REVIEW

CURRENT
OPINION Genetic origins of low birth weight
Hanieh Yaghootkar a and Rachel M. Freathy a,b

Purpose of review
Smaller size at birth is associated with a higher risk of type 2 diabetes in later life, but the mechanisms
behind this association are poorly understood. Genetic variants which influence susceptibility to type 2
diabetes via effects on insulin secretion or action are good candidates for association with birth weight
because foetal insulin is a key foetal growth factor. This review will focus on recent progress in identifying
associations between common genetic variants and birth weight.
Recent findings
Foetal genetic variants at two loci (near CCNL1 and in ADCY5) were robustly associated with birth weight
via the foetal genotype in the first genome-wide association study of birth weight. The birth weight-lowering
allele at ADCY5 also predisposes to type 2 diabetes. In addition, evidence from studies of other type 2
diabetes loci is accumulating for association between the foetal risk alleles at CDKAL1 and HHEX-IDE and
lower birth weight, and the maternal risk alleles at GCK and TCF7L2 and higher birth weight.
Summary
The associations with birth weight at ADCY5, CDKAL1 and HHEX-IDE support the foetal insulin hypothesis,
which proposed that type 2 diabetes and lower birth weight could be two phenotypes of the same
genotype. The associations at GCK and TCF7L2 illustrate that maternal genes are also important
determinants of birth weight.
Keywords
birth weight, genome-wide association, single nucleotide polymorphism, type 2 diabetes

INTRODUCTION variants which influence insulin secretion or action

A clear understanding of the genetic and environ-
mental determinants of foetal growth is important
because it is correlated with considerable morbidity
in both the short term and long term. Babies with
extremely high or low birth weights are at consider-
able risk of perinatal complications [1,2], and there
is a large body of evidence linking lower size at birth
with a higher risk of chronic adult diseases such as
type 2 diabetes [3]. However, the mechanisms
underlying the associations with disease in later life
are unclear. It is likely that both genetic and non-
genetic factors are responsible. Much work has
focused on the developmental origins of health
and disease hypothesis, which attributes poor adult
health outcomes to adaptations made by the foetus
under adverse conditions in utero, at critical devel-
opmental periods [3]. Twin studies have shown
support for a nongenetic basis of the association
[4,5]. However, evidence from monogenic con-
ditions and epidemiological studies (reviewed in
detail by Shields et al. [6]) suggests that genetic
factors are also likely to play a role. Foetal insulin
is a key foetal growth factor, and the foetal insulin
hypothesis [7] proposes that common genetic
may both reduce foetal growth and increase the risk
of type 2 diabetes in later life. Such an association
between a genotype and the two temporally distinct
phenotypes would explain some of the observed
correlation between lower birth weight and type 2
diabetes.
This review will focus on recent progress in
identifying associations between common genetic
variants and birth weight, and how some of these
associations have enhanced our understanding of
the relationship between birth weight and type 2
diabetes.
a
Genetics of Complex Traits, Peninsula College of Medicine and
Dentistry, University of Exeter, Exeter and bThe MRC Centre for Causal
Analyses in Translational Epidemiology, School of Social and Community
Medicine, University of Bristol, Bristol, UK
Correspondence to Dr Rachel M. Freathy, Genetics of Complex Traits
Peninsula Medical School St. Luke’s Campus, University of Exeter, EX1
2LU, UK. Tel: +44 1392 722925; e-mail: rachel.freathy@pms.ac.uk
Curr Opin Clin Nutr Metab Care 2012, 15:258–264
DOI:10.1097/MCO.0b013e328351f543

www.co-clinicalnutrition.com Volume 15
Number 3
May 2012

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

KEY POINTS

To date, two genetic loci (ADCY5 and CCNL1) have
been robustly associated with birth weight (P < 5  10 –8)
via the foetal genotype.

Variants at four other loci show promising evidence of
association: CDKAL1 and HHEX-IDE via the foetal
genotype; GCK and TCF7L2 via the
maternal genotype.

The foetal associations at ADCY5, CDKAL1 and
HHEX-IDE support the foetal insulin hypothesis, which
proposes that lower birth weight and type 2 diabetes
may be two phenotypes of one genotype.

