Content uploaded by Marion Taine
Author content
All content in this area was uploaded by Marion Taine on Jun 04, 2020
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=iahb20
Annals of Human Biology
ISSN: 0301-4460 (Print) 1464-5033 (Online) Journal homepage: https://www.tandfonline.com/loi/iahb20
Does cord blood leptin level mediate the
association between neonatal body size and
postnatal growth? Results from the EDEN
mother–child cohort study
Marion Taine, Olfa Khalfallah, Anne Forhan, Nicolas Glaichenhaus, Marie-
Aline Charles & Barbara Heude
To cite this article: Marion Taine, Olfa Khalfallah, Anne Forhan, Nicolas Glaichenhaus, Marie-Aline
Charles & Barbara Heude (2020) Does cord blood leptin level mediate the association between
neonatal body size and postnatal growth? Results from the EDEN mother–child cohort study,
Annals of Human Biology, 47:2, 159-165, DOI: 10.1080/03014460.2020.1748712
To link to this article: https://doi.org/10.1080/03014460.2020.1748712
View supplementary material
Published online: 26 Apr 2020.
Submit your article to this journal
Article views: 35
View related articles
View Crossmark data
RESEARCH PAPER
Does cord blood leptin level mediate the association between neonatal body
size and postnatal growth? Results from the EDEN mother–child cohort study
Marion Taine
a,b
, Olfa Khalfallah
c
, Anne Forhan
a
, Nicolas Glaichenhaus
c
, Marie-Aline Charles
a,d
and
Barbara Heude
a
a
Centre for Research in Epidemiology and Statistics, INSERM, Universit
e de Paris, INRAe, Paris, France;
b
Department of Paediatric
endocrinology, Necker Enfants Malades Hospital, Paris, France;
c
Institut de Pharmacologie Mol
eculaire et Cellulaire, CNRS, INSERM,
Universit
e de Nice-Sophia Antipolis, UMR7275, UMR_S, Valbonne, France;
d
Unit
e mixte Inserm-Ined-EFS ELFE, Paris, France
ABSTRACT
Background: Leptin is potentially involved in the correction of early postnatal growth of infants hav-
ing deviated from their genetic trajectory in utero.
Aim: To analyse the potential mediating role of cord blood leptin level in the association between
neonatal anthropometry and early postnatal growth in the mother–child EDEN cohort.
Subjects and methods: We included term newborns with information on leptin, birth weight and
length, and weight and length SD-score changes over the first 2 months. The Baron and Kenny
method was used to quantify the mediation contribution of leptin in the association between neo-
natal anthropometry and postnatal growth, considering several confounders. Analyses were stratified
to consider sexual dimorphism.
Results: A 1 SD higher birth weight was associated with a lower 2-months weight variation of 0.27
(0.18; 0.36) SD and a 0.16 (0.06; 0.26) SD, in boys and girls, respectively. Leptin explained 20% and
25% of these associations, respectively. Leptin did not mediate the association between birth length
and birth-to-2 months length variation.
Conclusion: Our results suggest that cord blood leptin may not be involved in the negative associ-
ation between birth length and postnatal length growth but may play a modest mediating role in
early postnatal catch-up or catch-down in weight.
ARTICLE HISTORY
Received 23 September 2019
Revised 13 March 2020
Accepted 18 March 2020
KEYWORDS
Cord blood leptin; neonatal
body size; weight-for-
length; catch-up growth;
mediation analysis
Introduction
Growth in the perinatal period has attracted increasing inter-
est from clinicians and epidemiologists (Fenton and Kim
2013; Gat-Yablonski and Phillip 2015). In utero, certain situa-
tions of deficiency (Brodsky and Christou 2004) or nutritional
excess (Zhang et al. 2018) lead to growth deviations from
the genetic foetal growth trajectory and are associated with
changes in foetal metabolism. This foetal metabolic program-
ming (Godfrey and Barker 2001) as well as the genetic influ-
ence on physical growth (Prader 1978; Jelenkovic et al. 2016)
allow the infant to reach, at least in part, his/her genetic
growth trajectory in the first year of life (Claris et al. 2010).
This phenomenon is well illustrated by the catch-up of post-
natal growth in infants born with intrauterine growth retard-
ation (Boersma and Wit 1997) or by the initial growth
deceleration in infants born with macrosomia (Hediger et al.
1998; Regnault et al. 2011).
These physiological mechanisms deserve to be better
understood (Gat-Yablonski and Phillip 2015). If several hor-
mones seem to be involved, the importance of their role to
help the infant to reach his/her growth potential has not
been assessed yet (Gat-Yablonski and Phillip 2015). Several
studies have shown high birth weight associated with high
concentrations of certain cord blood hormones such as lep-
tin, insulin-like growth factor 1, and c-peptide (Leger et al.
1996; Ong et al. 1999; Verkauskiene et al. 2007; Mantzoros
et al. 2009; Karakosta et al. 2011). Conversely, low birth
weight was associated with low concentrations of these
three hormones (Leger et al. 1996; Ong et al. 1999;
Verkauskiene et al. 2007; Mantzoros et al. 2009; Karakosta
et al. 2011). Insulin-like growth factor 1 is one of the main
growth factors (Laron 2001); its low levels in newborns with
intrauterine growth retardation cannot explain a growth
acceleration over the first months of life (Leger et al. 1996).
However, low cord blood leptin level could be responsible
for such early growth acceleration in newborns with intra-
uterine growth retardation and early deceleration in new-
borns with macrosomia.
