(DUO\+XPDQ'HYHORSPHQW

Contents lists available at ScienceDirect

Early Human Development

journal homepage: www.elsevier.com/locate/earlhumdev

Long-term attention decits combined with subcortical and cortical 0$5.

structural central nervous system alterations in young adults born small for
gestational age
Sabrina Surena, Diego Angulob,c,d, Yang Dinga, Pablo Reyesc,e, Jorge Marinf,
Jose T. Hernandezd, Nathalie Charpakc, Gregory A. Lodygenskya,g,
a
Research Center, Sainte-Justine University Hospital, 3175, Chemin de la Cote-Sainte-Catherine, Montreal, Quebec H3T1C5, Canada
b
ImExHS, Imex-Research, Calle 93 # 16 46, Of 502, Edicio Zen Oce, Bogot, Colombia
c
Kangaroo Foundation, Bogot, Colombia
d
IMAGINE, Systems and Computing Engineering, Universidad de los Andes, Bogot, Colombia
e
Hospital Universitario San Ignacio, Bogot, Colombia
f
Fundacin Hospital Infantil Universitario de San Jos, Bogot, Colombia
g
Department of Pediatrics and Pharmacology, University of Montreal, C.P.6128, succursale Centre-ville, Montreal, Quebec H3C3J7, Canada

A R T I C L E I N F O A B S T R A C T

Keywords: Background: Being born small for gestational age has been associated with neurodevelopmental disabilities and
Small for gestational age smaller gray matter volumes in childhood. However, it is not known if these changes persist in adults and
Adults whether SGA has any impact on attention memory and IQ.
Cerebral MRI Aims: The goal of this study was to evaluate the association between birth weight and gray matter anatomy in
Voxel-based morphometry
adults born small for gestational age at term, in relation to IQ, attention and memory.
FreeSurfer
Materials and methods: This prospective follow-up study at age 20 included 39 adults born small for gestational
Attention
age at term and 37 adults born appropriate for gestational age at term. Detailed neurocognitive skills were
assessed (IQ, attention and memory). Anatomical images were analyzed using Voxel-Based-Morphometry and
FreeSurfer.
Results: Adults born small for gestational age at term had lower performances in subtests assessing attention and
executive functions. They also showed smaller total intracranial volume; smaller volumes and surface areas in
the frontal lobe, inferior/middle parietal and temporal gyrus; smaller cerebellum, thalamus and basal ganglia
volumes. Interestingly, all these structures correlated with attention subtests.
Conclusion: These results highlight the persistent eects of being born small for gestational age on attention and
associated brain structures.

1. Introduction priate stimulation and enhances bonding and breastfeeding. Kangaroo

Being born small for gestational age (SGA), either being born
preterm or at term, was linked to decreased IQ, academic achievement
and learning abilities; visomotors skills; as well as diculties in
attention, memory, executive function and emotional regulation in
children and adults [18]. Moreover, decreased total brain volume has
been observed in SGA children and adults [9].
The Kangaroo Mother Care program is a human-based care inter-
vention devised to complement neonatal care for low birth weight and
premature infants. Kangaroo position (skin-to-skin contact on the
mother's chest) oers thermal regulation, physiological stability, appro-
nutrition is based on breastfeeding and kangaroo discharge policy relies
on family empowerment and early discharge in kangaroo position with
close ambulatory follow-up [10].
One study in 4 to 7 years-old SGA children born at term reported
smaller cerebellum, globus pallidus, putamen and hippocampus vo-
lumes [11]. One study in 15 years-old SGA adolescents born at term
showed smaller cerebellum and thalamus volumes, but no dierence in
the amygdala and hippocampus volumes [12]. In 20 years-old SGA
adults born at term, Ostgard et al. (2014) reported smaller frontal lobe,
cingulate, temporal and parietal lobes surface area, as well as smaller
putamen volume. However, they found no dierence regarding cortical

Corresponding author at: Pediatrics and Pharmacology, University of Montreal, Research Centre, Ste-Justine University Hospital, 3175 Cote Ste-Catherine, Montreal, Quebec H3T
1C5, Canada.
E-mail address: ga.lodygensky@umontreal.ca (G.A. Lodygensky).

