New York 1987).
Engvig A, Fjell Am, Westlye Lt, Moberget T, Sundseth O, Larsen Va, Walhovd Kb (2010)
Effects of memory training on cortical thickness in the elderly. Neuroimage 52:1667–1676.
Engvig A, Fjell Am, Westlye Lt, Moberget T, Sundseth O, Larsen Va, Walhovd Kb (2012)
Memory training impacts short-term changes in aging white matter: A Longitudinal
Diffusion Tensor Imaging Study. Hum Brain Mapp 33:2390–2406.
Erickson Ki, Colcombe Sj, Wadhwa R, Bherer L, Peterson Ms, Scalf Pe, Kim Js, Alvarado M,
Kramer Af (2007) Training-induced functional activation changes in dual-task process-
ing: an FMRI study. Cereb Cortex 17:192–204.
Erickson Ki, Voss Mw, Prakash Rs, Basak C, Szabo A, Chaddock L, Kim Js, Heo S, Alves H,
White Sm, Wojcicki Tr, Mailey E, Vieira Vj, Martin Sa, Pence Bd, Woods Ja, Mcauley E,
Kramer Af (2011) Exercise training increases size of hippocampus and improves memory.
P Natl Acad Sci USA 108:3017–3022.
Faisal Aa, Selen Lp, Wolpert Dm (2008) Noise in the nervous system. Nat Rev Neurosci
9:292–303.
Faul F, Erdfelder E, Buchner A, Lang Ag (2009) Statistical power analyses using G*Power
3.1: tests for correlation and regression analyses. Behav Res Methods 41:1149–1160.
Ferencz B, Laukka Ej, Welmer Ak, Kalpouzos G, Angleman S, Keller L, Graff C, Lovden M,
Backman L (2014) The benefits of staying active in old age: physical activity counteracts
the negative influence of PICALM, BIN1, and CLU risk alleles on episodic memory func-
tioning. Psychol Aging 29:440–449.
Fields Rd (2008) White matter in learning, cognition and psychiatric disorders. Trends Neurosci
31:361–370.
Finkel D, Andel R, Gatz M, Pedersen Nl (2009) The role of occupational complexity in
trajectories of cognitive aging before and after retirement. Psychology and Aging
24:563–573.
Fisher Gg, Stachowski A, Infurna Fj, Faul Jd, Grosch J, Tetrick Le (2014) Mental work demands,
retirement, and longitudinal trajectories of cognitive functioning. Journal of occupational
health psychology 19:231–242.
Flynn Jr (1984) The mean IQ of Americans: Massive gains from 1932 to 1978. Psychological
Bulletin 95:29–51.
Fratiglioni L, Qiu C (2011) Prevention of cognitive decline in ageing: dementia as the target,
delayed onset as the goal. Lancet Neurol 10:778–779.
479
Ghisletta P, Bickel Jf, Lovden M (2006) Does activity engagement protect against cognitive
decline in old age? Methodological and analytical considerations. J Gerontol B-Psychol
61:P253-P261.
Giedd Jn, Blumenthal J, Jeffries No, Castellanos Fx, Liu H, Zijdenbos A, Paus T, Evans Ac,
Rapoport Jl (1999) Brain development during childhood and adolescence: a longitudinal
MRI study. Nat Neurosci 2:861–863.
Glymour Mm, Weuve J, Berkman Lf, Kawachi I, Robins Jm (2005) When is baseline adjust-
ment useful in analyses of change? An example with education and cognitive change. Am
J Epidemiol 162:267–278.
Gow Aj, Mortensen El, Avlund K (2012a) Activity participation and cognitive aging from age
50 to 80 in the glostrup 1914 cohort. J Am Geriatr Soc 60:1831–1838.
Gow Aj, Avlund K, Mortensen El (2014) Occupational characteristics and cognitive aging in
the Glostrup 1914 Cohort. J Gerontol B Psychol Sci Soc Sci 69:228–236.
Gow Aj, Corley J, Starr Jm, Deary Ij (2012b) Reverse causation in activity-cognitive ability
associations: the Lothian Birth Cohort 1936. Psychol Aging 27:250–255.
Gow Aj, Bastin Me, Munoz Maniega S, Valdes Hernandez Mc, Morris Z, Murray C, Royle
Na, Starr Jm, Deary Ij, Wardlaw Jm (2012c) Neuroprotective lifestyles and the aging
brain: activity, atrophy, and white matter integrity. Neurology 79:1802–1808.
Hafsteinsdottir Sh, Eiriksdottir G, Sigurdsson S, Aspelund T, Harris Tb, Launer Lj, Gudnason V
(2012) Brain tissue volumes by APOE genotype and leisure activity-the AGES-Reykjavik
Study. Neurobiol Aging 33:829 e821–e828.
Hall Cb, Lipton Rb, Sliwinski M, Katz Mj, Derby Ca, Verghese J (2009) Cognitive activities
delay onset of memory decline in persons who develop dementia. Neurology 73:356–361.
Head D, Bugg Jm, Goate Am, Fagan Am, Mintun Ma, Benzinger T, Holtzman Dm, Morris
Jc (2012) Exercise Engagement as a Moderator of the Effects of APOE Genotype on
Amyloid Deposition. Arch Neurol 69:636–643.
Heishman Sj, Kleykamp Ba, Singleton Eg (2010) Meta-analysis of the acute effects of nicotine
and smoking on human performance. Psychopharmacology (Berl) 210:453–469.
Helzner Ep, Scarmeas N, Cosentino S, Portet F, Stern Y (2007) Leisure activity and cognitive
decline in incident Alzheimer disease. Arch Neurol 64:1749–1754.
Hertzog C, Kramer Af, Wilson Rs, Lindenberger U (2009) Enrichment effects on adults cogni-
tive development. Psychological Science in the Public Interest 9:1–65.
Hoyland A, Dye L, Lawton Cl (2009) A systematic review of the effect of breakfast on the cog-
nitive performance of children and adolescents. Nutrition research reviews 22:220–243.
Hultsch Df, Hertzog C, Small Bj, Dixon Ra (1999) Use it or lose it: Engaged lifestyle as a buffer
of cognitive decline in aging? Psychology and Aging 14:245–263.
Jones He, Conrad Hs (1933) The growth and decline of intelli-gence: A study of a homogeneous
group between the ages of ten and sixty. Genetic Psychology Monographs 13:223–275.
Jorm Af, Rodgers B, Henderson As, Korten Ae, Jacomb Pa, Christensen H, Mackinnon A
(1998) Occupation type as a predictor of cognitive decline and dementia in old age. Age
Ageing 27:477–483.
Karbach J, Verhaeghen P (2014). Making working memory work: a meta-analysis of exec-
utive-control and working memory training in older adults. Psychological Science
25(11):2027–2037.
Karr JE, Areshenkoff CN, Rast P, Garcia-Barrera MA (2014) An empirical comparison of the
therapeutic benefits of physical exercise and cognitive training on the executive functions
of older adults: a meta-analysis of controlled trials. Neuropsychology 28:829–845.
Kelly Am, Garavan H (2005) Human functional neuroimaging of brain changes associated with
practice. Cereb Cortex 15:1089–1102.
480
Kelly Me, Loughrey D, Lawlor Ba, Robertson Ih, Walsh C, Brennan S (2014) The impact
of cognitive training and mental stimulation on cognitive and everyday functioning of
healthy older adults: a systematic review and meta-analysis. Ageing Res Rev 15:28–43.
Kirkwood Tb (2005) Understanding the odd science of aging. Cell 120:437–447.
Kivipelto M, Rovio S, Ngandu T, Kareholt I, Eskelinen M, Winblad B, Hachinski V, Cedazo-
Minguez A, Soininen H, Tuomilehto J, Nissinen A (2008) Apolipoprotein E epsilon4
magnifies lifestyle risks for dementia: a population-based study. Journal of Cellular and
Molecular Medicine 12:2762–2771.
Kueider Am, Parisi Jm, Gross Al, Rebok Gw (2012) Computerized cognitive training with older
adults: a systematic review. Plos One 7:e40588.
Kuhn S, Schmiedek F, Noack H, Wenger E, Bodammer Nc, Lindenberger U, Lovden M (2013)
The dynamics of change in striatal activity following updating training. Hum Brain Mapp
34:1530–1541.
Landau SM, Marks SM, Mormino EC, Rabinovici GD, Oh H, O’neil JP, Wilson RS, Jagust WJ
(2012) Association of lifetime cognitive engagement and low beta-amyloid deposition.
Lautrey J (2003) A pluralistic approach to cognitive differentiation and development. In
(Sternberg, Robert J, Lautrey, Jacques, Lubart, Todd I. eds.); Models of intelligence:
International perspectives (pp. 117–131). Washington, DC, US: American Psychological
Association xvi:368.
Li R, Zhu X, Yin S, Niu Y, Zheng Z, Huang X, Wang B, Li J (2014) Multimodal intervention in
older adults improves resting-state functional connectivity between the medial prefrontal
cortex and medial temporal lobe. Front Aging Neurosci 6:39.
Lindenberger U (2014) Human cognitive aging: Corriger la fortune? Science 346:572–578.
Lustig C, Shah P, Seidler R, Reuter-Lorenz PA (2009) Aging, training, and the brain: a review
and future directions. Neuropsychol Rev 19:504–522.
Lövdén M, Ghisletta P, Lindenberger U (2005) Social participation attenuates decline in per-
ceptual speed in old and very old age. Psychology and Aging 20:423–434.
Lövdén M, Bäckman L, Lindenberger U, Schaefer S, Schmiedek F (2010a) A theoretical frame-
work for the study of adult cognitive plasticity. Psychological Bulletin 136:659–676.
Lövdén M, Wenger E, Mårtensson J, Lindenberger U, Bäckman L (2013) Structural brain plas-
ticity in adult learning and development. Neurosci Biobehav Rev 37:2296–2310.
Lövdén M, Ronnlund M, Wahlin A, Backman L, Nyberg L, Nilsson Lg (2004) The extent of
stability and change in episodic and semantic memory in old age: Demographic predictors
of level and change. Journals of Gerontology Series B: Psychological Sciences and Social
Sciences 59:P130-P134.
Lövdén M, Köhncke Y, Laukka Ej, Kalpouzos G, Salami A, Li Tq, Fratiglioni L, Bäckman L
tract are associated in very old age. Neuroimage 102P2:520–530.
Lövdén M, Schaefer S, Noack H, Bodammer Nc, Kuhn S, Heinze Hj, Duzel E, Backman L,
Lindenberger U (2012) Spatial navigation training protects the hippocampus against age-
related changes during early and late adulthood. Neurobiol Aging 33:620 e629-620 e622.
Lövdén M, Bodammer Nc, Kuhn S, Kaufmann J, Schutze H, Tempelmann C, Heinze Hj, Duzel
E, Schmiedek F, Lindenberger U (2010b) Experience-dependent plasticity of white-matter
microstructure extends into old age. Neuropsychologia 48:3878–3883.
Lövdén M, Schaefer S, Noack H, Kanowski M, Kaufmann J, Tempelmann C, Bodammer Nc,
Kuhn S, Heinze Hj, Lindenberger U, Duzel E, Backman L (2011) Performance-related
increases in hippocampal N-acetylaspartate (NAA) induced by spatial navigation training
are restricted to BDNF Val homozygotes. Cereb Cortex 21:1435–1442.
481
Mackey Ap, Whitaker Kj, Bunge Sa (2012) Experience-dependent plasticity in white matter
microstructure: reasoning training alters structural connectivity. Front Neuroanat 6:32.
Madden Dj, Bennett Ij, Burzynska A, Potter Gg, Chen Nk, Song Aw (2012) Diffusion ten-
1822:386–400.
Marquie Jc, Duarte Lr, Bessieres P, Dalm C, Gentil C, Ruidavets Jb (2010) Higher mental stim-
ulation at work is associated with improved cognitive functioning in both young and older
workers. Ergonomics 53:1287–1301.
May A (2011) Experience-dependent structural plasticity in the adult human brain. Trends
Cogn Sci 15:475–482.
Mazzonna F, Peracchi F (2012) Ageing, cognitive abilities and retirement. European Economic
Review 56:691–710.
Mcardle Jj, Hamagami F (2001) Latent difference score structural models for linear dynamic
analyses with incomplete data. In: New methods for the analysis of change (Collins Lm,
Sayer Ag, eds), pp 139–175. Washington, DC.: American Psychological Association.
Mitchell MB, Cimino CR, Benitez A, Brown CL, Gibbons LE, Kennison RF, Shirk SD, Atri A,
Robitaille A, Macdonald SW, Lindwall M, Zelinski EM, Willis SL, Schaie KW, Johansson
B, Dixon RA, Mungas DM, Hofer SM, Piccinin AM (2012) Cognitively stimulating activ-
ities: effects on cognition across four studies with up to 21 years of longitudinal data.
Journal of Aging Research 2012:461592.
Nehlig A (2010) Is caffeine a cognitive enhancer? J Alzheimers Dis 20 Suppl 1:S85–S94.
Noack H, Lovden M, Schmiedek F (2014) On the validity and generality of transfer effects in
cognitive training research. Psychological research.
Noack H, Lovden M, Schmiedek F, Lindenberger U (2009) Cognitive plasticity in adulthood
and old age: Gauging the generality of cognitive intervention effects. Restor Neurol
Neuros 27:435–453.
O’sullivan M, Jones Dk, Summers Pe, Morris Rg, Williams Scr, Markus Hs (2001) Evidence
57:632–638.
Park Dc, Reuter-Lorenz P (2009) The adaptive brain: aging and neurocognitive scaffolding.
Park Dc, Lodi-Smith J, Drew L, Haber S, Hebrank A, Bischof Gn, Aamodt W (2014) The
impact of sustained engagement on cognitive function in older adults: the Synapse Project.
Psychol Sci 25:103–112.
Rabbitt P, Osman P, Moore B, Stollery B (2001) There are stable individual differences in
performance variability, both from moment to moment and from day to day. Q J Exp
Psychol-A 54:981–1003.
Raichlen Da, Alexander Ge (2014) Exercise, APOE genotype, and the evolution of the human
lifespan. Trends Neurosci 37:247–255.
Raz N, Schmiedek F, Rodrigue Km, Kennedy Km, Lindenberger U, Lövdén M (2013)
Differential brain shrinkage over 6 months shows limited association with cognitive prac-
tice. Brain Cogn 82:171–180.
ual differences and modifiers. Cereb Cortex 15:1676–1689.
Reijnders J, Van Heugten C, Van Boxtel M (2013) Cognitive interventions in healthy older
adults and people with mild cognitive impairment: a systematic review. Ageing Res Rev
12:263–275.
482
Roberts Ba, Fuhrer R, Marmot M, Richards M (2011) Does retirement influence cognitive
performance? The Whitehall II Study. Journal of epidemiology and community health
65:958–963.
Rohwedder S, Willis RJ (2010) Mental retirement. The Journal of Economic Perspectives: A
Journal of the American Economic Association 24:119–138.
Salthouse Ta (2006) Mental exercise and mental aging: Evaluating the evidence of the “use it or
lose it” hypothesis Perspectives on Psychological Science 1:68–87.
Salthouse Ta (2011) Neuroanatomical substrates of age-related cognitive decline. Psychological
Bulletin 137:753–784.
Salthouse Ta, Berish De, Miles Jd (2002) The role of cognitive stimulation on the relations
between age and cognitive functioning. Psychology and Aging 17:548–557.
Satz P (1993) Brain reserve capacity on symptom onset after brain injury: A formulation and
review of evidence for threshold theory,. Neuropsychology 7:273–295.
Scarmeas N, Zarahn E, Anderson Ke, Habeck Cg, Hilton J, Flynn J, Marder Ks, Bell Kl,
Sackeim Ha, Van Heertum Rl, Moeller Jr, Stern Y (2003) Association of life activities with
cerebral blood flow in Alzheimer disease -Implications for the cognitive reserve hypothe-
sis. Arch Neurol-Chicago 60:359–365.
Schaie Kw (1994) The Course of Adult Intellectual Development. Am Psychol 49:304–313.
Schaie Kw (2005) Developmental influences on adult intelligence: The Seattle longitudinal
study. New York: Oxford University Press.
Schmiedek F, Lovden M, Lindenberger U (2010) Hundred days of cognitive training enhance
broad cognitive abilities in adulthood: findings from the COGITO study. Frontiers in
Aging Neuroscience 2.
Schmiedek F, Lovden M, Lindenberger U (2013) Keeping it steady: older adults perform more
consistently on cognitive tasks than younger adults. Psychol Sci 24:1747–1754.
Scholz J, Klein Mc, Behrens Te, Johansen-Berg H (2009) Training induces changes in white-
matter architecture. Nat Neurosci 12:1370–1371.
Schooler C, Mulatu Ms (2001) The reciprocal effects of leisure time activities and intellectual
functioning in older people: a longitudinal analysis. Psychol Aging 16:466–482.
Schooler C, Mulatu Ms, Oates G (1999) The continuing effects of substantively complex work
on the intellectual functioning of older workers. Psychol Aging 14:483–506.
Small Bj, Dixon Ra, Mcardle Jj, Grimm Kj (2012) Do changes in lifestyle engagement mod-
erate cognitive decline in normal aging? Evidence from the Victoria Longitudinal Study.
cept. J Int Neuropsychol Soc 8:448–460.
Stern Y (2009) Cognitive reserve. Neuropsychologia 47:2015–2028.
Sterne Ja, Egger M (2001) Funnel plots for detecting bias in meta-analysis: guidelines on
choice of axis. Journal of clinical epidemiology 54:1046–1055.
Stine-Morrow Ea, Parisi Jm, Morrow Dg, Park Dc (2008) The effects of an engaged lifestyle on
cognitive vitality: a field experiment. Psychol Aging 23:778–786.
Strenziok M, Parasuraman R, Clarke E, Cisler Ds, Thompson Jc, Greenwood Pm (2014)
Neurocognitive enhancement in older adults: comparison of three cognitive training
tasks to test a hypothesis of training transfer in brain connectivity. Neuroimage 85 Pt
3:1027–1039.
Tang Y-Y, Lu Q, Geng X, Stein Ea, Yang Y, Posner Mi (2010) Short-term meditation induces
white matter changes in the anterior cingulate. P Natl Acad Sci USA 107:15649–15652.
Tranter Lj, Koutstaal W (2008) Age and flexible thinking: an experimental demonstration of
the beneficial effects of increased cognitively stimulating activity on fluid intelligence in
healthy older adults. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn 15:184–207.
483
Van Der Elst W, Van Boxtel Mp, Jolles J (2012) Occupational activity and cognitive aging: a
case-control study based on the Maastricht Aging Study. Exp Aging Res 38:315–329.
Vaughan L, Erickson KI, Espeland MA, Smith JC, Tindle HA, Rapp SR (2014) Concurrent and
longitudinal relationships between cognitive activity, cognitive performance, and brain
volume in older adult women. J Gerontol B Psychol Sci Soc Sci 69(6):826–836.
Verhaeghen P, Marcoen A, Goossens L (1992) Improving memory performance in the aged
through mnemonic training -a meta-analytic study. Psychology and Aging 7:242–251.
Viswanathan A, Rocca Wa, Tzourio C (2009) Vacular risk factors and dementia. Neurology
72:368–374.
Wake H, Lee Pr, Fields Rd (2011) Control of local protein synthesis and initial events in myelin-
ation by action potentials. Science 333:1647–1651.
Wallace Bc, Issa Jd, Thomas Ak, Lau J, Trow P, Schmid Ch (2012) Closing the gap between
methodologists and end-users: R as a computational back-end. Journal of Statistical
Software 49:5.
Wang Hx, Jin Y, Hendrie Hc, Liang C, Yang L, Cheng Y, Unverzagt Fw, Ma F, Hall Ks, Murrell
Jr, Li P, Bian J, Pei Jj, Gao S (2013) Late life leisure activities and risk of cognitive
decline. The journals of gerontology Series A, Biological sciences and medical sciences
68:205–213.
Wenger E, Schäfer S, Noack H, Kühn S, Mårtensson J, Heinze Hj, Duzel E, Bäckman L,
Lindenberger U, Lövdén M (2012) Cortical thickness changes following spatial navigation
training in adulthood and aging. Neuroimage 59:3389–3397.
Wilson Rs, Bennett Da, Bienias Jl, Mendes De Leon Cf, Morris Mc, Evans Da (2003) Cognitive
activity and cognitive decline in a biracial community population. Neurology 61:812–816.
Wilson Rs, Li Y, Aggarwal Nt, Barnes Ll, Mccann Jj, Gilley Dw, Evans Da (2004) Education
and the course of cognitive decline in Alzheimer disease. Neurology 63:1198–1202.
Wirth M, Villeneuve S, La Joie R, Marks Sm, Jagust Wj (2014) Gene-environment interac-
tions: lifetime cognitive activity, APOE genotype, and beta-amyloid burden. J Neurosci
34:8612–8617.
Zahodne Lb, Glymour Mm, Sparks C, Bontempo D, Dixon Ra, Macdonald Sw, Manly Jj (2011)
Education does not slow cognitive decline with aging: 12-year evidence from the victoria
longitudinal study. J Int Neuropsychol Soc 17:1039–1046.
Zatorre Rj, Fields Rd, Johansen-Berg H (2012) Plasticity in gray and white: neuroimaging
changes in brain structure during learning. Nat Neurosci 15:528–536.
Papers Included in Meta-Analysis

Ackerman Pl, Kanfer R, Calderwood C (2010) Use it or lose it? Wii brain exercise practice and
reading for domain knowledge. Psychol Aging 25:753–766.
Basak C, Boot Wr, Voss Mw, Kramer Af (2008) Can training in a real-time strategy video game
attenuate cognitive decline in older adults? Psychol Aging 23:765–777.
Boot Wr, Champion M, Blakely Dp, Wright T, Souders Dj, Charness N (2013) Video games as
a means to reduce age-related cognitive decline: attitudes, compliance, and effectiveness.
Frontiers in psychology 4:31.
Borella E, Carretti B, Riboldi F, De Beni R (2010) Working memory training in older adults: evi-
dence of transfer and maintenance effects. Psychol Aging 25:767–778.
Borella E, Carretti B, Zanoni G, Zavagnin M, De Beni R (2013) Working memory training
in old age: an examination of transfer and maintenance effects. Arch Clin Neuropsychol
28:331–347.
Borella E, Carretti B, Cantarella A, Riboldi F, Zavagnin M, De Beni R (2014) Benefits of train-
ing visuospatial working memory in young-old and old-old. Dev Psychol 50:714–727.
484
Brehmer Y, Westerberg H, Backman L (2012) Working-memory training in younger and older

adults: training gains, transfer, and maintenance. Front Hum Neurosci 6:63.
Burki Cn, Ludwig C, Chicherio C, De Ribaupierre A (2014) Individual differences in cognitive
plasticity: an investigation of training curves in younger and older adults. Psychological
research.
Carretti B, Borella E, Zavagnin M, De Beni R (2013) Gains in language comprehension relating
to working memory training in healthy older adults. Int J Geriatr Psychiatry 28:539–546.
Dahlin E, Nyberg L, Backman L, Neely As (2008) Plasticity of executive functioning in young
and older adults: immediate training gains, transfer, and long-term maintenance. Psychol
Aging 23:720–730.
Heinzel S, Schulte S, Onken J, Duong Ql, Riemer Tg, Heinz A, Kathmann N, Rapp Ma
(2014) Working memory training improvements and gains in non-trained cognitive
tasks in young and older adults. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn
21:146–173.
Karbach J, Kray J (2009) How useful is executive control training? Age differences in near and
far transfer of task-switching training. Developmental Sci 12:978–990.
Richmond Ll, Morrison Ab, Chein Jm, Olson Ir (2011) Working memory training and transfer
in older adults. Psychol Aging 26:813–822.
Schmiedek F, Lovden M, Lindenberger U (2010) Hundred Days of Cognitive Training Enhance
Broad Cognitive Abilities in Adulthood: Findings from the COGITO Study. Frontiers in
Aging Neuroscience 2.
Stepankova H, Lukavsky J, Buschkuehl M, Kopecek M, Ripova D, Jaeggi Sm (2014) The mal-
leability of working memory and visuospatial skills: a randomized controlled study in
older adults. Dev Psychol 50:1049–1059.
Theill N, Schumacher V, Adelsberger R, Martin M, Jancke L (2013) Effects of simultaneously
performed cognitive and physical training in older adults. Bmc Neurosci 14:103.
Van Muijden J, Band Gp, Hommel B (2012) Online games training aging brains: limited trans-
fer to cognitive control functions. Front Hum Neurosci 6:221.
Von Bastian Cc, Langer N, Jancke L, Oberauer K (2013) Effects of working memory training
in young and old adults. Mem Cognit 41:611–624.
Wilkinson Aj, Yang L (2012) Plasticity of inhibition in older adults: retest practice and transfer
effects. Psychol Aging 27:606–615.
Zinke K, Zeintl M, Rose Ns, Putzmann J, Pydde A, Kliegel M (2014) Working memory training
and transfer in older adults: effects of age, baseline performance, and training gains. Dev
Psychol 50:304–315.
485
19
The Challenges of Disambiguating

Preclinical Alzheimer’s Disease
from Cognitive Aging
Reisa Sperling
G radual decline in cognitive function, particularly in episodic mem-

ory performance, is commonly associated with advanced aging and
often raises concern about incipient Alzheimer’s disease (AD). Despite remarkable
advances in the field of AD over the past decade, it remains extremely challenging
to differentiate the earliest cognitive changes on the AD trajectory from the cognitive
decline observed in “typical” aging. This chapter will review recent progress in the
quest to define the preclinical stages of AD, and highlight the questions that remain to
be elucidated with further research.
AD dementia is now estimated to affect more than one out of every nine individuals
over the age of 65, and is currently the only leading cause of death for which there is
no effective disease-modifying therapy. Converging evidence suggests that the patho-
physiologic process of AD likely begins well more than a decade prior to mild cog-
nitive impairment and dementia (see Figure 19.1). Recent advances in cerebrospinal
fluid (CSF) assays and molecular neuroimaging with positron emission tomography
(PET) have greatly facilitated our ability to detect evidence of the hallmark pathol-
ogies of AD—that is, aggregates of amyloid-beta (Aβ) in neuritic plaques and tau
in neurofibrillary tangles in vivo (see Figure 19.2). Convergent findings from large
autopsy cohorts, PET amyloid imaging, and cerebrospinal fluid (CSF) studies suggest
that 25%–35% of clinically normal (CN) older individuals over the age of 65 have
evidence of substantial Aβ plaque burden (Arriagada et al., 1992; Morris et al., 1996;
Bennett et al., 2006; Mintun et al., 2006; Jack et al., 2008; Gomperts et al., 2008;
Rowe et al., 2010; De Meyer et al., 2010; Fagan et al., 2009; Montine, 2011; Kantarci
485
486
The continuum of Alzheimer’s disease
“Normal” Aging
Preclinical
Cognition
Mild cognitive
impairment
Dementia
Years
Figure 19.1 The continuum of Alzheimer’s disease is postulated to begin with a “preclinical”

phase during which neuropathological changes are beginning in the brain but clinically evident
symptoms are not yet manifest, and then progresses to mild cognitive impairment (sometimes
referred to as prodromal AD) and dementia.
Figure 19.2 The neuropathology of Alzheimer’s disease. A microscopic section from an AD

patient autopsy demonstrating extracellular aggregates of amyloid-beta in a neuritic plaque and
intraneuronal aggregates of tau in neurofibrillary tangles. (See color plate also)
â•‡ 487
The Challenges of Disambiguating Preclinical Alzheimer’s Diseaseâ•… 487
et al., 2012; Hampel, 2013; Villemagne et al., 2013; van Harten et al., 2013; Jack et al.,
2014). Interestingly, this proportion of normal individuals with evidence of Aβ plaque
over age 65 is nearly identical to the population estimates for clinical AD dementia
above age 80, and is consonant with the hypothesis that Aβ accumulation may pre-
cede dementia by 15–â•‰20 years (Rowe et al., 2010). Recent studies in rare forms of
autosomal dominant AD that affect individuals in mid-â•‰life also demonstrate imaging
and CSF biomarker evidence of Aβ accumulation approximately 20 years prior to the
estimated age of dementia in these mutation carriers (Bateman et al., 2012; Reiman et
al., 2012; Benzinger et al., 2013; Fleisher et al., 2012). The challenge is that whereas
autosomal dominant forms of AD are virtually 100% penetrant, it remains unknown
whether the majority of clinically normal older individuals with evidence of Aβ accu-
mulation will in fact progress to develop AD dementia, and over what time frame.
The Amyloid Cascade Hypothesis
The amyloid cascade hypothesis (Hardy and Selkoe, 2002, Selkoe, 2011) posits that
Aβ accumulation is the key pathogenic event in AD and remains somewhat contro-
versial, despite compelling genetic evidence. All of the known early-onset, autosomal
dominant forms of AD are related to mutations in production of cleavage of the amy-
loid-â•‰precursor protein (APP, PSEN-â•‰1 and PSEN-â•‰2). Similarly, the APP coding region
is located on chromosome 21, and the vast majority of individuals with Trisomy-â•‰21
develop AD pathology and eventual AD dementia if they live beyond the age of 60.
Apolipoprotein E (APOE), the major genetic risk factor for late-onset AD, has long
been implicated in amyloid plaque pathology, likely related to impaired Aβ clear-
ance mechanisms. The recent identification of the first protective gene mutation for
late-onset AD provides some additional support of the amyloid hypothesis (Jonsson
et al., 2012). A coding mutation in the APP gene (A673T), near the beta-â•‰secretase
splice region, was recently associated with a significant reduction in the prevalence
of AD dementia in Scandanavia and found to be protective against “age-â•‰related” cog-
nitive decline. Interestingly, this alteration in APP was also associated in vitro with
an approximately 40% reduction in Aβ1–â•‰42 formation, and has reinvigorated work on
beta-â•‰secretase inhibitors as a therapeutic target for AD.
Recent PET amyloid imaging studies in autosomal dominant AD have demon-
strated Aβ deposition beginning up to 20 years prior to the average age of symptom
onset in these families (Benzinger et al., 2013; Fleisher et al., 2012). There have been
reports of early synaptic change, however, detectable prior to Aβ deposition, in autoso-
mal dominant mutation carriers (Reiman et al., 2012) and APOE e4 carriers (Filippini
et al., 2009) that might reflect toxicity of Aβ oligmeric forms or other synaptic, mito-
chondrial, metabolic, or neuronal cytoskeletal alterations in the pathogenesis of AD.
Whereas overproduction of amyloidogenic species of Aβ is thought to be the driv-
ing etiologic agent in autosomal dominant early-onset AD, aberrant clearance of Ae42
aggregates may be a key etiologic event in sporadic, late-â•‰onset AD (Mawuenyega et
al., 2010). It has been suggested, however, that sequestration of Aβ into fibrillar forms
might serve as a protective mechanism, as it may be smaller oligomeric forms of Aβ
that are particularly synaptotoxic (Lee et al., 2004; Shankar et al., 2008). Both autopsy
488
and biomarker studies (see below) suggest that Aβ42 accumulation increases with
advanced aging and is associated with other “AD-â•‰like” markers of neurodegeneration.
Although recent laboratory data are supportive of this hypothesis (Choi et al., 2014), it
remains to be proven that Aβ accumulation is sufficient in humans to incite the down-
stream pathological cascade of AD and lead to cognitive impairment and dementia.
Toward Defining Preclinical Alzheimer’s Disease
The National Institute on Aging and the Alzheimer’s Association (NIA-â•‰AA) convened
a workgroup in 2010 to develop guidelines for research on the earliest “preclinical”
stage of AD (Sperling et al., 2011a). The NIA-â•‰AA workgroup proposed a hypothetical
staging framework, based primarily on biomarker status, utilizing CSF and imaging
markers of Aβ accumulation (Aβ) and markers of neurodegeneration (ND), including
FDG temporo-â•‰parietal hypometabolism, volumetric MRI evidence of hippocampal
atrophy and/â•‰or cortical thinning, and/â•‰or CSF tau. The NIA-â•‰AA Staging framework
initially proposed 3 stages: Stage 1 = Aβ+/â•‰ND-â•‰, Stage 2 = Aβ+/â•‰ND+, and Stage
3 = Aβ+/â•‰ND+ plus evidence of subtle cognitive and behavioral decline that is not yet
sufficient to meet criteria for mild cognitive impairment (See Figure 19.3).
Amyloid and Tau PET imaging

PiB Aβ T807 Tau
Clinically
Normal
Aβ–neg
Clinically
Normal
Aβ–pos
AD
Dementia
Aβ–pos
Figure 19.3â•‡ PET Amyloid and Tau imaging. Coronal PET images superimposed on structural
magnetic resonance) of PiB Aβ (left) and T807 Tau (right) acquired on 3 participants in the
Harvard Aging Brain Study. The top row is a clinically normal older individual with low PiB
retention and minimal T807 binding in the medial temporal lobe (MTL). The middle row shows
a clinically normal older individual with elevated PiB retention and T807 binding extending
beyond the MTL into inferior temporal neocortex. The bottom row shows images from an AD
dementia patient with extensive PiB and T807 binding in the neocortex. Images courtesy of Dr.
Keith Johnson. (See color plate also)
489
The Challenges of Disambiguating Preclinical Alzheimer’s Disease 489
In the past few years since the NIA-AA Preclinical Workgroup staging schema
was published, several additions to the proposed classifications have been suggested,
including a “Stage 0” group, individuals who have no evidence of abnormal biomark-
ers to serve as a reference group (Jack et al., 2012, see Figure 19.4). Perhaps the most
important concept to be considered is “suspected non-alzheimer pathology” (SNAP)
suggested by the Mayo Clinic (Jack et al., 2012), which is comprised of older individ-
uals with biomarker evidence of neurodegeneration (ND) but without clear biomarker
evidence of Aβ accumulation. Approximately 40%–50% of clinically normal older
normals are classified as Stage 0, 10%–15% as Stage 1, 15% as Stage 2 (Aβ+/ND+),
and approximately 25% are classified as SNAP (Aβ-/ND+) (Jack et al., 2012; Vos et
al., 2013; Wirth et al., 2013b; Mormino et al., 2014a).
Stage 3 of preclinical AD remains to be fully delineated. We refer to individuals
with evidence of preclinical AD as “clinically normal” to connote that these indi-
viduals in general are not seeking medical care for cognitive concerns, and that they
Figure 19.4 Staging framework for preclinical AD (adapted from Sperling et al. 2011a with
updates from Jack et al., 2012). Stage 0 represents individuals without biomarker abnormalities
who are not thought to be on the AD trajectory. SNAP or suspected non-alzheimer pathology
has evidence of neurodegeneration without apparent amyloidosis. Stage 1 begins with cerebral
amyloidosis; Stage 2 is amyloidosis plus markers of neurodegeneration; Stage 3 is amyloidosis +
neurodegeneration + evidence of subtle cognitive and behavioral decline that is not yet suffi-
cient to meet criteria for mild cognitive impairment or dementia due to AD.
490
generally test in the “normal range” on neuropsychological test scores. However,

there is increasing evidence of subtle cognitive and behavioral change associated with
Aβ and ND markers that occurs several years preceding a diagnosis of MCI, that
is detectable on longitudinal testing and/â•‰or with subjective report (Amariglio et al.
2015). Thus, Stage 3 likely represents a transitional stage between the asymptomatic
and symptomatic stages of AD, and may be a particularly important group for second-
ary prevention trials.
Cognitive Changes Associated with Amyloid Accumulation
One of the most compelling remaining questions is whether there are clear cogni-
tive correlates of Aβ accumulation among clinically normal older individuals. The
early literature with cross-â•‰sectional Aβ PET data was quite mixed, with some studies
reporting subtle associations between amyloid burden and neuropsychological tests
(Rentz et al., 2010) and other studies reporting no difference in cognitive performance
between older individuals with and without amyloid, particularly when using stan-
dard neuropsychological tests used to detect evidence of dementia (Aizenstein et al.,
2008). Perhaps this should not be surprising as, by definition, selecting a cognitively
normal population substantially truncates the variance in test performance. As the
field has matured with large sample sizes, there have been increasing reports of sig-
nificant amyloid-â•‰associated deficits in cognitive performance, even among the limited
range observed in clinically normal older individuals (Lim et al., 2012 [Epub ahead
of print]; Sperling et al., 2013; Li et al., 2014) but the cross-â•‰sectional effects remain
relatively subtle. The most convincing findings, however, stem from longitudinal
Aβ associated decline among normals (Resnick et al., 2010; Storandt et al., 2009;
Lim et al., 2012; Doraiswamy et al., 2012; Kawas et al., 2013; Lim et al., 2014b;
Mormino et al., 2014b; Doraiswamy et al., 2014; Mormino et al., 2014a; Donohue et
al., 2014). The majority of these studies have indicated that episodic memory shows
the greatest vulnerability to Aβ-â•‰associated cognitive decline, but other domains of
cognition including working memory and executive function have also demonstrated
Aβ-â•‰associated change (Rodrigue et al., 2012; Storandt et al., 2009). The data thus far
suggest that normal older individuals with markers of Aβ in combination with mark-
ers of ND drive the majority of this decline (see below). In addition, several reports
have suggested that Aβ-â•‰related cognitive deficits may be accelerated in the presence
of one of more APOE e4 alleles (Kantarci et al., 2012; Mormino et al., 2014b) and
other genetic variants, including BDNF (Lim et al., 2013), suggesting that there may
be important interactions between Aβ accumulation and genetic background in risk
of cognitive decline. More sensitive cognitive and behavioral measures are currently
under development in many of these longitudinal cohorts. Recent work with chal-
lenging associative memory measures (Rentz et al., 2011) and iPAD computerized
testing (Rentz et al., 2013) may yield useful measures that will improve our ability to
track the earliest clinical changes that appear to begin years prior to the stage of mild
cognitive impairment.
Also of particular interest are several reports regarding the association of Aβ accu-
mulation with subjective cognitive concerns. Two recent papers have reported that
â•‡ 491
greater PET amyloid burden was related to increased report of subjective memory
concerns, even when controlling for concurrent symptoms of depression (Amariglio
et al., 2012; Perrotin et al., 2012). Another report from a clinic population of sub-
jective memory complaint patients revealed increased cognitive decline in Aβ+ with
subjective concerns compared to Aβ-â•‰with similar levels of subjective cognitive con-
cerns (van Harten et al., 2013). Most recently, a large autopsy study found that greater
subjective memory complaints among individuals who died prior to diagnosis of cog-
nitive impairment were associated with greater neuritic plaque burden (Kryscio et al.,
2014).
Impact of Reserve
The observation that amount of AD pathology does not always align well with the
degree of clinical impairment led to the concept of “reserve,” which might be con-
ceptualized as the ability to tolerate given levels of brain injury without exhibiting
clinical symptoms. The concept of reserve has continued to evolve to include both
“brain reserve” and “cognitive reserve” (Stern, 2009). Brain reserve generally refers
to the capacity of the brain to withstand pathologic insult, perhaps due to greater
synaptic density or a larger number of neurons, such that sufficient healthy neural
substrate can continue to support normal function. Cognitive reserve represents the
ability to engage alternate brain networks or cognitive strategies to withstand the
effects of encroaching pathology. It is not clear, however, that the data support a sharp
demarcation between these two constructs, as many factors, such as higher socio-â•‰
economic status or engagement in cognitively stimulating activities, may contribute
to both forms of reserve, and have been associated with lower age-â•‰adjusted incidence
of clinical AD diagnosis (Wilson et al., 2007). Recent studies suggest that increased
reserve may influence the capability of individuals to tolerate their AD pathology
for longer periods of time, but may be associated with rapid decline once a “tipping
point” is reached and compensatory mechanisms begin to fail (Fotenos et al., 2008;
Wilson et al.). The relationship between cognition and Aβ burden is stronger in lower
reserve individuals (Rentz et al., 2010; Roe et al., 2011), suggesting that high reserve
individuals may be able to maintain high levels of cognitive abilities despite underly-
ing pathology. Furthermore, reserve has been shown to be independently associated
with cognitive decline, above and beyond what is explained by AD markers such
as amyloid, hippocampus volume and glucose metabolism in AD vulnerable regions
(Vemuri et al., 2012), as well as common age-â•‰related pathologies as measured during
post mortem examination (Wilson et al., 2013). Thus, reserve is an important deter-
minant of cognitive abilities in aging, and may interact with Aβ to infer additional
risk of decline. Greater lifeâ•‰time participation in cognitively stimulating activities has
also been associated with reduced levels of Aβ in one study (Landau et al., 2012),
consistent with studies reporting that an enriching environment is associated with
reduced Aβ in mice (Costa et al., 2007), and/â•‰or protection against Aβ toxicity (Li
et al., 2013). Recent neuropathologic studies have also highlighted specific factors
that may confer resilience at the synaptic and neuronal level (Perez-â•‰Nievas et al.,
2013). Thus, higher reserve may exert protective effects directly through influencing
492
Aβ accumulation, and also by through synaptic and neuronal capacity to influence the
interaction between Aβ and downstream neurodegenerative processes.
The Relationships Among Amyloid, Neurodegeneration,

and Cognitive Decline
Although an increasing number of longitudinal studies suggest that clinically normal

older individuals with elevated Aβ burden are at increased risk for cognitive decline
(Lim et al., 2014b; Lim et al., 2014c; Knopman et al., 2012; Vos et al., 2013; Mormino
et al., 2014a), it is clear that Aβ is only one piece of the puzzle. Convergent reports
suggest that Aβ is necessary, but not sufficient in isolation to determine the clini-
cal progression toward MCI and AD dementia. Many recent studies have suggested
that the combination of Aβ and ND markers may be particularly valuable to predict
imminent decline along the AD trajectory, but the relationship between these markers
remains to be fully elucidated.
The reports of cross-sectional association between Aβ and ND markers have been
particularly variable, and may be influenced by methodological issues. Several stud-
ies have reported that Aβ+ normals are much more likely to demonstrate evidence of
ND than Aβ-(Becker et al., 2011; Dickerson et al., 2008; Rowe et al., 2010; Storandt
et al., 2009; Mormino et al., 2014a). However, other studies have not found a cross-
sectional relationship between Aβ and ND (Chetelat et al., 2010; Wirth et al., 2013a).
These inconsistencies may be related to cohort size and characteristics, as well as the
specific metric of ND. Also, commonly employed markers of ND are not necessarily
consistent with one another, and the associations are dependent on the specific mark-
ers and the thresholds employed for determining evidence of ND (Toledo et al., 2014;
Whitwell et al., 2013). Markers of ND, such as hippocampal volume, likely repre-
sent multiple influences, including developmental and reserve factors, comorbidities,
and other age-related processes that are unrelated to neurodegenerative disease. Both
autopsy studies (Nelson et al., 2012) and preliminary Tau PET imaging studies sug-
gest that majority of individuals over the age of 60 have tau accumulation in neuro-
fibrillary tangles in the medial temporal lobe. Thus it is possible that the mechanisms
resulting in amyloidosis and various ND markers arise independently, and merely
conspire to result in greater vulnerability to cognitive decline. However, even if Aβ
and ND may begin via separate pathways, it is possible that Aβ then exacerbates
underlying ND. Longitudinal studies relating baseline levels of Aβ to change in ND
biomarker over time are supportive of this hypothesis (Chetelat et al., 2012; Dore et
al., 2013; Nosheny et al., 2014; Schott et al., 2010; Storandt et al., 2009; Knopman et
al., 2013).
Longitudinal studies do suggest that normal individuals with markers of CN both
Aβ and ND (Stage 2 of the NIA-AA criteria) are more likely to demonstrate cognitive
decline over time (Mormino et al., 2014a; Wirth et al., 2013b) and to progress to MCI
or dementia (Fagan et al., 2007; Desikan et al., 2012; Knopman et al., 2012; Rowe
et al., 2013; Vos et al., 2013; Wirth et al., 2013b; Mormino et al., 2014a). Individuals
classified as SNAP (Aβ-/ND+) also show a greater rate of decline than normals with-
out evidence of any biomarker abnormality (Stage 0 Aβ-/ND-) but significantly less
â•‡ 493
decline than Stage 2/â•‰3 (Aβ+/â•‰ND+) (Vos et al., 2013; Mormino et al., 2014a), again
suggesting that both amyloid and neurodegeneration are required for cognitive decline
along the AD trajectory.
One promising recent development is Tau PET imaging (see Color Figure 19.2).
Although tau and neurofibrillary tangles are not specific to AD, there is a characteris-
tic anatomic spread in AD proposed by Braak and colleagues (Braak et al., 2006) that
may be detectable with PET imaging. The earliest of Braak stages (Stages I–â•‰II) in the
medial temporal lobe, particularly transentorhinal cortex occurs in ~80% of all indi-
viduals by age 60 (Nelson et al., 2012). Extension of tangle pathology to limbic and
neighboring temporal neocortices (Braak Stage III–â•‰IV and higher) is thought to mark
the transition from asymptomatic to symptomatic pathology (Price and Morris, 1999;
Nelson et al., 2012; Hyman et al., 2012). Until very recently, this critical spread and
intensification of Tau pathology has been invisible to all but the neuropathologist, but
remarkable advances in PET imaging now allow us to image Tau pathology in vivo
(Maruyama et al., 2013; Zhang et al., 2012; Chien et al., 2013; Johnson et al., 2015).
The selectivity of at least one of the compounds (18F-â•‰T807/â•‰AV-â•‰1451) for Tau over Aβ
is estimated to be approximately 27-fold by autoradiographic comparison of human
cortical brain sections (Xia et al., 2013), but additional validation work on these novel
tracers is ongoing. The early Tau PET reports confirm previous autopsy reports that
medial temporal lobe Tau accumulation is extremely common after age 60 (Nelson et
al., 2012), but it remains unknown whether this pathology contributes to age-â•‰related
memory change, at any level of Aβ. If tau accumulation in the entorhinal cortex is
found to be associated with worse memory performance, particularly in the absence of
supra-â•‰threshold levels of Aβ accumulation—â•‰should this be considered part of normal
aging or one of the earliest “hits” in the AD pathophysiological process? Indeed, the
early Tau PET data suggest that distinction between the process of brain aging and
at least one of the hallmark pathologies of AD may be particularly difficult to disam-
biguate. Consistent with Braak staging, our preliminary Tau PET experience in MCI
and dementia patients who have extensive Aβ deposition, demonstrates Tau binding
in multiple regions of neocortex, particularly in the inferior temporal cortices, lateral
and medial parietal cortices (Johnson et al., 2015; see Color Figure 19.3). Based on
our early preliminary work with Tau PET and previous evidence from animal studies
and cross-â•‰sectional autopsy studies, we postulate that Aβ may accelerate the spread of
Tau both within and beyond the MTL, disrupting function and initiating neurodegen-
eration in distributed brain networks, resulting in cognitive decline.
Memory Activity and Network Dysfunction in Preclinical AD
One of the intriguing early findings with PET amyloid imaging was the observation
that Aβ preferentially accumulates in brain regions overlapping the topography of the
default mode network (Buckner et al., 2005), although subsequent work noted that
Aβ deposition occurs in multiple regions of high connectivity, designated as “cortical
hubs”(Buckner et al., 2009). A number of groups have now investigated the rela-
tionship of Aβ accumulation to disruption of functional networks in clinically nor-
mal older individuals, using both task-â•‰related and resting state (task-â•‰free) functional
494
MRI. The majority of these studies have reported decreases in default network activity
associated with Aβ deposition (Hedden et al., 2009; Sheline et al., 2009; Brier et al.,
2014b; Wang et al., 2013; Brier et al., 2014a). A similar pattern of default network
disruption has also been reported in carriers of autosomal dominant genetic mutations
(Chhatwal et al., 2012; Thomas et al., 2014). Interestingly, there have also been reports
of at least transiently related increased connectivity in the default network associated
with Aβ in clinically normal older individuals (Jack et al., 2013; Lim et al., 2014a),
in prefrontal regions (Mormino et al., 2011) and in the salience network of APOE e4
carriers (Machulda et al., 2011). Resting-state functional connectivity MRI is already
being implemented in several large-scale, secondary prevention AD trials (see below),
as these relatively short MR sequences can be acquired on most scanners during safety
MRI sessions. It should be noted, however, that additional work is needed to optimize
these methodologies, as all functional MRI techniques are still fraught with sources
of variability, such as head movement, physiological noise, and analytic approaches,
that may limit the utility of these techniques in prevention trials.
Using a variety of paradigms with task fMRI, clinically normal older individuals
with PET evidence of Aβ accumulation have demonstrated impaired ability to modu-
late activity in the default mode network (Sperling et al., 2009; Kennedy et al., 2012;
Rami et al., 2012), similar to earlier reports in AD dementia (Lustig et al., 2003). In
particular, the failure to deactivate the posterior cingulate/precuneus regions during
episodic memory encoding tasks has been associated with elevated Aβ accumulation
in these regions. Interestingly, a similar pattern of disrupted posterior cingulate activ-
ity during memory encoding has been reported in presymptomatic presenilin 1 carri-
ers (Reiman et al., 2012) and APOE e4 carriers (Pihlajamaki et al., 2010; Fleisher et
al., 2009), suggesting that deactivation of the posteromedial cortices may be an early
indicator of synaptic dysfunction in preclinical AD.
Markers of Aβ accumulation have been less consistently associated with alterations
in MTL function in preclinical AD. Previous cross-modality imaging work suggested
that alterations in hippocampal function, manifesting as hippocampal hyperactivity,
became evident at the point of the earliest clinical symptoms, perhaps at the stage of
early mild cognitive impairment (MCI) (Celone et al., 2006; Sperling et al., 2009;
O’Brien et al., 2010). Interestingly, however, hippocampal hyperactivity has been
reported in asymptomatic carriers of autosomal dominant mutations (Quiroz et al.,
2010, Reiman et al., 2012) and younger APOE e4 carriers (Filippini et al., 2009) sug-
gesting that hippocampal hyperactivity might occur earlier in the setting of virulent
Aβ pathology. More recently, we reported evidence of Aβ associated disruptions in
entorhinal activity (Huijbers et al., 2014), specifically failure to deactivate the entorhi-
nal cortices, similar to the reports of altered function in the neocortical regions of the
default network. Of note, the early Aβ associated disruptions in task-related activity,
whether in default network regions that typically deactivate during memory encoding
or in “task-positive” regions that typically activate during memory processes, tend to
manifest as relative increases in activity (Sperling et al., 2010). This finding is consist-
ent with the hypothesis that Aβ might accelerate excitotoxicity through excessive or
aberrant neuronal firing (Palop and Mucke, 2010). It is also possible that the increased
activity may represent evidence of attempted compensation, at least early in the patho-
physiological process (Elman et al., 2014). Unfortunately, hippocampal hyperactivity
â•‡ 495
may also be an indicator of network degeneration (Putcha et al., 2011) and subsequent
cognitive decline (O’Brien et al., 2010). It is also possible that this is a perpetuating
cycle of increasing neural activity that in turn increases Aβ production that further
drives hyperactivity (Cirrito et al., 2005; Jagust and Mormino, 2011). It is of particular
interest that the two hallmark pathologies of AD seem to begin to accumulate at the
opposite ends of the distributed brain networks supporting memory function, with tau
accumulation beginning in transentorhinal cortex and related structures in the MTL
that are functionally connected to the heteromodal cortical regions of default network
particularly vulnerable to Aβ deposition (Celone et al., 2006). Future multimodality
imaging studies should provide insights into influence of network topography on the
spread of molecular pathology, synaptic alterations, and neuronal loss in early AD.
Caveats
The delineation of a preclinical stage of AD has understandably raised concerns about
inappropriately labeling of individuals with “Alzheimer’s disease” who might never
develop dementia. Although convergent studies suggest that older individuals with
evidence of preclinical AD, particularly those with Aβ and ND markers, have a statis-
tically greater risk of manifesting subsequent cognitive and progression to the symp-
tomatic stages of AD at a group level, there are insufficient data at this time to make
accurate predictions at an individual level. The NIA-â•‰AA criteria emphasize the fact
that some individuals with biomarker evidence of preclinical AD will not progress to
develop AD dementia during their lifetime. However, the NIA-â•‰AA workgroup felt it
was important to put forth the concept that there is strong evidence for a preclinical
phase of AD that is detectable prior to clinically evident symptoms, in particular, to
enable clinical trials aimed at preventing the clinical expression of AD.
Implications for Prevention Trial Design
The vast majority of potential disease-â•‰modifying interventions in AD have been

tested in patients with mild to moderate dementia, when there is already substan-
tial synaptic and neuronal damage. It is likely, as in the other fields of medicine
for which we have made significant inroads, such as cancer, cardiovascular disease,
stroke, HIV/â•‰AIDS, and diabetes; we would have a greater chance for success tar-
geting much earlier intervention in AD (Sperling et al., 2011b). Fortunately, recent
advances in molecular imaging and fluid markers have greatly facilitated our ability
to detect evidence of AD pathology in vivo, and have led to the initiation of several
secondary prevention trials. Secondary prevention in this context refers to interven-
tions at the point when the disease process has begun in the brain aimed at delaying
the onset of symptoms or progression to the clinical stages of AD. Several trials are
targeting presymptomatic individuals with autosomal dominant AD, including the
Dominantly Inherited Alzheimer Network (DIAN), an international consortium trial
in families with autosomal dominant mutations in PSEN1, PSEN2, and APP muta-
tions (Mills et al., 2013), and the Alzheimer’s Prevention Initiative (API) trial in the
496
large Colombian PSEN1 cohort (Reiman et al., 2011). These trials are testing mono-
clonal antibodies against various forms of Aβ. API also plans to begin another trial
in apolipoprotein E ε4 homozygotes next year. The TOMMORROW trial is testing
pioglitazone, aimed at glucose metabolism, in carriers of the high-risk variant of the
TOMM-40 gene.
A complementary approach to the trials in genetic risk cohorts is the Anti-Amyloid
Treatment in Asymptomatic Alzheimer’s disease (A4) Trial. The A4 Study is a large
Phase 3 secondary prevention trial that identifies older individuals at risk for cognitive
decline on the basis of biomarker evidence of Aβ accumulation (Sperling et al., 2014).
The A4 Study is enrolling clinically normal older individuals (ages 65–85) into a 3.3
year trial with a monoclonal anti-Aβ antibody versus placebo. The primary outcome
of the A4 trial is a cognitive composite (Donohue et al., 2014), but also incorporates a
number of participant reported outcomes and computerized testing on an IPAD using
CogState technology augmented with two episodic memory measures adapted for the
A4 trial (Rentz et al., 2013; Stark et al., 2013). Importantly, the A4 Study also includes
an observational comparison arm of older CN who “screen-fail” on the basis of fall-
ing below the threshold for Aβ positivity on screening PET amyloid imaging in the
Longitudinal Evaluation of Amyloid Risk and Neurodegeneration (LEARN) study.
Additional trials with other anti-amyloid agents in a similar biomarker-risk defined
older population are currently being planned. It is likely that we will ultimately
require combination therapies to fully bend the curve of clinical decline due to AD.
These combinations may include several mechanisms to lower Aβ burden, such as a
beta-secretase inhibitor to decrease the production of Aβ1–42 and an antibody against
aggregated forms of Aβ. Ideally, we will eventually combine anti-Aβ and anti-Tau
treatments to target multiple mechanisms simultaneously, as we do in many complex
medical illnesses.
Prevention trial efforts are also beginning in other neurodegenerative diseases;
in particular, Huntington’s disease as risk can be reliably identified through genetic
mutation testing, and an estimated date of symptom onset predicted through the num-
ber of trinucleotide repeats. Both functional and structural abnormalities have been
identified in preclinical stages of the Huntington’s disease (Rosas et al., 2006; Wolf
et al., 2012; Wolf et al., 2013). Interestingly, although the participants were identified
as pre-Huntington’s, these studies found relationship between the imaging and cogni-
tive measures (Wolf et al., 2013), supporting the hypothesis that there is a continuum
of very subtle symptoms that may emerge in the years prior to clinical recognition in
Huntington’s disease similar to the concept of “Stage 3” in the preclinical AD criteria
(Sperling et al., 2011a). As Parkinson’s disease does not manifest clinically until more
than 90% of the substantia nigra neurons are lost, Parkinson’s also offers a tremendous
opportunity for secondary prevention trials. Ongoing biomarkers efforts in Parkinson’s
disease are working to identify reliable markers that can be used to select asympto-
matic individuals at risk for motor progression and to track response to therapies in
the preclinical stages of Parkinson’s disease. As multiple pathologies are common in
late-life cognitive decline and dementia, and it is likely that there may be common
mechanisms underlying protein misfolding in these neurodegenerative diseases, it is
likely that this research will also benefit the AD field, and hopefully one day serve to
prevent cognitive decline related to multiple age-related neuropathological processes.
â•‡ 497
References
Aizenstein, H. J., Nebes, R. D., Saxton, J. A., Price, J. C., Mathis, C. A., Tsopelas, N. D.,
Ziolko, S. K., James, J. A., Snitz, B. E., Houck, P. R., Bi, W., Cohen, A. D., Lopresti, B.
J., Dekosky, S. T., Halligan, E. M. & Klunk, W. E. 2008. Frequent amyloid deposition
without significant cognitive impairment among the elderly. Arch Neurol, 65, 1509–â•‰1517.
Amariglio, R. E., Becker, J. A., Carmasin, J., Wadsworth, L. P., Lorius, N., Sullivan, C., Maye,
J. E., Gidicsin, C., Pepin, L. C., Sperling, R. A., Johnson, K. A. & Rentz, D. M. 2012.
Subjective cognitive complaints and amyloid burden in cognitively normal older individu-
als. Neuropsychologia, 50, 2880–â•‰2886.
Amariglio, R., Mormino, E., Pietras, A. C., Marshall, G. A., Vannini, P, Johnson, K. A., Sperling,
R. A., Rentz, D. M. Subjective cognitive concerns, amyloid-â•‰β, and neurodegeneration in
clinically normal elderly. Neurology, 2015;85(1):56–â•‰62.
Arriagada, P. V., Marzloff, K. & Hyman, B. T. 1992. Distribution of Alzheimer-â•‰type pathologic
changes in nondemented elderly individuals matches the pattern in Alzheimer’s disease.
Neurology, 42, 1681–â•‰1688.
Bateman, R. J., Xiong, C., Benzinger, T. L., Fagan, A. M., Goate, A., Fox, N. C., Marcus, D.
S., Cairns, N. J., Xie, X., Blazey, T. M., Holtzman, D. M., Santacruz, A., Buckles, V.,
Oliver, A., Moulder, K., Aisen, P. S., Ghetti, B., Klunk, W. E., Mcdade, E., Martins, R.
N., Masters, C. L., Mayeux, R., Ringman, J. M., Rossor, M. N., Schofield, P. R., Sperling,
R. A., Salloway, S. & Morris, J. C. 2012. Clinical and Biomarker Changes in Dominantly
Inherited Alzheimer’s Disease. N Engl J Med, 367, 795–â•‰804.
Becker, J. A., Hedden, T., Carmasin, J., Maye, J., Rentz, D. M., Putcha, D., Fischl, B., Greve, D.
N., Marshall, G. A., Salloway, S., Marks, D., Buckner, R. L., Sperling, R. A. & Johnson,
K. A. 2011. Amyloid-â•‰beta associated cortical thinning in clinically normal elderly. Ann
Neurol, 69, 1032–â•‰1042.
Bennett, D., Schneider, J., Arvanitakis, Z., Kelly, J., Aggarwal, N., Shah, R. & Wilson, R. 2006.
Neuropathology of older persons without cognitive impairment from two community-â•‰
based studies Neurology, 66, 1837–â•‰1844.
Benzinger, T. L., Blazey, T., Jack, C. R., Jr., Koeppe, R. A., Su, Y., Xiong, C., Raichle, M.
E., Snyder, A. Z., Ances, B. M., Bateman, R. J., Cairns, N. J., Fagan, A. M., Goate, A.,
Marcus, D. S., Aisen, P. S., Christensen, J. J., Ercole, L., Hornbeck, R. C., Farrar, A.
M., Aldea, P., Jasielec, M. S., Owen, C. J., Xie, X., Mayeux, R., Brickman, A., Mcdade,
E., Klunk, W., Mathis, C. A., Ringman, J., Thompson, P. M., Ghetti, B., Saykin, A. J.,
Sperling, R. A., Johnson, K. A., Salloway, S., Correia, S., Schofield, P. R., Masters, C. L.,
Rowe, C., Villemagne, V. L., Martins, R., Ourselin, S., Rossor, M. N., Fox, N. C., Cash, D.
M., Weiner, M. W., Holtzman, D. M., Buckles, V. D., Moulder, K. & Morris, J. C. 2013.
Regional variability of imaging biomarkers in autosomal dominant Alzheimer’s disease.
Proc Natl Acad Sci U S A, 110, E4502–â•‰E4509.
Braak, H., Alafuzoff, I., Arzberger, T., Kretzschmar, H. & Del Tredici, K. 2006. Staging of
Alzheimer disease-â•‰associated neurofibrillary pathology using paraffin sections and immu-
nocytochemistry. Acta Neuropathol, 112, 389–â•‰404.
Brier, M. R., Thomas, J. B., Fagan, A. M., Hassenstab, J., Holtzman, D. M., Benzinger, T. L.,
Morris, J. C. & Ances, B. M. 2014a. Functional connectivity and graph theory in preclini-
cal Alzheimer’s disease. Neurobiol Aging, 35, 757–â•‰768.
Brier, M. R., Thomas, J. B., Snyder, A. Z., Wang, L., Fagan, A. M., Benzinger, T., Morris, J. C. &
Ances, B. M. 2014b. Unrecognized preclinical Alzheimer disease confounds rs-â•‰fcMRI
studies of normal aging. Neurology.
Buckner, R. L., Sepulcre, J., Talukdar, T., Krienen, F. M., Liu, H., Hedden, T., Andrews-â•‰Hanna,
J. R., Sperling, R. A. & Johnson, K. A. 2009. Cortical hubs revealed by intrinsic functional
498
connectivity: mapping, assessment of stability, and relation to Alzheimer’s disease. J

