Sex Differences in Outcome and Associations with Neonatal Brain

Morphology in Extremely Preterm Children

B
eatrice Ski€old, MD, PhD1,2, Georgios Alexandrou, MD, PhD1, Nelly Padilla, MD, PhD1, Mats Blennow, MD, PhD2,3,
Brigitte Vollmer, MD, PhD1,*, and Ulrika 
Ad
en, MD, PhD1,2

Objective To investigate sex differences in neurologic and developmental outcomes in extremely preterm (EPT)
children and explore associations with neonatal brain morphology.
Study design A population-based cohort of infants born at <27 weeks gestation underwent magnetic resonance im-
aging (MRI) at term equivalent age (n = 107). Voxel-based morphometry (n = 27) and tract-based spatial statistics (n = 29)
were performed in infants with normal MRI findings. Neurologic and developmental assessment (using the Bayley Scales
of Infant and Toddler Development–Third Edition [BSITD-III]) was performed at 30 months corrected age (n = 91).
Results EPT boys had lower mean cognitive composite scores (P = .03) and lower mean language composite
scores (P = .04) compared with EPT girls. Rates of cerebral palsy were similar in the 2 sexes. No perinatal factor
explained the variance in outcomes. Visual inspection of T1- and T2-weighted MRI images found that delayed mye-
lination was found more frequently in boys, whereas cerebellar abnormalities were more common in girls. In the
subgroup of children with normal MRI findings (n = 27), boys had poorer cognitive function (P = .015) and language
function (P = .008), despite larger volumes of cerebellar tissue (P = .029). In boys, cerebellar volume was positively
correlated with BSITD-III cognitive and motor scores (P = .04 for both). In girls, white matter volume (P = .02) and
cortical gray matter volume (P = .03) were positively correlated with BSITD-III language score. At the regional level,
significant correlations with outcomes were found only in girls.
Conclusion Cognitive and language outcomes at age 30 months were poorer in boys. Sex-related differences
were observed on neonatal structural MRI, including differences in the patterns of correlations between brain vol-
umes and developmental scores at both global and regional levels. (J Pediatr 2014;164:1012-8).

T
he male disadvantage with regard to perinatal morbidity and mortality is well known.1 In preterm infants, follow-up data
from several cohorts have revealed significant differences in outcomes between the sexes.2,3 The higher incidence of brain
lesions, such as intraventricular hemorrhage and periventricular leukomalacia, in male preterms2,4 explains some, but
not all, of these differences. For example, in extremely preterm (EPT) infants with normal head ultrasound, male sex was shown
to be an independent risk factor for a low mental developmental index at age 18-22 months.5 Magnetic resonance imaging
(MRI) in preterm adolescents has demonstrated sex-based differences in brain microstructure and volume in the absence of
overt perinatal injury.6
We previously reported MRI data in a population-based cohort of EPT infants7 and found significant associations between
neonatal white matter abnormalites identified by visual inspection of structural MRI and adverse outcomes.8 Others have
demonstrated relationships between neonatal brain tissue volumes and neurodevelopmental outcomes.9 Furthermore, a study
using Tract-Based Spatial Statistics (TBSS, a tool for analyzing diffusion tensor imaging [DTI] data) performed in the neonatal
period suggested an association between white matter structural alterations and toddler age outcomes in preterm infants.10 The
aim of the present study was to investigate sex differences in neurologic and developmental outcomes in a cohort of EPT chil-
dren, and explore associations with brain structure as assessed by MRI at term equivalent age.

Methods
From the Departments of 1Women’s and Children’s
Health, 2Neonatology, and 3Clinical Sciences,
Study subjects were participants in the Extremely Preterm Infants in Sweden Study Intervention, and Technology, Karolinska Institutet,
(EXPRESS) project,11 a prospective population-based study of EPT infants in Stockholm, Sweden
*Present address: Clinical and Experimental Sciences,
University of Southampton, Southampton, United
Kingdom.
BSITD-III Bayley Scales of Infant and Toddler Development–Third Edition Supported by Swedish Medical Research Council (2011-
3981), Marianne and Marcus Wallenberg Foundation
CP Cerebral palsy (2011-MWW.0085), the regional agreement on medical
CPAP Continuous positive airway pressure training and clinical research between Stockholm County
Council and Karolinska Institutet (ALF 20120450), Linnea
DTI Diffusion tensor imaging and Josef Carlsson’s Foundation, Swedish Order of
EPT Extremely preterm Freemasons in Stockholm, Swedish Medical Society,
Swedish Brain Foundation, and Sa €llskapet Barnav
ard.
EXPRESS Extremely Preterm Infants in Sweden Study
The authors declare no conflicts of interest.
MRI Magnetic resonance imaging
TBSS Tract-Based Spatial Statistics 0022-3476/$ - see front matter. Copyright ª 2014 Elsevier Inc.
All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2013.12.051