Leptin is mainly secreted by white adipose tissue as early
as 19 weeks of gestation (Cetin et al. 2000) and has an anor-
exic function after birth (Auwerx and Staels 1998; Friedman
2009). It binds to its receptors in the hypothalamic arched
nucleus and regulates appetite by stimulating the secretion
CONTACT Marion Taine mariontaine@gmail.com, Inserm UMR1153 Centre for Research in Epidemiology and Statistics (CRESS), EAROH team, B^
atiment
15-16, 16 av Paul Vaillant Couturier Villejuif 94807, France
Supplemental data for this article are available online at https://doi.org/10.1080/03014460.2020.1748712.
ß2020 Informa UK Limited, trading as Taylor & Francis Group
ANNALS OF HUMAN BIOLOGY
2020, VOL. 47, NO. 2, 159–165
https://doi.org/10.1080/03014460.2020.1748712
of anorexigenic peptides (Auwerx and Staels 1998; Bouret
and Simerly 2006). Neonates with higher body weight will
have higher leptin concentrations at birth and feel more sati-
ety. Reciprocally, infants with lower fat mass and lower leptin
concentrations would experience lower satiety and thus
increase their nutritional intake (Auwerx and Staels 1998;
Friedman 2009). This situation would lead to an acceleration
of early postnatal growth compared with other children. This
hypothesis is consistent with previous results by Ong et al.’s
observational study (Ong et al. 1999) and other studies
(Mantzoros et al. 2009; Popova et al. 2018; Yeung et al. 2018)
showing that low cord blood leptin level could predict
greater weight gain in the first months of life as compared
to other children.
In this context, we hypothesised that cord blood leptin
level could be responsible for part of the growth variability
in the first 2 months of life. The objective of this work was
to analyse, in the mother-child EDEN cohort, the potential
mediating role of leptin concentration in cord blood in the
association between neonatal anthropometry and early post-
natal growth.
Subjects and methods
Study design and population
The mother-child EDEN cohort was designed to assess the pre-
and postnatal determinants of children’s growth, health, and
development (Heude et al. 2016). Study participation was pro-
posed to all women visiting the prenatal clinic before 24
weeks. Exclusion criteria were multiple pregnancies, known dia-
betes before pregnancy, French illiteracy or planning to move
out of the region within the next 3 years. Among the 3758
women invited to participate, 2002 (53%) were enrolled in the
study (1034 women from Nancy and 968 from Poitiers).
Recruitment extended from 2003 to 2006. Women were
included at 15 weeks on average (range: 8–26). As compared
with the 2003 French National Perinatal Survey (Enqu^
ete
Nationale P
erinatale [ENP]), a national sample of births (Blondel
et al. 2012), women included in EDEN and still followed up at
delivery had a higher level of education (results not shown);
however, percentages of preterm births or admissions of the
newborn to a neonatal unit were similar.
Written consent was obtained from both parents for inclu-
sion of the newborn at study inclusion and after delivery.
The study received approval from the ethics committee
(CCPPRB) of Kremlin-Bic^
etre on 12 December 2002 and from
CNIL (Commission Nationale Informatique et Libert
e), the
French data privacy institution.
Overall, 107 preterm infants were excluded, leaving 1366
full-term infants with cord blood leptin level, neonatal
anthropometry and 2-month growth data available
for analysis.
Data collection
Mothers were followed up from inclusion to delivery, and
offspring were followed up postnatally by visits to research
centres and questionnaires mailed to parents. Clinical exami-
nations were conducted by specifically trained midwives for
the mothers at 26 weeks on average and for the newborns
at birth. To collect further information on pregnancy, parents
and their lifestyle, mothers were interviewed by using stand-
ardised questionnaires during pregnancy and at delivery.
Anthropometric data from the parents were collected, includ-
ing their own birth weight and adult height and weight
(weight before pregnancy reported by the mother). The
mothers also reported the number of cigarettes smoked per
day during the pregnancy. Other information about preg-
nancy and the newborn was collected from obstetric records:
parity (primiparous/multiparous), mother’s age at delivery,
blood glucose 1 hour after glucose administration by the
O’Sullivan test, the presence of treated hypertension during
pregnancy, mother’s gestational weight gain, child’s sex, ges-
tational age at birth, and birth weight and length. At birth,
the placenta was weighed.
The child was examined clinically four times: at birth, at
age 1 year and 3 years, and during the 6th year (at 5 years
and 8 months, on average). Concomitantly, and inbetween
clinical visits, mothers and fathers completed questionnaires
about their offspring, themselves and their household that
were self-administered or administered by midwives.
Questionnaires were completed twice during the child’s first
year (at age 4 and 8 months), at 1 year and every year there-
after until age 5 to 6 years. Consequently, from birth to age
5 to 6 years, children had 22 weight measurements, on aver-
age (interquartile range 16–26).
Biological samples were collected from the mother during
pregnancy and at delivery and from the child (or cord) at
birth. Leptin and c-peptide assays were performed in 2018
on all cryopreserved cord serum samples by using an
immunoassay multiplexed sandwich technique at the Centre
National de Recherche Scientifique in Nice. C-peptide level is
considered an excellent reflection of insulin level and more
stable than insulin (Retnakaran et al. 2019). For the sake of
simplicity, cord blood leptin, or cord blood leptin level will
be referred to as cord leptin in this study.
