Early Postnatal Weight Gain, Intellectual Performance, and Body Mass

Index at 7 Years of Age in Term Infants with Intrauterine

Growth Restriction
MARY PYLIPOW, MD, LOGAN G. SPECTOR, PHD, SUSAN E. PUUMALA, MS, CHRISTOPHER BOYS, PHD, JESSICA COHEN, BS,
AND MICHAEL K. GEORGIEFF, MD

Objective To determine whether the postnatal growth rate of infants with intrauterine growth restriction (IUGR) is
associated with later cognitive function and body mass index (BMI).
Study design Infants with IUGR (<2211 g at >37 weeks’ gestation) were identified in data from the Collaborative Perinatal
Project, excluding those with diagnoses affecting cognition or growth. Wechsler Scale of Children’s Intelligence (WISC) scores
at age 7 years and data on postnatal growth at 16 weeks were available for 463 infants with IUGR. Linear regression relating
postnatal growth and WISC score, adjusting for potential confounders, was performed for these infants. BMI at 7 years also was
examined.
Results Weight gain at 16 postnatal weeks ranged from 1059 to 5119 g in the infants with IUGR, with lower achieved
cognitive testing scores apparent at both extremes (ie, an inverted J-shape; P < .001). Infants gaining 1200 and 5000 g scored
15.5 and 2.4 fewer points, respectively, on the full scale compared with infants with score-maximizing growth. In contrast, BMI
at 7 years was linearly related to postnatal weight gain (P < .001).
Conclusions Growth in the first 4 postnatal months is an independent risk factor for cognitive outcome at age 7 years, with
both extremes associated with negative effects. (J Pediatr 2009;154:201-6)

ntrauterine growth restriction (IUGR) is the most common cause for small for gestational age (SGA) status in term and

I preterm births worldwide. Although most infants with IUGR demonstrate catch-up growth in the first 3 postnatal months,
approximately 10% remain small for life. Infants with IUGR are at increased risk for worse cognitive outcomes, metabolic
syndrome (comprising hypertension, diabetes, and cardiovascular disease),1 precocious puberty, and short stature2-5 compared
with appropriate for gestational age (AGA) infants. Concerns about the combined effect of IUGR and postnatal growth failure
have driven the common postnatal nutritional strategy of inducing the most rapid postnatal catch-up growth possible.
Although poor postnatal growth after IUGR further increases the cognitive risks of IUGR,6 the potential adverse effects of
excessive weight gain after prolonged IUGR on neurodevelopment have not yet been assessed. An increased risk of cardiovascular and
metabolic diseases when IUGR is coupled with rapid weight gain has been reported.2 The
developmental origins of health and disease (DOHaD) hypothesis states that the nutritional
environment present during times of active fetal development may program metabolism for
life.1 Also known as the “thrifty phenotype hypothesis,”7-9 the DOHaD hypothesis posits that From the Divisions of Neonatology (M.P.,
the early adaptive alterations in neurologic and endocrine mechanisms that promote survival M.G.), Epidemiology/Clinical Research (L.S.,
of the fetus during adverse intrauterine conditions are maladaptive later when living in S.P.), and Academic General Pediatrics
(C.B.), Department of Pediatrics, University
enriched postnatal conditions. The mechanisms of altered regulation are only now being of Minnesota Medical School, Minneapolis,
10,11
understood and appear to involve alterations in the hypothalamic-pituitary-adrenal axis MN and School of Medicine, Emory Uni-
associated with activation of inflammatory cytokines.12,13 Because they cross the blood-brain versity, Atlanta, GA (J.C.).
The authors declare no conflicts of interest,
barrier, these factors have potential effects on higher central nervous system regions and real or perceived.
processes that develop in the early postnatal period. For example, excessive glucocorticoid Submitted for publication Jun 10, 2008; last
exposure reduces hippocampal neurogenesis, differentiation and volume, and proinflammatory revision received Jul 15, 2008; accepted
Aug 7, 2008.
cytokines alter myelination.14,15 Both mechanisms therefore present potential risks to developing Reprint requests: Michael Georgieff, MD,
brain structures and may result in altered long-term cognitive function. Our aim was to analyze the Division of Neonatology, University of Min-
nesota, MMC 39, D-136 Mayo, 420 Dela-
ware St SE, Minneapolis, MN 55455. E-mail:
AGA Appropriate for gestational age IUGR Intrauterine growth restriction georg001@umn.edu.