1012

Sweden investigating neurodevelopmental outcomes at 30
months corrected age, with a subcohort in Stockholm undergo-
ing MRI at term equivalent age.8 All infants born between
January 2004 and April 2007 at gestational age <27 weeks +
0 days were eligible for inclusion in the study (Figure 1).
Children with malformations, chromosome aberrations,
malignant disorders, or congenital infections were excluded.
Perinatal data were collected prospectively (Table I). The
regional Ethics Committee in Stockholm approved the study,
and informed consent was obtained from parents of all subjects.
MRI at Term Equivalent Age
MRI was performed with a Philips Intera 1.5-T system
(Philips, Eindhoven, The Netherlands) and consisted of
sagittal T1- weighted and T2-weighted turbo spin echo im-
ages, axial T2*-weighted images, axial inversion recovery im-
ages, and a 3-dimensional T1-weighted gradient echo image.
The diffusion-weighted MRI images were acquired with 15
directions, b value = 700 s/mm2, voxel size = 1.4  1.4 
2.2 mm3, and one b = 0 image in the axial plane, as described
previously.7 The T1-weighted and T2-weighted images were
visually inspected by 4 independent observers according to a
standardized scoring system for structural abnormalities.12
Five separate white matter variables were assessed: abnormal
white matter signal, reduction in white matter volume, cystic
changes, ventricular dilatation, and thinning of the corpus
callosum/myelination (with myelination deemed normal if
reaching the central corona radiata, caudate nucleus, and
centrum semiovale; mildly delayed if reaching the posterior
limb of the internal capsule, lenticular nucleus, and thalamus;
and severely delayed if reaching only the brainstem and cer-
Figure 1. Study population.
Vol. 164, No. 5  May 2014
ebellum). From the summed individual item scores, a
composite score was derived. Based on this composite score,
the study group was divided into 4 subgroups: (1) no white
matter abnormalities; (2) mild white matter abnormalities;
(3) moderate white matter abnormalities; and (4) severe
white matter abnormalities. Similarly, gray matter was as-
sessed for abnormal signal in the cortical or deep gray matter,
enlargement of the subarachnoid spaces, and delayed gyral
maturation. Infants were then divided into those with normal
gray matter and those with abnormal gray matter based on
the calculated composite score. The agreement rate between
observers was 98% for group assignment, and consensus was
reached after discussion when opinions differed.
In addition to the scoring, a standard radiologic assess-
ment of MRI images was performed, assessing for malforma-
tions, the presence of hemosiderin, and cerebellar
abnormalities, such as cysts, hemosiderin, atrophy, and
abnormal signal intensity. Quantitative analyses were per-
formed on MRI data of the infants without moderate to se-
vere white matter abnormalities or gray matter
abnormalities based on the scoring system, or without cere-
bellar abnormalities. Automatic segmentation of brain tis-
sues was performed using SPM8 software (http://www.fil.
ion.ucl.ac.uk/spm) running in MATLAB version 7.5 (Math-
Works, Natrick, Massachusetts) with specific neonatal
priors,13 including white matter, cortical and deep gray mat-
ter, cerebrospinal fluid, the whole cerebellum, and brainstem.
(The latter 2 structures were not segmented into gray matter
and white matter.) A Diffeomorphic Anatomical Registration
through an Exponentiated Lie algebra algorithm was applied
to improve the intersubject registration. Images were modu-
lated and smoothed with a full width at half-maximum of
3-mm Gaussian kernel for white matter, cortical and deep
gray matter, whole cerebellum, and brainstem. Volumes
(mm3) for the segmented/normalized/modulated images
were obtained with the Easy Volume toolbox.14
Diffusion-weighted MRI data were corrected for move-
ment and eddy current artifacts and fitted to a tensor model,
and fractional anisotropy maps and apparent diffusion coef-
ficient maps were calculated. To investigate differences in
apparent diffusion coefficient and fractional anisotropy be-
tween groups across the whole brain, diffusion data were pro-
cessed using FMRIB’s Diffusion Toolbox (version 2.0;
Analysis Group, Oxford, United Kingdom) and analyzed
with TBSS version 1.2 in FSL version 4.1.4 (Analysis
Group).15 The fractional anisotropy (and apparent diffusion
coefficient) maps of all subjects were aligned into a standard
space using nonlinear registration, and a mean fractional
anisotropy skeleton was created representing the centers of
major tracts common to the group of subjects. The fractional
anisotropy skeleton was thresholded to fractional anisotropy
$ 0.20 to include the major white matter tracts but exclude
peripheral tracts where there was signficant intersubject vari-
ability and/or partial volume effects with gray matter. The
aligned fractional anisotropy data of each subject were pro-
jected onto this skeleton perpendicular to the local tract di-
rection, and the resulting data were fed into voxel-wise
1013
THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. 164, No. 5