The associations between maternal GCK and TCF7L2
variants and higher offspring birth weight are probably
mediated by a primary effect on maternal glucose
levels, which result in increased foetal insulin secretion
and increased foetal growth.
BIRTH WEIGHT IS INFLUENCED BY
FOETAL GENES, AND BY MATERNAL
GENES ACTING VIA THE INTRAUTERINE
ENVIRONMENT
Birth weight is a complex trait influenced by
multiple genetic and environmental factors. Esti-
mates of the heritability of birth weight are generally
between 10 and 40% [8,9]. Importantly, two
distinct, but related, genomes contribute to this
genetic variance: foetal genes influence foetal
growth directly, whereas maternal genes influence
the intrauterine environment, which in turn affects
the growth of the foetus (Fig. 1). This principle has
been clearly illustrated by a study of families with
rare mutations in the glucokinase (GCK) gene [10].
Maternal genes Paternal genes
External
Intra-uterine environment
environment
Foetal genes
Foetal insulin levels
Foetal growth
FIGURE 1. Genetic and environmental factors influencing
foetal insulin levels and foetal growth. Reproduced with
permission from [6].
1363-1950 ß 2012 Wolters Kluwer Health | Lippincott Williams & Wilkins
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Genetic origins of low birth weight Yaghootkar and Freathy
Heterozygous GCK mutations cause impaired glu-
cose sensing, which reduces an affected individual’s
ability to secrete insulin and results in stable fasting
hyperglycemia throughout life [11]. In an analysis
of mother–offspring pairs, unaffected offspring
born to mothers with a GCK mutation were, on
average, 600 g heavier at birth than the offspring
of unaffected mothers [10]. This was due to higher
foetal insulin secretion in response to maternal
hyperglycemia. Conversely, offspring with a GCK
mutation born to an unaffected mother secreted less
foetal insulin and were approximately 530 g lighter
at birth. In cases in which both mother and foetus
had a GCK mutation, the two opposing effects
resulted in birth weights that were no different from
those of unaffected mother–offspring pairs [10].
To identify genetic loci that influence birth
weight, it is, therefore, necessary to study both
maternal and foetal genetics. In addition, studies
need to be large to provide sufficient statistical
power to detect common variants of modest effect,
which typically underlie common multifactorial
human traits.
GENOME-WIDE ASSOCIATION STUDIES
HAVE IDENTIFIED FOETAL GENETIC
VARIANTS ASSOCIATED WITH BIRTH
WEIGHT, SOME OF WHICH ARE ALSO
ASSOCIATED WITH TYPE 2 DIABETES
Success in identifying strong, reproducible associ-
ations between common genetic variants and birth
weight has been made possible by genome-wide
association (GWA) studies. Such studies typically
enable 2.5 million or more common single nucleo-
tide polymorphisms (SNPs) to be analysed in
thousands of individuals in a single experiment.
The level of significance generally agreed to consti-
tute robust evidence of association (‘genome-wide
significance’) in such studies is a ¼ 5  10 –8, and to
achieve the necessary statistical power, meta-
analyses of tens to hundreds of thousands of indi-
viduals are performed by large collaborations of
investigators [12].
The Early Growth Genetics (EGG) Consortium
performed the first meta-analysis of GWA studies of
birth weight (n ¼ 10 623 individuals, with replica-
tion in n ¼ 27 591 European individuals) and iden-
tified two robust associations at P < 5  10 –8 [13 ].
&
The first associated locus contained a cluster of
correlated SNPs near the CCNL1 and LEKR1 genes.
Each additional foetal C-allele of SNP rs900400 was
associated with an approximately 42 g [95% confi-
dence interval (CI): 35–48 g] lower birth weight
(P ¼ 2  10 –35). It was also strongly associated with
reduced ponderal index at birth (P ¼ 5  10 –21), with
www.co-clinicalnutrition.com 259
 Paediatrics
an effect size similar to that of the FTO locus on BMI
[14]. Effect size estimates of the associations
between rs900400 and head circumference or birth
length were comparatively much smaller, sug-
gesting this locus influences soft tissue, rather than
skeletal, mass. An independent follow-up study by
& &
Ryckman et al. [15 ] showed that the same SNP was
associated with birth weight in a study of preterm
neonates, thus demonstrating a clear influence on
birth weight irrespective of gestational age. In
another follow-up study of repeated foetal ultra-
&
sound examinations, Mook-Kanamori et al. [16 ]
found evidence of association between rs900400
and foetal growth measures from the second trimes-
ter onward, demonstrating its effects on growth
from early in pregnancy. Despite the very strong
evidence of association with birth weight, the
CCNL1 and LEKR1 genes have no known link with
foetal growth, and the biological mechanism has yet
to be elucidated.
The second cluster of SNPs to be robustly associ-
ated with birth weight centred on the ADCY5 gene
& &
[13 ,17 ]. Each additional foetal C-allele at SNP
rs9883204 is associated with a lower birth weight
by approximately 30 g (95% CI: 23–38 g)
(P ¼ 7  10 –15 [13 ]). Importantly, this SNP is in link-
&
age disequilibrium with SNP rs11708067 (r2 ¼ 0.75),
which was previously shown to predispose to type 2
diabetes (P ¼ 1  10 –20 [18]). The C-allele of
rs9883204 is associated both with lower birth weight
and an increased risk of type 2 diabetes. Therefore,
genetic variation at ADCY5 must explain part of the
observed correlation between lower birth weight
and type 2 diabetes, which supports the foetal insu-
lin hypothesis [7]. It is not known how genetic
variation at ADCY5 alters foetal growth or risk of
type 2 diabetes. Robust associations (P < 5  10 –8)
have been demonstrated between the type 2 dia-
betes risk allele and higher fasting glucose levels
[18], higher 2-h glucose during an oral glucose
tolerance test [19] and lower Homeostatic Model
Assessment (HOMA)-derived index of beta-cell func-
tion (HOMA-B) [18] in adults, in addition to a strong
association (P ¼ 1  10 –6) with lower 2-h serum insu-
lin levels, adjusted for 2-h plasma glucose levels [19].
Together, these suggest a role in insulin secretion,
though it is unlikely that the locus is directly
involved in early insulin response to a glucose
challenge (P > 0.1 for insulinogenic index [19,20]).
Further studies will be needed to determine whether
the ADCY5 locus influences birth weight by affect-
ing foetal insulin secretion in utero.
No common genetic variants apart from those at
the CCNL1 and ADCY5 loci have been associated
with birth weight at genome-wide levels of signifi-
cance. However, evidence is accumulating from
260 www.co-clinicalnutrition.com
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
candidate gene studies that foetal variants in the
CDKAL1 gene and near the HHEX and IDE genes are
also associated with birth weight. Several groups of
investigators have performed large genetic associ-
ation studies of birth weight using risk variants
discovered in GWA studies of type 2 diabetes
([17 ,21–23] Table 1). These are considered good
candidates for association with birth weight due
to their potential effects on insulin secretion or
action. Evidence of association between the type 2
diabetes risk alleles at CDKAL1 or HHEX-IDE and
lower birth weight has been reported in a number of
&
independent studies [17 ,21,24–26], with the larg-
est and most recent meta-analyses showing associ-
ations of 20 g (95% CI: 29 to 11 g) (P ¼ 5  10 –6)
and -16 g (95% CI: 24 to 8 g) (P ¼ 8  10 –5) per risk
&
allele, respectively [17 ]. As with the association at
ADCY5, these associations between lower birth
weight and type 2 diabetes risk alleles support the
foetal insulin hypothesis [7]. Both the CDKAL1 and
HHEX-IDE have been associated with reduced
pancreatic b-cell function and reduced insulin
secretion in studies of nondiabetic adults [27,28].
Further studies will be needed to confirm their
associations with birth weight at genome-wide sig-
nificance and to determine whether they influence
foetal insulin secretion.
NOT ALL TYPE 2 DIABETES LOCI ARE
ASSOCIATED WITH BIRTH WEIGHT
It is clear that association with birth weight is not a
characteristic of all type 2 diabetes loci (Table 1).
Genetic studies of birth weight can, therefore, give
clues as to whether a defect in b-cell function exists
early in development or appears later in life. For
example, TCF7L2, which confers the largest risk of
type 2 diabetes in Europeans and strongly influences
b-cell function and insulin secretion in adulthood,
has been studied extensively in relation to birth
weight and shows no association via the foetal
&
genotype [17 ,22,24,26,29]. This suggests that, in
contrast to the likely effects of ADCY5, CDKAL1
and HHEX-IDE, any b-cell defect conferred by
TCF7L2 is not present in foetal life.
GENETIC LOCI ASSOCIATED WITH ADULT
BMI DO NOT INFLUENCE BIRTH WEIGHT
High birth weight is correlated with high BMI in
adulthood [30]. To investigate whether common
genetic variants are associated with body weight
in both foetal and adult life, two recent studies
tested associations with birth weight of loci known
to influence adult BMI and measures of adiposity.
Between 12 and 24 loci were tested in sample sizes
Volume 15
Number 3
May 2012
 Table 1. Published associations between birth weight and common foetal genetic variants (primarily associated either with birth weight or type 2
diabetes in genome-wide association studies at P < 5  108), analysed in at least 10 000 individuals
Changec in birth Change in birth weight
weight per foetal per foetal risk allele in n in analysis
Known Total n in largest risk allele in grams grams, adjusted for adjusted for Source(s)
a
Gene(s) marking associated Risk Nonrisk reported (95% confidence maternal genotype maternal of data
association signal SNP(s) adult trait(s) allele allele meta-analysisb interval) P-value (95% confidence interval) genotype (references)