BMI Body mass index SES Socioeconomic status 0022-3476/$ - see front matter
CPP Collaborative Perinatal Project SGA Small for gestational age Copyright © 2009 Mosby Inc. All rights
DOHaD Developmental origins of health and disease WISC Wechsler Intelligence Scale for Children reserved.
FTT Failure to thrive
10.1016/j.jpeds.2008.08.015

201
long-term effects on growth parameters and cognitive tasks of early Table I. Mean WISC component and summary
excessively rapid weight gain on specific cognitive tasks evaluated by scores at age 7 years for term children with IUGR
the WISC subscores using data from the large, multicenter, lon- and AGA children in the CPP
gitudinal Collaborative Perinatal Project (CPP).
IUGR AGA
METHODS WISC (n ⴝ 503) (n ⴝ 30 412)
The CPP was a prospective cohort study of children component Mean SD Mean SD P value*
born between 1959 and 1965; the methods are described in Full 89.97 13.62 97.25 14.27 ⬍.0001
detail elsewhere.16 Neurologic and intellectual development Verbal 88.92 12.71 95.45 14.11 ⬍.0001
was the primary outcome of interest. Approximately 45 000 Performance 92.96 15.05 99.72 15.12 ⬍.0001
women were enrolled during pregnancy at 12 United States Information 8.26 2.75 9.39 3.02 ⬍.0001
academic institutions. A total of 59 843 pregnancies, resulting Comprehension 7.92 2.50 8.89 2.75 ⬍.0001
in 54 795 live births, were recorded among these women. Vocabulary 7.90 2.82 8.97 3.07 ⬍.0001
Extensive maternal, pregnancy, social, demographic and fam- Digit span 8.75 2.93 9.77 2.94 ⬍.0001
Picture 8.34 2.94 9.81 3.21 ⬍.0001
ily history data were collected at multiple time points by
Block 8.78 2.72 9.89 2.82 ⬍.0001
trained personnel using standardized methods. Children were Code 9.88 3.02 10.19 2.88 .02
examined several times during the newborn period and at age
SD, standard deviation.
4, 8, and 12 months. Investigators continued to collect de- *P value from the t-test for 2 means.
tailed data at several time points up to age 7 or 8 years,
depending on the institution. Analysis of the CPP dataset,
which is deidentified and in the public domain, did not full-scale IQ and individual subtest scaled scores were com-
require Institutional Review Board approval. pared between children with IUGR and AGA children using
We conducted analyses on 2 segments of the CPP study the t-test for 2 means.
population. In the first analysis, we compared all component Linear regression was then used to assess the influence
and full-scale scores of the original Wechsler Intelligence of weight gain at 16 weeks postpartum on cognitive and
Scale (WISC)17,18 at age 7 years in children born at full term anthropometric outcomes at age 7 years while controlling for
who had IUGR and those who were AGA. A total of 34 460 race, SES, sex, and birth weight as categorical variables. The
children were born at ⱖ37 weeks’ gestation, completed a natural logs of weight and BMI at age 7 years were used to
WISC assessment at age 7 years, and had allowable values for meet the normality assumption of the regression model. SES
all WISC component scores. IUGR was defined as the 5th comprised 5 categories, as assessed by the original CPP in-
percentile of birth weight in this cohort (⬍2211 g). We chose vestigators,20 and race comprised 3 categories (Caucasian,
a more stringent criterion than the standard definition of African American, and other). Birth weight was categorized
⬍2500 g,19 in an effort to eliminate small infants who may using natural cutpoints (ⱕ2000 g, 2001 to 2150 g, and 2151
not have had IUGR. We excluded children with conditions to 2211 g), which divided the IUGR cohort approximately
likely to diminish intellectual capacity, including those with into tertiles.