http://dx.doi.org/10.1016/j.earlhumdev.2017.04.016
Received 11 January 2017; Received in revised form 26 April 2017; Accepted 27 April 2017
(OVHYLHU%9$OOULJKWVUHVHUYHG
S. Suren et al.
thickness and they did not assess cortical volumes [13].
The objective of this study was to investigate subcortical volumes,
as well as cortical volume, thickness and surface area in 20 years-old
SGA adults born at term. Participants were from the Colombian
Kangaroo Mother Care cohort [10]. In addition to cortical thickness
and surface area measurements [13], cortical volumes were assessed
using two complementary methods for anatomical structural measure-
ments (voxel-based and surface-based morphometric analyses). More-
over, to test the relationship between structure and function, we
measured IQ, attention/executive functions and memory of the parti-
cipants.
2. Methods
2.1. Participants
This study is part of a larger prospective research program on low
birth weight, prematurity and the Kangaroo Mother Care program in
Bogota, Colombia [14]. This program involved fetal assessment and
long-term postnatal follow-up of several hundred newborns at the
Clinica San Pedro Claver. In this cohort, newborns were born between
1993 and 1994 with a weight of < 2001 g. Among these newborns, 39
term infants participated in the follow-up study until adulthood. The
present study included 76 adults, with 39 adults born at term, small for
gestational age (SGA, mean birth weight: 1817 g 400 g) and 37
adults born at term, appropriate for gestational age (AGA, mean birth
weight: 3032 g 800 g) and equivalent in terms of socio-economic
status. The SGA group was dened by a birth weight below the 10th
percentile, adjusted for gestational age [15]. Control subjects were
dened as adults with birth weight between the 10th and 90th
percentile. We dened SGA adults whose height at 12 months of age
was equal to or above the 3rd percentile as having experienced catch-up
growth [16,17]. In the SGA adults, 15 experienced catch-up growth, 22
not. The study protocol was approved by an ethical committee, and
written informed consent was obtained for all participants.
2.2. Measures
Maternal age at birth; gender and parental education and incomes
were recorded. Developmental physical data included weight, height
and head perimeter at birth, 41 weeks of GA, 3, 6, 9, 12 months of
corrected age and 20 years.
Cognitive and neuropsychological tests were administered to all
participants at the time of testing. Intelligent Quotient (IQ) was
measured with the Weschler Abbreviated Scale of Intelligence (WASI-II)
[18]. The California Verbal Learning Test-2nd edition (CVLT-2) was used
to assess memory [19]. Nine subtests of the Test-battery of Attentional
Performance (TAP 2.2.) [20] were used to evaluate attention/executive
function of the youth: alertness (focus attention), divided attention,
sustained attention, working memory and go/nogo (selective attention/
inhibition) subtests.
2.3. Image acquisition and processing
Structural brain images were acquired on a 3 T MRI scanner, with
whole-brain, high resolution T1-weighted MPRAGE images acquired in
the sagittal plane (TR = 8,52 ms, TE = 4.13 ms, matrix si-
ze = 250 256 160 mm, voxel size = 0.97 0.97 1 mm,
FOV = 74 mm). The two-dimensional DICOM les of each brain were
organized into volumetric three-dimensional les using the MRIcron
software package (http://www.mccauslandcenter.sc.edu/mricro/
mricron/).
2.3.1. Voxel-based morphometry (VBM)
The structural T1 images were rstly processed using the
Computational Anatomy Toolbox (CAT12; developed by Christian

(DUO\+XPDQ'HYHORSPHQW
Gaser, University of Jena) within the SPM12 software package
(Wellcome Department of Cognitive Neurology, London, UK) and
MATLAB R2016a, 9.0.0 (MathWorks, Natick, MA, USA) [21,22]. First,
all T1-weighted anatomical images were manually reoriented to place
the anterior commissure (AC) at the origin of the three-dimensional
Montreal Neurological Institute (MNI) space. The images were then
segmented into gray matter, white matter, and cerebrospinal uid [23].
Segmentations were then inspected for their quality and homogeneity
was checked with the CAT12 toolbox. One participant was excluded
because of the anatomical image inhomogeneity and motion. Images
were normalized to Montreal Neurological Institute (MNI) space using a
dieomorphic non-linear registration algorithm (dieomorphic anato-
mical registration through exponentiated lie algebra toolbox-DARTEL)
[24]. Images were modulated by the Jacobian transformed tissue
probability maps (to obtain volume dierences in gray matter) and
smoothed with a Gaussian kernel of 8 mm full width at half maximum
before statistical analyses.
2.3.2. Surface-based morphometry (SBM)
We assessed cortical volumes, thickness and surface area, as well as
sub-cortical volumes using the FreeSurfer software (version 5.3.0) in
the Canadian Brain Imaging Research Platform (CBRAIN) [25]. Free-
Surfer, a fully automated surface-based pipeline, was used to process
the T1 images into a common stereotactic space, in which volumes,
cortical thickness and surface area values could be derived on a
participant-by-participant basis [26] Cortical and sub-cortical segmen-
tation procedure involved the assignment of a neuroanatomical label to
each voxel in a MRI volume using voxel intensity, a probabilistic atlas
estimated from a manually labeled training set, and Bayesian classica-
tion rules [26]. This technique was previously shown to be comparable
in accuracy to manual tracing [27,28]. Cortical thickness was auto-
matically quantied within FreeSurfer on a vertex-by-vertex basis by
computing the average shortest distance between the white matter
boundary and the pial surface (i.e. the cerebral spinal uid boundary)
at each point on the cortex [29]. Segmentation boundaries were
visually inspected by a trained rater and, if necessary, errors due to
segmentation miss classication were reprocessed.
2.4. Statistical analyses
Developmental, demographic, cognitive and neuropsychological
data were analyzed in SPSS v.20 (Armonk, NY) with 2 sample t-tests
comparing the two groups (small for gestational age, SGA vs. appro-
priate for gestational age, AGA), when data were normally distributed.
If not, a non-parametric U Mann-Whitney test was used. With catego-
rical data, a non-parametric Chi2 test was used.
A whole brain VBM analyses was performed with a Gaussian
random eld threshold set at = 0.001 with an extent of at least 130
contiguous voxels (based on the expected voxels per clusters < k >
= 128.81 for this analysis). Clusters were considered signicant at
p < 0.05, FWE-corrected. For FreeSurfer analyses, we extracted ROIs
based on anatomical atlas and evaluated their gray matter volume,
cortical thickness and surface area. Bonferroni's correction for multiple
comparisons was applied with a signicance level of p < 0.0022
(p < 0.05/23 number of comparisons in each hemisphere) to sub-
cortical volume; cortical volume, thickness and surface area analysis of
19 regions. Two sample t-tests between groups (SGA vs. AGA), and sex
as covariate of no interest, were performed in SPM12 and SPSS 20,
respectively.
With Spearman's correlations, we then assessed how signicant
extracted VBM peak voxels and signicant FreeSurfer gray matter
volume, cortical thickness and surface area were related to the
neurocognitive signicant dierences between groups. Correlations
were considered signicant at p < 0.05.
Exploratory, we performed a whole brain VBM analyze to dier-
entiate participants who presented catch-up growth or not.
S. Suren et al. (DUO\+XPDQ'HYHORSPHQW