&
CCNL1 rs900400 – C T 37 745 42 (48 to 35) 2  1035 55 (72 to 39) 7659 [13 ]
15 &
ADCY5 rs9883204 T2D, FPG C T 38 214 30 (38 to 23) 7  10 43 (60 to 24) 7910 [13 ]
d 6 &
CDKAL1 rs10946398, T2D C A 24 885 20 (29 to 11) 5  10 36 (56 to 16) 5399 [17 ,21]
rs7756992,
rs7754840
&
HHEX-IDE rs1111875 T2D C T 25 164 16 (24 to 8) 8  10 – 5 29 (48 to 10) 5480 [17 ,21]
d e &
GCK rs4607517, T2D, FPG A G 10 623 21 (4–39) 0.01 18 (NA) 2102 [13 ], 23
rs1799884
&
TCF7L2 rs7903146 T2D, FPG T C 19 745 12 (3–22) 0.01 9 (9 to 31) 6044 [17 ,22]
&
CDKN2A,CDKN2B rs10811661 T2D T C 25 957 12 (2–22) 0.02 2 (27 to 23) 5493 [17 ,21]
&
MTNR1B rs10830963 T2D, FPG G C 10 623 16 (1–31) 0.04 NA NA [13 ]
&
SLC30A8 rs13266634 T2D, FPG C T 24 908 7 (2 to 15) 0.10 7 (28 to 14) 5342 [17 ,21]
&
CDC123,CAMK1D rs12779790 T2D G A 10 623 11 (5 to 27) 0.19 NA NA [13 ]
&
IGF2BP2 rs4402960 T2D T G 24 393 6 (3 to 14) 0.20 10 (31 to 10) 5507 [17 ,21]
&
GCKR rs780094 T2D, FPG C T 10 623 8 (21 to 5) 0.21 NA NA [13 ]
&
KCNQ1 rs2237892 T2D C T 10 623 12 (40 to 15) 0.37 NA NA [13 ]