Table I. Perinatal characteristics

Characteristic EPT boys (n = 60) EPT girls (n = 47) P value
Gestational age at birth, wk + d, mean  SD 25 + 4  1 25 + 5  1 NS
Birth weight, g, mean  SD 819  146 777  169 NS
Gestational age at MRI, wk + d, mean  SD 40 + 5  1 40 + 4  1 NS
Prenatal steroids, n (%) 56 (93) 44 (94) NS
Cesarean delivery, n (%) 29 (48) 18 (38) NS
Premature rupture of membranes, n (%) 18 (30) 11 (23) NS
Maternal signs of infection, n (%) 22 (37) 16 (34) NS
Inotropic support, n (%) 33 (55) 22 (47) NS
Duration of ventilation, d, mean (range) 14 (0-47) 14 (0-55) NS
Duration of CPAP, d, mean (range) 43 (19-115) 36 (21-58) .015
Postnatal steroids, n (%) 13 (22) 6 (13) NS
Bronchopulmonary dysplasia, oxygen at age 28 d, n (%) 56 (93) 45 (96) NS
Bronchopulmonary dysplasia, oxygen at age 36 wk, n (%) 26 (43) 20 (42) NS
Necrotizing enterocholitis. Bell grade 2-3, n (%) 5 (8) 2 (4) NS
Retinopathy of prematurity, laser-treated, n (%) 14 (23) 6 (13) NS
Days until full enteral nutrition, mean (range) 24 (8-79) 26 (7-112) NS
Patent ductus arteriosus, Ibuprofen 45 (75) 27 (57) NS
Patent ductus arteriosus, surgical ligation, n (%) 16 (27) 17 (36) NS
IVH grade 1-2, n (%)* 17 (28) 20 (43) NS
IVH grade 3-4/periventricular hemorrhagic infarction, n (%)* 7 (12) 4 (9) NS
Periventricular leukomalacia, n (%)* 0 (0) 3 (6) .047

IVH, intraventricular hemorrhage; NS, not significant.

*Cranial ultrasound examinations were performed repeatedly during the hospital stay in all infants.

cross-subject statistics. Recently proposed modifications for
neonatal data were applied.16
Neurodevelopmental Assessment at 30 Months
At corrected age 30 months, children were examined neuro-
logically (ie, movements, posture, reflexes, and muscle tone).
Neurologic status was categorized as normal (completely
normal neurologic status), unspecific signs (eg, asymmetric
muscle tone or reflexes, muscular hypotonia, or muscular hy-
pertonia without definite signs of cerebral palsy [CP]), or
abnormal (signs of CP present as defined by the Surveillance
of Cerebral Palsy in Europe Working Group). On the same
day, children completed the Bayley Scales of Infant and
Toddler Development–Third Edition (BSITD-III) for assess-
ment of cognitive, language, and motor development.
Statistical Analyses
Statistical analysis was performed with PASW Statistics 21.0
(SPSS Inc, Chicago, Illinois). Neurologic status was
compared between boys and girls (n = 91; 48 boys and 43
girls) using the Pearson c2 test, and performance on the
BSITD-III was compared using the Student t test (n = 86;
45 boys and 41 girls). To investigate differences in outcomes
between boys and girls in relation to MRI findings, 2 sets of
analyses were performed, the first set on all infants with
MRI data available, and the second set limited to those
without moderate to severe white matter abnormalities or
gray matter abnormalities based on the scoring system, or
cerebellar abnormalities (exclusion group; n = 14; 8 boys
and 6 girls).
Comparisons of perinatal characteristics (Table I) and
MRI findings in boys and girls were performed using the
Student t test for continuous variables and the Pearson c2
test for dichotomous data. Based on these analyses, linear
regression models including sex, number of days on
1014
continuous positive airway pressure (CPAP), cerebellar
abnormalities, and delayed myelination detected on MRI
were used to investigate the contribution of these variables
to the variance in BSITD-III outcomes. Pearson
correlations were calculated separately for boys and girls to
investigate the possible relationship between global brain
volume and BSITD-III score.
Differences in global volumetric measures between boys
(n = 12) and girls (n = 15) were assessed using a general-
ized linear model (multivariate), with brain volume as the
dependent variable and sex as the independent factor. To
control for generalized scaling effects, total intracranial
volume (all brain tissues, including cerebrospinal fluid)
served as the covariate. For voxel-based morphometry Dif-
feomorphic Anatomical Registration through an Exponen-
tiated Lie algebra analyses (n = 27; 12 boys and 15 girls),
group comparisons were performed using the t test to eval-
uate sex differences, and multiple regression analyses were
performed in boys and girls separately to test for possible
relationships between brain volume at the voxel level and
developmental score. Familywise error correction
(P < .05) was applied.
In the DTI analyses, t tests were performed within TBSS to
identify group differences in apparent diffusion coefficient
and fractional anisotropy between boys and girls (n = 29;
19 boys and 10 girls), and Threshold-Free Cluster Enhance-
ment in the “randomize” function was applied. A P value
of <.05 was chosen as the cutoff level for significance.
Results
Sex and Outcomes at 30 Months
At corrected age 30 months, 6 of the 91 children (3 boys
and 3 girls) met the criteria for CP. Unspecific neurologic
signs were distributed similarly between boys and girls,
€ ld et al
Skio
May 2014 ORIGINAL ARTICLES

