Cognitive Abilities in Preterm and Term-Born Adolescents

Luke A. Schneider, PhD1, Nicholas R. Burns, PhD2, Lynne C. Giles, PhD3, Ryan D. Higgins, BSc1, Theodore J. Nettelbeck, PhD2,
Michael C. Ridding, PhD1, and Julia B. Pitcher, PhD1

Objective To investigate the influence of a range of prenatal and postnatal factors on cognitive development in
preterm and term-born adolescents.
Study design Woodcock-Johnson III Tests of Cognitive Abilities were used to assess general intellectual ability
and 6 broad cognitive abilities in 145 young adolescents aged approximately 12.5 years and born 25-41 weeks
gestational age (GA). To study potential links between neurophysiologic and cognitive outcomes, corticomotor
excitability was measured using transcranial magnetic stimulation and surface electromyography. The influence
of various prenatal and postnatal factors on cognitive development was investigated using relative importance
regression modeling.
Results Adolescents with greater GA tended to have better cognitive abilities (particularly general intellectual abil-
ity, working memory, and cognitive efficiency) and higher corticomotor excitability. Corticomotor excitability ex-
plained a higher proportion of the variance in cognitive outcome than GA. But the strongest predictors of
cognitive outcome were combinations of prenatal and postnatal factors, particularly degree of social disadvantage
at the time of birth, birthweight percentile, and height at assessment.
Conclusions In otherwise neurologically healthy adolescents, GA accounts for little interindividual variability in
cognitive abilities. The association between corticomotor excitability and cognitive performance suggests that
reduced connectivity, potentially associated with brain microstructural abnormalities, may contribute to cognitive
deficits in preterm children. It remains to be determined if the effects of low GA on cognitive outcomes attenuate
over childhood in favor of a concomitant increase in the relative importance of heritability, or alternatively, if
cognitive development is more heavily influenced by the quality of the postnatal environment. (J Pediatr
2014;165:170-7).

I
n developed countries, 6%-12% of all births annually are preterm (ie, <37 completed weeks gestation).1 A plethora of
studies have shown associations between preterm birth and later suboptimal neurodevelopmental outcomes. In terms
of identifying the actual effects of reduced gestational age (GA) on neurodevelopment, most have arguably been
confounded by not differentiating GA from birthweight percentile (BW%), and/or including children with clinical histories
of brain lesions or other neurosensory impairments, and rarely including late preterm children (33-37 weeks GA), who
comprise over 70% of all preterm births. Compared with their term-born peers, the late preterm exhibit a high prevalence
of low severity motor, cognitive, and behavioral impairments.2-6 They account for up to 74% of the total burden of dysfunc-
tion because of preterm birth,7 a greater need for special education,2,8,9 lower net income, and a reduced likelihood of
completing a university education.7 These outcomes are not explained by perinatal brain lesions that affect <1% of children
(<10% in those born <32 weeks GA), but more likely by microstructural brain abnormalities not readily detected with stan-
dard magnetic resonance imaging (MRI).10-13
Using transcranial magnetic stimulation (TMS), we previously showed relationships between preterm birth and reduced cor-
ticomotor excitability, neuroplasticity, and functional motor development in early adolescence.4,14 The motor cortex (M1)
contributes to at least some cognitive functions15,16 and a basic TMS measure of corticomotor excitability, the resting motor
threshold (rMT), also correlates with cortical white matter maturation and integrity.17 Here, we investigated if there are also
links between GA, corticomotor excitability, and cognitive abilities, in adolescents born across a range of GAs but without
known brain lesions or neurosensory disabilities. We hypothesized that increased

From the 1Research Center for Early Origins of Health

and Disease, Robinson Institute, School of Pediatrics
BW% Birthweight percentile and Reproductive Health, 2School of Psychology, and
3
GA Gestational age Discipline of Public Health, University of Adelaide,
Adelaide, Australia
GIA General intellectual ability
Supported by the National Health and Medical Research
IRSD Index of relative socioeconomic disadvantage Council of Australia (565344 and 299087 both to J.P.), the
M1 Motor cortex South Australian Channel 7 Childrens Research Foun-
dation, the Womens and Childrens Hospital Research
MRI Magnetic resonance imaging Foundation, and the Faculty of Health Sciences, Univer-
rMT Resting motor threshold sity of Adelaide. The authors declare no conflicts of
interest.
SES Socioeconomic status
TMS Transcranial magnetic stimulation 0022-3476/$ - see front matter. Copyright 2014 Elsevier Inc.
WCH Womens and Childrens Hospital All rights reserved.
http://dx.doi.org/10.1016/j.jpeds.2014.03.030