5-minute APGAR scores ⬍6 (n ⫽ 606), congenital syn- The significance of 16-week weight gain as a quadratic
dromes (n ⫽ 633), or microcephaly (n ⫽ 40). We also variable was evaluated by entering both continuous values of
eliminated children with missing or invalid data points, leav- centered weight gain and the square of centered weight gain
ing a total of 30 915 children available for the first analysis. in linear regression models. Using centered weight gain (cal-
Our second analysis related early postnatal weight gain culated as the difference between each observation and the
to weight, WISC score, and 2 anthropometric measures (ie, mean) avoided the problem of collinearity between the linear
weight and body mass index [BMI]) at age 7 years in the and quadratic terms. The t-test of the quadratic term was used
infants with IUGR. The CPP protocol called for weighing to evaluate the fit of a parabola, to describe the relationship
infants at age 16 weeks, but because the actual age at weighing between 16-week weight gain and each dependent variable.
varied (mean age, 16.4 weeks ⫾ 2.4 weeks), we calculated the The quadratic term was dropped from the model if it was not
grams gained per week between birth and age at weighing and significant at ␣ ⫽ 0.05, and the linear term was then evalu-
multiplied by 16 to obtain a consistent metric. Forty children ated.
were excluded because of missing growth data, and 1 child When either the quadratic or linear terms improved
was excluded because of missing WISC data, leaving 463 model fit significantly, estimated mean WISC component
children with IUGR available for analysis. scores with 95% confidence intervals were calculated for low,
middle, and high values of 16-week weight gain using a “base
Statistical Analysis case” scenario. The “base case” values of all covariates were set
Characteristics of the cohort, including race, socioeco- to the following referents: male sex, Caucasian race, middle
nomic status (SES), and sex, were tabulated by birth weight SES category, and birth weight 2151 to 2211 g. The rounded
status using contingency tables for categorical variables and extremes and median of the observed distribution of 16-week
means for continuous variables. Mean WISC scores and weight gain were used as the low, middle, and high values

202 Pylipow et al The Journal of Pediatrics • February 2009

Table II. Variable estimates (␤) from linear regression of independent variables in relation to WISC
summary scores and anthropometric measures at age 7 years in infants with IUGR
WISC score Anthropometric measures
Full score Performance Verbal ln (weight) ln (BMI)
Variable n ⴝ 463 % ␤ P value B P value ␤ P value ␤ P value ␤ P value
Intercept 62.03 ⬍.001 67.96 ⬍.001 63.30 ⬍.001 3.55 ⬍.001 2.60 ⬍.001
Sex
Male 172 37.1 Ref Ref Ref Ref Ref
Female 291 62.9 1.25 .29 2.10 .13 0.36 .75 ⫺0.001 .95 ⫺0.01 .29
Race
Caucasian 169 36.5 Ref Ref Ref Ref Ref
African American 274 59.2 ⫺7.84 ⬍.001 ⫺7.64 ⬍.001 ⫺6.75 ⬍.001 ⫺0.01 .41 ⫺0.05 ⬍.001
Other 20 4.3 ⫺8.59 .003 ⫺1.27 .71 ⫺13.71 ⬍.001 ⫺0.07 .08 ⫺0.06 .03
SES index
1 52 11.2 ⫺0.33 .87 0.60 .80 ⫺1.21 .52 0.03 .24 0.02 .18
2 158 34.1 ⫺0.20 .89 0.21 .90 ⫺0.52 .69 0.01 .76 0.01 .29
3 153 33.0 Ref Ref Ref Ref Ref
4 74 16.0 7.15 ⬍.001 7.26 ⬍.001 5.65 ⬍.001 ⫺0.01 .65 0.01 .68
5 26 5.6 10.39 ⬍.001 9.39 .002 9.49 ⬍.001 ⫺0.08 .02 ⫺0.04 .09
Birth weight, g*
2151-2211 152 32.8 Ref Ref Ref Ref Ref
2000-2150 159 34.3 1.74 .20 0.99 .53 2.14 .09 0.01 .65 0.01 .55
⬍2000 152 32.8 ⫺0.17 .90 0.37 .82 ⫺0.69 .59 ⫺0.01 .50 ⫺0.02 .21
Weight gain at 16
weeks, g†
Increase per 100 g 1.67 .003 1.47 .02 1.55 .003 0.01 ⬍.001 0.005 ⬍.001
Increase per 100 g2 ⫺0.02 .01 ⫺0.019 .05 ⫺0.02 .01 NS NS
Ref, referent value; NS, not significant at P ⬍ .05.