Table II. BSITD-III results in EPT boys and girls and correlations with brain tissue volumes
Correlations between developmental
scores and brain tissue volumes
Boys with normal Girls with normal Boys with normal Girls with normal
BSITD-III score, Boys Girls structural structural structural structural
mean ± SD (n = 45) (n = 41) P value MRI (n = 36) MRI (n = 34) P value MRI (n = 12) MRI (n = 15)
Cognitive
Composite score 93.6  7.0 97.9  11.3 .03 93.8  7.1 99.7  9.4 .004 Cerebellum: NS
r = 0.59; P = .04
Cognitive scaled score 8.7  1.4 9.6  2.3 .04 8.8  1.5 9.9  1.9 .005 NS NS
Language
Composite score 94.4  13.1 100.7  15.2 .04 95.1  13.5 103.2  14.1 .018 NS Cortical gray matter:
r = 0.54; P = .03
Receptive language 9.2  2.2 10.4  2.6 .03 9.3  2.2 10.7  2.6 .014 NS NS
scaled score
Expressive language 8.9  2.5 9.9  2.8 NS 8.9  2.7 10.3  2.7 .034 NS White matter:
scaled score r = 0.58; P = .02
Deep gray matter:
r = 0.60; P = .01
Brainstem:
r = 0.51; P = .03
Motor
Composite score 101.3  15.5 104.2  14.5 NS 102.0  16.6 104.9  13.7 NS Cortical gray NS
matter: r = 0.66;
P = .02
Cerebellum: r = 0.62;
P = .04
Fine motor scaled 9.4  2.6 10.3  2.9 NS 9.5  2.7 10.8  2.6 NS NS Brainstem:
score r = 0.55; P = .03
Gross motor scaled 11.3  4.2 11.0  3.3 NS 11.1  4.6 11.0  3.3 NS NS NS
score

