A Large Multiethnic Genome-Wide Association Study of Adult Body Mass Index Identifies Novel Loci

Body mass index (BMI), a proxy measure for obesity, is determined by both environmental (including ethnicity, age, and sex) and genetic factors, with > 400 BMI-associated loci identified to date. However, the impact, interplay, and underlying biological mechanisms among BMI, environment, genetics, and ancestry are not completely understood. To further examine these relationships, we utilized 427,509 calendar year-averaged BMI measurements from 100,418 adults from the single large multiethnic Genetic Epidemiology Research on Adult Health and Aging (GERA) cohort. We observed substantial independent ancestry and nationality differences, including ancestry principal component interactions and nonlinear effects. To increase the list of BMI-associated variants before assessing other differences, we conducted a genome-wide association study (GWAS) in GERA, with replication in the Genetic Investigation of Anthropomorphic Traits (GIANT) consortium combined with the UK Biobank (UKB), followed by GWAS in GERA combined with GIANT, with replication in the UKB. We discovered 30 novel independent BMI loci (<i>P</i> < 5.0 × 10^-8) that replicated. We then assessed the proportion of BMI variance explained by sex in the UKB using previously identified loci compared to previously and newly identified loci and found slight increases: from 3.0 to 3.3% for males and from 2.7 to 3.0% for females. Further, the variance explained by previously and newly identified variants decreased with increasing age in the GERA and UKB cohorts, echoed in the variance explained by the entire genome, which also showed gene-age interaction effects. Finally, we conducted a tissue expression QTL enrichment analysis, which revealed that GWAS BMI-associated variants were enriched in the cerebellum, consistent with prior work in humans and mice.