We examined the effect of youth DRD4 status on susceptibility to the intervention. Analysis of the unconditional model showed that the largest share of the variance in past-month substance use was within individuals (21%), reinforcing the importance of within- individual variation. As Table 2 shows, a significant overall treatment effect on past-month substance use emerged (γSAAF = -.20, p < 0.05). Main effects for both the DRD4 7-repeat allele (γ g = .16, p < .10) and the short 5HTT allele (γ g = .20, p < .10) approached significance. A significant gene × intervention interaction emerged for DRD4 (γS×G = -1.04, p < .01) but not for 5HTT (γS×G = - 0.18, NS). To explicate the significant interaction with DRD4, we plotted the means across time for past-month substance use for each of the four subgroups. As Figure 1 shows, SAAF reduced growth in past-month substance use only for youths with one or two copies of the 7-repeat allele for DRD4. The effect size of 0.26 for the interaction of genotype with intervention accounted for 2% of the explained