1363-1950 ß 2012 Wolters Kluwer Health | Lippincott Williams & Wilkins

&
KCNJ11 rs5215 T2D C T 10 623 6 (18 to 7) 0.38 NA NA [13 ]
&
ADAMTS9 rs4607103 T2D C T 10 623 7 (8 to 22) 0.39 NA NA [13 ]
&
WFS1 rs10010131 T2D G A 10 623 6 (18 to 7) 0.39 NA NA [13 ]
&
PROX1 rs340874 T2D, FPG C T 10 623 5 (18 to 7) 0.40 NA NA [13 ]
&
FTO rs8050136 T2D A C 10 623 5 (8 to 18) 0.43 NA NA [13 ]
&
TSPAN8 rs7961581 T2D C T 10 623 6 (21 to 9) 0.45 NA NA [13 ]
&
TCF2 rs4430796 T2D G A 10 623 4 (10 to 19) 0.55 NA NA [13 ]
&
PPARG rs1801282 T2D C G 10 623 3 (21 to 15) 0.74 NA NA [13 ]
&
NOTCH2 rs10923931 T2D T G 10 623 2 (22 to 17) 0.82 NA NA [13 ]
&
DGKB rs2191349 T2D, FPG T G 10 623 1 (11 to 14) 0.87 NA NA [13 ]
&
THADA rs7578597 T2D T C 10 623 2 (24 to 20) 0.88 NA NA [13 ]
&
JAZF1 rs864745 T2D T C 10 623 1 (13 to 12) 0.91 NA NA [13 ]

NA, not available; SNP, single nucleotide polymorphism; T2D, type 2 diabetes; FPG, fasting plasma glucose.
a
The risk allele is the type 2 diabetes susceptibility allele. There are no known associations between common genetic variation at CCNL1 and traits other than birth weight, so the effect is reported for the birth weight-
lowering allele.
b & & & &
The samples meta-analysed [13 ,17 ] are not independent (83% of the individuals meta-analysed in [13 ] are also in [17 ]). Therefore, for each locus analysed in both studies (CDKAL1, HHEX-IDE, TCF7L2, SLC30A8,

www.co-clinicalnutrition.com
IGF2BP2, PPARG, KCNJ11 and JAZF1), the result with the larger sample size is presented.
c &
The effect sizes in [13 ] were reported as z-scores. To facilitate comparison, they have been converted to the equivalent effect sizes in grams by multiplying by 484, which is the median standard deviation of birth
&
weight in grams, from studies included in [13 ].
d
Different SNPs have been (meta)-analysed at the CDKAL1 and GCK loci, depending on the study. However, the SNPs within each locus are highly correlated: r2 > 0.67 at CDKAL1; r2 ¼ 1 at GCK. The risk/nonrisk
Genetic origins of low birth weight Yaghootkar and Freathy

261
alleles relate to the underlined SNPs.
e
The reported effect size in [23] compares birth weights of individuals who carried one or two risk alleles with those who did not (P > 0.05).

