Sie sind auf Seite 1von 5



Myocardial edemaa new clinical entity?

Matthias G. Friedrich
Abstract | Edema is a generic component of the tissue response to acute injury and,
therefore, an important diagnostic target for assessing the acuity of tissue damage
in vivo. In the past, edema could not be used as a diagnostic target, because even
histological techniques failed to provide reliable qualitative or even quantitative
data on its presence, extent, and regional distribution. Cardiac MRI is about to
change that. Using water-sensitive MRI, visualization of myocardial edema in vivo
is possible within a few breath holds, without using radiation or contrast agents.
Edema imaging provides useful incremental diagnostic and prognostic information
in a variety of clinical settings associated with suspected acute myocardial injury. In
combination with scar imaging, MRI differentiates reversible from irreversible injury
and can quantify myocardial salvage after coronary revascularization. With this unique
contribution, MRI of edema should be considered an essential diagnostic tool and,
as part of a multitarget MRI scan, illustrates the exceptional role of comprehensive
cardiac MRI in diagnosing and staging myocardial diseases safely and efficiently.
Friedrich, M. G. Nat. Rev. Cardiol. 7, 292296 (2010); published online 23 March 2010;

It is not only that there is water in the world,

but there is a world in water.
Dgen Zenji (12001253)

Myocardial edema formation

Water is a major constituent of all living

tissues and is essential for virtually all
protein interactions. Water molecules can
permeate cellular or vascular membranes,
but under physiological conditions these
compartments are kept in equilibrium by
the carefully regulated binding of water to
proteins or other complex molecules1 and
by osmotic forces, which are strongly influenced by the concentration of proteins and
Na+ (Figure 1a). Almost 80% of the myocardium is water and about 78% of this
water is held in intracellular compartments.
However, changes in the transmembranous
osmotic gradientfor example by altered
molecular fluid compositioncan result in
compartmental water loss or accumulation.
Since cell membranes do not generate
or sustain hydrostatic pressure gradients,
disturbances typically cause cell swelling,
increased interstitial water accumulation,
or both. The resulting edema (from idema,
Competing interests
The author declares no competing interests.

the Greek word for swelling) is an integral

part of a generic tissue response to any
insult, regardless of its etiology (mechanical, toxic, or ischemic). In no-flow ischemia
of 3 min or more, the breakdown of the
energy-dependent Na+/K+ pump leads to a
net cellular uptake of Na+ with a subsequent
increase in the intracellular osmotic pressure, resulting in net inflow of water and
cellular edema (Figure 1b).2 Acute ischemia
has another important consequence; usually
92% of water molecules in vivo are bound
to proteins, ribonucleic acids, and other
molecules.1 This is the reason why jellyfish
do not simply burst when injured, despite
their water content of about 95%. Ischemiainduced acidosis, however, causes changes
in ionization and spatial conformation of
plasma proteo glycans and intracellular
proteins with subsequent release of water
molecules.3 The fraction of free (unbound)
water, therefore, increases.
Ischemia of more than 6090 min results in
damage to capillary membranes, which are less
susceptible to ischemia than cardiomyocytes,
and leakage of water from the intravascular
space into the interstitial space,4 leading to
interstitial edema (Figure 1c). Finally, cardiomyocyte death (oncosis) is followed by
necrosis, a cascade of mostly inflammatory

292 | MAY 2010 | voluMe 7

reactions for removing dead cells and debris.5

In acute coronary reperfusion after prolonged (for more than 3060 min) ischemia, cellular edema is augmented by
the sudden normalization of the extracellular osmotic pressure, which promotes
extracellular water accumulation.6

Consequences of myocardial edema

Myocardial edema has important consequences; increased hydrostatic pressure

within the interstitial space can exacerbate
the extent of necrosis by capillary compression.7 In addition, myocardial edema
decreases cardiac output and, when chronic,
can cause interstitial fibrosis.8 Edema may
also contribute to postischemic myocardial
dysfunction (stunning), arrhythmia,6 and
reduced ventri cular compliance. 9 The
precise mechanisms of how edema affects
function, long-term tissue composition,
and electrical stability remain to be studied
further. However, increased stiffness and
reduced compliance of edematous myocardium will clearly affect diastolic and
systolic function substantially, and altered
fluid composition has an impact on intramyocardial conduction and hence arrhythmia. The formation of myocardial fibrosis
through edema could be related to changes
in intramyocardial pressure and to the high
protein content of edema. 8 Myocardial
edema is, therefore, an important diagnostic target for assessing the acuity, extent, and
severity of tissue damage in vivo.