Table 1 Table 2
Demographic and physical development between small for gestational age (SGA) and Neurocognitive scores between 20 years-old young adults born small for gestational age
appropriate for gestational age (AGA) groups. (SGA) and appropriate for gestational age (AGA) (mean (SD)).

SGA (n = 39) AGA (n = 37) p SGA (n = 39) AGA (n = 37) p

Maternal age at birth (years, 28(0.93) 24(0.90) 0.01 IQ (WASI-II)

mean (SD)) Verbal comprehension (composite 87(2.15) 88(2.72) 0.984
Gender (males/females) 15/24 18/19 0.37 score)
Socio-economic status Perceptual reasoning (composite 84(2.09) 89(2.52) 0.171
Mother's education level 12/27 8/29 0.37 score)
(primary/more than Full scale 4 (composite score) 84(1.83) 87(2.61) 0.484
primary) Memory
Father's education level 12/27 6/31 0.14 CVLT-II immediate correct free recall 46(1.67) 49(1.85) 0.307
(primary/more than Attention/executive functions (T-
primary) scores)
Monthly income per capita 100,428(74887) 117,107(82784) 0.35 TAP alertnessa 49(1.49) 53(1.52) 0.032
(mean (SD)) TAP divided attention: omissionsa 45(1.18) 48(1.02) 0.139
Weight (g, mean (SD)) TAP sustained attention: false 39(1.87) 46(2.66) 0.024
Birth 1817(31) 3032(57) < 0.001 alarmsa
41 weeks of GA 2308(74) 3344(64) < 0.001 TAP sustained attention: omissionsa 35(1.79) 41(2.15) 0.049
3 months 4869(93) 5929(120) < 0.001 TAP working memory: omissionsa 48(1.69) 53(2.06) 0.051
6 months 6213(115) 7437(160) < 0.001 TAP working memory: RT 56(1.63) 56(2.31) 0.533
9 months 7058(130) 8292(184) < 0.001 TAP Go/noGo: false alarmsa 47(1.37) 52(0.93) 0.005
12 months 7987(159) 9234(210) < 0.001 TAP Go/noGo: RT (median) 49(1.32) 44(1.64) 0.005
20 years 54,928(1645) 58,216(1421) 0.085
Height (cm, mean (SD)) Note. IQ, Intellectual Quotient; WASI, Wechsler Abbreviated Scale of Intelligence; CVLT,
Birth 43.49(0.37) 48.83(0.34) < 0.001 Californian Verbal Learning Test; TAP, Test-battery of Attentional Performance; RT,
41 weeks of GA 44.98(0.41) 49.89(0.41) < 0.001 Reaction Time.
3 months 55.33(0.43) 58.73(0.50) < 0.001 a
Highest TAP scores indicate better outcome.
6 months 61.45(0.39) 64.21(0.48) < 0.001
T-test between the two groups, p < 0.05.
9 months 65.82(0.40) 68.78(0.41) < 0.001
12 months 69.66(0.45) 73.14(0.38) < 0.001
20 years 158.90(1.24) 165.14(1.34) 0.001 Among the participants in the SGA group, there were no dierence
Head perimeter (cm, mean (SD)) in outcome between those who had a catch-up growth and those who
did not (data not shown).
Birth 31.26(0.28) 39.15(1.32) < 0.001
41 weeks of GA 33.03(0.22) 35.34(0.2) < 0.001
3 months 38.92(0.19) 39.83(0.30) 0.002 3.2. Imaging data
6 months 41.56(0.21) 42.97(0.26) < 0.001
9 months 43.16(0.24) 44.65(0.26) < 0.001 3.2.1. Voxel-based morphometry
12 months 44.45(0.25) 45.99(0.27) < 0.001
Whole brain VBM analysis revealed 4 clusters signicantly smaller
20 years 53.85(0.26) 54.91(0.29) 0.006
Total intracranial volume at 1397(17.54) 1545(26.88) < 0.001
in SGA compared to AGA adults: one cluster in the left frontal lobe
20 years (mean (SD)) [9, 36, 45] (T72 = 6.97, p < 0.001, FWE-corrected; cluster size:
38,822), including parts of the lateral and medial superior frontal
Note. Chi2 test;