170
 Vol. 165, No. 1  July 2014

cortical excitability is associated with increased cognitive per-
formance. To better characterize any associations, we also
examined the influence of a range of pre- and postnatal vari-
ables known to influence cognitive abilities, including fetal
growth18 and socioeconomic factors.19 Preliminary results
have been presented in abstract.20
Methods
Stratified recruitment was used to recruit 145 early adoles-
cents (78 males) with parent/primary caregiver written
informed consent (Table). GAs ranged from 25-41 weeks
(34.5  3.5 weeks) and the mean uncorrected age at
assessment was 148.7  9.3 months (ie, 12 years and
5 months, range: 128-168 months). All preterm adolescents
(N = 101) were born between January 1996 and December
1997 at the Womens and Childrens Hospital (WCH),
Adelaide, Australia. Term-born adolescents (N = 44) were
recruited from the preterm childrens schools and from
community newspaper advertisements. Exclusion criteria
were any abnormality on perinatal cranial ultrasound (no
MRI available), any genetic or chromosomal disorder, an
identifiable syndrome, or physical or intellectual disability
that rendered participants unable to follow simple
instructions, in addition to the exclusion criteria
recommended for the safe use of TMS.21 Ineligible children
were screened and removed from the database lists prior to
recruitment. Ethical approval was provided by local WCH,
university, government, and Catholic education human
research ethics committees. All procedures were performed
in accordance with the Declaration of Helsinki (2008
revision).
As this study was part of the broader Preterm Motor and
Cognitive Development study,14 all data collection was per-
formed by investigators blinded to GA, BW%, etc. Each childs
current height, weight, and percentage of body fat, determined
using bio-impedance scales (body composition analyzer, Ta-
nita, Kewdale, Australia) were recorded. Characteristics per-
taining to each preterm (and some term) participants birth
were obtained from WCH Perinatal Statistics collection with
parental written consent. Gestation Related Optimal Weight
software22 was used to calculate each childs actual birthweight
relative to their predicted optimal term weight adjusted for
GA, sex, maternal size, ethnicity, and parity. This BW% is a
marker of fetal growth. The Australian Bureau of Statistics in-
dex of relative socioeconomic disadvantage (IRSD) was calcu-
lated for the address each child went home to following their
birth (1996 National Census; IRSDbirth) and for their current
address (2006 National Census; IRSDcurrent). This composite
measure, which includes educational attainment, occupation,
employment, and income, is a summary of economic and so-
cial conditions of people and households within small
geographic areas (ie, census districts).
Cognitive Abilities Assessment
The age-normed Woodcock-Johnson III Tests of Cognitive
Abilities23 were administered to each participant according
to standardized procedures.24 The Woodcock-Johnson III
Tests of Cognitive Abilities is explicitly linked to the
Cattell-Horn-Carroll theory, which provide a model of the
structure of cognitive abilities.25 We included tests 1-9
from the standard and test 14 from the extended batteries
(see www.assess.nelson.com/pdf/asb-7.pdf for more specific
test details). Combinations of the subtests contribute to

Table. Characteristics and cognitive abilities of the participants by GA group

Early preterm, 32 wk GA Late preterm, 33-36 wk GA Term, 37-41 wk GA
(N = 38) (N = 63) (N = 44) Total (N = 145)
GA (wk) 29.7  2.2* ,
34.8  1.1* 38.1  1.5 34.5  3.5
BW% 37.7  33.0* 37.1  31.5* 56.2  30.0 43.3  32.4
Sex
Males 19 (50%) 36 (57%) 23 (52%) 78 (54%)
Females 19 (50%) 27 (43%) 21 (48%) 67 (46%)
Parity 0.8  1.3 0.8  0.9 11 0.8  1
Birth head circumference (cm) 27.8  2.3*, 32.4  1.9* 34.8  3.5 31.8  3.7
Birth length (cm) 39.0  3.4*, 45.9  2.6* 48.5  4.5 44.8  5
Apgar score 1 min 6.6  1.8*, 7.9  1.5 8.1  1.1 7.6  1.7
Apgar score 5 min 8.6  1.4*, 9.1  0.7 9.2  0.6 91
Child weight at assessment (kg) 40.8  11 44.6  10.4 45.8  11.4 44  11
Child height at assessment (m) 1.5  0.1*, 1.5  0.1 1.5  0.1 1.5  0.1
Child % body fat at assessment 20.6  8.8 20.4  6.6 21.2  7.9 20.7  7.6
IRSD score birth 996.2  109.3 960.8  109.3 995.7  88.5 980  104.8
IRSD score current 1006.5  76 1007.1  88.5 993.0  91.6 1002.7  86.1
GIA 93.8  13 100.7  14.2 99.3  11.2 98.5  13.2
Verbal ability 97.3  10.3 99.5  9.9 97.0  11.9 98.2  10.6
Thinking ability 99.8  14.1 104.8  14 104.3  12.2 103.3  13.6
Cognitive efficiency 89  13.9 98.1  15.6 94.7  16.8 94.7  15.9
Auditory processing 106.3  14.5 110.7  13.4 111.2  16 109.7  14.6
Phonemic awareness 102.5  16.2 106.3  18.4 107.1  17.6 105.5  17.6
Working memory 94.8  13.4*, 102.3  15.2 102.5  12.7 100.4  14.3

Data are mean  SD for each GA group, except for sex N (%) of sample in each GA group.
*Denotes P < .05 compared with the term-born group.
Denotes P < .05 compared with the late preterm group.

