To evaluate whether identified SNPs contribute independently to MaxDrinks, we repeated the analyses, conditioning on the most significant ADH1B SNPs in each population: rs1229984 in EAs and rs2066702 in AAs. In EAs, after adjusting for rs1229984, rs2309169 on chromosome 2 remained highly significant (p=2.12 x 10-9) and rs1799876 on chromosome 1 was nearly significant (p=1.99 x 10-7). In AAs, after adjusting for rs2066702, among the eight SNPs identified as significant in the combined Yale-UPenn and the SAGE samples, only one SNP remained significant. Rs200475889 on chromosome 4 showed significant association in the Yale-UPenn AAs after adjusting for rs2066702 (p=4.80 x 10-8), suggesting that both of these SNPs contributed to genetic risk for MaxDrinks in this population. Thus, independent loci other than ADH1B contributed to Maxdrinks in both AAs and EAs.