Mann-Whitney test;
t-test. gyrus, orbitofrontal cortex and anterior cingulate cortex, putamen,
caudate nucleus and thalamus (see Fig. 1a). Smaller clusters in the right
3. Results inferior/supramarginal parietal gyrus [58.5, 45, 49.5] (T72 = 4.32,
p < 0.05, FWE-corrected; cluster size: 1086); right inferior/middle
3.1. Clinical, developmental, demographic and neurocognitive data temporal gyrus [61.5, 36, 25.5] (T72 = 5. 25, p < 0.05, FWE-
corrected; cluster size: 1429); and right cerebellum [10.5, 42,
All participants were born between 37 and 41 weeks of gestational 19.5] (T72 = 4.39, p < 0.001, FWE-corrected; cluster size: 2500)
age. However, as we can see in Table 1, maternal age at birth was were also observed (see Fig. 1b, c, d, respectively). When we corrected
higher in the SGA than in the AGA group (T74 = 2.74, p = 0.008). for total intracranial volume, the left superior frontal cluster was shown
There were no signicant dierences in gender and socioeconomic to be signicantly reduced [10.5, 34.5, 45] (T71 = 4.85, p < 0.05,
measures between the two groups. Weight, height and head perimeter FWE-corrected; cluster size: 1272).
were signicantly lower in SGA compared to AGA adults, at all data
points (at birth, 41 weeks of GA, 3, 6, 9, 12 months of corrected age and 3.2.2. Surface-based morphometry
20 years) (all ps < 0.006, nonparametric U Mann-Whitney tests), As we can see in Table 3, FreeSurfer cortical results are similar to
except for the weight at 20 years. Total intracranial volume was also results observed with VBM. Indeed, we observed a reduction in the
lower in SGA compared to AGA adults at the time of testing volume or surface area in several frontal regions (ROIs were based on
(T73 = 4.62, p < 0.001). the Desikan-Killiany Atlas [30]), in SGA compared to AGA adults: the
The participants' neurocognitive data are summarized in Table 2. left superior frontal gyrus volume; left orbitofrontal cortex volume and
Attention/executive functions decits were shown using the Test- surface area; and left and right anterior cingulate cortex surface areas.
battery of Attentional Performance with the following subtests signi- Smaller surface areas in the right inferior/supramarginal parietal gyrus
cantly reduced in the SGA group, as measured (see Table 2): phasic and the right inferior/middle temporal gyrus were also observed in SGA
alertness (T74 = 2.19, p = 0.03); sustained attention, false alarms compared to AGA adults (see Table 3). No cortical thickness dierences
(T53 = 2.32, p = 0.02) and omissions (T53 = 2.01, p = 0.05); were observed.
working memory, false alarms (T32 = 2.40, p = 0.009) and omissions At a subcortical level, and in agreement with what has been
(T62 = 1.99, p = 0.05); and go/nogo, false alarms (T74 = 2.91, observed with the VBM method, we obtained smaller right cerebellum
p = 0.005) and reaction time (T74 = 2.95, p = 0.004) subtests. gray matter volume; left and right thalamus; right caudate nucleus;
There were no signicant dierences in IQ and memory subtest (all right pallidum; left and right putamen volumes in SGA compared to
ps > 0.05). AGA adults (see Table 3). There was no dierence of volumes between