171
THE JOURNAL OF PEDIATRICS  www.jpeds.com
cluster scores for 6 broad cognitive abilities: verbal ability,
thinking ability, cognitive efficiency, auditory processing,
phonemic awareness, and working memory; and a global in-
telligence score, general intellectual ability (GIA). GIA is a
composite score derived from principal component analysis
of the narrow cognitive abilities to provide a single best
predictor of cognitive ability, academic performance, career
success, and exceptional achievement.25
TMS
Corticomotor excitability was assessed prior to the cognitive
assessment. Participants were seated in an arm chair with
their hands and forearms supported. Adhesive silversilver
chloride electrodes were applied over the first dorsal inteross-
eous muscles of both hands using a muscle belly-tendon
montage. Electromyogram signals were amplified (1000)
and band-pass filtered (20 Hz-1 kHz) (D360; Digitimer,
Welwyn Garden City, United Kingdom) then digitized at
5.1 kHz with a laboratory interface (CED 1401; Cambridge
Electronic Design, Cambridge, United Kingdom). The
optimal M1 scalp site for TMS was determined for both
hemispheres separately using a hunting procedure.26 The
coil was oriented so that the current induced in the cortex
flowed in a plane perpendicular to the estimated alignment
of the central sulcus in a posterior-to-anterior direction.
Single pulse TMS was used to determine rMT and was
applied with a 70 mm figure-of-eight stimulating coil con-
nected to a monophasic Magstim 2002 magnetic stimulator
(Magstim Co, Whitland, United Kingdom). The rMT was
determined as the lowest TMS intensity that evoked motor
evoked potentials of at least 50 mV peak-to-peak amplitude
in the resting first dorsal interosseous in 5 of 10 trials.
Statistical Analyses
Data were analysed using R statistical analysis freeware
(v 2.13.1; http://www.r-project.org/). Statistical significance
was accepted at a # 0.05. Cognitive outcome variables
were included in the analyses as standard scores based on
age-norms. In separate analyses, we considered weeks GA
as a continuous variable and as a categorical variable (GA
group). For the latter, we used the World Health Organiza-
tion definitions (ie, 37-41 weeks GA: term-born;
33-36 weeks GA: late preterm; 28-32 weeks GA: very preterm;
#27 weeks GA: extremely preterm). GA group comparisons
were made using 1-way ANOVA with polynomial contrasts
and post hoc tests adjusted for multiple comparisons.
Explanatory variables were included in the main analysis if
they correlated with the broad cognitive ability cluster score
of interest and included head circumference and body length
at birth, Apgar score at 1/5 minutes, singleton or multiple
birth, parity, maternal age at childs birth, maternal body
mass index at first antenatal visit, the IRSDbirth and IRSDcur-
rent, the childs current height, weight and percentage of body
fat, laterality quotient (Edinburgh Handedness Inventory27),
and the rMT for both hands. The influence of the explanatory
variables on each of the cognitive cluster scores was assessed
with relative importance regression modeling. We averaged
172
Vol. 165, No. 1
over all orderings of regressors in the model, facilitating the
interpretation of the variance allocations of each regressor
to the total R2.28,29
Results
As there were only 7 extremely preterm children, the very
(ie, 28-32 weeks GA) and extremely (#27 weeks GA) preterm
groups were combined into an early preterm group
(#32 weeks). However, it should be noted that the mean
BW% in the extremely preterm group was 83.7%  8.2%
(range: 72.9%-93.1%) compared with 26.8%  27.0%
(0.05%-84.0%) in the very preterm; that is, the only
extremely preterm children who fulfilled the criteria for
inclusion in the study were those with good fetal growth at
the time of their preterm birth (relative to their individual
predicted potential). Thus, the sample is not representative
of all preterm children, only the most neurologically normal
(ie, those with no known brain lesions or neurosensory dis-
abilities). BW% data are missing for 22 participants for
whom all data necessary for the calculation were not avail-
able. Thirty-three children were excluded; 29 on the basis
of the TMS safety screen and 4 because of ipsilateral
responses to TMS (indicating a possible corticospinal tract
lesion).
GA Group Comparison of Cognitive Abilities and
Corticomotor Excitability
GIA (F[1,142] = 3.49, P = .03), cognitive efficiency
(F[1,142] = 4.12, P = .02), and corticomotor excitability
(rMT-right hand; F[1,135] = 4.24, P = .02) (rMT-left hand;
F[1,129] = 3.17, P = .05) were lower in the early preterm
compared with the late preterm group, but not the term-
born group. They had poorer working memory
(F[1,142] = 4.13, P = .02) than both the late preterm
(P = .03) and term-born (P = .04) children. There were no
differences between any of the groups in verbal ability,
thinking ability, auditory processing, or phonemic aware-
ness. There were no differences between late preterm and
term-born groups in any cognitive domain or in corticomo-
tor excitability. Combining these groups did not alter the dif-
ferences found for the early preterm group.
Relationships between GA, BW%, Corticomotor
Excitability, and Cognitive Abilities Scores
Positive linear relationships were evident between GIA and
GA (R2 = 0.03, F[1,143] = 4.2, P = .04), and also between
GIA and BW% (R2 = 0.05, F[1,121] = 6.2, P = .01), although
GA and BW% explained only a limited amount of the vari-
ability (Figure 1, A and B). An rMT could be obtained
from the left hemisphere (right hand) of 133, and the right
hemisphere (left hand) of 128 participants. The lower the
GA, the higher the stimulus intensity required to obtain
rMT in the left (R2 = 0.06, F[1,126] = 8.1, P = .005) and
right hands (R2 = 0.09, F[1,131] = 13.0, P # .001). BW%
explained a proportion of the variance in rMT of the left
(R2 = 0.05, F[1,105] = 5.5, P = .02), but not the right hand.14
Schneider et al
July 2014 ORIGINAL ARTICLES

Figure 1. Individual relationships between GIA, and A, GA, B, BW%, and rMT in the C, left and D, right hands. Data points are
individual subject results. Dotted lines are 95% CIs.