Neurosci, 29, 1860–1873.
Buckner, R. L., Snyder, A. Z., Shannon, B. J., Larossa, G., Sachs, R., Fotenos, A. F., Sheline,
Y. I., Klunk, W. E., Mathis, C. A., Morris, J. C. & Mintun, M. A. 2005. Molecular, struc-
tural, and functional characterization of Alzheimer’s disease: evidence for a relationship
between default activity, amyloid, and memory. J Neurosci, 25, 7709–7717.
Celone, K. A., Calhoun, V. D., Dickerson, B. C., Atri, A., Chua, E. F., Miller, S. L., Depeau, K.,
Rentz, D. M., Selkoe, D. J., Blacker, D., Albert, M. S. & Sperling, R. A. 2006. Alterations
in memory networks in mild cognitive impairment and Alzheimer’s disease: an indepen-
dent component analysis. J Neurosci, 26, 10222–10231.
Chetelat, G., Villemagne, V. L., Pike, K. E., Baron, J. C., Bourgeat, P., Jones, G., Faux, N. G.,
Ellis, K. A., Salvado, O., Szoeke, C., Martins, R. N., Ames, D., Masters, C. L. &
Rowe, C. C. 2010. Larger temporal volume in elderly with high versus low beta-amyloid
deposition. Brain, 133, 3349–3358.
Chetelat, G., Villemagne, V. L., Villain, N., Jones, G., Ellis, K. A., Ames, D., Martins, R. N.,
Masters, C. L. & Rowe, C. C. 2012. Accelerated cortical atrophy in cognitively normal
elderly with high beta-amyloid deposition. Neurology, 78, 477–484.
Chhatwal, J., Schultz, A., Johnson, K. A., Benzinger, T., Jack, C., Salloway, S., Ringman, J., Koeppe,
R., Marcus, D., Thompson, P., Saykin, A., Correia, S., Schofield, P., Rowe, C. C., Fox, N. C.,
Brickman, A. M., Mayeux, R., Mathis, C. A., Weiner, M., Fagan, A. M., Goate, A., Cairns,
N., Xiong, C., Masters, C., Ghetti, B. F., Moulder, K., Martins, R., Bateman, R., Buckles, V.,
Morris, J. C. & Sperling, R. A. Impaired default network functional connectivity in autoso-
mal dominant Alzheimer’s disease: Findings from the DIAN study. Alzheimer’s Association
International Conference, July 14–19 2012 Vancouver, British Columbia, Canada.
Chien, D. T., Bahri, S., Szardenings, A. K., Walsh, J. C., Mu, F., Su, M. Y., Shankle, W. R.,
Elizarov, A. & Kolb, H. C. 2013. Early clinical PET imaging results with the novel PHF-
tau radioligand [F-18]-T807. J Alzheimers Dis, 34, 457–468.
Choi, S. H., Kim, Y. H., Hebisch, M., Sliwinski, C., Lee, S., D’avanzo, C., Chen, H., Hooli, B.,
Asselin, C., Muffat, J., Klee, J. B., Zhang, C., Wainger, B. J., Peitz, M., Kovacs, D. M.,
Woolf, C. J., Wagner, S. L., Tanzi, R. E. & Kim, D. Y. 2014. A three-dimensional human
neural cell culture model of Alzheimer’s disease. Nature.
Cirrito, J. R., Yamada, K. A., Finn, M. B., Sloviter, R. S., Bales, K. R., May, P. C., Schoepp,
D. D., Paul, S. M., Mennerick, S. & Holtzman, D. M. 2005. Synaptic activity regulates
interstitial fluid amyloid-beta levels in vivo. Neuron, 48, 913–922.
Costa, D. A., Cracchiolo, J. R., Bachstetter, A. D., Hughes, T. F., Bales, K. R., Paul, S. M.,
Mervis, R. F., Arendash, G. W. & Potter, H. 2007. Enrichment improves cognition in AD
mice by amyloid-related and unrelated mechanisms. Neurobiol Aging, 28, 831–844.
De Meyer, G., Shapiro, F., Vanderstichele, H., Vanmechelen, E., Engelborghs, S., De Deyn,
P. P., Coart, E., Hansson, O., Minthon, L., Zetterberg, H., Blennow, K., Shaw, L. &
Trojanowski, J. Q. 2010. Diagnosis-independent Alzheimer disease biomarker signature
in cognitively normal elderly people. Arch Neurol, 67, 949–956.
Desikan, R. S., Mcevoy, L. K., Thompson, W. K., Holland, D., Brewer, J. B., Aisen, P. S.,
Sperling, R. A. & Dale, A. M. 2012. Amyloid-beta-Associated Clinical Decline Occurs
Only in the Presence of Elevated P-tau. Arch Neurol, 69, 709–713.
Dickerson, B. C., Bakkour, A., Salat, D. H., Feczko, E., Pacheco, J., Greve, D. N., Grodstein,
F., Wright, C. I., Blacker, D., Rosas, H. D., Sperling, R. A., Atri, A., Growdon, J. H.,
Hyman, B. T., Morris, J. C., Fischl, B. & Buckner, R. L. 2008. The Cortical Signature of
Alzheimer’s Disease: Regionally Specific Cortical Thinning Relates to Symptom Severity
in Very Mild to Mild AD Dementia and is Detectable in Asymptomatic Amyloid-Positive
Individuals. Cereb Cortex.
499
Donohue, M. C., Sperling, R. A., Salmon, D. P., Rentz, D. M., Raman, R., Thomas, R. G.,
Weiner, M., Aisen, P. S., For The Australian Imaging, B., Lifestyle Flagship Study Of, A., The
Alzheimer’s Disease Neuroimaging, I. & The Alzheimer’s Disease Cooperative, S. 2014. The
Preclinical Alzheimer Cognitive Composite: Measuring Amyloid-Related Decline. JAMA
Neurol.
Doraiswamy, P. M., Sperling, R. A., Coleman, R. E., Johnson, K. A., Reiman, E. M., Davis,
M. D., Grundman, M., Sabbagh, M. N., Sadowsky, C. H., Fleisher, A. S., Carpenter, A.,
Clark, C. M., Joshi, A. D., Mintun, M. A., Skovronsky, D. M. & Pontecorvo, M. J. 2012.
Amyloid-beta assessed by florbetapir F 18 PET and 18-month cognitive decline: A multi-
center study. Neurology.
Doraiswamy, P. M., Sperling, R. A., Johnson, K., Reiman, E. M., Wong, T. Z., Sabbagh, M. N.,
Sadowsky, C. H., Fleisher, A. S., Carpenter, A., Joshi, A. D., Lu, M., Grundman, M.,
Mintun, M. A., Skovronsky, D. M., Pontecorvo, M. J. & Group, A. A. S. 2014. Florbetapir
F 18 amyloid PET and 36-month cognitive decline:a prospective multicenter study. Mol
Psychiatry.
Dore, V., Villemagne, V. L., Bourgeat, P., Fripp, J., Acosta, O., Chetelat, G., Zhou, L.,
Martins, R., Ellis, K. A., Masters, C. L., Ames, D., Salvado, O. & Rowe, C. C. 2013.
Cross-sectional and longitudinal analysis of the relationship between Abeta deposition,
cortical thickness, and memory in cognitively unimpaired individuals and in Alzheimer
disease. JAMA Neurol, 70, 903–911.
Elman, J. A., Oh, H., Madison, C. M., Baker, S. L., Vogel, J. W., Marks, S. M., Crowley, S.,
O’neil, J. P. & Jagust, W. J. 2014. Neural compensation in older people with brain amyloid-
beta deposition. Nat Neurosci, 17, 1316–1318.
Fagan, A. M., Mintun, M. A., Shah, A. R., Aldea, P., Roe, C. M., Mach, R. H., Marcus, D.,
Morris, J. C. & Holtzman, D. M. 2009. Cerebrospinal fluid tau and ptau(181) increase
with cortical amyloid deposition in cognitively normal individuals: implications for future
clinical trials of Alzheimer’s disease. EMBO Mol Med, 1, 371–380.
Fagan, A. M., Roe, C. M., Xiong, C., Mintun, M. A., Morris, J. C. & Holtzman, D. M. 2007.
Cerebrospinal fluid tau/beta-amyloid(42) ratio as a prediction of cognitive decline in non-
demented older adults. Arch Neurol, 64, 343–349.
Filippini, N., Macintosh, B. J., Hough, M. G., Goodwin, G. M., Frisoni, G. B., Smith, S. M.,
Matthews, P. M., Beckmann, C. F. & Mackay, C. E. 2009. Distinct patterns of brain activity
in young carriers of the APOE-epsilon4 allele. Proc Natl Acad Sci U S A, 106, 7209–7214.
Fleisher, A. S., Chen, K., Quiroz, Y. T., Jakimovich, L. J., Gomez, M. G., Langois, C. M.,
Langbaum, J. B., Ayutyanont, N., Roontiva, A., Thiyyagura, P., Lee, W., Mo, H., Lopez, L.,
Moreno, S., Acosta-Baena, N., Giraldo, M., Garcia, G., Reiman, R. A., Huentelman,
M. J., Kosik, K. S., Tariot, P. N., Lopera, F. & Reiman, E. M. 2012. Florbetapir PET analy-
sis of amyloid-beta deposition in the presenilin 1 E280A autosomal dominant Alzheimer’s
disease kindred: a cross-sectional study. Lancet Neurol, 11, 1057–1065.
Fleisher, A. S., Sherzai, A., Taylor, C., Langbaum, J. B., Chen, K. & Buxton, R. B. 2009.
Resting-state BOLD networks versus task-associated functional MRI for distinguishing
Alzheimer’s disease risk groups. Neuroimage, 47, 1678–1690.
Fotenos, A. F., Mintun, M. A., Snyder, A. Z., Morris, J. C. & Buckner, R. L. 2008. Brain vol-
ume decline in aging: evidence for a relation between socioeconomic status, preclinical
Alzheimer disease, and reserve. Arch Neurol, 65, 113–120.
Gomperts, S. N., Rentz, D. M., Moran, E., Becker, J. A., Locascio, J. J., Klunk, W. E., Mathis, C.
A., Elmaleh, D. R., Shoup, T., Fischman, A. J., Hyman, B. T., Growdon, J. H. & Johnson,
K. A. 2008. Imaging amyloid deposition in Lewy body diseases. Neurology, 71, 903–910.
Hampel, H. 2013. Amyloid-beta and cognition in aging and Alzheimer’s disease: molecular and
neurophysiological mechanisms. J Alzheimers Dis, 33 Suppl 1, S79–S86.
500
Hardy, J. & Selkoe, D. J. 2002. The amyloid hypothesis of Alzheimer’s disease: progress and
problems on the road to therapeutics. Science, 297, 353–356.
Hedden, T., Van Dijk, K. R., Becker, J. A., Mehta, A., Sperling, R. A., Johnson, K. A. &
Buckner, R. L. 2009. Disruption of functional connectivity in clinically normal older
adults harboring amyloid burden. J Neurosci, 29, 12686–12694.
Huijbers, W., Mormino, E. C., Wigman, S. E., Ward, A. M., Vannini, P., Mclaren, D. G.,
Becker, J. A., Schultz, A. P., Hedden, T., Johnson, K. A. & Sperling, R. A. 2014. Amyloid
Deposition Is Linked to Aberrant Entorhinal Activity among Cognitively Normal Older
Adults. J Neurosci, 34, 5200–5210.
Hyman, B. T., Phelps, C. H., Beach, T. G., Bigio, E. H., Cairns, N. J., Carrillo, M. C.,
Dickson, D. W., Duyckaerts, C., Frosch, M. P., Masliah, E., Mirra, S. S., Nelson, P. T.,
Schneider, J. A., Thal, D. R., Thies, B., Trojanowski, J. Q., Vinters, H. V. & Montine, T. J.
2012. National Institute on Aging-Alzheimer’s Association guidelines for the neuropatho-
logic assessment of Alzheimer’s disease. Alzheimers Dement, 8, 1–13.
Jack, C. R., Jr., Knopman, D. S., Weigand, S. D., Wiste, H. J., Vemuri, P., Lowe, V., Kantarci, K.,
Gunter, J. L., Senjem, M. L., Ivnik, R. J., Roberts, R. O., Rocca, W. A., Boeve, B. F. &
Petersen, R. C. 2012. An operational approach to National Institute on Aging-Alzheimer’s
Association criteria for preclinical Alzheimer disease. Ann Neurol, 71, 765–775.
Jack, C. R., Jr., Lowe, V. J., Senjem, M. L., Weigand, S. D., Kemp, B. J., Shiung, M. M.,
Knopman, D. S., Boeve, B. F., Klunk, W. E., Mathis, C. A. & Petersen, R. C. 2008. 11C
PiB and structural MRI provide complementary information in imaging of Alzheimer’s
disease and amnestic mild cognitive impairment. Brain, 131, 665–680.
Jack, C. R., Jr., Wiste, H. J., Knopman, D. S., Vemuri, P., Mielke, M. M., Weigand, S. D.,
Senjem, M. L., Gunter, J. L., Lowe, V., Gregg, B. E., Pankratz, V. S. & Petersen, R. C.
2014. Rates of beta-amyloid accumulation are independent of hippocampal neurodegener-
ation. Neurology, 82, 1605–1612.
Jack, C. R., Jr., Wiste, H. J., Weigand, S. D., Knopman, D. S., Lowe, V., Vemuri, P., Mielke, M.
M., Jones, D. T., Senjem, M. L., Gunter, J. L., Gregg, B. E., Pankratz, V. S. & Petersen, R.
C. 2013. Amyloid-first and neurodegeneration-first profiles characterize incident amyloid
PET positivity. Neurology, 81, 1732–1740.
Jagust, W. J. & Mormino, E. C. 2011. Lifespan brain activity, beta-amyloid, and Alzheimer’s
disease. Trends Cogn Sci, 15, 520–526.
Johnson K.a, Schultz A., Betensky R.a., Becker J.a., Sepulcre, J., Rentz, D., Mormino, E.,
Chhatwal, J., Amariglio, R., Papp, K, Marshall, G., Albers, M., Mauro, S., Pepin, L.,
Alverio, J., Judge, K., Phillossaint, M, Shoup, T., Yokell, D., Dickerson, B., Gomez-Isla,
T., Hyman, B., Vasdev, N., Sperling, R. Tau PET imaging in aging and early Alzheimer’s
disease. Annals of Neurology, 2015 Oct 27. doi: 10.1002/ana.24546. [Epub ahead of print].
Jonsson, T., Atwal, J. K., Steinberg, S., Snaedal, J., Jonsson, P. V., Bjornsson, S., Stefansson,
H., Sulem, P., Gudbjartsson, D., Maloney, J., Hoyte, K., Gustafson, A., Liu, Y., Lu, Y.,
Bhangale, T., Graham, R. R., Huttenlocher, J., Bjornsdottir, G., Andreassen, O. A., Jonsson,
E. G., Palotie, A., Behrens, T. W., Magnusson, O. T., Kong, A., Thorsteinsdottir, U., Watts,
R. J. & Stefansson, K. 2012. A mutation in APP protects against Alzheimer’s disease and
age-related cognitive decline. Nature, 488, 96–99.
Kantarci, K., Lowe, V., Przybelski, S. A., Weigand, S. D., Senjem, M. L., Ivnik, R. J., Preboske,
G. M., Roberts, R., Geda, Y. E., Boeve, B. F., Knopman, D. S., Petersen, R. C. & Jack, C.
R. 2012. APOE modifies the association between Abeta load and cognition in cognitively
normal older adults. Neurology, 78, 232–240.
Kawas, C. H., Greenia, D. E., Bullain, S. S., Clark, C. M., Pontecorvo, M. J., Joshi, A. D. &
Corrada, M. M. 2013. Amyloid imaging and cognitive decline in nondemented oldest-
old: the 90+ Study. Alzheimers Dement, 9, 199–203.
501
Kennedy, K. M., Rodrigue, K. M., Devous, M. D., SR., Hebrank, A. C., Bischof, G. N. & Park,
D. C. 2012. Effects of beta-amyloid accumulation on neural function during encoding
across the adult lifespan. Neuroimage, 62, 1–8.
Knopman, D. S., Jack, C. R., Jr., Wiste, H. J., Weigand, S. D., Vemuri, P., Lowe, V., Kantarci,
K., Gunter, J. L., Senjem, M. L., Ivnik, R. J., Roberts, R. O., Boeve, B. F. & Petersen, R. C.
2012. Short-term clinical outcomes for stages of NIA-AA preclinical Alzheimer disease.
Neurology, 78, 1576–1582.
Knopman, D. S., Jack, C. R., Jr., Wiste, H. J., Weigand, S. D., Vemuri, P., Lowe, V. J., Kantarci,
K., Gunter, J. L., Senjem, M. L., Mielke, M. M., Roberts, R. O., Boeve, B. F. & Petersen,
R. C. 2013. Selective worsening of brain injury biomarker abnormalities in cognitively
normal elderly persons with beta-amyloidosis. JAMA Neurol, 70, 1030–1038.
Kryscio, R. J., Abner, E. L., Cooper, G. E., Fardo, D. W., Jicha, G. A., Nelson, P. T., Smith,
C. D., Van Eldik, L. J., Wan, L. & Schmitt, F. A. 2014. Self-reported memory com-
plaints: Implications from a longitudinal cohort with autopsies. Neurology.
Landau, S. M., Marks, S. M., Mormino, E. C., Rabinovici, G. D., Oh, H., O’neil, J. P., Wilson,
R. S. & Jagust, W. J. 2012. Association of lifetime cognitive engagement and low beta-
amyloid deposition. Arch Neurol, 69, 623–629.
Lee, H. G., Casadesus, G., Zhu, X., Takeda, A., Perry, G. & Smith, M. A. 2004. Challenging the
amyloid cascade hypothesis: senile plaques and amyloid-beta as protective adaptations to
Alzheimer disease. Ann N Y Acad Sci, 1019, 1–4.
Li, G., Millard, S. P., Peskind, E. R., Zhang, J., Yu, C. E., Leverenz, J. B., Mayer, C., Shofer, J.
S., Raskind, M. A., Quinn, J. F., Galasko, D. R. & Montine, T. J. 2014. Cross-sectional and
longitudinal relationships between cerebrospinal fluid biomarkers and cognitive function
in people without cognitive impairment from across the adult life span. JAMA Neurol, 71,
742–751.
Li, S., Jin, M., Zhang, D., Yang, T., Koeglsperger, T., Fu, H. & Selkoe, D. J. 2013. Environmental
novelty activates beta2-adrenergic signaling to prevent the impairment of hippocampal
LTP by Abeta oligomers. Neuron, 77, 929–941.
Lim, H. K., Nebes, R., Snitz, B., Cohen, A., Mathis, C., Price, J., Weissfeld, L., Klunk, W. &
Aizenstein, H. J. 2014a. Regional amyloid burden and intrinsic connectivity networks in
cognitively normal elderly subjects. Brain.
Lim, Y. Y., Ellis, K. A., Ames, D., Darby, D., Harrington, K., Martins, R. N., Masters, C. L.,
Rowe, C., Savage, G., Szoeke, C., Villemagne, V. L. & Maruff, P. 2012 [Epub ahead of
print]. Abeta amyloid, cognition, and APOE genotype in healthy older adults. Alzheimers
Dement.
Lim, Y. Y., Maruff, P., Pietrzak, R. H., Ames, D., Ellis, K. A., Harrington, K., Lautenschlager,
N. T., Szoeke, C., Martins, R. N., Masters, C. L., Villemagne, V. L., Rowe, C. C. & Group,
A. R. 2014b. Effect of amyloid on memory and non-memory decline from preclinical to
clinical Alzheimer’s disease. Brain, 137, 221–231.
Lim, Y. Y., Maruff, P., Pietrzak, R. H., Ellis, K. A., Darby, D., Ames, D., Harrington, K., Martins,
R. N., Masters, C. L., Szoeke, C., Savage, G., Villemagne, V. L., Rowe, C. C. & Group, A.
R. 2014c. Abeta and cognitive change: Examining the preclinical and prodromal stages of
Alzheimer’s disease. Alzheimers Dement.
Lim, Y. Y., Pietrzak, R. H., Ellis, K. A., Jaeger, J., Harrington, K., Ashwood, T., Szoeke, C.,
Martins, R. N., Bush, A. I., Masters, C. L., Rowe, C. C., Villemagne, V. L., Ames, D.,
Darby, D. & Maruff, P. 2012. Rapid Decline in Episodic Memory in Healthy Older Adults
with High Amyloid-beta. J Alzheimers Dis.
Lim, Y. Y., Villemagne, V. L., Laws, S. M., Ames, D., Pietrzak, R. H., Ellis, K. A., Harrington,
K. D., Bourgeat, P., Salvado, O., Darby, D., Snyder, P. J., Bush, A. I., Martins, R. N.,
Masters, C. L., Rowe, C. C., Nathan, P. J., Maruff, P., Australian Imaging, B. & Lifestyle
502
Research, G. 2013. BDNF Val66Met, Abeta amyloid, and cognitive decline in preclinical
Alzheimer’s disease. Neurobiol Aging, 34, 2457–2464.
Lustig, C., Snyder, A. Z., Bhakta, M., O’brien, K. C., Mcavoy, M., Raichle, M. E., Morris, J.
C. & Buckner, R. L. 2003. Functional deactivations: change with age and dementia of the
Alzheimer type. Proc Natl Acad Sci U S A, 100, 14504–14509.
Machulda, M. M., Jones, D. T., Vemuri, P., Mcdade, E., Avula, R., Przybelski, S., Boeve, B. F.,
Knopman, D. S., Petersen, R. C. & Jack, C. R., Jr. 2011. Effect of APOE epsilon4 status
on intrinsic network connectivity in cognitively normal elderly subjects. Arch Neurol, 68,
1131–1136.
Maruyama, M., Shimada, H., Suhara, T., Shinotoh, H., Ji, B., Maeda, J., Zhang, M. R.,
Trojanowski, J. Q., Lee, V. M., Ono, M., Masamoto, K., Takano, H., Sahara, N., Iwata,
N., Okamura, N., Furumoto, S., Kudo, Y., Chang, Q., Saido, T. C., Takashima, A., Lewis,
J., Jang, M. K., Aoki, I., Ito, H. & Higuchi, M. 2013. Imaging of tau pathology in a tauop-
athy mouse model and in Alzheimer patients compared to normal controls. Neuron, 79,
1094–1108.
Mawuenyega, K. G., Sigurdson, W., Ovod, V., Munsell, L., Kasten, T., Morris, J. C., Yarasheski,
K. E. & Bateman, R. J. 2010. Decreased clearance of CNS beta-amyloid in Alzheimer’s
disease. Science, 330, 1774.
Mills, S. M., Mallmann, J., Santacruz, A. M., Fuqua, A., Carril, M., Aisen, P. S., Althage, M.
C., Belyew, S., Benzinger, T. L., Brooks, W. S., Buckles, V. D., Cairns, N. J., Clifford, D.,
Danek, A., Fagan, A. M., Farlow, M., Fox, N., Ghetti, B., Goate, A. M., Heinrichs, D.,
Hornbeck, R., Jack, C., Jucker, M., Klunk, W. E., Marcus, D. S., Martins, R. N., Masters,
C. M., Mayeux, R., Mcdade, E., Morris, J. C., Oliver, A., Ringman, J. M., Rossor, M.
N., Salloway, S., Schofield, P. R., Snider, J., Snyder, P., Sperling, R. A., Stewart, C.,
Thomas, R. G., Xiong, C. & Bateman, R. J. 2013. Preclinical trials in autosomal dominant
AD: implementation of the DIAN-TU trial. Rev Neurol (Paris), 169, 737–743.
Mintun, M. A., Larossa, G. N., Sheline, Y. I., Dence, C. S., Lee, S. Y., Mach, R. H., Klunk,
W. E., Mathis, C. A., Dekosky, S. T. & Morris, J. C. 2006. [11C]PIB in a nondemented
population: potential antecedent marker of Alzheimer disease. Neurology, 67, 446–452.
Montine, T. J. 2011. Prevalence estimates for latent neurodegenerative disease. Toxicol Pathol,
39, 99–102.
Mormino, E. C., Betensky, R. A., Hedden, T., Schultz, A. P., Amariglio, R. E., Rentz, D.
M., Johnson, K. A. & Sperling, R. A. 2014a. Synergistic Effect of beta-Amyloid and
Neurodegeneration on Cognitive Decline in Clinically Normal Individuals. JAMA Neurol.
Mormino, E. C., Betensky, R. A., Hedden, T., Schultz, A. P., Ward, A., Huijbers, W.,
Rentz, D. M., Johnson, K. A., Sperling, R. A., Alzheimer’s Disease Neuroimaging, I.,
Australian Imaging, B., Lifestyle Flagship Study Of, A. & Harvard Aging Brain, S.
2014b. Amyloid and APOE epsilon4 interact to influence short-term decline in preclin-
ical Alzheimer disease. Neurology, 82, 1760–1767.
Mormino, E. C., Smiljic, A., Hayenga, A. O., Onami, S. H., Greicius, M. D., Rabinovici, G.
D., Janabi, M., Baker, S. L., Yen, I. V., Madison, C. M., Miller, B. L. & Jagust, W. J. 2011.
Relationships between beta-amyloid and functional connectivity in different components
of the default mode network in aging. Cereb Cortex, 21, 2399–2407.
Morris, J. C., Storandt, M., Mckeel, D. W., Jr., Rubin, E. H., Price, J. L., Grant, E. A. & Berg,
L. 1996. Cerebral amyloid deposition and diffuse plaques in “normal” aging: Evidence for
presymptomatic and very mild Alzheimer’s disease. Neurology, 46, 707–719.
Nelson, P. T., Alafuzoff, I., Bigio, E. H., Bouras, C., Braak, H., Cairns, N. J., Castellani, R. J.,
Crain, B. J., Davies, P., Del Tredici, K., Duyckaerts, C., Frosch, M. P., Haroutunian, V.,
Hof, P. R., Hulette, C. M., Hyman, B. T., Iwatsubo, T., Jellinger, K. A., Jicha, G. A., Kovari,
E., Kukull, W. A., Leverenz, J. B., Love, S., Mackenzie, I. R., Mann, D. M., Masliah, E.,
503
Mckee, A. C., Montine, T. J., Morris, J. C., Schneider, J. A., Sonnen, J. A., Thal, D. R.,
Trojanowski, J. Q., Troncoso, J. C., Wisniewski, T., Woltjer, R. L. & Beach, T. G. 2012.
Correlation of Alzheimer disease neuropathologic changes with cognitive status: a review
of the literature. J Neuropathol Exp Neurol, 71, 362–381.
Nosheny, R. L., Insel, P. S., Truran, D., Schuff, N., Jack, C. R., Jr., Aisen, P. S., Shaw, L. M.,
Trojanowski, J. Q. & Weiner, M. W. 2014. Variables associated with hippocampal atrophy
rate in normal aging and mild cognitive impairment. Neurobiol Aging.
O’brien, J. L., O’keefe, K. M., Laviolette, P. S., Deluca, A. N., Blacker, D., Dickerson, B. C. &
Sperling, R. A. 2010. Longitudinal fMRI in elderly reveals loss of hippocampal activation
with clinical decline. Neurology, 74, 1969–1976.
Palop, J. J. & Mucke, L. 2010. Amyloid-beta-induced neuronal dysfunction in Alzheimer’s
disease: from synapses toward neural networks. Nat Neurosci, 13, 812–818.
Perez-Nievas, B. G., Stein, T. D., Tai, H. C., Dols-Icardo, O., Scotton, T. C., Barroeta-Espar,
I., Fernandez-Carballo, L., De Munain, E. L., Perez, J., Marquie, M., Serrano-Pozo, A.,
Frosch, M. P., Lowe, V., Parisi, J. E., Petersen, R. C., Ikonomovic, M. D., Lopez, O. L.,
Klunk, W., Hyman, B. T. & Gomez-Isla, T. 2013. Dissecting phenotypic traits linked to
human resilience to Alzheimer’s pathology. Brain, 136, 2510–2526.
Perrotin, A., Mormino, E. C., Madison, C. M., Hayenga, A. O. & Jagust, W. J. 2012. Subjective
cognition and amyloid deposition imaging: a pittsburgh compound B positron emission
tomography study in normal elderly individuals. Arch Neurol, 69, 223–229.
Pihlajamaki, M., K, O. K., Bertram, L., Tanzi, R. E., Dickerson, B. C., Blacker, D., Albert, M.
S. & Sperling, R. A. 2010. Evidence of altered posteromedial cortical FMRI activity in
subjects at risk for Alzheimer disease. Alzheimer Dis Assoc Disord, 24, 28–36.
Price, J. L. & Morris, J. C. 1999. Tangles and plaques in nondemented aging and “preclinical”
Alzheimer’s disease. Ann Neurol, 45, 358–368.
Putcha, D., Brickhouse, M., O’keefe, K., Sullivan, C., Rentz, D., Marshall, G., Dickerson, B.
& Sperling, R. 2011. Hippocampal hyperactivation associated with cortical thinning in
Alzheimer’s disease signature regions in non-demented elderly adults. J Neurosci, 31,
17680–17688.
Quiroz, Y. T., Budson, A. E., Celone, K., Ruiz, A., Newmark, R., Castrillon, G., Lopera, F. &
Stern, C. E. 2010. Hippocampal hyperactivation in presymptomatic familial Alzheimer’s
disease. Ann Neurol, 68, 865–875.
Rami, L., Sala-Llonch, R., Sole-Padulles, C., Fortea, J., Olives, J., Llado, A., Pena-Gomez, C.,
Balasa, M., Bosch, B., Antonell, A., Sanchez-Valle, R., Bartres-Faz, D. & Molinuevo, J.
L. 2012. Distinct functional activity of the precuneus and posterior cingulate cortex during
encoding in the preclinical stage of Alzheimer’s disease. J Alzheimers Dis, 31, 517–526.
Reiman, E. M., Langbaum, J. B., Fleisher, A. S., Caselli, R. J., Chen, K., Ayutyanont, N., Quiroz,
Y. T., Kosik, K. S., Lopera, F. & Tariot, P. N. 2011. Alzheimer’s Prevention Initiative: a
plan to accelerate the evaluation of presymptomatic treatments. J Alzheimers Dis, 26 Suppl
3, 321–329.
Reiman, E. M., Quiroz, Y. T., Fleisher, A. S., Chen, K., Velez-Pardo, C., Jimenez-Del-Rio,
M., Fagan, A. M., Shah, A. R., Alvarez, S., Arbelaez, A., Giraldo, M., Acosta-Baena,
N., Sperling, R. A., Dickerson, B., Stern, C. E., Tirado, V., Munoz, C., Reiman, R. A.,
Huentelman, M. J., Alexander, G. E., Langbaum, J. B., Kosik, K. S., Tariot, P. N. &
Lopera, F. 2012. Brain imaging and fluid biomarker analysis in young adults at genetic
risk for autosomal dominant Alzheimer’s disease in the presenilin 1 E280A kindred: a
case-control study. Lancet Neurol, 11, 1048–1056.
Rentz, D. M., Amariglio, R. E., Becker, J. A., Frey, M., Olson, L. E., Frishe, K., Carmasin, J.,
Maye, J. E., Johnson, K. A. & Sperling, R. A. 2011. Face-name associative memory perfor-
mance is related to amyloid burden in normal elderly. Neuropsychologia, 49, 2776–2783.
504
Rentz, D. M., Locascio, J. J., Becker, J. A., Moran, E. K., Eng, E., Buckner, R. L., Sperling, R.
A. & Johnson, K. A. 2010. Cognition, reserve, and amyloid deposition in normal aging.
Ann Neurol, 67, 353–364.
Rentz, D. M., Parra Rodriguez, M. A., Amariglio, R., Stern, Y., Sperling, R. & Ferris, S. 2013.
Promising developments in neuropsychological approaches for the detection of preclinical
Alzheimer’s disease: a selective review. Alzheimers Res Ther, 5, 58.
Resnick, S. M., Sojkova, J., Zhou, Y., An, Y., Ye, W., Holt, D. P., Dannals, R. F., Mathis, C.
A., Klunk, W. E., Ferrucci, L., Kraut, M. A. & Wong, D. F. 2010. Longitudinal cognitive
decline is associated with fibrillar amyloid-beta measured by [11C]PiB. Neurology, 74,
807–815.
Rodrigue, K. M., Kennedy, K. M., Devous, M. D., Sr., Rieck, J. R., Hebrank, A. C.,
Diaz-Arrastia, R., Mathews, D. & Park, D. C. 2012. β-Amyloid burden in healthy
aging: Regional distribution and cognitive consequences. Neurology, 78, 387–395.
Roe, C. M., Fagan, A. M., Grant, E. A., Marcus, D. S., Benzinger, T. L., Mintun, M. A.,
Holtzman, D. M. & Morris, J. C. 2011. Cerebrospinal fluid biomarkers, education, brain
volume, and future cognition. Arch Neurol, 68, 1145–1151.
Rosas, H. D., Tuch, D. S., Hevelone, N. D., Zaleta, A. K., Vangel, M., Hersch, S. M. & Salat,
D. H. 2006. Diffusion tensor imaging in presymptomatic and early Huntington’s dis-
ease: Selective white matter pathology and its relationship to clinical measures. Mov
Disord, 21, 1317–1325.
Rowe, C. C., Bourgeat, P., Ellis, K. A., Brown, B., Lim, Y. Y., Mulligan, R., Jones, G., Maruff,
P., Woodward, M., Price, R., Robins, P., Tochon-Danguy, H., O’keefe, G., Pike, K. E.,
Yates, P., Szoeke, C., Salvado, O., Macaulay, S. L., O’meara, T., Head, R., Cobiac, L.,
Savage, G., Martins, R., Masters, C. L., Ames, D. & Villemagne, V. L. 2013. Predicting
Alzheimer disease with beta-amyloid imaging: results from the Australian imaging, bio-
markers, and lifestyle study of ageing. Ann Neurol, 74, 905–913.
Rowe, C. C., Ellis, K. A., Rimajova, M., Bourgeat, P., Pike, K. E., Jones, G., Fripp, J., Tochon-
Danguy, H., Morandeau, L., O’keefe, G., Price, R., Raniga, P., Robins, P., Acosta, O.,
Lenzo, N., Szoeke, C., Salvado, O., Head, R., Martins, R., Masters, C. L., Ames, D. &
Villemagne, V. L. 2010. Amyloid imaging results from the Australian Imaging, Biomarkers
and Lifestyle (AIBL) study of aging. Neurobiol Aging, 31, 1275–1283.
Schott, J. M., Bartlett, J. W., Fox, N. C. & Barnes, J. 2010. Increased brain atrophy rates in
cognitively normal older adults with low cerebrospinal fluid Abeta1-42. Ann Neurol, 68,
825–834.
Selkoe, D. J. 2011. Resolving controversies on the path to Alzheimer’s therapeutics. Nat Med,
17, 1060–1065.
Shankar, G. M., Li, S., Mehta, T. H., Garcia-Munoz, A., Shepardson, N. E., Smith, I., Brett, F.
M., Farrell, M. A., Rowan, M. J., Lemere, C. A., Regan, C. M., Walsh, D. M., Sabatini, B.
L. & Selkoe, D. J. 2008. Amyloid-beta protein dimers isolated directly from Alzheimer’s
brains impair synaptic plasticity and memory. Nat Med, 14, 837–842.
Sheline, Y. I., Raichle, M. E., Snyder, A. Z., Morris, J. C., Head, D., Wang, S. & Mintun, M. A.
2009. Amyloid plaques disrupt resting state default mode network connectivity in cogni-
tively normal elderly. Biol Psychiatry, 67, 584–587.
Sperling, R. A., Aisen, P. S., Beckett, L. A., Bennett, D. A., Craft, S., Fagan, A. M., Iwatsubo, T.,
Jack, C. R., Jr., Kaye, J., Montine, T. J., Park, D. C., Reiman, E. M., Rowe, C. C., Siemers,
E., Stern, Y., Yaffe, K., Carrillo, M. C., Thies, B., Morrison-Bogorad, M., Wagster, M.
V. & Phelps, C. H. 2011a. Toward defining the preclinical stages of Alzheimer’s dis-
ease: Recommendations from the National Institute on Aging-Alzheimer’s Association
workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement, 7,
280–292.
505
Sperling, R. A., Dickerson, B. C., Pihlajamaki, M., Vannini, P., Laviolette, P. S., Vitolo, O. V.,
Hedden, T., Becker, J. A., Rentz, D. M., Selkoe, D. J. & Johnson, K. A. 2010. Functional
alterations in memory networks in early Alzheimer’s disease. Neuromolecular Med,
12, 27–43.
Sperling, R. A., Jack, C. R., Jr. & Aisen, P. S. 2011b. Testing the right target and right drug at
the right stage. Sci Transl Med, 3, 111cm33.
Sperling, R. A., Johnson, K. A., Doraiswamy, P. M., Reiman, E. M., Fleisher, A. S., Sabbagh, M.
N., Sadowsky, C. H., Carpenter, A., Davis, M. D., Lu, M., Flitter, M., Joshi, A. D., Clark, C.
M., Grundman, M., Mintun, M. A., Skovronsky, D. M. & Pontecorvo, M. J. 2013. Amyloid
deposition detected with florbetapir F 18 ((18)F-AV-45) is related to lower episodic mem-
ory performance in clinically normal older individuals. Neurobiol Aging, 34, 822–831.
Sperling, R. A., Laviolette, P. S., O’keefe, K., O’brien, J., Rentz, D. M., Pihlajamaki, M.,
Marshall, G., Hyman, B. T., Selkoe, D. J., Hedden, T., Buckner, R. L., Becker, J. A. &
Johnson, K. A. 2009. Amyloid deposition is associated with impaired default network
function in older persons without dementia. Neuron, 63, 178–188.
Sperling, R. A., Rentz, D. M., Johnson, K. A., Karlawish, J., Donohue, M., Salmon, D. P. &
Aisen, P. 2014. The A4 Study: Stopping AD before Symptoms Begin? Sci Trans Med, 6,
228fs13.
Stark, S. M., Yassa, M. A., Lacy, J. W. & Stark, C. E. 2013. A task to assess behavioral pattern
separation (BPS) in humans: Data from healthy aging and mild cognitive impairment.
Neuropsychologia, 51, 2442–2449.
Stern, Y. 2009. Cognitive reserve. Neuropsychologia, 47, 2015–2028.
Storandt, M., Mintun, M. A., Head, D. & Morris, J. C. 2009. Cognitive decline and brain
volume loss as signatures of cerebral amyloid-beta peptide deposition identified with
Pittsburgh compound B: cognitive decline associated with Abeta deposition. Arch Neurol,
66, 1476–1481.
Thomas, J. B., Brier, M. R., Bateman, R. J., Snyder, A. Z., Benzinger, T. L., Xiong, C., Raichle,
M., Holtzman, D. M., Sperling, R. A., Mayeux, R., Ghetti, B., Ringman, J. M., Salloway,
S., Mcdade, E., Rossor, M. N., Ourselin, S., Schofield, P. R., Masters, C. L., Martins, R. N.,
Weiner, M. W., Thompson, P. M., Fox, N. C., Koeppe, R. A., Jack, C. R., Jr., Mathis, C. A.,
Oliver, A., Blazey, T. M., Moulder, K., Buckles, V., Hornbeck, R., Chhatwal, J., Schultz, A.
P., Goate, A. M., Fagan, A. M., Cairns, N. J., Marcus, D. S., Morris, J. C. & Ances, B. M.
2014. Functional connectivity in autosomal dominant and late-onset Alzheimer disease.
JAMA Neurol, 71, 1111–1122.
Toledo, J. B., Weiner, M. W., Wolk, D. A., Da, X., Chen, K., Arnold, S. E., Jagust, W., Jack, C.,
Reiman, E. M., Davatzikos, C., Shaw, L. M. & Trojanowski, J. Q. 2014. Neuronal injury
biomarkers and prognosis in ADNI subjects with normal cognition. Acta Neuropathol
Commun, 2, 26.
Van Harten, A. C., Smits, L. L., Teunissen, C. E., Visser, P. J., Koene, T., Blankenstein, M. A.,
Scheltens, P. & Van Der Flier, W. M. 2013. Preclinical AD predicts decline in memory and
executive functions in subjective complaints. Neurology, 81, 1409–1416.
Vemuri, P., Lesnick, T. G., Przybelski, S. A., Knopman, D. S., Roberts, R. O., Lowe, V. J.,
Kantarci, K., Senjem, M. L., Gunter, J. L., Boeve, B. F., Petersen, R. C. & Jack, C. R., Jr.
2012. Effect of lifestyle activities on Alzheimer disease biomarkers and cognition. Ann
Neurol, 72, 730–738.
Villemagne, V. L., Burnham, S., Bourgeat, P., Brown, B., Ellis, K. A., Salvado, O., Szoeke, C.,
Macaulay, S. L., Martins, R., Maruff, P., Ames, D., Rowe, C. C., Masters, C. L., Australian
Imaging, B. & Lifestyle Research, G. 2013. Amyloid beta deposition, neurodegeneration,
and cognitive decline in sporadic Alzheimer’s disease: a prospective cohort study. Lancet
Neurol, 12, 357–367.
506
Vos, S. J., Xiong, C., Visser, P. J., Jasielec, M. S., Hassenstab, J., Grant, E. A., Cairns, N. J.,
Morris, J. C., Holtzman, D. M. & Fagan, A. M. 2013. Preclinical Alzheimer’s disease and
its outcome: a longitudinal cohort study. Lancet Neurol, 12, 957–965.
Wang, L., Brier, M. R., Snyder, A. Z., Thomas, J. B., Fagan, A. M., Xiong, C., Benzinger, T.
L., Holtzman, D. M., Morris, J. C. & Ances, B. M. 2013. Cerebrospinal fluid Abeta42,
phosphorylated Tau181, and resting- state functional connectivity. JAMA Neurol, 70,
1242–1248.
Whitwell, J. L., Tosakulwong, N., Weigand, S. D., Senjem, M. L., Lowe, V. J., Gunter, J. L.,
Boeve, B. F., Knopman, D. S., Dickerson, B. C., Petersen, R. C. & Jack, C. R., Jr. 2013.
Does amyloid deposition produce a specific atrophic signature in cognitively normal sub-
jects? Neuroimage Clin, 2, 249–257.
Wilson, R. S., Barnes, L. L., Aggarwal, N. T., Boyle, P. A., Hebert, L. E., Mendes De Leon, C.
F. & Evans, D. A. Cognitive activity and the cognitive morbidity of Alzheimer disease.
Neurology, 75, 990–996.
Wilson, R. S., Boyle, P. A., Yu, L., Barnes, L. L., Schneider, J. A. & Bennett, D. A. 2013.
Life-span cognitive activity, neuropathologic burden, and cognitive aging. Neurology, 81,
314–321.
Wilson, R. S., Scherr, P. A., Schneider, J. A., Tang, Y. & Bennett, D. A. 2007. Relation of cogni-
tive activity to risk of developing Alzheimer disease. Neurology, 69, 1911–1920.
Wirth, M., Madison, C. M., Rabinovici, G. D., Oh, H., Landau, S. M. & Jagust, W. J. 2013a.
Alzheimer’s disease neurodegenerative biomarkers are associated with decreased cogni-
tive function but not beta-amyloid in cognitively normal older individuals. J Neurosci, 33,
5553–5563.
Wirth, M., Villeneuve, S., Haase, C. M., Madison, C. M., Oh, H., Landau, S. M., Rabinovici,
G. D. & Jagust, W. J. 2013b. Associations between Alzheimer disease biomarkers, neu-
rodegeneration, and cognition in cognitively normal older people. JAMA Neurol, 70,
1512–1519.
Wolf, R. C., Gron, G., Sambataro, F., Vasic, N., Wolf, N. D., Thomann, P. A., Saft, C.,
Landwehrmeyer, G. B. & Orth, M. 2012. Brain activation and functional connectivity
in premanifest Huntington’s disease during states of intrinsic and phasic alertness. Hum
Brain Mapp, 33, 2161–2173.
Wolf, R. C., Thomann, P. A., Thomann, A. K., Vasic, N., Wolf, N. D., Landwehrmeyer, G.
B. & Orth, M. 2013. Brain structure in preclinical Huntington’s disease: a multi-method
approach. Neurodegener Dis, 12, 13–22.
Xia, C. F., Arteaga, J., Chen, G., Gangadharmath, U., Gomez, L. F., Kasi, D., Lam, C., Liang,
Q., Liu, C., Mocharla, V. P., Mu, F., Sinha, A., Su, H., Szardenings, A. K., Walsh, J. C.,
Wang, E., Yu, C., Zhang, W., Zhao, T. & Kolb, H. C. 2013. [(18)F]T807, a novel tau pos-
itron emission tomography imaging agent for Alzheimer’s disease. Alzheimers Dement.
Zhang, W., Arteaga, J., Cashion, D. K., Chen, G., Gangadharmath, U., Gomez, L. F., Kasi, D.,
Lam, C., Liang, Q., Liu, C., Mocharla, V. P., Mu, F., Sinha, A., Szardenings, A. K., Wang,
E., Walsh, J. C., Xia, C., Yu, C., Zhao, T. & Kolb, H. C. 2012. A highly selective and spe-
cific PET tracer for imaging of tau pathologies. J Alzheimers Dis, 31, 601–612.
507
20
Late-Life Depression
Translating Neurobiological Hypotheses
into Novel Treatments
George S. Alexopoulos
Robert E. Kelly
D epression is a health hazard because it is common and has devastating

outcomes for the health and well-being of people of all ages. In older
people, it can also be fatal, as it increases substantially both suicide and non-suicide
mortality (Schulz et al., 2000; Conwell et al., 2002). Many pharmacological and non-
pharmacological treatments exist for late-life depression. Each of these treatments
helps only a fraction of patients (Fava et al., 2003). So far, the empirical basis for
investigating “which treatments best fit which patients” principally consists of post-
hoc analyses of unitary treatments (e.g., comparing the impact of an antidepressant
to a psychotherapy) tested with the “ideal” subject who has little medical burden,
disability or cognitive impairment (Zimmerman et al., 2002). However, this ideal sub-
ject does not reflect the reality of most depressed older adults. Rather, the typical
depressed elderly person faces a bewildering constellation of interacting health threats
and chronic psychosocial stress compromising his or her outcomes (National Institute
of Mental Health, 1999; National Advisory Mental Health Council Workgroup, 2006).
These complexities introduce significant barriers to the “reduction” inherent in
experimental work. Nonetheless, significant progress has been made in understanding
some of the mechanisms of late-life depression and in treatment development. In the
present chapter, we present an overall model of late-life depression that has guided
our research. We present three major hypotheses derived from the theoretical platform
provided by the model. Each hypothesis is introduced and followed by a select sum-
mary of studies from our laboratory, as well as contributions by others, who evaluated
507
508
the hypotheses. In our view, the reported work followed the classical process of
empirical sciences. Accordingly, it relied initially on theory (omnibus hypotheses)
that led to targeted hypotheses, which after experimentation, led to explanations of
specific events. This process allows theoretical and experimental bias at all three lev-
els (for comment see Alexopoulos, 2004). Although we recognize that these findings
should be seen as incomplete knowledge with only relative value, they nevertheless
have guided the development of novel treatments for late-â•‰life depression that are now
at various stages of testing, which we summarize in the final section of the chapter.
A Model of Late-â•‰Life Depression
A model of depression is presented in Figure 20.1 that provides conceptual order

to overarching theories and delineates specific hypotheses of the path that leads to
late-â•‰life depression (Figure 20.1) (Alexopoulos, 2005). This model postulates that
etiological factors, including cerebrovascular insults and inflammatory processes,
alone or exacerbated by neurobiological changes from chronic psychosocial stress,
lead to brain changes conferring vulnerability to depression (predisposing factors) and
worsening its course. Etiological factors may operate directly to activate mechanisms
that mediate depression (i.e., reward and conflict systems abnormalities) or they
may operate indirectly by causing brain abnormalities that enhance neurobiological
response to stress and in turn activate reward and conflict system dysfunction medi-
ating the expression of depressive symptoms. Note that the relationship between etio-
logical factors and neurobiological responses to stress is bidirectional, such that stress
from chronic adversity may promote processes serving as etiological factors that then
directly induce reward and conflict system changes resulting in depressive symptoms
Etiological factors
Aging-related neuronal changes, vascular changes,
repair responses, inflammation, heredity, other
Predisposing Neurobiological
factors responses to
Frontolimbic stress
compromise Inflammation
reactive oxygen
species
dendritic remodeling
neurogenesis
altered functional
connectivity
Mechanisms mediating depression

Reward & conflict systems’ abnormalities
FigureÂ€20.1â•‡ Model for Late-â•‰Life Depression.