*Mean, 2 025.9 ⫾ 182.2 g.
†Mean, 3 152.0 ⫾ 674.4 g.

when the linear model was significant for WISC component
scores. When the quadratic model was significant, the 16-
week weight gain that maximized the WISC score (ie, the
vertex of the parabola) was used as the middle value. Similar
modeling was performed for weight gain over the first 4 weeks
and over the first 12 months (data not shown). The results
were similar across all 3 time periods but were somewhat
stronger using the first 16 weeks.
RESULTS
Mean IQ scores were consistently lower in the IUGR
group compared with the AGA group, with deficits of 7.28
points for full-scale IQ score, 6.53 points for verbal IQ score,
and 6.76 points for performance IQ score (Table I). Individ-
ual subtest scores also were significantly lower in the IUGR
group. The statistical significance of these comparisons was
not surprising given the large sample size, but the differences
also were qualitatively quite substantial.
Characteristics of the children with IUGR included in
linear regression analyses are summarized in Table II. A
majority of the cohort was male, African American, and from
the middle of the SES scale. Mean birth weight was 2025.5 g ⫾
182.1 g. Mode of feeding was not included as a covariate,
because a large majority (96%) of infants with IUGR were
formula-fed. Table II also gives parameter estimates (␤) from
linear regression models. Among the covariates, SES and race
demonstrated significant associations with WISC full, per-
formance, and verbal scores, whereas birth weight within the
IUGR group was unrelated.
Postnatal weight gain between birth and age 16 weeks
was the main variable of interest. The mean weight gain in the
IUGR group during this period was 3152 ⫾ 675 g (90th
percentile, 3992 g); these values are similar to those in AGA
group (mean, 3143 ⫾ 706 g; 90th percentile, 4050 g). Mod-
eling weight gain as a quadratic term provided a significantly
better fit to the data compared with modeling it as a linear
term for the WISC full IQ score (P ⫽ .01) and verbal IQ
score (P ⫽ .01), with a borderline significant quadratic term
(P ⫽ .05) for the performance IQ score (Table II). These
results remained significant after exclusion of influential ob-
servations identified by studentized deleted residuals and
DFFITS (data not shown) Among the WISC subtests, qua-
dratic terms were significant for the vocabulary (P ⫽ .03) and
information (P ⫽ .02) scaled scores, and linear terms were at
or near significance for the comprehension (P ⫽ .06) and digit
span (P ⫽ .005) scaled scores; no association was apparent
between postnatal weight gain and the block, coding, or
picture score. Postnatal weight gain was significantly associ-
ated with the log of weight and BMI at age 7 years as a linear,
but not a quadratic, term (P ⬍ .001 for both). Modeling

Early Postnatal Weight Gain, Intellectual Performance, and Body Mass Index at 7 Years of Age
in Term Infants with Intrauterine Growth Restriction 203
Table III. Mean WISC scores at age 7 years in infants with IUGR with low, middle, and high 16-week
postnatal weight gain*
WISC score
component Weight gain form Weight gain at 4 months, g† Mean WISC score‡ 95% CI
Full Quadratic 1200 78.99 71.32-86.66
3887 94.52 91.35-97.69
5000 92.14 85.81-97.89
Performance§ Quadratic 1200 82.87 73.96-91.78
3801 96.00 92.33-99.66
5000 93.21 86.19-100.23
Verbal Quadratic 1200 79.09 72.00-86.19
3975 94.11 91.14-97.07
5000 92.06 86.47-97.65
Information Quadratic 1200 6.13 4.50-7.77
3891 8.94 8.27-9.61
5000 8.23 6.94-9.51
Comprehension¶ Linear 1200 7.97 6.98-8.96
3119 8.59 7.96-9.22
5000 9.20 8.39-10.01
Vocabulary Quadratic 1200 6.40 4.85-7.95
3924 9.19 8.55-9.83
5000 8.76 7.54-9.98
Digit span Linear 1200 7.60 6.45-8.75
3119 8.71 7.97-9.44
5000 9.79 8.84-10.73
CI, Confidence interval.