Assessment of myocardial edema

Interestingly, the exact physiological pathways of edema are still not well understood;5
this is likely to result from a lack of appropriate in vivo visualization methods. Cellular
edema is difficult to assess under the microscope; even with a 100% increase in cell
volume, the increase in cellular diameter
is only 26%.5 Although electron microscopy may show features such as membrane
blebbing, detachment of actin filaments,
mitochondrial condensation, and swelling of the endoplasmic reticulum and the
Golgi apparatus, this method is not suitable
for assessing the regional extent and distribution of edema. Colloidal suspensions,
such as carbon black, which are retained
in the vessel wall and leaking agents, such

2010 Macmillan Publishers Limited. All rights reserved







Bound water

Free water

Plasma cell




Na /K pump


Dysconformational protein

Figure 1 | A simplified schematic view of edema evolution in ischemic injury to the myocardium. a | Normal equilibrium. Under normal physiological
conditions, the intracellularextracellular water balance is kept in equilibrium by an active, ATp-dependent Na +/K+ exchange and by the binding of
water to complex molecules such as intracellular proteins. b | Early phase of injury: intracellular edema. very early after the onset of hypoxia, water
is released from proteins because of lactate-induced acidosis, increasing the intracellular fraction of free (unbound) water. Furthermore, the
mismatch between O2 supply and demand causes a drop in ATp and leads to Na+/K+ pump failure. The increase in intracellular Na+ raises
the intracellular osmotic pressure and pulls water from the interstitial compartment into the cells, leading to swelling and blebbing. At this stage
of intracellular edema, capillary membranes are still intact because they are less sensitive to ischemia than are cardiomyocytes. c | Late phase of
injury: interstitial edema. If ischemia persists, capillary membranes become permeable to complex molecules including proteins and plasma cells.
Extravasation of fluids and inflammation secondary to microhemorrhage lead to interstitial edema.

as Evans blue, are being used as surrogate

markers for increased membrane permeability. These substances do not, however,
allow the in vivo assessment of edema and
its dynamic evolution.


Infarcted myocardium
Normal myocardium

T2 relaxation time (ms)

MrI for the assessment of edema

MrI is unique among in vivo cardiovascular
imaging methods in its capability to generate
a native pathology-specific tissue contrast.10
This modality is also free from radiation and
does not require the use of contrast agents
for visualizing edema. unlike ultrasound or
CT, MrI primarily reflects chemical tissue
composition instead of translucency or
reflection. In MrI, the relaxation properties of protons after radiofrequency pulses
are the main determinants of tissue contrast.
Importantly, proton relaxation time is a
function of the molecular environment. For
example, the transverse component of proton
relaxation time (T2) is short (6080 ms at
1.5 T) in a lipid-rich environment, but long
(~2,000 ms at 1.5 T) when protons are part
of water molecules (that is, in effusion or
edema).11 Importantly, a large difference in
T2 has also been observed between proteinbound water and free water.3 using specific
MrI scanner settings, images with a high
sensitivity to long T2 can be acquired. Tissue
with high water content will appear bright
on these T2-weighted images, because of
the stronger signal intensity feedback from
protons within water molecules.
A close correlation has been shown
between T2 and edema in the heart.12 As in
the brain, T2 lengthening in the myocardium



5 min

30 min

60 min 120 min 180 min 240 min 300 min

414 days

Time after occlusion

Figure 2 | Change in T2 relaxation after experimental coronary occlusion. very early after the
onset of hypoxia, there is an increase in T2 relaxation times in the tissue, reflecting edema.
Abbreviation: pRE, before occlusion. permission obtained from the Radiological society of North
America Tscholakoff, D. et al. Radiology 159, 667672 (1986).