Individuals with greater corticomotor excitability (ie, lower (parity = 27.6%, b = 2.05, P = .01). Thinking ability
rMT) had better GIA, and this was evident for both the left (R2 = 0.20, F[3,114] = 9.3, P < .001) was better in children
(R2 = 0.04, F[1,126] = 4.7, P = .03) and right (R2 = 0.06, who went home to less social disadvantage after birth
F[1,131] = 8.2, P = .005) rMTs. Lower rMTs were also (38.3%, b = 0.04, P = .002), were taller at assessment
associated with better verbal ability (right hand only,
R2 = 0.04, F[1,131] = 4.9, P = .03), thinking ability (right
hand, R2 = 0.06, F[1,131] = 7.6, P = .006; left hand,
R2 = 0.04, F[1,126] = 4.8, P = .03), and cognitive efficiency
(right hand only, R2 = 0.03, F[1,131] = 4.7, P = .03) but not
auditory processing, phonemic awareness, or working
memory (Figure 1, C and D).

Factors Influencing GIA

The best model for GIA (R2 = 0.18, F[3,124] = 9.4, P < .001)
(Figure 2) indicated higher GIA scores in those children
who went home to less social disadvantage after birth
(IRSDbirth = 59.4%, b = 0.04, P < .001), who had a lower
right hand rMT (rMT-right hand = 21.2%, b = 0.18,
P = .05), and who were taller at assessment
(height = 19.4%, b = 26.99, P = .03).

Factors Influencing Broad Cognitive Abilities

Verbal ability (R2 = 0.17, F[3,135] = 9.4, P # .001) (Figure 3,
A) was better in taller children (height = 36.3%, b = 28.30,
P = .001), who, at birth, went home to less social
disadvantage (IRSDbirth = 36.1%, b = 0.02, P = .004), with
a mother who had given birth to fewer children previously
Cognitive Abilities in Preterm and Term-Born Adolescents
Figure 2. Relative importance regression model of explana-
tory variables contributing to GIA. The P value is given for each
individual regressor in the model. The relative importance of
regressors sums to 100% of the total variance accounted for
by the best regression model.
173
THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. 165, No. 1

Figure 3. Relative importance regression models for explanatory variables contributing to A, verbal ability, B, thinking ability,
C, cognitive efficiency, D, auditory processing, E, phonemic awareness, and F, working memory. The P value is given for indi-
vidual regressors in each model. The relative importance of regressors sums to 100% of the total variance accounted for by the
best regression model for each cognitive ability.

