024). In longitudinal analyses of each abstinence outcome stratified by genotype Seliciclib Cdc2 (Supplementary Tables 6 and 7), we observed significant association of (a) treatment in individuals with a VNTR L+ genotype for both abstinence outcomes (point prevalence abstinence [OR = 2.74, 95% CI: 1.14�C6.59, p = .025] and continuous abstinence [OR = 3.15, 95% CI: 1.12�C8.88, p = .030] respectively); (b) time in individuals with a VNTR SS genotype at 6MO with point prevalence abstinence and at 6MO and 12MO with continuous abstinence; (c) gender in individuals with a VNTR SS genotype with continuous abstinence; and (d) age (age squared) in individuals with a VNTR L+ genotype with continuous abstinence. We also observed three principal components of population genetic variation significantly associated with continuous abstinence in individuals with a SS genotype (ps < .

033). The effect sizes of treatment in individuals with a VNTR SS genotype in multivariate (EOT, 6MO, and 12MO) and longitudinal analyses of point prevalence and continuous abstinence outcomes were nonsignificant: OR = 1.39, 95% CI: 0.78�C2.48, p = .262; OR = 1.29, 95% CI: 0.69�C2.41, p = .428; OR = .96, 95% CI: 0.48�C1.90, p = .896; OR = 1.48, 95% CI: 0.80�C2.77, p = .215; OR = 1.92, 95% CI: 0.91�C4.06, p = .087; OR = 1.42, 95% CI: 0.58�C3.47, p = .441; OR = 1.38, 95% CI: 0.78�C2.44, p = .27; and OR = 1.47, 95% CI: 0.78�C2.75, p = .23, respectively. The power to detect the gene by environment effect reported by Leventhal et al. (2012) with the abstinent (case) and nonabstinent (control) sample sizes, dominantly coded genotype prevalence, and treatment prevalence of the Lerman et al.

(2003) sample, and the average placebo abstinence rate of 13.8% from 80 placebo arms (Fiore et al., 2008), was 11.5%. Under the same assumptions, a sample size of 3,362 individuals would be required to detect the gene by treatment interaction effect reported by Leventhal et al. (2012) with 80% power. DISCUSSION In analyses of the relations between bupropion treatment and VNTR genotype in N = 416 self-identified White treatment-seeking smokers, we observed statistically significant effects of treatment in multivariate analyses of both abstinence outcomes at EOT and at 6MO, and in longitudinal analyses of both abstinence outcomes, but only in individuals with a VNTR L+ genotype.

We did not observe statistically significant associations of genotype or genotype by treatment interaction. Time was observed to be significantly associated with both abstinence outcomes, as expected. There Drug_discovery were a few covariate associations with both abstinence outcomes, including principal components of population genetic variation. The magnitude of the statistically significant association of active treatment with both abstinence outcomes was not significantly different from the magnitude of effects previously observed in samples not stratified by DRD4 VNTR genotype (Brown et al., 2007; Hurt et al.