Preliminary statistical treatment of variables of interest
Because the distributions of cord leptin and c-peptide levels
were skewed, logarithmic decimal transformations of these
variables were preliminarily performed.
Individual growth curves for weight and length from
birth to 5–6 years were obtained by using the Jenss-Bayley
growth curve model (Botton et al. 2014). The Jenss-Bayley
model fits growth from birth to age 8 years, before growth
starts to accelerate again (start of puberty). One part of the
model is an exponential term that fits weight growth decel-
eration during the first 2 years after birth and the other part
is the equation of a straight line that can describe growth
when it has a linear trend (from about age 2 to 8 years). We
used mixed-effects nonlinear models to assess individual
weight growth curves with the SAEMIX package in R. This
method can provide many individual characteristics of the
growth curve, including predictions of weight and length. It
160 M. TAINE ET AL.
allows for taking repeated measurements and missing data
into account and does not require the same number of
measurements or the same ages at measurement for all
individuals. Of note, the Jenss-Bayley model necessarily has
a monotonic shape, which is not compatible with neonatal
weight loss observed for most individual weight growth
curves (Macdonald et al. 2003). Therefore, we did not
include birth weight or any of the values of weight
recorded during the first 3 days in the dataset for this
model. In the context of the present study, this growth
modelling was used to predict neonatal length, weight at
4 days of life and weight and length at two months of age.
Indeed, we preferred to use predicted length at birth
instead of measured birth length because of the lack of
accuracy of this measurement at birth. For each anthropo-
metric parameter, we calculated SD scores based on study
means and SDs at birth, 4 days and 2 months of age, strati-
fying by sex. We calculated the change in SD-score between
birth and 2 months for predicted length and between
4 days of life and 2 months for predicted weight as previ-
ously described. For simplicity, these differences in SD-score
will be named early weight or length changes in the rest of
the manuscript.
The body mass index (BMI) of the pre-pregnant mother
and the father were obtained by dividing the weight (in kilo-
grams) by the squared height (in metres).
Statistical analysis
First, we compared participants included or not in the ana-
lysis by using Student’st-test or chi-square tests as appropri-
ate. We also used Student’st-test to compare
anthropometric parameters and cord leptin and c-peptide
level between boys and girls. All the analyses described here-
after were performed separately in girls and boys to consider
a potential sexual dimorphism (Ashley-Martin et al. 2020).
We investigated the correlations between cord leptin,
anthropometry at birth, 4-day weight, 2-month anthropom-
etry, and early weight and length changes using partial
Pearson’s correlation coefficients after a-minima adjustment
for gestational age and inclusion centre. Then, we used the
Baron and Kenny method (Baron and Kenny 1986) to quan-
tify the mediation contribution of cord leptin in the associ-
ation between neonatal body size and 2-month changes.
As illustrated in Figure 1, the association between neo-
natal body size and 2-month changes corresponds to the
total effect (path c). The direct effect, or non-mediated effect
(path c0) is the effect of neonatal size on 2-month changes
via causal pathways other than cord leptin, while the indirect
effect is the effect that operates via the effect of neonatal
size on cord leptin. The coefficient b relates the mediator to
the dependent variable adjusted for the independent vari-
able. The product ab corresponds to the mediation, or indir-
ect effect. For 95% standard normal confidence limits of the
indirect effect, a critical value of 1.96 was used for the stand-
ard error. All the coefficients are standardised. We assessed
the percentage of mediation as the ratio between the indir-
ect (ab) and the total effect (c ¼c0þab). The same method
was applied to the study of length.
All these associations were adjusted for the follow-
ing factors:
Parental factors: birth weight and adult length, paternal
and maternal pre-gestational BMI
Pregnancy-related factors: parity, maternal age at the
beginning of pregnancy, glucose level at 1-h post-charge
from the O’Sullivan test, treated hypertension during
pregnancy, gestational weight gain, number of cigarettes
smoked during the last trimester of pregnancy
Neonatal factors: sex, gestational age at birth, cord blood
c-peptide level and placental weight at delivery
For multivariable analyses, missing data on adjustment
variables were imputed by using the chained-equation
Cord blood lepn
Birth to 2 months size
variaon
c’
c
a b
Neonatal anthropometry
Figure 1. Mediation analysis diagram. (a) association between neonatal anthro-
pometry and the potential mediator: cord leptin; (b) association between cord
leptin and the outcome size growth variation (ab) will quantify the indirect
effect; (c’) is the direct effect; (c) is the total effect: overall association between
neonatal anthropometry and size variation.
Table 1. Study population characteristics (N¼1366).