(height = 32.1%, b = 34.00, P = .008), and had a higher BW%
(BW% = 29.6%, b = 0.09, P = .01) (Figure 3, B). Cognitive
efficiency (R2 = 0.16, F[3,135] = 8.6, P < .001) (Figure 3, C)
was highest in females (sex = 33.6%, b = 6.94, P = .004),
with less social disadvantage at birth (IRSD birth = 38.4%,
b = 0.03, P = .004) and greater length at birth
(length = 28.0%, b = 0.65, P = .007). Taller adolescents
(height = 66.1%, b = 41.89, P = .002) with a higher BW%
(BW% = 33.9%, b = 0.08, P = .03) had the best auditory
processing abilities (R2 = 0.12, F[2,120] = 8.5, P < .001)
(Figure 3, D). Phonemic awareness (R2 = 0.20,
F[3,119] = 10.1, P < .001) (Figure 3, E) was better in taller
adolescents (height = 62.3%, b = 65.82, P < .001), with a
greater BW% (BW% = 21.3%, b = 0.10, P = .02), who
were born to a mother who had given birth to fewer
children previously (parity = 16.4%, b = 3.60, P = .01).
174
Working memory was influenced by IRSDbirth, head
circumference at birth, and sex (R2 = 0.17, F[3,135] = 9.5,
P < .001) (Figure 3, F) and was highest in children who
went home to less social disadvantage after birth
(IRSDbirth = 57.4%, b = 0.04, P < .001), had a greater head
circumference at birth (head circumference = 23.2%,
b = 0.85, P = .006), and were female (sex = 19.4%,
b = 5.19, P = .02).
Influence of Social Disadvantage at Birth vs at
Assessment
GA did not correlate with IRSDbirth (r = 0.015, P = .86, N = 143)
as might have been expected.30 IRSDbirth and IRSDcurrent corre-
lated (r = 0.633, P # .0001). Although IRSDcurrent was higher
(ie, less disadvantage) (mean  SD = 1002.7  86.0) than
IRSDbirth (980.0  104.8) (t(142) = 3.61, P # .0001), both
Schneider et al
July 2014
scores lie within the fifth percentile for national IRSD scores
(1996 and 2006). Even though statistically different, this is un-
likely to indicate a tangible shift in overall disadvantage.
Discussion
In individuals born preterm but with no clinical history of
brain lesions or neurosensory disability, corticomotor excit-
ability and a number of other factors, most notably relative
social disadvantage at birth, BW%, and height, are stronger
predictors of cognitive abilities than GA, when assessed in
early adolescence.
As we reported previously, corticomotor excitability was
reduced in preterm adolescents14 and this correlated with
their GIA, as well as a number of more specific cognitive
abilities. Taken together with other imaging evidence,17,31
this suggests that these adolescents may have reduced
cortical thickness and/or reduced white matter maturation,
integrity, and/or connectivity in cortical regions associated
with M1. However, there is also limited evidence that reduc-
tions in M1 gray matter are associated with concurrent in-
creases in rMT.32 Many children for whom we report data
participated in both studies, and readers are referred to
Pitcher et al14 for more detailed interpretation. Whether
or not reduced corticomotor excitability contributes directly
to reduced cognitive abilities in preterm adolescents, or is
simply a marker of overall cortical development is unknown.
Impairments in motor and cognitive function commonly
co-occur in preterm children,33 suggesting that cortical def-
icits because of preterm birth are global rather than confined
to discrete cortical regions. However, white matter and
cortical excitability are plastic, particularly in childhood,
and can be modulated by factors associated with environ-
mental enrichment, such as learning to play a musical in-
strument.34 A postnatal environment that promotes
cortical development (eg, good nutrition, stimulating, low
psychosocial stress) may ameliorate some of the effects of
a low GA on corticomotor excitability. This may also
explain, at least in part, why corticomotor excitability and
IRSDbirth, but not GA, consistently featured in our models
of cognitive abilities. The relationship between corticomotor
excitability and GIA may also be partly explained by genetic
variability. Increased cortical thickness is associated with
increased corticomotor excitability,31 and there is an overlap
in the genes that influence intelligence and cortical thick-
ness, such that greater cortical thickness is associated with
higher IQ.35 Regardless, real-time neurophysiologic TMS
measures of corticomotor excitability would appear to be
a useful addition to assessment of these individuals, at least
in early adolescence.
Increased height at the time of assessment was a consistent
predictor of cognitive scores, in accord with the existing large
literature indicating height correlating with IQ, health, and
economic status in adulthood.36 However, the mechanisms
underlying this association are still poorly understood. Adult
height is determined by a combination of genetics, environ-
ment (including fetal and postnatal nutrition and psychoso-
Cognitive Abilities in Preterm and Term-Born Adolescents
ORIGINAL ARTICLES
cial stress), and gene-environment interactions, although the
variance explained by the latter is unknown.36 In our sample,
BW% correlated with length at birth but not current height
(absolute and z-scores). Neither birth length nor current
height correlated with IRSD at any age. However, we had
no accurate data for postnatal growth rate, nor of the timing
and magnitude of growth spurts at 0-3 years and early adoles-
cence, both of which have been implicated in mediating the
association between height and cognitive ability.36 Eluci-
dating these mechanisms will require prospective, fetal, and
long-term postnatal growth evaluation studies.
It is well-known that socioeconomic status (SES) has ef-
fects on cognitive development and on the risk of preterm
birth, even in developed countries like Australia.30 Although
IRSDbirth did not correlate with GA in our sample, it was a