S. Suren et al. (DUO\+XPDQ'HYHORSPHQW

Fig. 1. Clusters represented a smaller volume in the participants born small for gestational age, compared to those born appropriate for gestational age, from the VBM analyses, for (a) the
frontal lobe (b) the cerebellum (c) the temporal lobe and (d) the parietal lobe. The color bars represent the t scores.

Table 3 Concerning the frontal lobe, the left medial frontal gyrus was
Smaller cortical volumes and surface areas and subcortical volumes between small for positively correlated with the sustained attention, false alarms subtest
gestational age (SGA) and appropriate for gestational age (AGA) groups.
(r = 0.24, p = 0.01) and omissions subtest (r = 0.26, p = 0.008); the
Side SGA (n = 38) AGA (n = 37) p left lateral orbitofrontal cortex (r = 0.20, p = 0.04) and the anterior
cingulate cortex (r = 0.20, p = 0.04) were also positively correlated to
Cortical the sustained attention, false alarms subtest. Concerning the parietal
Frontal lobe, the right inferoparietal gyrus was positively correlated with the
Superior volume L 18,564(351) 20,259(346) 0.001
Lateral orbitofrontal L 10,084(197) 10,980(199) 0.002
sustained attention, false alarms subtest (r = 0.20, p = 0.04).
volume Concerning the temporal lobe, the inferior temporal gyrus volume
Lateral orbitofrontal L 3209(59) 3473(60) 0.002 was positively correlated with the working memory, omissions subtest
surface area (r = 0.26, p = 0.04).
Anterior cingulate cortex L 1304(35) 1481(36) 0.001
Concerning subcortical structures, the right cerebellum was posi-
surface area
Anterior cingulate cortex R 1342(42) 1537(42) 0.002 tively correlated with the sustained attention, false alarms subtests
surface area (r = 0.20, p = 0.04); the left and right thalamus were positively
Temporal correlated with the working memory, omissions subtest (r = 0.35,
Middle surface area R 3108(63) 3439(63) < 0.001 p = 0.005; r = 0.26, p = 0.04, respectively). Concerning basal ganglia,
Parietal
Inferior surface area R 4954(94) 5449(95) < 0.001
the right putamen was positively correlated with the working memory,
Middle surface area L 3513(74) 3849(75) 0.002 omissions subtest (r = 0.27, p = 0.03); the caudate nucleus was posi-
Subcortical tively correlated with the sustained attention, omissions subtests
Thalamus L 6518(103) 7106(104) < 0.001 (r = 0.20, p = 0.04); the right pallidum was negatively correlated with
Thalamus R 6747(103) 7360(104) < 0.001
the Go/NoGo, reaction time subtest (r = 0.17, p = 0.03).
Caudate nucleus R 3658(79) 4020(80) 0.002
Putamen L 6281(93) 6773(94) < 0.001
Putamen R 6015(95) 6516(96) 0.001 3.2.4. Catch-up growth
Globus pallidus R 1620(39) 1866(40) < 0.001 SGA adults with a catch-up growth had a higher right superior
Cerebellar gray matter R 40,764(734) 44,196(744) 0.002
frontal gyrus compared to SGA adults without catch-up growth. This
Note. Results are from the FreeSurfer analyses.
cluster is signicant with [29, 18, 48] (T32 = 4.36, p < 0.05, FWE
corrected) and without [26, 17, 44] (T33 = 4.44, p < 0.05, FWE
the two groups using FreeSurfer when we corrected for total intracra- corrected) correction for total intracranial volume (see Fig. 2).
nial volume.
4. Discussion

3.2.3. Correlation with neurocognition
Correlation analyses were done on structures known to be asso-
ciated with attention and executive functions. Frontal and parietal lobes
seem very important in the control of attentional processes and
executive functions [3133]. Temporal lobes were also linked to
attention/executive functions [34]. The basal ganglia, the thalamus
and the cerebellum seem to be also activated during attentional tasks
[3133].
To our knowledge, this study is the rst to compare adults born at
term Small for Gestational Age (SGA, < 2001 g) and Appropriate for
Gestational Age (AGA, > 2500 g), with a VBM approach. In addition,
no study has used both methods (VBM and SBM) simultaneously.
Surface-based morphometry alone, had already been used to measure
the surface area, cortical thickness and subcortical volumes of adoles-
cents and SGA adults [12,13]. The simultaneous use of both methods
allows an internal validation of our results and underlines the strength


S. Suren et al.
Fig. 2. Cluster represented a higher volume in the participants born small for gestational age with catch-up growth, compared to those born small for gestational age without catch-up
growth. The color bar represents the t score.
of this study. In our cohort, SGA adults had lower attention/executive
functions scores and smaller brain volumes and surface areas compared
with AGA adults.
The brain regions that exhibit the greatest reductions in volume
and/or surface area were the frontal, temporal and parietal lobes, the
cerebellum, thalamus and basal ganglia, similarly to Ostgard et al.
(2014). When corrected for total intracranial volume, the left superior
frontal lobe volume was shown to be reduced with the VBM method.
The anatomical cerebral studies on being born SGA generally do not
correct for the total intracranial volume, except for the subcortical
volumes. In this study, we reported the corrected and uncorrected
results. As being born SGA is accompanied by a smaller total intracra-
nial volume, we interpret results uncorrected for the total intracranial
volume, so as not to correct for the disease itself.
Our results are consistent with smaller anterior cingulate, frontal,
temporal and parietal lobes surface areas reported in young SGA adults
born at term [13]. Our results are all the more important that we found
similar alterations with both methods (VBM and FreeSurfer). Although
the results with FreeSurfer related alterations in the surface area more
than in the volume, damaged structures remain similar with both
methods.
On the other hand, as Ostgard et al. (2014), we found no dierence
in cortical thickness. Surface area and volume measures seem more
vulnerable to the eects of being born SGA than the cortical thickness
measures [35,36]. Moreover, gray matter volume dierences seem
more closely related and driven by surface area rather than cortical
thickness dierences [3739]. The radial unit hypothesis of cortical
development suggests that cortical thickness and surface area measures
are genetically and phenotypically independent, and are driven by
distinct cellular mechanisms [3840]. It is therefore not surprising that
our results show similar dierences in the measurements of cortical
gray matter volume and surface area, but absence of dierences in the
cortical thickness measures.
At a subcortical level, smaller cerebellum volume in the SGA adults
born at term is similar to results observed in SGA children [11] and
adolescents [12] born at term. Smaller thalamus volume in SGA adults
born at term is similar to results in SGA adolescents born at term [12].
Smaller basal ganglia (caudate nucleus, nucleus accumbens, pallidum,
putamen) volumes in SGA adults born at term is similar to results of
other studies with SGA children [11] and adults [13] born at term.
Thalamus, cerebellum and basal ganglia seem to be particularly
vulnerable to injury/toxin exposure during fetal brain development
[4143].
Attention/executive functions were already shown to be aected in
very low birth weight children [1] and young adults [34]. Moreover,
Ostgard et al. (2016) found correlations between executive functions
scores and surface areas of the frontal and temporal lobes in a cohort of
very low birthweight young adults [34]. Interestingly, our study shown
these decits are also present in young adults born SGA, and shown
similar correlations. Indeed, smaller frontal lobe volume and surface