Maternal characteristics
Age (years) 29 ± 4.9
Smoking during pregnancy 26.5 (353)
Primiparity 43.4 (592)
Height (cm) 163.7 ± 6
BMI (kg/m
2
) 23.1 ± 4.5
Gestational diabetes 5.9 (80)
Gestational hypertension requiring treatment 2.1 (29)
Gestational weight gain (kg) 9.1 ± 5.2
Type of delivery –
Vaginal 77.0 (1049)
Instrumental vaginal 10.3 (142)
Caesarian 12.7 (175)
Placental weight (g) 539.5 (115)
Infant characteristics
Girl 48.0 (657)
Gestational age (weeks) 39.5 ± 1.2
Birth weight <10th percentile 11.6 (155)
Birth weight >90th percentile 8.5 (114)
Infant characteristics by gender Girls (N¼657) Boys (N¼709)
Birth weight (g) 3269 (403) 3416 (449)
4-day weight (g) 3107 (318) 3273 (363)
Birth length (cm) 49.8 (1.8) 50.8 (1.8)
2-month weight (g) 4878 (489) 5297 (513)
2-month length (cm) 56.1 (1.7) 57.7 (1.8)
Cord blood leptin level (ng/ml) 10.5 (6.0–18.2) 5.4 (3.0–10.9)
Cord blood c-peptide level (ng/ml) 0.61 (0.41–0.90) 0.56 (0.39–0.85)
Data are mean ± SD or % (n) or mean (Q1–Q3) for hormones levels
BMI, body mass index
ANNALS OF HUMAN BIOLOGY 161
multiple imputation method according to Rubin’s rules
(Sterne et al. 2009). We generated 10 independent imputed
tables with 10 iterations each. The missing values pattern is
described in Supplementary Table 1.
Statistical analyses were conducted using SAS 9.3 (SAS
Inst., Inc., Cary, NC).
Results
As shown in Supplementary Table 2, mothers of newborns
not included in the analysis were shorter (162.8 vs 163.7 cm,
p¼0.008), had a lower level of education (48% vs 56%
of >A-level, p¼0.006) and tended to have a higher BMI
(23.6 vs 23.1 kg/m
2
,p<0.07) and more small-for-gestational-
age neonates (15.6% vs 11.6%, p¼0.06) than their counter-
parts. Among the 1366 infants included, 657 (48%) were girls
(Table 1).
As shown in Table 1, boys had lower cord leptin and cord
blood c-peptide level and higher neonatal or 2 months
weight and length than girls. Early weight and length
changes were null on average for both girls and boys.
Table 2 describes correlations between anthropometric
parameters and cord leptin by sex, after a-minima adjust-
ment for gestational age and inclusion centre. Correlations
between cord leptin and weight, positive and strong at birth,
were much lower at two months of age (from 0.41,
p<0.0001 with birth weight to 0.13, p<0.001 with 2-month
weight in girls and from 0.42, p<0.0001 to 0.07, p¼0.07 in
boys). Correlation of cord leptin with birth length was weak
and also lower at 2 months (from 0.20, p<0.0001 with birth
length to 0.11, p<0.05 with 2-month length in girls and
from 0.16, p<0.0001 with birth length to 0.09, p<0.05 with
2-month length in boys). Cord leptin was negatively and
moderately correlated with early weight and length changes:
the strongest correlation was the one with early weight
change in boys (r¼0.28, p<0.0001 in boys vs 0.19,
p<0.0001 in girls) and the weakest ones with the early
length change regardless the sex (r¼0.12 in boys vs
r¼0.11 in girls). Birth weight and length negatively and
moderately correlated with their early change in SD score
(r¼0.28, p<0.0001 in boys vs r¼0.20, p<0.0001 in girls
for weight and r¼0.21, p<0.0001 in boys vs r¼0.18,
p<0.0001 in girls for length).
Results of the mediation analyses of the association
between neonatal body size and early changes in weight
and length through cord leptin are displayed in Figure 2.
Total, direct and indirect effects were all negative for both
sexes. In boys, a birth weight increase of 1 SD was associated
with a decrease of 0.27 (95%IC 0.18;0.36) SD of early weight
change. The direct effect was 0.21 (95%IC 0.31; 0.12)
and the indirect effect via cord leptin was 0.06 (95%IC
0.09; 0.02). In girls, a birth weight increase of 1SD was
associated with a decrease of 0.16 (95%IC 0.06;0.26) SD of
early weight change. The direct effect was 0.12 (95%IC
0.23; 0.01) and the indirect effect via cord leptin was
0.04 (95%IC 0.07; 0.01). Therefore, 20% and 25% of the
association between birth weight early weight change was
mediated by cord leptin in boys and girls, respectively.
Results of the mediation analysis related to length growth
were very similar in girls and boys (Figure 2) and showed
that the indirect effects were close to zero, suggesting no
mediating role of cord leptin in the association between
birth length and early length change.
Discussion
This study in the EDEN cohort showed that cord blood leptin
level was positively and strongly correlated with neonatal
anthropometry. These two factors were negatively and mod-
erately correlated with change in SD score of weight and
length over the first 2 months of life. Leptin explained 20%
and 25% of the total effect of birth weight on early weight
change, in girls and boys, respectively, but did not mediate
the association between birth length and early
length change.
As established in the literature, we found that cord leptin
(Matsuda et al. 1997) (Ong et al. JCEM 1999), c-peptide levels
(Shields et al. 2007) and neonatal anthropometric parameters
differed by sex. C-peptide level is an excellent reflection of
insulin level (Retnakaran et al. 2019), which plays a major
role as a growth factor in prenatal life (Verhaeghe et al.
1993). Girls had a higher level of neonatal c-peptide than
boys, but paradoxically, their birth weight and length were
lower. These results support the hypothesis of some insulin
resistance from the prenatal period in girls (Verhaeghe et al.
1993). Indeed, insulin level sex difference is transmitted to
leptin as insulin modulates the expression and secretion of
leptin in circulating blood (Kim-Motoyama et al. 1997). The
transient increase in sex hormones before birth also might
contribute to this sex difference of leptin level (Ong et al.
1999), a stimulating role played by oestrogens (Demerath
et al. 1999) and an inhibiting role played by androgens
(Demerath et al. 1999; Isidori et al. 1999) having been
observed on circulating leptin concentrations in a study car-
ried out on transsexuals (Elbers et al. 1997).
