We conducted GWAS, on 7,784,968 SNPs in the EUR sample and 16,100,604 SNPs in the AFR sample using a mixed model incorporating a genetic relationship matrix to control for relatedness [59] in the GWAF package in R [60]. We included sex, age, the first three ancestral PCs (PC1-PC3), genotype array, and birth cohort (prior to 1930, 1930–1949, 1950–1969, and 1970 and after) as covariates. GWAS were stratified by ancestries, using identical phenotypic definitions, covariates, SNP QC standards, MAF thresholds and imputation protocols. Subsequently, we meta-analyzed across the AFR and EUR results using inverse-variance fixed-effects weighting and genomic control in METAL [61]. We used established thresholds for genome wide significance (p< 5 × 10−8). We also conducted a post-hoc GWAS analysis covarying for depression, given the high prevalence.