during ischemia is associated with a

3% increase in overall water content. 13
Interestingly, a study of brief ischemic
episodes (15 min of ischemia followed by
2 min of reflow after reperfusion) showed
no increase in water content 14 and raised
questions about whether a relatively small
increase in total water content can induce
changes in T2. The absolute water content,
however, is not the only mechanism that
can cause T2 changes; the movement of
water molecules from the extracellular to
the intracellular compartmentas in pure


cellular edemaincreases T2 of the cytoplasm and nucleus.15 Furthermore, the T2

of intracellular water increases substantially when it is released from proteins (as
in acute ischemia).3 In fact, as the T2 of free
water is about 40 times longer than that of
bound water,16 the overall T2 of ischemic
tissue is significantly prolonged, even in the
absence of an increase in net water content.
Experimental data have confirmed that this
intracellular water fractionation into free
and bound molecules has a stronger impact
on T2 than the overall water content.17
voluME 7 | MAY 2010 | 293

2010 Macmillan Publishers Limited. All rights reserved


Figure 3 | Myocardial edema in a patient 2 days after acute, reperfused myocardial infarction (mid-ventricular, short-axis views). a | Image from
T2-weighted MRI showing high signal intensity of the anteroseptal segment (arrow). b | Computer-aided signal intensity analysis of the T2-weighted
image with color-coded display of relative signal intensity, normalized to skeletal muscle. Blue indicates a signal intensity ratio of myocardium/
skeletal muscle of 2.0, indicating edema, green indicates normal signal intensity (1.41.9). c | Contrast-enhanced image (late gadolinium
enhancement) showing high signal intensity reflecting increased contrast accumulation in necrotic myocardium (arrow). d | Computer-aided signal
intensity analysis with color-coded display of relative signal intensity, normalized to remote myocardium. Red indicates a signal intensity of
5 standard deviations above remote myocardium. The comparison of edema with necrosis shows a complete overlap, indicating lack of any
relevant myocardial salvage.

Figure 4 | Regional myocardial edema in a patient with active myocarditis. a | T2-weighted MRI image showing a focus with high signal intensity in
the inferoseptal segment (arrows). b | Computer-aided signal intensity analysis of the T2-weighted image with color-coded display of relative signal
intensity, normalized to skeletal muscle. Blue indicates a signal intensity ratio of myocardium/skeletal muscle of 2.0, indicating edema, green
indicates normal signal intensity (1.41.9). c | Contrast-enhanced image (late gadolinium enhancement) showing a small focus with high signal
intensity in the same region indicating a small necrosis (arrow). d | Computer-aided signal intensity analysis of the necrosis image with color-coded
display of relative signal intensity, normalized to remote myocardium. Red indicates a signal intensity of 5 standard deviations above remote
myocardium. The comparison of edema with necrosis shows that only part of the inflamed tissue is irreversibly injured.

Figure 5 | Diffuse myocardial edema in a

patient with active myocarditis. Image from
T2-weighted MRI showing extensive
myocardial edema as visible by its high signal
intensity (arrows). The fact that the regional
distribution is not reflective of a coronary
perfusion bed and is mainly subepicardial
indicates a nonischemic etiology.

Contrast-media-enhanced MrI, with

selective suppression of signal in normal
myocardium (late gadolinium enhancement; lGE), is the in vivo gold standard
for visualizing myocardial infarction.
Data obtained with electron probe X-ray
microanalysis, however, indicate that lGE
exclusively reflects irreversible injury.18 In
acute, reperfused myocardial infarction
(with irreversible damage surrounded by
salvaged, edematous myocardium), a consistent mismatch between T2-weighted
imaging (reflecting edema in both areas) and
lGE (reflecting necrosis) has been repeatedly reported.19,20 In acute stress-induced
(Takotsubo) cardiomyopathythe prime
example of acute, reversible myocardial
diseaseedema as defined by high T2 signal
intensity is a consistent feature, whereas

294 | MAY 2010 | voluMe 7

areas with increased signal in lGE images

are observed very infrequently.21 Therefore,
contrast-enhanced MrI (lGE) cannot be
used as a marker for myocardial edema, but
as a complementary component of cardiac
MrI to visualize myocardial injury.