*Derived from linear regression models described in Methods and adjusted for sex, race/ethnicity, SES, and birth weight.
†For linear models, the middle value is the median 16-week weight gain. Otherwise, the middle value was calculated as the vertex of the parabola described by parameter estimates
for the quadratic 16-week weight gain.
‡Mean score setting covariates to reference values of race, SES, and birth weight.
§Quadratic term P value ⫽ .05. Results for the quadratic model are still presented, because the P value represents a marginal association.
¶Linear term P value ⫽ .06. Results for the linear model are still presented, because the P value represents a marginal association.

predicted that term infants with IUGR who gain 1200 g and
5000 g in the first 16 weeks of life will score 15.53 and 2.38
fewer points, respectively, on the WISC full-scale IQ score at
age 7 years compared with those who gain the score-maxi-
mizing weight of 3887 g (Table III). Fitted quadratic regres-
sion curves for WISC summary scores and linear regression
curves for BMI, with the common x-axis of postnatal weight
gain, are depicted in the Figure.
DISCUSSION
In a large cohort of infants with IUGR, we demon-
strated that excessively rapid weight gain in the first 16
postnatal weeks after prenatal growth failure predicted both
increased BMI and decreased cognitive scores at age 7 years.
Early weight gain and BMI had a direct linear relationship,
but cognitive functions generated an inverted J-shaped risk
curve. These effects remained after controlling for SES, sex,
race, and birth weight. Consistent with previous studies, slow
postnatal growth predicted the worst long-term outcome, and
the effect was dose-responsive. As might be expected, the
effect of excessive weight gain was less pronounced than that
of suboptimal weight gain. Interestingly, the degree of IUGR
below a birth weight of 2211 g (5th percentile) did not
correlate with either the rate of postnatal catch-up growth or
later cognitive outcomes. The strengths of the present analysis
are the use of a large longitudinal study, a more stringent
definition of IUGR that eliminates most other causes of
smallness, and regression techniques that control for potential
confounding variables.
Most previous studies of either SGA infants or infants
with IUGR have assessed neurodevelopmental outcomes with-
out accounting for rates of postnatal growth.6,19,21-23 Impair-
ments in verbal ability and spatial memory, as well as modest
losses in general IQ, were reported for small infants compared
with AGA infants. Reduced attention, lower reading scores
(Wide Range Achievement Test),21 IQ, and school achieve-
ment19 have all been reported in term SGA infants. Neonatal
hippocampal function, as indexed by auditory recognition mem-
ory, is known to be compromised after IUGR.24 Magnetic
resonance imaging of cortical gray matter at term has demon-
strated volume losses that correlate with later functional deficits
in attention, cognition, and school performance.23 Regions af-
fected by prenatal nutrient deficiencies include the striatum,
cortex, and hippocampus.25-27
The neurobehavioral implications of slow catch-up post-
natal growth after IUGR has been the subject of fewer studies,
but the data suggest that poor postnatal catch-up growth in-
creases neurodevelopmental risk after IUGR. The timing of
growth failure was addressed by Casey et al,6 in a cohort study
comparing the effects at age 8 years of antenatal growth failure

204 Pylipow et al The Journal of Pediatrics • February 2009


Figure. Fitted regression curves and 95% confidence intervals for BMI
and A, full, B, verbal, and C, performance WISC scores at age 7 years by
16-week postnatal weight gain in infants with IUGR.
and postnatal failure to thrive (FTT) occurring between 4 and 36
months corrected gestational age. Low birth weight preterm
infants who were SGA and had postnatal FTT had the worst
outcomes, followed by those with only postnatal FTT. Math and
reading capabilities were at particular risk. Infants who were
SGA without postnatal FTT had normal academic achievement
and cognitive function as assessed by the full-scale, performance,
and verbal IQ scales of the WISC, and also had somewhat
higher scores on the Woodcock-Johnson Broad Math Achieve-
Early Postnatal Weight Gain, Intellectual Performance, and Body Mass Index at 7 Years of Age
in Term Infants with Intrauterine Growth Restriction
ment Test.6 Our study confirmed that slow postnatal growth in
the first 4 months after IUGR confers a dose-dependent risk to
cognitive function.