(DUO\+XPDQ'HYHORSPHQW
area, as well as smaller temporal lobe surface area were linked to
attention/executive functions decits observed in the SGA adults. In
addition to Ostgard et al. (2016), smaller parietal volume; smaller
cerebellum, thalamus and basal ganglia volumes were also linked to
attention/executive functions decits observed in the SGA adults.
On the other hand, it was interesting to rediscover the known
positive eect of catch-up growth in the rst year as previously
described by Ostgard et al. (2014). In our study, catch-up growth was
shown to have a clear protective eect on frontal lobe volume.
One limitation of this study is the lack of data on parental weight
and height. Indeed, we cannot arm here that our participants are not
genetically small. Future studies should take these variables into
account. However, before speaking of small genetic size, it would be
necessary to eliminate the bad nutrition present in this poor Colombian
population of 20 years ago.
In conclusion, being born small for gestational age, without being
premature, is linked to smaller frontal lobe volume and attention/
executive function decits in adulthood.
Conict of interest
All authors declare that they have no nancial interests or potential
conicts of interest.
Acknowledgements
This work was supported by Grand Challenge Canada (< grant
number 0059-03 > [to < NC > ]); Fundacion Canguro y
COLCIENCIAS (National Colombia Organization for Science and
Technology), (< grant number 714-2013 > [to < NC > ]); Brain
Canada, the Research Center of the Sainte Justine Hospital, Sainte-
Justine Foundation and the Cercle des Jeunes Leaders [to < GL > ].
We thank Alan Evans and CBRAIN for their support in this project.
We also thank all the families who participated in the research, as
well as the members of the Kangaroo research team.
References
[1] C.S.H. Aarnoudse-Moens, N. Weisglas-Kuperus, J. van Goudoever, J. Oosterlaan,
Meta-analysis of neurobehavioral outcomes in very preterm and/or very low birth
weight children, Pediatrics 124 (2009) 717728, http://dx.doi.org/10.1542/peds.
2008-2816.
[2] O. Hollo, P. Rautava, T. Korhonen, H. Helenius, P. Kero, M. Sillanp, Academic
achievement of small-for-gestational-age children at age 10 years, Arch. Pediatr.
Adolesc. Med. 156 (2002) 179187, http://dx.doi.org/10.1001/archpedi.156.2.
179.
[3] J. Kutschera, B. Urlesberger, U. Maurer, W. Mller, Small for gestational age -
somatic, neurological and cognitive development until adulthood, Zeitschrift Fr
Geburtshilfe Und Neonatol. 206 (2002) 6571, http://dx.doi.org/10.1055/s-2002-
30139.
[4] B. Larroque, S. Bertrais, P. Czernichow, J. Lger, School diculties in 20-year-olds
who were born small for gestational age at term in a regional cohort study,
Pediatrics 108 (2001) 111115, http://dx.doi.org/10.1542/peds.108.1.111.
S. Suren et al.
[5] G.C.C. Lhaugen, H.F. stgrd, S. Andreassen, G.W. Jacobsen, T. Vik, A.-
M. Brubakk, J. Skranes, M. Martinussen, Small for gestational age and intrauterine
growth restriction decreases cognitive function in young adults, J. Pediatr. 163
(2013) 447453, http://dx.doi.org/10.1016/j.jpeds.2013.01.060.
[6] I. Paz, R. Gale, A. Laor, Y.L. Danon, D.K. Stevenson, D.S. Seidman, The cognitive
outcome of full-term small for gestational age infants at late adolescence, Obstet.
Gynecol. 85 (1995) 452456.
[7] R.S. Strauss, Adult functional outcome of those born small for gestational age:
twenty-six-year follow-up of the 1970 British birth cohort, J. Am. Med. Assoc. 283
(2000) 625632.
[8] R.F. Theodore, J.M.D. Thompson, K.E. Waldie, D.M.O. Becroft, E. Robinson,
C.J. Wild, P.M. Clark, E.A. Mitchell, Determinants of cognitive ability at 7 years: a