â•‡ 509
Late-Life Depressionâ•… 509
and signs, or that compromise brain functions, ultimately resulting in increased vul-
nerability to depression.
The above neurobiological model serves as a blueprint organizing specific testable
hypotheses of relationships among etiological, predisposing, and stress-related factors
on the one hand and mechanisms mediating behavioral expressions of late-life depres-
sion, response to antidepressants, and course of illness.
A Hypothesis on Factors Predisposing to Late-â•‰Life Depression
The Depression–â•‰Executive Dysfunction Hypothesis

We proposed that a “depression–â•‰executive dysfunction” (DED) syndrome exists
with a clinical presentation resembling medial frontal lobe syndrome and hav-
ing poor response to antidepressant drugs (Alexopoulos, 2001). Implicit in this
hypothesis has been a mechanistic relationship whereby frontostriatal and fronto-
limbic dysfunction (clinically expressed as executive dysfunction) predisposes to
late-â•‰life depression.
The depression-â•‰executive dysfunction syndrome is common and has a presentation
consistent with frontolimbic and frontostriatal abnormalities, i.e., psychomotor retar-
dation, reduced interest in activities, impaired insight, and severe behavioral disabil-
ity, but less pronounced depressive ideation and vegetative signs (Alexopoulos et al.,
2002a). Studies using different samples documented that patients with depression and
executive dysfunction often have poor, slow, and unstable response to antidepressants
prescribed at therapeutic dosages (Kalayam and Alexopoulos, 1999; Alexopoulos et
al., 2000; Sneed et al., 2007; Kalayam and Alexopoulos, 1999; Alexopoulos et al.,
2004; Baldwin et al., 2004; Potter et al., 2004; Bogner et al., 2007; Sneed et al., 2007;
Bella et al., 2010; Sheline et al., 2010).
Neuroimaging studies on DED have targeted frontal and frontal subcortical struc-
tures, as structural impairment or functional disruption in these regions is likely to
cause executive dysfunction and also influence mood regulation. At the structural
level, we hypothesized that white matter abnormalities may interfere with limbic-â•‰
cortical balance and lead to chronic depressive syndromes, a mechanism consistent
with both the DED hypothesis as well as the vascular depression hypothesis (dis-
cussed later). Specifically, it was postulated that depressed elders who fail to achieve
remission have microstructural white matter abnormalities in cortico-â•‰striato-â•‰limbic
networks implicated in geriatric depression, and that these abnormalities may repre-
sent a neuroanatomical substrate predisposing to geriatric depression.
In support of this hypothesis, subcortical white matter hyperintensities have been
associated with non-â•‰remission of late-â•‰life depression (Gunning-â•‰Dixon et al., 2010).
Complementing these findings were studies from our lab utilizing diffusion tensor
imaging, a technique that can assess the integrity of white matter tracks, reflected by
higher fractional anisotropy (FA) scores. In two independent samples (Alexopoulos
et al., 2002b; Alexopoulos et al., 2008b), we reported that lower fractional anisot-
ropy in distributed cerebral networks (dorsal and rostral ACC, dorsolateral prefrontal
510
cortex, hippocampus, posterior cingulate, insula, neostriatum, and the midbrain as

well as select temporal and parietal regions) was associated with poor response of
late-life major depression to a serotonin reuptake inhibitor.
We next conducted genetic studies to examine the relationship of compromised
white matter integrity to the serotonin transporter gene 5-HTTLPR and the gene tran-
scribing brain-derived neurotrophic factor (BDNF) to determine their joint relation-
ship to antidepressant response. A question implicit in these studies was: Are genetic
influences on antidepressant response mediated by the frontolimbic abnormalities
identified in non-responders?
Results indicated that depressed, elderly were short 5-HTTLPR allele carriers
and had lower fractional anisotropy than long allele homozygotes in frontolimbic
brain areas, including the dorsal and rostral anterior cingulate, posterior cingulate,
dorsolateral prefrontal and medial prefrontal regions, thalamus, and in other regions
(Alexopoulos et al., 2009). Short allele carriers also had a lower remission rate than
long allele homozygotes. A potential interpretation of these findings is that the risk
for chronicity of geriatric depression in short allele carriers is in part caused by fron-
tolimbic compromise, an abnormality most frequently found in short allele carriers.
Another study focused on the BDNFval/met polymorphism (Alexopoulos et al.,
2010b). This polymorphism is common, may increase the risk for depression, and
affects BDNF secretion, critical for neuronal survival, plasticity, neurogenesis, and
synaptic connectivity. The objectives of the BDNF study were 1) to test the hypothesis
that BDNFval/met status influences the remission rate of geriatric depression; and 2) to
explore whether the relationship between BDNF allelic status to remission is influ-
enced by the presence of microstructural white matter abnormalities. Results indi-
cated that depressed older BDNFmet carriers had a higher remission rate than BDNFval/
val
homozygotes. This effect was not related to microstructural white matter abnor-
malities, which predicted remission independently. These findings suggested that the
relationship between BDNFval66met and remission is due to effects of BDNF in different
brain structures related to mood regulation.
Taken together, these two genetic studies showed that the 5-HTTLPT serotonin
transporter and the BDNFval/met polymorphism each predict treatment responsiveness
of late-life depression. However, only the relationship of 5-HTTLPR to remission
may be mediated by frontolimbic abnormalities.
In addition to a focus on white matter, the DED hypothesis, with its focus on brain
dysfunction, lends itself to functional connectivity (FC) investigation. FC is based
on the observation that spontaneous blood oxygen level dependent (BOLD) signal
fluctuations tend to be correlated across distributed brain regions to form networks
subserved by a specifc process or task. For example, a fronto-parietal-cortical con-
trol network (CCN) was isolated when cognitively normal younger adults showed
intercorrelations in BOLD signal strength when the brain was at rest across regions
associated with cognitive effort (Fox and Raichle, 2007).
We examined two networks in elder depressives that have been isolated with FC
studies, the cognitive control network (CCN) and the default mode network (DMN).
The CCN, mentioned above, includes the dorsal anterior cingulate, dorsolateral pre-
frontal cortex, and parts of the parietal lobe; and the DMN includes the posterior cin-
gulate/precuneus, ventromedial prefrontal cortex, ventral anterior cingulate, inferior
â•‡ 511
lateral parietal lobes, and parts of the temporal lobe. We targeted the CCN because
anatomical and functional abnormalities of its structures have been identified in late-â•‰
life depression and because some of these abnormalities have been linked to poor
response to antidepressants. The DMN consists of regions that are more active at rest
than during task performance. The DMN regions consistently decrease their activ-
ity during cognitive task performance (Fox and Raichle, 2007; Raichle and Snyder,
2007). The DMN has been associated with self-â•‰referential musing in the absence of
a cognitive task when the brain is theoretically “at rest” (Raichle et al., 2001). Thus,
structures of the DMN are central to mood regulation and have been found excessively
activated during depressive episodes.
Results indicated that low resting FC within the CCN and high FC within the DMN
distinguished depressed from normal elderly subjects (Alexopoulos et al., 2012).
Beyond this “double dissociation,” low resting FC within the CCN predicted low
remission rate and persistence of depressive symptoms and signs, apathy, and dysex-
ecutive behavior after treatment with escitalopram. In contrast, resting FC within the
DMN was correlated with pessimism but did not predict treatment response. If con-
firmed, these findings may serve as a signature of the brain’s functional topography
characterizing late-â•‰life depression and sustaining its symptoms. By identifying the
network abnormalities underlying biologically meaningful characteristics (apathy,
dysexecutive behavior, pessimism) and sustaining late-â•‰life depression, these findings
can provide a novel target on which new somatic and psychosocial treatments can be
tested.
Apathy is common in late-â•‰life depression and is another symptom associated with
executive dysfunction and poor antidepressant response. In a preliminary study, we
examined whether resting FC of the nucleus accumbens (NAcc) and the dorsal ante-
rior cingulate (dACC) with other structures can distinguish apathetic depressed older
patients from non-â•‰apathetic depressed patients and normal subjects (Alexopoulos et
al., 2013a). Apathetic depressed patients had lower FC of the NAcc with the amyg-
dala, caudate, putamen, globus pallidus, and thalamus, and increased FC with the
dorsomedial prefrontal cortex, the superior frontal cortex, and the insula than non-â•‰
apathetic patients. Further, apathetic patients had lower FC of the dACC with dor-
solateral and ventrolateral prefrontal cortices and higher FC with the insula and the
orbitofrontal cortex than non-â•‰apathetic patients. This study suggests that FC between
the NAcc and the dACC and structures related to reward and related behavioral
responses characterize apathy of late-life depression; and it offers support for the
hypothesis that frontolimbic dysfunctions are factors predisposing to distinct depres-
sive syndromes.
Hypotheses on Etiological Factors of Late-â•‰Life Depression
The Late-â•‰Onset Hypothesis

The “late-â•‰onset hypothesis” postulates that major depression with first onset in late
life includes a large subgroup of patients in whom neurological brain abnormalities
contribute to the development of the depressive syndrome (Alexopoulos, 1990).
512
The brain abnormalities underlying late-onset depression may or may not be clin-
ically evident when the depressive syndrome first appears. Clinical studies offered
support for the “late-onset hypothesis.” Indeed, compared to elders with early-onset
major depression, individuals with late-onset major depression have been shown to
have a less frequent family history of mood disorders (Baron et al., 1981), a higher
prevalence of dementing disorders and higher rate of dementia development on fol-
low-up (Alexopoulos et al., 1993), greater enlargement in lateral brain ventricles
(Alexopoulos et al., 1992), and more white matter hyperintensities (Coffey et al.,
1988). However, some disagreement exists(Conwell et al., 1989; Herrmann et al.,
1989).
The “late-onset hypothesis” has been useful in generating mechanistic hypotheses.
However, methodological and conceptual concerns limit its value. On a methodolog-
ical level, onset of depression is difficult to identify, especially when early episodes
are of mild severity (Weiner et al., 1994). Moreover, recruitment of older patients is
biased; in the most severe cases of early-onset depression, patients often die early
due to vascular diseases or suicide. On a conceptual level, neurological changes may
contribute to a late-life episode regardless of other depressive episodes in early life.
Moreover, early-onset depression may be a risk factor for brain abnormalities similar
to those hypothesized by the “late-onset hypothesis.” One such mechanism involves
stress-related hormones leading to reduction of neurotrophic factor secretion, and ulti-
mately decreasing neurogenesis at the dentate nucleus of the hippocampus. Another
may be the increased vascular vulnerability which results from repeated episodes of
depression in early life. These concerns, along with the development of imaging tech-
nology that allowed us to explore more sophisticated hypotheses, led us to conclude
our work on the “late-onset hypothesis.”
Nonetheless, it is important to recognize that work on the “late-onset hypothe-
sis” led to findings of clinical and heuristic significance. First, it was shown that
elderly patients with major depression with an initially reversible dementia (a syn-
drome often called “depressive pseudodementia”) have a high risk of developing
irreversible dementia over a period of approximately three years (Alexopoulos et
al., 1993). Most of these patients were diagnosed with late-onset depression. This
work replaced the classical view of depressive pseudodementia as a benign syn-
drome with the new understanding that a high percentage of depressed “pseudo-
demented” patients, indeed, have a dementing disorder at an early stage. Another
study showed that patients with late-onset depression are less likely to recover from
their depressive symptoms compared to elderly patients with early-onset recur-
rent depression (Alexopoulos et al., 1996). This study did not control treatment,
and the intensity of antidepressant drug therapy was weak. Therefore, its findings
might reflect the natural course of “late-onset depression” with little influence by
pharmacotherapy.
Taken together, findings of studies on the “late-onset hypothesis” raised two ques-
tions. First, can cognitive dysfunction serve as a guide to develop studies aiming to
identify brain abnormalities contributing to depression in late life? Second, what
are the specific brain abnormalities contributing to chronicity and/or recurrence of
late-life depression? Experiments aimed at addressing these two questions led to the
hypotheses described below.
â•‡ 513
The Vascular Depression Hypothesis

Late-life depression and its relationship to cerebrovascular disease have been described
by Gaupp as early as 1905 (Gaupp, 1905). Developments in cognitive neuroscience
and in imaging generated new interest in this area. Vascular disease is a possible cause
of neurological abnormalities, through loss of brain tissue and damage to white matter
tracts. These changes in brain structure are frequently observable with MRI long before
the development of clinically observable neurological abnormalities and, importantly,
the exact locations of these changes can be recorded. Thus, a focus on the relationship
between vascular disease and late-â•‰life depression provides a fertile ground from which
to pursue some of the questions generated by the “late-â•‰onset hypothesis.” To this end,
we formulated the “vascular depression hypothesis,” postulating that cerebrovascular
disease predisposes, precipitates or perpetuates some late-â•‰life depressive syndromes
(Alexopoulos et al., 1997; Alexopoulos, 2006; Taylor et al., 2013a).
When proposed, the “vascular depression” hypothesis was supported by (1) the
high comorbidity of depression and vascular risk factors, (2) the high incidence of
depression in stroke, (3) the high prevalence of white matter (WM) hyperintensities
in late-â•‰onset depression, (4) the high frequency of cognitive impairment in depressed
patients with vascular risk factors, and (5) the similarity of cognitive abnormalities of
depression to those associated with WM hyperintensities. The “vascular depression”
hypothesis cannot be directly tested since the mechanisms of depression are unknown.
However, this hypothesis stimulated research that has improved our understanding
of the complex relationships between late-â•‰life depression, vascular risk factors, and
cognition.
There are well-â•‰established relationships between late-â•‰life depression, vascular risk
factors, and cerebral hyperintensities, most of which are related to vascular impair-
ment. Cognitive dysfunction is common in late-â•‰life depression, particularly executive
dysfunction, a finding predictive of poor antidepressant response. Over time, progres-
sion of hyperintensities and cognitive deficits predicts a poor course of depression
and may reflect underlying worsening of vascular disease. Patients with “vascular
depression” have a clinical presentation reminiscent of medial frontal lobe syndrome:
They have more apathy, retardation, and lack of insight, and less agitation and guilt
than depressed elders without vascular risk factors.
Several mechanisms have been proposed to explain how vascular disease leads to
depression. A “disconnection mechanism” argues that vascular damage to specific fiber
tracts impairs the tract’s structural connectivity and contributes to circuitry changes that
mediate clinical symptoms and influence antidepressant response. Supporting the dis-
connection mechanism are studies showing that that late-â•‰life depression is associated
with white matter hyperintensities in specific tracts including the cingulum bundle,
uncinate fasciculus, and superior longitudinal fasciculus (Sheline et al., 2008; Dalby et
al., 2010b; Taylor et al., 2013b). Greater severity of white matter hyperintensities in the
uncinate and superior longitudinal fasciculi is associated both with executive dysfunc-
tion (Bocti et al., 2005; Sheline et al., 2008; Smith et al., 2011) and greater depression
severity(Dalby et al., 2010a). Diffusion tensor imaging studies offer further support to
the disconnection mechanism by demonstrating white matter microstructural abnor-
malities in the uncinate fasciculus, cingulum bundle, anterior thalamic radiation, and
514
superior longitudinal fasciculus (Taylor et al., 2007; Sexton et al., 2012). The uncinate
fasciculus and cingulum bundle are structures important for processing both cognition
and emotions (Gaffan and Wilson, 2008; Rudrauf et al., 2008), although other tracts
are also associated with antidepressant response of late-â•‰life depression (Alexopoulos
et al., 2010a). Fiber tract structural connectivity is correlated with resting-state FC of
connected regions (Skudlarski et al., 2008; van den Heuvel et al., 2008; Damoiseaux
and Greicius, 2009; Teipel et al., 2010; Steffens et al., 2011).
Hypoperfusion may be another mechanism by which vascular abnormalities lead to
late-â•‰life depression. Reduced cerebral blood flow and abnormalities in vascular endothe-
lium are common in late-â•‰life depression (Rajagopalan et al., 2001; Broadley et al., 2002;
Tiemeier et al., 2003; Chen et al., 2006; Greenstein et al., 2010; Paranthaman et al., 2010)
and can impair regional brain function, contributing to affective and cognitive symptoms.
Cerebral blood flow is regulated both by systemic hemodynamics and local autoregula-
tion. Both these processes are impaired by vascular diseases and similar changes have
been identified in late-â•‰life depression (de la Torre, 2012). Patients with late-â•‰life depression
have increased intima media thickness, increased arterial stiffness, and endothelial dys-
function (Rajagopalan et al., 2001; Broadley et al., 2002; Tiemeier et al., 2003; Chen et al.,
2006; Greenstein et al., 2010; Paranthaman et al., 2010); and endothelial function may be
further impaired in poor responders to antidepressant agents (Paranthaman et al., 2012).
Even in the absence of ischemia, reduced perfusion can impair protein synthesis (Mies et
al., 1991) required for cognitive processing (Martin et al., 2000; Debiec et al., 2002) and
for preserving the integrity of cortical functional networks (Kleim et al., 2003). The sub-
cortical white matter is most vulnerable to cerebral blood flow changes and to ischemia,
where its terminal arterioles have limited collateral flow (Moody et al., 1990; Matsushita
et al., 1994). Aging reduces cerebral blood flow in frontotemporal regions (Asllani et al.,
2009). Late-â•‰life depression is associated with perfusion deficits in subcortical, medial and
lateral prefrontal cortex, and temporal structures (Lesser et al., 1994; Vasile et al., 1996;
Ebmeier et al., 1998; Oda et al., 2003; Ishizaki et al., 2008; Dotson et al., 2009). Perfusion
may improve in the dorsolateral prefrontal cortex, but not in other areas, after improve-
ment of depression (Ishizaki et al., 2008). Persistently reduced regional perfusion is asso-
ciated with reduced processing speed (Rabbitt et al., 2006) and may be a biomarker of
nonresponse to antidepressants (Bench et al., 1995; Milo et al., 2001; Vangu et al., 2003;
Vlassenko et al., 2004; Kohn et al., 2007).
Abnormal inflammatory processes may be a third mechanism by which cerebro-
vascular dysfunction leads to depression (described below). Impairment in structural
and functional connectivity, hypoperfusion, and inflammatory responses may act in
a synergistic way and lead to the brain metabolic changes associated with depres-
sion either directly or by compromising frontolimbic and frontostriatal networks, thus
increasing vulnerability to depression.
The Inflammation Hypothesis in Late-Life Depression
Aging-â•‰related and disease-â•‰related processes result in CNS inflammatory changes that

may lead to depressive syndromes (Alexopoulos and Morimoto, 2011). Depressed
older adults have lower familial prevalence of mood disorders than younger adults and
â•‡ 515
greater cognitive symptoms, structural brain abnormalities, medical morbidity, disabil-

ity and mortality (Alexopoulos and Kelly, 2009). Some of these medical and cognitive
events are thought to predispose patients to geriatric depression and may be fueled by
a CNS pro-â•‰inflammatory state (Copeland et al., 2012; Dantzer, 2012). Accordingly,
we have proposed that immune processes are likely to promote changes in the emo-
tional and cognitive neural systems predisposing to geriatric depression or trigger the
metabolic brain changes mediating the depressive syndrome in late life (Alexopoulos
and Morimoto, 2011). We based this assertion on converging findings suggesting that
1) aging results in increased peripheral immune responses, impaired peripheral-â•‰CNS
immune communication, and a shift of the CNS into a pro-â•‰inflammatory state with
exaggerated and prolonged responses to immune challenge; 2) exaggerated and pro-
longed immune responses of the CNS can influence the function of some of the emo-
tional and cognitive networks pertinent to geriatric depression; 3) aging of the brain’s
inflammatory responses leads to behavioral changes reminiscent of the depressive and
cognitive symptoms of geriatric depression; 4) some antidepressants reduce the expres-
sion of several inflammation markers in the periphery; and 5) limited data suggest that
some anti-â•‰inflammatory agents may have antidepressant properties.
Translation of Empirical Findings into Novel Therapeutics
Targeting the DED Syndrome

Pharmacotherapy for DED Syndrome
The “depression–â•‰executive dysfunction syndrome” of late life has prognostic and
therapeutic implications. The poor response of this syndrome to antidepressants, com-
bined with early relapse and recurrence rates, and accompanying disability, suggest
that novel pharmacological and non-â•‰pharmacological approaches need to be consid-
ered. The depression–â•‰executive dysfunction syndrome is an appropriate target for
treatment with dopamine receptor 3 (D3) agonists because frontostriatal dysfunction
may be mediated by dopaminergic systems (Table 20.1). We have argued that existing
findings permitted us to generate the hypothesis that D3 agonists (e.g., pramipexole,
ropinirole), alone or in combination with classical antidepressants would improve
depressive symptomatology as well as executive dysfunction and disability in patients
with the depression–â•‰executive dysfunction syndrome of late life (Alexopoulos, 2001).
Other dopaminergic agents may play a similar role, including partial dopamine
receptor agonists (e.g., aripiprazole, lurasidone), COMT inhibitors (e.g., tolcapone),
selective MAO inhibitors (e.g., selegiline), and bupropion. Some of these drugs are
antidepressants. However, they may be particularly efficacious in DED where dopa-
minergic dysfunction is likely. Therefore, controlled efficacy studies targeting the
DED are warranted.
While dopamine-acting agents should be one of the first pharmacological interven-
tions to be studied in the depression–â•‰executive dysfunction syndrome, other pharma-
cological investigations may also be considered. Agents modifying neurotransmitters
participating in frontostriatal pathways, e.g., acetylcholine and opiates, are candidates
516
for such studies, as drugs influencing these systems are available for human use. These
include cholinesterase inhibitors, as well as opiate receptor agonists and antagonists.
Neuroplasticity-â•‰Based Computerized Cognitive Remediation (nCCR) for DED Syndrome

nCCR may be another novel approach for treating DED. nCCR for humans adapts
training parameters that have been shown to be effective in inducing neuroplastic
change in aging animals (Mahncke et al., 2006; Erickson et al., 2007). Neuronal
mechanisms of neuroplasticity include Hebbian learning, dendritic modifications (in
length, branching, or spine density), synaptic modifications, glial size and number,
axonal and neuronal growth/â•‰regrowth, and metabolic activity (Morimoto, Wexler, and
Alexopoulos, 2012). Earlier, we had observed that semantic clustering of items in
both a verbal learning test and a semantic fluency test predicted poor response to a
serotonin reuptake inhibitor (Morimoto et al., 2011; Morimoto et al., 2012). These
findings are consistent with dysfunction in the semantic network. The SSCN net-
work is adversely affected by aging and depression. Older adults perform worse than
younger adults on tasks requiring semantic strategy (Brickman et al., 2005; Meinzer
et al., 2009). Moreover, aging preferentially damages semantic network structures
(Raz et al., 1997; Raz et al., 2004) and induces region-â•‰specific alterations in dendritic
morphology, cellular connectivity, gene expression and other neuroplastic processes
that ultimately affect semantic network function (Burke and Barnes, 2006). Geriatric
depression occurs in the context of abnormalities in brain structures central to the
semantic network and in their connections (Ballmaier et al., 2004; Gunning-â•‰Dixon
et al., 2008). Further, older adults with major depression have greater deficits in ver-
bal fluency and cognitive control than depressed young adults, which persist after
symptom improvements (Lockwood et al., 2002; Elderkin-â•‰Thompson et al., 2003;
Herrmann et al., 2007; Gualtieri and Johnson, 2008; Pisljar et al., 2008). Based on
the assumption that networks related to semantic organization are critical for anti-
depressant response, we devised an nCCR technique targeting the semantic net-
work with increasingly demanding paradigms that were adaptive to the individual
patient’s learning curve, i.e., the task difficulty increased as the patient’s performance
improved. Each paradigm required attention and offered reward cues. We compared
nCCR to a gold-â•‰standard treatment (escitalopram: target dose 20 mg per 12 weeks)
in 11 treatment-â•‰resistant older adults with major depression, with 33 matched histor-
ical controls (Morimoto et al., 2014). Approximately 91% of participants completed
nCCR. nCCR is equally effective at reducing depressive symptoms as escitalopram,
but does so in 4 weeks instead of 12. In addition, nCCR improved executive function
more than the escitalopram.
Problem-Solving Therapy for DED (PST-â•‰ED)

In patients with the DED syndrome, the behavioral disability resulting from executive
dysfunction and from the motivational disturbance of depression serves as a chronic
stressor. Assuming that persistent exposure to stress contributes to the persistence of
the DED syndrome, we modified problem-solving therapy (PST-â•‰ED) by tailoring it
to the needs of DED patients with the goal of improving their behavioral competence
â•‡ 517
and reducing the experience of stress. To this end, PST-â•‰ED trains patients to use the
problem solving method to address “real life” problems. The problem solving method
teaches the patient to 1) define the problem, 2) establish realistic and achievable goals,
3) generate alternative solutions, 4) generate a list of pros and cons for each solution,
5) evaluate and select a solution, 6) implement the selected solution, and 7) evaluate
the outcome.
We tested PST-â•‰ED (12 sessions) in 221 older adults with DED syndrome and
compared its efficacy with manual-â•‰based supportive therapy (Arean et al., 2010;
Alexopoulos et al., 2011). PST-â•‰ED and supportive therapy led to comparable reduc-
tion of depressive symptoms and of disability in the first 6 weeks of treatment, but
there was a greater reduction in depression and disability in PST-â•‰ED patients at weeks
9 and 12. PST-â•‰ED patients had greater response rates than supportive therapy patients
at weeks 9 (47% vs. 22%) and 12 (57% vs. 34%), and higher remission rates during
the same times (9th week: 38% vs. 22%; 12th week: 46% vs. 28%). PST-â•‰ED yielded
one additional response or remission over supportive therapy for every 4.5th patient
by the end of the trial. The therapeutic advantage of PST-â•‰ED over supportive therapy
in reducing depression was in part due to greater reduction of disability by PST-â•‰ED.
While disability increased during the 24 weeks following the end of treatment, the
advantage of PST-â•‰ED over supportive-therapy treated patients was retained. In non-â•‰
demented patients, depression appears to contribute to behavioral disability mainly
in the context of executive dysfunction, while the impact of depression is limited in
patients with unimpaired executive functions (Kiosses et al., 2000). Therefore, appro-
priately focused interventions may break the downward spiral of behavioral deterio-
ration created by the interaction of depressive symptoms and executive dysfunction.
Targeting “Vascular Depression”

The “vascular depression” hypothesis provides the impetus for focused pharmaco-
logical research (Alexopoulos et al., 1997). Effective treatment of hypertension and
hypecholesterolemia reduces cerebrovascular morbidity and mortality. Ticlopidine,
aspirin, and dipyridamole may prevent future stroke in patients with transient isch-
emic attacks or ischemic stroke. Studies can ascertain whether antihypertensive,
antihypercholesterolemia, and antiplatelet agents improve the outcome of “vascular
depression.” Antiplatelet agents may prove effective in preventing further vascular
damage occurring during depressive episodes, when the serotonin-â•‰mediated thrombo-
genic platelet response is enhanced. In addition, longitudinal assessment of depressive
symptomatology during antiplatelet secondary prevention trials can evaluate the effi-
cacy of these agents in improving the course of “vascular depression.”
Agents with cytoprotective action during ischemia and reperfusion may be rel-
evant to “vascular depression” and require focused investigation. These include
thrombolytic agents, calcium and sodium-â•‰channel antagonists, N-â•‰methyl-â•‰D-â•‰aspar-
tate (NMDA) receptor antagonists, glutamate-synthesis inhibitors, glutamate-â•‰release
inhibitors, gamma-â•‰aminobenzoic acid (GABA) antagonists, gangliocides, aminos-
teroids, antioxidants, growth factors, and antiapoptotic agents. Angiotensin receptor
blockers improve cerebral perfusion (Nagata et al., 2010) and may improve cognition
518
(Fogari et al., 2003; Saxby et al., 2008; Hajjar et al., 2012), although their ability to
reduce depressive symptoms and signs had not been investigated.
Tricyclic antidepressants and other psychotropic agents with alpha-â•‰blocking
action may inhibit behavioral recovery following ischemic lesions, whereas psy-
chotropic drugs increasing cathecholaminergic activity (e.g., dopamine or norepi-
nephrine enhancing agents) may promote recovery. Appropriately targeted studies
may guide the selection of antidepressant approaches in patients with “vascular
depression.”
Targeting the “Inflammation of Late-â•‰Life Depression”

Despite the conceptual appeal of the inflammation theory of depression, few
studies have investigated the role of anti-â•‰inflammatory agents (Alexopoulos and
Morimoto, 2011). However, numerous anti-â•‰ inflammatory agents have become
available recently, though limited data exist on their use or effects in psychiatric
disorders.
A proof-of-concept study of the tumor necrosis factor (TNF) antagonist inflix-
imab may improve depressive symptoms in patients with high baseline inflammatory
biomarkers but does not have generalized efficacy in treatment-â•‰resistant depression.
Etanercept, a soluble TNF-â•‰α receptor, prevents TNF-â•‰α mediated cellular response
by competitively inhibiting the interaction of TNF-â•‰α with cell-â•‰surface receptors.
Etanercept 50 mg twice weekly reduced symptoms of depression in a placebo-â•‰con-
trolled study of patients with psoriasis (Tyring et al., 2006).
The cyclo-â•‰oxygenase-â•‰2 inhibitor celecoxib has been found to augment the efficacy
of reboxetine and fluoxetine in patients with major depression (Chen et al., 2000;
Plane et al., 2010). In a study of osteoarthritis patients, the cyclo-â•‰oxygenase-â•‰2 inhib-
itor rofecoxib reduced symptoms of depression and improved cognition (Collantes-â•‰
Estevez et al., 2003). The putative mechanism of celecoxib’s action is inhibition of
prostaglandin E2 (PGE2). PGE2-â•‰stimulated production of IL-â•‰6 is increased in depres-
sion (Chen et al., 2000; Plane et al., 2010).
Minocycline, a semi-â•‰synthetic, second-â•‰generation tetracycline analog, crosses the
blood-brain barrier and has anti-â•‰inflammatory, anti-â•‰apoptotic, and antioxidant prop-
erties in addition to its bacteriostatic action. The anti-â•‰inflammatory effects are both
direct and indirect through suppression of microglia activation and subsequent release
of cytokines IL-â•‰1β, IL-â•‰6 and TNF-â•‰α (Chen et al., 2000; Plane et al., 2010). Finally,
minocycline inhibits poly-â•‰ADP polymerase 1, a molecule that when activated by
DNA damage contributes to excitotoxicity. In an open study of a small number of
subjects, we observed improvement of depression symptoms in older patients with
major depression who failed a trial of escitalopram.
Targeting Chronic Stress
Our model of depression postulates that stress originating from chronic experiences of
adversity may promote etiological factors, compromise brain functions that increase
â•‡ 519
vulnerability to depression, or directly lead to a depressive state. To break the path

from stress to depression, we developed two brief, manualized therapies targeting
causes of stress in distinct populations with late-â•‰life depression.
Personalized Intervention for Depressed Patients with COPD (PID-â•‰C)

Chronic obstructive pulmonary disease (COPD) exemplifies the health problems
of aging adults living with chronically deteriorating conditions and can serve as a
model for development of interventions addressing their needs. Approximately 25%
of COPD patients suffer from depression. Disability stemming from COPD and exac-
erbated by depression makes every task effortful. The hopelessness of depression
adds to resignation. Ironically, pulmonary rehabilitation consists of strengthening,
breathing, and endurance exercises requiring a persistent effort by the debilitated and
demoralized depressed COPD patient. This constellation of factors undermines treat-
ment adherence, compromises the care of depressed COPD patients and contributes
to chronic stress perpetuating their depression.
We developed a manualized intervention targeting barriers to treatment adherence
of depressed COPD patients (PID-â•‰C) (Sirey et al., 2007; Alexopoulos et al., 2008a).
PID-â•‰C drew from the Theory of Reasoned Action, according to which patients weigh
risks and benefits of treatment (Ajzden, 1996), aimed to shift the balance in favor of
treatment engagement. PID-â•‰C is administered by trained care managers who work
with each patient and the patient’s treatment team. The care managers 1) identify
adherence barriers specific to each patient, and through education and support, help
patients adhere to their exercise regimens and to antidepressants; and 2) work with the
patients’ physicians to facilitate treatment adherence.
The efficacy of PID-â•‰C was compared to that of usual care. PID-â•‰C led to higher
depression remission rate, and greater reduction in depression and dyspnea-â•‰related
disability than usual care over 28 weeks of treatment, and 6 months after the last
session (Alexopoulos et al., 2013b). Lower severity of dyspnea-â•‰related disability and
greater adherence to antidepressants predicted subsequent improvement of depression
(Alexopoulos et al., 2014). Exercise and low depression severity predicted improve-
ment of dyspnea-â•‰related disability. The inter-â•‰relationship of the course of depression
and dyspnea-â•‰related disability underscores the need to target adherence to both anti-
depressant treatment and COPD rehabilitation.
Ecosystem Focused Therapy (EFT) for Post-â•‰Stroke Depression

Stroke exemplifies the problems of aging adults living with disability occurring after
an acute medical event and can serve as a model for development of interventions
addressing their needs (Alexopoulos and Bruce, 2009). More than 20% of stroke
patients suffer from major depression, which worsens their outcomes, undermines
treatment adherence, and compromises their care. A Cochrane meta-â•‰analysis con-
cluded that antidepressants are only weakly efficacious in post-â•‰stroke depression
(Hackett et al., 2008), suggesting that while some patients improve, many derive lim-
ited, if any, benefit.
520
We developed Ecosystem Focused Therapy (EFT) (Alexopoulos and Bruce, 2009),

a manualized intervention that targets the “psychosocial storm” experienced by the
PSD patient and his/her family through five integrated components: 1) It offers to
patients and families an action-oriented, “new perspective” about recovery and the
new physical state; 2) it helps patients to form a treatment “adherence enhancement
structure”; 3) it provides a “problem solving structure” to the patient, focusing on
solvable problems valued by the patient and pertinent to daily function; 4) it helps the
family “reengineer its goals, involvement, and plans” to accommodate the patient’s
disability and its impact on the family (e.g., finances, time commitment); and 5) it
“coordinates care with specialized therapists” to arrive at a synergistic approach
increasing patient participation in treatment and rehabilitation and utilization of com-
munity resources.
We conducted a preliminary study in 24 patients with post-stroke major depression
who were randomly assigned to receive weekly sessions of EFT or a comparison
Table 20.1 Theory-Generated Approaches to Targeting Vulnerability Factors, Etiological

Factors, and Stress Responses in Late-Life Depression
Depression-Executive Dysfunction Syndrome

Dopamine receptor 3 (D3) agonists
Partial dopamine receptor agonists (some atypical antipsychotics)
Catechol-ortho-methyl transferase (COMT) inhibitors
Monoamine oxidase (MAO) inhibitors selective for MAO-B
Bupropion
Neuroplasticity-based Computerized Cognitive Remediation
Problem Solving Therapy for the Depression-Executive Dysfunction Syndrome
Vascular Depression
Prevention of Ischemic Lesions
Treatment of hypertension
Anti-cholesterolemia agents
Antiplatelet agents
Improvement of Perfusion
Angiotensin receptor blockers
After Ischemic Lesions

Cytoprotective agents (thrombolytic agents, calcium and sodium-channel antagonists, N-methyl-D-
aspartate receptor antagonists, glutamate synthesis inhibitors, glutamate-release inhibitors, gamma-
aminobenzoic acid (GABA) antagonists, gangliocides, aminosteroids, antioxidants, growth factors, and
antiapoptotic agents.
Selection of Antidepressants
Antidepressants without alpha receptor blocking properties
Inflammatory Processes
Infliximab
Etanercept
Minocycline
Stress Responses
Personalized Intervention for Depressed Patients with Severe COPD
Ecosystem Focused Therapy for Post Stroke Depression
â•‡ 521
condition consisting of systematic Education on Stroke and Depression and their treat-
ment (ESD) for 12 weeks. In this small sample, EFT may be more efficacious than
ESD in reducing depressive symptoms and signs, in leading to a higher remission rate,
and in ameliorating disability in PSD. Reduction of disability in the early part of the
trial mediated later improvement in depressive symptomatology. Similarly, reduction
in depressive symptoms and signs early on mediated later improvement in disability.
These encouraging findings require replication. Beyond its potential direct benefits
in post-â•‰stroke depression, EFT may provide an appropriate context for efficient and
timely administration of pharmacotherapy and of physical, speech, and occupational
therapies, thus maximizing their efficacy.
Conclusion
This review reflects a historical, dialectic process between theory (omnibus hypoth-
eses), specific hypotheses, and experiments using the available, contemporary tools.
Earlier theories guided studies that generated findings that led to new theories with
greater explanatory power and greater potential for treatment development.
We have argued elsewhere that epistemological criteria do not justify excessive
reliance on observation and experimentation even when complemented by a mean-
ingful theory (Alexopoulos, 2004). The work on late-â•‰life depression outlined in this
chapter reflects the pragmatic (availability of technology, need for novel treatment
development) and social factors (evolving scientific weltanschauung) that have deter-
mined the choices of theories, hypotheses, and studies. We believe that this approach
does not weaken the epistemic value of this work. Awareness of the relative value of
findings and a critical stance toward the factors that influenced them can enhance our
ability to advance further. Clearly, the findings outlined above have at least temporary
usefulness and hopefully will lead to new hypotheses and experiments that will give
us a clearer view of the mechanisms of late-â•‰life depression and better ways to treat
our patients.
References
Ajzden I (1996) The directive influence of attitudes on health behavior. In: The psychology of
action:linking cognition and motivation to behavior (Gollwitzer PM, Bargh JA, eds). New
York: The Guilford Press.
Alexopoulos Gs (1990) Clinical and biological findings in late-â•‰onset depression. In: Review
of Psychiatry (Tasman A, Goldfinger Sm, Kauffmann Ca, eds), pp 249–â•‰262. Washington,
DC: American Psychiatric Press.
Alexopoulos Gs (2001) “The depression-â•‰executive dysfunction syndrome of late life”: a spe-
cific target for D3 agonists? Am J Geriatr Psychiatry 9:22–â•‰29.
Alexopoulos Gs (2004) On the “infallibility” of psychopathology and its implications for
action. Am J Psychiatry 161:2151–â•‰2154.
Alexopoulos Gs (2005) Depression in the elderly. Lancet 365:1961–â•‰1970.
Alexopoulos Gs (2006) The vascular depression hypothesis: 10 years later. Biol Psychiatry
60:1304–â•‰1305.
522
Alexopoulos Gs, Kelly Re, Jr. (2009) Research advances in geriatric depression. World
Alexopoulos Gs, Bruce Ml (2009) A model for intervention research in late-life depression. Int
J Geriatr Psychiatry 24:1325–1334.
Alexopoulos Gs, Morimoto Ss (2011) The inflammation hypothesis in geriatric depression. Int
J Geriatr Psychiatry:1109–1118.
Alexopoulos Gs, Young Rc, Shindledecker Rd (1992) Brain computed tomography findings
in geriatric depression and primary degenerative dementia. Biol Psychiatry 31:591–599.
Alexopoulos Gs, Kiosses Dn, Murphy C, Heo M (2004) Executive dysfunction, heart dis-
ease burden, and remission of geriatric depression. Neuropsychopharmacology 29:
2278–2284.
Alexopoulos Gs, Raue Pj, Sirey Ja, Arean Pa (2008a) Developing an intervention for depressed,
chronically medically ill elders: a model from COPD. Int J Geriatr Psychiatry 23:447–453.
Alexopoulos Gs, Meyers Bs, Young Rc, Mattis S, Kakuma T (1993) The course of geriatric
depression with “reversible dementia”: a controlled study. Am J Psychiatry 150:1693–1699.
Alexopoulos Gs, Kiosses Dn, Klimstra S, Kalayam B, Bruce Ml (2002a) Clinical presentation
of the “depression-executive dysfunction syndrome” of late life. Am J Geriatr Psychiatry
10:98–106.
Alexopoulos Gs, Kiosses Dn, Choi Sj, Murphy Cf, Lim KO (2002b) Frontal white matter
microstructure and treatment response of late-life depression: a preliminary study. Am J
Alexopoulos Gs, Meyers Bs, Young Rc, Campbell S, Silbersweig D, Charlson M (1997)
“Vascular depression” hypothesis. Archives of general psychiatry 54:915–922.
Alexopoulos Gs, Hoptman Mj, Kanellopoulos D, Murphy Cf, Lim Ko, Gunning FM (2012)
Functional Connectivity in the Cognitive Control Network and the Default Mode Network
in Late-life Depression. J Affect Disord 139:56–65.
Alexopoulos Gs, Meyers Bs, Young Rc, Kakuma T, Feder M, Einhorn A, Rosendahl E (1996)
Recovery in geriatric depression. Archives of general psychiatry 53:305–312.
Alexopoulos Gs, Raue Pj, Kiosses Dn, Mackin Rs, Kanellopoulos D, Mcculloch C, Arean
Pa (2011) Problem-solving therapy and supportive therapy in older adults with major
depression and executive dysfunction: effect on disability. Archives of general psychiatry
68:33–41.
Alexopoulos Gs, Hoptman Mj, Yuen G, Kanellopoulos D, Seirup Jk, Lim Ko, Gunning Fm
(2013a) Functional connectivity in apathy of late-life depression: a preliminary study.
J Affect Disord 149:398–405.
Alexopoulos Gs, Meyers Bs, Young Rc, Kalayam B, Kakuma T, Gabrielle M, Sirey Ja, Hull J
(2000) Executive dysfunction and long-term outcomes of geriatric depression. Archives of
general psychiatry 57:285–290.
Alexopoulos Gs, Murphy Cf, Gunning-Dixon Fm, Latoussakis V, Kanellopoulos D, Klimstra S,
Lim Ko, Hoptman MJ (2008b) Microstructural white matter abnormalities and remission
of geriatric depression. Am J Psychiatry 165:238–244.
Alexopoulos Gs, Kiosses Dn, Sirey Ja, Kanellopoulos D, Novitch Rs, Ghosh S, Seirup Jk, Raue
Pj (2013b) Personalised intervention for people with depression and severe COPD. Br J
Alexopoulos Gs, Glatt Ce, Hoptman Mj, Kanellopoulos D, Murphy Cf, Kelly Re, Jr., Morimoto
Ss, Lim Ko, Gunning FM (2010a) BDNF Val66met polymorphism, white matter abnor-
malities and remission of geriatric depression. J Affect Disord 125:262–268.
Alexopoulos Gs, Glatt Ce, Hoptman Mj, Kanellopoulos D, Murphy Cf, Kelly Re, Jr.,
Morimoto Ss, Lim Ko, Gunning FM (2010b) BDNF val66met polymorphism, white
523
Late-Life Depression 523
matter abnormalities and remission of geriatric depression. J Affect Disord 125:

262–268.
Alexopoulos Gs, Kiosses Dn, Sirey Ja, Kanellopoulos D, Seirup Jk, Novitch Rs, Ghosh S,
Banerjee S, Raue PJ (2014) Untangling Therapeutic Ingredients of a Personalized
Intervention for Patients with Depression and Severe COPD. Am J Geriatr Psychiatry
22:1316–1324.
Alexopoulos Gs, Murphy Cf, Gunning-Dixon Fm, Glatt Ce, Latoussakis V, Kelly Re, Jr.,
Kanellopoulos D, Klimstra S, Lim Ko, Young Rc, Hoptman Mj (2009) Serotonin trans-
porter polymorphisms, microstructural white matter abnormalities and remission of geri-
atric depression. J Affect Disord 119:132–141.
Arean Pa, Raue P, Mackin Rs, Kanellopoulos D, Mcculloch C, Alexopoulos Gs (2010) Problem-
solving therapy and supportive therapy in older adults with major depression and executive
dysfunction. Am J Psychiatry 167:1391–1398.
Asllani I, Habeck C, Borogovac A, Brown Tr, Brickman Am, Stern Y (2009) Separating function
from structure in perfusion imaging of the aging brain. Hum Brain Mapp 30:2927–2935.
Baldwin R, Jeffries S, Jackson A, Sutcliffe C, Thacker N, Scott M, Burns A (2004) Treatment
response in late-onset depression: relationship to neuropsychological, neuroradiological
and vascular risk factors. Psychological medicine 34:125–136.
Ballmaier M, Toga Aw, Blanton Re, Sowell Er, Lavretsky H, Peterson J, Pham D, Kumar
A (2004) Anterior cingulate, gyrus rectus, and orbitofrontal abnormalities in elderly
depressed patients: an MRI-based parcellation of the prefrontal cortex. Am J Psychiatry
161:99–108.
Baron M, Mendlewicz J, Klotz J (1981) Age-of-onset and genetic transmission in affective
disorders. Acta psychiatrica Scandinavica 64:373–380.
Bella R, Pennisi G, Cantone M, Palermo F, Pennisi M, Lanza G, Zappia M, Paolucci S (2010)
Clinical presentation and outcome of geriatric depression in subcortical ischemic vascular
disease. Gerontology 56:298–302.
Bench Cj, Frackowiak Rs, Dolan Rj (1995) Changes in regional cerebral blood flow on recov-
ery from depression. Psychol Med 25:247–261.
Bocti C, Swartz Rh, Gao Fq, Sahlas Dj, Behl P, Black Se (2005) A new visual rating scale to
assess strategic white matter hyperintensities within cholinergic pathways in dementia.
Stroke 36:2126–2131.
Bogner Hr, Bruce Ml, Reynolds Cf, 3Rd, Mulsant Bh, Cary Ms, Morales K, Alexopoulos Gs
(2007) The effects of memory, attention, and executive dysfunction on outcomes of depres-
sion in a primary care intervention trial: the PROSPECT study. Int J Geriatr Psychiatry
22:922–929.
Brickman Am, Paul Rh, Cohen Ra, Williams Lm, Macgregor Kl, Jefferson Al, Tate Df, Gunstad
J, Gordon E (2005) Category and letter verbal fluency across the adult lifespan: relation-
ship to EEG theta power. Archives of clinical neuropsychology: the official journal of the
National Academy of Neuropsychologists 20:561–573.
Broadley Aj, Korszun A, Jones Cj, Frenneaux Mp (2002) Arterial endothelial function is
impaired in treated depression. Heart 88:521–523.
Burke Sn, Barnes Ca (2006) Neural plasticity in the ageing brain. Nat Rev Neurosci 7:30–40.
Chen Cs, Chen Cc, Kuo Yt, Chiang Ic, Ko Ch, Lin Hf (2006) Carotid intima-media thickness in
late-onset major depressive disorder. Int J Geriatr Psychiatry 21:36–42.
Chen M, Ona Vo, Li M, Ferrante Rj, Fink Kb, Zhu S, Bian J, Guo L, Farrell La, Hersch Sm,
Hobbs W, Vonsattel Jp, Cha Jh, Friedlander Rm (2000) Minocycline inhibits caspase-1
and caspase-3 expression and delays mortality in a transgenic mouse model of Huntington
disease. Nat Med 6:797–801.
524
Coffey Ce, Figiel Gs, Djang Wt, Cress M, Saunders Wb, Weiner Rd (1988) Leukoencephalopathy
in elderly depressed patients referred for ECT. Biol Psychiatry 24:143–161.
Collantes-Estevez E, Munoz-Villanueva Mc, Canete-Crespillo Jd, Sanmarti-Sala R, Gratacos-
Masmitja J, Zarco-Montejo P, Torre-Alonso Jc, Gonzalez-Fernandez C (2003) Infliximab
in refractory spondyloarthropathies: a multicentre 38 week open study. Ann Rheum Dis
62:1239–1240.
Conwell Y, Duberstein Pr, Caine Ed (2002) Risk factors for suicide in later life. Biol Psychiatry
52:193–204.
Conwell Y, Nelson Jc, Kim Km, Mazure Cm (1989) Depression in late life: age of onset as
marker of a subtype. J Affect Disord 17:189–195.
Copeland We, Shanahan L, Worthman C, Angold A, Costello Ej (2012) Cumulative depression
episodes predict later C-reactive protein levels: a prospective analysis. Biol Psychiatry
71:15–21.
Dalby Rb, Frandsen J, Chakravarty Mm, Ahdidan J, Sorensen L, Rosenberg R, Videbech P,
Ostergaard L (2010a) Depression severity is correlated to the integrity of white matter
fiber tracts in late-onset major depression. Psychiatry Res 184:38–48.
Dalby Rb, Chakravarty Mm, Ahdidan J, Sorensen L, Frandsen J, Jonsdottir Ky, Tehrani E,
Rosenberg R, Ostergaard L, Videbech P (2010b) Localization of white-matter lesions
and effect of vascular risk factors in late-onset major depression. Psychol Med 40:
1389–1399.
Damoiseaux Js, Greicius Md (2009) Greater than the sum of its parts: a review of studies com-
bining structural connectivity and resting-state functional connectivity. Brain Struct Funct
213:525–533.
Dantzer R (2012) Depression and inflammation: an intricate relationship. Biol Psychiatry
71:4–5.
De La Torre Jc (2012) Cerebral Hemodynamics and Vascular Risk Factors: Setting the Stage for
Alzheimer’s Disease. J Alzheimers Dis 32:553–567.
Debiec J, Ledoux Je, Nader K (2002) Cellular and systems reconsolidation in the hippocampus.
Neuron 36:527–538.
Dotson Vm, Beason-Held L, Kraut Ma, Resnick Sm (2009) Longitudinal study of chronic
depressive symptoms and regional cerebral blood flow in older men and women. Int J
Geriatr Psychiatry 24:809–819.
Ebmeier Kp, Glabus Mf, Prentice N, Ryman A, Goodwin Gm (1998) A voxel-based
analysis of cerebral perfusion in dementia and depression of old age. Neuroimage
7:199–208.
Elderkin-Thompson V, Kumar A, Bilker Wb, Dunkin Jj, Mintz J, Moberg Pj, Mesholam Ri, Gur
Re (2003) Neuropsychological deficits among patients with late-onset minor and major
depression. Arch Clin Neuropsychol 18:529–549.
Erickson Ki, Colcombe Sj, Wadhwa R, Bherer L, Peterson Ms, Scalf Pe, Kim Js, Alvarado M,
Kramer Af (2007) Training-induced plasticity in older adults: effects of training on hemi-
spheric asymmetry. Neurobiol Aging 28:272–283.
Fava M, Rush Aj, Trivedi Mh, Nierenberg Aa, Thase Me, Sackeim Ha, Quitkin Fm, Wisniewski
S, Lavori Pw, Rosenbaum Jf, Kupfer Dj (2003) Background and rationale for the sequenced
treatment alternatives to relieve depression (STAR*D) study. The Psychiatric clinics of
North America 26:457–494, x.
Feldman De (2009) Synaptic Mechanisms for Plasticity in Neocortex. Annu Rev Neurosci 32:
33–55.
Fogari R, Mugellini A, Zoppi A, Derosa G, Pasotti C, Fogari E, Preti P (2003) Influence of
losartan and atenolol on memory function in very elderly hypertensive patients. J Hum
Hypertens 17:781–785.
525
Fox Md, Raichle Me (2007) Spontaneous fluctuations in brain activity observed with functional
magnetic resonance imaging. Nat Rev Neurosci 8:700–711.
Gaffan D, Wilson Cr (2008) Medial temporal and prefrontal function: recent behavioural dis-
connection studies in the macaque monkey. Cortex 44:928–935.
Gaupp Re (1905) Depressive states in older age [in German]. Medizinische Wochenschrift
52:1531–1537.
Greenstein As, Paranthaman R, Burns A, Jackson A, Malik Ra, Baldwin Rc, Heagerty Am
(2010) Cerebrovascular damage in late-life depression is associated with structural and
functional abnormalities of subcutaneous small arteries. Hypertension 56:734–740.
Gualtieri Ct, Johnson Lg (2008) Age-related cognitive decline in patients with mood disorders.
Prog Neuropsychopharmacol Biol Psychiatry 32:962–967.
Gunning-Dixon Fm, Hoptman Mj, Lim Ko, Murphy Cf, Klimstra S, Latoussakis V, Majcher-
Tascio M, Hrabe J, Ardekani Ba, Alexopoulos Gs (2008) Macromolecular white mat-
ter abnormalities in geriatric depression: a magnetization transfer imaging study. Am J
Geriatr Psychiatry 16:255–262.
Gunning-Dixon Fm, Walton M, Cheng J, Acuna J, Klimstra S, Zimmerman Me, Brickman Am,
Hoptman Mj, Young Rc, Alexopoulos Gs (2010) MRI signal hyperintensities and treat-
ment remission of geriatric depression. J Affect Disord 126:395–401.
Hackett Ml, Anderson Cs, House A, Xia J (2008) Interventions for treating depression after
stroke. Cochrane Database Syst Rev:CD003437.
Hajjar I, Hart M, Chen Yl, Mack W, Milberg W, Chui H, Lipsitz L (2012) Effect of antihyper-
tensive therapy on cognitive function in early executive cognitive impairment: a double-
blind randomized clinical trial. Arch Intern Med 172:442–444.
Herrmann Ll, Goodwin Gm, Ebmeier Kp (2007) The cognitive neuropsychology of depression
in the elderly. Psychol Med 37:1693–1702.
Herrmann N, Lieff S, Silberfeld M (1989) The effect of age of onset on depression in the
elderly. Journal of geriatric psychiatry and neurology 2:182–187.
Ishizaki J, Yamamoto H, Takahashi T, Takeda M, Yano M, Mimura M (2008) Changes in
regional cerebral blood flow following antidepressant treatment in late-life depression. Int
J Geriatr Psychiatry 23:805–811.
Kalayam B, Alexopoulos Gs (1999) Prefrontal dysfunction and treatment response in geriatric
depression. Archives of general psychiatry 56:713–718.
Kiosses Dn, Alexopoulos Gs, Murphy C (2000) Symptoms of striatofrontal dysfunction con-
tribute to disability in geriatric depression. Int J Geriatr Psychiatry 15:992–999.
Kleim Ja, Bruneau R, Calder K, Pocock D, Vandenberg Pm, Macdonald E, Monfils Mh,
Sutherland Rj, Nader K (2003) Functional organization of adult motor cortex is dependent
upon continued protein synthesis. Neuron 40:167–176.
Kohn Y, Freedman N, Lester H, Krausz Y, Chisin R, Lerer B, Bonne O (2007) 99mTc-HMPAO
SPECT study of cerebral perfusion after treatment with medication and electroconvulsive
therapy in major depression. J Nucl Med 48:1273–1278.
Lesser Im, Mena I, Boone Kb, Miller Bl, Mehringer Mc, Wohl M (1994) Reduction of cerebral
blood flow in older depressed patients. Arch Gen Psychiatry 51:677–686.
Lockwood Ka, Alexopoulos Gs, Van Gorp Wg (2002) Executive dysfunction in geriatric
depression. Am J Psychiatry 159:1119–1126.
Mahncke Hw, Connor Bb, Appelman J, Ahsanuddin On, Hardy Jl, Wood Ra, Joyce Nm, Boniske
T, Atkins Sm, Merzenich Mm (2006) Memory enhancement in healthy older adults using
a brain plasticity-based training program: a randomized, controlled study. Proc Natl Acad
Sci U S A 103:12523–12528.
Martin Kc, Barad M, Kandel Er (2000) Local protein synthesis and its role in synapse-specific
plasticity. Curr Opin Neurobiol 10:587–592.
526
Matsushita K, Kuriyama Y, Nagatsuka K, Nakamura M, Sawada T, Omae T (1994)

