A consistent pattern of results was observed in the regressions predicting drinking variables as a function of implicit coping and ALDH2 (see Table 2). Main effects for ALDH2 and the cope IAT were significant for all three drinking variables (all p’s < .004). In addition, there was a significant ALDH2 × cope IAT interaction for all three drinking variables (all p’s < .038), supporting our moderation hypothesis. Consistent with predictions, associations of cope IAT scores were stronger among ALDH2*1/*1 individuals and weaker in ALDH2*1/*2 individuals (see Figure 1).
