The interaction between the ADH1B variant and peer drinking was tested by adding GxE product term to models of drinking milestones in adolescent drinkers (N=1,550), which illustrated a significant statistical interaction for first intoxication (p=.002) and first DSM-5 symptom (p=.01) (Table 2). Among individuals who reported none/few best friends drinking, the ADH1B GA/AA genotypes had a strong protective effect for first intoxication (HR=0.56, 95% CI 0.41–0.76) and first DSM-5 symptom (HR=0.45, 95% CI 0.26–0.77). In individuals who reported most/all best friends drinking, however, this protective effect was not observed for either first intoxication (HR=1.16, 95% CI 0.82–1.65) or first DSM-5 symptom (HR=1.03, 95% CI 0.73–1.45), as illustrated by the point estimates close to 1. Figure 2 more clearly illustrates this GxE interaction by presenting the survival estimates.