Our study showed that, as expected, birth weight and
height were negatively correlated with their corresponding
SD-scores changes over the first two months of life. These
results support the observation of Hediger et al. (Hediger
et al. 1998): infants born with lower birth weight, especially
due to intra-uterine growth restriction, have a period of
catch-up growth in early infancy; infants with higher birth
weight, due to abnormalities of maternal glucose metabolism
or gestational diabetes mellitus, slow down in growth. The
magnitude of these negative correlations (r¼0.20,
p<0.0001 in girls and r¼0.28, p<0.0001 in boys) were
weaker than the one found by Hemachandra et al. (r¼0.40,
p<0.0001) (Hemachandra et al. 2007). This could be partly
explained by the fact that we considered the predicted
weight on the 4th day of life in the difference and not the
measured birth weight. This allowed limiting the regression
to the mean issue, by avoiding having the same measure-
ment error (related to birthweight) in both terms of the cor-
relation (Cameron et al. 2005). Also, focussing on the nadir
of neonatal weight on the 4th day of life, allowed for a more
162 M. TAINE ET AL.
relevant initial value to investigate the change of SD score,
especially for catch-up growth.
After adjustment for numerous confounders, Figure 2
shows a strengthened difference between sexes of this
inverse association with total effects twice as high in boys
than in girls: a 1 SD higher birth weight was associated with
the decrease of weight variation over the 2 months of 0.27
(0.18;0.36) SD in boys vs 0.16 (0.06;0.26) SD in girls, suggest-
ing a sexual dimorphism of the association.
Cord blood leptin seemed to play a modest mediating
role in the total effect observed for weight: for a 1 SD
increase in birth weight, cord blood leptin level was respon-
sible for a 0.06 SD decrease in weight variation over the 2
months in boys and 0.04 SD in girls. Regarding length, our
results suggest that cord leptin does not play any mediating
role in the observed association.
Several other hormones besides leptin might be involved
in regulating early postnatal growth and explain the sexual
dimorphism found for weight (Saad et al. 1997; Dearden
et al. 2018). The hormones regulating hunger and satiety are
particularly interesting because growth is under nutritional
influence during the first months of life (Victora et al. 2008).
The sex hormones potently control food intake and body
weight (Asarian and Geary 2013). In general, male animals
consume more food than females because oestrogen exerts
an inhibitory effect on meal size and daily food intake
(Asarian and Geary 2006; Chen et al. 2015). Ghrelin also has
a very interesting profile by being the direct cause of hunger
sensation and could be responsible for sexual dimorphism
(Dearden et al. 2018). Its concentration is inversely correlated
with birth weight (Onal et al. 2004) and seems to be a pre-
dictor of weight gain in the first year of life (Gohlke
Table 2. Pearson’s partial correlation coefficients between anthropometric parameters and cord blood leptin, adjusted for centre and gestational age, for girls
(below left triangle (dark grey), N¼657) and boys (upper right triangle (light grey), N¼709).
Birth weight 4-day weight 2-month weight Early weight change
a
Birth length 2-month length Early length change
b
Cord leptin
c
Birth weight 0.84 0.60 0.28 0.67 0.61 0.07 0.42
4-day weight 0.80 0.69 0.36 0.67 0.65 0.02 0.29
2-month weight 0.65 0.72 0.42 0.53 0.66 0.32 0.07
Early weight change
a
0.20 0.37 0.38 0.15 0.07 0.39 0.28
Birth length 0.67 0.63 0.52 0.13 0.84 0.21 0.16
2-month length 0.63 0.61 0.66 0.07 0.86 0.35 0.09
Early length change
b
0.02 0.03 0.32 0.38 0.18 0.35 0.12
Cord leptin
c
0.41 0.27 0.13 0.19 0.20 0.11 0.11
All partial correlations pValues under 0.08 had a related pValues >0.05
a
Change in predicted weight SD scores between 4 days and 2 months.
b
Change in predicted length SD scores between birth and 2 months.
c
Cord blood leptin level.
−0.158
−0.118
−0.039
−0.400
−0.350
−0.300
−0.250
−0.200
−0.150
−0.100
−0.050
0.000
0.050
−0.273
−0.217
−0.056
−0.400
−0.350
−0.300
−0.250
−0.200
−0.150
−0.100
−0.050
0.000
0.050
−0.400
−0.350
−0.300
−0.250
−0.200
−0.150
−0.100
−0.050
0.000
0.050
−0.400
−0.350
−0.300
−0.250
−0.200
−0.150
−0.100
−0.050
0.000
0.050
−0.153 −0.148
−0.004
−0.183
−0.176
−0.006
Weight Δ (SD) Weight Δ (SD)
Length Δ (SD) Length Δ (SD)
Figure 2. Mediation analysis through cord leptin of the association between birth anthropometry and early change in weight (upper) and length (lower) in girls
(left) and boys (right). Effect size b(95% CI). Total effect in black, direct effect in white, indirect effect in grey.
All models were adjusted for parental, pregnancy-related and neonatal factors.