longitudinal casecontrol study of children born small-for-gestational age at term,
Eur. J. Pediatr. 168 (2009) 12171224, http://dx.doi.org/10.1007/s00431-008-
0913-9.
[9] M. Martinussen, D.W. Flanders, B. Fischl, E. Busa, G.C. Lhaugen, J. Skranes,
T.R. Vangberg, A.-M. Brubakk, O. Haraldseth, A.M. Dale, Segmental brain volumes
and cognitive and perceptual correlates in 15-year-old adolescents with low birth
weight, J. Pediatr. 155 (2009) 848853.e1, http://dx.doi.org/10.1016/j.jpeds.
2009.06.015.
[10] N. Charpak, J.G. Ruiz, The Kangaroo Mother Care Method: from scientic evidence
generated in Colombia to worldwide practice, J. Clin. Epidemiol. (2016), http://dx.
doi.org/10.1016/j.jclinepi.2016.05.019.
[11] H.M.A. De Bie, K.J. Oostrom, M. Boersma, D.J. Veltman, F. Barkhof,
H.A. Delemarre-van de Waal, M.P. van den Heuvel, Global and regional dierences
in brain anatomy of young children born small for gestational age, PLoS One 6
(2011) 111, http://dx.doi.org/10.1371/journal.pone.0024116.
[12] T. Rogne, A.A. Engstrm, G.W. Jacobsen, J. Skranes, H.F. stgrd, M. Martinussen,
Fetal growth, cognitive function, and brain volumes in childhood and adolescence,
Obstet. Gynecol. 125 (2015) 673682, http://dx.doi.org/10.1097/aog.
0000000000000694.
[13] H.F. stgrd, G.C.C. Lhaugen, K.J. Bjuland, L.M. Rimol, A.-M. Brubakk,
M. Martinussen, T. Vik, A.K. Hberg, J. Skranes, Brain morphometry and cognition
in young adults born small for gestational age at term, J. Pediatr. 165 (2014)
921927, http://dx.doi.org/10.1016/j.jpeds.2014.07.045.
[14] N. Charpak, J.G. Ruiz-Pelaez, C.Z. Figueroa de, Kangaroo mother versus
Traditional care for newborn infants < 2000 grams: a randomized control trial, J.
Clin. Epidemiol. 51 (1997) S12, http://dx.doi.org/10.1016/S0895-4356(98)
90036-1.
[15] F.C. Battaglia, L.O. Lubchenco, A practical classication of newborn infants by
weight and gestational age, J. Pediatr. 71 (1967) 159163, http://dx.doi.org/10.
1016/S0022-3476(67)80066-0.
[16] J.M. Wit, B. Boersma, Catch-up growth: denition, mechanism, and models, J.
Pediatr. Endocrinol. Metab. 15 (2002) 12291241.
[17] WHO, WHO Child Growth Standards: Length/Height-for-age, Weight-for-age,
Weight-for-length, Weight-for-Height and Body Mass Index-for-age: Methods and
Development, (2006).
[18] D. Wechsler, C. Hsiao-pin, WASI-II: Wechsler Abbreviated Scale of Intelligence,
Pearson, 2011.
[19] D.C. Delis, J.H. Kramer, E. Kaplan, B.A. Ober, CVLT-II California Verbal Learning
Test Manual Adult Version, San Antonio, Texas Psychol. Corp, 2000.
[20] P. Zimmermann, B. Fimm, Test d'valuation de l'attention (TAP 2.2), Herzogenrath,
Ger. PsyTest, 2009.
[21] J. Ashburner, K.J. Friston, Voxel-based morphometry-the methods, NeuroImage 11
(2000) 805821, http://dx.doi.org/10.1006/nimg.2000.0582.
[22] J. Ashburner, Computational anatomy with the SPM software, Magn. Reson.
Imaging 27 (2009) 11631174, http://dx.doi.org/10.1016/j.mri.2009.01.006.
[23] J. Ashburner, K.J. Friston, Unied segmentation, NeuroImage 26 (2005) 839851,
http://dx.doi.org/10.1016/j.neuroimage.2005.02.018.
[24] J. Ashburner, A fast dieomorphic image registration algorithm, NeuroImage 38
(2007) 95113, http://dx.doi.org/10.1016/j.neuroimage.2007.07.007.
[25] T. Sherif, P. Rioux, M.-E. Rousseau, N. Kassis, N. Beck, R. Adalat, S. Das, T. Glatard,
A.C. Evans, CBRAIN: a web-based, distributed computing platform for collaborative
neuroimaging research, Front. Neuroinform. 8 (2014) 54, http://dx.doi.org/10.
3389/fninf.2014.00054.
[26] B. Fischl, FreeSurfer, NeuroImage 62 (2012) 774781, http://dx.doi.org/10.1016/j.
neuroimage.2012.01.021.
[27] R.A. Morey, C.M. Petty, Y. Xu, J. Pannu Hayes, H.R. Wagner Ii, D.V. Lewis,