Periventricular white matter lucency and cerebral blood flow autoregulation in hyperten-
sive patients. Hypertension 23:565–568.
Meinzer M, Flaisch T, Wilser L, Eulitz C, Rockstroh B, Conway T, Gonzalez-Rothi L, Crosson
B (2009) Neural signatures of semantic and phonemic fluency in young and old adults. J
Mies G, Ishimaru S, Xie Y, Seo K, Hossmann Ka (1991) Ischemic thresholds of cerebral protein
synthesis and energy state following middle cerebral artery occlusion in rat. J Cereb Blood
Flow Metab 11:753–761.
Milo Tj, Kaufman Ge, Barnes We, Konopka Lm, Crayton Jw, Ringelstein Jg, Shirazi Ph (2001)
Changes in regional cerebral blood flow after electroconvulsive therapy for depression. J
ECT 17:15–21.
Moody Dm, Bell Ma, Challa Vr (1990) Features of the cerebral vascular pattern that predict
vulnerability to perfusion or oxygenation deficiency: an anatomic study. AJNR Am J
Neuroradiol 11:431–439.
Morimoto Ss, Gunning Fm, Murphy Cf, Kanellopoulos D, Kelly Re, Alexopoulos Gs (2011)
Executive function and short-term remission of geriatric depression: the role of semantic
strategy. The American journal of geriatric psychiatry: official journal of the American
Association for Geriatric Psychiatry 19:115–122.
Morimoto Ss, Wexler Be, Liu J, Hu W, Seirup J, Alexopoulos Gs (2014) Neuroplasticity-based
computerized cognitive remediation for treatment-resistant geriatric depression. Nature
communications 5:4579.
Morimoto Ss, Gunning Fm, Kanellopoulos D, Murphy Cf, Klimstra Sa, Kelly Re, Jr.,
Alexopoulos Gs (2012) Semantic organizational strategy predicts verbal memory and
remission rate of geriatric depression. Int J Geriatr Psychiatry 27:506–512.
Morimoto Ss, Wexler, B. E., Alexopoulos, G.s. (2012) Neuroplasticity-based computerized cog-
nitive remediation for geriatric depression. International Journal of Geriatric Psychiatry
27:1239–1247.
Nagata R, Kawabe K, Ikeda K (2010) Olmesartan, an angiotensin II receptor blocker, restores
cerebral hypoperfusion in elderly patients with hypertension. J Stroke Cerebrovasc Dis
19:236–240.
National Advisory Mental Health Council Workgroup (2006) The Road Ahead: Research
Parterships to Transform Services. In: (National Institute of Mental Health, ed).
National Institute of Mental Health (1999) Bridging Science and Service: A Report by the
National Advisory Mental Health Council’s Clinical Treatment and Services Research
Workgroup (NIH Publication No. 99-4353). In. Rockville, MD: National Institute of
Mental Health.
Oda K, Okubo Y, Ishida R, Murata Y, Ohta K, Matsuda T, Matsushima E, Ichimiya T, Suhara T,
Shibuya H, Nishikawa T (2003) Regional cerebral blood flow in depressed patients with
white matter magnetic resonance hyperintensity. Biol Psychiatry 53:150–156.
Paranthaman R, Greenstein A, Burns As, Heagerty Am, Malik Ra, Baldwin Rc (2012)
Relationship of endothelial function and atherosclerosis to treatment response in late-life
depression. Int J Geriatr Psychiatry 27:967–973.
Paranthaman R, Greenstein As, Burns As, Cruickshank Jk, Heagerty Am, Jackson A, Malik Ra,
Scott Ml, Baldwin Rc (2010) Vascular function in older adults with depressive disorder.
Biol Psychiatry 68:133–139.
Pisljar M, Pirtosek Z, Repovs G, Grgic M (2008) Executive dysfunction in late-onset depres-
sion. Psychiatr Danub 20:231–235.
Plane Jm, Shen Y, Pleasure De, Deng W (2010) Prospects for minocycline neuroprotection.
527
Potter Gg, Kittinger Jd, Wagner Hr, Steffens Dc, Krishnan Kr (2004) Prefrontal neuropsycholog-
ical predictors of treatment remission in late-life depression. Neuropsychopharmacology
29:2266–2271.
Rabbitt P, Scott M, Thacker N, Lowe C, Jackson A, Horan M, Pendleton N (2006) Losses in
gross brain volume and cerebral blood flow account for age-related differences in speed
but not in fluid intelligence. Neuropsychology 20:549–557.
Raichle Me, Snyder Az (2007) A default mode of brain function: a brief history of an evolving
idea. Neuroimage 37:1083–1090; discussion 1097–1089.
Raichle Me, Macleod Am, Snyder Az, Powers Wj, Gusnard Da, Shulman Gl (2001) A default
mode of brain function. Proc Natl Acad Sci U S A 98:676–682.
Rajagopalan S, Brook R, Rubenfire M, Pitt E, Young E, Pitt B (2001) Abnormal brachial
artery flow-mediated vasodilation in young adults with major depression. Am J Cardiol
88:196–198, A197.
Raz N, Gunning-Dixon F, Head D, Rodrigue Km, Williamson A, Acker Jd (2004) Aging, sexual
dimorphism, and hemispheric asymmetry of the cerebral cortex: replicability of regional
differences in volume. Neurobiol Aging 25:377–396.
Raz N, Gunning Fm, Head D, Dupuis Jh, Mcquain J, Briggs Sd, Loken Wj, Thornton Ae, Acker
Jd (1997) Selective aging of the human cerebral cortex observed in vivo: differential vul-
nerability of the prefrontal gray matter. Cereb Cortex 7:268–282.
Rudrauf D, Mehta S, Grabowski Tj (2008) Disconnection’s renaissance takes shape: Formal
incorporation in group-level lesion studies. Cortex 44:1084–1096.
Saxby Bk, Harrington F, Wesnes Ka, Mckeith Ig, Ford Ga (2008) Candesartan and cognitive
decline in older patients with hypertension: a substudy of the SCOPE trial. Neurology
70:1858–1866.
Schulz R, Beach Sr, Ives Dg, Martire Lm, Ariyo Aa, Kop Wj (2000) Association between
depression and mortality in older adults: the Cardiovascular Health Study. Arch Intern
Med 160:1761–1768.
Sexton Ce, Le Masurier M, Allan Cl, Jenkinson M, Mcdermott L, Kalu Ug, Herrmann Ll,
Bradley Km, Mackay Ce, Ebmeier Kp (2012) Magnetic resonance imaging in late-life
depression: vascular and glucocorticoid cascade hypotheses. Br J Psychiatry 201:46–51.
Sheline Yi, Price Jl, Vaishnavi Sn, Mintun Ma, Barch Dm, Epstein Aa, Wilkins Ch, Snyder Az,
Couture L, Schechtman K, Mckinstry Rc (2008) Regional white matter hyperintensity bur-
den in automated segmentation distinguishes late-life depressed subjects from comparison
subjects matched for vascular risk factors. Am J Psychiatry 165:524–532.
Sheline Yi, Pieper Cf, Barch Dm, Welsh-Bohmer K, Mckinstry Rc, Macfall Jr, D’angelo G,
Garcia Ks, Gersing K, Wilkins C, Taylor W, Steffens Dc, Krishnan Rr, Doraiswamy Pm
(2010) Support for the vascular depression hypothesis in late-life depression: results
of a 2-site, prospective, antidepressant treatment trial. Archives of general psychiatry
67:277–285.
Sirey Ja, Raue Pj, Alexopoulos Gs (2007) An intervention to improve depression care in older
adults with COPD. Int J Geriatr Psychiatry 22:154–159.
Skudlarski P, Jagannathan K, Calhoun Vd, Hampson M, Skudlarska Ba, Pearlson G (2008)
Measuring brain connectivity: diffusion tensor imaging validates resting state temporal
correlations. Neuroimage 43:554–561.
Smith Ee, Salat Dh, Jeng J, Mccreary Cr, Fischl B, Schmahmann Jd, Dickerson Bc, Viswanathan
A, Albert Ms, Blacker D, Greenberg Sm (2011) Correlations between MRI white matter
lesion location and executive function and episodic memory. Neurology 76:1492–1499.
Sneed Jr, Roose Sp, Keilp Jg, Krishnan Kr, Alexopoulos Gs, Sackeim Ha (2007) Response
inhibition predicts poor antidepressant treatment response in very old depressed patients.
Am J Geriatr Psychiatry 15:553–563.
528
Steffens Dc, Taylor Wd, Denny Kl, Bergman Sr, Wang L (2011) Structural integrity of the unci-
nate fasciculus and resting state functional connectivity of the ventral prefrontal cortex in
late life depression. PLoS ONE 6:e22697.
Taylor Wd, Aizenstein Hj, Alexopoulos Gs (2013a) The vascular depression hypothesis: mech-
anisms linking vascular disease with depression. Mol Psychiatry 18:963–974.
Taylor Wd, Macfall Jr, Gerig G, Krishnan Kr (2007) Structural integrity of the uncinate fas-
ciculus in geriatric depression: Relationship with age of onset. Neuropsychiatr Dis Treat
3:669–674.
Taylor Wd, Zhao Z, Ashley-Koch A, Payne Me, Steffens Dc, Krishnan Rr, Hauser E, Macfall
Jr (2013b) Fiber tract-specific white matter lesion severity Findings in late-life depression
and by AGTR1 A1166C genotype. Hum Brain Mapp 34:295–303.
Teipel Sj, Bokde Al, Meindl T, Amaro E, Jr., Soldner J, Reiser Mf, Herpertz Sc, Moller Hj,
tivity in the human brain. Neuroimage 49:2021–2032.
Tiemeier H, Breteler Mm, Van Popele Nm, Hofman A, Witteman Jc (2003) Late-life depres-
sion is associated with arterial stiffness: a population-based study. J Am Geriatr Soc
51:1105–1110.
Tyring S, Gottlieb A, Papp K, Gordon K, Leonardi C, Wang A, Lalla D, Woolley M, Jahreis
A, Zitnik R, Cella D, Krishnan R (2006) Etanercept and clinical outcomes, fatigue, and
depression in psoriasis: double-blind placebo-controlled randomised phase III trial. Lancet
367:29–35.
Van Den Heuvel M, Mandl R, Luigjes J, Hulshoff Pol H (2008) Microstructural organization
of the cingulum tract and the level of default mode functional connectivity. J Neurosci
28:10844–10851.
Vangu Md, Esser Jd, Boyd Ih, Berk M (2003) Effects of electroconvulsive therapy on
regional cerebral blood flow measured by 99mtechnetium HMPAO SPECT. Prog
Neuropsychopharmacol Biol Psychiatry 27:15–19.
Vasile Rg, Schwartz Rb, Garada B, Holman Bl, Alpert M, Davidson Pb, Schildkraut Jj (1996)
Focal cerebral perfusion defects demonstrated by 99mTc- hexamethylpropyleneamine
oxime SPECT in elderly depressed patients. Psychiatry Res 67:59–70.
Vlassenko A, Sheline Yi, Fischer K, Mintun Ma (2004) Cerebral perfusion response to success-
ful treatment of depression with different serotoninergic agents. J Neuropsychiatry Clin
Neurosci 16:360–363.
Weiner Mf, Edland Sd, Luszczynska H (1994) Prevalence and incidence of major depression in
Alzheimer’s disease. Am J Psychiatry 151:1006–1009.
Zimmerman M, Mattia Ji, Posternak Ma (2002) Are subjects in pharmacological treatment
trials of depression representative of patients in routine clinical practice? Am J Psychiatry
159:469–473.
â•‡ 529
Author Index
Aamodt, W., 465 Adcock, R. A., 340, 341

Aartsen, M. J., 462 Addis, D. R., 113, 114, 317, 318, 324, 340,
Aasland, H. W., 131, 144 341, 344, 345
Abbamonte, M., 308 Adelsheim, S., 107, 108, 110, 117f
Abbate, R., 440 Ader, H., 23
Abdi, H., 226, 245 Adinoff, B., 27
Abduljalil, A., 394, 396 Adluru, N., 44
Abe, O., 159 Admiraal-Behloul, F., 260t, 265
Abi-Dargham, A., 424 Adolfsdottir, S., 260t, 264
Abner, E. L., 289, 491 Adolfsson, R., 377, 417, 420, 421, 424,
Abrahamowicz, M., 146 426, 429
Abrahams, S., 243 Adolphs, R., 72, 339, 340
Abrous, D. N., 280 Adriaanse, S. M., 43, 45
Achard, S., 108, 117f Aerni, A., 417, 426
Achim, A. M., 302 Afshin-Pour, B., 106, 107, 108, 117f, 226
Achten, E., 22, 23, 89 Aganj, I., 22
Acker, J. D., 15, 89, 157, 158, 224, 278, Agartz, I., 72, 157, 237, 371
303, 313, 321, 367t, 368, 369, 371, 373, Aggarwal, N., 485
377, 390, 392, 393, 395, 397, 416, 444, Aggarwal, N. T., 471, 491
461, 516 Aggleton, J. P., 276, 313, 317
Ackerman, B. P., 303, 322 Agrigoroaei, S., 365, 366, 379
Ackermann, U., 372 Aguirre, G. K., 25, 250
Acosta-Baena, N., 145, 487 Ahdidan, J., 513
Acosta-Baena, N., 487, 494 Ahern, G. L., 376, 404
Acosta, O., 42, 485, 492 Ahlström, H., 447
Adair, J. C., 223 Ahsanuddin, O. N., 516
Adalsteinsson, E., 28, 78, 80, 82, 92, 217 Aine, C. J., 223
Adam, S., 463 Aisen, P., 496
Adamczuk, K., 43, 145 Aisen, P. S., 40, 42, 48, 134, 145, 372, 487,
Adams, I., 280 490, 492, 495, 496
Adams, M. M., 278, 284 Aizenman, A., 345, 346
529
530
530 Author Index
Aizenstein H. J., 116 Almeida, O. P., 425

Aizenstein, H. J., 43, 45, 393, 395, 396, 398, Alperin, B. R., 215
449, 454, 490, 494, 513 Alpert, G. F., 212
Ajzden, I., 519 Alpert, M., 514
Akatsu, H., 48 Alsop, D. C., 24, 25, 394, 396
Akbaraly, T. N., 377, 378 Altar, C. A., 422
Akbudak, E., 78, 85, 106, 226, 446 Althage, M. C., 495
Akil, M., 424 Altmann-Schneider, I., 110
Akitsuki, Y., 311t, 312, 318 Alvarado, M., 473, 516
Akiyama, H., 37 Alvarado, M. C., 281
Akl, S. L., 454 Alvarez, G. A., 244
Akoudad, S., 395 Alvarez, P., 302
Akram, I., 79 Alvarez, S., 145, 487, 494
Akshoomoff, N., 134, 146 Alverio, J., 493
Al-Hashimi, O., 473 Alves, H., 119, 444, 445f, 449, 450, 451,
Al-Sarraf, H., 394 452, 471
Alafuzoff, I., 492, 493 Alzheimer, A., 40
Alain, C., 114, 209, 216 Amaral, D. G., 134, 146, 275f, 276, 277, 281,
Alavi, A., 37 283
Albensi, B. C., 44 Amariglio, R. E., 43, 44, 489, 490, 491, 492,
Albers, M., 493 493, 496
Albert, M. S., 161, 173, 174, 264, 286, 287, Amaro, E., Jr., 139, 514
288, 418, 420, 421, 494, 495, 513 Amenedo, E., 209, 210, 211
Albert, S., 118 Amenta, F., 393, 394, 404
Albin, R. L., 48, 447 Amer, T., 248
Alcolea, D., 42 Ames, D., 42, 45, 46, 372, 423, 485, 487,
Aldea, P., 485, 487 490, 492
Aldroubi, A., 75 Amlien, I. K., 42, 84, 134, 137, 138, 140,
Alessandrini, F., 137, 288 157, 158, 159, 237, 371
Alexander, A. L., 44, 260t, 264 Amodio, P., 215, 225
Alexander, D. C., 76 An, Y., 157, 164, 165, 172, 377, 490
Alexander, G. E., 39, 118, 145, 238, 376, Ances, B. M., 26, 27, 110, 487, 494
393, 471, 487, 493 Andel, R., 463
Alexopoulos, G. S., 6, 89, 395, 508, 509, Andersen, P., 276
510, 511, 512, 513, 514, 515, 516, 517, Andersen, S. K., 215
518, 519, 520, 521 Anderson, C. S., 520
Alger, J. R., 27 Anderson, C. V., 157
Alhainen, K., 378, 403 Anderson, H. S., 446, 447, 448, 454
Alkass, A., 280 Anderson, J. A. E., 226
Allali, G., 393 Anderson, J. R., 197, 302
Allan, C. L., 514 Anderson, K., 422
Allard, E. S., 339, 346 Anderson, K. E., 471
Allen, A. N., 146 Anderson, N. D., 87, 162, 185, 321
Allen, E. A., 105, 107, 108, 110, 117f Anderson, V., 116
Allen, H. A., 210 Andersson, J., 40
Allen, J. S., 140 Andersson, J. L., 22
Allen, P., 308 Andersson, M., 48, 50, 52, 53, 132, 137,
Allen, T. A., 275 161, 162, 163, 164, 166, 170, 174, 184,
Ally, B. A., 25, 243 185, 188, 190, 191, 240, 249, 304, 326,
Almasy, L., 159 424, 474
531
Author Index 531
Andrawis, J. P., 42 Armstrong, R. C., 76, 80

Andreasen, N., 40 Arnold, A. M., 89, 395
Andreassen, O. A., 487 Arnold, S. E., 372, 471
Andreescu, C., 116 Arns, M., 366
Andreeva, T., 428 Arnsten, A. F. T., 50, 53, 238
Andrews-Hanna, J. R., 105, 106, 108, 109, Aronen, H. J., 145
110, 116, 117f, 139, 307, 315, 472, 493 Arriagada, P. V., 485
Andrykowski, M. A., 430 Arteaga, J., 493
Angel, L., 191, 311t, 312, 314t Artegiani, B., 280
Angell, K. E., 273 Artero, S., 395
Angleman, S., 470 Arthurs, O. J., 18, 25
Angold, A., 515 Artiges, E., 146
Anguera, J. A., 473 Arvanitakis, Z., 485
Ankri, J., 377, 378 Arzberger, T., 493
Annerbrink, K., 424 Asami, T., 171
Annweiler, C., 393 Ashburner, J., 15, 157, 159
Ansado, J., 216 Ashley-Koch, A., 513
Ansari, M. S., 52 Aslan, S., 25, 450, 472, 473
Ansel, C., 26 Asllani, I., 514
Ansiau, D., 377 Aspelund, T., 380, 393, 470
Anstey, K. J., 84, 260t, 263, 393, 462 Assaf, Y., 76, 260t
Anticevic, A., 106 Asselin, C., 488
Antón-Aguirre, S., 42 Asselin, M-C., 50
Antonell, A., 43, 494 Astafiev, S. V., 107
Antonello, R. M., 404 Asthana, S., 44, 376, 420, 421
Antonenko, D., 108, 117f, 139 Atapattu, R. K., 340
Antoni, G., 156 Atiya, M., 288
Antuono, P., 450 Atkins, S. M., 516
Anzai, Y., 37 Atkinson, J., 18, 19, 25, 26, 27
Aoki, I., 493 Atri, A., 42, 161, 462, 492, 494, 495
Aoki, S., 159 Atti, A. R., 430
Apostolova, L. G., 42 Attwell, D., 135
Apparsundaram, S., 284 Atwal, J. K., 487
Appelman, J., 516 Atwood, C., 420, 421
Arai, H., 48 Au, J., 470
Arakawa, R., 52 Au, R., 377, 380, 391, 453
Arbelaez, A., 145, 487, 494 Auer, S. R., 79
Archibald, S. L., 446 Auerbach, E. J., 17, 22
Archie, J. J., 288 Auffray, E., 392, 395
Ardekani, B. A., 509, 516 Augustin, M., 23
Arean, P. A., 517, 519 Aulchenko, Y. S., 419
Arenaza-Urquijo, E. M., 108, 113, 117f, 118, Aurell, M., 389
134, 143 Aurousseau, C., 280
Arendash, G. W., 491 Aurtenetxe, S., 218, 225
Areshenkoff, C. N., 466 Austin, B. P., 396
Aribisala, B. S., 89, 260t Avery, R. A., 50, 53
Arigita, E. J. S., 135, 226 Avlund, K., 187, 187f, 379, 462, 463, 464
Ariyo, A. A., 507 Avula, R., 118, 494
Armilio, M. L., 116 Awh, E., 237
Armony, J. L., 308 Axelsson, J., 50, 52, 132, 137
532
532 Author Index
Axenovich, T. I., 419 Balsters, J. H., 215

Ayakta, N., 42 Balteau, E., 191, 311t, 312, 314t
Ayaz, M., 92 Baltes, P. B., 321, 367t, 370, 391, 473
Aycock, J., 339 Bammer, R., 214, 260t
Aylward, E. H., 134, 260t, 264, 266 Banaschewski, T., 146
Ayutyanont, N., 145, 487, 496 Bandettini, P. A., 105
Aziz, T. Z., 75 Banducci, S. E., 429, 439, 452, 454
Banerjee, S., 519
Babakchanian, S., 42 Bangen, K. J., 306t, 308, 314t, 324, 420
Babb, J. S., 15, 371 Bangert, A., 161, 306t, 314t, 324
Bacci, D., 440 Banich, M. T., 217, 441, 442
Bach, J., 161 Banner, H., 422
Bachevalier, J., 274, 281 Bano-Otalora, B., 274
Bachstetter, A. D., 491 Bansal, R., 288
Backes, W. H., 196 Bantick, R. A., 50
Bäckman, L., 5, 6, 39, 48, 50, 52, 53, 54, 80, Barad, M., 514
81, 83, 84, 88, 111, 112, 116, 118, 132, Barale, F., 308
134, 137, 145, 156, 159, 170, 174, 186, Barber, R., 447
212, 213, 239, 249, 260t, 263, 264, 266, Barch, D. M., 219, 245, 323, 430, 509, 513
303, 306t, 341, 349, 354, 367t, 368, 369, Bardell, L., 441, 442
370, 377, 415, 416, 416f, 417, 418, 420, Bargallo, N., 108, 113, 117f, 240
421, 422, 423, 424, 425, 426, 427, 428, Barkai, E., 284
429, 439, 462, 464, 465, 470, 471, 472, Barker, G. J., 85, 146
473, 474, 475 Barkhof, F., 23, 43, 45, 108, 110, 117f, 135,
Badaracco, M., 390, 391 145, 226
Baddeley, A. D., 235 Barletta, J., 156
Badre, D., 246, 321 Barnes, C. A., 277, 278, 279, 280, 281, 282,
Bahri, M. A., 39 283, 287, 289, 516
Bahri, S., 48, 493 Barnes, D. E., 403, 440
Baker, C., 354 Barnes, J., 42, 44, 492
Baker, S. L. V., 37, 40, 42, 43, 48, 50, 53, Barnes, L. L., 161, 471, 492
118, 157, 494 Barnes, W. E., 514
Bakken, R. C., 442 Barnett, A. G., 75, 191
Bakker, A., 277, 288 Barnett, J. H., 424
Bakkour, A., 42, 110, 492 Barnhart, T. E., 44
Balanos, G, M., 26 Baron, J. C., 37, 134, 157, 492
Balasa, M., 43, 494 Baron, M., 512
Baldermann, C., 246 Barr, L. L., 454
Baldo, J. V., 214, 260t Barral, S., 417
Baldwin, R., 509 Barrall, K. A., 260t, 265
Baldwin, R. C., 514 Barré, L., 37, 157
Baldwin, R. M., 50, 52, 53 Barrett-Connor, E., 453
Bales, K. R., 138, 491, 495 Barrett, L. F., 340
Ball, K., 465 Barrick, T. R., 72, 83, 174, 260t, 264,
Ballard, C., 447 303, 446
Ballard, D. J., 403 Barroeta-Espar, I., 491
Ballinger, J. R., 50 Barroso, J., 261t, 266
Ballmaier, M., 516 Bartha, R., 393
Baloh, R. W., 157 Bartlett, F. C., 461
Balota, D. A., 110, 261t, 263 Bartlett, J. W., 42, 44, 492
533
Author Index 533
Bartolomeo, P., 72 Behrens, T. W., 487

Bartres-Faz, D., 108, 113, 117f, 118, 134, Beiser, A., 377, 380, 391, 453
143, 240, 494 Bell, K. L., 471
Bartz, E. K., 25, 472, 473 Bell, M. A., 514
Bartzokis, G., 72, 76, 83, 91, 92, 259, 260t, Bella, R., 509
261t, 265, 266, 447, 472, 474 Bellec, P., 26, 27
Barulli, D., 240, 473 Belliveau, J. W., 17
Basak, C., 119, 444, 445f, 452, 471 Belsky, J., 465, 471
Basak, J. M., 37 Beltramello, A., 137, 288
Basser, P. J., 72, 73, 75, 78 Belyew, S., 495
Basser, Q., 75 Bench, C., 50
Bassett, D. S., 119 Bench, C. J., 52, 514
Bassett, S. S., 172, 288, 420 Bender, A. R., 249
Bastin, C., 39, 191, 311t, 312, 314t Bendlin, B. B., 44, 90, 137, 260t, 264, 376,
Bastin, M. E., 81, 83, 89, 92, 260t, 261, 263, 420, 421
264, 446, 447, 448, 471 Bendriem, B., 49, 50, 51f
Bateman, R. J., 41, 487, 494, 495 Benedetti, F., 308
Bates, J., 288 Benedict, C., 447
Bates, J. F., 288 Bengtsson, S. L., 472
Bates, T. C., 417 Benitez, A., 462
Batterham, P. J., 260t, 263 Benjamin, E. J., 389
Batty, G. D., 337 Benner, T., 134, 144, 279
Bauer, A. Q., 43, 52 Bennett, D. A., 161, 273, 372, 391, 417, 418,
Bauer, E., 212, 241 462, 471, 485, 491, 496
Baum, A., 379 Bennett, I. J., 39, 73, 78, 80, 81, 82, 83, 87,
Baxendale, S., 185 88, 134, 139, 140, 211, 264, 266, 287,
Baxter, L. C., 376 417, 446, 474
Baxter, M. G., 273, 281, 283 Bennett, P. J., 111
Bays, P. M., 243 Bennion, K. A., 340
Beach, S. R., 507 Bentin, S., 223
Beach, T. G., 492, 493 Benzinger, T. L., 42, 76, 80, 110, 118, 138,
Beard, C. M., 403 139, 470, 487, 491, 494, 495
Beason-Held, L. L., 132, 161, 163, 165, 165, Berch, D. B., 465
171, 377, 394, 397 Berchtold, N. C., 453
Beauchamp, N. J., Jr., 89, 447 Berg, L., 41, 485
Beauchet, O., 393 Berger, J. S., 400
Beaulieu, C., 72, 73, 76, 80, 141, 159 Bergman, S. R., 514
Beauregard, M., 281 Bergmann, O., 280
Beavers, K. M., 453 Bergstro, M., 156
Beblo, T., 212, 241 Bergstrom, Z., 223
Becker, J. A., 41, 42, 43, 44, 118, 139, 197, Bergui, M., 394
372, 485, 490, 491, 492, 493, 494 Berish, D. E., 84, 463
Becker, J. T., 146, 394, 397, 446, 447, Berk, M., 514
448, 454 Berkman, L. F., 463
Beckett, L. A., 40, 90, 134, 161, 260t, 264, Berman, K. F., 162, 239, 421
392, 496 Berman, M. G., 116, 236, 237, 247
Beckmann, C. F., 106, 107, 108, 109, 110, Bernal-Rusiel, J. L., 169, 172, 176
117f, 135, 226, 377, 421, 487, 494 Bernard, M., 146
Behl, P., 513 Bernard, S., 280
Behrens, T. E. J., 22, 75, 85, 472 Bernstein, M. A., 15, 41, 42
534
534 Author Index
Berntsen, D., 344 Bisiacchi, P. S., 215, 225

Bero, A. W., 43, 138 Biswal, B. B., 26, 106, 107, 108, 110, 116,
Berr, C., 155 117f
Berry, A. S., 216, 223 Bizon, J. L., 238, 280
Berry, E. M., 339 Bjornerud, A., 76, 80, 131, 134, 135, 139,
Berry, J. D., 389 140, 141, 142f, 143f, 144
Berry, R. W., 278, 280 Bjørnerud, A., 159, 446
Berti, V., 37 Bjornsdottir, G., 487
Bertolino, A., 422 Bjornsson, S., 380, 487
Bertram, L., 288, 420, 421, 422, 424, 425, Black, S. E., 513
427, 428, 494 Blacker, D., 42, 161, 165, 176, 264, 288,
Besançon, V., 392, 395 418, 420, 421, 492, 494, 495, 513
Bessette-Symons, B., 113, 114, 340, 341, Blair, C. K., 376, 378
344, 345, 347 Blair, S. N., 453
Bessieres, P., 463 Blamire, A. M., 398
Betensky, R. A., 42, 45, 48, 489, 490, 492, 493 Blanchard-Fields, F., 352
Bettcher, B. M., 37, 48 Blanchard, M. M., 424
Bettens, K., 43 Blangero, J., 76, 159
Beuthien-Baumann, B., 37 Blankenstein, M. A., 487, 491
Bhagat, Y. A., 80 Blanton, R. E., 516
Bhakta, M., 132, 165, 226, 307, 316, 494 Blatter, D. D., 157
Bhangale, T., 487 Blauw, G. J., 260t, 265
Bhat, V., 422 Blazey, T. M., 487, 494
Bhattacharjee, N., 452 Blennow, K., 40, 145, 485
Bhattacharyya, S., 422 Blesa, R., 42
Bherer, L., 26, 27, 473, 516 Bliss, T. V. P., 276
Bi, W., 490 Bloise, S. M., 273
Bialystok, E., 118 Blom, O., 261t
Bian, J., 462, 518 Blomqvist, G., 156
Bickel, J. F., 463, 464 Bloss, C. S., 134, 146
Bielak, A. A. M., 260t, 263, 462, 463 Bluck, S., 348, 354
Bieliauskas, L. A., 207 Blumenfeld, R. S., 302, 337, 347
Bienias, J. L., 273, 372, 392, 462, 471 Blumenthal, J. A., 79, 390, 442, 461
Bienkowska, K., 39 Boberg, M., 392, 403
Biessels, G. J., 44, 390, 392 Boccanfuso, J., 223, 473
Bigio, E. H., 492, 493 Bocklage, A., 218, 225
Bigler, E. D., 157 Bocti, C., 513
Bilgic, B., 28, 92 Bodammer, N. C., 393, 472, 474
Bilker, W. B., 516 Boegehold, M. A., 389
Binder, D. K., 422 Boellaard, R., 43, 45
Binder, J., 340 Boerwinkle, E., 376, 378
Biney, F. O., 378, 379 Boesiger, P., 24, 340
Binnewijzend, M. A. A., 43 Boeve, B. F., 37, 39, 42, 118, 485, 489, 490,
Binns, M. A., 118 491, 492, 494
Bird, T. D., 417 Bogdahn, U., 471
Birdsill, A. C., 44, 90 Bohanek, J. G., 343
Birn, R. M., 108, 110, 117f Bohbot, V. D., 422
Bisbocci, D., 394 Bohn, A., 344
Bischof, G. N., 6, 43, 239, 245, 304, 372, Bohnen, N. I., 37, 447
373, 375, 375, 377, 465, 494 Bohr, V. A., 453
535
Author Index 535
Boileau, R. A., 441, 442 Brady, T. F., 244

Bokde, A. L. W., 39, 139, 226, 514 Braff, D. L., 284
Bokde, A. W., 215 Brandel, M. G., 43
Boland, L. L., 377, 392, 393 Brandmaier, A. M., 422
Bolandzadeh, N., 446, 454 Braskie, M. N., 50, 53, 400, 424
Bollen, E. L. E. M., 260t, 265 Braun, H., 394
Bollinger, J., 113, 114, 216, 219, 223, 224, Brautigam, H., 404
225f Braver, T. S., 52, 106, 119, 219, 245, 323
Bolmont, T., 372 Bray, M. S., 376, 378
Bolton, J. L., 424 Brayne, C., 444
Boly, M., 108, 117f Brehmer, Y., 50, 53, 54, 466, 473
Bondad, A., 226, 245 Breitenstein, C., 53, 443
Bondareff, W., 278 Breiter, H. C., 308
Bondi, M. W., 25, 420, 421 Bremner, J. D., 378
Boniface, S., 18, 25 Brennan, S., 466
Boniske, T., 516 Brenner, C., 273
Bonne, O., 514 Breteler, M. M., 22, 23, 27, 89, 90, 260t,
Bonner, J., 446 261t, 264, 265, 393, 394, 403, 448, 514
Bonsang, E., 463 Bretsky, P., 418
Bontempo, D., 471 Brett, C. E., 419
Bookheimer, S. Y., 400, 418, 420, 421 Brett, F. M., 41, 487
Boomsma, D. I., 198, 380 Brewer, J., 134
Boone, K. B., 514 Brewer, J. B., 145, 173, 492
Booth, T., 260t, 263, 264 Brickhouse, M., 495
Borden, W. B., 389 Brickman, A. M., 43, 44, 79, 89, 249, 261t,
Borella, E., 466 264, 289, 366, 395, 396, 453, 487, 494,
Borghesani, P. R., 134, 260t, 264, 266 514, 516
Bories, C., 281 Brier, M. R., 494
Borjesson-Hanson, A., 429 Brierley, B., 339
Borogovac, A., 514 Briggs, S. D., 224, 303, 516
Borsboom, D., 263 Brinch, C. N., 461, 471
Bosch, B., 240, 494 Brinkley, T. E., 453
Bosma, H., 462, 463 Britton, J. C., 169, 172, 176
Bosnell, R., 72 Brizzee, K. R., 278
Boss, B. D., 277 Broadley, A. J., 514
Bostrom, E., 280 Brockmole, J. R., 243
Boudreau, R. M., 449 Brodaty, H., 157, 260t, 264
Bouix, S., 171 Broderick, J. E., 114
Bouras, C., 492, 493 Bronge, L., 23
Bourgeat, P., 42, 372, 423, 485, 487, 492 Brook, R., 514
Bowman, C. R., 311t, 315 Brookeman, J. R., 16
Boyd, I. H., 514 Brooks, A. F., 48
Boyle, P. A., 248, 418, 474, 491 Brooks, D. J., 52
Braak, E., 48, 79, 372 Brooks, K. P., 116
Braak, H., 48, 79, 372, 492, 493 Brooks, S. J., 447
Bradley, K. M., 514 Brooks, W. M., 446, 447, 448, 454
Bradley, T., 119 Brooks, W. S., 495
Bradley, W. G., Jr., 12 Broughton, M., 219
Brady, C. B., 390 Brouwer, R. M., 198
Brady, D. R., 118 Brown, A., 279, 284
536
536 Author Index
Brown, B., 487, 492 135, 138, 139, 157, 159, 162, 165, 170,
Brown, C. K., 140 175, 176, 197, 218, 226, 250, 307, 308,
Brown, C. L., 462 315, 372, 373, 446, 472, 490, 491, 492,
Brown, G. G., 420, 421 493, 494
Brown, K. J., 341, 354 Bucur, B., 88, 111, 118, 210, 260t, 390
Brown, M. W., 276, 313, 317 Budde, M., 238, 239, 304
Brown, R., 29, 49, 337, 339 Budde, M. D., 75, 76, 80
Brown, T., 41 Budge, M. M., 377, 392
Brown, T. H., 279, 280, 289 Budson, A. E., 494
Brown, T. R., 514 Bueche, C. Z., 394
Brown, T. T., 134, 140, 146 Buell, S. J., 278
Browndyke, J. N., 442 Buerger, K., 226
Brozoski, T., 49 Bugg, J. M., 443, 444, 448, 452, 470
Bruce, M. L., 519, 520 Buist, R., 44
Bruhl, R., 146 Bullain, S. S., 490
Bruhlmann, M., 72 Bullmore, E. T., 106, 107, 108, 117f,
Brumback-Peltz, C. R., 238, 239, 250 306t, 308
Brumback, C. R., 448 Bulte, D. P., 26
Brummett, B., 465, 471 Bunce, D., 84, 223, 260t, 263, 429
Brun, A., 43 Bundgaard, M. J., 137
Brun, V. H., 284 Bunge, S. A., 472
Bruneau, R., 514 Buonanno, F. S., 24
Brunner, E. J., 378 Buonocore, M. H., 279, 280, 283, 287
Brunner, F., 417, 426 Burdette, J. H., 72, 451
Brunnereau, L., 392, 395 Burdorf, A., 462, 463
Brunni-Hakala, S., 391 Burgmans, S., 83, 118, 260t, 261, 266, 394,
Bruss, J., 140 395, 396
Bruss, P. J., 367t, 369 Burgos, J., 447
Bryan, A. D., 107, 108, 110, 117f Buri, C., 287
Bryan, D., 279 Burianová, H., 114, 115f, 219, 304
Bryan, N., 395 Burke, D. M., 283
Bryan, R. N., 23 Burke, G. L., 395
Bryck, R. L., 243, 245 Burke, S. N., 278, 279, 280, 281, 516
Bu, G., 418 Burnett, B. M., 157
Buchanan, J. A., 116 Burnett, M. M., 176
Buchanan, T. W., 339, 340, 349 Burnham, S., 487
Buchel, C., 146, 471 Burnight, K. P., 367t, 368, 369
Buchholz, B. A., 280 Burns, A. S., 509, 514
Buchler, N. G., 75f, 79–80, 112, 118, 219, Burns, J. M., 446, 447, 448, 454
303, 311t, 312, 319, 323, 323f Burrows, C. L., 345
Buchman, A. S., 418 Burt, V., 389
Buchmann, A., 417, 426 Burton, H., 85
Buchner, A., 470 Burwell, R. D., 280
Buchsbaum, B. R., 226, 237, 245 Burzynska, A., 73, 78, 82, 417, 474
Buckholtz, J. W., 52 Burzynska, A. Z., 80, 415, 418, 425
Buckles, V., 494 Busa, E., 131, 173, 174, 371, 373
Buckles, V. D., 487, 495 Busatto, G. F., 105, 139
Buckner, R. L., 3, 5, 36, 39, 41, 42, 43, 44, Busch, V., 471
46, 47, 47f, 48, 52, 76, 78, 105, 106, 108, Buschkuehl, M., 470
109, 110, 116, 117f, 118, 131, 132, 134, Bush, A. I., 423
537
Author Index 537
Bustillo, J., 107, 108, 110, 117f Cappaert, N. L., 276

Butters, M. A., 116 Cappelen, T., 159
Butts, A. M., 450 Cappell, K. A., 87, 145, 162, 209, 211, 212,
Buxton, R. B., 24, 25, 26, 27, 420, 494 239, 240, 240f, 241, 241f, 242f, 245, 288,
Byrd, M., 303, 322 373, 400
Caprihan, A., 107, 108, 110, 117f
Caballero-Bieda, M., 274 Caracciolo, B., 444
Cabeza, R., 4, 5, 75f, 79–80, 84, 87, 88, Carass, A., 172
111, 112, 113, 114, 118, 139, 145, 162, Carbonell, F., 78
183, 185, 192, 208, 210, 212, 219, 236, Cardoner, N., 106
238, 239, 240, 246, 247, 260t, 266, 302, Carey, J. R., 442
303, 304, 306t, 307, 308, 309, 310, 310f, Carletti, F., 308
311t, 312, 312f, 313, 314t, 315, 316, 317, Carlson, C., 339
318, 318f, 319, 320, 321, 323, 323f, 324, Carlson, M. C., 392,
340, 341, 344, 345, 347, 349, 351, 352, Carlsson, A., 49
373, 400 Carlsson, C. M., 44, 376, 378, 396, 420
Cadaveira, F., 209, 210, 211 Carmasin, J., 42, 490, 491, 492
Caffo, B., 172 Carmelli, D., 390, 391, 392, 393, 395, 396
Caffo, B. S., 420 Carmichael, O. T., 44, 73, 78, 82, 90, 91,
Cahn-Weiner, D. A., 248 110, 138, 146, 159, 174, 260t, 264, 287,
Cai, N., 422 394, 397, 453
Caine, E. D., 507 Carmona-Iragui, M., 42
Cairns, N. J., 134, 487, 489, 492, 493, 494, Carnahan, J., 422
495 Carp, J., 114, 240, 245, 304, 374
Cairns, N., 494 Carpenter, A., 490
Calder, K., 514 Carpenter, G. M., 48
Calegari, F., 280 Carpenter, J. R., 172
Calhoun, V. D., 105, 107, 108, 110, 117f, Carretti, B., 466
118, 161, 261t, 266, 494, 495, 514 Carril, M., 495
Callahan, C. M., 403 Carrillo, M. C., 493, 496
Callahan, J. L., 418 Carro, E., 453
Callicott, J. H., 108, 113, 114, 117f, 119, Carroll, D. C., 106
162, 226, 239, 317, 324, 341, 349, 351, Carroll, S., 453
417, 422, 425, 426, 427f Carson, R. E., 213, 219
Calne, D. B. D., 52 Carstensen, L. L., 114, 116, 337, 339, 341,
Caltagirone, C., 288 344, 345, 346, 348, 349, 352, 354
Cam, C. A. N. R. T., 83 Casadesus, G., 487
Cambois, E., 378 Casanova, R., 72
Camfield, D., 212 Caselli, R., 37
Camicioli, R., 159 Caselli, R. J., 376, 404, 417, 426, 487, 496
Cammoun, L., 118 Casement, M., 25
Campbell, K. L., 108, 109, 109f, 113, 117f, Casey, B. J., 134, 146
119, 219, 248 Cash, D. M., 487
Campbell, S., 513, 517 Cashdollar, N., 218, 225
Campbell, W., 215 Cashion, D. K., 493
Cansino, S., 313, 317 Casini, A., 440
Cantone, M., 509 Caskie, G. I. L., 250
Canu, E., 137, 288 Cassel, J. C., 279, 284
Cao, M., 108, 117f Cassel, S., 279, 284
Cao, X. Y., 108, 117f Cassidy, B. S., 345, 346
538
538 Author Index
Castel, A. D., 303 Charlton, R. A., 72, 83, 114, 174, 238, 260t,
Castellani, R. J., 492, 493 264, 303, 446
Castellanos, F. X., 106, 108, 117f, Charney, D. S., 378
119, 461 Chase, W. G., 461
Castello, C., 396 Chason, J., 441, 442
Castrillon, G., 494 Chawla, M. K., 277, 279, 280, 287
Catana, C., 39 Checksfield, K., 114, 224, 225f
Catani, M., 72 Chen, A. C., 119
Cattapan-Ludewig, K., 287 Chen, C. C., 514
Cauley, S., 76 Chen, C. S., 514
Caviness, V. S., 279 Chen, G., 169, 172, 176, 493
Ceaser, A., 430 Chen, H., 488
Ceci, S. J., 471 Chen, J. J., 25
Cedazo-Minguez, A., 470 Chen, K., 145, 403, 487, 494, 496
Cella, D., 518 Chen, M., 518
Celle, S., 393 Chen, N. K., 73, 78, 82, 84, 88, 106, 108,
Celone, K. A., 161, 197, 288, 308, 314t, 117f, 139, 261, 266, 417, 474
494, 495 Chen, Q., 417, 426, 427f
Cercignani, M., 76 Chen, R., 75
Cerrato, P., 394 Chen, S. H. A., 170, 223
Cervantes, S. N., 195, 197, 304, 311t, 312 Chen, W., 29, 39
Cervenka, S., 54 Chen, W. X., 108, 117f
Cha, J. H., 518 Chen, X., 393
Chabriat, H., 19, 260t Chen, Y. L., 518
Chacko, V. P., 19 Chen, Z. J., 78
Chaddock, L., 119, 250, 444, 445f, 449, 450, Chêne, G., 168, 169, 173
451, 452, 471 Cheng, B., 424
Chadick, J. Z., 219, 222f Cheng, F. H. T., 223
Chakravarty, M. M., 146, 513 Cheng, J. C., 89, 395
Chalela, J. A., 25 Cheng, N. Y., 92
Chalfonte, B. L., 242, 284 Cheng, Y. C., 25, 27, 28, 241, 250, 325, 462
Challa, V. R., 514 Cherbuin, N., 260t, 263
Chamberlain, J. P., 107 Chesler, D. A., 17
Chamberlain, S. R., 424 Chetelat, G., 108, 110, 117f, 118, 134, 143,
Chan, G. L. G., 52 145, 372, 492
Chan, M. Y., 108, 110, 114, 117f, 118, 373, Chételat, G., 37, 39, 42, 46, 157
375, 375f Chetty, S., 40, 212
Chance, S. A., 75 Chhatwal, J. P., 43, 108, 110, 117f,
Chandra, V., 403 493, 494
Chang, C., 76, 80, 105, 119 Chiang, I. C., 514
Chang, L., 134, 146 Chiarelli, P. A., 26
Chang, M., 380 Chiba, A. A., 278
Chang, Q., 493 Chicherio, C., 249, 415, 422, 424
Chanraud, S., 216 Chien, D. T., 48, 493
Chao, L. L., 158, 216 Chiew, K. S., 247, 316, 319
Chapman, C. E., 393, 395 Chipchase, S. Y., 343
Chapman, P., 343 Chisin, R., 514
Chapman, S. B., 25, 450, 472, 473 Choi, S. H., 488
Charles, S., 339 Choi, S. J., 509
Charlson, M., 513, 517 Chonde, D. B., 39
539
Author Index 539
Chou, Y. H., 84, 88, 92, 93, 106, 108, 117f, Coenen, H. H., 52
139, 261, 266 Coffey, C. E., 512
Chouinard, M. L., 279 Cohen, A. D., 36, 43, 45, 490, 494
Christensen, H., 84, 463 Cohen, A. L., 107
Christensen, J. J., 487 Cohen, J., 45, 49, 176
Christensen, U., 379 Cohen, J. D., 251
Christian, B. T., 44 Cohen, M. S., 17, 418, 420, 421
Christie, L-A., 453 Cohen, N. J., 217, 276, 441, 442, 443, 447,
Christman, D. R., 51f 449, 452
Chua, E. F., 161, 288, 494, 495 Cohen, R. A., 249, 366, 394, 453, 516
Chuang, P-C., 453 Cohen, R. M., 421
Chuang, Y. F., 401, 401f Cohen, Y., 76
Chui, H., 518 Colantuoni, C., 286
Chui, H. C., 23, 42, 43, 112, 158, 176, 404 Colcombe, A., 441, 442
Chung, Y., 134, 146 Colcombe, S. J., 88, 119, 217, 442, 443, 444,
Churchill, N., 107, 108, 109, 109f, 111f, 114, 447, 448, 449, 452, 473, 516
117f, 226 Cole, M. W., 106, 119
Ciaramelli, E., 302, 307, 313, 321 Coleman, P. D., 278
Cifelli, G., 404 Coleman, R. E., 37, 490
Cimino, C. R., 462 Collette, F., 191, 311t, 312, 314t
Cirrito, J. R., 43, 138, 495 Collier, D. A., 422
Cisler, D. S., 119, 472, 473 Collins, D. L., 79
Claassen, D., 36 Colman, D. R., 71
Claiborne, B., 277 Colom, R., 72
Clapp, W. C., 113, 209, 216, 219, 220, 221f, Coluccia, D., 417, 426
222, 223, 225, 245, 247 Colzato, L. S., 427
Clarimón, J., 42 Compagnone, M., 223
Clark, C. A., 19, 72, 83, 85, 260t, 446 Cona, G., 215, 225
Clark, C. M., 490 Concha, L., 75, 141
Clark, J., 319 Connally, E., 349
Clark, L. R., 25 Conneally, P. M., 429
Clark, R. E., 273, 283 Connor, B. B., 516
Clark, V. H., 171 Connor, D. J., 376
Clark, V. P., 107, 108, 110, 117f Connor, J. R., 27, 91, 92
Clarke, C. L., 425 Conrod, P., 146
Clarke, E., 119, 472, 473 Constable, R. T., 105, 108, 117f
Clavaguera, F., 372 Constans, J. M., 37, 157
Clayden, J. D., 81, 83, 260t, 261, 263, 264 Conturo, T. E., 78, 85, 446
Cleary, C. A., 44 Conway, M. A., 339
Clemente, I. C., 240 Conway, T., 516
Clerici, F., 37 Conwell, Y., 507
Clifford, D., 495 Cools, R., 52
Clifford, P., 440 Coon, K. D., 376
Cliffordson, C., 461, 471 Cooney, J. W., 219, 220f
Clore, G. L., 343 Cooper, G. E., 491
Coart, E., 485 Cooper, H., 442
Coates, U., 226 Cooper, R. A., 379
Cobb, W., 223 Cooper, S. R., 81, 263
Cobiac, L., 492 Copeland, W. E., 515
Cochran, E. J., 372 Copen, W. A., 24
540
540 Author Index
Corbetta, M., 72, 105, 107, 140 Cull, T. S., 85

Corbin, K. B., 71 Culver, J. P., 43
Corder, E. H., 429 Cummings, J. L., 42, 447
Corey-Bloom, J., 420, 421 Cummins, T. D. R., 219
Corkin, S., 78, 139, 209, 339, 347, 349, 372 Cunningham, V. J., 52
Corley, J., 263, 264, 462 Cupples, L. A., 418
Cornell, J., 76 Curcio, C. A., 278
Corneveaux, J. J., 146, 417, 426 Curran, T., 209
Corrada, M. M., 490 Curtis, C. E., 243
Correia, R., 261t, 266 Curtis, W. J., 283
Correia, S., 487, 494 Czanner, S., 173, 174
Cosentino, S., 471
Costa, D. A., 491 D’Agostino, R. B., 392, 453
Costello, E. J., 515 D’Angelo, G., 118, 138, 139, 509
Costello, M. C., 111, 118 D’avanzo, C., 488
Cotman, C. W., 451, 453 D’Esposito, M., 25, 52, 53, 113, 162, 209,
Cournot, M., 377 219, 220f, 221f, 236, 237, 239, 245, 247,
Court, J. A., 279, 284 249, 250, 304, 309, 326, 400
Couture, L., 513 D’Hooge, R., 282
Cowan, N., 213, 236, 243 Daffner, K. R., 215
Cowan, R. L., 52 Daffner, R., 215
Cowan, W. M., 277 Dagher, A., 52
Cox, R. W., 169, 172, 176 Dahle, C., 15, 157, 158, 278, 303, 313, 321,
Coyle, T., 39, 260t, 261t 373, 416, 444, 461
Cracchiolo, J. R., 491 Dahlin, E., 465, 473, 474
Craft, S., 496 Dahnke, R., 158
Craig, D. W., 146, 417, 426 Dai, W., 24, 394, 397
Craik, F. I. M., 82, 111, 112, 114, 116, 118, Dai, Z. J., 108, 117f
183, 303, 304, 321, 322, 341 Dalby, R. B., 513
Crain, B. J., 19, 37, 492, 493 Dale, A. M., 42, 52, 72, 76, 78, 80, 83, 131,
Crawford, J., 157, 260t, 264 134, 135, 138, 139, 140, 141, 142f, 143f,
Crawley, A. P., 396 144, 145, 146, 157, 158, 159, 173, 174,
Crayton, J. W., 514 237, 249, 279, 288, 372, 373, 420,
Crelier, G. R., 18, 19, 25, 26, 27 446, 492
Creso Moyano, J., 319 Dalm, C., 463
Cress, M., 512 Daly, E. M., 263, 393
Crichton, R. R., 27, 91, 92 Damaraju, E., 107, 108, 110, 117f
Cristinzio, C., 420 Damasio, H., 72, 140
Crivello, F., 421 Damoiseaux, J. S., 87, 105, 107, 108, 110,
Crone, E. A., 170, 176 117f, 135, 139, 226, 514
Cronk, B. B., 447, 448 Danek, A., 495
Cross, A. H., 76, 80, 141 Daneman, M., 321
Crosson, B., 516 Dang, L. C., 106, 113, 116
Crowley, S., 43, 118, 494 Daniel, S., 223
Cruickshank, J. K., 514 Daniels, M. J., 173
Cruickshank, J. M., 398 Danker, J. F., 197, 302
Cruts, M., 377, 420, 421 Dannals, R. F., 49, 490
Cruz, L., 137 Dantzer, R., 515
Cselényi, Z., 40 Darby, D., 45, 490, 492
Csernansky, J. G., 430 Darst, B. F., 134, 146
541
Author Index 541
Das, D., 422, 423, de Frias, C. M., 250, 415, 424

Das, S., 108, 113, 114, 117f, 119, 162, 226, De Geus, E. J. C., 380
239, 317, 324, 341, 344, 349, 351, 417, de Groot, J. C., 22, 23, 89, 90, 394
422, 425, 426, 427f de Groot, M., 89, 90, 395
Daselaar, S. M., 88, 112, 145, 183, 192, 208, De Kersaint-Gilly, A., 392, 395
210, 238, 239, 302, 303, 304, 306t, 308, De Koninck, Y., 281
310, 311t, 312, 312f, 313, 314t, 316, 317, De Koning, I., 419
324, 341, 351 de La Sayette, V., 134
Daugherty, A. M., 91, 92, 93 De Laat, K. F., 448
Daulatzai, M. A., 404 de Leeuw F. E., 43
Dautoff, R., 340 de Leeuw, F. E., 22, 23, 89, 90, 394, 448
Davachi, L., 110, 302, 303, 313, 315, 324 De Leon, A. A., 226
Davatzikos, C., 23, 118, 156, 157, 161, 171 de Leon, M. J., 37, 138, 176
Davegardh, C., 429 De Luca, C. R., 116
Davey, A., 390, 391 de Luna, X., 162, 185, 364, 424
Davidsen, L., 37 De Meyer, G., 485
Davidson, N. S., 155, 186, 366, 367t, 369, De Munain, E. L., 491
370, 391 de Pasquale, F., 105
Davidson, P. B., 514 De Quervain, D. F., 417, 418, 422
Davidson, P. S. R., 303, 349 De Quervain, D. J., 340, 417, 426
Davidsson, P., 40 de Rochefort, L., 29
Davies, P., 41, 492, 493 De Rosa, E., 223
Davis, J. C., 446, 454 De Santi, S., 37
Davis, M. D., 490 De Stefano, N., 72
Davis, P. R., 289 De Weer, A. S., 43
Davis, S. W., 75f, 79–80, 83, 84, 87, 88, 111, Dean, M., 422
118, 139, 145, 188, 208, 210, 241, 246, Dear, K., 84
250, 266, 303, 304, 319, 325, 341, 351 Deary, I. I., 417
Davis, T., 195 Deary, I. J., 81, 83, 89, 92, 186, 260t, 261,
Davis, T. L., 18, 19, 25, 26, 27 263, 264, 337, 371, 397, 416, 417, 419,
Davis,M. D., 490 424, 446, 447, 448, 462, 471
Dawant, B. M., 23 Deaton, A., 114
Dawes, L. L. C., 303 Debette, S., 377, 380, 447, 453
Dayeh, T., 429 Debiec, J., 514
de Bazelaire, C., 24, 25 Debnath, M. L., 40, 156
De Beni, R., 466 Decarli, C. C., 404
de Boer, R., 23 DeCarli, C. S., 43, 146
de Brabander, J. M., 278 DeChairo, B. M., 146
De Carli, C., 39, 42, 44, 89, 90, 91, 110, Deco, G., 105
138, 139, 159, 174, 248, 249, 260t, 264, Dederen, P. J., 43
287, 289, 371, 377, 380, 393, 394, 395, Deeg, D. J., 462
403, 453 Defina, L. F., 25, 450, 472, 473
De Carli, F., 396 Deib, G., 398
de Chastelaine, M., 190, 191, 192f, 193, Deiber, M. P., 219
193f, 194, 194f, 196, 197, 198f, 306t, 308, Deibler, A. R., 72
312, 314t, 324, 325 Deisseroth, K., 119
de Craen, A. J. M., 260t, 265 Deistung, A., 92
De Deyn, P. P., 282, 485 Dekker, J. M., 392
De Dominicis, P., 404 DeKosky, S. T., 40, 41, 49, 391, 485, 490
de Fockert, J. W., 223 Del Arco, A., 50
542
542 Author Index
Del Bigio, M. R., 44 Devous, M. D., Sr., 43, 372, 405, 405f,
Del Tredici, K., 372, 492, 493 490, 494
Delacour, J., 283 Dew, I. T. Z., 112, 219, 304, 311t, 312, 314t,
Delacourte, A., 48 320, 323, 323f
Delahunt, P., 216 Dewey, S. L., 49, 50, 51f
DeLaPaz, R., 287 Di Curzio, D. L., 44
Delcroix, N., 421 Di Fabio, R. P., 442
Deldin, P. J., 116 Di Pardo, A., 404
Della Penna, S., 105 Diamond, E., 197, 288, 308, 314t
Della Sala, S., 243, 367t, 368, 369 Diana, R. A., 276
Delproposto, Z., 27 Diaz-Arrastia, R., 372, 404, 405f, 447,
Deluca, A. N., 418, 420, 421, 494, 495 449, 490
DeLuca, A. N., 161, 165, 176 Diaz-Flores, L., 261t, 266
Delvin, M. C., 225 Diaz-Ruiz, C., 404
Dempster, F. N., 208, 215, 217 Diaz, M. T., 81, 84, 266
den Heijer, T., 90, 260t, 261t, 264, 265, 393 Dichgans, M., 260t
Denburg, N. L., 339, 340 Dickerson, B. C., 42, 43, 110, 145, 161,
Dence, C. S., 41, 485 165, 173, 174, 176, 197, 264, 288, 308,
Deng, W., 518 314t, 340, 418, 420, 421, 487, 492, 493,
Denmuth, I., 250 495, 513
Denney, N. W., 367t, 369, 370 Dickson, D. W., 493
Dennis, N. A., 75f, 79–80, 111, 112, 118, Dickstein, D. L., 404
145, 162, 208, 210, 225, 240, 246, 247, Didehbani, N., 25, 450, 472, 473
303, 304, 306t, 308, 309, 310, 310f, Diehl-Schmid, J., 37
311t, 315, 316, 317, 318f, 319, 324, 341, Diggle, P., 168, 170, 176
351, 352 Ding, Y. S., 50, 52, 53
Denny, K. L., 514 Dinov, I. D., 146, 447, 448
Deouell, L. Y., 326 DiProspero, N., 286
DePeau, K., 161, 197, 288, 308, 314t, Distefano, P. S., 422
494, 495 Dixit, S., 118, 138, 139
Derby, C. A., 471 Dixon, R. A., 52, 53, 156, 250, 348,
Deriche, R., 22 367t, 368, 369, 378, 379, 428, 462,
Deroche, P. S., 280 463, 464, 471
Derosa, G., 517 Djang, D. S. W., 37
Descoteaux, M., 22 Djang, W. T., 512
Desgranges, B., 37, 39, 108, 110, 117f, 118, Dobbins, I. G., 112, 183, 192, 219, 308, 311t,
134, 143, 157 312, 312f, 313, 314t, 317, 323, 323f, 324
Desikan, R. S., 131, 371, 373, 492 Dobko, T. T., 52
Desjardins-Crepeau, L., 26, 27 Dobson, A. J., 191
Deslauriers, J., 216 Dodart, J. C., 274
Desmond, D. W., 391 Dodge, H. H., 391
Desmond, J. E., 170, 304 Dodge, J. T., 278
Desmond, T. J., 48 Doeller, C. F., 425
Desphande, A. S., 108, 110, 117f Doi, T., 444, 446
DeTeresa, R., 41 Dolan, R. J., 195, 308, 340, 514
deToledo-Morrell, L., 280, 284 Dolcos, F., 113, 238, 239, 304, 318, 318f,
Detre, J. A., 24, 25 324, 340, 341, 344, 346
Deus, J., 106 Dolcos, S., 348
Dev, S. I., 25 Dols-Icardo, O., 491
Devanand, D. P., 288 Donahue, M. J., 25
543
Author Index 543
Dong, Q., 108, 117f Duering, M., 260t