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Paediatrics
ranging from n ¼ 4744 to n ¼ 28 219 [31 ,32]. How-
&
ever, there was no strong evidence of association,
suggesting little overlap between the foetal genetics
of foetal growth and adult adiposity.
GENETIC ASSOCIATION STUDIES OF
BIRTH WEIGHT SHOULD ADJUST FOR THE
CORRELATION BETWEEN FOETAL AND
MATERNAL GENOTYPES
The opposing effects on birth weight of maternal
and foetal heterozygous GCK mutations [10] demon-
strate that an analysis of birth weight which considers
only the foetal, or only the maternal, genotype
could produce spurious results. For example, if the
foetal genotype only is analysed when there are truly
equal but opposing effects of maternal and foetal
genotype, there is a high probability that the analysis
would find no evidence of association with birth
weight. Alternatively, if there is no effect of foetal
genotype, but a true association between maternal
genotype and birth weight, an analysis of only the
foetal genotype might find evidence of association,
but this would be purely attributable to the corre-
lation between maternal and foetal genotypes. An
example of the latter was observed in a study
of TCF7L2: an initially strong association between
foetal genotype and higher birth weight [18 g (95% CI
7–29 g); P ¼ 0.001] was greatly reduced after adjust-
ing for maternal genotype [9 g (95% CI 9 to 31 g);
P ¼ 0.29] [22]. Analysis of foetal genotype, con-
ditional on maternal genotype, has been carried
&
out for common variants at several loci [13 ,21,
22,33]. The effect size estimates at CCNL1, ADCY5,
CDKAL1 and HHEX-IDE do not decrease on adjust-
ment for maternal genotype (Table 1), confirming
that these associations are with the foetal genotype.
Moreover, at each of these loci, the maternal-
adjusted effect size estimates are larger than the
unadjusted estimates. For example, an analysis of
ADCY5 in n ¼ 7910 mother–offspring pairs showed
a 44 g (95% CI: 25–62 g) lower birth weight per
foetal C-allele after adjustment for maternal geno-
type [compared with 33 g(95% CI: 17–49 g) before
adjustment]. This is consistent with an opposing
effect of maternal genotype, which, given the
known glycemic associations at this locus, may
be due to higher maternal glucose levels. However,
the sample sizes available to date for analysis of
both maternal and foetal genotypes are relatively
small (typically between n  5000 and n  8000
mother–offspring pairs). Larger samples will be
necessary to provide sufficient statistical evidence
to distinguish between changes in effect size that
are due to adjustment for maternal genotype or due
to chance.
262 www.co-clinicalnutrition.com
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
COMMON MATERNAL GENETIC VARIANTS
AT THE GCK AND TCF7L2 LOCI ARE
ASSOCIATED WITH OFFSPRING BIRTH
WEIGHT VIA MATERNAL GLUCOSE LEVELS
Studies of the associations between common
maternal genetic variants and offspring birth weight
have been less extensive than those of foetal genetic
variants, both in terms of the number of loci studied
and the available sample sizes ([21,22,33,34 ];
&
Table 2). Despite this, common maternal variants
at two loci have shown strong evidence (P < 0.001)
of association in large meta-analyses (n > 12 500).
Common maternal variants at GCK and TCF7L2
predispose both to type 2 diabetes and higher fasting
glucose levels, and their estimated effects on
offspring birth weight are increases of 30 g (95%
CI: 15–44 g) and 23 g (95% CI: 11–35 g) per
&
maternal risk allele, respectively [34 ]. The common
maternal GCK and TCF7L2 variants are both asso-
ciated with maternal glucose levels in pregnancy
and with the recent consensus definition of gesta-
tional diabetes from the International Association of
&
Diabetes and Pregnancy Study Groups [34 ,35]. As
maternal glucose levels in pregnancy are strongly
predictive of offspring birth weight [36], it is likely
that the associations at GCK and TCF7L2 are medi-
ated by raised fasting glucose, which crosses the
placenta, causing the foetus to respond by secreting
more foetal insulin, which acts as a growth factor.
The associations between maternal GCK or
TCF7L2 and offspring birth weight are little altered
on adjustment for foetal genotype (Table 2). This is
in line with the results of foetal genotype analyses of
these loci, which show no association with birth
&
weight [22,23,33,34 ]. For the common GCK variant
this lack of foetal effect contrasts with the strong
foetal effect of the rare mutations at the locus [10].
Given the clear importance of maternal glucose
levels on birth weight, we might expect all genetic
loci which influence maternal glucose levels to be
also associated with offspring birth weight. This
hypothesis remains to be tested in well powered
analyses.
NEXT STEPS
For a clearer understanding of the genetic contri-
bution to foetal growth, it will be necessary to per-
form larger GWA studies, which include separate and
conditional analyses of foetal and maternal geno-
types. ‘Next generation’ genetic approaches, includ-
ing whole genome and exome sequencing and denser
or exome-specific genotyping microarrays should
additionally prove useful for capturing lower fre-
quency genetic variation. In addition, associations
between confirmed birth weight loci and key-related
Volume 15
Number 3
May 2012
 Genetic origins of low birth weight Yaghootkar and Freathy

