Genome-wide associations for birth weight and correlations with adult disease

Abstract

Multi-ancestry genome-wide association analyses for birth weight in 153,781 individuals identified 60 genomic loci in which birth weight and fetal genotype were associated and found an inverse genetic correlation between birth weight and cardiometabolic risk. Mark McCarthy and colleagues from the Early Growth Genetics (EGG) Consortium report multi-ancestry genome-wide association analyses for birth weight in 153,781 individuals, identifying 60 genomic loci where the fetal genotype was associated with birth weight. They demonstrate an inverse genetic correlation between birth weight and developing type 2 diabetes or coronary artery disease as an adult, suggesting that previous epidemiological associations may be explained in part by shared genetic variation influencing both early growth and adult disease. Birth weight (BW) has been shown to be influenced by both fetal and maternal factors and in observational studies is reproducibly associated with future risk of adult metabolic diseases including type 2 diabetes (T2D) and cardiovascular disease1. These life-course associations have often been attributed to the impact of an adverse early life environment. Here, we performed a multi-ancestry genome-wide association study (GWAS) meta-analysis of BW in 153,781 individuals, identifying 60 loci where fetal genotype was associated with BW (P < 5 × 10−8). Overall, approximately 15% of variance in BW was captured by assays of fetal genetic variation. Using genetic association alone, we found strong inverse genetic correlations between BW and systolic blood pressure (Rg = −0.22, P = 5.5 × 10−13), T2D (Rg = −0.27, P = 1.1 × 10−6) and coronary artery disease (Rg = −0.30, P = 6.5 × 10−9). In addition, using large -cohort datasets, we demonstrated that genetic factors were the major contributor to the negative covariance between BW and future cardiometabolic risk. Pathway analyses indicated that the protein products of genes within BW-associated regions were enriched for diverse processes including insulin signalling, glucose homeostasis, glycogen biosynthesis and chromatin remodelling. There was also enrichment of associations with BW in known imprinted regions (P = 1.9 × 10−4). We demonstrate that life-course associations between early growth phenotypes and adult cardiometabolic disease are in part the result of shared genetic effects and identify some of the pathways through which these causal genetic effects are mediated.