The following additional time-varying covariates were added, each separately, to the models in order to further probe for the genotype effects observed: (a) setting (i.e., where individuals were drinking), (b) whether or not others were drinking; and (c) time since last cigarette (among smokers only). Gene × Environment interactions were examined with the time-varying covariates. In light of the significant gender imbalance in the DRD4 and OPRM1 groups (shown in Table 1), and the higher frequency of females among consenters to DNA collection, gender was used as a covariate in all analyses. Analyses were repeated with smoking status (yes/no) in the model to explore its possible moderating effect on the hypothesized relationships.