References

[33,34 ]
[22,34 ]
&

&

.4
[21]
[21]
[21]
[21]
[21]

Log (odds ratio) for type 2 diabetes

TCF7L2

.3
Table 2. Published associations between birth weight and common maternal genetic variants, analysed in at least 5000 mother-offspring pairs

foetal genotype
n in analysis
adjusted for

>6500
6044
5480
5493
5342
5399
5507

.2
CDKAL1

HHEX-IDE

.1
weight per maternal

ADCY5
risk allele in grams,

confidence interval)
adjusted for foetal

31 (9–48)
7 (12 to 26)
16 (9 to 41)
20 (0–41)
21 (1–42)
7 (13 to 27)
b

GCK
26 (NA)
Change in birth

genotype (95%

The reported effect size in [33] compares birth weights of offspring born to mothers who carried one or two risk alleles with those born to mothers who had no risk alleles.

0
−40 −20 0 20 40
Change in birth weight per risk allele, in grams

FIGURE 2. Five type 2 diabetes risk loci associated with

birth weight (P < 0.001). The plot shows the type 2 diabetes
–5

–4
5  10
3  10
0.05
0.06
0.29
0.60
0.83
P-value

effect size (y-axis) against the birth weight effect size (x-axis).
White bars indicate associations with foetal genotype, and
shaded bars indicate associations with maternal genotype.
The underlined (ADCY5) locus is associated with birth
(95% confidence interval)

weight at P < 5  10 – 8. The error bars show 95% confidence

30 (15, 44)
23 (11, 35)

17 (1 to 35)
8 (7 to 23)
4 (11 to 19)
2 (13 to 17)
14 (28 to 0)
weight per maternal

intervals. The effect sizes for birth weight relate to the SNPs
risk allele in grams
Change in birth

shown in Tables 1 and 2. The effect sizes for type 2

diabetes are from published genome-wide association meta-
analyses [18,38] [for ADCY5 and HHEX-IDE, the type 2
diabetes SNPs were rs11708067 and rs5015480,
respectively (r2 > 0.8 with the birth weight SNPs)].
FPG, fasting plasma glucose; NA, not available; SNP, single nucleotide polymorphisms; T2D, type 2 diabetes.
argest reported
(meta)-analysis

traits such as foetal insulin levels and maternal meta-

12 643
13 406
7799
7817
7655
7760
7821
Total n in l

bolic traits in pregnancy will enable us to begin to

The risk allele is the type 2 diabetes susceptibility allele and/or the fasting glucose-raising allele.

understand the mechanisms underlying the associ-

ations. Finally, most genetic association studies of
birth weight have focused on individuals of European
ancestry, and to date, there have been only a few
Nonrisk
allele

G
C
C
C
C

&
T

studies of non-European populations [13 ,37]. Large

studies will be necessary to assess whether the
observed associations are present in other ethnic
groups, and may help to narrow down the regions
allele
Riska

C
C
T
T
T
T

of association to locate the causal variants.

associated

T2D, FPG
T2D, FPG

T2D, FPG
adult trait

CONCLUSION
T2D
T2D

T2D
T2D
Known

To date, two genetic loci (ADCY5 and CCNL1) have

been robustly associated with birth weight in large
studies of Europeans, and four others (CDKAL1,
rs10811661
rs13266634
rs10946398
rs1799884
rs7903146
rs1111875

rs4402960

HHEX-IDE, GCK, TCF7L2) show promising evidence

of association. Five of the six loci are also associated
SNP

with type 2 diabetes, with birth weight effects from

either the foetal genotype or the maternal genotype
(Fig. 2). This demonstrates that there is a clear
CDKN2A,CDKN2B

overlap between the genetics of type 2 diabetes

association signal
Gene(s) marking

and foetal growth. The foetal associations at ADCY5,

CDKAL1 and HHEX-IDE support the foetal insulin
SLC30A8
HHEX-IDE

IGF2BP2
CDKAL1
TCF7L2

hypothesis, which proposes that lower birth weight

GCK

and type 2 diabetes may be two phenotypes of one

a
b

1363-1950 ß 2012 Wolters Kluwer Health | Lippincott Williams & Wilkins www.co-clinicalnutrition.com 263