Doniger, G. M., 260t Duezel, S., 250
Donix, M., 421 Duff, E. P., 144
Donley, B. E., 191, 192f, 193, 194, 194f, 197, Duff, K., 48
198f, 312, 325 Duffin, J., 26
Donnelly, J. E., 446, 447, 448, 454 Dufouil, C., 168, 169, 173, 392, 395, 421
Donohue, M. C., 42, 145, 372, 490, 496 Duggirala, R., 159
Doraiswamy, P. M., 118, 490, 509 Dugravot, A., 155
Dore, G. A., 390, 391 Dulas, M. R., 306t, 308, 309, 311t, 312, 314t,
Dore, V., 492 315, 324, 353
Doré, V., 42 Dumas, J. A., 27
Doricchi, F., 72 Dumitriu, D., 283
Dosenbach, N. U., 107 Duncan, G. J., 470
Dosenbach, R. A., 107 Duncan, J., 83
Doshi, J., 171 Duning, K., 425
Dostrovsky, J., 282 Duning, T., 444
Douaud, G., 75, 143, 144 Dunkin, J. J., 516
Dougherty, R. F., 85, 118, 264 Dunn, A. L., 453
Douglass, K. H., 49 Dunstan, D. W., 453
Dowell, N. G., 249 Dupont, P., 43, 145
Dowling, N. M., 37, 44, 45 Dupret, D., 280
Downing, P. E., 325 Dupuis, J. H., 303, 516
Drachman, D., 475 Durbin, T., 42
Drag, L. L., 207 Durgerian, S., 450, 452
Draganski, B., 471 Durstewitz, D., 49
Drapeau, E., 280 Durston, S., 198
Drew, L., 465 Dusek, T., 260t, 264
Drewelies, J., 250 Duvall, M., 424
Driessen, M., 212, 241 Duverne, S., 190, 191, 306t, 308, 311t, 313,
Driver, J., 308 314t, 324
Dronkers, N. F., 214, 260t Duyckaerts, C., 492, 493
Druid, H., 280 Duyn, J., 105
Drzezga, A., 37, 39, 43, 139 Duyn, J. H., 24
Du, A. T., 23, 158 Duzel, E., 52, 306t, 308, 314t, 324, 472, 474
Duan, Y., 108, 110 Dye, L., 461
Duarte, A., 194, 225, 306t, 308, 309, 310, Dyson, B. J., 223
311t, 312, 314t, 315, 324, 353
Duarte, L. R., 463 Eagan, D. E., 378, 379
Dube, C., 214, 216 Earl, N., 37
Duberstein, P. R., 507 Eber, B., 260t, 264
DuBois Bowman, F., 172 Ebmeier, K. P., 155, 421, 512, 514, 516
Duchek, J. M., 110, 261t, 263 Ebner, N. C., 428
Duckwiler, G., 27 Ecker, C., 263
Duda, B., 243 Eckert, M. A., 261t, 266
Duda, J., 75 Edgerton, V. R., 453
Dudfield, M., 453 Edland, S. D., 512
Dudukovic, N. M., 288 Edvinsson, L., 397
Due-Tønnessen, P., 76, 80, 131, 134, 135, Efimova, I. Y., 398
139, 140, 141, 142f, 143f, 144, 159, 446 Efimova, N. Y., 398
Dueck, A. C., 376, 404 Egan, J., 377
544
544 Author Index
Egan, M. F., 422, 424 Erami, S. S., 50

Ehmke, R., 108, 117f Ercole, L., 487
Ehrin, E., 50 Erdfelder, E., 470
Eibich, P., 250 Erhardt, E. B., 107, 108, 110, 117f
Eichele, T., 107, 108, 110, 117f, 260t, 264 Erhart, M., 134, 146
Eichenbaum, H., 276, 280, 283, 286, 288, Erickson, K. I., 6, 88, 119, 217, 250,
302, 303, 313, 324 380, 401, 401f, 429, 439, 443, 444, 445f,
Eickhoff, S. B., 107 446, 447, 448, 449, 451, 452, 454, 471,
Eilertsen, D. E., 52, 139, 140, 157 473, 516
Einhorn, A., 512 Eriksdotter, M., 40
Eiriksdottir, G., 380, 470 Eriksson, E., 417, 424, 426, 429
Eklund, K., 88, 303 Eriksson, J., 48, 53, 83, 159, 161, 162, 163,
Ekman, P., 339 164, 166, 170, 174, 184, 185, 188, 190,
Ekstrom, A. D., 282 191, 240, 260t, 264, 266, 326
Elavsky, S., 444, 447, 448, 449 Eriksson, K. F., 429
Elbaz, A., 155 Erixon-Lindroth, N., 53
Elderkin-Thompson, V., 516 Erkinjuntti, T., 23, 145
Eldreth, D. A., 304, 401, 401f Erlandson, T. J., 442
Elfving, B., 444 Ernemann, U., 174
Elias, M. F., 377, 390, 391, 392, 453 Ernst, T. M., 134, 146
Elias, P., 453 Ersner-Hershfield, H., 116
Elias, P. K., 377, 390, 391, 392 Esiri, M. M., 75
Elizarov, A., 493 Eskelinen, M. H., 380, 470
Elkind, M. S., 395 Espeland, M. A., 451, 471
Ellermann, J. M., 17 Espeseth, T., 42, 72, 108, 110, 117f, 135,
Ellis, K. A., 42, 45, 46, 212, 372, 423, 485, 138, 157, 158, 159, 237, 260t, 264, 371,
487, 490, 492 376, 421
Ellwanger, J., 284 Esser, J. D., 514
Elmaleh, D. R., 485 Etchamendy, N., 422
Elman, J. A., 43, 118, 494 Etcoff, N. L., 308
Elmenhorst, D., 52 Etkin, A., 119
Elvevåg, B., 421 Etnier, J. L., 440
Emch, J., 219 Eulitz, C., 516
Emery, M., 417, 421, 426, 427f Eustache, F., 37, 39, 108, 110, 117f, 118,
Emsley, C. L., 392 134, 143, 157
Emsley, H. C. A., 217 Evans, A. C., 78, 79, 174, 198, 393, 461
Endo, H., 444 Evans, D. A., 161, 391, 391, 462, 471, 491
Eng, E., 490, 491 Evans, S., 249
Engel, A. K., 140 Evensmoen, H. R., 108
Engelborghs, S., 485 Eyler, L. T., 306t, 308, 314t, 324, 420, 421
Engelman, C. D., 376 Ezekiel, F., 23
Engler, H., 156
Englund, E., 43 Fabi, K., 243
Engvig, A., 76, 80, 131, 135, 139, 140, 141, Fabiani, M., 210, 214, 218, 238, 239, 240,
142f, 143f, 144, 159, 446, 471, 472 245, 250
Ennabil, N., 216 Fagan, A. M., 37, 110, 118, 138, 139,
Ennis, D. B., 76 145, 470, 485, 487, 489, 491, 492, 494,
Eppinger, B., 211 495, 496
Epstein, A. A., 72, 513 Fagan, S. C., 398
Epstein, R., 325 Fagerlund, B., 187, 187f, 379
545
Author Index 545
Fain, S., 15 Ferrell, R. E., 429, 439, 452

Fair, D. A., 107 Ferrer-Caja, E., 263
Fairchild, A. J., 266, 267 Ferrie, J. E., 155, 377, 378
Faisal, A. A., 461 Ferrieres, J., 377
Falcon, C., 240 Ferris, S. H., 176, 490, 496
Fan, F. M., 108, 117f Ferrucci, L., 157, 165, 171, 377, 490
Fan, Q., 76 Festini, S. B., 239, 248
Fandakova, Y., 83, 118, 224, 260t, 261, 266, Festini, S. L., 6
393, 394, 396 Feuerstein, R., 346
Faraji, F., 472 Ffytche, D. H., 72
Farde, L., 40, 50, 52, 53, 54, 156, 303, 416 Fields, R. D., 141, 471, 472, 474
Fardo, D. W., 491 Figiel, G. S., 512
Farias, S. T., 90, 248, 249, 264 Figueroa, C. M., 452
Farlow, M., 495 Figurski, J., 116
Farovik, A., 288 Filbey, F. M., 27, 107, 108, 110, 117f
Farrar, A. M., 487 Filippi, M., 174
Farrazano, P., 289 Filippini, N., 106, 377, 421, 487, 494
Farrell, L. A., 518 Filley, C. M., 72
Farrell, M. A., 41, 487 Fink, D. J., 37
Farrer, L. A., 418 Fink, G. R., 212, 306t, 309, 311t, 313, 314t
Faul, F., 470 Fink, K. B., 518
Faure, S., 216 Finkel, D., 367t, 370, 415, 417, 463
Faux, N. G., 492 Finn, J. P., 260t, 265
Fava, M., 507 Finn, M. B., 138, 495
Fawaz, M., 48 Finnigan, S., 219
Fawaz, M. V., 48 Firbank, M. J., 398
Fazekas, F., 23, 92, 260t, 264 Fischer, F. U., 82
Fearnley, J. M., 49 Fischer, H., 50, 53, 54, 111, 112, 116, 118,
Feczko, E., 42, 492 306t, 341, 349, 354, 428
Feder, M., 512 Fischer, K., 514
Federspiel, A., 287 Fischl, B., 42, 52, 72, 78, 131, 134, 139, 140,
Fegen, D., 237 144, 157, 159, 169, 172, 173, 174, 176,
Feige, B., 246 237, 264, 372, 373, 492, 513
Feinberg, D. A., 17, 22 Fischman, A. J., 41, 485
Feiwell, R. J., 107 Fisher, J. A., 26, 396
Felder, C., 50 Fitzgerald, M. E., 137, 260t, 264, 420, 421
Feldstein Ewing, S. W., 107, 108, 110, 117f Fitzmaurice, G. M., 169, 170, 172, 173, 176
Feldstein, C. A., 404 Fivush, R., 343
Fellgiebel, A., 39, 82, 217 Fjaer, S., 260t, 264
Fennema-Notestine, C., 134, 145, 173, Fjell, A. M., 3, 23, 41, 42, 52, 71, 72, 76, 80,
249, 446 84, 85, 131, 134, 135, 137, 138, 139, 140,
Fenstermacher, E., 288 141, 142f, 143, 143f, 144, 145, 146, 157,
Fera, F., 162, 239, 341, 344 158, 159, 159f, 161, 173, 174, 197, 237,
Ferencz, B., 416, 420, 429, 470 261t, 371, 376, 446, 471, 472
Ferguson, S. A., 349 Fladby, T., 159
Fernandes, M., 349 Flaisch, T., 108, 117f, 516
Fernandez-Carballo, L., 491 Fleck, M. S., 112, 145, 183, 192, 208, 210,
Ferrante, R. J., 518 246, 304, 308, 311t, 312, 312f, 313, 314t,
Ferreira D. Molina, Y., 261t, 266 317, 324, 341, 351
Ferreira, L. K., 105, 139 Fleischman, D. A., 273
546
546 Author Index
Fleisher, A. S., 26, 27, 145, 403, 420, 421, Fouquet, M., 108, 110, 117f, 134
487, 490, 494, 496 Fowkes, F. G. R., 424
Flensborg-Madsen, T., 379 Fowler, J. S., 39, 49, 50, 51f, 52, 53
Fletcher, E., 89, 90, 91, 110, 138, 159, 174, Fox, H. C., 186
264, 287, 453 Fox, M. D., 39, 105, 107, 510, 511
Fletcher, J. M., 185 Fox, N. C., 42, 44, 157, 487, 492, 494, 495
Fletcher, P. C., 306t, 308 Fox, P., 39, 76
Flicker, B., 25, 27, 241, 250, 325 Fox, P. M., 106, 107
Flicker, L., 425 Fox, P. T., 106, 107, 159, 260t, 261t
Fling, B. W., 118 Fozard, J. L., 259
Flitter, M., 490 Frackowiak, R. S., 15, 157, 159, 185, 190,
Flöel, A., 53, 108, 117f, 139, 424, 443, 444 306t, 308, 324, 514
Flood, D. G., 278 Franchow, E., 170
Flor, H., 146 Frandsen, J., 513
Flory, J. D., 429, 439, 452 Frank, J. A., 24, 75
Flory, J. M., 452 Frank, L. R., 24, 76
Flynn, J., 471 Frank, M. J., 49, 52
Flynn, J. R., 461 Franke, D., 107
Fodero-Tavoletti, M. T., 48, 49 Frankle, W. G., 424
Fogari, E., 517 Franssen, E. H., 79
Fogari, R., 517 Fratiglioni, L., 39, 48, 81, 83, 84, 134, 159,
Fokin, V., 428 174, 260t, 263, 264, 266, 391, 392, 416,
Foley, D., 377, 391, 393, 403 418, 420, 425, 430, 444, 453, 472, 475
Foley, J. M., 137 Frazier, J., 134, 146
Folkow, B., 389 Fredieu, A., 27
Folsom, A. R., 376, 378 Freeborough, P. A., 157
Fonds, H., 377 Freedman, M., 118
Fong, S. G., 283 Freedman, N., 514
Fong, T., 25 Freeman, S. H., 137
Ford, C. C., 107, 108, 110, 117f Freidl, W., 260t, 264
Ford, G. A., 391, 398, 518 Frenneaux, M. P., 514
Ford, J. H., 5, 340, 349, 351, 352, 353, 354 Frensch, P. A., 425
Ford, J. M., 114 Freunberger, R., 220
Foreman, N., 283 Freund-Levi, Y., 40
Foroud, T., 146 Frey, K. A., 48
Foroud, T. M., 417 Frey, M., 490
Forsberg, A., 40 Friberg, P., 389
Forsgren, L., 52 Fried, L. P., 395, 401, 401f
Forsman, L., 261t, 472 Friedlander, R. M., 518
Forssberg, H., 472 Friedman, D., 284, 367t, 369
Förster, S., 39 Fries, J., 107, 108, 110, 117f
Forsyth, J. K., 174 Friese, M., 340
Fortea, J., 42, 494 Friesen, W. V., 339
Fortin, N., 288 Frings, L., 171
Fortin, N. J., 275 Fripp, J., 42, 485, 492
Foster, N. L., 37 Frisen, J., 280
Foster, T. C., 238 Frishe, K., 490
Fotenos, A., 157, 159 Frisoni, G. B., 137, 288, 377, 421, 487, 494
Fotenos, A. F., 41, 43, 118, 491 Friston, K. J., 15, 39, 157, 159
Fougnie, D., 244 Fritz, M. S., 266, 267
547
Author Index 547
Friz, J., 421 Gao, F. Q., 513

Frölich, L., 37 Gao, S., 392, 403, 462
Frosch, M. P., 137, 491, 492, 493 Garada, B., 514
Frost, C., 42 Garavan, H., 146, 217, 473
Frost, J. J., 49, 54 Garcia-Barrera, M. A., 466
Frouin, V., 146 Garcia-Munoz, A., 41, 487
Frye, S., 420 Garcia, C. I., 91
Fryxell, K. J., 428 Garcia, D., 24
Fu, H., 491 Garcia, G., 145, 487
Fu, S. M., 211 Garcia, J. H., 394
Fuchs, E., 278 Garcia, K. S., 509
Fuhrer, R., 463 Garcia, M. E., 453
Fujii, K., 397 Garde, E., 174
Fujikawa, A., 39 Garden, S. E., 367t, 368, 370
Fujishima, M., 397 Gardener, H., 395
Fukuda, H., 39, 54, 157, 263, 392, 393 Gardner, C. M., 25, 472, 473
Fukuda, K., 218, 225 Gardner, E. D., 71
Fuld, P., 41 Garner, J., 393, 395
Fuqua, A., 495 Garofalo, J. P., 379
Furberg, C. D., 89 Garoff-Eaton, R. J., 343, 344
Furey, M. L., 39, 118, 325 Garraux, G., 53
Furst, A. J., 40, 212 Garrett, D. D., 84, 119
Furtak, S. C., 279, 280 Garvin, E., 343
Furumoto, S., 48, 493 Garz, C., 394
Fusar-Poli, P., 159, 308 Gaser, C., 158, 471
Gaskell, P. C., 429
G., M., 279 Gasparotti, R., 308
Gabrieli, J. D., 198, 273 Gasper, K., 343
Gabrieli, J. D. E., 36, 214, 260t, 304 Gass, A., 75, 144
Gach, H. M., 394, 397, 446 Gatley, J., 52, 53
Gage, F. H., 280, 444 Gatley, S. J., 50, 51f
Gagoski, B. A., 17 Gatz, M., 367t, 370, 463
Gainer, H., 37 Gaudet, D., 146
Galasko, D. R., 281, 490 Gaunt, J. T., 354
Gale, S. D., 157 Gaupp, R. E., 513
Gallagher, C. L., 44, 376 Gauthier, C. J., 26, 27
Gallagher, M., 274, 278, 279, 280, 281, 282, Gauthier, S., 444
283, 284, 286, 287, 288 Gazes, Y., 304
Gallagher, P., 263, 266 Gaziano, J. M., 390
Gallhofer, B., 212 Gazzaley, A., 4, 113, 114, 173, 176, 207,
Gallinat, J., 146, 287 209, 213, 214, 215, 216, 218, 219, 220,
Gallo, D. A., 195, 197, 304, 311t, 312, 349 220f, 221f, 222, 222f, 223, 224, 225, 225f,
Galloway, T. A., 461, 471 226, 245, 247, 326, 472
Gambaro, M., 396 Gchwendtner, A., 260t
Gamo, N. J., 238 Ge, Y., 15, 371
Gamst, A. C., 446 Gebhardt, H., 212, 241
Gandy, S., 404 Geda, Y. E., 485, 490
Gangadharmath, U., 493 Gee, M., 159
Ganguli, M., 391 Geerligs, L., 108, 110, 111, 117f, 210
Ganzola, R., 288 Geerlings, M. I., 264, 397, 399f
548
548 Author Index
Geinisman, Y., 278, 280, 284 Giorgio, A., 72

Geng, J., 90, 260t Giovanello, K., 248
Geng, X., 146, 472 Giovanello, K. S., 281, 308, 311t, 312, 324
Genon, S., 191, 311t, 312, 314t Giraldo, M., 145, 487, 494
Gensini, G. F., 440 Girton, L., 157, 159
Gentil, C., 463 Gitcho, A., 37
Gentile, A., 390, 391 Gjedde, A., 49
Gentile, M. T., 404 Gjerstad, L., 159
George, A. E., 176 Glabus, M. F., 514
Gerbaldo, H., 40 Glahn, D. C., 76, 106, 107, 159
Gerber, J. D., 146 Glascher, J., 72
Gerhardt, G. A., 284 Glasser, M. F., 17, 22, 141
Gerig, G., 514 Glatt, C. E., 510
Germine, L. T., 366, 367t, 370 Gleason, C. E., 37, 45, 376
Gerritsen, L., 416, 429 Glisky, E. L., 238, 303
Gersing, K., 509 Glodzik, L., 37, 138
Gerstorf, D., 15, 157, 158, 250, 278, 303, Glover, G. H., 25, 26, 105, 106, 107, 119,
313, 321, 373, 416, 422, 444, 461, 462 288, 304
Geschwind, N., 72 Gluck, M. A., 319
Geurts-van Bon, L., 43 Glymour, M. M., 155, 168, 169, 173,
Geyer, M. A., 284 463, 471
Gharapetian, L., 83 Glynn, N. W., 447, 449
Ghebremedhin, E., 372 Glynn, R. J., 391
Ghetti, B. F., 487, 494, 495 Gmeindl, L., 239, 240, 240f, 241, 241f,
Ghisletta, P. P., 48, 53, 84, 118, 155, 188, 242f, 245
237, 266, 371, 422, 444, 462, 463, 464 Go, A. S., 389
Gholkar, A., 447 Goate, A. M., 37, 118, 138, 139, 470, 487,
Ghosh, P. M., 37, 48, 118 494, 495
Ghosh, S., 519 Gobbini, M. I., 325
Gianaros, P. J., 393, 395, 396, 398, 447, 453 Gobin, Y. P., 27
Giannakopoulos, P., 219 Goddard, N. H., 277
Gibbons, L. E., 462 Godde, B., 218, 444
Gibbs, S. E., 52 Godeau, C., 37, 157
Gidicsin, C., 490, 491 Goessler, W., 92
Giedd, J. N., 79, 365, 461 Goh, J. O., 145, 164, 165, 218, 304, 373
Gifford, K. A., 390, 391 Golay, X., 24, 25
Gigandet, X., 118 Gold, B. T., 82, 83, 260t, 266, 422, 446
Giggey, P. P., 391 Gold, E., 75
Gilbert, P., 281 Gold, G., 219
Gilbert, P. E., 286 Goldberg, I. E., 17
Gilewski, M. J., 308 Goldberg, T. E., 162, 239, 317, 324, 341,
Gill, B., 18, 19, 25, 26, 27 349, 351, 422, 424
Gillette, J. A., 39, 90, 372 Goldman-Rakic, P. S., 49, 236, 281
Gilley, D. W., 471 Goldman, D., 422, 424
Gillis, J., 146 Goldman, P., 49
Gilmore, A. W., 107 Goldszal, A. F., 156, 157, 161
Gilmore, G. C., 321 Golomb, J., 176
Gilmore, J. H., 146 Golski, S., 156, 157, 161
Ginovart, N., 52, 53, 156 Goltsov, A., 428
Ginsberg, S. D., 284 Gomez-Isla, T., 491, 493
549
Author Index 549
Gómez-Pinilla, F., 453 Graham, K. S., 310, 311t, 312, 314t

Gomez, L. F., 493 Graham, R. R., 487
Gomez, M. G., 145, 487 Grambaite, R., 159
Gomperts, S. N., 485 Grande, L. J., 137, 393, 395
Gong, G., 78 Grandjean, J., 43, 44
Gonik, M., 260t Grant, E. A., 41, 485, 489, 491, 492
Gons, R. A., 448 Grasby, P. M., 50, 52
Gonzales-Aviles, A., 39, 90, 372 Grassiot, B., 421
Gonzales, M. M., 378, 379 Gratton, C., 210
Gonzalez-Atavales, J. B., 25 Gratton, G., 214, 218, 238, 239, 240, 245,
Gonzalez-Castillo, J., 105 448
Gonzalez-Rothi, L., 516 Gray, S. J., 195, 197, 304, 311t, 312
Gonzalez, R. G., 24 Green, A. E., 42
González, S., 42 Green, R. C., 134
Good, C. D., 15, 157, 159, 185, 190, 306t, Greenberg, S. M., 41, 264, 513
308, 324 Greene, E., 279, 287
Good, N., 46 Greengard, P., 49
Goodwin, G. M., 377, 487, 494, 512, Greenhouse, I., 223
514, 516 Greenia, D. E., 490
Goodwin, J. A., 217 Greenop, K. R., 425
Goossens, L., 465 Greenstein, A. S., 514
Gordinou de Gouberville, M. C., 43 Greenwood, P. M., 72, 119, 211, 376, 428,
Gordon, B. A., 214, 238, 239, 250, 394, 448 472, 473
Gordon, E. M., 105, 249, 341, 354, 366, 516 Greer, P., 400
Gordon, K., 518 Greer, P. J., 393
Gore, J. C., 25 Gregas, M., 41
Gorelick, P. B., 23 Gregg, B. E., 43, 47, 487, 494
Gorno-Tempini, M. L., 37, 48, 118 Greicius, M. D., 43, 85, 87, 105, 106, 107,
Göthberg, G., 389 118, 139, 494, 514
Gothe, N., 119 Greve, D., 43, 139
Goto, R., 157, 393 Greve, D. N., 42, 72, 78, 131, 134, 144,
Gottlieb, A., 518 157, 159, 169, 172, 176, 237, 288, 371,
Gottlob, L. R., 209 373, 492
Gough, K., 27 Grgic, M., 516
Gouw, A. A., 260t, 261t Grieshofer, P., 260t, 264
Govoni, R., 321 Grieve, S. M., 79, 249, 261t, 264, 453
Gow, A. J., 81, 83, 92, 260t, 263, 264, 371, Griffith, E. Y., 43
419, 446, 447, 448, 462, 463, 464, 471 Grigg, O., 106, 108, 109, 109f, 110, 111f,
Goyal, M. S., 141 114, 117f
Grabowski, T. J., 108, 110, 117f, 514 Grill-Spector, K., 170
Grace, A. A., 284 Grill, J. D., 278
Grady, C., 3, 5, 106, 118, 145, 418, 421 Grimm, K. J., 378, 379, 463, 464
Grady, C. L., 39, 84, 90, 106, 107, 108, 109, Grimmer, T., 39
109f, 110, 111, 111f, 112, 113, 114, 115f, Grodstein, F., 42, 492
117f, 118, 119, 165, 166f, 209, 213, 216, Groenewegen, H. J., 281
219, 226, 239, 247, 248, 304, 316, 319, Grol, M. J., 170, 176
321, 326, 341, 349, 372, 400 Gron, G., 496
Graff-Radford, N., 417 Gronenschild, E. H., 83, 118, 260t, 261, 266,
Graff, C., 416, 420, 425, 430, 470 394, 395, 396
Grafman, J., 340, 344 Gronlund, S. D., 339
550
550 Author Index
Groop, L., 429 Gusella, J. F., 429

Gross, A. L., 466 Gusnard, D. A., 106, 197, 226
Gross, D. W., 141 Gustafson, A., 487
Gross, J. J., 346, 352 Gustafson, D., 429, 453
Gross, M. S., 273 Gustafsson, J., 461, 471
Grossman, R. I., 15, 371 Gutchess, A. H., 161, 306t, 314t, 319, 324,
Grossmann, I., 367t, 370 340, 345, 346
Group, A. A. S., 490 Gutierrez, A., 284
Group, A. R., 423, 490, 492 Gutierrez, R., 211
Groves, A. R., 143, 144 Guttmann, C. R. G., 260t, 261t
Growdon, J. H., 42, 339, 485, 492 Guzman, V. A., 43, 44
Gruen, J. R., 134, 146 Guzowski, J. F., 277
Grühn, D., 339 Gyldensted, C., 24
Grundman, M., 490
Gruner, W., 107, 108, 110, 117f Haacke, E. M., 27, 28, 92, 93
Grunwald, M., 215 Haan, M., 37, 453
Grydeland, H., 23, 42, 76, 80, 85, 131, 134, Haase, C. M., 42, 489, 492
135, 138, 139, 140, 141, 142f, 143f, 144, Haasz, J., 260t, 264
158, 159, 446 Habak, C., 219
Gu, Q., 389 Habeck, C. G., 42, 106, 113, 283, 304, 396,
Guallar, E., 447 471, 514
Gualtieri, C. T., 516 Habedank, B., 135
Gudbjartsson, D., 487 Haber, S., 465
Gudnason, V., 380, 470 Haberkamp, A., 212
Guez, J., 281, 303 Haberman, R. P., 280, 281, 286
Guillozet, A. L., 372 Habibi, A., 191, 306t, 311t, 313, 314t
Guitton, M. J., 281 Hachinski, V., 470
Gujral, S., 452 Hachinski, V. C., 394
Gundapuneedi, T., 447, 449 Hackett, M. L., 520
Gundersen, H. J., 137, 278 Haddar, D., 27
Gunderson, E., 453 Hafkemeijer, A., 110
Gungor, N. Z., 246 Hafsteinsdottir, S. H., 470
Gunn, R. N., 52 Hagger-Johnson, G., 378
Gunnar, M. R., 379 Hagler, D. J., Jr., 83, 134, 145, 146, 173, 249
Gunning-Dixon, F. M., 22, 89, 90, 341, 392, Hagmann, P., 118, 448
395, 509, 510, 516 Hahn, S., 339, 441, 442
Gunning, F. M., 303, 510, 511, 516 Haight, T. J., 44
Gunstad, J. J., 366, 453, 516 Haines, J. L., 418, 429
Gunter, J. L., 37, 39, 41, 42, 43, 46, 47, 145, Hajjar, I., 394, 396, 518
485, 487, 489, 491, 492, 494 Hake, J. L., 442
Gunther, M., 17, 22, 25 Halder, I., 429
Guo, L., 518 Halder, L., 439, 452
Guo, X., 429 Hale, L. A., 212
Guo, Z., 391, 392 Hale, S., 82, 242, 260t, 264, 266
Gupta, M., 23 Haley, A. P., 378, 379
Gupta, N., 280 Hall, C. B., 471
Gur, R. C., 50, 52, 53, 341 Hall, E., 429
Gur, R. E., 37, 53, 341, 516 Hall, H., 50
Guralnik, J. M., 418, 447 Hall, K. S., 392, 462
Gurevicius, K., 280 Hallbäck-Nordlander, M., 389
551
Author Index 551
Halldin, C., 40, 52, 53, 156 Harris, T. B., 380, 449, 453, 470
Hallgren, B., 91–92 Harrison, B. J., 106
Halliday, G. M., 89 Harrison, C. R., 441, 442
Halligan, E. M., 45, 490 Harrop, R., 37
Hallikainen, M., 378, 403 Harsch, N., 339
Hama, H., 280 Harsha, A., 446, 447, 448, 454
Hamagami, F., 463 Hart, J. J., Jr., 25, 472, 473
Hamann, S., 340 Hart, M., 518
Hamberger, M., 367t, 369 Hartikainen, P., 157
Hamer, R. M., 146 Hartley, A., 304
Hamilton, R. L., 40 Hartley, A. A., 209
Hampel, H., 139, 174, 226, 487, 514 Hartline, D. K., 71
Hampson, M., 105, 108, 117f, 514 Hartshorne, J. K., 366, 367t, 370
Han, H., 349 Hartzell, A. L., 280
Han, J., 26 Harvey, D., 37, 89, 90, 134, 248, 264
Han, J. S., 396 Hasher, L., 113, 116, 208, 215, 217, 219,
Han, L. Y., 372 244, 245, 247, 248, 284, 316, 319, 339
Han, S. D., 420, 421 Hashimoto, R., 422
Han, X., 173, 174 Hashtroudi, S., 273, 349
Hancock, P. A., 259 Hasselmo, M. E., 284
Handwerker, D. A., 105 Hassenstab, J., 489, 492, 494
Hanggi, J., 417, 426 Hatazawa, J., 325, 393
Hanks, S. D., 278 Hauenstein, K. H., 174
Hänninen, T., 377, 403 Haughton, V. M., 107
Hansen, N. L., 187, 187f Hauser, E., 513
Hanson, M., 37 Havlicek, M., 107, 108, 110, 117f
Hanson, R. J., 450 Havlik, R. J., 377, 391, 393, 403
Hansson, L., 392 Hawellek, D. J., 140
Hansson, O., 485 Hawk, T. C., 210
Hantke, N. C., 450 Hawkins, K. A., 418
Hao, J., 283 Hawkins, R. D., 282
Hao, X., 288 Haxby, J. V., 39, 90, 213, 219, 304, 321, 325,
Hara, Y., 238, 280, 283 372, 393
Harada, R., 48 Hay, J. F., 339
Hardies, J., 260t, 261t Hayama, H. R., 302
Hardy, J., 216, 372, 487 Hayasaka, S., 451
Hardy, J. L., 516 Hayashi, M., 52
Harel, N., 22 Hayashi, N., 173, 174, 176
Hargreaves, E., 282, 287 Hayenga, A. O., 43, 118, 491, 494
Haring, A. E., 215 Hayes, A. F., 83
Hariri, A. R., 162, 239, 341, 344, 422 Hayes, S. M., 88, 112, 113, 114, 303,
Haroutunian, V., 492, 493 306t, 309, 310f, 311t, 317, 318f, 324,
Harper, S., 196 341, 344, 345
Harrington, F., 391, 518 Hazlett, K. E., 452
Harrington, K., 45, 423, 490, 492 He, B. J., 72
Harris, A., 325 He, J., 110, 138, 264, 398
Harris, G., 401, 401f He, P., 43, 44
Harris, J. J., 135 He, X., 108, 110
Harris, K. C., 261t, 266 He, Y., 78, 108, 112, 114, 117f, 209, 260t,
Harris, S. E., 417, 419, 424 264, 319
552
552 Author Index
Head, D., 15, 43, 78, 108, 110, 116, 117f, Hendrikse, J., 25
118, 139, 157, 158, 261t, 263, 278, 303, Henerson, R. D., 263, 264
313, 321, 373, 392, 393, 416, 443, 444, Henke, K., 417, 426
446, 448, 452, 461, 470, 472, 490, 492, Henkenius, A. L., 15, 26, 157
494, 516 Henson, R. N., 15, 83, 157, 159, 188, 195,
Head, E., 289 241, 250, 310, 311t, 312, 313, 314t, 317,
Head, R., 485, 492 325, 337
Heagerty, A. M., 514 Heo, M., 508
Healey, M. K., 113 Heo, S., 119, 219, 444, 445f, 449, 452, 471
Healy, G. N., 453 Herholz, K., 37, 135
Hearons, P., 225 Herlitz, A., 416, 420, 429
Heatherton, T. F., 340, 346 Herman, M. M., 424
Hebert, L. E., 391, 391, 491 Hermann, B. P., 90, 420
Hebisch, M., 488 Hernandez-Garcia, L., 25
Hebrank, A. C., 25, 26, 27, 43, 188, 189f, Hernandez-Ribas, R., 106
195, 239, 240, 306t, 314t, 316, 316f, 319, Hernandez, M. C., 22, 81, 83
325, 326, 367t, 368, 370, 372, 374f, 375, Hernández, M. C., 447, 448
403, 465, 490, 494 Herpertz, S. C., 139, 514
Heck, A., 425 Herrera, S. L., 44
Hedden, T., 3, 36, 39, 42, 43, 44, 45, 46, 46f, Herrmann, L. L., 512, 514, 516
47, 47f, 48, 52, 108, 110, 117f, 118, 134, Herron, J. E., 313, 317
135, 155, 161, 174, 186, 197, 218, 306t, Herrup, K., 145
314t, 324, 366, 367t, 369, 370, 372, 391, Hersch, S. M., 496, 518
489, 490, 492, 493, 494 Herscovitch, P., 23, 213, 219
Hedehus, M., 21, 78, 82 Hertzog, C. C., 48, 53, 84, 155, 188, 266,
Hedley-Whyte, E. T., 137 367t, 368, 369, 462, 463, 465
Hedman, A. M., 158, 174 Herve, D., 260t
Heekeren, H. R., 80, 212, 213, 214, 239, 249, Herzig, M. C., 372
415, 422, 424, 425, 427, 428 Hess, T. M., 352
Heidrich, S. M., 367t, 370 Hesse, C., 40
Heijboer, R., 22, 23, 89 Hesselink, J. R., 446
Heim, A. F., 393, 397 Hevelone, N. D., 78, 139, 372, 496
Heim, C., 379 Hibar, D. P., 146
Heinrichs, D., 495 Higuchi, M., 493
Heinz, A., 146 Hillman, C. H., 250, 380
Heinze, H. J., 52, 306t, 308, 314t, 324, 339, Hillmer, A. T., 44
425, 472, 474 Hills, A., 209
Heise, V., 421 Hilton, J., 471
Heishman, S. J., 461 Himali, J. J., 377, 380, 453
Heiss, W. D., 135 Hindin, S. B., 442
Helkala, E. L., 378, 380, 403, 453 Hinds, J. W., 278
Hellwig, B., 246 Hipp, J. F., 140
Helmer, K. G., 159 Hirsch, C. H., 89
Helmers, K. F., 465 Hiscock, M., 185
Helpern, J. A., 22 Hitch, G., 235
Helson, R., 366, 370 Hitzemann, R. J., 50, 51f, 52, 53
Helzner, E. P., 471 HJ, L., 280
Hemby, S. E., 284 Ho, A. J., 146, 446, 447, 448
Henderson, A. S., 463 Ho, H. V., 444
Hendrie, H. C., 392, 403, 462 Ho, L., 404
553
Author Index 553
Hobbs, W., 518 Hou, L., 24

Hockley, B. G., 48 Houck, P. R., 490
Hodge, S. M., 279 Hough, M. G., 377, 487, 494
Hodneland, E., 260t, 264 Houle, S., 304
Hodzic, J. C., 452 House, A., 520
Hoerndli, F. J., 417, 426 House, M. J., 92
Hof, P. R., 283, 404, 492, 493 Houston, F. P., 279
Hofer, S. M., 84, 184, 462 Houston, W. S., 420, 421
Hoffman, B. M., 442 Houtekier, S. C., 398
Hoffman, E. J., 36 Houx, P., 462
Hoffman, J. M., 37, 210 Houx, P. J., 462, 463
Hoffman, K. L., 277 Howard, D. V., 80, 81, 266
Hofman, A., 22, 23, 27, 89, 90, 260t, 261t, Howard, G., 377, 392, 393
264, 265, 393, 394, 395, 448, 514 Howard, J. H., Jr., 80, 81, 266
Hofman, P., 395 Howard, R., 50
Hogan, J. W., 173 Howard, R. J., 50
Hoge, R. D., 18, 19, 25, 26, 27 Howe, F., 446
Holcomb, P. J., 215 Howe, F. A., 72, 260t
Holden, H. M., 281 Hoyland, A., 461
Holden, J. E. J., 52 Hoyte, K., 487
Holland, A. C., 339, 344 Hrabe, J., 509, 516
Holland, C. M., 396 Hsu, C. L., 446, 454
Holland, D., 42, 131, 144, 145, 173, 237, 492 Hsu, P. J., 42
Holland, F., 170 Hu, D., 106, 108, 117f
Holman, B. L., 514 Hu, L., 119, 444, 447, 448, 451
Holmes, K., 347 Hu, W., 516
Holt, D. P., 40, 156, 490 Hu, X., 283
Holthoff, V., 37 Hua, X., 146, 447, 448
Holtzman, D. M., 37, 43, 110, 138, 470, 485, Huang, C-M., 304
487, 489, 491, 492, 494, 495 Huang, C. M., 218, 245, 374
Hommel, B., 427 Huang, G-F., 40
Honea, R. A., 446, 454 Huang, G., 396
Honey, C. J., 118 Huang, H., 447, 449
Hongwanishkul, D., 165, 166f Huang, S. C., 36
Hooker, J. M., 39 Huang, X., 473
Hooli, B., 488 Hubbard, P. L., 75
Hoppel, B. E., 17 Hudson, D., 223
Hoptman, M. J., 509, 510, 511, 516 Huebner, P., 159, 174
Horan, M., 90, 214, 261t, 264, 266, 422, 514 Hueluer, G., 250
Hori, H., 422 Huentelman, M. J., 145, 146, 417, 426,
Horie, K., 39 487, 494
Horn, S. D., 157 Huettel, S. A., 88, 112, 118, 210, 250,
Hornbeck, R. C., 487, 494, 495 260t, 306t, 309, 310f, 311t, 317, 318f,
Horne, M. J., 241 324, 346
Horsfield, M. A., 21 Huffman, D. J., 287
Horwitz, B., 39, 90, 213, 219, 304, 321, Hugdahl, K., 214
372, 393 Hugenschmidt, C. E., 72
Horwitz, G. J., 278 Hughes, T. F., 430, 462, 491
Hoshida, T., 287 Huijbers, W., 42, 43, 45, 48, 108, 110, 117f,
Hossmann, K. A., 514 134, 490, 494
554
554 Author Index
Hulette, C., 37 Ino, H., 422

Hulette, C. M., 492, 493 Inoue, K., 39
Hüll, M., 171 Inoue, M., 49
Hulsey, K. M., 23 Inoue, O., 54
Hulshoff Pol, H. E., 85, 87, 107, 158, Insel, P. S., 42, 492
174, 198 Intihar, T. W., 442
Hulshoff, P., 514 Intrieri, R. C., 415
Hultsch, D. F., 84, 367t, 368, 369, 462 Iosif, A. M., 90, 110, 138, 249, 264
Humala, N., 404 Iriki, A., 280
Huot, J. R., 349 Isaacs, A., 419
Huppert, F. A., 306t, 308 Iser, L., 430
Huppertz, H-J., 171 Ishai, A., 325
Hurlemann, R., 340 Ishibashi, K., 50
Husain, M., 243 Ishida, R., 514
Huser, J. M. J., 52 Ishihara, M., 108, 110, 117f, 135
Hussey, E., 25, 243 Ishii, K., 50
Husson, Z., 281 Ishimaru, S., 514
Huston, J., 15 Ishiwata, K., 50
Hutcheson, N. L., 212 Ishizaki, J., 514
Hutchison, J. L., 19, 26 Ishizuka, N., 277
Hutchison, K., 107, 108, 110, 117f Issa, J. D., 466
Hutchison, R. M., 105 Ito, H., 39, 40, 52, 325, 493
Huttenlocher, J., 487 Ito, K., 37, 444
Huttner, H. B., 280 Ito, M., 39, 393
Hutton, M., 372 Itoh, T., 54
Huynh, K. D., 417, 426 Ives, D. G., 507
Hwang, K. S., 42 Ivnik, R. J., 278, 485, 489, 490, 492
Hyde, J. S., 107 Iwata, N., 493
Hyde, T. M., 424 Iwata, R., 48
Hyman, B. T., 42, 43, 48, 137, 197, 281, 372, Iwatsubo, T., 492, 493, 496
418, 485, 491, 492, 493, 494 Iyengar, V., 88, 303
Hyman, S. E., 308
Jack, C. R., Jr., 37, 39, 41, 42, 43, 46, 47,
Iadecola, C., 397 118, 134, 145, 146, 278, 372, 485, 487,
Ibanez, V., 219 489, 490, 491, 492, 494, 495, 496
Ibáñez, V., 39 Jackson, A., 90, 214, 261t, 264, 266, 422,
Ibaraki, M., 325 509, 514
Ibayashi, S., 397 Jackson, J. D., 261t, 263
Ichimiya, T., 49, 514 Jacobs, H. I., 260t, 264, 266, 395
Ichise, M., 50 Jacobsen, P. B., 430
Ikeda, K., 418 Jacoby, L. L., 322
Ikeda, O., 422 Jacomb, P. A., 463
Ikonen, S., 279, 280, 283 Jaeggi, S. M., 470
Ikonomovic, M. D., 40, 491 Jagannathan, K., 514
Ikram, M. A., 23, 27, 89, 90, 260t, 395, 448 Jagust, W. J., 23, 36, 37, 40, 42, 43, 44, 45,
Immenroth, M., 17 48, 50, 52, 53, 89, 106, 113, 116, 118,
Imperati, D., 108, 117f 134, 138, 144, 145, 157, 158, 168, 306t,
Ince, P. G., 279, 284, 447 314t, 317, 324, 372, 404, 424, 470, 489,
Ingvar, M., 162, 377, 420, 421 491, 492, 494, 495
Innis, R. B., 50, 53 Jahan, R., 27
555
Author Index 555
Jahreis, A., 518 Jicha, G. A., 82, 83, 260t, 266, 289, 446,
Jain, E., 339 491, 492, 493
Jak, A. J., 420, 421 Jim, H. S., 430
Jakimovich, L. J., 145, 487 Jimenez-Del-Rio, M., 145, 403, 487, 494
Jakobsson-Mo, S., 52 Jin, L. E., 238
James, A., 144 Jin, M., 491
James, J. A., 490 Jin, Y., 462
Janabi, M., 37, 43, 48, 118, 494 Jing, Y., 280
Jäncke, L., 158 Jirsa, V. K., 105
Janes, L., 75 Joanette, Y., 216
Jang, M. K., 493 Jobe, J. B., 465
Janowich, J., 472 Johansen-Berg, H., 72, 74f, 75, 76, 77f, 141,
Janowsky, J. S., 341, 345 144, 159f, 161, 471, 472
Jansen, A., 311t, 317, 319, 324 Johansson, B., 429, 462
Janssen, I., 453 Johansson, J., 474
Janssen, W. G., 280, 283 Johansson, L., 447
Janssens, A. C., 419 Johnson K. A, Schultz A., 493
Jantz, T., 241, 245, 248 Johnson, G. A., 92
Jaques, D., 219 Johnson, J. D., 194, 194f, 196, 197, 312, 325
Jarvik, L. F., 370 Johnson, J. K., 223
Jasielec, M. S., 487, 489, 492 Johnson, K. A., 39, 41, 42, 44, 45, 46, 47,
Jawad, A. F., 23 47f, 48, 49, 52, 118, 134, 135, 139, 174,
Jbabdi, S., 22, 75 197, 218, 372, 485, 487, 489, 490, 491,
Jefferson, A. L., 390, 391, 516 492, 493, 494, 496
Jeffries, N. O., 461 Johnson, L. G., 516
Jeffries, S., 509 Johnson, M., 279, 284
Jelic, V., 444 Johnson, M. A., 81, 84
Jelicic, M., 462 Johnson, M. K., 242, 273, 284, 309, 349
Jellinger, K. A., 492, 493 Johnson, S. C., 37, 44, 45, 90, 137, 157,
Jeng, J., 264, 513 260t, 264, 376, 378, 420, 421
Jenkins, C., 281, 420 Johnson, W., 263, 371, 416
Jenkinson, M., 22, 75, 514 Johnsrude, I. S., 15, 157, 159
Jenkinson, N., 75 Johnston, E., 472
Jennings, J. M., 304, 451 Johnston, P., 263
Jennings, J. R., 389, 393, 395, 396, 397, 400, Johnström, P., 40
404, 406 Jolles, D. D., 170, 176
Jennings, R. G., 134, 249 Jolles, J., 22, 23, 83, 89, 90, 118, 196, 260t,
Jensen, J. H., 22 261, 261t, 264, 265, 266, 395, 462, 463
Jeppsson, F., 40 Jonaitis, E. M., 90
Jernigan, T. L., 134, 140, 446 Jonassaint, C., 465, 471
Jerome, G. J., 449 Jonasson, L. S., 52, 448
Jeste, D. V., 306t, 308, 314t, 324 Jones, C. J., 514
Jezzard, P., 25, 26 Jones, C. K., 287
Jhamandas, J., 428 Jones, D. G., 280
Ji, B., 493 Jones, D. K., 21, 72, 73, 75, 78, 82, 260t,
Jiang, J., 157 261, 319, 472
Jiang, L. L., 108, 117f Jones, D. T., 43, 47, 118, 494
Jiang, T., 108, 110 Jones, G., 372, 485, 492
Jiang, Y., 325 Jones, H. A., 429
Jiang, Y. S., 225 Jones, L. R., 288
556
556 Author Index
Jones, P. B., 424 Kalu, U. G., 514

Jones, T., 52 Kalus, P., 287
Jonides, J., 116, 236, 237, 247, 304 Kalyanam, R., 107, 108, 110, 117f
Jonker, C., 310, 311t, 314t, 324 Kambara, T., 311t, 312, 318
Jonnson, P. V., 393 Kambeitz-Ilankovic, L. M., 422
Jonsdottir, K. Y., 513 Kambeitz, J. P., 422
Jonsdottir, M. K., 380 Kamboh, M. I., 45
Jonsson, P. V., 380, 487 Kammerer, E., 280
Jonsson, T., 260t, 263, 264, 266, 487 Kamo, H., 37
Jonsson, E. G., 487 Kamourieh, S., 106, 108, 109
Joober, R., 422 Kampert, J. B., 453
Jordan, L., 37, 39 Kan, K-J., 263
Jorm, A. F., 463 Kanazawa, I., 418
Josefsson, M., 162, 171, 173, 185, 364, 424 Kandah, C. C., 452
Josephs, O., 195 Kandel, E. R., 49, 282, 514
Joshi, A. D., 48, 490 Kandel, R., 137
Jost, K., 243, 245 Kane, M. J., 245, 247
Jouvent, E., 260t Kanellopoulos, D., 509, 510, 511, 516,
Jovicich, J., 173, 174 517, 519
Joyce, N. M., 516 Kang, C., 347
Ju, C., 78, 82 Kanno, I., 325
Ju, W. K., 141 Kannurpatti, S. S., 26
Jubault, T., 216 Kanowski, M., 474
Jucker, M., 372, 495 Kantarci, K., 37, 39, 42, 43, 46, 485, 489,
Judge, K., 493 490, 491, 492
Julin, P., 40 Kanwisher, N., 325
Julkunen, I., 424 Kaplan, R. F., 260t, 261t
Jung, R. E., 107, 108, 110, 117f Kappelle, L. J., 390, 392
Jungreis, C. A., 89, 395 Kapucu, A., 349
Junque, C., 108, 113, 117f, 240, 261t, 266 Kapur, S., 304
Juraska, J. M., 278 Karabanov, A., 261t
Juréus, A., 40 Karbach, J., 465, 467, 470
Jurick, S. M., 25 Kåreholt, I., 453, 463, 470
Karlamangla, A. S., 249, 379
K, O. K., 420, 421, 494 Karlawish, J., 496
Kaczynski, K., 22 Karlsson, P., 50, 52, 53, 54
Kadish, I., 274 Karlsson, S., 50, 52, 53, 54, 118, 416,
Kahn, I., 106 420, 429
Kahn, R. S., 85, 87, 158, 174, 198 Karlström, G., 389
Kähönen-Väre, M., 391 Karow, D., 134
Kakuma, T., 512 Karr, J. E., 466
Kalaria, R. N., 145 Kasai, K., 159
Kalayam, B., 509 Kasi, D., 493
Kalbe, E., 37 Kasner, S. E., 25
Kalkstein, J., 113, 114, 224, 225f Kassner, A., 396
Kaller, C., 245 Kasten, T., 487
Kalpouzos, G., 37, 39, 48, 53, 81, 83, 84, Kastman, E. K., 260t, 264
134, 157, 159, 161, 163, 164, 166, 170, Kastrup, A., 25, 26
174, 184, 185, 188, 190, 191, 240, 260t, Kastyak-Ibrahim, M. Z., 44
263, 264, 266, 326, 349, 354, 425, 428, Kathmann, N., 353
466, 470, 472, 473 Kato, M., 52
557
Author Index 557
Kato, N., 159 Kennedy, K. M., 15, 25, 26, 27, 43, 82, 89,
Kato, T., 444 118, 157, 158, 188, 189f, 195, 217, 237,
Katsumi, Y., 348 238, 239, 241, 245, 250, 260t, 264, 278,
Katula, J. A., 451 303, 304, 306t, 313, 314t, 316, 316f, 321,
Katz, L., 448, 449 325, 326, 367t, 368, 369, 370, 371, 373,
Katz, M. J., 471 374, 374f, 375, 377, 390, 392, 393, 394,
Katzev, M., 246 397, 403, 404, 405f, 416, 444, 461, 472,
Katzman, R., 41 490, 494
Kaufman, G. E., 514 Kennedy, Q., 348, 349, 354
Kaufmann, J., 283, 472, 474 Kennedy, W. A., 308
Kaufmann, W. E., 134, 146 Kennelly, S. P., 403
Kaup, A. R., 306t, 308, 314t, 324 Kennison, R. F., 462
Kauppi, K., 48, 53, 161, 162, 163, 164, 165, Kenny, R. A., 403
166, 170, 174, 176, 184, 185, 188, 190, Kenowsky, S., 79
191, 240, 249, 304, 326, 417, 420, 421, Kensinger, E. A., 5, 113, 114, 306t, 308,
424, 426, 428, 429 311t, 312, 317, 318, 324, 339, 340, 341,
Kausler, D. H., 303 343, 344, 345, 346, 347, 348, 349, 351,
Kawachi, I., 463 352, 353, 354
Kawas, C. H., 490 Kent, C., 279
Kawasaki, K., 50 Kentros, C., 282
Kawashima, R., 39, 157, 263, 311t, Kentros, J., 212
312, 318 Kenward, M. G., 172, 173
Kay, A. R., 17 Kerchner, G. A., 82, 260t, 264, 266
Kay, C. D., 452 Keren, N. I., 261t, 266
Kaye, J., 496 Kern, R. P., 347
Kǲad‫״‬tir, R. T, 74f, 76, 77f Kerwin, J. M., 279, 284
Keebler, M. W., 25, 450, 472, 473 Kessels, R. P., 260t, 261t, 390
Keidel, J. L., 311t, 317, 324 Kessler, R. M., 52
Keil, B., 76 Kety, S. S., 397
Keilholz, S. D., 105, 107 Ketz, N., 110
Keilp, J. G., 509 Keuker, J. I., 278
Keist, R., 43, 44 Khalidov, I., 29
Kelche, C., 279, 284 Khan, A., 92
Keller, J., 106, 107 Khan, M., 447, 449
Keller, L., 416, 420, 425, 430, 470 Kharitonova, M., 198
Kelley, J., 209, 219, 221f, 245, 247 Khoo, T., 259, 260t, 266
Kellison, I. L., 137 Kido, D. K., 27
Kelly, A. M., 106, 116, 473 Kidwell, C. S., 27
Kelly, C., 119 Kiefer, C., 287
Kelly, J., 485 Kiehl, K. A., 107, 108, 110, 117f
Kelly, J. F., 418 Kieley, J. M., 209
Kelly, M. E., 466 Kievit, R. A., 83, 263
Kelly, R. E., Jr., 6, 510, 515, 516 Kilander, L., 392, 403, 447
Kemp, A. H., 341, 354 Kilbourn, M. R., 48
Kemper, T. L., 279 Kilcullen, S. M., 215
Kempermann, G., 280, 471 Kiliaan, A. J., 43
Kenet, T., 134 Kilk, A., 243
Kenna, H., 106, 107 Killiany, R. J., 173, 174, 279, 288, 321
Kennedy, A. M., 157 Kilts, C., 172
Kennedy, D. N., 17, 279 Kim, A. S. N., 349
Kennedy, J. L., 83 Kim, D. Y., 488
558
558 Author Index
Kim, H., 306t, 308, 310, 321, 324 Klunk, W. E., 485
Kim, J. S., 37, 119, 217, 444, 445f, 447, 448, Kluth, J. T., 42, 50, 53, 157
450, 451, 452, 471, 473, 516 Klyszejko, Z., 243
Kim, S. G., 17, 24 Kmiecik, J., 23
Kim, S. Y., 308, 324 Knecht, S., 53, 443, 444
Kim, Y. H., 488 Knickmeyer, R. C., 146
Kim, Y. J., 50 Knight, M. R., 339, 345, 352, 354
Kindlmann, G., 75, 76 Knight, R., 209, 219, 221f
King, H. E., 390 Knight, R. T., 194, 210, 212, 216, 245, 247
King, K. S., 23 Knipscheer, K. C., 462
King, S. L., 249 Knische, T. R., 75
Kinnecom, C., 41 Knoefel, J. E., 223
Kinomura, S., 39, 157 Knopman, D. S., 37, 39, 41, 42, 43, 46, 47,
Kiosses, D. N., 508, 509, 517, 519 48, 118, 145, 372, 376, 377, 392, 393,
Kippenhan, S., 421 485, 487, 489, 490, 491, 492, 494
Kircher, T. T., 311t, 317, 319, 324 Knorr, S., 353
Kirkwood, T. B., 474 Knutson, B., 264
Kirsch, P., 425 Ko, C. H., 514
Kirsche, W., 92, 280 Ko, J. Y., 392
Kirschen, M. P., 170 Ko, P. C., 243
Kirwan, C. B., 277, 286, 287 Kobayashi, S., 209
Kisley, M. A., 345 Koch, H. E., 339
Kitani, M., 392 Koch, M., 260t, 264
Kitayama, S., 367t, 370 Koch, W., 226
Kitchener, P., 280 Kochan, N. A., 260t, 264
Kittinger, J. D., 509 Kochunov, P., 39, 76, 159, 173, 260t, 261t
Kivimaki, M., 155, 377, 379 Kochunov, V., 260t, 261t
Kiviniemi, V., 105 Koda, M., 422
Kivipelto, M., 40, 378, 380, 380, 403, 429, Kodaka, F., 52
446, 447, 453, 470 Kodituwakku, P., 107, 108, 110, 117f
Kjartansson, O., 393 Koeglsperger, T., 491
Klamm, E. L., 119, 450, 451 Koen, J. D., 195, 196, 302, 303
Klatt, M., 226 Koene, T., 487, 491
Klawiter, E. C., 76, 80 Koepp, M. J., 52
Klee, J. B., 488 Koeppe, R. A., 37, 42, 48, 157, 304, 487, 494
Kleim, J. A., 514 Koeppe, R., 494
Klein, J. C., 135 Kohl, I. II, 453
Klein, M. C., 472 Kohn, P., 421
Kleinman, J. E., 424 Kohn, Y., 514
Kleykamp, B. A., 461 Köhncke, Y., 39, 48, 81, 83, 84, 159, 174,
Klimesch, W., 220 425, 472
Klimstra, S. A., 509, 510, 516 Koivisto, P., 156
Klohs, J., 43, 44 Kojima, M., 422
Klöppel, S., 246 Kok, A., 210
Kloppenborg, R. P., 264, 390 Kokmen, E., 403
Klostermann, E. C., 53, 424 Kokmen, R., 278
Klotz, J., 512 Kolachana, B. S., 417, 421, 422, 424, 426,
Kluger, A., 176 427f
Klunk, W. E., 36, 40, 41, 43, 45, 118, 156, Kolb, B., 281, 439
372, 485, 487, 490, 491, 494, 495 Kolb, H. C., 493
559
Author Index 559
Kolson, D. L., 15, 371 Kretzschmar, H., 493