Edema in acute myocardial injury

In the early 1980s, Higgins et al. reported that

experimental coronary occlusion in dogs is
followed by an increase in T2 within minutes
after the onset of ischemia (Figure 2).13,22
These findings were confirmed in humans,23
but during the late 1980s and 1990s, clinical
imaging of edema was challenged by limited
availability of cardiac-specific MrI hardware
and software, and a lower signal-to-noise
ratio of T2-weighted imaging as compared
with other MrI protocols. Improved scanner

2010 Macmillan Publishers Limited. All rights reserved



Figure 6 | Myocardial edema of myocardial salvage after acute, reperfused myocardial infarction (midventricular short-axis views). a | Image from
T2-weighted MRI showing high signal intensity of the anterior, anteroseptal, and inferoseptal segments (arrows). b | Computer-aided signal intensity
analysis of the T2-weighted image with color-coded display of relative signal intensity, normalized to skeletal muscle. Blue indicates a signal intensity
ratio of myocardium/skeletal muscle of 2.0, indicating edema, green indicates normal signal intensity (1.41.9). c | Contrast-enhanced image (late
gadolinium enhancement) showing high signal intensity reflecting increased contrast accumulation in necrotic myocardium (arrow). d | Computeraided signal intensity analysis of the late gadolinium enhancement image with color-coded display of relative signal intensity, normalized to remote
myocardium. Red indicates a signal intensity of 5 standard deviations above remote, uninjured myocardium. The comparison of edema (panels a
and b) with necrosis (panels c and d) shows myocardial edema in areas without necrosis, indicating myocardial salvage (asterisks in panel b).
Images courtesy of Dr Ingo Eitel.

and sequence technology,24 however, helped

achieve an image quality suitable for clinical
application of edema imaging in the heart
(Figure 3).
Early clinical applications of this technology included the detection of peri-infarct
edema in reperfused infarction.25 Myocardial
edemaas defined by a high signal intensity on T2-weighted MrIwas introduced
as a specific marker for the acuity of myocardial injury, being apparent in acute, but
not in chronic, myocardial infarction.19 In
patients, the addition of edema-sensitive
MrI increased the diagnostic accuracy of
a comprehensive cardiac MrI protocol to
detect acute coronary syndrome.26
Edema imaging is also useful in other
acute myocardial diseases, such as transplant rejection,27 myocarditis,28 and stressinduced (Takotsubo) cardiomyopathy. 29
Notably, as for necrosis, the regional distribution of edema in these entities typically
includes the middle or subepicardial layer
of the myocardium (Figures 4 and 5) and is
thus distinct from ischemic injury, which
predominantly affects the subendocardium
or is transmural (Figures 3 and 6).

Edema in reversible myocardial injury

Myocardial T2 increases after a few minutes

of ischemia, long before tissue injury
becomes irreversible.22 In animals, a signal
intensity increase in T2-weighted MrI was
observed within less than 30 min of intermittent coronary artery occlusion in the absence
of detectable necrosis.30 Additional evidence
for the potential reversibility of edematous
tissue changes comes from MrI studies of
myocardial salvage after acute myocardial

infarction. This approach has been validated

in animal studies of reperfused infarction31
and shown to consistently visualize reversible ischemic injury in reper fused myocardial infarction (Figure 6).20,32 Data from
Eitel et al. show that the amount of myocardial salvage as defined by this approach
also predicts patient outcome.33

Future challenges

Technical challenges for MrI of myocardial edema include the susceptibility of

T2-weighted MrI to blood flow and arrhythmias, which alter the signal by the motion of
the spins during the relatively long period
between excitation of spins and signal
read-out, and the limited signal-to-noise
ratio of standard protocols. Mapping of T2
could overcome such limitations by allowing for a direct measurement of T2, instead
of using signal intensity as a marker of T2
changes.34,35 Furthermore, new sequences
using shorter acquisition protocols, such as
single-shot techniques,35 are likely to lead to
a more robust image quality. Furthermore,
MrI settings and quantitative image analysis of diffuse myocardial edema have to be
simplified and standardized. Finally, more
research is required to better understand
the dynamic pathophysiology of myocardial
edema and further specify the contribution
of edema imaging in various clinical settings.
Examples include suspected, but not confirmed, acute reocclusion of stents, unstable
angina, and cardiotoxicity.
Myocardial edema imaging is ready for
prime time, because it enables myocardial
salvage to be used as a marker for assessing
the success of coronary revascularization in


