Methods
Cambridge baby growth study
The first phase of the prospective, longitudinal Cambridge Baby Growth Study recruited mothers (and their partners and offspring) attending early pregnancy ultrasound clinics at the Rosie Maternity Hospital, Cambridge, UK between the years 2001–2009 [6, 10]. At around 28 weeks of gestation the mothers underwent a 75 g oral glucose tolerance test (OGTT) after fasting overnight. Plasma glucose concentrations were measured using a standard glucose oxidase-based procedure on samples collected when fasting and 60 min after the consumption of the glucose load. Offspring birth weight and gestational age at birth data were collected from hospital notes. Large for gestational age (LGA) at birth was defined as being in the top decile for birth weight adjusted for gestational age. Similarly small for gestational age at birth (SGA) was defined as being in the bottom decile. In this cohort 96.9% of the offspring were Caucasian, 0.8% were mixed race, 0.6% were Afro-Caribbean, 0.8% were Oriental and 0.9% were Indo-Asian. The current analysis was restricted to pregnancies where both fasting and 60 min OGTT glucose concentrations were available.
Genotyping and fetal allele score formulation
Blood or mouth swab samples for DNA extraction were collected from 845 family (mother, father and baby) trios of the 1074 families where maternal OGTT data were available. Genomic DNA was extracted from these samples using an Autopure LS Machine (Qiagen Ltd., Crawley, UK). Allelic transmission to the fetus was imputed from the DNA family trio genotypes [6, 10], with the genotyping performed using Kompetitive Allele Specific PCR assays (by LGC Genomics, Hoddesdon, UK). The genotypes that were used in this study were all consistent with Hardy–Weinberg equilibrium (p > 0.05 using the χ2 test) and had a repeat genotyping discordancy rate of < 1.0%. The unweighted fetal allele score was formulated as previously described [6] using the fetal paternally-transmitted INS-IGF2 rs10770125 and rs2585, and maternally-transmitted KCNQ1 rs231841 and rs7929804 alleles.
Statistical analysis
Associations with offspring birth weight were tested using linear regression, adjusted for established co-variates. Associations with LGA and SGA were tested using both logistic and linear regression. P < 0.05 was considered statistically significant throughout. Data are mean (95% confidence interval) unless stated otherwise.
Results
Associations with offspring birth weight
The covariates in the regression models (gestational age at birth, sex, parity, maternal pre-pregnancy body mass index, pregnancy weight gain and maternal smoking during pregnancy) explained 31.5% of the variance in the offspring birth weights by themselves. OGTT fasting glucose concentrations were significantly associated with offspring birth weights when added to the model (β = 0.16 (0.09–0.23) g l/mmol, β′ = 0.150, p = 1.5 × 10−5, n = 609), as were OGTT 60 min glucose concentrations (β = 0.05 (0.03, 0.07) g l/mmol, β′ = 0.161, p = 4.4 × 10−6, n = 602). The fetal allele score was also positively associated with birth weight (β = 63 (17–109) g/risk allele, β′ = 0.113, p = 7.6 × 10−3, n = 405) (Fig. 1). This association was partially attenuated when adjusting for the OGTT fasting and 60 min glucose concentrations (β = 50 (4–95) g/risk allele, β′ = 0.089, p = 0.03, n = 405), shown by the flatter slope of the predicted line of best fit of the model (Fig. 1).
Associations with being large or small for gestational age at birth
The fetal allele score was positively associated with risk of being LGA [odds ratio (OR) 1.60 (1.19–2.15) per risk allele, p = 2.1 × 10−3] (Fig. 2a). The association was only partially attenuated by adjusting for OGTT fasting and 60 min glucose concentrations [OR 1.47 (1.09–1.98) per risk allele, p = 0.01)]. The fetal allele score was also negatively associated with risk of being SGA [OR 0.65 (0.44–0.96) per risk allele, p = 0.03] (Fig. 2b). The association was not attenuated when adjusting for OGTT fasting and 60 min glucose concentrations [OR 0.67 (0.49–0.99) per risk allele, p = 0.04].
Discussion
In this study the fetal imprinted gene allele score that we had previously shown to be associated with both maternal glucose concentrations and gestational diabetes risk [6], was additionally associated with offspring birth weight and risk of being LGA or SGA. This is perhaps not surprising given the enrichment of imprinted gene regions identified in the largest birth weight-related GWAS using fetal genotypes [8]. The effect sizes were partially attenuated when the associations were further adjusted for week 28 OGTT fasting and 60 min maternal glucose concentrations, suggesting that the link between the fetal allele score and birth weight is mediated through both glucose-dependent and glucose independent mechanisms. These findings are therefore somewhat inconsistent with the associations reported by Hughes et al. [9] where their fetal gene score was associated with birth weight completely independently of maternal glucose concentrations. The difference in the results of the two studies is probably due to the way that the two fetal gene scores were formulated: that used by Hughes et al. [9] being put together from fetal polymorphisms strongly associated with birth weights and our allele score being established using fetal alleles found to be associated with maternal glucose concentrations. Whilst increased maternal glucose concentrations are known to lead to increased birth weights [4], glucose-independent pathways upregulated in gestational diabetes that could potentially affect offspring birth weight include increased placental transport of both fatty acids and certain amino acids [11, 12].
The main strengths of this study are its prospective nature and the use of a novel fetal imprinted gene allele score that we found to be robustly associated with maternal glucose concentrations, even to genome wide significance levels by meta-analysis of three different birth cohorts [6]. Its conclusion is that some of the principal fetal imprinted gene variants that are associated with maternal glucose concentrations in late pregnancy in our studies (fetal paternally-transmitted INS-IGF2 rs10770125 and rs2585, and maternally-transmitted KCNQ1 rs231841 and rs7929804 [6, 10]), are collectively associated with birth weight through both maternal glucose-dependent and glucose-independent mechanisms.