Komesu, Y. M., 107, 108, 110, 117f Krienen, F. M., 39, 493
Kong, A., 487 Kril, J. J, 89
Kong, E., 472 Krishnan, K. R., 509, 514
Konietzko, U., 43, 44 Krishnan, R. R., 509, 513, 518
Konneker, T., 146 Kritchevsky, S., 453
Konopack, J. F., 448 Kritchevsky, S. B., 449
Konopka, L. M., 514 Krogh, J., 444
Kop, W. J., 507 Kross, E., 116
Koretsky, A. P., 24 Krstic, D., 43, 44
Koretz, B. K., 249 Kruger, K., 443, 444
Korf, E. S., 393, 393 Kryscio, R. J., 289, 491
Korgaonkar, M. S., 79, 261t, 264, 453 Ksiezak-Reding, H., 404
Kornak, J., 158 Ku, Y., 219, 220, 222, 223, 225
Koroshetz, W. J., 24 Kubie, J. L., 283
Korszun, A., 514 Kuceyeski, A., 138
Korten, A. E., 463 Kuczynski, B., 37
Koscik, R. L., 90 Kudo, Y., 48, 493
Kosik, K. S., 145, 487, 494, 496 Kueider, A. M., 466
Koudstaal, P. J., 90, 260t, 261t, 264, 265 Kuhl, C. K., 12
Koutstaal, W., 170, 176, 273, 465 Kuhl, D. E., 36
Kovacevic, N., 84, 107, 108, 117f, 119, 226 Kuhn, H. G., 280, 471
Kovacevic, S., 223 Kuhn, S., 472, 474
Kovacs, D. M., 488 Kühn, S., 472
Kovari, E., 492, 493 Kukolja, J., 306t, 309, 311t, 313, 314t
Koyama, A. K., 83 Kukull, W. A., 418, 492, 493
Krabbe, K., 174 Kulik, J., 337, 339
Kraft, R. A., 72, 451 Kullberg, J., 447
Kragel, J. E., 87, 88, 118, 139 Kuller, L. H., 394, 397, 446, 447, 448, 453
Krajina, K., 453 Kumar, A., 304, 516
Kramer, A. F., 88, 119, 207, 217, 219, Kumazawa-Manita, N., 280
250, 380, 441, 442, 443, 444, 445f, 447, Kunugi, H., 422
448, 449, 451, 452, 462, 463, 465, 471, Kunz, D., 287
473, 516 Kunzmann, U., 339
Kramer, J. H., 23, 82, 118, 158, 176, 260t, Kuo, Y. T., 514
264, 266 Kuperman, J. M., 83, 134, 146
Kramers, R. J., 278 Kupfer, D. J., 49, 507
Krasuski, J. S., 393 Kuriyama, Y., 514
Krause, D. N., 397 Kurth, S., 157
Krauss, G. L., 288 Kurz, A., 37
Krausz, Y., 514 Kusevic, I., 420
Kraut, M. A., 89, 118, 132, 156, 157, 161, Kuswanto, C., 159
163, 165, 171, 377, 394, 397, 490 Kuzminski, S., 92
Kraybill, M. L., 170 Kwak, Y., 118
Krebs, N., 92 Kwong, K. K., 17, 18, 19, 24, 25, 26, 27
Krebs, R. M., 52
Kremerskothen, J., 425 La Joie, R., 108, 110, 117f, 118, 134,
Krendl, A. C., 339, 340, 346, 349 143, 470
Kressler, B., 29 Laakso, M. P., 378, 403
Krestin, G. P., 27, 89, 90, 260t, 395 448 Labar, K. S., 340, 341, 346
560
560 Author Index
Labouvie-Vief, G., 339 Larue, A., 90, 370, 390, 391, 392, 393, 396
Lachman, M. E., 249, 365, 367t, 368, 369, Laruelle, M., 50, 53, 424
370, 379, 379 Lathrop, M., 146
Lacy, J. W., 277, 286, 287, 288, 496 Latoussakis, V., 509, 510, 516
Laere, K. V., 145 Lattka, E., 146
LaHue, S. C., 81, 263 Lau, J., 466
Lai, R., 215 Laubach, M., 238
Laine, M., 474 Laukka, E. J., 39, 48, 81, 83, 84, 134, 159,
Laird, A. R., 106, 107 174, 260t, 263, 264, 266, 416, 420, 425,
Laird, N. M., 169, 170, 172, 173, 176 430, 470, 472
Lal, R., 53 Launer, L. J., 23, 379, 380, 391, 393, 403,
Lalla, D., 518 418, 447, 453, 470
Lalonde, F., 421 Laurienti, P. J, 72
Laluz, V., 82, 260t, 264, 266 Laurienti, P. J., 451
Lam, C., 493 Lauritzen, M., 187, 187f
Lambert, C., 376 Lautenschlager, G., 155, 186, 366, 367t, 369,
Lammertsma, A. A., 43, 45 370, 391
Lanari, A., 404 Lautenschlager, N. T., 425, 490, 492
Lancaster, J. L., 39, 76, 159, 260t, 261t Lavenex, P., 275f, 276
Lancaster, M. A., 450 LaViolette, P. S, 110
Land, S. J., 423 LaViolette, P. S., 43, 161, 165, 176, 197, 372,
Landau, S. M., 37, 42, 44, 48, 50, 53, 83, 418, 420, 421, 494, 495
138, 145, 424, 470, 489, 491, 492 Lavori, P. W., 507
Landeau, B., 37, 39, 108, 110, 117f, 118, Lavretsky, H., 516
134, 143, 157 Lawlor, B. A., 403, 466
Landers, D. M., 440 Lawrence, A. D., 52
Landi, P., 308 Lawrence, J. J., 366
Landman, B. A., 398 Laws, S. M., 46, 423
Landwehrmeyer, B. G., 171 Lawton, C. L., 461
Landwehrmeyer, G. B., 496 Lazar, M., 260t, 264
Lang, A. E., 50 Lazarus, C., 279, 284
Lang, A. G., 470 Le Bihan, D., 19, 22
Lang, M., 52 Le Grevès, M., 447
Langbaum, J. B., 145, 487, 494, 496 Le Masurier, M., 514
Langenecker, S. A., 217 Le Moal, M., 280
Langeslag, S. J. E., 345 Le, T. Q., 76, 80
Langkammer, C., 92, 93 Leal, J. P., 54
Langois, C. M., 145, 487 Leary, M. C., 27
Långstrom, B., 156 Lebel, C., 159
Lansing, A. E., 26, 27 Lebert, F., 48
Lanza, G., 509 Lebon, V., 29
Lao, Z., 23 Lechner, H., 260t, 264
LaRossa, G., 41, 43, 118 Leckie, R. L., 443, 452, 454
Larossa, G. N., 41, 485 Leclerc, C. M., 113, 114, 318, 324, 339, 341,
Larraburo, Y., 472 344, 345, 346, 347, 349, 351
Larsen, V. A., 471, 472 Lecours, A. R., 79
Larsson E. M., 447 Ledoux, J. E., 29, 514
Larsson, A., 52, 162, 377, 420, 421, 424, Lee, D., 238
465, 473 Lee, D. Y., 264
Larsson, H. B., 174 Lee, G. J., 83, 259, 260t, 266
561
Author Index 561
Lee, H. C., 393 Levine, B., 339, 447, 449

Lee, H. G., 487 Levine, L. J., 348, 354
Lee, J. H., 394, 417 Levy, C., 392, 395
Lee, J. M., 138 Lewandrowski, K. U., 37
Lee, P. R., 474 Lewis-Amezcua, K., 26
Lee, R. R., 223 Lewis, C. M., 140
Lee, S. Y., 41, 146, 485, 488 Lewis, J. M., 372, 450
Lee, T. M., 17 Lewis, R. L., 236, 237, 247
Lee, V. M., 372, 493 Lewis, J., 493
Lee, W., 145, 487 Lewkowicz, C., 365
Lee, Y., 114, 115f, 219, 238, 239, 304, 448 Leys, D., 23
Leech, R., 106, 108, 109 Li TQ, Jonsson, T., 134
Lees, A. J., 49 Li, G., 490
Lefkowitz, D., 89 Li, J., 191, 304, 311t, 312, 314t,
Lehmann, M., 118 324, 473
Lei, H., 44 Li, K. Z., 108, 110, 284, 321
Leigh, J. S., 24, 28 Li, L., 28, 210
Leighton, E. A., 339 Li, M-W., 44
Lein, E. S., 108 Li, M., 518
Leirer, V. O., 308 Li, P., 462
Lemaitre, H., 417, 421, 426, 427f Li, R., 52, 473
Lemaitre, H. S., 422, Li, S. C., 41, 50, 80, 156, 211, 212, 213, 214,
Lemere, C. A., 41, 487 239, 240, 249, 304, 415, 416, 422, 424,
Lenglet, C., 22 425, 427, 428, 487, 491
Lennox, B., 306t, 308 Li, T. Q., 21, 25, 26, 39, 48, 81, 83, 84, 159,
Lenox, M., 135 174, 260t, 263, 264, 266, 425, 472
Lenzo, N., 485, 492 Li, W., 92, 93
Leonard, G. T., 146 Li, Y. O., 37, 81, 112, 138, 263, 280, 319,
Leonardi, C., 518 471
Leontiev, O., 26, 27 Li. S. C., 475
Leopold, D. A., 105 Liang, C. L., 26, 27, 462
Leow, A. D., 146 Liang, H. F., 44, 76, 80
Lepage, M., 302 Liang, Q., 493
Leplow, B., 283 Liao, D., 377, 392, 393
Lerch, J., 393 Libiger, O., 134, 146
Lerer, B., 514 Libkuman, T. M., 347
Leritz, E. C., 137, 260t, 264, 266, 393, 395 Liddell, B. J., 341, 354
Leshikar, E. D., 306t, 309, 311t, 312, 319, Lieb, K., 217
345, 346 Liebeskind, D. S., 27
Lesnick, T. G., 43, 46, 145, 372, 491 Light, L. L., 301, 303
Lesser, I. M., 514 Lim, A., 453
Lester, H., 514 Lim, H. K., 43, 494
Leube, D. T., 311t, 317, 319, 324 Lim, K. O., 15, 78, 82, 509, 510, 511, 516
Leung, A. W. S., 209 Lim, Y. Y., 45, 423, 490, 492
Leung, K. K., 44 Lin, C., 15
Leurgans, S. E., 417 Lin, H. F., 514
Leveck, M. D., 465 Lin, M. K., 376, 428
Levens, S. M., 264 Lin, S. J., 76, 80, 141
Levenson, R. W., 339 Lin, W., 146
Leverenz, J. B., 490, 492, 493 Lin, Y-T., 453
562
562 Author Index
Lind, J., 48, 53, 161, 162, 163, 164, 165, Locascio, J. J., 137, 485, 490, 491
166, 170, 174, 176, 184, 185, 188, 190, Locke, D. E., 376, 404
191, 240, 326, 377, 420, 421 Locke, P. A., 429
Lindenberg, R., 108, 117f Löckenhoff, C. E., 346
Lindenberger, U., 5, 6, 15, 48, 50, 52, 53, 80, Lockhart, S. N., 73, 78, 82, 90, 91, 139, 260t,
84, 88, 116, 118, 132, 137, 145, 155, 156, 264, 287
157, 158, 184, 188, 212, 213, 214, 220, Lodi-Smith, J., 465
224, 237, 239, 240, 242, 243, 244, 245, Loessner, A., 37
249, 266, 278, 303, 304, 313, 321, 371, Logan, J., 49, 50, 51f, 52, 53
373, 391, 392, 393, 415, 416, 416f417, Logan, J. M., 162
418, 422, 423, 424, 425, 427, 428, 439, Logie, R. H., 241, 242, 243
444, 461, 462, 463, 464, 465, 470, 471, Logothetis, N. K., 287
472, 473, 474, 475 Lohmann, H., 443
Lindquist, K., 453 Loken, W. J., 303, 516
Lindroos, M., 391 Longstreth, W. T., 89, 395
Lindsay, D. S., 273 Loosli, S. V., 245
Lindsay, R. M., 422 Lopera, F., 145, 403, 487, 494, 496
Lindwall, M., 462 Lopez-Sola, M., 106
Ling, C., 429 Lopez, I., 157
Links, J. M., 49 Lopez, L., 145, 487
Linnington, H., 306t, 308 Lopez, O. L., 146, 380, 393, 397, 446, 447,
Lipa, P., 277 448, 454, 491
Lipnicki, D. M., 157 Lopresti, B. J., 156, 490
Lipsitz, L. A., 137, 395, 518 Lord, C., 174
Lipska, B. K., 424 Lorenzo-Lopez, L., 209, 210, 211
Lipton, R. B., 471 Lori, N. F., 85
Lishmanov, Y. B., 398 Lorist, M. M., 108, 110, 111, 117f, 210
Lisman, J. E., 284 Lorius, N., 490, 491
Lithell, H., 392, 403 Loth, E., 146
Little, R., 168, 176 Lotta, T., 424
Liu-Ambrose, T., 446, 454 Lou, W., 283
Liu, A., 159, 174 Loughrey, D., 466
Liu, C., 76, 92, 93, 118, 138, 139, 493 Lourdusamy, A., 146
Liu, D., 23, 390, 391 Lövdén, M., 5, 6, 39, 48, 50, 52, 81, 83, 84,
Liu, E., 134 88, 116, 132, 134, 137, 145, 159, 174,
Liu, F., 419 239, 240, 243, 260t, 263, 264, 266, 415,
Liu, H., 219, 224, 461, 493 416, 417, 420, 423, 425, 429, 439, 461,
Liu, J., 29, 288, 516 462, 463, 464, 465, 466, 470, 471, 472,
Liu, L. L., 161, 306t, 314t, 324 473, 474, 475
Liu, P., 3, 25, 26, 27, 188, 189f, 195, 280, Love, S., 492, 493
326, 403 Low, K. A., 245, 250
Liu, Q., 92 Lowe, C., 90, 214, 261t, 264, 266, 514
Liu, T. T., 25, 29, 157, 420, 422 Lowe, V. J., 37, 39, 42, 43, 46, 47, 145, 485,
Liu, Y., 108, 110, 487 487, 489, 490, 491, 492, 494
Livy, D. J., 141 Loy, C. T., 42
Lladó, A., 43, 494 Lu, Q., 472
Lleó, A., 42 Lu, B., 422, 453
Lo, R. Y., 43, 168 Lu, H., 3, 19, 22, 25, 26, 27, 188, 189f, 195,
Lobaugh, N. J., 83 241, 250, 325, 326, 403, 447, 449, 450,
Locantore, J. K., 87, 162, 185 472, 473
563
Author Index 563
Lu, M., 490 MacDonald, S. W., 53, 54, 84, 349, 354, 429,
Lu, P. H., 83, 259, 260t, 265, 266, 447 462, 471
Lü, W., 347 MacFall, J. R., 509, 513, 514
Lu, Y., 487 Macgregor, K. L., 516
Lu, Z., 390, 391 MacGregor, R. R., 49, 50, 51f
Luchsinger, J. A., 146, 289 Mach, R. H., 41, 485
Lucignani, G., 37 Machado, A., 261t, 266
Luck, S. J., 90, 242, 243, 260t Machizawa, M. G., 243
Luechinger, R., 340 Machulda, M. M., 43, 46, 118, 494
Luigjes, J., 514 Macintosh, B. J., 25, 377, 421, 487, 494
Luiten, P. G., 278 Mack, W., 42, 404, 518
Luk, G., 118 Mackay, C. E., 106, 377, 421, 487, 494, 514
Luks, T. L., 107 Mackay, D. G., 283
Lumley, G., 217 Mackenzie, I. R., 492, 493
Lumley, M. A., 339 Mackey, A. P., 472
Lund, R., 379 Mackin, R. S., 42, 517
Lundervold, A. J., 52, 110, 139, 140, 157, Mackinnon, A., 463
260t, 264, 376 MacKinnon, D. P., 266, 267
Lundquist, A., 3, 162, 171, 173, 249, Maclin, E. L., 238, 239, 250
304, 424 Macpherson, H. N., 212
Lundt, E. S., 37, 39 MacPherson, S. E., 367t, 368, 369, 370
Lunn, D., 214 Madden, D. J., 3, 39, 73, 75f, 78, 79–80, 81,
Lunn, M., 90 82, 83, 84, 87, 88, 92, 93, 106, 108, 111,
Luo, L., 183 112, 117f, 118, 134, 139, 183, 192, 207,
Luo, V., 425 209, 210, 260t, 261, 266, 303, 306t, 308,
Luo, Y. J., 211 309, 310f, 311t, 312, 312f, 313, 314t, 317,
Lupien, S. J., 174, 379 318f, 319, 324, 346, 390, 417, 446, 474
Lustig, C. A., 4, 108, 110, 113, 116, 117f, Mader, I., 171
132, 139, 165, 208, 209, 211, 212, 215, Madhyastha, T. M., 108, 110, 117f, 134,
217, 219, 226, 236, 237, 239, 240, 241, 260t, 264, 266
245, 247, 284, 304, 307, 316, 472, 494 Madison, C. M., 37, 42, 43, 50, 53, 118, 145,
Luszcz, M. A., 462 157, 489, 491, 492, 494
Luszczynska, H., 512 Madison, G., 261t
Luu, T., 281 Madjar, C., 26, 27
Luxen, A., 191, 311t, 312, 314t Madrid, J. A., 274
Luzzi, S., 243 Maeda, J., 493
Ly, M., 286 Maeder, P., 448
Lyketsos, C., 447 Maercker, A., 367t, 370
Lyle, K. B., 273 Maestu, F., 211
Lyons, J. A., 450 Maffei, A., 404
Magnusson, O. T., 487
M, G., 278, 279 Maguire, P., 173, 174
Ma, F., 462 Mahlberg, R., 287
Macaulay, S. L., 487, 492 Mahncke, H. W., 216, 516
MacAvoy, M. G., 50, 53 Mahony, K., 116
Macchi, C., 440 Mailey, E., 250, 444, 445f, 452, 471
MacDonald, A. W., 429 Mailey, E. L., 119, 449
MacDonald, A. W. III, 439, 452 Maillard, P., 90, 91, 264, 287, 453
MacDonald, E., 514 Maillet, D., 183, 193, 316, 326
MacDonald, K., 428 Maisog, J. M., 213, 219, 304, 321
564
564 Author Index
Maitland, S. B., 415 Markowitsch, H. J., 241

Majcher-Tascio, M., 509, 516 Marks, B., 448, 449
Mak, E. E., 52 Marks, D., 42, 492
Makizako, H., 444, 446 Marks, S. M., 43, 118, 470, 491, 494
Makris, N., 52, 139, 140, 157, 173, 174, 279 Marks, W., 224
Malarkey, W., 226 Markus, E. J., 279
Maldjian, J. A., 25, 72 Markus, H. S., 72, 72, 78, 82, 83, 174, 260t,
Mali, W. P., 397, 399f 261, 264, 303, 319, 446, 447, 472
Malik, R. A., 260t, 514 Marmot, M., 377, 379, 463
Malina, D., 428 Marmot, M. G., 377
Malkowski, E., 119, 450, 451 Marner, L., 76, 137
Mallmann, J., 495 Marquez, D. X., 444, 447, 448, 449
Malloy, P. F., 248 Marquie, J. C., 377, 463
Malone, I. B., 44 Marquie, M., 491
Maloney, J., 487 Marr, D., 276
Mandell, D. M., 396 Marrett, S., 18, 19, 25, 26, 27
Mandeville, J. B., 39 Marriott, L. K., 277
Mandl, R. C., 85, 87, 514 Marshall, G. A., 42, 43, 197, 372, 492,
Mang, S. C., 76 493, 495
Manganaro, D., 404 Marshuetz, C., 240, 304
Mangialasche, F., 40 Marsiske, M., 465
Mangin, J. F., 19, 22 Marsland, A. L., 447, 453
Maniega, S. M., 89, 92, 260t, 263, 264, Mårtensson, J., 439, 471, 472, 473
447, 448 Martin, A. M., 303
Manly, J. J., 471 Martin, K. C., 514
Mann, D. M., 492, 493 Martin, M., 44, 366, 367t, 368
Mann, K., 146 Martin, R. C., 246, 247
Mannon, L. J., 15, 371 Martin, R. E., 198
Manolio, T. A., 89, 395 Martin, S. A., 444, 445f, 452, 471
Manrique, A., 39 Martin, W., 159
Manuck, S. B., 390, 400, 429, 439, 452 Martin, W. R., 23
Mao, Y., 280 Martinaud, C., 377
Maquet, P., 191, 311t, 312, 314t Martinez, O., 110, 138, 264
Marchant, N. L., 43, 404 Martins, R., 42, 372, 485, 487, 492, 494
Marcoen, A., 465 Martins, R. N., 45, 46, 423, 490, 492,
Marcus, A., 365 494, 495
Marcus, B. H., 453 Martire, L. M., 507
Marcus, D. S., 42, 485, 487, 491, 494, 495 Maruff, P., 41, 45, 423, 487, 490, 492
Marcus, J., 395 Maruyama, M., 493
Marder, K. S., 471 Maruyama, N., 287
Mardimae, A., 26 Marzloff, K., 485
Marek, K. L., 50, 53 Masaki, K., 377, 391, 393, 403
Marenco, S., 396 Masamoto, K., 493
Margolin, R. A., 23 Masangkay, E., 113, 223
Maril, A., 170, 176 Masdeu, J. C., 421
Marin, K., 343 Mash, D. C., 280, 372
Marioni, R. E., 424 Masliah, E., 492, 493
Markham, J. A., 278 Mason, E., 243
Markley, C., 138 Masters, C. L., 41, 42, 45, 46, 48, 49, 372,
Marklund, P., 52, 424 423, 485, 487, 490, 492, 494
565
Author Index 565
Masters, C. M., 495 Mayhew, S. D., 26

Masutani, Y., 159 Maylor, E. A., 242
Mata, M., 37 Mayr, U., 243, 245
Mathalon, D. H., 15 Mazer, J. A., 238
Mather, M., 114, 339, 340, 341, 344, 345, Mazoyer, B., 421
346, 348, 349, 352, 354 Mazzanti, C. M., 424
Mathews, D., 372, 490 Mazziotta, J. C., 418, 420, 421
Mathis, C. A., 40, 41, 42, 43, 45, 49, 118, Mazzonna, F., 463
156, 157, 372, 485, 487, 490, 494 McAdams, C. J., 27
Matsuda, H., 39 Mcardle, J. J., 367t, 370, 378, 379, 417, 463,
Matsuda, I., 345 464
Matsuda, T., 514 Mcauley, E., 441, 442, 443, 444, 445f, 447,
Matsumoto, M., 424 448, 449, 451, 452, 471
Matsumoto, S., 23 McAuley, E., 119, 250
Matsunari, I., 39 McAvoy, M. P., 78, 107, 132, 165, 226, 307,
Matsuo, K., 223 316, 446, 494
Matsushima, E., 514 McCabe, C., 134, 146
Matsushita, K., 514 McCann, J. J., 471
Matsuzawa, T., 393 McCarley, R. W., 171
Mattay, V. S., 108, 113, 114, 117f, 119, 162, McCarthy, G., 250
226, 239, 317, 324, 341, 344, 349, 351, McClelland, J. L., 277, 302
417, 421, 422, 425, 426, 427f McColl, R. W., 23
Mattfeld, A. T., 286f, 287 McCollum, L. A., 212
Matthews, M. A., 452 McCreary, C. R., 264, 513
Matthews, P. M., 75, 85, 144, 377, 421, 487, Mcculloch, C., 517
494 McDade, E., 118, 494, 495
Matthies, H., 280 Mcdade, E., 487
Mattia, J. I., 507 Mcdaniel, M. A., 394
Mattis, S., 512 Mcdermott, L., 514
Mattson, J. T., 191, 192f, 193, 196, 197, 198f Mcdonald-Miszczak, L., 367t, 368, 369
Mattson, M. P., 453 McDonald, C. R., 83
Mattsson, N., 42 McDonough, I. M., 195, 197, 304, 311t, 312
Maura, G. M., 428 McEvoy, K., 209, 219, 221f, 245, 247
Maurage, C-A., 48 McEvoy, L. K., 42, 83, 131, 134, 144, 145,
Maurer, K., 40 173, 237, 249, 492
Maurits, N. M., 108, 110, 111, 117f, 210 McEwen, B. S., 249, 379
Mauro, S., 493 McFarland, C. P., 137
Mawuenyega, K. G., 487 McGann, J. P., 279, 280
May, A., 158, 471 McGeer, E. G., 37
May, C. P., 113, 116, 244, 245, 247, 248, 339 McGeer, P. L., 37
May, P. C., 138, 274, 495 McGlinchey, R. E., 137, 260t, 264, 266, 393,
Mayberg, H. S., 54 395
Mayda, A. B., 287 Mcgovern, P., 377
Maye, J. E., 42, 490, 491, 492 Mcguire, P., 422
Mayer, A. R., 107, 108, 110, 117f McHugh, P., 138
Mayer, C., 490 McInnes, L., 170
Mayes, A., 422 McIntosh, A. R., 72, 84, 87, 105, 107, 108,
Mayes, A. R., 313 111, 112, 114, 117f, 119, 162, 165, 166f,
Mayeux, R., 146, 287, 289, 417, 418, 487, 185, 196, 226, 239, 304, 341
494, 495 Mcintyre, D., 446
566
566 Author Index
McIntyre, D. J., 72, 260t Mennes, M., 108, 117f, 119

McKay, D. R., 107 Menon, R., 17
McKay, H. M., 278 Menon, V., 85, 106, 107, 118
Mckee, A. C., 492, 493 Mentis, M. J., 118, 213, 219, 304, 321, 393
McKeel, D. W., 41, 485 Merkle, H., 17
McKeith, I. G., 391, 447, 518 Merrill, D. A., 278
McKian, K. P., 278 Merskey, H., 394
McKinney, M., 279 Mervis, R. F., 491
Mckinstry, R. C., 509, 513 Merzenich, M. M., 516
McKinstry, R. C., 85 Mesholam, R. I., 516
Mclaren, D. G., 420, 421, 494 Mesulam, M. M., 72, 372
McLaren, D. G., 42, 43, 108, 110, 117f, 137 Metsemakers, J., 462
McLaughlin, A. C., 24 Metter, E. J., 156, 157, 161
McMahan, R. W., 279, 280, 283 Metter, J., 377
McMillan, K. M., 107 Metzak, P., 112, 319
McNab, J. A., 75 Meuli, R., 118, 448
McNaughton, B. L., 276, 277, 279, 282, 302 Meunier, D., 108, 138, 146, 240
Mcneil, C. J., 397 Mevel, K., 37, 108, 110, 117f, 118, 134,
McQuain, J., 303, 516 143, 157
Mecocci, P., 40, 403 Meyer-Lindenberg, A., 162, 239, 317, 324,
Medford, N., 339 341, 349, 351, 425
Medrano, S., 42 Meyer-Luehmann, M., 372
Meghpara, M., 83 Meyerand, M. E., 108, 110, 117f
Mehringer, M. C., 514 Meyers, B. S., 512, 513, 517
Mehta, A., 43, 118, 494 Mézenge, F., 39, 108, 110, 117f, 118,
Mehta, M. A., 50 134, 143
Mehta, S., 514 Michael, A. M., 107, 108, 110, 117f
Mehta, T. H., 41, 487 Middelkoop, H. A. M., 110
Mei, J., 44 Mielke, M. M., 37, 39, 42, 43, 47, 145, 392,
Meier-Ruge, W., 72 487, 492, 494
Meier, D., 24 Mielke, R., 37
Meier, E., 72 Mier, D., 425
Meier, I. B., 43, 79, 261t, 264 Mies, G., 514
Meier, T. B., 108, 110, 117f, 396 Miezin, F. M., 107
Meindl, T., 139, 174, 226, 514 Mikulis, D. J., 396
Meinzer, M., 108, 117f, 516 Milberg, W. P., 137, 260t, 264, 266, 393,
Melchior, M., 377, 378 395, 518
Melen, K., 424 Miles, J., 174
Melhem, E. R., 23 Miles, J. D., 463
Melka, M. G., 146 Milham, M. P., 106, 108, 116, 117f, 119, 217
Meltzer, C. C., 49, 54, 400 Millard, S. P., 490
Meltzer, J., 282 Miller-Martinez, D. M., 249, 379
Mena, I., 514 Miller, A., 304
Mende-Siedlecki, P., 346 Miller, A. C., 239
Mendelson, D. N., 393, 395, 396, 398 Miller, B. L., 23, 37, 40, 42, 43, 48, 82, 118,
Mendes De Leon, C. F., 462, 471, 491 157, 158, 176, 260t, 264, 266, 393, 395,
Mendez, M. F., 260t, 265 494, 514
Mendlewicz, J., 512 Miller, D. H., 85
Menke, R. A., 75 Miller, E. K., 251
Mennerick, S., 138, 495 Miller, J., 422
567
Author Index 567
Miller, K. L., 17, 22, 75, 106 Moeller, J. R., 471

Miller, M., 277 Moeller, S., 17, 22
Miller, M. I., 287 Mogapi, O., 214
Miller, S. L., 161, 197, 288, 308, 314t, Moghekar, A., 394, 397
494, 495 Mohtasib, R. S., 217
Miller, W. L., 107 Molbo, D., 379
Miller, Z. A., 37, 48 Molden, S., 284
Millet, P., 39, 219 Molenberghs, G., 173
Mills, S. M., 495 Molinuevo, J. L., 43, 240, 494
Milner, B., 274 Molitor, R. J., 243
Milnik, A., 425 Molko, N., 19
Milo, T. J., 514 Moller, H. J., 139, 514
Mimura, M., 514 Molteni, R., 453
Minato, K., 287 Monchi, O., 216
Minear, M., 161, 241, 304, 306t, 314t, 324 Mondadori, C. R., 417, 426
Ming, G. L., 280 Monfils, M. H., 514
Minoshima, S., 37 Monk, C. S., 283
Minthon, L., 40, 485 Monsch, A. U., 75, 144
Minton, B., 190, 191, 193, 193f, 306t, 308, Montaldi, D., 313
314t, 324 Montaron, M. F., 280
Mintun, M. A., 23, 37, 41, 43, 48, 118, 138, Montero-Odasso, M., 393
139, 157, 400, 470, 485, 490, 491, 492, Montine, T. J., 485, 490, 492, 493, 496
494, 513, 514 Moody, D. M., 514
Mintz, J., 260t, 265, 447, 516 Moore, B., 461
Minuzzi, L., 52 Moore, K. S., 236, 237, 247
Miran-Khan, K., 446, 454 Moore, R. Y., 49
Miranda, D., 83 Moore, T. L., 279
Mirra, S. S., 493 Mooren, F., 443, 444
Mirzakhanian, H., 306t, 308, 314t, 324 Mora, F., 50
Mirzazade, S., 306t, 309, 311t, 313, 314t Moran, E. K., 41, 485, 490, 491
Mishkin, M., 213, 219, 283 Morandeau, L., 485, 492
Missonnier, P., 219 Morar, B., 425
Mistur, R., 37 Moratti, S., 211
Mitchell, D. B., 367t, 369, 369 Morcom, A. M., 108, 185, 190, 191, 196,
Mitchell, D. J., 83 304, 306t, 308, 311t, 312, 314t, 324
Mitchell, K. J., 309 Morenas, E., 42
Mitchell, M. B., 462 Moreno, S., 145, 487
Mitchell, T. W., 372 Moretti, R., 404
Mitrofanov, A., 428 Morgan, A., 451
Mittal, S., 27 Mori, S., 19, 72, 73, 397
Mittelman, M. P., 176 Mori, T., 422
Miyajima, F., 422 Moriguchi, Y., 340, 422
Miyakawa, A., 52 Morimoto, S. S., 510, 514, 515, 516, 518
Miyakoshi, M., 223 Moriya, H., 422
Miyawaki, A., 280 Mormino, E. C., 39, 40, 42, 43, 44, 45, 46,
Mizusawa, H., 50, 418 47, 47f, 48, 49, 52, 118, 134, 135, 138,
Mo, H., 145, 487 157, 174, 470, 489, 490, 491, 492, 493,
Moberg, P. J., 52, 53, 516 494, 495
Moberget, T., 471, 472 Morra, J. H., 42
Mocharla, V. P., 493 Morrell, F., 280, 284
568
568 Author Index
Morris, J. A., 157, 351, 353, 354 Mugellini, A., 517

Morris, J. C., 37, 41, 42, 43, 78, 110, 118, Mugler, J. P. III, 16
131, 132, 134, 138, 139, 162, 165, 226, Muhammad, A., 439
250, 307, 316, 371, 373, 446, 470, 485, Mukherjee, P., 263
487, 489, 490, 491, 492, 493, 494, 495 Mukherjee, Q., 81
Morris, J. N., 465 Mulatu, M. S., 462, 463, 464
Morris, K. S., 119, 444, 451 Muldoon, M. F., 390, 393, 395, 396, 397,
Morris, M. C., 391, 462 400, 452
Morris, R. G., 72, 72, 78, 82, 83, 174, 238, Muller-Oehring, E. M., 211, 216
260t, 261, 264, 276, 277, 282, 303, 319, Muller, M. M., 215, 393, 397, 399f
446, 472 Muller, R. U., 283
Morris, Z., 260t, 446, 447, 448, 471 Muller, R. V., 282
Morrison-Bogorad, M., 496 Muller, U., 424
Morrison, J. H., 137, 249, 273, 278, 280, Müller, U., 424
281, 283, 284 Mulligan, R. S., 48, 372, 492
Morrow, D. G., 308, 465 Mullin, K., 288
Morse, J. K., 422 Mulsant, B. H., 83
Mortensen, E. L., 174, 187, 187f, 379, 462, Munafo, M. R., 424
463, 464 Mungas, D. M., 37, 42, 90, 91, 112, 159,
Mosconi, L., 36, 37, 138 174, 176, 249, 264, 462
Moscovitch, M., 108, 112, 114, 115f, 219, Munoz Maniega, S., 81, 83, 263, 264, 446,
302, 304, 307, 313, 321, 339, 349 471
Moscufo, N., 260t, 261t Munoz, C., 145, 487, 494
Moseley, M. E., 21, 25, 26, 76, 78, 82 Munsell, L., 487
Moser, E. I., 108, 284 Murakami, M., 422
Moser, M. B., 284 Murali, D., 44
Mosley, T. H., 376, 378, 392, 393 Muraskin, J., 43, 396
Moss, M. B., 279, 287, 321 Murata, Y., 514
Moss, S. A., 41 Murgas, D., 248
Mostofsky, S., 134, 146 Murphy, C. F., 379, 508, 509, 510, 516, 517
Motamedinia, S., 190, 308, 314t, 324 Murphy, D. G. M., 39, 90, 263, 372, 393
Motes, M. A., 26, 211, 264 Murphy, D. R., 321
Motl, R. W., 380 Murphy, K. J., 116
Mott, K. K., 215 Murphy, M. C., 43
Moulder, K., 487, 494 Murray, A. D., 397
Mowinckel, A. M., 108, 110, 117f, 135 Murray, B. D., 339, 340, 341
Moyer, J. R. J., 279, 280 Murray, C., 81, 83, 89, 92, 260t, 263, 264,
Mozaffarian, D., 389 446, 447, 448, 471
Mozley, D., 37 Murray, E. A., 283, 286
Mozley, L. H., 53 Murray, G. D., 424
Mozley, P. D., 53 Murray, H. M., 260t, 265
Mu, F., 493 Murray, J., 138
Mu, Y., 280 Murray, M. D., 392
Mucke, L., 494 Murray, S. S., 134, 146
Mueller, D., 17 Murrell, J. R., 462
Mueller, S. G., 226, 288 Murty, V. P., 108, 113, 114, 117f, 119, 226,
Muffat, J., 488 317, 324, 340, 341, 349, 351, 422, 425
Mufson, E. J., 278, 287, 372 Muscatell, K. A., 113, 114, 318, 324,
Muftuler, L. T., 190, 193, 193f, 277, 287, 344, 345
306t, 308, 314t, 324 Muse, J., 417, 426, 427f
569
Author Index 569
Myers, C., 319 Neumann, J., 218

Myers, R. H., 418 Nevalainen, N., 50, 52, 132, 137
Myerson, J., 242 Newell, K., 137
Newman, A. B., 446, 447, 449
Na, J., 367t, 369 Newmark, R. E., 288, 494
Nadel, L., 108, 112, 282, 302 Newsome, R. N., 225, 353
Nader, K., 514 Ng, C., 113, 114, 117f, 219
Nagamatsu, L. S., 446, 454 Ng, J., 81, 263
Nagatsuka, K., 514 Ng, S., 41
Nagel, B. J., 420, 421 Nga, L., 345
Nagel, I. E., 212, 213, 239, 249, 415, 418, Ngandu, T., 380, 453, 470
422, 424, 425, 427, 428 Nguyen, C. T., 263
Nagy, Z., 472 Nichols, L. M., 421
Nair, V. A., 108, 110, 117f, 396 Nicholson, D. A., 278, 280
Naismith, R. T., 76, 80 Nick, R., 156, 157, 161
Nakagawa, S., 311t, 312, 318 Nicklas, B. J., 453
Nakai, T., 223 Nielson, K. A., 217, 450, 452
Nakamura, M., 514 Niemann, C., 444
Nakayama, C., 422 Nieminen, L. J., 380, 446, 447
Namba, Y., 418 Nierenberg, A. A., 507
Nanko, S., 418 Niessen, W. J., 23, 89, 90, 260t, 448
Naranjo, N., 279, 287 Nietfeld, W., 424, 427, 428
Narayana, S., 39 Nieto, A., 261t, 266
Narkhede, A., 43 Niinisto, L., 145
Nashiro, K., 339, 340, 341, 345 Nijpels, G., 392
Nathan, P. J., 425 Nikolich, K., 425
Naveh-Benjamin, M., 191, 242, 243, 281, Nilsson, E., 429
303, 321, 324, 326 Nilsson, J., 263, 266
Neaton, J. D., 389 Nilsson, L. G., 48, 52, 53, 83, 84, 132, 137,
Nebes, R., 43, 494 144, 159, 161, 162, 163, 164, 165, 166,
Nebes, R. D., 45, 490 170, 174, 176, 183, 184, 185, 186, 188,
Nederkoorn, P. J., 264 190, 191, 240, 249, 260t, 264, 266, 304,
Nee, D. E., 116, 236, 237, 247 326, 364, 367t, 368, 369, 370, 373, 377,
Neelavalli, J., 27, 28, 92 415, 417, 420, 421, 424, 426, 428,
Neelin, P., 393 429, 471
Neely, A. S., 465, 473, 474 Nisbett, R. E., 367t, 370
Negreira, A., 340 Nishikawa, T., 514
Nehlig, A., 461 Nishimura, S., 39
Neisser, U., 339 Niskanen, E., 380, 446, 447
Nelissen, N., 43 Nissanov, J., 372
Nelson, C. A., 283 Nissen, M. J., 209
Nelson, J. K., 307, 316 Nissinen, A., 378, 380, 380, 403, 446, 447,
Nelson, P. T., 48, 289, 491, 492, 493 453, 470
Neltner, J. H., 289 Nitert, M. D., 429
Nematollahi, S., 280 Nitsch, R. M., 43, 44
Nemoto, K., 422 Nittono, H., 345
Nesmith, K., 354 Niu, Y., 473
Nesselroade, J. R., 84, 116 Niznik, H. B., 50
Neufeld, A. H., 141 Noack, H., 240, 243, 466, 470, 472, 474
Neugebauer, E. A., 17 Nobili, F., 396
570
570 Author Index
Nobre, A. C., 213, 219, 224 O’Meara, T., 492

Noda, A., 422, 423 O’Neil, J. P., 37, 40, 43, 48, 50, 53, 106, 113,
Noguchi, H., 422 116, 118, 424, 470, 491, 494
Nooyens, A. C., 380 O’Reilly, R. C., 49, 276, 277, 302
Nordberg, A., 156 O’Sullivan, M., 72, 72, 78, 82, 260t, 261,
Nordenskjöld, R., 447 319, 446, 472
Nordentoft, M., 444 Oakley, M., 454
Norman, D., 158 Oates, G., 463
Norman, K. A., 277, 302 Oathes, D. J., 119
Norris, D. G., 448 Obenaus, A., 92
Nosheny, R., 42 Oberlin, L. E., 6
Nosheny, R. L., 492 Ochsner, K. N., 346, 352
Nothelle, S., 392 Oda, K., 50, 514
Novak, P., 394, 396 Oedekoven, C. S., 311t, 317, 319, 324
Novak, V., 394, 396 Ofenloch, I. T., 283
Novitch, R. S., 519 Offenbacher, H., 260t, 264
Nowell, P. M., 440 Offord, K. P., 403
Nuechterlein, K. H., 447 Oga, T., 212
Null, M., 260t, 261t Ogawa, S., 17
Nummenmaa, A., 76 Ogg, R. J., 92
Nutt, R., 135 Ögren, M., 50, 52, 132, 137
Nwankwo, T., 389 Oh, H., 42, 43, 45, 46f, 138, 145, 306t, 314t,
Nyberg, L., 3, 5, 48, 50, 52, 53, 80, 83, 84, 317, 324, 372, 470, 489, 491, 492, 494
87, 88, 108, 110, 116, 117f, 132, 137, 144, Oh, J., 44
145, 156, 159, 161, 162, 163, 164, 165, Ohnishi, T., 422
166, 170, 171, 173, 174, 176, 184, 185, Ohta, K., 514
186, 188, 190, 191, 212, 213, 236, 238, Ohtomo, K., 159, 173, 174, 176
239, 240, 249, 260t, 264, 266, 304, 306t, Oinas, M., 145
326, 341, 349, 364, 367t, 368, 369, 370, Oishi, K., 44
373, 377, 415, 416, 417, 420, 421, 422, Okada, T., 422
423, 424, 426, 428, 429, 465, 471, 473, Okamura, N., 48, 493
474, 475 Okonkwo, O. C., 44, 90
Nyengaard, J. R., 76, 137 Okubo, Y., 49, 52, 514
Nyffeler, T., 287 Okuizumi, K., 418
Nyman, H., 392, 403 Okumura, M., 52
Old, S. R., 191
O’Brien, J. L., 43, 161, 165, 176, 197, 344, Oleksik, A. M., 110
372, 418, 420, 421, 447, 494, 495 Olichney, J., 264
O’Brien, J. T., 263, 266, 398 Oliver, A., 494, 495
O’Brien, K. C., 132, 165, 226, 307, 316, 494 Olives, J., 43, 494
O’Brien, P. C., 278 Olivieri, G., 341, 354
O’Brien, R. J., 165, 377 Ollier, W., 422
O’Connell, R. G., 215, 219 Olofsen, H., 260t, 265
O’Connor, D. F., 26 Olson, A. K., 379
O’Hara, R., 304 Olson, E. A., 119
O’Keefe, G., 372, 485, 492 Olson, I. R., 302, 307, 313, 321
O’Keefe, J., 282 Olson, L. E., 490
O’Keefe, K. M., 43, 110, 161, 165, 176, 197, Olsson, A. H., 429
372, 418, 420, 421, 495 Oluwadara, B., 260t, 265
O’Leary, D. H., 89 Omae, T., 514
571
Author Index 571
Oman, L., 424 Paller, K. A., 304

Ona, V. O., 518 Palmer, A. C., 23
Onami, S. H., 43, 118, 494 Palmer, D., 341, 354
Ono, M., 493 Palop, J. J., 494
Onoda, K., 108, 110, 117f, 135 Palotie, A., 487
Onodera, O., 418 Palumbo, C., 377, 380, 453
Onur, O. A., 212 Pan, P., 219, 224
Oosterman, J. M., 260t, 261t Pandav, R., 391
Oostra, B. A., 419 Pang, P. T., 453
Opherk, C., 260t Pankratz, N., 146
Opitz, P. C., 346 Pankratz, V. S., 41, 42, 43, 47, 145, 372, 487,
Orr, G., 279 494
Orth, M., 496 Pantelis, C., 106, 116
Ortiz, H., 106 Pantoni, L., 23, 394
Osborne, D., 376, 417, 426 Pantzar, A., 416, 420, 430
Osler, M., 187, 187f, 379 Paolucci, S., 509
Osman, P., 461 Papademetris, X., 105, 108, 117f
Osorio, R. S., 138 Papadimitriou, G. M., 279
Ossenkoppele, R., 37, 43, 45, 48 Papassotiropoulos, A., 417, 418, 422, 426
Ossher, L., 118 Papenberg, G., 6, 416f, 422, 424, 425,
Østby, Y., 76, 80, 131, 135, 139, 140, 141, 428, 430
142f, 143f, 144, 159, 446 Papp, K. V., 260t, 261t, 493, 518
Ostergaard, L., 24, 513 Pappas, N., 50, 52, 53
Ostling, S., 429 Pappata, S., 19
Otani, H., 347 Paranthaman, R., 514
Otnass, M. K., 284 Parasuraman, R., 119, 207, 211, 376, 428,
Otsuka, T., 52 472, 473
Otten, L. J., 315, 337 Pardo, L. M., 419
Otto, T., 276 Pare-Blagoev, E. J., 319
Oudkerk, M., 22, 23, 89, 90 Pariante, G. M., 308
Ourselin, S., 487, 494 Parisi, J. E., 491
Overbye, K., 131, 144 Parisi, J. M., 465, 466
Ovod, V., 487 Park, D. C., 6, 25, 26, 27, 43, 87, 108, 110,
Owen, A. M., 107 114, 117f, 118, 145, 155, 157, 161, 162,
Owen, C. J., 487 185, 186, 188, 189f, 195, 209, 211, 212,
Owen, N., 453 214, 218, 236, 239, 240, 241, 250, 251,
Oyanagi, K., 418 251f, 304, 306t, 314t, 316, 316f, 319, 324,
Oztekin, I., 246 325, 326, 363, 364f, 366, 367t, 369, 370,
372, 373, 374f, 375, 375f, 376, 377, 379,
Pa, J., 173, 176, 223 391, 403, 404, 405f, 465, 473, 490,
Pacheco, J. L., 42, 159, 173, 174, 492 494, 496
Padilla, D., 223 Park, H., 306t, 314t, 316, 316f, 367t, 369,
Paez, P. M., 91 370, 374, 374f, 376, 446
Pagliaro, T. A., 40 Park, J., 114, 245, 304, 373
Paglieri, C., 394 Park, R., 241, 304
Pahor, M., 447, 453 Parker, G. J. M., 75, 85
Paik, M., 391 Parker, M. G., 463
Pajevic, S., 75 Parker, N., 157
Pakkenberg, B., 76, 137, 278 Parkes, L. M., 25, 217
Palermo, F., 509 Parks, C. M., 90, 249
572
572 Author Index
Parks, C. S., 280 Pendleton, N., 90, 214, 261t, 264, 266,
Parks, E. L., 3, 84, 266 422, 514
Parnetti, L., 404 Penfield, W., 274
Parra Rodriguez, M. A., 490, 496 Peng, S. L., 27
Parra, M. A., 243 Peng, Y., 365
Pascual-Marqui, R. D., 210 Penix, L. R., 75
Pasinetti, G. M., 404 Penke, L., 81, 83, 92, 260t, 263, 264, 416
Pasotti, C., 517 Penninx, B., 453
Pasquier, F., 23, 48 Pennisi, G., 509
Pasquini, J. M., 91 Pennisi, M., 509
Passow, S., 214 Penny, W., 85
Pasternak, O., 260t Penpeci, C., 111
Patel, G., 107 Pepin, L., 493
Patel, P., 40 Pepin, L. C., 490, 491
Patel, T. A., 45, 46f, 372 Peracchi, F., 463
Patterson, B., 219, 226 Perani, D., 37
Patterson, M. B., 209 Perelman, S., 463
Pattie, A., 260t, 261, 264, 371, 419 Perez-Nievas, B. G., 491
Pattinson, K. T., 26 Perez, J., 491
Paul, L. K., 72 Pericak-Vance, M. A., 418, 420, 421
Paul, R. H., 249, 366, 453, 516 Pericakvance, M. A., 429
Paul, S. M., 138, 491, 495 Perkins, A. C., 341
Paus, T., 79, 146, 461 Perneczky, R., 37
Pausova, Z., 146 Perron, M., 146
Pavlidis, P., 146 Perrotin, A., 39, 108, 110, 117f, 118, 134,
Paxton, J. L., 219, 323 143, 491
Payer, D., 240 Perry, E. K., 279, 284
Payne, H., 210 Perry, G., 487
Payne, L. W., 398 Perry, L. M., 264
Payne, M. E., 513 Perry, M. E., 83
Payton, A., 422 Perry, R., 447
Pearce, K. A., 370 Perry, R. H., 279, 284
Pearlson, G., 514 Persina, I. S., 280, 284
Pearlson, G. D., 89, 107, 108, 110, 117f Personett, D., 279
Pearson, J. V., 417, 426 Persson, J., 48, 53, 132, 137, 144, 161, 162,
Pearson, M. A., 250 163, 164, 165, 166, 170, 174, 176, 184,
Peavy, G. M., 281 185, 188, 190, 191, 240, 249, 304, 307,
Peck, A., 41 316, 326, 373, 377, 420, 421, 424, 428
Pedersen, N. L., 367t, 370, 415, 417, 463 Perthen, J. E., 26, 27
Peduto, A., 341, 354 Peshock, R. M., 23
Pegueroles, J., 42 Peskind, E. R., 490
Pei, J. J., 462 Peters, A., 72, 76, 137, 279, 281, 287, 321
Peich, M. C., 243 Peters, D. G., 260t, 265
Peiffer, A. M., 72 Petersen, R. C., 37, 39, 41, 42, 43, 46, 47, 48,
Peitz, M., 488 118, 134, 145, 278, 372, 444, 485, 487,
Peltier, S. J., 116, 118 489, 490, 491, 492, 494
Peltz, C. B., 218 Petersen, S. E., 107, 108, 110, 114, 117f,
Pelvig, D., 137 118, 119, 373, 375, 375f
Pena-Gomez, C., 494 Peterson, B. S., 15, 26, 79, 157, 288
Pence, B. D., 444, 445f, 452, 471 Peterson, G. M., 277
573
Author Index 573
Peterson, J., 516 Pitt, E., 514