A unique finding of our study is that excessive postnatal
weight gain also confers a risk to the developing brain after
IUGR. The magnitude of this effect is smaller than that of
postnatal malnutrition, resulting in an inverted J-shaped risk
curve. IUGR, especially when followed by rapid postnatal weight
gain, also is associated with increased risk of subsequent coronary
events1,7,9 and diabetes mellitus with a similarly inverted J shaped
risk curve.28,29 The DOHaD hypothesis suggests that these
chronic disease outcomes may be due to early endocrinologic
alterations in response to incongruent prenatal and postnatal
metabolic milieus.30 Cortisol and proinflammatory cytokines are
proposed mediators of the relationship between rapid postnatal
growth and chronic disease.31 These factors are known to cross
the blood-brain barrier and affect the developing brain; for
example, animal models have demonstrated that excess cortisol
alters early hippocampal development and function,32 and that
proinflammatory cytokines damage oligodendrocytes and result
in leukomalacia.33 Although we cannot directly test this hypoth-
esis in our data set, we speculate that these factors may be
involved in mediating the reduced cognitive performance in the
infants with overly rapid postnatal weight gain. These findings
may have implications for other groups in which severe growth
restriction is followed by overfeeding and rapid catch-up growth,
including preterm infants requiring intensive care.
The timing, severity, and duration of an insult to the
developing brain determine the risk to future function and
explain regional variability in brain dysfunction after timed
insults.34 Consistent with the concept of regional brain effects
due to the timing of key developmental windows,27,33,35,36 the
effect of postnatal weight gain on later cognitive function is
not uniform and maps onto brain structures that are vulner-
able in the perinatal period.27 This finding of a heterogeneous
neurologic effect reinforces that the small effect that we
detected is indeed real. The cognitive scores that were inde-
pendently altered by early growth were verbal skills (informa-
tion and vocabulary) and working memory (digit span). In
addition, both the performance IQ and full-scale IQ at age 7
years were affected. Although full-scale IQ scores provide
only a broad assessment of cognitive development, individual
subscales reflect specific tasks implicating specific brain re-
gions.17 Verbal comprehension tasks include information, vo-
cabulary, and comprehension tests, which require both im-
mediate and long-term memory, as well as experience and
judgment. Tests of perceptual organization include picture
arrangement, block design, and object assembly using short-
term spatial, visual, and working memory; visuomotor orga-
nization; concept formation; and nonverbal reasoning. Digit
span tasks require verbal working memory and freedom from
distraction. Coding tests probe processing speed, which is
historically correlated with myelination. Even though these
tasks are not specific to just one brain region, they do suggest
involvement of such rapidly developing regions as the hip-
pocampus, basal ganglia, and prefrontal cortex, which are
205
potentially more vulnerable to insult in the late fetal and early
neonatal periods.27
The present analysis also had demonstrated that the
rate of early postnatal weight gain predicts BMI in childhood,
suggesting that the pathophysiologic mechanisms of the
DOHaD hypothesis are relevant to this population. BMI is
commonly used to monitor growth in children over age 2
years, providing a general assessment of body habitus or
weight in relation to linear growth. Excessive BMI carried
into adulthood confers an increased risk of metabolic syn-
drome,37 and infants with IUGR are themselves at increased
risk for metabolic syndrome and short stature.7,8 Prospective
studies are needed to delineate optimal early feeding practices
and ideal growth patterns, but absent these, avoiding excessive
weight gain in relationship to linear growth may be the most
prudent strategy. Although nutrients are essential for growth,
linear growth rate may be programmed or more highly gov-
erned by other factors, including genetics and endocrine (hy-
pothalamic-pituitary) function.