present in 12 boys and 14 girls. Boys had lower mean
composite scores for cognitive function (94  7 vs
98  11; P = .03) and language function (94  13 vs
101  15; P = .04) compared with girls (Table II),
whereas motor scores were not significantly different
between the sexes.
Sex and Findings on Visual Analysis of MRI
According to the composite white matter score, 14% of the
cohort had moderate to severe white matter abnormalities
on MRI at term equivalent age, 86% had no or mild white
matter abnormalities,7 and 8% had gray matter abnormal-
ities. Boys and girls were similar in terms of MRI group
assignment (white matter group: no/mild/moderate/severe
abnormalities; gray matter group: normal/abnormal).
Analyzing separate scoring items, delayed myelination was
more common in boys than in girls (P = .019). Other white
matter and gray matter scoring items were similarly distrib-
uted between the sexes. Six girls had cerebellar abnormalities
(eg, atrophy, hemosiderin, cysts), compared with only 1 boy.
Sex, Perinatal Characteristics, and Neonatal
Morbidities
Boys spent longer time on CPAP compared with girls
(P = .015), but had a similar rate of bronchopulmonary
dysplasia. Three infants, all girls, had cystic periventricular
leukomalacia detected on neonatal ultrasound. Other
neonatal variables and perinatal characteristics were similar
in boys and girls (Table I).
Sex Differences in Outcome and Associations with Neonatal Brain Morphology in Extremely Preterm Children
Outcomes in EPT Infants with Normal Structural
MRI
After excluding infants with moderate to severe white matter
abnormalities, gray matter abnormalities, or cerebellar ab-
normalities on T1- and T2-weighted MRI (exclusion group;
n = 14), 2 infants with CP (1 boy and 1 girl) remained in the
sample, and no difference between boys and girls in the fre-
quency of unspecific neurologic signs was seen. Differences
in the mean cognitive composite scores and the mean lan-
guage composite scores between boys and girls remained
(Table II). However, delayed myelination was still more
frequent in boys, and cerebellar abnormalities remained
more frequent in girls. When repeating the analyses
excluding infants with delayed myelination and/or
cerebellar abnormalities, we found that the differences in
mean cognitive composite score (P = .015) and mean
language composite score (P = .008) between boys and girls
persisted. Regression analyses confirmed that sex
significantly influenced the variance in cognitive and
language outcomes on the BSITD-III scales, whereas delayed
myelination, cerebellar abnormalities, and number of days
on CPAP did not. The BSITD-III motor composite score
was influenced by the number of days on CPAP but not by
sex, delayed myelination, or cerebellar abnormalities.
Sex, Diffusion Measures, and Outcomes
In the subsample of infants included in the TBSS analysis (n =
29), no differences were seen in diffusion measures, neuro-
logic status, or BSITD-III scores between boys and girls.
1015
THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. 164, No. 5

Figure 2. A, Gray matter involving the dorsal anterior cingulum correlated positively to BSITD-III language composite scores. B,
Gray matter involving the supplementary motor area correlated positively to BSITD-III fine motor scores. C, White matter
involving the medial parietal areas (precuneus) correlated positively to BSITD-III expressive language score. In boys, no signif-
icant correlations were found at regional level. Plot points indicate real data. The line indicates data adjusted to the theoretical
model. GA, gestational age.