Peterson, K. M., 311t Pizzini, F. B., 137, 288
Peterson, M. S., 473, 516 Pizzolato, G., 404
Petersson, K. M., 377, 420, 421 Placentino, A., 308
Petrella, J. R., 118 Plane, J. M., 518
Petrican, R., 339 Plange, K., 280
Petrovic, K., 92, 93 Pleasure, D. E., 518
Petrovitch, H., 377, 391, 393, 403 Pluess, M., 465, 471
Pettigrew, C., 246, 247 Pocock, D., 514
Pettit, L. D., 241 Podewils, L., 447
Pfefferbaum, A., 15, 28, 73, 78, 79, 79f, 80, Podraza, K. M., 420
82, 92, 211, 216, 217, 264, 372, 448 Poels, M. M., 27
Pham, D. L., 118, 157, 516 Poggesi, A., 260t, 261t
Phelps, C. H., 493, 496 Poldrack, R. A., 195, 319
Phelps, M. E., 36 Polidori, M. C., 403
Philip, L., 394 Polimeni, J. R., 17, 39, 76
Phillips, C., 191, 311t, 312, 314t Polk, T. A., 114, 218, 241, 304, 374
Phillips, J. P., 107, 108, 110, 117f Pollock, B. G., 49, 83
Phillips, L. H., 367t, 368, 369, 370 Polster, M. R., 303
Phillossaint, M., 493 Polvikoski, T. M., 145
Piazza, P. V., 280 Poncelet, B. P., 17
Piccinin, A. M., 462 Ponds, R. W., 462, 463
Pichiule, P., 289 Ponomareva, N., 428
Pichora-Fuller, M. K., 321 Pontecorvo, M. J., 490
Pickard, N., 212 Pontifex, M. B., 380
Piefke, M., 212 Pope, D. L., 107
Pientka, L., 403 Popovic, N., 274
Pieper, C. F., 509 Poppenk, J., 108
Pierce, B. L., 428 Porras, A., 50
Pierpaoli, C., 72, 73, 75, 78 Portet, F., 471
Pietrini, P., 39, 118, 213, 219, 304, 321, 325 Posner, M. I., 472
Pietrzak, R. H., 45, 423, 490, 492 Possnert, G., 280
Piggott, M. A., 279, 284 Posternak, M. A., 507
Piguet, O., 349 Posthuma, D., 380
Pihlajamaki, M., 43, 110, 197, 288, 308, Postle, B. R., 236, 237, 249
314t, 372, 420, 421, 494 Potkin, S. G., 146
Pihlajamäki, M., 197 Potter, G. G., 73, 78, 82, 84, 266, 417,
Pike, B., 174 474, 509
Pike, G. B., 18, 19, 25, 26, 27 Potter, H., 491
Pike, K. E., 41, 46, 372, 485, 492 Potter, P., 394
Pine, D. S., 169, 172, 176 Poublanc, J., 396
Pinkham, A. E., 27 Poulet, R., 404
Piper, R. J., 89 Poulin, R., 105, 106, 108, 109
Pipingas, A., 212 Poupon, C., 19, 22
Pirogovsky, E., 281, 286 Powell, D. K., 82, 83, 260t, 266, 446
Pirraglia, E., 37, 138 Power, J. D., 106, 119
Pirtosek, Z., 516 Powless, M. R., 379
Pisljar, M., 516 Prabhakaran, V., 108, 110, 117f
Pitkala, K., 391 Pradhaban, G., 288
Pitt, B., 514 Pragnell, T. R., 26
574
574 Author Index
Prakash, R. S., 119, 217, 219, 226, 250, 444, Qi, H., 263
445f, 447, 448, 449, 450, 451, 452, 471 Qian, X., 404
Praseedom, A., 306t, 308 Qin, W., 108, 110
Preacher, K. J., 83 Qin, Y-Y., 44
Preboske, G. M., 145, 485, 490 Qiu, C., 475
Preiss, D., 26 Qualls, C., 223
Prentice, N., 514 Quesada, C. A., 48
Press, D. Z., 25 Quigley, C., 215
Preston, A. R., 288 Quilter, R. E., 259
Preti, P., 517 Quinlan, D. M., 50, 53
Preuschhof, C., 80, 212, 213, 239, 424, Quinn, B. T., 52, 139, 140, 157, 159,
425, 427 173, 174
Preuss, T. M., 141 Quinn, J. F., 490
Price, D. A., 278, 287 Quiroz, Y. T., 145, 403, 487, 494, 496
Price, J., 43, 494 Quitkin, F. M., 507
Price, J. C., 41, 45, 156, 400, 490
Price, J. F., 424 Raaijmakers, J. G., 310, 311t, 314t, 324
Price, J. L., 41, 118, 138, 139, 485, 493, 513 Rabbia, F., 394
Price, R., 485, 492 Rabbitt, P., 90, 170, 214, 261t, 264, 266, 422,
Prince, J. L., 172 461, 514
Prince, M. R., 29 Rabin, M. L., 15, 371
Prince, S. E., 112, 118, 238, 239, 302, 304, Rabinovici, G. D., 37, 40, 42, 43, 48, 118,
306t, 309, 310f, 311t, 316, 317, 318f, 324 145, 157, 470, 489, 491, 492, 494
Prins, N. D., 90, 260t, 261t, 264, 265, 393, Rabinowitz, J. C., 303, 322
395 Race, E. A., 304
Printz, H., 49 Rachakonda, S., 107, 108, 110, 117f
Prisman, E., 26 Racine, A. M., 44
Proffitt, T. M., 116 Racine, C. A., 40, 82, 219, 260t, 264,
Protasova, M., 428 266, 323
Protzner, A. B., 107, 108, 117f, 226 Radde, R., 372
Provenzale, J. M., 88, 118, 210, 260t Rademakers, R., 376
Provenzano, F. A., 43, 79, 261t, 264 Rahhal, T., 339
Pruessmann, K. P., 24 Rahm, B., 245
Pruessner, J. C., 174 Rahmati, M., 243
Pruessner, M., 174 Raichle, M. E., 23, 39, 43, 85, 106, 107, 108,
Pruis, T. A., 341, 345 110, 116, 117f, 118, 132, 138, 139, 141,
Prull, M. W., 303, 304 165, 197, 226, 250, 307, 316, 472, 487,
Przybelski, S. A., 43, 46, 118, 485, 490, 494, 510, 511
491, 494 Raichlen, D. A., 471
Pudas, S., 3, 48, 53, 87, 108, 110, 116, Raj, A., 138
117f, 132, 137, 144, 161, 162, 163, Rajagopalan, S., 514
164, 165, 166, 170, 174, 176, 184, 185, Rajah, M. N., 53, 111, 113, 162, 183, 193,
188, 190, 191, 240, 249, 304, 326, 364, 316, 326
373, 424 Raji, C. A., 146, 446, 447, 448
Puglielli, L., 376 Rajji, T. K., 83
Pujol, J., 106 Rakitin, B., 106
Punsoni, M., 280 Rakitin, B. C., 304
Pupi, A., 37 Ram, N., 116
Purnell, C., 403 Ramage, A. E., 39
Putcha, D., 42, 43, 110, 139, 492, 495 Raman, M., 159, 174
575
Author Index 575
Raman, R., 490, 496 Ready, R. E., 349

Ramani, A., 22 Reagh, Z. M., 286
Ramanna, S., 17, 22 Rebbechi, D., 304
Ramchurn, A., 223 Rebok, G. W., 465, 466
Ramey, V., 89 Rebucal, K. A., 339
Rami, L., 43, 494 Reed, A. E., 341, 345, 346
Ramirez-Amaya, V., 277 Reed, B. R., 42, 43, 90, 91, 110, 112, 138,
Ramnani, N., 85 158, 176, 248, 249, 264, 404
Ramos, L. M., 22, 23, 89 Reed, J. D., 119, 422, 425
Ramsbottom, M. J., 76, 80 Reed, T., 393, 395
Rand-Giovannetti, E., 288 Reese, T. G., 22
Randall, C., 138 Reeves, S. J., 50
Randolph, J. J., 380 Regan, C. M., 41, 487
Randolph, J. S., 380 Regeur, L., 137
Rane, S., 25 Reichenbach, J. R., 27, 92
Ranganath, C., 194, 276, 287, 302, 303, 313, Reid, R. I., 43, 46
324, 337, 340, 347 Reidler, J. S., 105, 106, 108, 109
Raniga, P., 485, 492 Reijmer, Y. D., 392
Rao, A., 75 Reijnders, J., 466
Rao, G., 279 Reiman, E. M., 37, 145, 376, 403, 404, 417,
Rao, N. K., 211, 264 426, 487, 490, 494, 496
Rao, S. M., 217, 450, 452 Reiman, R. A., 145, 487, 494
Rapcsak, S. Z., 376, 404 Reinhart, B., 260t, 264
Rapoport, J. L., 79, 461 Reinvang, I., 42, 52, 72, 138, 139, 140, 157,
Rapoport, S. I., 39, 90, 213, 219, 372, 393 158, 159, 237, 371, 376, 421
Rapp, M. A., 249 Reisberg, B., 37, 79
Rapp, P. R., 274, 278, 279, 280, 281, 282, Reischies, F. M., 249
283, 284, 286, 287 Reiser, M. F., 139, 174, 226, 514
Rapp, S. R., 471 Reishofer, G., 92
Rasch, B., 340, 417, 418, 422 Reiss, A. L., 106, 107, 118
Raskind, M. A., 490 Reiter, M. A., 134, 260t, 264, 266
Rast, P., 466 Reitz, C., 146
Rastas, S., 145 Rejeski, W. J., 451
Rauch, L. C., 346 Renbing, X., 41
Rauch, S. L., 308 Renken, R. J., 108, 110, 111, 117f, 210
Raue, P. J., 517, 519 Rentz, D. M., 41, 42, 43, 44, 45, 48, 161,
Raven, E. P., 259, 260t, 266 197, 215, 288, 308, 314t, 372, 485, 489,
Ravert, H. T., 49 490, 491, 492, 493, 495, 496
Rawles, J. M., 15 Repovs, G., 106, 516
Rawson, K. S., 430 Reppermund, S., 157, 260t, 264
Ray, K. L., 107 Resnick, S. M., 118, 132, 156, 157, 161, 163,
Ray, R. D., 341, 345 165, 171, 172, 377, 394, 397, 490
Raye, C. L., 309 Reuben, A., 304
Raz, N., 15, 22, 28, 42, 48, 72, 82, 83, 89, Reuter-Lorenz, P. A., 4, 87, 145, 162, 185,
90, 91, 92, 93, 118, 138, 157, 158, 159, 209, 211, 212, 214, 236, 238, 239, 240,
184, 188, 217, 224, 237, 238, 249, 260t, 240f, 241, 241f, 242f, 245, 248, 251, 251f,
261, 264, 266, 278, 303, 305, 313, 321, 288, 304, 307, 316, 363, 364f, 373, 376,
339, 367t, 368, 369, 371, 372, 377, 390, 379, 400, 473
392, 392, 393, 394, 395, 397, 398, 416, Reuter-Lorenz, Q., 87
443, 444, 447, 452, 461, 472, 516 Reuter, E. M., 218
576
576 Author Index
Reuter, M., 169, 172, 176 Roberts, D. R., 261t, 266

Reuter, S., 174 Roberts, J., 392
Reyes, S., 260t Roberts, J. A., 278
Reynolds, C. A., 367t, 370, 415, 417 Roberts, J. M., 286
Reynolds, C. F., 49, 116 Roberts, N., 313
Reynolds, J. R., 106 Roberts, R., 485, 490
Reynolds, S. L., 259 Roberts, R. O., 37, 39, 42, 485, 489, 491, 492
Riboldi, F., 466 Robertson, I. H., 215, 219, 466
Ribot, T., 373 Robin, D. A., 39, 260t, 261t
Ricciardi, G. K., 137 Robins, J. M., 463
Rich, P., 110 Robins, P., 485, 492
Richard, A., 260t, 265 Robitaille, A., 462
Richards, M., 463 Robitsek, R. J., 288
Richards, S. S., 392 Roc, A. C., 24
Richardson, M. P., 340 Rocca, W. A., 145, 475, 485, 489
Richer, L., 146 Roche, F., 393
Ridderinkhof, K. R., 210 Rockstroh, B., 516
Riddle, D. R., 278 Roddey, J. C., 134, 146
Riddle, W., 36 Rodgers, B., 463
Ridgway, G. R., 44 Rodnick, M. E., 48
Rieck, J. R., 304, 372, 374, 405, 405f, 490 Rodrigue, K. M., 6, 15, 25, 26, 27, 28, 43,
Rieckmann, A., 3, 39, 43, 46, 47, 47f, 48, 50, 48, 89, 93, 118, 157, 158, 188, 189f, 195,
52, 53, 54, 111, 112, 116, 118, 134, 135, 237, 239, 241, 245, 250, 278, 303, 304,
174, 263, 264, 349, 354, 428 306t, 313, 314t, 316, 316f, 321, 325, 326,
Riederer, P., 27, 49, 91, 92 367t, 368, 369, 370, 371, 373, 374, 374f,
Riedl, V., 39 375, 377, 390, 392, , 393, 395, 397, 403,
Ries, M. L., 260t, 264, 376, 420, 421 404, 405f, 416, 444, 461, 472, 490,
Rietschel, M., 146 494, 516
Riklund, K., 50, 52, 84, 88, 116, 132, 137, Rodriguez-Fornells, A., 43
145, 239, 415, 417, 423, 473, 474 Rodriguez, C., 219
Rimajova, M., 485, 492 Rodriguez, G., 396
Rimmler, J. B., 429 Roe, C. M., 37, 485, 491, 492
Ringelstein, J. G., 514 Roe, D., 366
Ringman, J. M., 487, 494, 495 Rogaev, E., 428
Rinne, J. O., 474 Roger, V. L., 389
Rintoul, J. L., 472 Roh, J. H., 138
Ripolles, P., 43 Rohlfing, T., 28, 79, 82, 92, 211, 264, 372
Risch, N., 418 Rohwedder, S., 463
Rissman, J., 219, 220f Roiser, J., 424
Rissman, R. A., 25 Rol, M. A., 274
Ritchey, M., 113, 114, 340, 341, 344, 345 Rolle, C., 472
Ritchie, K., 395 Rolls, E. T., 277
Rizio, A. A., 225, 316, 319 Rombouts, S. A., 43, 108, 110, 117f, 135,
Roach, A. E., 90, 91, 260t, 264, 287 170, 176, 226, 310, 311t, 314t, 324
Roalf, D. R., 341, 345 Rönn, T., 429
Robb, W. G. K., 313, 317 Rönnlund, M., 48, 84, 170, 185, 186, 367t,
Robbins, M. A., 279, 377, 390, 391, 392 368, 369, 370, 415, 471
Robbins, P. A., 26 Roontiva, A., 145, 487
Robbins, T. W., 424 Roose, S. P., 509
Roberts, B. A., 463 Rootwelt, H., 110, 376, 421
577
Author Index 577
Ropele, S., 92, 93 Rubin, D. C., 112, 168

Rordorf, G., 24 Rubin, E. H., 41, 485
Rosa-Neto, P., 78 Rubin, S. R., 303, 453
Rosadini, G., 396 Rudin, M., 43, 44
Rosand, J., 41 Rudolph, J. L., 395
Rosano, C., 446, 447, 448, 453 Rudrauf, D., 72, 514
Rosas, D., 173, 174 Rudy, J. W., 276
Rosas, H. D., 25, 42, 78, 139, 172, 372, Rugg, M. D., 3, 185, 190, 191, 192f, 193,
492, 496 193f, 194, 194f, 195, 196, 197, 198f, 302,
Rose, N. S., 242 304, 306t, 308, 311t, 312, 313, 314t, 315,
Rosen, A. C., 304 317, 324, 325, 337
Rosen, B. R., 18, 19, 24, 25, 26, 27, 39, 76, Ruidavets, J. B., 377, 463
78, 308 Ruiz, A., 494
Rosen, H. J., 40 Rundek, T., 395
Rosenbaum, A. E., 49 Ruscheweyh, R., 443, 444
Rosenbaum, J. F., 507 Rush, A. J., 507
Rosenberg, H. F., 303 Rusinek, H., 138
Rosenberg, R., 513 Russell, J., 76, 80
Rosenbloom, M. J., 216 Russell, J. A., 339
Rosendahl, E., 512 Rusted, J. M., 249
Rosene, D. L., 137, 279, 281 Ruth, T. J. T., 52
Roses, A. D., 429 Rutledge, S., 286
Rosnick, C. B., 418 Rutman, A. M., 216
Ross, G. W., 393, 403 Rutten-Jacobs, L. C. A., 43
Ross, M., 349 Ryan, C. M., 390, 393, 395, 396, 397, 400
Ross, R. S., 288, 453 Ryan, L., 112, 281, 302
Ross, T. J., 217 Ryberg, M., 161
Rossi, M., 280, 284 Rykhlevskaia, E. I., 448
Rossor, M. N., 44, 157, 487, 494, 495 Ryman, A., 514
Rostrup, E., 174, 187, 187f, 444 Rypma, B., 19, 25, 26, 73, 78, 82, 87, 88, 140,
Rosvold, H., 49 211, 239, 250, 264, 304, 400
Rotello, C. M., 349
Routhieaux, B. C., 303 Saad, Z. S., 169, 172, 176
Rovio, S., 380, 446, 447, 470 Sabatier, R., 395
Rowan, M. J., 41, 487 Sabatini, B. L., 41, 487
Rowe, C. C., 41, 42, 45, 46, 372, 423, 485, Sabbagh, M. N., 376, 404, 490
487, 490, 492, 494, 496 Sabbatini, M., 393
Rowe, G., 245 Sabharwal, J., 113, 219, 222
Rowe, J. B., 188, 241, 250, 325 Sabia, S., 377, 378
Rowley, H. A., 44, 260t, 264, 396, 420, 421 Sable, J. J., 245
Roy, M., 346 Sabuncu, M. R., 169, 172, 176
Roy, R. R., 453 Sacco, R. L., 395
Royall, D. R., 39, 260t, 261t Sachdev, P., 157, 260t, 263, 264, 451
Royle, N. A., 89, 260t, 263, 264, 446, 447, Sachdev, P. S., 393
448, 471 Sachdev, Q., 84
Rozkalne, A., 137 Sachs, R., 41, 43, 118
Ruano, S., 284 Sackeim, H. A., 471, 507, 509
Rub, U., 288 Sacuiu, S., 429
Rubenfire, M., 514 Saczynski, J. S., 380
Rubens, M. T., 113, 216, 219, 222, 223 Sadek, J. R., 107, 108, 110, 117f
578
578 Author Index
Sadowsky, C. H., 490 Sander, M. C., 212, 220, 237, 242, 243,
Saft, C., 496 244, 245
Sager, M. A., 44, 90, 378, 420, 421 Sander, T., 249, 422, 424
Sahara, N., 493 Sanders, A. L., 162, 250
Sahlas, D. J., 513 Santacruz, A. M., 495
Saido, T. C., 493 Santelli, L., 72
Sakaki, M., 340, 341, 345 Santo, N. M., 454
Sala-Llonch, R., 108, 113, 117f, 494 Santos, M., 37, 48
Sala, I., 42 Sanz-Arigita, E. J., 43, 108, 110, 117f
Salami, A., 39, 48, 53, 81, 83, 84, 87, 108, Sapiro, G., 22
110, 111, 112, 116, 117f, 159, 161, 162, Sapkota, S., 428
163, 164, 166, 170, 174, 184, 185, 188, Sasson, E., 260t
190, 191, 240, 260t, 264, 266, 326, Sato, K., 157, 263
421, 472 Sato, S., 39
Salarirad, S., 397 Sato, T., 287
Salat, C. E., 288 Satz, P., 474
Salat, D. H., 25, 41, 42, 52, 72, 73, 75, 78, Sauerland, S., 17
80, 82, 131, 134, 137, 139, 140, 144, 157, Saults, J. S., 243
159, 173, 174, 237, 260t, 264, 266, 372, Saunders, A. M., 418, 420, 421, 429
373, 393, 395, 492, 496, 513 Saunders, W. B., 512
Salehi, A., 422, 423 Sauvage, M., 288
Salehpour, M., 280 Sava, S., 279
Salerno, J. A., 39, 90, 213, 219, 304, 321, Savage, A., 241, 304
372, 393 Savage, G., 41, 45, 490, 492
Saliasi, E., 108, 110, 111, 117f, 210 Savaki, H., 37
Salisbury, D. F., 171 Savalia, N. K., 108, 110, 114, 117f, 118, 373,
Salloway, S., 40, 42, 487, 492, 494, 495 375, 375f
Salmon, D. P., 281, 420, 421, 490, 496 Saver, J. L., 27
Salmon, E., 37, 39, 191, 311t, 312, 314t Saverino, C., 108, 109, 109f, 111f, 114,
Salmon, J., 453 117f, 119
Salthouse, T. A., 46, 46f, 52, 82, 83, 84, 88, Savitcheva, I., 156
89, 90, 155, 157, 158, 174, 183, 184, 208, Sawada, T., 514
211, 219, 259, 263, 264, 266, 267, 301, Sawaguchi, T., 49
366, 367t, 369, 371, 463, 464, 474 Saxby, B. K., 391, 518
Salvado, O., 42, 372, 485, 487, 492 Saxton, J. A., 45, 454, 490
Salvatore, M. E., 284 Saykin, A., 494
Samanez-Larkin, G. R., 116, 264 Saykin, A. J., 134, 146, 487
Sambataro, F., 108, 113, 114, 117f, 119, 226, Sayres, R., 170
317, 324, 341, 349, 351, 417, 421, 422, Scahill, R. I., 42
425, 426, 427f, 496 Scalaidhe, S. P., 281
Sammer, G., 212, 241 Scalf, P. E., 444, 447, 448, 449, 473, 516
Samsonovich, A., 112, 302 Scalf, Q., 88
Sanchez-Aldeguer, J., 240 Scarmeas, N., 471
Sanchez-Juan, P., 419 Schacter, D. L., 106, 107, 110, 112, 114, 119,
Sánchez-Saudinos, M-B., 42 170, 176, 273, 281, 306t, 307, 308, 311t,
Sanchez-Valle, R., 494 312, 315, 340, 341, 343, 344, 346
Sanchez, M. M., 422, 423 Schäfer, S., 462, 464, 465, 470, 472, 473,
Sandell, J., 40, 156 474, 475
Sander, C. Y., 39 Schaie, K. W., 52, 134, 155, 250, 259, 366,
Sander, D., 340, 344 367t, 368, 369, 370, 415, 461, 462, 463
579
Author Index 579
Schapiro, M. B., 39, 118, 213, 219, 372, 393 Schoenfeld, R., 283
Scharfman, H. E., 422 Schoepp, D. D., 138, 495
Schatzberg, A. F., 106, 107 Schofield, P. R., 487, 494, 495
Schechtman, K., 513 Scholte, H. S., 263
Scheckenberger, M., 39 Scholz, J., 472
Scheff, S. W., 278, 287, 289 Schon, K., 288
Schefter, M., 353 Schonhaut, D. R., 37, 48
Scheibe, S., 116, 337 Schönknecht, P., 37
Scheinost, D., 105, 108, 117f Schooler, C., 462, 463, 464
Scheller, E., 246 Schork, N. J., 134, 146
Scheltens, P., 23, 43, 45, 108, 110, 117f, 135, Schott, B. H., 52
145, 226, 260t, 261t, 393, 447, 487, 491 Schott, J. M., 44, 492
Schepens-Franke, A. N., 43 Schou, M., 40
Scherder, E. J. A., 260t, 261t Schouten, J. L., 325
Scherer, K. R., 339 Schraedley-Desmond, P., 170
Scherr, P. A., 391, 491 Schreiber, S., 394
Scheurer, E., 92 Schretlen, D. J, 89
Scheurich, A., 39, 82, 217 Schroder, J., 424, 425, 427, 428
Schiavone, F., 83, 174, 260t, 264 Schroder, L., 341
Schiepers, O. J., 419 Schroeter, A., 43, 44
Schiff, S., 215, 225 Schroeter, M. L., 218
Schildkraut, J. J., 514 Schroth, G., 287
Schimmack, U., 339 Schryer, E., 349
Schjeide, B. M., 425, 428 Schuck, N. W., 425
Schlaggar, B. L., 107 Schuff, N., 23, 158, 492
Schlosser, A. E., 260t, 261t Schuierer, G., 471
Schlyer, D., 50 Schulte, T., 211, 216
Schlyer, D. J., 49, 50, 51f Schultz, A. P., 39, 42, 43, 44, 45, 46, 47, 47f,
Schmader, K. E., 429 48, 52, 108, 110, 117f, 118, 134, 135, 139,
Schmahmann, J. D., 264, 513 174, 489, 490, 492, 494
Schmansky, N. J., 172 Schultz, M. S., 345, 346
Schmechel, D. E., 429 Schulz, R., 507
Schmid, C. H., 466 Schulze, L., 215
Schmidt, M. E., 134 Schulzer, M. M., 52
Schmidt, R., 92, 93, 260t, 264 Schumacher, M., 260t, 264
Schmidt, R. E., 76, 80 Schumann, G., 146
Schmiedek, F., 461, 462, 464, 465, 466, 470, Schutze, H., 306t, 308, 314t, 324, 472, 474
472, 473, 474, 475 Schuur, M., 419
Schmitt, F. A., 278, 287, 289, 491 Schwab, L. C., 425
Schmitz, F., 217 Schwab, S. G., 425
Schmitz, T. W., 223, 420 Schwamm, L. H., 24
Schnack, H. G., 158, 174, 198 Schwartz, J. E., 114
Schneider-Garces, N. J., 238, 239 Schwartz, R. B., 514
Schneider, A., 425 Schwartz, W. J., 37
Schneider, B. A., 321 Schwartz, Y., 146
Schneider, J. A., 161, 372, 471, 485, 491, Schwartzman, A. N., 52
492, 493 Schwarz, C. G., 43, 44, 46, 287
Schnell, E., 284 Schweser, F., 92
Schnyer, D. M., 395 Scoriels, L., 424
Schoenberg, B. S., 403 Scott, M., 90, 214, 261t, 264, 266, 509, 514
580
580 Author Index
Scott, M. L., 514 Shah, A. R., 145, 485, 487, 494

Scott, P. J. H., 48 Shah, K., 444
Scotton, T. C., 491 Shah, N. S., 403
Scoville, W. B., 274 Shah, R., 485
Scullin, M. K., 394 Shallice, T., 195
Seamans, J. K., 49 Shamy, J. L., 283
Sebastian, A., 246 Shanahan, L., 515
Section, J., 25, 26, 27, 188, 189f, 195, Shankar, G. M., 41, 487
326, 403 Shankle, W. R., 493
Sedvall, G., 50 Shannon, B. J., 41, 43, 118
Seeley, W. W., 106, 107, 118 Shao, L., 40
Seeman, P., 50 Shao, X., 48
Seeman, T. E., 249, 379, 418 Shapiro, F., 485
Segall, J. M., 107, 108, 110, 117f Shapiro, H., 283
Segovia, G., 50 Shapiro, M., 282
Sehgal, V., 27 Sharp, D. J., 106, 108, 109
Seibyl, J. P., 50, 53 Shaw, J. E., 453
Seidenbecher, C. I., 52 Shaw, L. M., 41, 42, 134, 145, 372, 485, 492
Seidenberg, M., 450, 452 Shaw, P. J., 279, 284
Seidl, A. H., 71, 82 Sheehan, E., 470
Seidl, K. N., 349 Sheikh, J. I., 308
Seidler, R. D., 118 Shelby, E. S., 52
Seirup, J., 516 Sheldon, F. C., 118
Seirup, J. K., 511, 519 Sheline, Y. I., 41, 42, 43, 118, 138, 139, 485,
Sejnowski, T. J., 49 494, 509, 513, 514
Seki, K., 418 Shelton, A. L., 288
Sekiguchi, A., 311t, 312, 318 Shelton, J. T., 394
Sekuler, A. B., 111, 321 Shen, D., 23
Sekuler, R., 111, 214, 216, 321 Shen, H., 106, 108, 117f
Selen, L. P., 461 Shen, J., 282
Selim, M., 394, 396 Shen, L., 146
Selin, C., 36 Shen, X., 105, 108, 117f
Selkoe, D. J., 41, 43, 161, 197, 372, 372, Shen, Y., 72, 260t, 446, 518
487, 491, 494, 495 Shenton, M. E., 83, 171
Senjem, M. L., 37, 39, 41, 42, 43, 46, 47, Shepardson, N. E., 41, 487
145, 485, 487, 489, 490, 491, 492, 494 Shepel, J., 137
Seo, K., 514 Shepler, A. M., 111
Sepulcre, J., 43, 105, 106, 108, 109, 119, Sheridan, M. A., 198
139, 493 Sherman, P., 48
Sergeant, N., 48 Sherr, E. H., 81, 263
Serrano-Pozo, A., 491 Sherwood, A., 442
Servan-Schreiber, D., 49 Sherzai, A., 420, 494
Seshadri, S., 377, 380, 391, 453 Shi, J., 376
Sethares, C., 76, 279, 287, 321 Shi, W. X., 44
Setsompop, K., 17 Shi, X., 396
Seunarine, K. K., 76 Shibuya, H., 514
Sexton, C. E., 159f, 161, 514 Shigemune, Y., 311t, 312, 318
Seymour, T. L., 354 Shih, J. Y., 345, 346
Shafto, M. A., 188, 241, 250, 325 Shimada, H., 40, 444, 446, 493
Shagam, L., 428 Shimokata, H., 444
581
Author Index 581
Shimony, J. S., 78, 85, 446 Sinha, A., 493

Shin, D. D., 25 Sipila, M., 283
Shin, E., 238, 239 Sirey, J. A., 519
Shindledecker, R. D., 512 Siuciak, J. A., 134
Shing, Y. L., 83, 118, 224, 260t, 261, 266, Sjöblom, N., 389
393, 394, 396 Sjogren, M., 23
Shinotoh, H., 40, 493 Skaggs, W. E., 282
Shipley, M. J., 377 Skare, S., 21, 472
Shirazi, P. H., 514 Skiba, E., 279
Shire, E. H., 218 Skoog, I., 429
Shirk, S. D., 462 Skovronsky, D. M., 490
Shofer, J. S., 490 Skrede, K. K., 276
Shohamy, D., 319, 346 Skudlarska, B. A., 514
Shokouhi, S., 36 Skudlarski, P., 514
Shoup, T., 485, 493 Slabach, E. H., 209
Shu, H., 42 Slavin, M. J., 260t, 264
Shu, N., 108, 117f Sleegers, K., 43, 419
Shubert, T., 280, 444 Slessarev, M., 26
Shulman, G. L., 72, 106, 107, 226 Slifstein, M., 424
Shulman, S., 281, 303 Sliwinski, C., 488
Shuman, M., 325 Sliwinski, M., 471
Sibley, B. A., 440 Sliwinski, M. J., 84, 184
Siddiqi, Z., 279 Slotboom, J., 287
Sidhar, K., 420 Sloviter, R. S., 138, 495
Sidhu, K., 451 Sluming, V., 217
Siedlecki, K. L., 79, 261t, 264 Small, B. J., 367t, 368, 369, 378, 379, 418,
Siemers, E., 496 429, 462, 463, 464
Siette, J., 451 Small, C., 421
Sigurdson, W., 487 Small, G. W., 400, 418, 420, 421, 430
Sigurdsson, S., 393, 470 Small, S. A., 48, 279, 280, 287, 289
Sikkes, S. A., 45 Smallwood, J., 106
Sikström, S., 304 Smiljic, A., 43, 118, 494
Silberstein, R., 212 Smith, A. D., 155, 186, 366, 367t, 369,
Silbersweig, D., 513, 517 370, 391
Silva, R. F., 107, 108, 110, 117f Smith, C. B., 37
Silvestrelli, G., 404 Smith, C. D., 82, 83, 260t, 266, 289, 446
Sim, K., 159 Smith, C. E., 52
Simmons, A., 21, 264 Smith, D. E., 278
Simó, M., 43 Smith, D. M., 465
Simon, H. A., 461 Smith, E. E., 41, 264, 513
Simons, J. S., 302 Smith, E. S., 304
Simonsick, E. M., 449, 453 Smith, G., 49, 52, 53
Simpson, E. H., 424 Smith, G. E., 278
Simpson, G. V., 107 Smith, I., 41, 487
Singer, S., 339 Smith, J., 367t, 370, 448, 449
Singerman, J. D., 250 Smith, J. C., 450, 452, 471
Singh-Manoux, A., 155, 378, 379 Smith, J. D., 418
Singh, B., 110, 138, 264 Smith, L. A., 391
Singh, T., 444, 452 Smith, M. A., 487
Singleton, E. G., 461 Smith, M. R., 241, 304
582
582 Author Index
Smith, P. J., 442 Sorensen, A. G., 24

Smith, P. K., 155, 186, 366, 367t, 369, Sorensen, L., 513
370, 391 Sorg, C., 39
Smith, S. M., 17, 22, 72, 75, 106, 107, 108, Soriano-Mas, C., 106
110, 117f, 135, 143, 144, 159, 226, 377, Sotiropoulos, S. N., 22, 75
421, 487, 494 Soto, C. J., 366, 370
Smith, T. D., 278, 284 Soubelet, A., 371
Smith, W. G., 341, 344 Souder, E., 37
Smith, C. D., 491 Sourander, Q., 91–92
Smits, C. H., 462 Souren, L. E., 79
Smits, L. L., 487, 491 Sovago, J., 53
Smits, M., 25 Soveri, A., 474
Smolka, M. N., 146 Sowell, E. R., 15, 26, 79, 134, 157, 516
Snaedal, J., 380, 487 Spalding, K. L., 280
Sneed, J. R., 509 Spaniol, J., 88, 111, 118, 210, 260t, 349
Snider, J., 495 Sparks, C., 471
Snitz, B. E., 43, 45, 490, 494 Speck, C. L., 288
Snow, B. J. B., 52 Spector, N., 138
Snudden, A. L., 379 Spehar, B., 242
Snyder, A. Z., 41, 43, 72, 78, 85, 106, 107, Spence, J. S., 25, 450, 472, 473
108, 110, 116, 117f, 118, 131, 132, 138, Spencer, T. J., 313
139, 157, 159, 162, 165, 226, 250, 307, Spencer, W. D., 303, 305, 339
316, 371, 373, 446, 472, 487, 491, 494, Sperling, R. A., 6, 39, 40, 41, 42, 43, 44, 45,
511, 513 46, 47, 47f, 48, 49, 52, 108, 110, 117f,
Snyder, C. H., 376 118, 134, 135, 139, 145, 161, 165, 174,
Snyder, P., 495 176, 197, 218, 288, 308, 314t, 372, 418,
Soares, H., 134 420, 421, 487, 489, 490, 491, 492, 493,
Sodhi, A., 376 495, 496
Soederberg Miller, L. M., 367t, 368, 369, Spiers, H. J., 302
370, 379 Spinath, F. M., 417
Sofi, F., 440 Spincemaille, P., 29
Soininen, H., 378, 380, 380, 403, 446, 447, Spinks, R. A., 321
453, 470 Spirduso, W. W., 440
Sojkova, J., 157, 398, 490 Spires-Jones, T. L., 48
Sokoloff, L., 36, 37 Spiro, A., 390
Solbakk, A. K., 212 Spliet, W. G. M., 43
Soldner, J., 139, 514 Spoelgen, R., 425
Sole-Padulles, C., 240, 494 Sporns, O., 78, 79, 83, 105, 106, 108,
Solomon, A., 40 117f, 118
Sommer, J., 444 Spreng, R. N., 106, 107, 112, 114, 119, 239,
Sommer, K., 92 246, 247, 326
Sommers, M., 242 Springate, B., 260t, 261t
Song, A. W., 39, 73, 78, 80, 82, 83, 88, 92, Springer, M. V., 165, 166f
93, 139, 261, 417, 446, 474 Spulber, G., 380, 446, 447
Song, H., 280 Squire, L. R., 112, 273, 276, 283, 302
Song, J., 108, 110, 117f Sreenivasan, A., 43, 139
Song, S. K., 76, 80, 141 Srivastava, G., 118
Song, X. W., 108, 117f St. Jacques, P. L., 112, 113, 318, 318f, 324,
Sonnen, J. A., 492, 493 340, 341, 344, 347
Sorbi, S., 37 St. Laurent, M., 226, 245
583
Author Index 583
Staff, R. T., 397 Stieltjes, B., 174

Stagg, C. J., 75 Stine-Morrow, E. A., 465
Stahl, D., 263 Stockton, S. D., 404
Stam, C. J., 108, 110, 117f, 135, 226 Stoessl, A. J. A., 52
Stamatakis, E. A., 138, 146, 240 Stollery, B., 461
Stamler, J., 389 Stone, A. A., 114
Stamler, R., 389 Storandt, M., 41, 485, 490, 492
Stanczak, L., 239 Stormer, V., 211
Stanek, K. M., 453 Stormer, V. S., 213
Stanley, D., 325 Storsve, A. B., 131, 144, 159f, 161
Stanley, J., 260t, 261t Stough, C., 212
Stanton, M., 465, 471 Stout, J. C., 446
Stark, C. E. L., 4, 183, 277, 286, 286f, 287, Strain, J. F., 25, 472, 473
288, 496 Stranahan, A. M., 280, 281
Stark, S. M., 4, 277, 286, 286f, 287, 288, 496 Strandberg, T., 391
Starkman, S., 27 Strange, B., 340
Starn, C. E., 288 Strange, B. A., 108, 340
Starr, J. M., 81, 83, 92, 186, 260t, 261, 263, Strassburger, T. L., 393
264, 337, 371, 397, 419, 446, 447, 448, Straub, R. E., 424
462, 471 Strauman, T. A., 442
Staudinger, U. M., 367t, 370, 391, 473 Strauss, M. E., 209
Staufenbiel, M., 372 Strauss, M. M., 308
Stefansson, K., 487 Strenziok, M., 119, 472, 473
Stefansson, H., 487 Stricker, N. H., 137
Steffenach, H. A., 284 Strittmatter, W. J., 429
Steffener, J., 106, 113, 304, 396 Strobel, C., 39
Steffens, D. C., 509, 513, 514 Strocchi, P., 393
Stehouwer, C. D., 392 Strohle, A., 146
Stein Merkin, S., 379 Strojwas, M. H., 418, 420, 421
Stein, E. A., 217, 472 Strother, S., 106
Stein, J. L., 146, 447, 448 Strother, S. C., 107, 108, 117f, 226
Stein, T. D., 491 Stroup, T. S., 278
Steinberg, S., 487 Strube, M. J., 242
Stemmelin, J., 279, 284 Struve, M., 146
Stenset, V., 159 Stuart, K., 424
Stephan, D. A., 146, 417, 426 Stubbe, J. H., 380
Stephen, J. M., 223 Stuber, M., 24
Stern, C. E., 145, 288, 487, 494 Studenski, S., 447
Stern, Y., 106, 113, 118, 209, 211, 212, 240, Studholme, C., 23
283, 287, 304, 346, 379, 391, 471, 474, Stuebing, K. K., 185
490, 491, 496, 514 Stuss, D. T., 116, 239
Stevens, A. A., 341, 345 Styner, M., 146, 448, 449
Stevens, J. M., 157 Su, H., 493
Stevens, M., 107, 108, 110, 117f Su, L., 106, 108, 117f
Stevens, W. D., 106, 107, 110, 119, 247, Su, M. Y., 493
316, 319 Su, Y., 487
Stevenson, R., 286 Suchow, J. W., 244
Stewart, C., 495 Suchy, Y., 170
Stewart, F. R., 43, 138 Sudo, Y., 49
Stewart, M. C., 424 Suga, M., 159
584
584 Author Index
Sugaya, K., 279 Szardenings, A. K., 48, 493

Suhara, T., 49, 52, 54, 493, 514 Szczepanik, J., 393
Sulem, P., 487 Szklo, M., 377, 392, 393
Sulkava, R., 145 Szoeke, C., 45, 46, 372, 485, 487, 490, 492
Sullivan, C., 42, 43, 139, 490, 491, 494
Sullivan, E. V., 15, 28, 73, 78, 79, 79f, 80, Tabares, P., 424
82, 92, 159, 211, 216, 217, 264, 372, 448 Tabet, N., 249
Sullivan, L. M., 392, 453 Tago, T., 48
Sultzer, D., 447 Tahmasian, M., 39
Sum, M. Y., 159 Tai, H. C., 491
Summers, M. J., 424 Takahashi, H., 52, 418
Summers, P. E., 72, 78, 82, 260t, 261, Takahashi, T., 514
319, 472 Takano, A., 49
Sun, F., 37 Takano, H., 40, 52, 493
Sun, S. W., 44, 76, 80, 141 Takao, H., 173, 174, 176
Sunaert, S., 43 Takashima, A., 493
Sundararajan, R., 428 Takeda, A., 487
Sunderland, T., 376, 421 Takeda, M., 514
Sundseth, O., 471, 472 Taki, Y., 39, 157, 263, 311t, 312, 318, 393
Suo, C., 157 Talmi, D., 337
Supekar, K., 85, 118 Talsma, D., 210
Surguladze, S., 308 Talukdar, T., 43, 139, 493
Surina, A., 347 Tambini, A., 110
Suster, M. S., 282 Tamnes, C. K., 23, 76, 80, 85, 131, 135, 139,
Sutcliffe, C., 509 140, 141, 142f, 143, 143f, 144, 159, 159f,
Sutherland, G. R., 302 161, 446
Sutherland, M. R., 340 Tan, H. Y., 108, 113, 114, 117f, 119, 226,
Sutherland, R. J., 514 317, 324, 341, 349, 351, 417, 425, 426,
Sutton-Tyrrell, K., 400 427f
Sutton, B., 240 Tanada, S., 49
Sutton, B. P., 238, 239, 319 Tanadini, M., 43, 44
Suzuki, A., 145, 223, 304 Tanaka, H., 378, 418
Suzuki, K., 54 Tang, J., 28
Suzuki, M., 196 Tang, M. X., 118
Suzuki, T., 444, 446 Tang, Y., 76, 157, 491
Suzuki, W. A., 287 Tang, Y. Y., 472
Svensson, S., 40 Tangalos, E. G., 278
Svoboda, E., 339 Tanila, H., 279, 280, 283, 286
Swahn, B. M., 40 Tank, D. W., 17
Swan, G. E., 390, 391, 392, 393, 395, 396 Tanzi, R. E., 288, 420, 421, 429, 488, 494
Swanberg, K., 344 Tariot, P. N., 145, 487, 494, 496
Swarny, B. R., 134, 260t, 264, 266 Tarrasch, R., 260t
Swartz, R. H., 513 Tarumi, T., 378
Sweatt, D. J., 430 Taskinen, J., 424
Sweet, R. A., 417 Tate, D. F., 249, 516
Swerdlow, R. H., 36, 447, 448 Tatemichi, T. K., 391
Switalski, R., 37 Tateno, Y., 54
Sylvester, C-Y. C., 236, 238 Tatschida, T., 50
Szabo, A. N., 119, 250, 444, 445f, 449, 450, Tayebati, S. K., 394
451, 452, 471 Taylor, C., 420, 494
585
Author Index 585
Taylor, C. A., 447 Thompson, W. K., 492

Taylor, J. L., 422, 423 Thomsen, C., 444
Taylor, J. R., 83, 188, 241, 250, 325 Thomson, A., 263
Taylor, W. D., 509, 513, 514 Thornton-Wells, T., 25
Teh, I., 159 Thornton, A. E., 303, 516
Tehrani, E., 513 Thorp, J. M., 398
Teichberg, D., 39 Thorslund, M., 463
Teipel, S., 226 Thorsteinsdottir, U., 487
Teipel, S. J., 139, 174, 514 Thorvaldsson, V., 429
Telesford, Q., 451 Thyreau, B., 157
Tempelmann, C., 472, 474 Tian, Q., 449
Ten Brinke, L. F., 446, 454 Tiemeier, H., 395, 514
ten Dam, V. H., 260t, 265 Tilgmann, C., 424
Tennstedt, S. L., 465 Tilley, C., 403
Terry, D. P., 346 Tilvis, R., 391
Terry, R., 41 Tindle, H. A., 471
Teshale, S., 365, 366, 378 Tingus, K., 259, 260t, 265, 266
Tessitore, A., 162, 239, 341, 344 Tirado, V., 145, 487, 494
Teunissen, C. E., 487, 491 Tisdall, D., 76
Teuscher, U., 107, 108, 110, 117f Tishler, T. A., 83
Thacker, N., 90, 214, 261t, 264, 266, 422, Tisserand, D. J., 196
509, 514 Tochon-Danguy, H., 485, 492
Thal, D. R., 372, 492, 493 Todorich, B., 91
Thal, L. J., 420, 421 Toepper, M., 212, 241
Thambisetty, M., 172, 377, 398 Tofts, P. S., 249
Thangavel, A., 216 Toga, A. W., 15, 26, 42, 134, 146, 157, 417,
Thase, M. E., 507 447, 448, 516
Thayer, J. F., 391 Tojanowski, J. Q., 284
Thiebaut de Schotten, M., 72 Tokoglu, F., 105, 108, 117f
Thiel, B. W., 260t, 264 Tolboom, N., 43
Thiel, C. M., 306t, 309, 311t, 313, 314t Tolentino, J. C., 281
Thies, B., 493, 496 Tolnay, M., 372
Thilers, P., 53 Tomasi, D., 108, 117f, 139
Thiran, J. P., 118, 448 Tomassini, V., 72
Thiyyagura, P., 145, 487 Tomassoni, D., 394, 404
Thoma, R. J., 107, 108, 110, 117f Toner, C. K., 281, 286
Thomann, A. K., 496 Tong, K. A., 27, 92
Thomann, P. A., 496 Torgerson, B. M., 420
Thomas, A. J., 263, 266 Tornberg, A., 429
Thomas, A. K., 466 Toro, R., 106
Thomas, C. W., 321 Torre, P., 404
Thomas, G. P., 446, 447, 448, 454 Torres-Alemán, L., 453
Thomas, J. B., 110, 494 Tosakulwong, N., 492
Thomas, R. G., 490, 495, 496 Tosto, G., 43, 44, 146
Thompson, A. J., 85 Tosun, D., 42
Thompson, J. C., 119, 472, 473 Tottenham, N., 283
Thompson, P., 42, 288, 494 Touyz, R. M., 389
Thompson, P. M., 15, 26, 42, 83, 146, Townsend, D. W., 400
157, 159, 259, 260t, 265, 266, 417, 446, Toy, B., 216
487, 494 Tracey, I., 26
586
586 Author Index
Trachtenberg, A. J., 421 Turner, J. A., 107, 108, 110, 117f

Trahan, L. H., 185 Turner, M., 75
Tranel, D., 72, 339, 340 Turner, R., 17, 75
Tranh, M., 39, 90, 372 Turner, T., 341
Tranter, L. J., 465 Turtzo, L. C., 75
Treadway, M. T., 52 Tuscher, O., 217
Trejo, J. L., 453 Tuszynski, M. H., 278
Treves, A., 277 Twisk, J., 45
Trinkaus, K., 76, 80 Tye-Murray, N., 242
Tripodis, Y., 390, 391 Tyler, L. K., 138, 146, 188, 240, 241,
Triss, S. V., 398 250, 325
Trivedi, M. A., 378, 420 Tyring, S., 518
Trivedi, M. H., 507 Tzourio-Mazoyer, N., 421
Trojanowski, J. Q., 41, 42, 134, 145, 372, Tzourio, C., 168, 169, 173, 392, 395, 421, 475
485, 492, 493
Trollor, J., 157 Uchida, S., 39
Troncoso, J. C., 492, 493 Uddin, L. Q., 106, 116
Trow, P., 466 Ueki, A., 418
Trujillo, C., 194 Uemura, E., 278
Truran, D., 492 Uemura, K., 446
Tryambake, D., 398 Ugurbil, K., 17, 22
Tsai, N., 470 Uh, J., 25, 27, 325
Tsai, W. Y., 118, 287 Ullen, F., 261t, 472
Tse, C. Y., 214 Ulmanen, I., 424
Tseng, B., 447, 449 Ulrich, J., 72
Tsetanov, K. A., 241, 250 Uncapher, M. R., 302, 307, 315, 321
Tsiouris, A. J., 29 Ungerleider, L. G., 213, 219, 304, 321
Tsopelas, N. D., 490 Unterrainer, J. M., 245
Tsuchiya, H., 209 Unverzagt, F. W., 462, 465
Tsui, W. H., 37, 138 Uprety, A. R., 280
Tsuji, I., 39 Upton, N., 215
Tsuji, S., 418 Urry, H. L., 346
Tsukiura, T., 302, 311t, 312, 318 Uylings, H. B., 83, 118, 260t, 261, 266, 278,
Tsutsumimoto, K., 446 281, 394, 395
Tsvetanov, K. A., 188, 325
Tuch, D. S., 22, 78, 139, 279, 372, 496 Vaccarino, V., 378
Tucker-Drob, E. M., 415, 417 Vaden, R. J., 212
Tucker, A. M., 346, 379 Vaidya, C. J., 80, 81, 105, 266
Tuladhar, A. M., 448 Vaishnavi, S. N., 513
Tulving, E., 195, 301, 304 Vakil, E., 441, 442
Tun, P. A., 249, 379, 379 Valdes Hernandez, M. C., 89, 92, 260t, 263,
Tuomilehto, J., 378, 380, 380, 403, 446, 447, 264, 446, 471
453, 470 Valdes Hernandez, M. D., 260t
Turan, B., 116 Valecchi, D., 440
Turcotte, J., 245 Valenzuela, M. J., 424, 451
Turetsky, B. I., 341 Vallabhajosula, S., 138
Turken, A. U., 214, 260t Vallesi, A., 239
Turkheimer, E., 430 Valli, I., 422
Turkington, T. G., 210 Valls-Pedret, C., 108, 113, 117f
Turner, G. R., 106, 119, 246, 247 Valls-Sole, J., 240
587
Author Index 587
van Assema, D. M. E., 43 Van Strien, J. W., 345

van Berckel, B. N., 45 Van Strien, N. M., 276
van Berckel, B. N. M., 43 Van Swieten, J. C., 419
van Boxtel, M. P., 83, 118, 260t, 261, 264, Van Tilburg, T., 462
266, 394, 395, 396, 462, 463, 466 van Velzen, J., 223
Van Broeckhoven, C., 43, 377, 420, 421 van Zijl, P. C., 19, 134, 146
van Buchem, M. A., 110, 170, 176, 260t, 265 Vandenberg, P. M., 514
van de Pol, L. A., 42 Vandenberghe, R., 43, 145
van de Wiel, L., 110 Vanderstichele, H., 40, 485
Van Den Berg, E., 390, 392 Vangel, M., 496
van den Heuvel, D. M. J., 260t, 265 Vangu, M. D., 514
van den Heuvel, M. P., 85, 87, 107, 118, 514 Vanmechelen, E., 40, 485
van den Wildenberg, W. P., 427 Vannini, P., 42, 43, 494
van der Berg, E., 264 Vannorsdall, T. D., 89
van der Elst, W., 260t, 264, 266, 395, 463 Vannucci, S., 289
van der Flier, W. M., 43, 45, 487, 491 Varghese, R., 26
Van Der Graaf, Y., 397, 399f Varma, V., 401, 401f
van der Grond, J., 110 Varnäs, K., 40
van der Kouwe, A. J., 78, 134, 144, 173, Varnum, M. E., 367t, 370
174, 279 Vasan, R. S., 377, 391
van der Lijn, F., 23 Vasdev, N., 493
van der Lugt, A., 23, 27, 89, 90, 260t, 395, Vasic, N., 496
448 Vasile, R. G., 514
van der Pols, J. C., 191 Vaughan, L., 248, 471
van der Veen, F. M., 196 Vaupel, J. W., 155
van Dijk, E. J., 90, 260t, 261t, 264, 265, Vaynman, S., 453
393, 448 Vecchione, C., 404
Van Dijk, K. R. A., 43, 76, 108, 110, 117f, Vechlekar, C. D., 451
118, 139, 218, 494 Veglio, F., 394
Van Duijn, C. M., 418, 419 Veillette, S., 146
van Dyck, C. H., 50, 53 Veitch, D. P., 134
Van Eldik, L. J., 289, 491 Vela, J., 284
Van Essen, D. C., 17, 22, 107, 141 Velez-Pardo, C., 145, 403, 487, 494
Van Gelder, B. M., 380 Veltman, D. J., 310, 311t, 314t, 324
van Gijn, J., 22, 23, 89, 90 Vemuri, P., 37, 39, 41, 42, 43, 47, 118, 145,
Van Groen, T., 274 372, 485, 487, 489, 491, 492, 494
van Haren, N. E. M., 158, 174, 198 Venkatraman, V. K., 134, 447
van Harten, A. C., 487, 491 Venugopal, S., 263
van Harten, B., 260t, 261t Verber, M. D., 450
van Heertum, R. L., 471 Vercruyssen, M., 259
van Heugten, C., 466 Verdile, G., 46
van Laere, K., 43 Verfaillie, S. C., 45
Van Norden, A. G., 448 Verghese, J., 471
van Osch, M. J., 25 Vergote, D., 428
Van Petten, C., 198 Vergun, S., 108, 110, 117f
Van Popele, N. M., 514 Verhaeghen, P., 46, 46f, 225, 245, 259, 273,
van Praag, H., 280, 444 465, 470
van Rooijen, H., 263 Verhey, F. R., 260t, 264, 266
van Schijndel, R. A., 45 Vermeer, S. E., 90, 260t, 261t, 264, 265, 393
van Straaten, E. C., 145 Vermetten, E., 378
588
588 Author Index
Vernooij, M. W., 23, 27, 89, 90, 260t, 395, 448 Volk, S., 40
Verschuren, W. M., 380 Volker, K., 443, 444
Versteeg, A., 45 Volkov, P., 429
Vespa, P., 27 Volkow, N. D., 49, 50, 51f, 52, 53, 108,
Viader, F., 37, 134, 157 117f, 139
Vial, C., 280 von Cramon, D. Y., 218
Vickers, J. C., 424 von Oertzen, T. T., 48, 53, 84, 155, 188, 249,
Vidal-Pineiro, D., 108, 113, 117f 266, 422, 424
Vidal, J. S., 403 VonDras, D. D., 379
Videbech, P., 444, 513 Vonsattel, J. P., 518
Vidgren, J., 424 Vos, S. J., 489, 492
Vieira, V. J., 444, 445f, 452, 471 Voss, M. W., 119, 217, 219, 250, 444, 445f,
Vieluf, S., 218 449, 450, 451, 452, 471
Viitanen, M., 391, 392, 453 Vrenken, H., 45
Vilaplana, E., 42 Vrooman, H. A., 23, 260t, 448
Vilberg, K. L., 302, 313 Vuilleumier, P., 308
Villablanca, J. P., 27 Vuurman, E. F., 83, 118, 260t, 261, 266, 394,
Villain, N., 108, 110, 117f, 134, 492 395, 396
Villareal, D. T., 444 Vythilingam, M., 378
Villblanca, P., 260t, 265
Villemagne, V. L., 41, 42, 45, 46, 48, 49, Wadhwa, R., 473, 516
372, 423, 485, 487, 490, 492 Wadsworth, L. P., 490, 491
Villeneuve, S., 42, 470, 489, 492 Wager, T. D., 346
Villien, M., 39 Wagner, A. D., 170, 176, 288, 302, 304, 307,
Villringer, A., 249, 424 315, 321
Vilotti, C., 404 Wagner, E., 213, 219, 304, 321
Vincent, J. L., 72, 106, 107, 108, 110, 116, Wagner, H. N., 49, 54
117f, 139, 472 Wagner, H. R., 509
Vines, D., 50 Wagner, M., 174
Vinters, H. V., 493 Wagner, S. L., 488
Vinuela, F., 27 Wagster, M. V., 496
Virta, A., 75 Wahl, M., 81, 263
Virta, J., 474 Wahlin, A., 52, 53, 156, 471
Visscher, K. M., 212 Wahlin, T. B., 52, 53, 156
Visser, P. J., 487, 489, 491, 492 Wahlovd, K. B., 261t
Visseren, F. L., 397, 399f Wahlund, L. O., 23, 134, 260t, 261t, 263,
Viswanathan, A., 264, 475, 513 264, 266
Vitaioli, L., 393 Wainger, B. J., 488
Vitolo, O. V., 43, 494 Wais, P. E., 226
Vitorica, J., 284 Wake, H., 474
Vivanco, P., 274 Wakefield, D. B., 260t, 261t
Vlassenko, A., 514 Wald, L. L., 17, 76, 279
Vnek, N., 349 Waldenberger, M., 146
Voelcker-Rehage, C., 218, 444 Waldinger, R. J., 345, 346
Voets, N. L., 75 Waldorp, L. J., 263
Vogel, E. K., 218, 225, 242, 243, 245 Waldstein, S. R., 89, 390, 391
Vogel, J. W., 43, 118, 494 Walhovd, K. B., 3, 23, 41, 42, 52, 71, 72, 74f,
Vogels, R. L. C., 260t, 261t 76, 77f, 80, 84, 85, 131, 134, 135, 138,
Vogler, C., 425 139, 140, 141, 142f, 143, 143f, 144, 145,
Voineskos, A. N., 83 146, 157, 158, 159, 159f, 161, 173, 174,
Volgyesi, G., 26 197, 237, 371, 376, 446, 471, 472
589
Author Index 589
Wall, A., 156 Weaver, K. L., 282