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Paediatrics
genotype [7]. However, as they explain only a small
amount of variation, much of the low birth weight –
type 2 diabetes correlation remains to be explained.
It will be important to understand clearly the
relative contributions of genetic and nongenetic
factors in order to develop effective intervention
strategies to prevent type 2 diabetes.
Acknowledgements
We are grateful to Professor Tim Frayling for helpful
discussions. R.M.F. is a Sir Henry Wellcome Postdoctoral
Fellow (Wellcome Trust grant: 085541/Z/08/Z).
Conflicts of interest
Funding: This work was supported by a Sir Henry Well-
come Postdoctoral Fellowship, awarded to R.M.F (Well-
come Trust grant: 085541/Z/08/Z).
There are no conflicts of interest.
REFERENCES AND RECOMMENDED
READING
Papers of particular interest, published within the annual period of review, have
been highlighted as:
& of special interest
&& of outstanding interest
Additional references related to this topic can also be found in the Current
World Literature section in this issue (p. 318).
1. Acker DB, Sachs BP, Friedman EA. Risk factors for shoulder dystocia. Obstet
gynecol 1985; 66:762–768.
2. Battaglia FC, Lubchenco LO. A practical classification of newborn infants by
weight and gestational age. J pediatr 1967; 71:159–163.
3. Godfrey KM, Barker DJ. Fetal nutrition and adult disease. Am J Clin Nutr 2000;
71:1344S–1352S.
4. Poulsen P, Grunnet LG, Pilgaard K, et al. Increased risk of type 2 diabetes in
elderly twins. Diabetes 2009; 58:1350–1355.
5. Poulsen P, Vaag AA, Kyvik KO, et al. Low birth weight is associated with
NIDDM in discordant monozygotic and dizygotic twin pairs. Diabetologia
1997; 40:439–446.
6. Shields B, Freathy RM, Hattersley AT. Genetic influences on the association
between fetal growth and susceptibility to type 2 diabetes. J Dev Origin
Health Dis 2010; 1:95–105.
7. Hattersley AT, Tooke JE. The fetal insulin hypothesis: an alternative explana-
tion of the association of low birthweight with diabetes and vascular disease.
Lancet 1999; 353:1789–1792.
8. Lunde A, Melve KK, Gjessing HK, et al. Genetic and environmental influences on
birth weight, birth length, head circumference, and gestational age by use of
population-based parent-offspring data. Am J Epidemiol 2007; 165:734–741.
9. van Baal CG, Boomsma DI. Etiology of individual differences in birth weight of
twins as a function of maternal smoking during pregnancy. Twin Res 1998;
1:123–130.
10. Hattersley AT, Beards F, Ballantyne E, et al. Mutations in the glucokinase gene
of the fetus result in reduced birth weight. Nat Genet 1998; 19:268–270.
11. Froguel P, Zouali H, Vionnet N, et al. Familial hyperglycemia due to mutations
in glucokinase. Definition of a subtype of diabetes mellitus. N Engl J Med
1993; 328:697–702.
12. de Bakker PI, Ferreira MA, Jia X, et al. Practical aspects of imputation-driven
meta-analysis of genome-wide association studies. Hum Mol Genet 2008;
17:R122–R128.
13. Freathy RM, Mook-Kanamori DO, Sovio U, et al. Variants in ADCY5 and near
& CCNL1 are associated with fetal growth and birth weight. Nat Genet 2010;
42:430–435.
This meta-analysis of GWA studies identified the first two genetic loci to be
associated with birth weight at genome-wide levels of significance.
14. Frayling TM, Timpson NJ, Weedon MN, et al. A common variant in the FTO
gene is associated with body mass index and predisposes to childhood and
adult obesity. Science 2007; 316:889–894.
15. Ryckman KK, Feenstra B, Shaffer JR, et al. Replication of a genome-wide
& association study of birth weight in preterm neonates. J Pediatr 2012;
160:19e4–24e4.
This study of preterm births showed that the birth weight-associated variant near
CCNL1 affects birth weight irrespective of gestational age.
264 www.co-clinicalnutrition.com
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
16. Mook-Kanamori DO, Marsh JA, Warrington NM, et al. Variants near CCNL1/
& LEKR1 and in ADCY5 and fetal growth characteristics in different trimesters.
J Clin Endocrinol Metabol 2011; 96:E810–E815.
This study showed that the birth weight-associated variant near CCNL1 influences
foetal growth from early in pregnancy, as it is associated with anthropometric
measures taken via foetal ultrasound from the second trimester.
17. Andersson EA, Pilgaard K, Pisinger C, et al. Type 2 diabetes risk alleles near
& ADCY5, CDKAL1 and HHEX-IDE are associated with reduced birthweight.
Diabetologia 2010; 53:1908–1916.
This is an important study because it meta-analyses previously published data on
associations between birth weight and foetal genotype at several type 2 diabetes
susceptibility loci. It shows that there is strong cumulative evidence of association
