−.30, t = −1.88, adjusted r2 = .06, stepwise F (1,36) = 3.52, p = .07). With the addition of OPRM1 (Step 2) the model accounted for 14% of the variance in peak BrAC (adjusted r2 = .14, stepwise F (2,35) = 4.00, p = .03), with genotype accounting for significant incremental variance in BrAC (r2Δ = .10, F change (1,35) = 4.16, p < .05). With sex and genotype estimated simultaneously, the parameter estimate for sex was not significant (β = −.23, t = −1.49, p = .15) but remained significant for OPRM1 (β = .32, t = 2.04, p < .05). The same findings were obtained in a model examining number of alcohol requests. Finally, a parallel set of regression models confirmed that the OPRM1 effect held while controlling for racial background (data not shown).
