in high LD with the core markers defining the associations with both SI and ND. Marker combination 1-3-5-6-7 produces very similar p values as that observed with the core markers (see Table 4, comparing combinations 1-3-5-6-7 and 5-6-7). These results suggest that the Asn40Asp mutation is not itself causative. The associations observed in some studies are likely a reflection of its high LD with other causative variant(s) not yet identified. In a recent study of alcohol dependence, this Asn40Asp polymorphism is not associated the phenotypes [19]. Similar to our study, while this polymorphism is partitioned in a different LD block, it does share substantial LD with a risk haplotype found in a separate LD block. Since several markers (rs1799971, rs609148 and rs648893) in the alcohol dependence study are also typed in our study, this allows us to compare the 3 marker haplotype directly. We find that the extended risk haplotype for SI in our sample is 1-1-1-2-1-2-2-2 or A-C-G-C-A-A-C-T for markers 1-3-5-6-7-8-9-10 or rs1799971-rs524731-rs9479757-rs2075572-rs10485057-rs9322447-rs609148-rs648893. Extracting markers 1-9-10 from this extended haplotype, we have 1-2-2 or A-C-T for rs1799971- rs609148-rs648893. This overlaps with one of the risk haplotypes, A-A-C-C-T for markers rs1799971-rs3823010-rs495491-rs609148-rs648893, identified in the alcoholism study. Given the high comorbidity