at the CDKAL1 and HHEX-IDE loci.
18. Dupuis J, Langenberg C, Prokopenko I, et al. New genetic loci implicated in
fasting glucose homeostasis and their impact on type 2 diabetes risk. Nat
Genet 2010; 42:105–116.
19. Saxena R, Hivert MF, Langenberg C, et al. Genetic variation in GIPR
influences the glucose and insulin responses to an oral glucose challenge.
Nat Genet 2010; 42:142–148.
20. Ingelsson E, Langenberg C, Hivert MF, et al. Detailed physiologic character-
ization reveals diverse mechanisms for novel genetic Loci regulating glucose
and insulin metabolism in humans. Diabetes 2010; 59:1266–1275.
21. Freathy RM, Bennett AJ, Ring SM, et al. Type 2 diabetes risk alleles are
associated with reduced size at birth. Diabetes 2009; 58:1428–1433.
22. Freathy RM, Weedon MN, Bennett A, et al. Type 2 diabetes TCF7L2 risk
genotypes alter birth weight: a study of 24 053 individuals. Am J Hum Genet
2007; 80:1150–1161.
23. Weedon MN, Frayling TM, Shields B, et al. Genetic regulation of birth weight
and fasting glucose by a common polymorphism in the islet cell promoter of
the glucokinase gene. Diabetes 2005; 54:576–581.
24. Pulizzi N, Lyssenko V, Jonsson A, et al. Interaction between prenatal growth
and high-risk genotypes in the development of type 2 diabetes. Diabetologia
2009; 52:825–829.
25. Winkler C, Illig T, Koczwara K, et al. HHEX-IDE polymorphism is associated
with low birth weight in offspring with a family history of type 1 diabetes. J Clin
Endocrinol Metabol 2009; 94:4113–4115.
26. Zhao J, Li M, Bradfield JP, et al. Examination of type 2 diabetes loci implicates
CDKAL1 as a birth weight gene. Diabetes 2009; 58:2414–2418.
27. Steinthorsdottir V, Thorleifsson G, Reynisdottir I, et al. A variant in CDKAL1
influences insulin response and risk of type 2 diabetes. Nat Genet 2007;
39:770–775.
28. Pascoe L, Tura A, Patel SK, et al. Common variants of the novel type 2
diabetes genes CDKAL1 and HHEX/IDE are associated with decreased
pancreatic beta-cell function. Diabetes 2007; 56:3101–3104.
29. Mook-Kanamori DO, de Kort SW, van Duijn CM, et al. Type 2 diabetes gene
TCF7L2 polymorphism is not associated with fetal and postnatal growth in
two birth cohort studies. BMC Med Genet 2009; 10:67.
30. Parsons TJ, Power C, Manor O. Fetal and early life growth and body mass
index from birth to early adulthood in 1958 British cohort: longitudinal study.
BMJ 2001; 323:1331–1335.
31. Kilpelainen TO, den Hoed M, Ong KK, et al. Obesity-susceptibility loci have a
& limited influence on birth weight: a meta-analysis of up to 28 219 individuals.
Am J Clin Nutr 2011; 93:851–860.
This study analysed the association between several loci known to influence adult
BMI and birth weight in up to 28 219 individuals. It showed that, in well powered
meta-analyses, there was little evidence that these loci influence birth weight.
32. Andersson EA, Pilgaard K, Pisinger C, et al. Do gene variants influencing adult
adiposity affect birth weight? A population-based study of 24 loci in
4 744 Danish individuals. PloS One 2010; 5:e14190.
33. Weedon MN, Clark VJ, Qian Y, et al. A common haplotype of the glucokinase
gene alters fasting glucose and birth weight: association in six studies and
population-genetics analyses. Am J Hum Genet 2006; 79:991–1001.
34. Freathy RM, Hayes MG, Urbanek M, et al. Hyperglycemia and Adverse
& Pregnancy Outcome (HAPO) study: common genetic variants in GCK and
TCF7L2 are associated with fasting and postchallenge glucose levels in
pregnancy and with the new consensus definition of gestational diabetes
mellitus from the International Association of Diabetes and Pregnancy Study
Groups. Diabetes 2010; 59:2682–2689.
This is an important study because it meta-analyses previously published data on
associations between birth weight and maternal genotype at the GCK and
TCF7L2 loci. It shows that there is strong cumulative evidence of association.
35. Metzger BE, Gabbe SG, Persson B, et al. International association of
diabetes and pregnancy study groups recommendations on the diagnosis
and classification of hyperglycemia in pregnancy. Diabetes care 2010;
33:676–682.
36. Metzger BE, Lowe LP, Dyer AR, et al. Hyperglycemia and adverse pregnancy
outcomes. N Engl J Med 2008; 358:1991–2002.
37. Vasan SK, Neville MJ, Antonisamy B, et al. Absence of birth-weight lowering
effect of ADCY5 and near CCNL, but association of impaired glucose-insulin
homeostasis with ADCY5 in Asian Indians. PloS One 2011; 6:e21331.
38. Voight BF, Scott LJ, Steinthorsdottir V, et al. Twelve type 2 diabetes
susceptibility loci identified through large-scale association analysis. Nat
Genet 2010; 42:579–589.
Volume 15
Number 3
May 2012