A meta‐analysis and polygenic score study identifies novel genetic markers for waist‐hip ratio in African populations

Objective Understanding the genetic underpinnings of anthropometric traits in diverse populations is crucial for gaining insights into their biological mechanisms and potential implications for health. Methods We conducted a genome‐wide association study, meta‐analysis, and gene set analysis of waist‐hip ratio (WHR), WHR adjusted for BMI (WHRadjBMI), waist circumference, BMI, and height using the African Collaborative Center for Microbiome and Genomics Research (ACCME) cohort (n = ~11,000) for discovery and polygenic score target analyses and the Africa America Diabetes Mellitus (AADM) study (n = ~5200) for replication and polygenic score validation. We generated and compared polygenic scores from European, African, Afro‐Caribbean, and multiethnic ancestry populations. Results The top loci associated with each trait in the meta‐analysis were in CD36 (rs3211826 [p = 5.90 × 10−12] for WHR and rs73709003 [p = 1.75 × 10−13] for WHRadjBMI), IFI27L1 (rs59775050 [p = 2.66 × 10−08] for waist circumference), INPP4B (rs2636629 [p = 1.44 × 10−09] for BMI), and HMGA1 (rs6937622 [p = 1.40 × 10−15] for height) gene regions. A novel variant rs7797157, near GNAT3, was also significantly associated with WHR (p = 2.50 × 10−10) and WHRadjBMI (p = 2.66 × 10−11). The ancestry‐specific parameters for the best predictive polygenic scores were European ancestry (R2 = 0.68%; p = 1.63 × 10−16) and multiethnic ancestry (R2 = 0.06%; p = 1.29 × 10−02) for WHR; European ancestry (R2 = 1.36%; p = 2.94 × 10−31) and multiethnic ancestry (R2 = 1.12%; p = 3.52 × 10−25) for BMI; and European ancestry (R2 = 3.16%; p = 2.95 × 10−73), African ancestry (R2 = 4.16%; p = 1.75 × 10−96), and African and Afro‐Caribbean ancestry (R2 = 2.67%; p = 4.35 × 10−62) for height. Conclusions The discovery of a novel locus for WHR and genetic signals for each trait and the assessment of polygenic score performance underscore the importance of conducting well‐powered studies in diverse populations.