strong predictor of poorer cognitive abilities, particularly
working memory and GIA. Conversely, IRSDcurrent had no
influence, suggesting that cognitive development is more sus-
ceptible to perturbation by disadvantage in infancy, a period
during which gray matter volume increases exponentially,
rather than in later childhood. This is supported by MRI
studies comparing gray matter development in children
from low, mid, and high SES groups.37 In infancy, gray mat-
ter volumes were similar across all SES groups, but by age
4 years, children with low SES had lower frontal and parietal
gray matter volumes than children with either middle or high
SES.37 These differences were not explained by differences in
weight or head circumference at birth, or infant health. The
frontal lobes are believed to mediate high level cognitive
functions including working memory,38 and the parietal
lobes play a variety of roles, particularly in sensory integra-
tion, spatial perception, and visuospatial aspects of attention
and working memory.39 Others have also reported associa-
tions between low SES, cognitive abilities, and gray matter
volumes in older children,40 but interestingly, neither of these
studies found effects on white matter volumes.37,40 Even
though the underlying mechanisms are not clear, our find-
ings indicate that the associations between low SES at birth
on cognitive abilities are still evident in early adolescence.
Rather than standardized birthweight charts, we used a
customized BW% calculator based on each individuals
optimal fetal growth potential. This varies with maternal
and pregnancy characteristics, is different for each infant,
and is not taken into account in standardized population-
based charts. A low BW% was associated with poorer
thinking ability, auditory processing, and phonemic aware-
ness, consistent with the well-documented effects of subopti-
mal fetal growth on neurodevelopment, even in children
born at term.41 The thinking ability cluster includes tasks
that call upon spatial abilities, which are known to be
impaired in growth restricted children.42 Critically, BW% ap-
pears to exert influences on cognitive outcomes that are
distinct from, and in some domains stronger than, the effects
of GA, highlighting the need to differentiate GA from BW%
in outcome studies. The serendipitous observation that only
children born <27 weeks GA with high BW% (ie, the lowest
was 72.9%) were neurologically-healthy enough for inclusion
175
THE JOURNAL OF PEDIATRICS  www.jpeds.com
in the study, may suggest a protective effect of good fetal
growth in this GA group. However, we did not have BW%
or neurologic details of those infants excluded.
Greater head circumference at birth was associated with
improved working memory. This finding is potentially
interesting; although head circumference is commonly asso-
ciated with degree of intrauterine growth restriction and
brain volume, it is less frequently directly associated with
neurodevelopmental outcome.43-45 However, when utilized
as a measure of postnatal growth, changes in head circum-
ference across early life are highly predictive of cognitive
outcome.44-47 Unfortunately, we did not have data pertain-
ing to early postnatal changes in head circumference to
investigate this finding further. Moreover, the birth mea-
surements (length, head circumference, ponderal index)
were not collected under research conditions, and this
observation requires more rigorous investigation before
valid conclusions can be drawn.
Females had better cognitive efficiency and working mem-
ory scores than males. Previous findings indicate that females
tend to perform better than males on tasks involving cogni-
tive processing speed, and this is further modulated by pu-
bertal stage, which we did not assess.48 Previous findings
regarding sex differences in working memory are inconsis-
tent, with reports of no differences,49 or a slight male advan-
tage.50 However, Kaufman51 noted a male advantage in
working memory only when the task involved spatial stimuli,
but not when the task was verbal, suggesting that the nature
of the task determines if sex differences are evident.
There is an intricate and changing balance between genetic
and environmental influences on cognitive abilities
throughout life, whereby the role of genetics increases across
the lifespan, while the role of environmental factors becomes
less important.35 That people essentially move toward their
genetic potential suggests that, in the absence of brain lesions,
the consequences of a reduced GA may diminish over child-
hood, potentially explaining our finding that GA has only
limited effects on cognitive outcomes when assessed in early
adolescence. Arguing against this is the weight of evidence
indicating that most preterm children do not catch up to
their peers,2,7,52,53 although this evidence may reflect age at
assessment and sample characteristics, particularly related
to perinatal neurological history.
There are some limitations to our findings. We did not
have access to MRI facilities to confirm the lesion-free status
of our participants, or to detect possible gray and white mat-
ter abnormalities. However, no participants had any history
of abnormal cranial ultrasounds, or had been diagnosed
with, or suspected of having, cerebral palsy.14 In addition,
up to 25% of preterm children with normal imaging go on
to exhibit later dysfunction.13 Many of the participants are
likely to have entered puberty. We did not determine puber-
tal stage and did not control for hormonal statuses, which
may have influenced our corticomotor data, as it is known
that sex steroid levels affect cortical excitability.54 Finally,
identifying which specific aspects of social disadvantage