We have reported that there is an optimal rate of weight
gain, somewhat below the maximum, that optimizes cognitive
outcome. We emphasize that whether postnatal weight gain
per se is the causal factor is not clear. In addition, differences
exist between the CPP population and contemporary popu-
lations. In the study population, 96% of the infants were
formula-fed. Current breast-feeding practices likely provide a
more moderate rate of early weight gain, perhaps contributing
to its beneficial effects on neurodevelopment. However, the
4% of CPP infants who were breast-fed could not have
accounted for all of the differences in outcome seen. Other
potentially important differences during that era include a
different nutrient composition of formula, a general goal of a
20-pound weight gain in pregnancy, better adherence to
feeding schedules, and propping bottles in infants’ mouths.
Nevertheless, the implications of our data are potentially
far-reaching. Health care providers must be cognizant of the
effects of early growth, nutrition, and stress on the developing
brains as well as the cardiovascular systems of their patients
and realize that early nutritional practices, including excessive
early growth, may affect long-term brain function.
REFERENCES
1. Barker DJ. Fetal origins of coronary heart disease. BMJ 1995;311:171-4.
2. Barker DJ, Osmond C, Forsen TJ, Kajantie E, Eriksson JG. Trajectories of growth
among children who have coronary events as adults. N Engl J Med 2005;353:1802-9.
3. Dahlgren J, Boguszewski M, Rosberg S, Albertsson-Wikland K. Adrenal steroid
hormones in short children born small for gestational age. Clin Endocrinol (Oxf)
1998;49:353-61.
4. Ibanez L, Potau N, Francois I, de Zegher F. Precocious pubarche, hyperinsulin-
ism, and ovarian hyperandrogenism in girls: relation to reduced fetal growth. J Clin
Endocrinol Metab 1998;83:3558-62.
5. Voordouw JJ, van Weissenbruch MM, Delemarre-van de Waal HA. Intrauterine
growth retardation and puberty in girls. Twin Res 2001;4:299-306.
6. Casey PH, Whiteside-Mansell L, Barrett K, Bradley RH, Gargus R. Impact of
prenatal and/or postnatal growth problems in low birth weight preterm infants on
school-age outcomes: an 8-year longitudinal evaluation. Pediatrics 2006;118:1078-86.
7. Hales CN, Barker DJ. Type 2 (non–insulin-dependent) diabetes mellitus: the
thrifty phenotype hypothesis. Diabetologia 1992;35:595-601.
206 Pylipow et al
8. Hales CN, Barker DJ, Clark PM, Cox LJ, Fall C, Osmond C, et al. Fetal and
infant growth and impaired glucose tolerance at age 64. BMJ 1991;303:1019-22.
9. Lucas A. Long-term programming effects of early nutrition: implications for the
preterm infant. J Perinatol 2005;25(Suppl 2):S2-6.
10. Clark PM. Programming of the hypothalamo-pituitary-adrenal axis and the fetal
origins of adult disease hypothesis. Eur J Pediatr 1998;157(Suppl 1):S7-10.
11. Fowden AL, Forhead AJ. Endocrine mechanisms of intrauterine programming.
Reproduction 2004;127:515-26.
12. Abraham NG, Brunner EJ, Eriksson JW, Robertson RP. Metabolic syndrome:
psychosocial, neuroendocrine and classical risk factors in type 2 diabetes. Ann N Y Acad
Sci 2007;1113:256-75 .
13. Bartha JL, Romero-Carmona R, Comino-Delgado R. Inflammatory cytokines in
intrauterine growth retardation. Acta Obstet Gynecol Scand 2003;82:1099-102.
14. Feldhaus B, Dietzel ID, Heumann R, Berger R. Effects of interferon-gamma and
tumor necrosis factor-alpha on survival and differentiation of oligodendrocyte progen-
itors. J Soc Gynecol Investig 2004;11:89-96.
15. Lupien SJ, Maheu F, Tu M, Fiocco A, Schramek TE. The effects of stress and
stress hormones on human cognition: Implications for the field of brain and cognition.
Brain Cogn 2007;65:209-37.
16. Hardy JB. The Collaborative Perinatal Project: lessons and legacy. Ann Epidemiol
2003;13:303-11.
17. Sattler JM. Assessment of Children’s Intelligence. Philadelphia: WB Saunders; 1974.
18. Wechsler D. Manual for the Wechsler Scale of Children’s Intelligence, 3rd edition
(WISC-III). New York: The Psychological Corporation; 1991.
