Multi-ancestry genome-wide association study of gestational diabetes mellitus highlights genetic links with type 2 diabetes

Abstract

Gestational diabetes mellitus (GDM) is associated with increased risk of pregnancy complications and adverse perinatal outcomes. GDM often reoccurs and is associated with increased risk of subsequent diagnosis of type 2 diabetes (T2D). To improve our understanding of the aetiological factors and molecular processes driving the occurrence of GDM, including the extent to which these overlap with T2D pathophysiology, the GENetics of Diabetes In Pregnancy Consortium assembled genome-wide association studies of diverse ancestry in a total of 5485 women with GDM and 347 856 without GDM. Through multi-ancestry meta-analysis, we identified five loci with genome-wide significant association (P < 5 × 10-8) with GDM, mapping to/near MTNR1B (P = 4.3 × 10-54), TCF7L2 (P = 4.0 × 10-16), CDKAL1 (P = 1.6 × 10-14), CDKN2A-CDKN2B (P = 4.1 × 10-9) and HKDC1 (P = 2.9 × 10-8). Multiple lines of evidence pointed to the shared pathophysiology of GDM and T2D: (i) four of the five GDM loci (not HKDC1) have been previously reported at genome-wide significance for T2D; (ii) significant enrichment for associations with GDM at previously reported T2D loci; (iii) strong genetic correlation between GDM and T2D and (iv) enrichment of GDM associations mapping to genomic annotations in diabetes-relevant tissues and transcription factor binding sites. Mendelian randomization analyses demonstrated significant causal association (5% false discovery rate) of higher body mass index on increased GDM risk. Our results provide support for the hypothesis that GDM and T2D are part of the same underlying pathology but that, as exemplified by the HKDC1 locus, there are genetic determinants of GDM that are specific to glucose regulation in pregnancy.

Figure 1

Correlation between GDM and T2D association summary statistics for lead SNVs at previously reported loci for T2D susceptibility. Association summary statistics for GDM were obtained from multi-ancestry GWAS meta-analyses of 5485 cases and 347 856 controls. Association summary statistics for T2D were obtained from multi-ancestry GWAS meta-analyses of 180 834 cases and 1 159 055 controls from the DIAMANTE Consortium. (a) Allelic effect sizes (log-ORs) for each disease, aligned to the T2D risk allele, from fixed-effects meta-analysis. The grey line represents log-OR of zero for GDM. (b) Association evidence (−log₁₀P-values) for each disease from meta-regression. The grey line represents genome-wide significance (P < 5 × 10⁻⁸) for GDM. The lead SNV at the TCF7L2 locus has been removed for ease of presentation (Supplementary Material, Table S4).

Figure 2

Effects of BMI on GDM from MR analyses. Each point corresponds to an independent SNV (genetic instrument), plotted according to the effect on BMI (on the x-axis) and the effect on GDM (log-OR, on the y-axis). Horizontal and vertical bars represent the standard errors of effect estimates. The coloured regression lines represent the effect of BMI on GDM from six MR models.