We hypothesised that some of the genetic risk for gestational diabetes (GDM) is due to the fetal genome affecting maternal glucose concentrations. Previously, we found associations between fetal IGF2 gene variants and maternal glucose concentrations in late pregnancy.

In the present study, we tested associations between SNP alleles from 15 fetal imprinted genes and maternal glucose concentrations in late pregnancy in the Cambridge Baby Growth and Wellbeing cohorts (1160 DNA trios).

Four fetal SNP alleles with the strongest univariate associations: paternally-transmitted IGF2 rs10770125 (P-value=2×10^–4) and INS rs2585 (P-value=7×10^–4), and maternally-transmitted KCNQ1(OT1) rs231841 (P-value=1×10^–3) and KCNQ1(OT1) rs7929804 (P-value=4×10^–3), were used to construct a composite fetal imprinted gene allele score which was associated with maternal glucose concentrations (P-value=4.3×10^–6, n=981, r²=2.0%) and GDM prevalence (odds ratio per allele 1.44 (1.15, 1.80), P-value=1×10^–3, n=89 cases and 899 controls). Meta-analysis of the associations including data from 1367 Hyperglycaemia and Adverse Pregnancy Outcome Study participants confirmed the paternally-transmitted fetal IGF2/INS SNP associations (rs10770125, P-value=3.2×10^–8, rs2585, P-value=3.6×10^–5) and the composite fetal imprinted gene allele score association (P-value=1.3×10^–8), but not the maternally-transmitted fetal KCNQ1(OT1) associations (rs231841, P-value=0.4; rs7929804, P-value=0.2).

This study suggests that polymorphic variation in fetal imprinted genes, particularly in the IGF2/INS region, contribute a small but significant part to the risk of raised late pregnancy maternal glucose concentrations.