ANNALS OF HUMAN BIOLOGY 163
et al. 2005). IGF1, a main growth factor, plays a major role in
catch-up growth (Rosenbloom 2007). Nevertheless, in neo-
nates born with intrauterine restriction, its low value at birth
cannot explain the inverse correlation observed between
neonatal birth size and SD-score variation in the first 2
months. We can assume that nutritional hormones act by
promoting feeding to contribute to correcting the under-
nourishment of infants born with such conditions, and cor-
rect the low values of IGF1 (Leger et al. 1996). The quick
increase of IGF1 value, though still low at 4 months of age,
in infants born with intrauterine growth restriction (Leger
et al. 1996), might contribute to the genetic growth potential
catch-up. In the EDEN cohort, we did not have measure-
ments of ghrelin, sex hormones or several IGF1 measure-
ments in the first months of life, which explains why we only
tested the leptin pathway hypothesis.
This study has several limitations. First, we investigated
weight as a proxy of adiposity (Chen et al. 2018). Other
markers more specific to adiposity would have been more
relevant for investigating the effect of leptin, but were not
collected in the EDEN cohort. Second, cord leptin reflected
not only foetal leptin but also placental leptin. However, the
short half-life of placental leptin makes its interference with
anthropometric data at 2 months of life unlikely (Cumin
et al. 1997). The third limitation was selection bias, limiting
the study population to children with a cord blood sample
collected at birth. In our sample, mothers had fewer compli-
cations during pregnancy than the original population and
small-for-gestational-age newborns were less represented.
In addition to the originality of the topic, this study has
several strengths. As far as we know, the EDEN cohort has
the largest sample of leptin in cord blood. In addition, the
modelling of growth provided values for each child for the
same time, which was particularly suited to the study of early
postnatal growth. Also, a wide range of confounding factors
were considered in the mediation analysis, which limits the
potential confounding bias of the study.
Conclusion
Our results suggest that cord blood leptin level mediated a
modest part of the negative association between birth
weight and weight variation over the first 2 months of life
but did not mediate the similar association observed for
length. Other hormones besides leptin may contribute to
help infants to achieve their genetic growth potential.
Further mediation analyses focussing on these candidate hor-
mones need to be replicated to better understand the
physiological mechanism regulating early postnatal growth.
Funding
French Nutrition Society; National Institute for Research in Public Health
(IRESP TGIR Cohorte Sant
e 2008 Programme); National Agency for
Research (ANR nonthematic programme); French Speaking Association
for the Study of Diabetes and Metabolism (Alfediam); Mutuelle G
en
erale
de l’
Education Nationale; Nestl
e; French National Institute for Health
Education (INPES); Paris-Sud University; French National Institute for
Population Health Surveillance (INVS); French Agency for Environment
Security (AFFSET); French Ministry of Health Perinatal Programme;
INSERM Nutrition Research Programme; Institut F
ed
eratif de Recherche
and Cohort Programme; French Ministry of Research; Fondation pour la
Recherche M
edicale
Acknowledgements
The authors are extremely grateful to all the families who took part in
this study, the midwives and psychologists who recruited and followed
them, and the whole EDEN team, including research scientists, engi-
neers, technicians and managers and especially Josiane Sahuquillo and
Edith Lesieux for their commitment and their role in the success of the
study. The authors also acknowledge the commitment of the members
of the EDEN Mother-Child Cohort Study Group: I. Annesi-Maesano, J.Y.
Bernard, J. Botton, M.A. Charles, P. Dargent-Molina, B. de Lauzon-Guillain,
P. Ducimeti
ere, M. de Agostini, B. Foliguet, A. Forhan, X. Fritel, A. Germa,
V. Goua, R. Hankard, B. Heude, M. Kaminski, B. Larroque†, N. Lelong, J.
Lepeule, G. Magnin, L. Marchand, C. Nabet, F Pierre, R. Slama, M.J.
Saurel-Cubizolles, M. Schweitzer, O. Thiebaugeorges.
Disclosure statement
The authors have no conflicts of interest relevant to this article
to disclose.
Funding
This work was supported by the Soci
et
e Franc¸aise de Nutrition.
References
Asarian L, Geary N. 2006. Modulation of appetite by gonadal steroid hor-
mones. Phil Trans R Soc B. 361(1471):1251–1263.
Asarian L, Geary N. 2013. Sex differences in the physiology of eating. Am
J Physiol Regul Integr Comp Physiol. 305(11):R1215–1267.
Ashley-Martin J, Karaceper M, Dodds L, Arbuckle TE, Ettinger AS, Fraser
WD, Muckle G, et al. 2020. An examination of sex differences in asso-
ciations between cord blood adipokines and childhood adiposity.
Pediatr Obes. 15(3):e12587.
Auwerx J, Staels B. 1998. Leptin. Leptin. Lancet. 351(9104):737–742.
Baron RM, Kenny DA. 1986. The moderator-mediator variable distinction
in social psychological research: conceptual, strategic, and statistical
considerations. J Pers Soc Psychol. 51(6):1173–1182.
Blondel B, Lelong N, Kermarrec M, Goffinet F, National CG,; of the
National Perinatal Surveys. 2012. Trends in perinatal health in France
from 1995 to 2010. Results from the French National Perinatal
Surveys. J Gynecol Obstet Biol Reprod (Paris). 41(4):e1–e15.
Boersma B, Wit JM. 1997. Catch-up growth. Endocr Rev. 18(5):646–661.
Botton J, Scherdel P, Regnault N, Heude B, Charles MA, Group E-C.;EDEN
Mother-Child Cohort Study Group. 2014. Postnatal weight and height
growth modeling and prediction of body mass index as a function of
time for the study of growth determinants. Ann Nutr Metab. 65(2–3):
156–166.