Sex, Brain Volumes, and Outcomes
The subsample for which morphometric data were available
(n = 27) was representative of the larger sample in terms of all
perinatal characteristics and neonatal varibles, as well as find-
ings on visual inspection of conventional structural MRI.
Consistent with our findings in the larger study sample,
cognitive (P = .007) and language (P = .026) function was
poorer in boys than girls in this subsample, whereas motor
function on the BSITD-III and neurologic status did not
differ between the sexes. Boys had overall larger brain vol-
umes than girls, although statistically significantly larger
1016
only for white matter (P = .039) and the cerebellum
(P = .029). After adjustment for total intracranial volume,
only the difference in cerebellar volume achieved statistical
significance. No significant sex differences were found at
the voxel level.
Correlations between brain volumes and developmental
scores differed between boys and girls on both global
(Table II) and regional levels (Figure 2). In girls,
cortical gray matter volume in precentral regions
(supplementary motor cortex and cingulate gyri
bilaterally) was positively correlated with BSITD-III
€ ld et al
Skio
May 2014
language composite scores (cluster size, 423 voxels; P <
.001) (Figure 2, A). Supplementary motor cortex vol-
ume also was positively correlated with BSITD-III fine
motor score (cluster size, 434 voxels; P < .001)
(Figure 2, B). White matter volume in medial parietal
areas (precuneus) positively correlated with BSITD-III
expressive language score (cluster size, 136 voxels; P <.01)
(Figure 2, C). In boys, no significant correlations were
found at the regional level.
Discussion
The purpose of this study was to investigate sex differences in
outcomes at 30 months corrected age in EPT infants and to
explore associations with brain structure revealed by
neonatal MRI. We found that boys had lower cognitive and
language scores compared with girls, whereas motor function
outcome was similar in the 2 sexes. Sex-related differences
were observed on MRI, including differences in the patterns
of correlation between brain volume and developmental
score at both global and regional levels.
The present study is a subcohort of the EXPRESS study, in
which boys were found to be at greater risk for moderate to
severe disabilities at age 30 months compared with girls.11
Male sex is a recognized risk factor for poor outcome after
preterm birth. In the EPICure study, survival to discharge
was lower and frequency of neonatal problems was higher
in EPT boys than in EPT girls. At age 30 months, boys
were more likely to have CP and impaired cognitive func-
tion.3 Similarly, in the same cohort at age 6 years, cognitive,
language, and educational difficulties were more pronounced
in boys,17 along with persistent greater motor difficulties.
However, apart from early findings on neonatal cranial ultra-
sound, possible explanations related to brain lesions were not
investigated. In the present study, we used advanced MRI
techniques to explore differences in brain morphology that
might possibly underpin these differences in outcomes.
In the large study of Hintz et al,2 boys had lower mental
developmental indices and a higher prevalence of CP
compared with girls at age 18-22 months. When adjusting
for perinatal risk factors and the higher incidence of neonatal
morbidities, preterm boys still had poorer outcomes. Similar
findings were reported by Peacock et al.18 The authors of
both studies concluded that preterm boys appear to have
an inherently greater risk for adverse outcomes. Correspond-
ingly, in the present study, sex differences in cognitive and
language function could not be entirely explained by peri-
natal risk factors or brain morphology on conventional
MRI. We found that boys had poorer cognitive and language
performance than girls, and that delayed myelination was
more frequent in boys. It is well known that myelination is
critical for normal neurodevelopment, that this process
may be adversely affected by preterm birth, and that
abnormal myelination often is associated with develop-
mental delay.19 Thus, we were surprised by our finding that
delayed myelination on visual inspection of MRI images
did not contribute to the differences in outcomes between
Sex Differences in Outcome and Associations with Neonatal Brain Morphology in Extremely Preterm Children
ORIGINAL ARTICLES
boys and girls in either univariate analysis or regression anal-
ysis. Whether these findings are related to compensatory, or
“catch-up,” mechanisms in boys between the time of imaging
and the time of neurodevelopmental assessment, or whether
EPT boys are truly at higher constitutional risk for adverse
outcomes, as suggested by Peacock et al18 and Hintz et al,2 re-
mains to be determined.
An intriguing question is at what stage the male vulnera-
bility arises. The higher number of spontaneous abortions
in male embryos20 may reflect in utero disadvantages of
male sex. Animal work has shown evidence of innate sex dif-
ferences due to different cell death pathways,21 independent
of hormonal influences.22 Most studies in humans are natu-
rally observational, leaving the underlying mechanisms
largely unexplored. DTI offers an opportunity to study brain
tissue microstructure in a noninvasive manner, and is associ-
ated with outcomes when performed at the same age as behav-
ioral testing.23 In prospective studies, DTI findings at term
equivalent age have been linked to later outcomes10,24-27 but
few have examined sex differences in that respect.28 In the pre-
sent study, outcomes were not different between boys and
girls in the subsample suitable for TBSS, and thus this could
not be investigated further. Similar challenges were reported
by van Kooij et el,10 where 56% of their sample had DTI im-
ages suitable for TBSS analysis. In that study, follow-up data
were available for 54% of the cohort, and the data showed a
trend toward better cognitive performance at corrected age
2 years in girls, but TBSS at term equivalent age revealed no
difference in fractional anisotropy between the sexes.
Our data support previously reported findings in preterm
and term newborns29,30 and in healthy older children and
adults31 demonstrating larger brain volumes in males than fe-
males. We found that lower global brain tissue volumes were
associated with lower developmental scores, similar to previ-
ous findings in preterm infants at term equivalent age9 and
older preterm subjects.32 An examination of how these global
brain volumes correlated with the 3 BSITD-III scales showed
that boys and girls displayed different patterns, and, interest-
ingly, on the regional level, positive correlations were seen
only in girls. There also may be some influence of sex-
based brain architectural differences, such as folding patterns
of the cortex, with preterm girls apparently having a larger
interface between cortical gray matter and white matter,29
which may affect later outcomes. Of note are the correlations
found in several regions that may subserve different language
functions, such as, for example, cortical gray matter in pre-
central regions. The role of the supplementary motor cortex
in language processing has been reported previously,33 and
the cingulate cortex is of particular interest, given its involve-
ment not only in language processing per se, but also in such
tasks as memory and attention,34 both of which are impor-
tant for early language acquisition.
The strengths of the present study include its prospective
population-based design, multimodal imaging approach,
and rigorous analyses of detailed clinical information. How-
ever, the small size of the subsamples suitable for volumetric
and TBSS analyses is a limitation that might have influenced
1017
THE JOURNAL OF PEDIATRICS  www.jpeds.com
our results. Moreover, outcomes were not different between
boys and girls in the TBSS subsample, and thus this could
not be investigated further. Further studies are needed to bet-
ter understand sexual dimorphism during brain development
and potential links to differences in early childhood outcomes
after preterm birth. n
The authors would like to thank all the participating children and their
parents. We are grateful to the EXPRESS Group, the psychologists per-
forming the Bayley assessments, the members of the Neonatal Brain
Imaging group, and our commited research nurses. Furthermore, we
acknowledge the MR physics group and the MR radiology staff at Astrid
Lindgren Children’s Hospital for valuable technical assistance.
Submitted for publication Jun 27, 2013; last revision received Nov 18, 2013;
accepted Dec 27, 2013.