mediate cognitive development will require more direct
176
Vol. 165, No. 1
measures of individual social and psychosocial disadvantage
than used here.
In summary, in individuals born preterm but with no his-
tory of brain lesion or neurosensory disability, GA appears to
have limited direct influence on their cognitive abilities in
early adolescence. Of greatest importance is the degree of
social disadvantage experienced in their home environment
at birth and in early infancy. Taken together with the findings
of others, the strong influence of current height, but not of
birth length and head circumference on cognitive abilities
in adolescence, supports the notion that growth rates (and
presumably nutrition) in early childhood have an important
mediating effect on cognitive development, although pro-
spective studies are required to confirm this in preterm chil-
dren. The association between corticomotor excitability and
cognitive performance suggests that reduced connectivity,
potentially associated with brain microstructural abnormal-
ities, may contribute to cognitive deficits in preterm children.
What is not clear is whether the effects of low GA on cognitive
outcomes attenuate over childhood in favor of a concomitant
increase in the relative importance of heritability, or whether
cognitive development trajectories are more heavily influ-
enced by the quality of the postnatal environment. n
Submitted for publication Sep 17, 2013; last revision received Jan 30, 2014;
accepted Mar 13, 2014.
Reprint requests: Julia B. Pitcher, PhD, Research Center for Early Origins of
Health and Disease, DX 640-517 Robinson Institute, School of Pediatrics and
Reproductive Health, University of Adelaide, Adelaide, SA 5005, Australia.
E-mail: julia.pitcher@adelaide.edu.au
References
1. Beck S, Wojdyla D, Say L, Betran AP, Merialdi M, Requejo JH, et al.
The worldwide incidence of preterm birth: a systematic review of
maternal mortality and morbidity. Bull World Health Organ 2010;
88:31-8.
2. Chyi LJ, Lee HC, Hintz SR, Gould JB, Sutcliffe TL. School outcomes of
late preterm infants: special needs and challenges for infants born at 32 to
36 weeks gestation. J Pediatr 2008;153:25-31.
3. Morse SB, Zheng H, Tang Y, Roth J. Early school-age outcomes of late
preterm infants. Pediatrics 2009;123:e622-9.
4. Pitcher JB, Riley AM, Doeltgen SH, Kurylowicz L, Rothwell JC,
McAllister SM, et al. Physiological evidence consistent with reduced neu-
roplasticity in human adolescents born preterm. J Neurosci 2012;32:
16410-6.
5. Woythaler MA, McCormick MC, Smith VC. Late preterm infants have
worse 24-month neurodevelopmental outcomes than term infants. Pedi-
atrics 2011;127:e622-9.
6. Kirkegaard I, Obel C, Hedegaard M, Henriksen TB. Gestational age and
birth weight in relation to school performance of 10-year-old children: a
follow-up study of children born after 32 completed Weeks. Pediatrics
2006;118:1600-6.
7. Lindstrom K, Winbladh B, Haglund B, Hjern A. Preterm infants as
young adults: a Swedish national cohort study. Pediatrics 2007;120:70-7.
8. van Baar AL, Vermaas J, Knots E, de Kleine MJ, Soons P. Functioning at
school age of moderately preterm children born at 32 to 36 weeks gesta-
tional age. Pediatrics 2009;124:251-7.
9. Anderson P, Doyle LW. Neurobehavioral outcomes of school-age chil-
dren born extremely low birth weight or very preterm in the 1990s.
JAMA 2003;289:3264-72.
10. Constable RT, Ment LR, Vohr BR, Kesler SR, Fulbright RK, Lacadie C,
et al. Prematurely born children demonstrate white matter
Schneider et al
July 2014
microstructural differences at 12 years of age, relative to term control
subjects: an investigation of group and gender effects. Pediatrics 2008;
121:306-16.
11. Kesler SR, Reiss AL, Vohr B, Watson C, Schneider KC, Katz KH, et al.
Brain volume reductions within multiple cognitive systems in male pre-
term children at age twelve. J Pediatr 2008;152:513-20.e1.
12. Counsell SJ, Edwards AD, Chew AT, Anjari M, Dyet LE, Srinivasan L,
et al. Specific relations between neurodevelopmental abilities and white
matter microstructure in children born preterm. Brain 2008;131:3201-8.
13. Ludeman NA, Berman JI, Wu YW, Jeremy RJ, Kornak J, Bartha AI, et al.
Diffusion tensor imaging of the pyramidal tracts in infants with motor
dysfunction. Neurology 2008;71:1676-82.
14. Pitcher JB, Schneider LA, Burns NR, Drysdale JL, Higgins RD,
Ridding MC, et al. Reduced corticomotor excitability and motor skills
development in children born preterm. J Physiol 2012;590:5827-44.
15. DAusilio A, Pulverm uller F, Salmas P, Bufalari I, Begliomini C, Fadiga L.
The motor somatotopy of speech perception. Curr Biol 2009;19:381-5.
16. Floel A, Cohen LG. Contribution of noninvasive cortical stimulation to
the study of memory functions. Brain Res Rev 2007;53:250-9.
17. Kloppel S, Baumer T, Kroeger J, Koch MA, Buchel C, Munchau A, et al.
The cortical motor threshold reflects microstructural properties of cere-
bral white matter. Neuroimage 2008;40:1782-91.
18. Corsi DJ, Davey Smith G, Subramanian SV. Commentary: challenges to
establishing the link between birthweight and cognitive development. Int
J Epidemiol 2013;42:172-5.
19. Hackman DA, Farah MJ, Meaney MJ. Socioeconomic status and the
brain: mechanistic insights from human and animal research. Nat Rev
Neurosci 2010;11:651-9.
20. Pitcher J, Schneider L, Higgins R, Burns N, Nettelbeck T, Ridding M,
et al. Motor cortex development and specific cognitive outcomes in chil-
dren born 25-41 weeks gestation; preliminary findings of the PREMO-
CODE study. Acta Paediatr 2009;98:210.
21. Wassermann EM. Risk and safety of repetitive transcranial magnetic stim-
ulation: report and suggested guidelines from the International Workshop
on the Safety of Repetitive Transcranial Magnetic Stimulation, June 5-7,
