Table 1c shows pattern of alcohol drinking as reported on the TLFB at assessment. Using the GLRM for genotype groups and their interaction for the three 90-day TLFB drinking measures, significant genotype and epistatic interactions for DAT1 and OPRM1 were found. The odds ratio of reporting more drinking days was significantly greater for A9 carriers vs A10 carriers (Table 2a). There was, however, a significant epistatic interaction, where OPRM1 *G allele appeared to attenuate the effects of the DAT1 A9 on number of drinking days (Table 2a). As shown in Table 2b, the relative risk for drinks per drinking day did not show significant differences between genotype groups. As shown in Table 2c, the odds ratio of heavy drinking days was significantly lower for OPRM1 *G carriers than for OPRM1 AA. There were no epistatic interactions between OPRM1 and DAT1 on number of heavy drinking days. Sex, which was included in all models as an apriori decision, was significant for number of drinking days (p=0.001), number of heavy drinking days (p=0.004) and number of drinks per drinking day (p=0.008).