atrice Skio
Reprint requests: Be € ld, MD, PhD, Neonatal Research Unit, Q2:07,
Astrid Lindgren Children’s Hospital, S-171 76 Stockholm, Sweden. E-mail:
beatrice.skiold@karolinska.se
References
1. Drevenstedt GL, Crimmins EM, Vasunilashorn S, Finch CE. The rise and
fall of excess male infant mortality. Proc Natl Acad Sci U S A 2008;105:
5016-21.
2. Hintz SR, Kendrick DE, Vohr BR, Poole WK, Higgins RD. Gender dif-
ferences in neurodevelopmental outcomes among extremely preterm,
extremely-low-birthweight infants. Acta Paediatr 2006;95:1239-48.
3. Wood NS, Costeloe K, Gibson AT, Hennessy EM, Marlow N,
Wilkinson AR. The EPICure study: associations and antecedents of
neurological and developmental disability at 30 months of age following
extremely preterm birth. Arch Dis Child Fetal Neonatal Ed 2005;90:
F134-40.
4. Leijser LM, Steggerda SJ, de Bruine FT, van der Grond J, Walther FJ, van
Wezel-Meijler G. Brain imaging findings in very preterm infants
throughout the neonatal period, part II: relation with perinatal clinical
data. Early Hum Dev 2009;85:111-5.
5. Laptook AR, O’Shea TM, Shankaran S, Bhaskar B. Adverse neurodeve-
lopmental outcomes among extremely low birth weight infants with a
normal head ultrasound: prevalence and antecedents. Pediatrics 2005;
115:673-80.
6. Constable RT, Ment LR, Vohr BR, Kesler SR, Fulbright RK,
Lacadie C, et al. Prematurely born children demonstrate white mat-
ter microstructural differences at 12 years of age, relative to term
control subjects: an investigation of group and gender effects. Pedi-
atrics 2008;121:306-16.
7. Ski€old B, Horsch S, Hallberg B, Engstrom M, Nagy Z, Mosskin M, et al.
White matter changes in extremely preterm infants, a population-based
diffusion tensor imaging study. Acta Paediatr 2010;99:842-9.
8. Ski€old B, Vollmer B, Bohm B, Hallberg B, Horsch S, Mosskin M, et al.
Neonatal magnetic resonance imaging and outcome at age 30 months
in extremely preterm infants. J Pediatr 2012;160:559-66.e1.
9. Jary S, De Carli A, Ramenghi LA, Whitelaw A. Impaired brain growth
and neurodevelopment in preterm infants with posthaemorrhagic ven-
tricular dilatation. Acta Paediatr 2012;101:743-8.
10. van Kooij BJ, de Vries LS, Ball G, van Haastert IC, Benders MJ,
Groenendaal F, et al. Neonatal tract-based spatial statistics findings and
outcome in preterm infants. AJNR Am J Neuroradiol 2012;33:188-94.
11. Serenius F, Kallen K, Blennow M, Ewald U, Fellman V, Holmstrom G,
et al. Neurodevelopmental outcome in extremely preterm infants at
2.5 years after active perinatal care in Sweden. JAMA 2013;309:1810-20.
12. Woodward LJ, Anderson PJ, Austin NC, Howard K, Inder TE. Neonatal
MRI to predict neurodevelopmental outcomes in preterm infants. N
Engl J Med 2006;355:685-94.
1018
Vol. 164, No. 5
13. Kuklisova-Murgasova M, Aljabar P, Srinivasan L, Counsell SJ, Doria V,
Serag A, et al. A dynamic 4D probabilistic atlas of the developing brain.
Neuroimage 2011;54:2750-63.
14. Pernet C, Andersson J, Paulesu E, Demonet JF. When all hypotheses
are right: a multifocal account of dyslexia. Hum Brain Mapp 2009;
30:2278-92.
15. Smith SM, Jemkinson M, Johansen-Berg H, Rueckert D, Nichols TE,
Mackay CE, et al. Tract-based spatial statistics: voxelwise analysis of
multi-subject diffusion data. Neuroimage 2006;31:1487-505.
16. Ball G, Counsell SJ, Anjari M, Merchant N, Arichi T, Doria V, et al. An
optimised tract-based spatial statistics protocol for neonates: applications