19. Goldenberg R, Hack M, Grantham M, Schurch B. Report of the IDECG/IUNS
Working Group on IUGR effects on neurological, sensory, cognitive, and behavioral
function. Eur J Clin Nutr 1998;52(Suppl 1):S100-1.
20. Myrianthopoulos NC, French KS. An application of the US Bureau of the Census
socioeconomic index to a large, diversified patient population. Soc Sci Med 1968;2:283-99.
21. O’Keeffe MJ, O’Callaghan M, Williams GM, Najman JM, Bor W. Learning,
cognitive, and attentional problems in adolescents born small for gestational age.
Pediatrics 2003;112:301-7.
22. Strauss RS, Dietz WH. Growth and development of term children born with low
birth weight: effects of genetic and environmental factors. J Pediatr 1998;133:67-72.
23. Tolsa CB, Zimine S, Warfield SK, Freschi M, Sancho Rossignol A, Lazeyras F,
et al. Early alteration of structural and functional brain development in premature infants
born with intrauterine growth restriction. Pediatr Res 2004;56:132-8.
24. Black LS, deRegnier RA, Long J, Georgieff MK, Nelson CA. Electrographic
imaging of recognition memory in 34-38 week gestation intrauterine growth–restricted
newborns. Exp Neurol 2004;190(Suppl 1):S72-83.
25. Lee KH, Calikoglu AS, Ye P, D’Ercole AJ. Insulin-like growth factor I (IGF-I)
ameliorates and IGF binding protein 1 (IGFBP-1) exacerbates the effects of undernu-
trition on brain growth during early postnatal life: studies in IGF-I and IGFBP-1
transgenic mice. Pediatr Res 1999;45:331-6.
26. Lister JP, Blatt GJ, DeBassio WA, Kemper TL, Tonkiss J, Galler JR, et al. Effect
of prenatal protein malnutrition on numbers of neurons in the principal cell layers of the
adult rat hippocampal formation. Hippocampus 2005;15:393-403.
27. Rao R, Georgieff MK. The role of nutrition in cognitive development. In: Nelson
CA, Luciana M, editors. Handbook in Developmental Cognitive Neuroscience. Cam-
bridge, MA: MIT Press; 2000. p. 491-504.
28. McCance DR, Pettitt DJ, Hanson RL, Jacobsson LT, Knowler WC, Bennett PH.
Birth weight and non–insulin-dependent diabetes: thrifty genotype, thrifty phenotype,
or surviving small baby genotype? BMJ 1994;308:942-5.
29. Regan FM, Cutfield WS, Jefferies C, Robinson E, Hofman PL. The impact of
early nutrition in premature infants on later childhood insulin sensitivity and growth.
Pediatrics 2006;118:1943-9.
30. Gluckman PD, Hanson MA. Developmental origins of disease paradigm: a
mechanistic and evolutionary perspective. Pediatr Res 2004;56:311-7.
31. Jones A, Godfrey KM, Wood P, Osmond C, Goulden P, Phillips DI. Fetal
growth and the adrenocortical response to psychological stress. J Clin Endocrinol Metab
2006;91:1868-71.
32. Setiawan E, Jackson MF, Macdonald JF, Matthews SG. Effects of repeated
prenatal glucocorticoid exposure on long-term potentiation in the juvenile guinea pig
hippocampus. J Physiol 2007;581(Pt 3):1033-42.
33. Folkerth RD. Periventricular leukomalacia: overview and recent findings. Pediatr
Dev Pathol 2006;9:3-13.
34. Kretchmer N, Beard JL, Carlson S. The role of nutrition in the development of
normal cognition. Am J Clin Nutr 1996;63:997S-1001S.
35. Godfrey KM, Barker DJ. Fetal nutrition and adult disease. Am J Clin Nutr
2000;71(5 Suppl):1344S-52S.
36. Rao R, Georgieff MK. Neonatal iron nutrition. Semin Neonatol 2001;6:425-35.
37. Ong KK. Size at birth, postnatal growth and risk of obesity. Horm Res 2006;
65(Suppl 3):65-9.
The Journal of Pediatrics • February 2009