clinical settings and as a primary end point

in clinical trials. While more mortality data
will have to be acquired, the obvious value
of myocardial edema imaging for patients
and for research should be emphasized.
Manufacturers of magnetic resonance scanners, writing groups for recommendations
and guidelines, as well as cardiologists and
cardiovascular imagers should see this goal
as a priority. Scanning protocols have to be
further simplified, and post-MrI processing
and evaluation procedures need to be standardized and made less time-consuming.
MrI protocols should be optimized for
more-consistent image quality. While new
developments, such as T2 mapping, may
change technical details they will not alter the
important contribution of edema imaging.
The safety profile, unparalleled versatility, and cost-efficiency of comprehensive
cardiac MrI protocols are very strong arguments for replacing techniques that use
radiation or radioactive material wherever
possible. Societal recommendations and
guidelines will need to reflect the importance of tissue characterization. Combined
cardiac MrI of ventricular function, and
myocardial edema and scar should be considered the tool of choice in patients with
suspected myocardial injury.
Stephenson Cardiovascular MR Centre, Libin
Cardiovascular Institute of Alberta, University of
Calgary, 1403 29th Street NW, Calgary,
AB T2N 2T9, Canada.
I want to cordially thank John v. Tyberg, James Hare,
and Jordin D. Green for their helpful review of the
manuscript, Marc Henry for his valuable thoughts,
and Ingo Eitel for providing the images for Figure 6.

voluME 7 | MAY 2010 | 295

2010 Macmillan Publishers Limited. All rights reserved














Henry, M. The state of water in living systems:

from the liquid to the jellyfish. Cell. Mol. Biol.
(Noisylegrand) 51, 677702 (2005).
Landis, E. M. Micro-injection studies of capillary
permeability: III. The effect of lack of oxygen on
the permeability of the capillary wall to fluid
and to the plasma proteins. Am. J. Physiol. 83,
528542 (1927).
Kuntz, I. D., Brassfield, T. s., Law, G. D. &
purcell, G. v. Hydration of macromolecules.
Science 163, 13291331 (1969).
Kloner, R. A. et al. Ultrastructural evidence of
microvascular damage and myocardial cell
injury after coronary artery occlusion: which
comes first? Circulation 62, 945952 (1980).
Majno, G. & Joris, I. Cells, Tissues, and Disease.
Principles of General Pathology (Oxford
University press, New York, 2004).
Garcia-Dorado, D. & Oliveras, J. Myocardial
edema: a preventable cause of reperfusion
injury? Cardiovasc. Res. 27, 15551563
DiBona, D. R. & powell, W. J. Jr. Quantitative
correlation between cell swelling and necrosis
in myocardial ischemia in dogs. Circ. Res. 47,
653665 (1980).
Laine, G. A. & Allen, s. J. Left ventricular
myocardial edema. Lymph flow, interstitial
fibrosis, and cardiac function. Circ. Res. 68,
17131721 (1991).
Desai, K. v. et al. Mechanics of the left
ventricular myocardial interstitium: effects
of acute and chronic myocardial edema.
Am. J. Physiol. Heart Circ. Physiol. 294,
H2428H2434 (2008).
Friedrich, M. G. Tissue characterization of acute
myocardial infarction and myocarditis by
cardiac magnetic resonance. JACC Cardiovasc.
Imaging 1, 652662 (2008).
Budinger, T. F. & Lauterbur, p. C. Nuclear
magnetic resonance technology for medical
studies. Science 226, 288298 (1984).
Williams, E. s., Kaplan, J. I., Thatcher, F.,
Zimmerman, G. & Knoebel, s. B. prolongation
of proton spin lattice relaxation times in
regionally ischemic tissue from dog hearts.
J. Nucl. Med. 21, 449453 (1980).
Higgins, C. B. et al. Nuclear magnetic
resonance imaging of acute myocardial
infarction in dogs: alterations in magnetic
relaxation times. Am. J. Cardiol. 52, 184188