to prematurity and chronic lung disease. Neuroimage 2010;53:94-102.
17. Wolke D, Samara M, Bracewell M, Marlow N. Specific language diffi-
culties and school achievement in children born at 25 weeks of gestation
or less. J Pediatr 2008;152:256-62.
18. Peacock JL, Marston L, Marlow N, Calvert SA, Greenough A. Neonatal
and infant outcome in boys and girls born very prematurely. Pediatr Res
2012;71:305-10.
19. Volpe JJ. Neurology of the newborn. 5th rev ed. Philadelphia: Saunders;
2008.
20. McMillen MM. Differential mortality by sex in fetal and neonatal deaths.
Science 1979;204:89-91.
21. Du L, Bayir H, Lai Y, Zhang X, Kochanek PM, Watkins SC, et al. Innate
gender-based proclivity in response to cytotoxicity and programmed cell
death pathway. J Biol Chem 2004;279:38563-70.
22. Dewing P, Shi T, Horvath S, Vilain E. Sexually dimorphic gene expres-
sion in mouse brain precedes gonadal differentiation. Brain Res Mol
Brain Res 2003;118:82-90.
23. Yung A, Poon G, Qiu DQ, Chu J, Lam B, Leung C, et al. White matter
volume and anisotropy in preterm children: a pilot study of neurocog-
nitive correlates. Pediatr Res 2007;61:732-6.
24. Krishnan ML, Dyet LE, Boardman JP, Kapellou O, Allsop JM, Cowan F,
et al. Relationship between white matter apparent diffusion coefficients
in preterm infants at term-equivalent age and developmental outcome at
2 years. Pediatrics 2007;120:E604-9.
25. Arzoumanian Y, Mirmiran M, Barnes PD, Woolley K, Ariagno RL,
Moseley ME, et al. Diffusion tensor brain imaging findings at term-
equivalent age may predict neurologic abnormalities in low birth weight
preterm infants. AJNR Am J Neuroradiol 2003;24:1646-53.
26. Rose J, Mirmiran M, Butler EE, Lin CY, Barnes PD, Kermoian R, et al.
Neonatal microstructural development of the internal capsule on diffu-
sion tensor imaging correlates with severity of gait and motor deficits.
Dev Med Child Neurol 2007;49:745-50.
27. Drobyshevsky A, Bregman J, Storey P, Meyer J, Prasad PV, Derrick M,
et al. Serial diffusion tensor imaging detects white matter changes that
correlate with motor outcome in premature infants. Dev Neurosci
2007;29:289-301.
28. Rose J, Butler EE, Lamont LE, Barnes PD, Atlas SW, Stevenson DK.
Neonatal brain structure on MRI and diffusion tensor imaging, sex,
and neurodevelopment in very-low-birthweight preterm children. Dev
Med Child Neurol 2009;51:526-35.
29. Vasileiadis GT, Thompson RT, Han VK, Gelman N. Females follow a
more “compact” early human brain development model than males: a
case-control study of preterm neonates. Pediatr Res 2009;66:551-5.
30. Knickmeyer RC, Wang J, Zhu H, Geng X, Woolson S, Hamer RM, et al.
Impact of sex and gonadal steroids on neonatal brain structure. Cereb
Cortex 2013; http://dx.doi.org/10.1093/cercor/bht125.
31. Groeschel S, Vollmer B, King MD, Connelly A. Developmental changes
in cerebral grey and white matter volume from infancy to adulthood. Int
J Dev Neurosci 2010;28:481-9.
32. Nosarti C, Giouroukou E, Healy E, Rifkin L, Walshe M, Reichenberg A,
et al. Gray and white matter distribution in very preterm adolescents me-
diates neurodevelopmental outcome. Brain 2008;131:205-17.
33. Friederici AD. The brain basis of language processing: from structure to
function. Physiol Rev 2011;91:1357-92.
34. Torta DM, Cauda F. Different functions in the cingulate cortex: a meta-
analytic connectivity modeling study. Neuroimage 2011;56:2157-72.
€ ld et al
Skio