1996. Electroencephalogr Clin Neurophysiol 1998;108:1-16.
22. Gardosi JO, Francis A. Customized Centile Calculator GROW-Centile.
5.1 ed: Gestation Network, Available at: www.gestation.net. Accessed
April 8 2008.
23. Woodcock R, McGrew K, Mather M. Woodcock-Johnson III Tests of
Cognitive Abilities. Itasca, IL: Riverside Publishing; 2001.
24. Mather N, Woodcock RW. Examiners Manual, Woodcock-Johnson III
Tests of Cognitive Abilities. Itasca, IL: Riverside Publishing; 2001.
25. Carroll J. Human cognitive abilities: a survey of factor analytic studies.
Cambridge: Cambridge University Press; 1993.
26. Rossini PM, Barker AT, Berardelli A, Caramia MD, Caruso G,
Cracco RQ, et al. Non-invasive electrical and magnetic stimulation of
the brain, spinal cord and rootsbasic principles and procedures for
routine clinical application - report of an IFCN committee. Electroence-
phalogr Clin Neurophysiol 1994;91:79-92.
27. Oldfield RC. The assessment and analysis of handedness: the Edinburgh
inventory. Neuropsychologia 1971;9:97-113.
28. Gromping U. Estimators of relative importance in linear regression
based on variance decomposition. Am Stat 2007;61:139-47.
29. Lindeman RH, Merenda PF, Gold RZ. Introduction to bivariate and
multivariate analysis. Glenview, IL: Scott, Foresman; 1980.
30. Donoghue DLD, Morgan G, Beard J. Influences on the degree of preterm
birth in New South Wales. Aust N Z J Public Health 2013;37:562-7.
31. Lepage JF, Clouchoux C, Lassonde M, Evans AC, Deal CL, Theoret H.
Abnormal motor cortex excitability is associated with reduced cortical
thickness in X monosomy. Hum Brain Mapp 2013;34:936-44.
32. Granert O, Peller M, Gaser C, Groppa S, Hallett M, Knutzen A, et al.
Manual activity shapes structure and function in contralateral human
motor hand area. Neuroimage 2011;54:32-41.
Cognitive Abilities in Preterm and Term-Born Adolescents
ORIGINAL ARTICLES
33. Bracewell M, Marlow N. Patterns of motor disability in very preterm
children. Ment Retard Dev Disabil Res Rev 2002;8:241-8.
34. Bengtsson SL, Nagy Z, Skare S, Forsman L, Forssberg H, Ullen F. Exten-
sive piano practicing has regionally specific effects on white matter devel-
opment. Nat Neurosci 2005;8:1148-50.
35. Brant AM, Munakata Y, Boomsma DI, Defries JC, Haworth CM,
Keller MC, et al. The nature and nurture of high IQ: an extended sensi-
tive period for intellectual development. Psychol Sci 2013;24:1487-95.
36. Case A, Paxson C. Stature and status: height, ability, and labor market
outcomes. J Polit Econ 2008;116:499-532.
37. Hanson JL, Hair N, Shen DG, Shi F, Gilmore JH, Wolfe BL, et al. Family
poverty affects the rate of human infant brain growth. PLoS One 2013;8:
e80954.
38. Jurado MB, Rosselli M. The elusive nature of executive functions: a re-
view of our current understanding. Neuropsychol Rev 2007;17:213-33.
39. Malhotra P, Coulthard EJ, Husain M. Role of right posterior parietal cor-
tex in maintaining attention to spatial locations over time. Brain 2009;
132:645-60.
40. Jednor og K, Altarelli I, Monzalvo K, Fluss J, Dubois J, Billard C, et al. The
influence of socioeconomic status on childrens brain structure. PLoS
One 2012;7:e42486.
41. Arcangeli T, Thilaganathan B, Hooper R, Khan KS, Bhide A. Neurode-
velopmental delay in small babies at term: a systematic review. Ultra-
sound Obstet Gynecol 2012;40:267-75.
42. Leitner Y, Heldman D, Harel S, Pick CG. Deficits in spatial orientation of
children with intrauterine growth retardation. Brain Res Bull 2005;67:
13-8.
43. Cheong JLY, Hunt RW, Anderson PJ, Howard K, Thompson DK,
Wang HX, et al. Head growth in preterm infants: correlation with mag-
netic resonance imaging and neurodevelopmental outcome. Pediatrics
2008;121:e1534-40.
44. Hack M, Breslau N, Weissman B, Aram D, Klein N, Borawski E. Effect of
very low birth weight and subnormal head size on cognitive abilities at
school age. N Engl J Med 1991;325:231-7.
45. Neubauer V, Griesmaier E, Pehbock-Walser N, Pupp-Peglow U, Kiechl-
Kohlendorfer U. Poor postnatal head growth in very preterm infants is
associated with impaired neurodevelopment outcome. Acta Paediatr
2013;102:883-8.
46. Belfort MB, Rifas-Shiman SL, Sullivan T, Collins CT, McPhee AJ,
Ryan P, et al. Infant growth before and after term: effects on neurodevel-
opment in preterm infants. Pediatrics 2011;128:e899-906.
47. Kuban KC, Allred EN, OShea TM, Paneth N, Westra S, Miller C, et al.
Developmental correlates of head circumference at birth and two years
in a cohort of extremely low gestational age newborns. J Pediatr 2009;
155:344-9.
48. Brumback TY, Arbel Y, Donchin E, Goldman MS. Efficiency of respond-
ing to unexpected information varies with sex, age, and pubertal devel-
opment in early adolescence. Psychophysiology 2012;49:1330-9.
49. Lambourne K. The relationship between working memory capacity and
physical activity rates in young adults. J Sports Sci Med 2006;5:149-53.
50. Longman RS, Saklofske DH, Fung TS. WAIS-III percentile scores by educa-
tion and sex for US and Canadian populations. Assessment 2007;14:426-32.
51. Kaufman SB. Sex differences in mental rotation and spatial visualization
ability: can they be accounted for by differences in working memory ca-
pacity? Intelligence 2007;35:211-23.
52. Mathiasen R, Hansen BM, Andersen AM, Forman JL, Greisen G. Gesta-
tional age and basic school achievements: a national follow-up study in
Denmark. Pediatrics 2010;126:e1553-61.
53. Marlow N, Wolke D, Bracewell MA, Samara M , the EPICure Study
Group. Neurologic and developmental disability at six years of age after
extremely preterm birth. N Engl J Med 2005;352:9-19.
54. Smith MJ, Keel JC, Greenberg BD, Adams LF, Schmidt PJ, Rubinow DA,
et al. Menstrual cycle effects on cortical excitability. Neurology 1999;53:
2069-72.
177