(DUO\+XPDQ'HYHORSPHQW
K.S. LaBar, M. Styner, G. McCarthy, A comparison of automated segmentation and
manual tracing for quantifying hippocampal and amygdala volumes, NeuroImage
45 (2009) 855866, http://dx.doi.org/10.1016/j.neuroimage.2008.12.033.
[28] B. Fischl, D.H. Salat, E. Busa, M. Albert, M. Dieterich, C. Haselgrove, A. van der
Kouwe, R. Killiany, D. Kennedy, S. Klaveness, A. Montillo, N. Makris, B. Rosen,
A.M. Dale, Whole brain segmentation: neurotechnique automated labeling of
neuroanatomical structures in the human brain, Neuron 33 (2002) 341355, http://
dx.doi.org/10.1016/S0896-6273(02)00569-X.
[29] B. Fischl, A.M. Dale, Measuring the thickness of the human cerebral cortex from
magnetic resonance images, Proc. Natl. Acad. Sci. 97 (2000) 1105011055, http://
dx.doi.org/10.1073/pnas.200033797.
[30] R.S. Desikan, F. Sgonne, B. Fischl, B.T. Quinn, B.C. Dickerson, D. Blacker,
R.L. Buckner, A.M. Dale, R.P. Maguire, B.T. Hyman, M.S. Albert, R.J. Killiany, An
automated labeling system for subdividing the human cerebral cortex on MRI scans
into gyral based regions of interest, NeuroImage (2006), http://dx.doi.org/10.
1016/j.neuroimage.2006.01.021.
[31] T. Yarkoni, R. Poldrack, T. Nichols, D. Van Essen, T. Wager, Large-scale automated
synthesis of human functional neuroimaging data, Nature 8 (2011) 665674.
[32] J. Duncan, The structure of cognition: attentional episodes in mind and brain,
Neuron 80 (2013) 3550, http://dx.doi.org/10.1016/j.neuron.2013.09.015.
[33] S. Llufriu, E. Martinez-Heras, E. Solana, N. Sola-Valls, M. Sepulveda, Y. Blanco,
E.H. Martinez-Lapiscina, M. Andorra, P. Villoslada, A. Prats-Galino, A. Saiz,
Structural networks involved in attention and executive functions in multiple
sclerosis, NeuroImage Clin. 13 (2017) 288296, http://dx.doi.org/10.1016/j.nicl.
2016.11.026.
[34] H.F. stgrd, A. Elisabeth Slsnes, K. Jrgen Bjuland, L. Morten Rimol,
M. Martinussen, A.-M. Brubakk, A. Kristine Hberg, J. Skranes, G. Christine
Christensen Lhaugen, Executive function relates to surface area of frontal and
temporal cortex in very-low-birth-weight late teenagers, Early Hum. Dev. 95 (2016)
4753, http://dx.doi.org/10.1016/j.earlhumdev.2016.01.023.
[35] A. Raznahan, D. Greenstein, N.R. Lee, L.S. Clasen, J.N. Giedd, Prenatal growth in
humans and postnatal brain maturation into late adolescence, Proc. Natl. Acad. Sci.
109 (2012) 1136611371, http://dx.doi.org/10.1073/pnas.1203350109.
[36] K.B. Walhovd, A.M. Fjell, T.T. Brown, J.M. Kuperman, Y. Chung, D.J. Hagler,
J.C. Roddey, M. Erhart, C. McCabe, N. Akshoomo, D.G. Amaral, C.S. Bloss,
O. Libiger, N.J. Schork, B.F. Darst, B.J. Casey, L. Chang, T.M. Ernst, J. Frazier,
J.R. Gruen, W.E. Kaufmann, S.S. Murray, P. van Zijl, S. Mostofsky, A.M. Dale, Long-
term inuence of normal variation in neonatal characteristics on human brain
development, PNAS 109 (2012) 2008920094, http://dx.doi.org/10.1073/pnas.
1208180109.
[37] P.A. Kelly, E. Viding, G.L. Wallace, M. Schaer, S.A. De Brito, B. Robustelli,
E.J. McCrory, Cortical thickness, surface area, and gyrication abnormalities in
children exposed to maltreatment: neural markers of vulnerability? Biol. Psychiatry
74 (2013) 845852, http://dx.doi.org/10.1016/j.biopsych.2013.06.020.
[38] M.S. Panizzon, C. Fennema-Notestine, L.T. Eyler, T.L. Jernigan, E. Prom-Wormley,
M. Neale, K. Jacobson, M.J. Lyons, M.D. Grant, C.E. Franz, H. Xian, M. Tsuang,
B. Fischl, L. Seidman, A. Dale, W.S. Kremen, Distinct genetic inuences on cortical
surface area and cortical thickness, Cereb. Cortex 19 (2009) 27282735, http://dx.
doi.org/10.1093/cercor/bhp026.
[39] A.M. Winkler, P. Kochunov, J. Blangero, L. Almasy, K. Zilles, P.T. Fox, R. Duggirala,
D.C. Glahn, Cortical thickness or grey matter volume? The importance of selecting
the phenotype for imaging genetics studies, NeuroImage 53 (2010) 11351146,
http://dx.doi.org/10.1016/j.neuroimage.2009.12.028.
[40] P. Rakic, Radial unit hypothesis of neocortical expansion, Evol. Dev. Biol. Cereb.
Cortex (2000) 3045.
[41] S.N. Mattson, E.P. Riley, T.L. Jernigan, A. Garcia, W.M. Kaneko, C.L. Ehlers,
K.L. Jones, A decrease in the size of the basal ganglia following prenatal alcohol
exposure: a preliminary report, Neurotoxicol. Teratol. 16 (1994) 283289, http://
dx.doi.org/10.1016/0892-0362(94)90050-7.
[42] C.H. Chang, F.M. Chang, C.H. Yu, H.C. Ko, H.Y. Chen, Assessment of fetal cerebellar
volume using three-dimensional ultrasound, Ultrasound Med. Biol. 26 (2000)
981988, http://dx.doi.org/10.1016/S0301-5629(00)00225-8.
[43] J.P. Boardman, S.J. Counsell, D. Rueckert, O. Kapellou, K.K. Bhatia, P. Aljabar,
J. Hajnal, J.M. Allsop, M.A. Rutherford, A.D. Edwards, Abnormal deep grey matter
development following preterm birth detected using deformation-based morpho-
metry, NeuroImage 32 (2006) 7078, http://dx.doi.org/10.1016/j.neuroimage.
2006.03.029.