Bouret SG, Simerly RB. 2006. Developmental programming of hypothal-
amic feeding circuits. Clin Genet. 70(4):295–301.
Brodsky D, Christou H. 2004. Current concepts in intrauterine growth
restriction. J Intensive Care Med. 19(6):307–319.
Cameron N, Preece MA, Cole TJ. 2005. Catch-up growth or regression to
the mean? Recovery from stunting revisited. Am J Hum Biol. 17(4):
412–417.
Cetin I, Morpurgo PS, Radaelli T, Taricco E, Cortelazzi D, Bellotti M, Pardi
G, et al. 2000. Fetal plasma leptin concentrations: relationship with
different intrauterine growth patterns from19 weeks to term. Pediatr
Res. 48(5):646–651.
Chen LW, Tint MT, Fortier MV, Aris IM, Shek LP, Tan KH, Chan SY, et al.
2018. Which anthropometric measures best reflect neonatal adiposity?
Int J Obes. 42(3):501–506.
164 M. TAINE ET AL.
Chen X, Wang L, Loh DH, Colwell CS, Tache Y, Reue K, Arnold AP. 2015.
Sex differences in diurnal rhythms of food intake in mice caused by
gonadal hormones and complement of sex chromosomes. Horm
Behav. 75:55–63.
Claris O, Beltrand J, Levy-Marchal C. 2010. Consequences of intrauterine
growth and early neonatal catch-up growth. Semin Perinatol. 34(3):
207–210.
Cumin F, Baum H-P, de Gasparo M, Levens N. 1997. Leptin is cleared
from the circulation primarily by the kidney. Int J Obes. 21 (6):
495–1126.
Dearden L, Bouret SG, Ozanne SE. 2018. Sex and gender differences in
developmental programming of metabolism. Mol Metab. 15:8–19.
Demerath EW, Towne B, Wisemandle W, Blangero J, Chumlea WC,
Siervogel RM. 1999. Serum leptin concentration, body composition,
and gonadal hormones during puberty. Int J Obes. 23(7):678–685.
Elbers JM, Asscheman H, Seidell JC, Frolich M, Meinders AE, Gooren LJ.
1997. Reversal of the sex difference in serum leptin levels upon cross-
sex hormone administration in transsexuals. J Clin Endocrinol Metab.
82(10):3267–3270.
Fenton TR, Kim JH. 2013. A systematic review and meta-analysis to revise
the Fenton growth chart for preterm infants. BMC Pediatr. 13(1):59.
Friedman JM. 2009. Leptin, leptin receptors, and the control of body
weight. Nutr Rev. 56(2 Pt 2):S38–S46.
Gat-Yablonski G, Phillip M. 2015. Nutritionally-induced catch-up growth.
Nutrients. 7(1):517–551.
Godfrey KM, Barker DJ. 2001. Fetal programming and adult health.
Public Health Nutr. 4(2B):611–624.
Gohlke BC, Huber A, Hecher K, Fimmers R, Bartmann P, Roth CL. 2005.
Fetal insulin-like growth factor (IGF)-I, IGF-II, and ghrelin in association
with birth weight and postnatal growth in monozygotic twins with
discordant growth. J Clin Endocrinol Metab. 90(4):2270–2274.
Hediger ML, Overpeck MD, Maurer KR, Kuczmarski RJ, McGlynn A, Davis
WW. 1998. Growth of infants and young children born small or large
for gestational age: findings from the Third National Health and
Nutrition Examination Survey. Arch Pediatr Adolesc Med. 152(12):
1225–1231.
Hemachandra AH, Howards PP, Furth SL, Klebanoff MA. 2007. Birth
weight, postnatal growth, and risk for high blood pressure at 7 years
of age: results from the Collaborative Perinatal Project [Comparative
Study Multicenter Study Research Support, N.I.H., Extramural Research
Support, N.I.H., Intramural]. Pediatrics. 119(6):e1264–1270.
Heude B, Forhan A, Slama R, Douhaud L, Bedel S, Saurel-Cubizolles MJ,
Hankard R, et al. 2016. Cohort Profile: The EDEN mother-child cohort
on the prenatal and early postnatal determinants of child health and
development. Int J Epidemiol. 45(2):353–363.
Isidori AM, Caprio M, Strollo F, Moretti C, Frajese G, Isidori A, Fabbri A.
1999. Leptin and androgens in male obesity: evidence for leptin con-
tribution to reduced androgen levels. J Clin Endocrinol Metab. 84(10):
3673–3680.
Jelenkovic A, Hur YM, Sund R, Yokoyama Y, Siribaddana SH, Hotopf M,
Sumathipala A, et al. 2016. Genetic and environmental influences on
adult human height across birth cohorts from 1886 to 1994. Elife. 5:
e20320.
Karakosta P, Chatzi L, Plana E, Margioris A, Castanas E, Kogevinas M.
2011. Leptin levels in cord blood and anthropometric measures at
birth: a systematic review and meta-analysis. Paediatr Perinat
Epidemiol. 25(2):150–163.
Kim-Motoyama H, Yamaguchi T, Katakura T, Miura M, Ohashi Y, Yazaki Y,
Kadawaki T. 1997. Serum leptin levels are associated with hyperinsuli-
nemia independent of body mass index but not with visceral obesity.
Biochem Biophys Res Commun. 239(1):340–344.
Laron Z. 2001. Insulin-like growth factor 1 (IGF-1): a growth hormone.