14. Whalen, D. A. Jr, Hamilton, D. G., Ganote, C. E.

& Jennings, R. B. Effect of a transient period of
ischemia on myocardial cells. I. Effects on cell
volume regulation. Am. J. Pathol. 74, 381397
15. Hsu, E. W., Aiken, N. R. & Blackband, s. J.
Nuclear magnetic resonance microscopy of
single neurons under hypotonic perturbation.
Am. J. Physiol. 271, C1895C1900 (1996).
16. Hazlewood, C. F., Chang, D. C., Nichols, B. L. &
Woessner, D. E. Nuclear magnetic resonance
transverse relaxation times of water protons in
skeletal muscle. Biophys. J. 14, 583606
17. Knight, R. A. et al. Temporal evolution of
ischemic damage in rat brain measured by
proton nuclear magnetic resonance imaging.
Stroke 22, 802808 (1991).
18. Rehwald, W. G., Fieno, D. s., Chen, E. L.,
Kim, R. J. & Judd, R. M. Myocardial magnetic
resonance imaging contrast agent
concentrations after reversible and irreversible
ischemic injury. Circulation 105, 224229
19. Abdel-Aty, H. et al. Delayed enhancement and
T2-weighted cardiovascular magnetic
resonance imaging differentiate acute from
chronic myocardial infarction. Circulation 109,
24112416 (2004).
20. ORegan, D. p. et al. Cardiac MRI of myocardial
salvage at the peri-infarct border zones after
primary coronary intervention. Am. J. Physiol.
Heart Circ. Physiol. 297, H340H346 (2009).
21. Eitel, I. et al. Differential diagnosis of
suspected apical ballooning syndrome using
contrast-enhanced magnetic resonance
imaging. Eur. Heart J. 29, 26512659 (2008).
22. Tscholakoff, D., Higgins, C. B., McNamara, M. T.
& Derugin, N. Early-phase myocardial infarction:
evaluation by MR imaging. Radiology 159,
667672 (1986).
23. McNamara, M. T. et al. Detection and
characterization of acute myocardial infarction
in man with use of gated magnetic resonance.
Circulation 71, 717724 (1985).
24. simonetti, O. p., Finn, J. p., White, R. D., Laub, G.
& Henry, D. A. Black blood T2-weighted
inversion-recovery MR imaging of the heart.
Radiology 199, 4957 (1996).
25. stork, A. et al. Comparison of an edemasensitive HAsTE-TIRM sequence with delayed
contrast enhancement in acute myocardial

296 | MAY 2010 | voluMe 7











infarcts [German]. Rofo 175, 194198

Cury, R. C. et al. Cardiac magnetic resonance
with T2-weighted imaging improves detection
of patients with acute coronary syndrome in
the emergency department. Circulation 118,
837844 (2008).
Marie, p. Y. et al. Detection and prediction of
acute heart transplant rejection with the
myocardial T2 determination provided by a blackblood magnetic resonance imaging sequence.
J. Am. Coll. Cardiol. 37, 825831 (2001).
Abdel-Aty, H. et al. Diagnostic performance of
cardiovascular magnetic resonance in patients
with suspected acute myocarditis: comparison
of different approaches. J. Am. Coll. Cardiol. 45,
18151822 (2005).
Abdel-Aty, H., Cocker, M. & Friedrich, M. G.
Myocardial edema is a feature of Tako-Tsubo
cardiomyopathy and is related to the severity of
systolic dysfunction: insights from T2-weighted
cardiovascular magnetic resonance. Int.
J. Cardiol. 132, 291293 (2009).
Abdel-Aty, H., Cocker, M., Meek, C., Tyberg, J. v.
& Friedrich, M. G. Edema as a very early marker
for acute myocardial ischemia: a cardiovascular
magnetic resonance study. J. Am. Coll. Cardiol.
53, 11941201 (2009).
Aletras, A. H. et al. Retrospective determination
of the area at risk for reperfused acute
myocardial infarction with T2-weighted cardiac
magnetic resonance imaging: histopathological
and displacement encoding with stimulated
echoes (DENsE) functional validations.
Circulation 113, 18651870 (2006).
Friedrich, M. G. et al. The salvaged area at risk in
reperfused acute myocardial infarction as
visualized by cardiovascular magnetic resonance.
J. Am. Coll. Cardiol. 51, 15811587 (2008).
Eitel, I. et al. prognostic significance and
determinants of myocardial salvage assessed
by cardiovascular magnetic resonance in acute
reperfused myocardial infarction. J. Am. Coll.
Cardiol. (in press).
Giri, s. et al. T2 quantification for improved
detection of myocardial edema. J. Cardiovasc.
Magn. Reson. 11, 56 (2009).
Green, J. D., Clarke, J. R., Flewitt, J. A. &
Friedrich, M. G. single-shot steady-state free
precession can detect myocardial edema in
patients: a feasibility study. J. Magn. Reson.
Imaging 30, 690695 (2009).
2010 Macmillan Publishers Limited. All rights reserved