Wallace, B. C., 466 Webb, A. G., 217, 443, 447, 449, 452
Wallace, J. L., 280 Wedeen, V. J., 17, 22, 76, 448
Wallin, A., 23, 40 Wee, E., 245
Walsh, C., 466 Weeks, J. C., 244
Walsh, D. M., 41, 487 Wehling, E., 376
Walsh, E., 430 Weickert, C. S., 424
Walsh, J., 404 Weierich, M., 340
Walsh, J. C., 48, 493 Weigand, S. D., 37, 39, 41, 42, 43, 47, 145,
Walter, M., 105 372, 485, 487, 489, 490, 492, 494
Wan, J., 172 Weiller, C., 171, 245
Wan, L., 491 Weinberger, D. R., 108, 113, 114, 117f, 119,
Wandall, B. A., 287 162, 226, 239, 317, 324, 341, 344, 349,
Wang, A., 518 351, 417, 421, 422, 425, 426, 427f
Wang, B., 473 Weiner, M. F., 490, 494, 496, 512
Wang, D. J., 25 Weiner, M. W., 23, 37, 41, 42, 43, 48, 112,
Wang, E., 493 134, 145, 146, 157, 158, 176, 288,
Wang, G. J., 50, 52, 53 372, 487, 492, 494
Wang, H. X., 462 Weiner, R. D., 512
Wang, J., 24, 146, 404 Weinstein, A. M., 250, 429, 439, 443,
Wang, J. H., 108, 117f 452, 454
Wang, L., 106, 108, 110, 112, 117f, 319, Weintraub, S., 372
494, 514 Weissfeld, L., 43, 494
Wang, M., 238 Weissfeld, L. A., 45
Wang, S., 43, 76, 118, 494 Weisskoff, R. M., 17, 18, 19, 24, 25, 26, 27
Wang, T. H., 190, 191, 192f, 193, 193f, 194, Welcome, S. E., 15, 26, 157
194f, 196, 197, 198f, 306t, 308, 312, 314t, Welmer, A. K., 470
324, 325 Welsh-Bohmer, K. A., 37, 442, 509
Wang, W. C., 304, 311t, 313, 314t, 320 Welsh, R. C., 161, 240, 306t, 314t, 324
Wang, X. J., 238 Wen, W., 84, 157, 260t, 263, 264, 393
Wang, Y. Y., 29, 40, 52, 156, 211 Wenger, E., 439, 466, 471, 472, 473
Wang, Z., 347 Wenger, K. K., 107
Ward, A., 42, 45, 48, 490 Wenzel, J., 280
Ward, A. M., 42, 43, 108, 110, 117f, 134, 494 Wenzel, M., 280
Ward, M. A., 378 Werheid, K., 353
Wardlaw, J. M., 81, 83, 89, 92, 260t, 261, Werkle-Bergner, M., 212, 220, 237, 242,
263, 264, 446, 447, 448, 471 243, 244, 245, 393
Ware, J. H., 169, 170, 172, 173, 176 Werring, D. J., 85
Waring, J. D., 317, 318, 324, 341, 344, Wesnes, K. A., 391, 518
345, 354 Wesseling, P., 43
Warnecke, T., 444 West, A. N., 283
Warner Schaie, K., 260t, 264, 266 West, M. J., 278
Warrington, E. K., 157 West, R., 116
Wartenburger, I., 214 West, R. L., 208, 215, 217
Washimi, Y., 444 Westaway, D., 428
Wasserman, B. T., 43, 79, 261t, 264 Westberg, L., 424, 428
Waters, T. E., 343 Westbrook, R. F., 451
Watkins, K. E., 106 Westendorp, R. G. J., 260t, 265
Wattez, A., 48 Westerhausen, R., 214
Watts, R. J., 487 Westerlund, I., 280
Waxman, S. G., 71, 82 Westlye, E. T., 110, 260t, 264
590
590 Author Index
Westlye, L. T., 23, 42, 72, 76, 80, 85, 108, Wilheim, R., 82, 260t, 264, 266
110, 117f, 131, 134, 135, 138, 139, 140, Wilkins, C. H., 509, 513
141, 142f, 143, 143f, 144, 157, 158, 159, Wilkinson, A. J., 223
159f, 161, 237, 371, 376, 421, 446, Willemer, C., 443, 444
471, 472 Williams, D. S., 24
Westman, E., 261t, 266 Williams, L. E., 78, 446
Wetzel, M. E., 176 Williams, L. M., 79, 119, 261t, 264, 341,
Weuve, J., 463 354, 516
Wexler, B. E., 516 Williams, N., 188, 241, 250, 325
Wey, H-Y., 39 Williams, R., 465, 471
Whalen, P. J., 308 Williams, S. C. R., 72, 78, 82, 138, 260t,
Whalley, L. J., 186, 260t, 261, 264, 337, 397 261, 319, 472
Wharton, W., 376 Williams, V. J., 137, 260t, 264, 266,
Wheatley, B. M., 141 393, 395
Wheeler-Kingshott, C. A., 76 Williamson, A., 15, 157, 158, 278, 303, 313,
Wheeler, D., 379 321, 373, 392, 393, 416, 444, 461, 516
Whitaker, K. J., 472 Williamson, D. E., 76
Whitcher, B., 75 Williamson, J., 447
White, B. R., 43 Willinek, W. A., 12
White, D. J., 212 Willis, R. J., 463
White, L. E., 75f, 79–80, 88, 118, 210, 260t, Willis, S. L., 134, 250, 260t, 264, 266, 366,
303, 319 367t, 369, 370, 415, 462, 465
White, L. R., 393, 393, 404 Wilms, M., 306t, 309, 311t, 313, 314t
White, S. M., 119, 250, 444, 445f, 449, 450, Wilser, L., 516
451, 452, 471 Wilson, A. A., 49
Whiteman, M. C., 186, 371 Wilson, F. A., 281
Whitfield-Gabrieli, S., 114, 214, 260t Wilson, H. R., 219
Whitford, T. J., 171 Wilson, I. A., 280, 283, 286
Whiting, W. L., 88, 210 Wilson, M. A., 282
Whitman, G. T., 157 Wilson, R. S., 161, 273, 391, 417, 418, 462,
Whitmer, R., 453 463, 465, 470, 471, 485, 491
Whittemore, A. R., 23 Wilson, T., 25
Whitwell, J. L., 492 Wilson, R. S., 491
Wicherts, J. M., 263 Winblad, B., 40, 49, 380, 391, 392, 429, 446,
Widaman, K., 112 447, 453, 470
Wiegand, M., 349 Windhorst, A. D., 43, 45
Wiegand, S. J., 422 Winecoff, A., 346
Wiegell, M. R., 22 Wink, A. M., 43
Wieler, M., 159 Winkler, A. M., 159
Wielopolski, P. A., 260t, 448 Winkler, J., 471
Wienhard, K., 135 Winneke, A. H., 218
Wierenga, C. E., 25, 306t, 308, 314t, 324 Winocur, G., 165, 166f, 339
Wiest, R., 287 Winter, B., 443
Wig, G. S., 108, 110, 114, 117f, 118, 373, Winz, O. H., 52
375, 375f Wirth, M., 42, 138, 145, 470, 489, 492
Wigman, S. E., 42, 43, 108, 110, 117f, 494 Wisdom, N. M., 418
Wijndaele, K., 453 Wise, R. G., 26
Wilcock, G. K., 421 Wisnieff, C., 29
Wilcox, C. E., 50, 53 Wisniewski, S., 507
Wildenauer, D., 425 Wisniewski, T., 37, 492, 493
591
Author Index 591
Wiste, H. J., 37, 39, 41, 42, 43, 47, 145, 372, Wszalek, T., 217
485, 487, 489, 492, 494 Wu, B., 92
Witte, A. V., 424 Wu, C. Y., 223, 286
Witteman, J. C., 419, 514 Wu, E. Y., 243
Witter, M. P., 108, 276, 284, 286 Wu, H. Z., 108, 117f
Witzel, T., 17, 76 Wu, J. T., 29, 108, 117f
Woermann, F. G., 241 Wu, K., 157, 263
Wohl, M., 514 Wu, M., 116
Wojcicki, T. R., 119, 250, 444, 445f, 449, Wu, O., 279
450, 451, 452, 471 Wu, W. E., 289
Wojtowicz, M., 107, 108, 117f, 226, 239, 326 Wu, Z., 27
Wolf, A. P., 50, 51f Wuketich, S., 49
Wolf, D., 82, 217 Wurm, L. H., 339
Wolf, N. D., 496 Wycliffe, N., 27
Wolf, P., 453
Wolf, P. A., 377, 380, 391, 392, 393, Xia, A., 260t, 264
395, 453 Xia, C. F., 493
Wolf, R. C., 496 Xia, J., 520
Wolf, R. L., 24 Xia, M. R., 108, 117f
Wolfe, B. B., 279 Xiao, G., 26
Wolfson, L., 260t, 261t Xie, X., 487
Wollmer, M. A., 417, 426 Xie, Y., 514
Wolpert, D. M., 461 Xiong, C., 37, 487, 489, 492, 494, 495
Woltjer, R. L., 492, 493 Xu, B., 53
Womelsdorf, T., 105 Xu, D., 44
Wong, A. T., 311t, 312 Xu, F., 25, 27, 241, 250, 325
Wong, C., 50 Xu, G., 260t, 264, 376, 396, 420, 421
Wong, D. F., 49, 157, 171, 490 Xu, J., 22
Wong, E. C., 24, 25 Xu, Y., 27, 278
Wong, T. Z., 490 Xuan, L., 82, 83, 260t, 266, 446
Wong, V. S., 264
Wood, R. A., 516 Yacoub, E., 17, 22
Wood, S. J., 116, 345 Yaffe, K., 403, 440, 449, 453, 496
Woodard, J. L., 450, 452 Yajima, K., 39
Woodman, G. F., 243 Yakovlev, P. I., 79
Woodruff, B. K., 376 Yakushev, I., 39
Woodruff, C. C., 223 Yali, A. M., 379
Woods, J. A., 444, 445f, 452, 471 Yamada, H., 159
Woodward, M., 492 Yamada, K. A., 138, 495
Woodward, N. D., 52 Yamada, M., 39
Woodward, T. S., 112, 319 Yamaguchi, S., 108, 110, 117f, 135, 209
Woolf, C. J., 488 Yamaguchi, T., 393
Woolley, M., 518 Yamamoto, H., 514
Woolrich, M. W., 143 Yamamoto, M., 49
Woolson, S., 146 Yamamoto, T., 48, 418
Worley, P. F., 277 Yamasaki, T., 54
Worsley, K., 79 Yamashita, F., 422
Worthman, C., 515 Yamasue, H., 159
Wright, C. B., 395 Yamaura, H., 393
Wright, C. I., 42, 340, 492 Yamazaki, M., 422
592
592 Author Index
Yan, C. G., 108, 117f Young, K. M., 76

Yan, P., 138 Young, R. C., 510, 512, 513, 517
Yan, Z., 118, 138, 139 Young, V. G., 89
Yanai, K., 48 Younger, A. P., 43, 44, 45, 46f, 372
Yanase, D., 39 Yousem, D. M., 420
Yang, F. G., 263 Ystad, M., 260t, 264
Yang, H. S., 108, 117f Yu, C. E., 108, 110, 490, 493
Yang, J. L., 453 Yu, H., 216
Yang, L. X., 223, 462 Yu, L., 491
Yang, T., 491 Yuan, P., 237
Yang, Y., 24, 238, 472 Yucel, M., 106
Yano, M., 514 Yuen, G., 511
Yap, M. J., 263
Yap, Q. J., 159 Zach, E., 260t, 264
Yaqub, M., 43 Zacharias, F. J., 398
Yarasheski, K. E., 487 Zacks, J. M., 208, 215, 217, 219
Yasar, S., 392 Zacks, R. T., 116, 208, 215, 217, 219, 244,
Yassa, M. A., 183, 277, 286, 286f, 287, 288, 248, 284
420, 496 Zaharchuk, G., 25
Yasuda, R. P., 279 Zahodne, L. B., 471
Yasuno, F., 49 Zahr, N. M., 79, 264
Yates, P., 492 Zahra, A., 396
Ye, F. Q., 24 Zaitsev, E., 422
Ye, W., 490 Zald, D. H., 52, 341, 345
Yegiyan, N. S., 343 Zaleta, A. K., 78, 496
Yen, I. V., 43, 118, 494 Zalla, T., 340, 344
Yen, K., 92 Zanstra, Y., 389, 404, 406
Yeo, B. T., 39 Zanto, T. P., 4, 173, 176, 207, 214, 216, 219,
Yeom, K. W., 92 220, 222, 222f, 223, 224, 225
Yesavage, J. A., 422, 423, Zanzonico, R., 76
Yetkin, F. Z., 107 Zappia, M., 509
Yezhuvath, U. S., 26 Zarahn, E., 25, 250, 471
Yin, S., 473 Zatorre, R. J., 141, 471
Ying, Z., 453 Zavagnin, M., 466
Yokell, D., 493 Zecca, L., 27, 91, 92
Yomogida, Y., 311t, 312, 318 Zeger, S. L., 420
Yonelinas, A. P., 110, 112, 138, 195, 196, Zelinski, E. M., 308, 367t, 368, 369,
276, 287, 302, 303, 306t, 308, 313, 314t, 442, 462
322, 324, 343 Zeng, G., 41, 42
Yoon, S. S., 389 Zetterberg, H., 485
Yoshida, D., 444 Zhang, C., 428, 488
Yoshida, R., 278, 280 Zhang, D., 491
Yoshikawa, K., 49 Zhang, H. Y., 108, 117f, 157, 280
Yoshikawa, T., 48 Zhang, J., 72, 73, 490
Yoshino, J., 76, 80 Zhang, M. R., 493
Yoshita, M., 89 Zhang, P., 396
Yotter, R. A., 158 Zhang, R., 27, 447, 449
You, Y., 422 Zhang, S., 44
Youdim, M. B., 27, 91, 92 Zhang, W., 493
Young, E., 514 Zhang, X., 108, 110
593
Author Index 593
Zhang, Y., 24, 44 Zimmet, P. Z., 453

Zhao, C., 280, 444 Zimprich, D., 366, 367t, 368
Zhao, L-Y., 44 Zink, T. A., 137
Zhao, P., 394, 396 Ziolko, S. K., 490
Zhao, T., 493 Zipursky, R. B., 15
Zhao, Z., 513 Zitnik, R., 518
Zheng, Z., 473 Zitzmann, M., 443
Zhou, L., 42, 492 Zlokovic, B. V., 23
Zhou, Y., 157, 171, 490 Zoccatelli, G., 137, 288
Zhou, Z. W., 119 Zola, S. M., 283
Zhu, H., 146 Zonderman, A. B., 118, 156, 157, 161, 391,
Zhu, S., 518 394, 397
Zhu, W. Z., 44, 157, 260t, 264, 451 Zoppi, A., 517
Zhu, X., 473, 487 Zorkoltseva, I. V., 419
Zhuang, J., 283 Zschutschke, L., 217
Zhuang, L., 157 Zuk, S. M., 43, 46
Zhuravleva, T. Y., 215 Zuo, X. N., 108, 117f
Ziegler, G., 158 Zürcher, N. R., 39
Zieren, N., 260t Zwan, M. D., 45
Zijdenbos, A. P., 23, 79, 393, 461 Zwiers, M. P., 448
Zilles, K., 52 Zysset, S., 218
Zimmerman, B., 240
Zimmerman, M. E., 43, 44, 79, 249, 261t,
264, 507
594
â•‡ 595
SubjectÂ€Index
Activities of daily living (ADLs), 248–â•‰49 Alzheimer’s Prevention Initiative (API),

Alzheimer’s disease. see also preclinical 495–â•‰96
Alzheimer’s disease Amygdala
amyloid burden, cognitively healthy aging, in depression, 511
40–â•‰48, 46–â•‰47f emotion, emotional memory, 5, 340–â•‰41,
amyloid-â•‰cognition relationships, 44–â•‰48, 342–â•‰43f, 344–â•‰48, 351
46–â•‰47f, 139, 145 episodic memory, 318, 318f
animal models of, 274 hypertension, 400
beta-â•‰amyloid accumulation, 372 item SMEs in, 308
bilingualism in protection from, 118 MTL anatomy, 275f
change detection in, 243 salience network (SLN), 106–â•‰7, 110,
continuum of, 486f 116–â•‰17, 117f
cortical thickness in, 42 in social cognition, 113–â•‰14
functional connectivity, amyloid burden Amyloid burden
in, 42–â•‰43 in Alzheimer’s disease, 372
functional connectivity in, 118 APOE4 in, 37, 404, 405f, 487
genes, heritability factors, 417, 429 cognitively healthy aging, 40–â•‰48, 46–â•‰47f
glucose hypometabolism in, 37 functional connectivity in, 42–â•‰43
hippocampal activity in, 41–â•‰42, 494–â•‰95 hippocampus, PET studies, 41–â•‰42
hypertension as risk factor, 403–â•‰5, 405f hypertension, 403–â•‰5, 405f
LME testing, 172 middle-â•‰aged brain, 371–â•‰73
neuropathology of, 486f MTL region, PET studies, 41–â•‰44
PET imaging, 488f PET studies, cognitively healthy
structural aging in, 237 aging, 40–â•‰41
tau imaging, 48–â•‰49, 488f, 492, 493 in preclinical Alzheimer’s disease, 41,
tau neurofibrillary tangles, 372, 492 485–â•‰87, 486f
temporal lobe atrophy in, 145 white matter abnormalities, Alzheimer’s
white matter abnormalities, amyloid bur- disease, 43–â•‰44
den in, 43–â•‰44 in white-â•‰matter integrity, 43–â•‰44, 134–â•‰35,
white matter volume, exercise and, 448 138–â•‰39, 145
595
596
596 Subject Index
Amyloid cascade hypothesis, 487–88, 488f brain activity changes, 287–88

Amyloid-cognition relationships cerebrovascular reactivity (CVR),
Alzheimer’s disease, 44–48, 46–47f, 188–89, 189f
139, 145 change detection, 212
episodic memory, 44–48, 46–47f cohort effects, 250
PET studies, 44–48, 46–47f discrimination, 217–18
processing speed, 44–48, 46–47f limitations of, 417
working memory, executive function, magnetic field strength, 12–13
44–48, 46–47f, 490 principles, applications of, 17–19, 18f
Antiplatelet agents, 517 structural-functional relations, age-related
APOE 4 differences in, 87–88
in amyloid deposition, 37, 405, 405f, 487 TaqIA polymorphism, 428
cerebral blood flow, 421 vascular aging effects, 25–27
characterization, 418–22, 419f white-matter integrity, structural-
episodic memory and, 419–21, 428, 429 functional relations, 85–87, 86f
exercise effects on, 452 Brain volumetrics. see also cortical thick-
exercise modulation of, 429 ness; white-matter integrity
fMRI studies, 420–22 cognitive training, 471–72
in functional connectivity, 118, 138–39 emotion, emotional memory, 340
middle-aged brain, 376 exercise effects, 439–40, 443–46, 445f,
MTL activity, 420–21 471–72
processing speed and, 429 hypertension effects, 393–95
temporal lobe volumes and, 146 longitudinal imaging, 157–59, 157f,
white-matter integrity, 421 174–75
in working memory, executive function, middle-aged brain, 140–42, 141–42f, 371
249, 428, 490 MRI studies, 15–17, 15f, 135–37,
APP, 487 136f, 145
Aripiprazole, 515 processing speed, 261, 261t
Arterial-spin-labeling (ASL) MRI, 12–13, selective attention, inhibitory control, 219,
23–25, 24f 222f, 224–25, 225f
Associative deficit hypothesis, 303 Val66Met polymorphism, 422–23, 423f
Associative memory (recollection), 186, Bupropion, 515
191–94, 192–94f, 197, 302–3, 306t,
309–10, 310f, 319–20 Celecoxib, 518
A4 Study, 496 Cerebral glucose metabolism, 36–39, 38f
A673T mutation, 487 Cholinesterase inhibitors, 516
Attention. see selective attention, inhibitory Chronic obstructive pulmonary disease
control (COPD), 519
Axial diffusivity (AD), 73, 74f, 76, Clusterin (CLU), 428
80–87, 86f COGITO study, 472
Cognitive ability, 186–88, 187f
BDNF Cognitive control network, 510–11
in depression, 510 Cognitive decline, longitudinal imaging,
exercise effects on, 452, 453 161–66, 162–64f, 166f
polymorphism, 422–23, 423f, 428, 429 Cognitive engagement hypothesis, 463, 464
working memory, executive function, 428, Cognitive function generally
452, 490 allostatic load in, 249
Betula Aging Study, 373 measures, level of analysis, 262–64, 262f
BOLD fMRI. see also fMRI studies Cognitive reserve hypothesis, 209
APOE4, 420–22 Cognitive training
597
Subject Index 597
brain maintenance, 474–75 D3 agonists, 515

brain mechanisms affecting, 473–76 Dallas Lifespan Brain Study, 373
brain volumetrics, 471–72 DAT interactions, 427–28
compensation, 474 Dedifferentiation
cortical thickness, 471–72 episodic memory, 304, 319–20
educational attainment, 471 functional connectivity, 114, 145–46,
engagement, between-person variation, 219, 224
462–65 in middle-aged brain, 373–74
engagement, brain structure effects, selective attention, inhibitory control,
470–73 219, 224
episodic memory, 463, 472 Default mode network (DMN)
functional connectivity, 472–73 APOE4, 420–21
leisure activities effects, 462, 463 associative memory (recollection),
level-change associations, 462–64 192–94, 193–94f, 197, 319
media consumption, 463 cognitive training, 473
occupational demands/retirement, 463–64 DED syndrome, 510–11
performance, effects on, 465–70, 467–69f episodic memory, 47f, 186–88, 187f, 307,
sedentary lifestyle, 470–71 315–17, 316f, 319
white-matter integrity, 472, 474 exercise effects on, 451
working memory, executive function, functional connectivity, 106–18, 109f,
466–70, 467–69f 111f, 117f, 134–39, 136f
Compensation imagery, 226
cognitive training, 474 longitudinal imaging, 165–66, 166f
emotion, emotional memory, 351 preclinical Alzheimer’s disease, 493–95
episodic memory, 303–5, 308, 312f task-negative effects, 197, 198f
hippocampal, 288 working memory, executive function, 247
hypothesis, 288 Dementia. see Alzheimer’s disease; preclinical
white-matter integrity, 87–88, 145–46, Alzheimer’s disease
164–65, 209, 219 Depression
working memory, executive function, amygdala in, 511
238–42, 240–42f, 251–52, 251f apathy, 511
Compensation-related utilization of neural BDNF in, 510
circuits hypothesis (CRUNCH), 209, cerebral perfusion, 514
239, 241, 246, 247 chronic stress targeting, 518–19
Complementary Learning Systems (CLS) COPD targeting, 519
theory, 277 depression-executive dysfunction syn-
Compound symmetry, 171 drome, 509–11, 515–17, 520t
COMT inhibitors, 515 disconnection mechanism, 513–14
COMT polymorphism, 424–30 5-HTTLPR in, 510
Context RSEs, 310, 311t, 312–13, 312f, inflammation hypothesis, 514–15,
322–23, 323f 518, 520t
Cortical thickness. see also brain late-onset hypothesis, 511–12
volumetrics middle-aged brain and, 378
in Alzheimer’s disease, 42 model of, 508–9, 508f
cognitive training, 471–72 post-stroke, 519–21
episodic memory, 197–99, 199f therapies, 515–21, 520t
longitudinal imaging, 158 vascular depression hypothesis, 513–14,
multimodal imaging, 141–42, 143f, 144 517–18, 520t
COX-2 inhibitors, 518 Depression-executive dysfunction syndrome,
Cytoprotective agents, 517–18 509–11, 515–17, 520t
598
598 Subject Index
Depressive pseudodementia, 512 prefrontal cortex, 341, 344–47, 349,

Diffusion tensor imaging (DTI) 351, 352
emotion, emotional memory, 350f processing, encoding, 337–39, 338f, 354
fractional anisotropy ( see fractional regulation, 352
anisotropy (FA)) response bias, 348–49
principles, applications of, 14t, 19–22, retrieval, 348–54, 350f
20f, 72 selective enhancements, 340
white matter integrity ( see white-matter self-referential processing, 346
integrity) valence effects, 341–43, 343f, 351
DMN. see default mode network (DMN) valence-only effects, 343f, 347–48
Dominantly Inherited Alzheimer Network valence-specific effects, 343–47, 343f,
(DIAN), 495–96 352–53
Dopamine visual detail, 344
cognitive functions associations, 52–55 Entorhinal cortex (EC), 278, 279, 284–85,
COMT/DBH interactions, 428 285f, 287
COMT polymorphism, 424–25 Episodic memory
PET studies, function decrease, 49–52, 51f age vs. aging effects, 183–84
prefrontal, functions of, 54–55 aging effects, 184–86, 190–94, 192–94f
production, hippocampus and, 279 amygdala, 318, 318f
TaqIA polymorphism, 428 amyloid-cognition relationships, 44–48,
in working memory, executive 46–47f
function, 249 APOE4 and, 419–21, 428, 429
Dorsal attention network (DAN), 107, associative deficit hypothesis, 303
110–12, 116–17, 117f associative memory (recollection), 186,
DRD2 interactions, 427–28 191–94, 192–94f, 197, 302–3, 306t,
309–10, 310f, 319–20
Early-to-late shift with aging (ELSA), associative recognition, 309–10, 311t
323, 323f associative RSEs, 310–13, 311t, 314t, 315,
Ecosystem focused therapy (EFT), 519–21 317, 325
Education on Stroke and Depression, 520–21 associative SMEs, 308–9, 314t
Emotion, emotional memory brain-behavior relationships, 194–95
ACC, 348 brain-memory relationships, 196–99,
age-related effects, 352–54 198–99f
amygdala, 5, 340–41, 342–43f, cerebrovascular reactivity, 325–26
344–48, 351 cognitive ability, 186–88, 187f
arousal effects, 339–41, 342–43f, 354 cognitive training, 463, 472
brain regions involved in, 342f cohort effects, 185–86
brain volumetrics, 340 compensation, 303–5, 308, 312f
cognitive control, 345, 352 context RSEs, 310, 311t, 312–13, 312f,
compensation, 351 322–23, 323f
DTI studies, 350f cortical disconnection, 174
ERP studies, 345, 349, 350f, 353 cortical thickness, 197–99, 199f
event detail, 343–45 cross-sectional design, 184, 189
flashbulb memory, 339 cross-sectional vs. longitudinal
fMRI studies, 348, 349, 350f methods, 326
functional connectivity, 340–41, 345 dedifferentiation, 304, 319–20
hippocampus, 340–41, 342f, 347 default mode network, 47f, 186–88, 187f,
motivational hypothesis, 345–46, 352 307, 315–17, 316f, 319
neural recruitment, 352 dopamine/cognitive functions
positivity effect, 345–48 associations, 52–53
599
Subject Index 599
fMRI studies, 305, 324–26 two-process MTL model, 303

frontoparietal control network (FPC), 47f Val66Met polymorphism, 422–23, 423f
functional connectivity, 110, 112, 134, variation partitioning, 186–88, 187f
138, 317–22, 318f, 324 Executive function. see working memory,
hemodynamic response function, executive function
188–89, 189f Exercise
hippocampus, 161–62, 162f, 302, 307–9, ACC effects, 443, 444, 450
312–13, 312f, 317–21, 318f, 323–24, APOE effects, 452
323f, 326 BDNF effects, 452, 453
hypertension and, 390–91, 397 brain structure effects, 439–40, 443–46,
inferences from cross-sectional data, 188 445f, 471–72
item memory (familiarity), 190–91, cognitive outcomes, 440–43
194–95, 302, 306t, 319–20 cross-sectional designs, 440, 443–44
item RSEs, 310–13, 311t, 312f, 314t, dose-response nature, 454
315–17, 316f effect mechanisms, 452–54
item SMEs, 307–8, 314t, 315–17, effect modifiers, 451–52
316f, 325 fMRI studies, 449–50
KIBRA T-allele, 417, 425–26, 427f functional connectivity, 450–51
longitudinal imaging, 155–56, 156f, 174 gender effect, 442, 452
middle-aged brain, 367t, 368–69 genetic effects modulation by, 429
monitoring effects, 195 gray matter volume, 443–46, 445f
negative RSEs, 313, 314t, 324–25 middle-aged brain, 379
negative SMEs, 308, 314t, 324–25 omega-3 effect, 452
neural inefficiency, 304 pro-inflammatory factors, 453
neural mechanisms, 301–5 prospective epidemiological
NMDA receptor interactions, 427–28 studies, 440
object studies, 309, 312, 325 randomized controlled trials, 440–43
over-recruitment, 185 risk factor reduction, 453
perirhinal cortex, 308, 312f, 320, 326 selective improvement hypothesis, 442
post-retrieval monitoring, 195 white-matter integrity, 446–49
in preclinical Alzheimer’s disease, in working memory, executive function,
490–91, 494 250, 442
prefrontal cortex, 302–4, 307–8, 312–13, Expanded inhibitory deficit hypothesis,
312f, 315–21, 316f, 318f, 323–24, 323f 247–48
repetition suppression, 317
resource deficit hypothesis, 303, 322 Fast field echo (FFE), 16
retrieval, recollection, 190–91, 194–99, Fast low angle shot (FLASH), 16
198–99f Feature binding, 242–44
retrieval effort, 323, 323f Florbetaben, 40
retrieval memory effects (RMEs), 304–5, Florbetapir, 40
309–15, 310f, 311t, 312f, 314t, 321–25 Flutemetamol, 40
statistical analysis, 191 Flynn Effect, 185
sub-groups matching, 190–91 FMRI studies. see also BOLD fMRI
subsequent memory effects (SMEs), BOLD signal, 17–19, 18f
304–9, 306t, 320–25 brain volumetrics, 15–17, 15f,
task-based effects, 317–20, 318f 135–37, 136f
task-independent effects, 303–5 change detection, 212–13
task-negative effects, 197, 198f compensation, 87–88
three-process MTL model, 302 COMT polymorphism, 424–25
true age effects, 190–94, 192–94f data quality assessment, 12
600
600 Subject Index
FMRI studies (Cont.) default mode network (see default mode

dopamine/cognitive functions network (DMN))
associations, 53–54 disconnection mechanism, 513–14
emotion, emotional memory, 348, dopamine/cognitive functions
349, 350f associations, 53–54
episodic memory, 305, 324–26 dorsal attention network (DAN), 107,
exercise, 449–50 110–12, 116–17, 117f
glucose hypometabolism, 37–39, 38t dorsal PCC subsystem, 108–9, 109f
hypertension, 401–3, 401f emotion, emotional memory,
KIBRA T-allele, 425–26, 427f 340–4 1, 345
principles, applications of, 14t, 17–19, 18f episodic memory, 110, 112, 134, 138,
resting-state, 17 317–22, 318f, 324
task-related, 17, 26 exercise, 450–51
top-down suppression selective deficit, factors affecting, 117–19
219, 220f fractional anisotropy (FA), 73–88, 79f,
vascular aging effects, 25–27 86f, 279
working memory, executive function, frontoparietal control network (FPC),
240–42, 240–42f 106–7, 110–19, 117f
Fractional anisotropy (FA) hippocampus, 278–79, 286–87, 451
age-related effects neurobiological hippocampus, longitudinal imaging, 161–
specificity, 80–81 62, 162f, 175
anterior-posterior gradient, 78–79, 79f interventions role in, 119
cognitive training effects, 472 iron accumulation in, 91–93
DED syndrome, 509–11 longitudinal imaging, 161–66, 162–64f,
DTI/behavior measure 166f, 175
relationships, 82–85 middle-aged brain, 110, 375
DTI/functional measure relationships, MTL subsystem, 108–9, 109f, 112
85–88, 86f network segregation, 110, 118
DTI tractography, 73–77, 74f parahippocampal place area
exercise effects, 448–49 (PPA), 113
functional connectivity, 73–88, 79f, PET studies, 42–43
86f, 279 prefrontal cortex (PFC), longitudinal
genes, heritability factors, 417 imaging, 162–65, 163–64f, 175
global vs. tract-specific effects, 81–82 processing speed, 261
multi-modal imaging, 137–38, resting, 107–10, 109f, 111f, 117f, 118,
140–41, 142f 134, 138–40
myelodegeneration hypothesis, 79–80 salience network (SLN), 106–7, 110,
Framingham Cardiovascular Risk Profile, 116–17, 117f
398, 405 selective attention, inhibitory control,
Frontoparietal control network (FPC), 47f, 220–24, 222f, 319, 321–22
106–7, 110–19, 117f self-ratings, 110, 111f
Functional connectivity shifting in, 119
amyloid burden in, 42–43 social cognition, 113–14, 115f
anti-correlation in, 107 study approaches, 105–6
autobiographical memory, 110 task-based, 110–16, 115f, 117f
behavior impacts, 110 task-based, longitudinal imaging, 161–65,
cognitive control, 111–12, 114, 116–18 162–64f
cognitive training, 472–73 ventral PCC subsystem, 108–9, 109f
dedifferentiation, 114, 145–46, 219, 224 white-matter integrity, 85–87, 86f, 138–40
DED syndrome, 510–11 (see also white-matter integrity)
601
Subject Index 601
whole brain studies, 107–8 entorhinal cortex (EC), 278, 279, 284–85,

working memory, 113 285f, 287
Functional Connectome Project, 139 episodic memory, 47f, 161–62, 162f, 302,
307–9, 312–13, 312f, 317–21, 318f,
General Electric MRI scanners, 16 323–24, 323f, 326
Genetics exercise effects, 444–46, 445f
age magnification, factors affecting, functional connectivity, 161–62, 162f, 175,
428–30 278–79, 286–87, 451
APOE4 ( see APOE4) hypertension, 393, 394, 400, 404
APP, 487 KIBRA T-allele, 425–26, 427f
BDNF polymorphism, 422–23, 423f, longitudinal imaging, 161–62, 162f, 175
428, 429 LTP induction, 279–82
candidate genes, 415–16, 418 memory changes, 282–83, 285–86, 286f
CLU polymorphism, 428 memory decline mediation, 273
COMT polymorphism, 424–30 middle-aged brain, 372, 373
DAT interactions, 427–28 mnemonic similarity task (MST),
DRD2 interactions, 427–28 285–88, 286f
KIBRA T-allele, 417, 425–26, 427f, 429 mossy fibers, 276, 277
OXTR polymorphism, 428 MTL structures, 274–76, 275f
PSEN-1/2, 487 multiple-synapse boutons (MSBs), 280
resource-modulation hypothesis, 416–18, neurobiological changes in, 278–82
416f, 425, 426, 429 neurogenesis, synaptogenesis, 280
TaqIA polymorphism, 428 neuromodulatory changes, 279
Glucose hypometabolism, 37–39, neurophysiological changes, 279–80
38t, 138 NMDA receptors, 282
pattern completion, separation, 276–77
Hemispheric asymmetry reduction in aging perirhinal cortex changes, 280–81
(HAROLD) model, 208, 212, 304, PET studies, amyloid burden in, 41–42
308, 324 prefrontal cortex interactions, 278, 281
Hindered diffusion, 73 trisynaptic loop, 276
Hippocampus Val66Met polymorphism, 422–23, 423f
in Alzheimer’s disease, 41–42, 494–95 white-matter integrity, 87
animal models, 273–74 working memory, executive function, 237
APOE4, 420–21 5-HTTLPR, 510
associative memory (recollection), Human Connectome Project, 13, 17, 22
191–92, 192f Huntington’s disease, 496
BDNF effects, 444–46, 445f Hypertension
brain activity changes, 287–88 age-dependent associations, 391–93
CA1/2/3, 275f, 276–80, 284–88, amygdala, 400
285–86f, 393 amyloid deposition, 403–5, 405f
communication pathways, 276 brain volumetrics effects, 393–95
compensation hypothesis, 288 cerebral perfusion, 394, 396, 397–400,
computational role of, 276–77 399f, 406
cross-species model, 283–88, 285–86f cerebrovascular reactivity, 403
dentate gyrus (DG), 275f, 276–80, 284–88, cognition, effects on, 390–91,
285–86f 396–97, 406–7
detonator synapses, 276 dose-response effect, 398–400
dopamine production and, 279 epidemiology, pathology, 389–90
emotion, emotional memory, 340–41, episodic memory and, 390–91, 397
342f, 347 fMRI studies, 401–3, 401f
602
602 Subject Index
Hypertension (Cont.) PET applications, 156–57

functional brain activation, 400–403, statistical analysis, 171–73
401–2f study design, 166–67
grey matter effects, 393–94, 396–97 test-retest effects, 167, 170, 173–75
hippocampus, 393, 394, 400, 404 white-matter integrity, 159–61, 159f,
middle-aged brain, 376–77 160t, 174
pathological aging role, 403–5, 405f Lurasidone, 515
PET studies, 398, 400, 403, 405
prefrontal cortex, 393, 394, 400 Magnetoencephalography (MEG), 223
processing speed and, 390–92, 397 MAO inhibitors, 515
resting cerebral blood flow, 397–400, 399f Mean diffusivity (MD), 73, 74f, 76, 80–87,
SPECT studies, 398 86f, 140–41, 142f, 417
white-matter integrity, 394–96 Middle-aged brain
working memory, executive function, amyloid deposition, 371–72
391, 392 APOE4, 376
APOE interactions, 428
Inflammation hypothesis, 514–15, 518, 520t BDNF interactions, 428
Inhibitory deficit hypothesis, 244–48 biological factors, 376–78
Intellectual engagement. see cognitive cholesterol and, 377
training cognition, 366, 367t, 381
IQ, iron accumulation in, 92–93 cognition, predictive value, 370
COMT interactions, 428
KIBRA T-allele, 417, 425–26, 427f, 429 contextual pressure, 380
crystallized knowledge, 367t, 370
Last value carried forward (LVCF) method, dedifferentiation in, 374
172–73 defined, 364–65
Late-onset hypothesis, 511–12 depression and, 378
LEARN study, 496 enrichment, depletion factors, 375–76
Linear mixed effect model (LME), 172–73 episodic memory, 367t, 368–69
Linked Independent Component Analysis everyday functioning, 367t, 370
(LICA), 143 functional changes, nonuniform, 381
Longitudinal imaging functional connectivity, 110, 375
benefits of, 155–56, 174 glucose tolerance, 377
brain network, 165, 166f gray matter volume, 371
brain volumetrics, 157–59, 157f, 174–75 hippocampus, 372, 373
cognitive decline, 161–66, 162–64f, 166f hypertension, 376–77
conjoint changes, 174 implicit memory, 367t, 369
cortical thickness, 158 lifestyle factors, 378–80
dropout effects, 167–68 longitudinal designs, 381
episodic memory, 155–56, 156f, 174 neural activity patterns, 372–75,
functional connectivity, 161–66, 162–64f, 373–74, 374f
166f, 175 neural plasticity in, 381
gray-matter changes, 157–59, 157f, peak performance, 380–81
174–75 prefrontal cortex, 372, 373
missing at random (MAR), 168–69, 175 processing speed, 366, 367t
missing completely at random (MCAR), psychosocial context of, 365
168, 172, 175 reasoning, 367t, 369
missingness, 167–73, 175 semantic encoding, 375, 375f
not missing at random (NMAR), 169–70, social, emotional processing,
172–73 367t, 369
603
Subject Index 603
spatial processing, 367t, 369 explained variance, 134–35

STAC-r model, 363–64, 364f, 373 life-span trajectories discovery, 140–42,
structural differences, 371–73 141–42f
tau neurofibrillary tangles, 372 modalities, 132–33, 133t
visual memory contrast, 374–75, 374f predictive value, 146
weight and, 377 temporal sequences, 145
white-matter integrity, 371–72 variable control, interpretation,
white matter volume trajectories, 140–42, 135–37, 136f
141–42f Myelodegeneration hypothesis, 79–80
working memory, executive function,
367t, 368 Nearest Neighbor imputation, 173
Middle temporal lobe (MTL) Neuroplasticity-based computerized cogni-
amyloid burden, PET studies, 41–44 tive remediation (nCCR), 516
anatomy, 274–76, 275f NMDA receptor interactions, episodic mem-
APOE4 activity, 420–21 ory, 427–28
episodic memory models, 302, 303 Not missing at random (NMAR), 169–70,
functional connectivity subsystem, 108–9, 172–73
109f, 112
Minimum intensity projection (mIP), 27–28, 28f Opiate receptor agonists, antagonists, 516
Minocycline, 518 Oxytocin receptor (OXTR), 428
Missing at random (MAR), 168–69, 175
Missing completely at random (MCAR), Parkinson’s disease, 50, 496
168, 172, 175 PASA model
Missingness, of data, 167–73, 175 emotion processing, 341
Motor performance, iron accumulation episodic memory, 303–4, 307, 308, 317,
in, 92–93 318f, 320–21, 324
MRI studies selective attention, inhibitory control, 208,
brain volumetrics, 15–17, 15f, 135–37, 210, 212, 216–17
136f, 145 PET studies
image quality, 11–12, 12f amyloid burden, cognitively healthy
magnetic field strength, 12–13 aging, 40–41
multislice acquisition, 16 amyloid-cognition relationships, 44–48,
overview, 3, 9 46–47f
principles, 10–12, 10f, 12f cerebral glucose metabolism, 36–39, 38f
7T, 13, 27–29, 28f disease progression, interindividual
signal origins, 10–11, 10f variability, 36
signal-to-noise ratio (SNR), 12 disease-related processes, 36
study design, 13 dopamine/cognitive functions
techniques, common, 14t associations, 52–55
3D acquisition, 16 dopamine function decrease, 49–52, 51f
MTL. see middle temporal lobe (MTL) dopamine imaging, 49
Multimodal imaging functional connectivity, amyloid burden
benefits of, 133–34, 144–46 in, 42–43
biological mechanism illumination, 137–40 hippocampus, amyloid burden in, 41–42
brain, change trajectories in, 131–32 hypertension, 398, 400, 403, 405
cognitive function, 134–37, 136f MTL region, amyloid burden in, 41–44
cortical thickness, 141–42, 143f, 144 overview, 3, 35–36
data analysis methods, 142–44 tau imaging, 48–49
dopamine/cognitive functions white matter abnormalities, amyloid bur-
associations, 53–54 den in, 43–44
604
604 Subject Index
Philips MRI scanners, 16 Problem solving therapy (PST-ED), 516–17

Physical activity. see exercise Processing speed
PID-C, 519 amyloid-cognition relationships, 44–48,
Pittsburgh Compound B (PIB), 40, 45, 138 46–47f
Positivity effect, 345–48 APOE4 and, 429
Posterior-to-anterior shift in aging (PASA) brain volumetrics, 261, 261t
model. see PASA model cognitive training effects, 466–70, 467–69f
Preclinical Alzheimer’s disease. see also COMT polymorphism, 429–30
Alzheimer’s disease discrimination, 217–18
amyloid burden, cognitive correlates to, functional connectivity, 261
490–92 hypertension and, 390–92, 397
amyloid burden in, 41, 485–87, 486f influences on age relations, 265–67, 265f
amyloid cascade hypothesis, 487–88, 488f level of analysis, 262–64, 262f
amyloid-neurodegeneration relationships, mediation, 266–67
492–93 middle-aged brain, 366, 367t
default mode network, 493–95 moderation, 266, 267
defined, 488–89, 495 neural correlates studies, 259–61, 260–61t
epidemiology, 485–87, 486f participant sampling, 267
episodic memory in, 490–91, 494 white-matter integrity, 260t
genes, heritability factors, 417, 427f, 429 Proton density MRI, 14t, 15, 22–23
prevention trial design, 495–96 PSEN-1/2, 487
reserve, impact of, 491–92 PSEN1 cohort trial, 495–96
staging, 489–90, 489f
tau imaging, 48–49, 488f, 492, 493 Quantitative susceptibility mapping (QSM),
Prefrontal cortex 28–29, 92
APOE4, 420–21
change detection (feature), 216–17 Recollection (associative memory), 186,
change detection (object), 219–22f, 191–94, 192–94f, 197, 302–3, 306t,
219–23 309–10, 310f, 319–20
change detection (spatial), 211–13 Repeated measures (RM-) ANOVA, 171
cognitive aging models, 208–9 Resource deficit hypothesis, 303, 322
cognitive training, 473 Resource-modulation hypothesis, 416–18,
COMT polymorphism, 424–25 416f, 425, 426, 429
discrimination, 217–18 Restricted diffusion, 73
emotion, emotional memory, 341, 344–47, Rofecoxib, 518
349, 351, 352
episodic memory, 302–4, 307–8, 312–13, Salience network (SLN), 106–7, 110,
312f, 315–21, 316f, 318f, 323–24, 323f 116–17, 117f
exercise effects, 443–44, 446 Scaffolding theory of aging and cognition
hippocampus interactions, 278, 281 (STAC), 209, 214
hypertension, 393, 394, 400 Scaffolding theory of aging and cognition-
imagery, 224–26, 225f revised. see STAC-r model
longitudinal imaging, 162–65, Schizophrenia
163–64f, 175 COMT polymorphism, 429–30
middle-aged brain, 372, 373 dopamine function in, 50, 52
over-recruitment, 113, 163–64, 185, 188, hypertension effects, 390
190, 341, 345, 421 RM-ANOVA testing, 171
target detection, 210–11, 215–16 Selective attention, inhibitory control
working memory, executive function, 236– activity vs. functional connectivity, 207–8
42, 240–42f, 249 brain volumetrics, 219, 222f, 224–25, 225f
605
Subject Index 605
change detection (object), 219–22f, principles, applications of, 14t, 22–23

219–23 tissue lesion detection, 22–23
change detection (spatial), 211–13, white-matter hyperintensities, 88–91,
242–43 89f, 138
cognitive aging models, 208–9 T1-MPRAGE, 14t, 15–16, 15f
COMT polymorphism, 429–30 Tolcapone, 515
dedifferentiation, 219, 224 TR/TE gradient-echo sequence, 16
discrimination (feature), 217–18 Tumor necrosis factor (TNF)
discrimination (object), 223–24 antagonists, 518
ERP components, 208–10, 221f T2-weighted MRI, 14t, 15, 22–23
expanded inhibitory deficit hypothesis,
247–48 Val66Met polymorphism, 422–23, 423f
feature, 213–18 Val158Met polymorphism, 424–25
functional connectivity, 220–24, 222f, 319, Vascular depression hypothesis, 513–14,
321–22 517–18, 520t
hypertension and, 391 Victoria Longitudinal Study, 462
inhibitory deficit hypothesis, 244–48 Voxel-based morphometry (VBM), 443, 444
internal, 224–26, 225f
object, 218–24, 220–22f White-matter hyperintensities
predictive cues, 223, 224 characterization, 88–91, 89f
retrospective cues, 225 DED syndrome, 509
spatial, 209–13 episodic memory, 47f
stimulus detection (feature), 214 exercise effects on, 447
stimulus detection (object), 218–19 functional connectivity, 137–38
stimulus detection (spatial), 209–10 hypertension, 395–97
target detection (feature), 215–16 middle-aged brain, 371, 377
target detection (spatial), 210–11 processing speed, 260–61t
task difficulty, 211, 212 selective attention, inhibitory control, 218
white matter integrity, 219, 222f vascular depression hypothesis, 513–14
Selective improvement hypothesis, 442 White-matter integrity
Selegiline, 515 amyloid burden in, 43–44, 134–35,
Siemens MRI scanners, 16 138–39, 145
Spoiled gradient echo (SPGR), 16 anterior-posterior gradient, 78–79, 79f
STAC-r model APOE4 and, 421
middle-aged brain, 363–64, 364f, 373 axial diffusivity (AD), 73, 74f, 76,
principles, 363–64, 364f 80–87, 86f
working memory, executive function, behavioral measures, variability vs. central
251–52, 251f tendency, 84–85
Stress, chronic, 518–19 cognitive decline, 137
Susceptibility weighted imaging (SWI), cognitive training, 472, 474
27–29, 28f compensation, 87–88, 145–46, 164–65,
Suspected non-AD pathology (SNAP), 37 209, 219
DED syndrome, 509–10
TaqIA polymorphism, 428 disconnection mechanism, 513–14
Tau imaging DTI/behavior measures, 82–85
Alzheimer’s disease, 48–49, 488f, DTI measures, age-related differences in,
492, 493 76, 77f
neurofibrillary tangles, 372, 492 DTI measures of, 73–76, 74–75f, 137,
principles, 48–49 395–96, 417, 448–49
T2-FLAIR episodic memory, 47f
606
606 Subject Index
White-matter integrity (Cont.) APOE interactions, 249, 428, 490

exercise effects on, 446–49 associative SMEs, 308–9
fornix, 87 auditory processing tasks, 247
fractional anisotropy (FA) ( see fractional BDNF interactions, 428, 452, 490
anisotropy (FA)) change detection (feature), 216–17
functional connectivity, 85–87, 86f, change detection (object), 219–22f,
138–40 219–23
global vs. tract-specific effects, 81–82 change detection (spatial), 211–13,
glucose hypometabolism in, 37–39, 242–43
38t, 138 cognitive training, 466–70, 467–69f
hippocampus, 87 cohort effects, 250
hyperintensities ( see white-matter compensation, 238–42, 240–42f,
hyperintensities) 251–52, 251f
hypertension, 394–96 COMT/DBH interactions, 428
iron accumulation, 91–93 concepts of, 235–36
longitudinal imaging, 159–61, 159f, 160t, 174 default mode network (DMN), 247
magnetic susceptibility, 91–93 dopamine/cognitive functions
mean diffusivity (MD), 73, 74f, 76, 80–87, associations, 53–54
86f, 140–41, 142f, 417 ERP studies, 244–45
middle-aged brain, 371 exercise in, 250, 442
myelination in, 71 expanded inhibitory deficit hypothesis,
myelodegeneration hypothesis, 79–80 247–48
network specificity, 82–83 feature binding, 242–44
neurobiological mechanisms, 80–81 fMRI studies, 240–42, 240–42f
oligodendrocytes, 76, 77f, 92 functional connectivity, 113
processing speed, 260t hippocampus, 237
radial diffusivity (RD), 73, 74f, 76, hypertension and, 391, 392
80–87, 86f inhibitory deficit hypothesis, 244–48
rodent models, 44 interference resolution, 246–47
selective attention, inhibitory control, item recognition, 238–41, 240f
219, 222f lifestyle, 248–50
structural-functional relations, 85–87, 86f middle-aged brain, 367t, 368
structural-functional relations, age-related modifiability of, 248–50
differences in, 87–88, 144 neural efficiency, 240–41, 240f
target detection, 211 neural enriching, depleting factors,
Val66Met polymorphism, 422–23, 423f 251–52, 251f
Val158Met polymorphism, 424–25 overactivity, task-related, 238–40,
working memory, executive function, 246–47
237–38 predictive cues, 223, 224
Working memory, executive function prefrontal cortex, 236–42, 240–42f, 249
activities of daily living (ADLs), 248–49 retrieval dynamics, 245–47
age group by load interaction, 240–42, retrospective cues, 225
241–42f risk factors, 248–50
aging vs. age-related effects, 250 structural aging, 236–38
allostatic load, 249 test-retest effects, 167, 170
amyloid-cognition relationships, 44–48, visual, 242–44
46–47f, 490 white-matter integrity, 237–38

Cabeza, Roberto - Nyberg, Lars - Park, Denise C-Cognitive Neuroscience of Aging Linking Cognitive and Cerebral Aging-Oxford University Press (2017)

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Cabeza, Roberto - Nyberg, Lars - Park, Denise C-Cognitive Neuroscience of Aging Linking Cognitive and Cerebral Aging-Oxford University Press (2017)

Hochgeladen von

Copyright:

Verfügbare Formate

i

Linking Cognitive and Cerebral Aging

Published in the United States of America by Oxford University Press

© Oxford University Press 2017

First Edition published in 2005

All rights reserved. No part of this publication may be reproduced, stored in

You must not circulate this work in any other form

Library of Congress Cataloging-​in-​Publication Data

1 MRI Measures of Aging: Methodological Issues â•… 9

2 Molecular Imaging of Aging and Neurodegenerative Disease â•… 35

3 Age Differences in Structural Connectivity: Diffusion Tensor Imaging

5 Multiâ•‰modal Imaging of the Aging Brain â•… 131

6 Structural and Functional Imaging of Aging: Longitudinal Studies â•… 155

7 Interpreting Age-â•‰Related Differences in Memory-â•‰Related Neural Activity â•… 183

8 Selective Attention and Inhibitory Control in the Aging Brain 207

9 Working Memory and Executive Functions In the Aging Brain 235

10 Neural Correlates of Age-​Related Slowing 259

11 The Aging Hippocampus: Linking Animal and Human Research 273

12 Episodic Memory Encoding and Retrieval in the Aging Brain 301

13 Emotion and Emotional Memory 337

III Health and Disease

14 The Middle-​Aged Brain: A Cognitive Neuroscience Perspective 363

15 The Modifying Role of Hypertension in Cognitive and Brain Aging 389

16 Genetics and Cognitive Neuroscience of Aging 415

17 Effects of Exercise on Cognition, Brain Structure, and Brain Function

18 The Link of Intellectual Engagement to Cognitive and Brain Aging 461

19 The Challenges of Disambiguating Preclinical Alzheimer’s Disease

20 Late-​Life Depression: Translating Neurobiological Hypotheses

George S. Alexopoulos Roberto Cabeza

Lars Bäckman Kirk I. Erickson

Jaclyn H. Ford Anders Lundquist

Adam Gazzaley Cindy Lustig

Cheryl L. Grady David J. Madden

Anna Rieckmann Craig E. L. Stark

Karen M. Rodrigue Shauna M. Stark

Michael D. Rugg Kristine B. Walhovd

Timothy A. Salthouse Wei-​Chun Wang

Reisa Sperling Theodore P. Zanto

BOLD signal (%)

Control Time delay Acquisition

Label Time delay Acquisition

D1 receptor (11C-SCH22390) D2 receptor (11C-Raclopride) Dopamine transporter (11C-Altropane)

Young Old Young Old Young Old

(A) (B) (C) λ1

(A) (B) (C) (D)

Young (< 52 years) Elderly (≥ 52 years)

Younger Middle Older

Resting functional connectivity

Left hippocampus Right hippocampus

Age decrease Age increase

P < 0.05 P < 0.001

(A) Anterior (B)

(A) L DLPFC R DLPFC (B) (C) L DLPFC

L IPL RSC HPC

MTL conn. HPC conn. Amygdala conn.

YAs > OAs OAs > YAs

Lateral PFC Hippocampus Ventral visual processing areas

(B) A IncLg > ConLg B 2.3 3.2

(A) (B) 0.8

(B) Caudate nucleus RIGHT CAUDATE NUCLEUS

Amyloid and Tau PET imaging

2 Cognitive Neuroscience of Aging

Number of papers Number of citations for

Library of Congress Cataloging-in-Publication Data

10 Neural Correlates of Age-Related Slowing 259

14 The Middle-Aged Brain: A Cognitive Neuroscience Perspective 363

20 Late-Life Depression: Translating Neurobiological Hypotheses

Timothy A. Salthouse Wei-Chun Wang

T1-MPRAGE Provide brain volumetric 4–7 min 3D Acquisition,

Functional Measurement of 5–10 min Multislice acquisition,

DTI Probe microstructural 4–5 min Multislice acquisition,

Proton density Help detect tissue 3–5 min Multislice acquisition,

T2-weighted Detect tissue lesions 2–4 min Multislice acquisition,

T2-FLAIR Detect tissue lesions 3–5 min 2D version:

Radial diffusivity ( RD ) = ( λ 2 + λ 3 ) / 2, [8]‌

Axial diffusivity ( AxD ) = λ1 ,[9]‌

Tissue Lesion Detection with Proton-Density, T2-weighted,

S fMRI,corr = S fMRI,uncorr / CVR,[10]