Mol Pathol. 54(5):311–316.
Leger J, Noel M, Limal JM, Czernichow P. 1996. Growth factors and intra-
uterine growth retardation. II. Serum growth hormone, insulin-like
growth factor (IGF) I, and IGF-binding protein 3 levels in children with
intrauterine growth retardation compared with normal control
subjects: prospective study from birth to two years of age. Study
Group of IUGR. Pediatr Res. 40(1):101–107.
Macdonald PD, Ross SR, Grant L, Young D. 2003. Neonatal weight loss in
breast and formula fed infants. Arch Dis Child Fetal Neonatal Ed.
88(6):F472–476.
Mantzoros CS, Rifas-Shiman SL, Williams CJ, Fargnoli JL, Kelesidis T,
Gillman MW. 2009. Cord blood leptin and adiponectin as predictors
of adiposity in children at 3 years of age: a prospective cohort study.
Pediatrics. 123(2):682–689.
Matsuda J, Yokota I, Iida M, Murakami T, Naito E, Ito M, Shima K, Kuroda
Y. 1997. Serum leptin concentration in cord blood: relationship to
birth weight and gender. J Clin Endocrinol Metab. 82(5):1642–1644.
Onal EE, Cinaz P, Atalay Y, Turkyilmaz C, Bideci A, Akturk A, Okumus N,
et al. 2004. Umbilical cord ghrelin concentrations in small- and appro-
priate-for-gestational age newborn infants: relationship to anthropo-
metric markers. J Endocrinol. 180(2):267–271.
Ong KK, Ahmed ML, Sherriff A, Woods KA, Watts A, Golding J, Dunger
DB. 1999. Cord blood leptin is associated with size at birth and pre-
dicts infancy weight gain in humans. ALSPAC Study Team. Avon
Longitudinal Study of Pregnancy and Childhood. J Clin Endocrinol
Metab. 84(3):1145–1148.
Popova P, Vasilyeva L, Tkachuck A, Puzanov M, Golovkin A, Bolotko Y,
Pustozerov E, et al. 2018. A randomised, controlled study of different
glycaemic targets during gestational diabetes treatment: effect on the
level of adipokines in cord blood and ANGPTL4 expression in human
umbilical vein endothelial cells. Int J Endocrinol. 2018:1–8.
Prader A. 1978. Catch-up growth. Postgrad Med J. 54 (Suppl 1):133–146.
Regnault N, Botton J, Heude B, Forhan A, Hankard R, Foliguet B, Hillier
TA, et al.; the EDEN Mother-Child Cohort Study Group. 2011. Higher
cord C-peptide concentrations are associated with slower growth rate
in the 1st year of life in girls but not in boys [Research Support, Non-
U.S. Gov’t]. Diabetes. 60(8):2152–2159.
Retnakaran R, Ye C, Kramer CK. 2019. Stability of insulin and C-peptide
measurement with long-term frozen storage of serum: Implications
for diabetes research studies. Diabetes Obes Metab. 21(4):1058–1060.
Rosenbloom AL. 2007. Recombinant human insulin-like growth factor I
(rhIGF-I) and rhIGF-I/rhIGF-binding-protein-3: new growth treatment
options? J Pediatr. 150(1):7–11.
Saad MF, Damani S, Gingerich RL, Riad-Gabriel MG, Khan A, Boyadjian R,
Jinagouda SD, et al. 1997. Sexual dimorphism in plasma leptin con-
centration. J Clin Endocrinol Metab. 82(2):579–584.
Shields BM, Knight B, Hopper H, Hill A, Powell RJ, Hattersley AT, Clark
PM. 2007. Measurement of cord insulin and insulin-related peptides
suggests that girls are more insulin resistant than boys at birth.
Diabetes Care. 30(10):2661–2666.
Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, Wood
AM, Carpenter JR. 2009. Multiple imputation for missing data in epi-
demiological and clinical research: potential and pitfalls. BMJ.
338(jun29 1):b2393–b2393.
Verhaeghe J, Van Bree R, Van Herck E, Laureys J, Bouillon R, Van Assche
FA. 1993. C-peptide, insulin-like growth factors I and II, and insulin-
like growth factor binding protein-1 in umbilical cord serum: correla-
tions with birth weight. Am J Obstet Gynecol. 169(1):89–97.
Verkauskiene R, Beltrand J, Claris O, Chevenne D, Deghmoun S, Dorgeret
S, Alison M, et al. 2007. Impact of fetal growth restriction on body
composition and hormonal status at birth in infants of small and
appropriate weight for gestational age. Eur J Endocrinol. 157(5):
605–612.
Victora CG, Adair L, Fall C, Hallal PC, Martorell R, Richter L, Sachdev HS.
Maternal Child Undernutrition Study G. 2008. Maternal and child
undernutrition: consequences for adult health and human capital.
Lancet. 371(9609):340–357.
Yeung EH, Sundaram R, Xie Y, Lawrence DA. 2018. Newborn adipokines
and early childhood growth. Pediatr Obes. 13(8):505–513.
Zhang C, Hediger ML, Albert PS, Grewal J, Sciscione A, Grobman WA,
Wing DA, et al. 2018. Association of maternal obesity with longitudinal
ultrasonographic measures of fetal growth: findings from the NICHD
fetal growth studies-singletons. JAMA Pediatr. 172(1):24–31.
ANNALS OF HUMAN BIOLOGY 165
