The primary phenotype was smoking abstinence at EOT, and the primary predictor was the rs16969968 genotype. We used logistic regression models. Subjects of European ancestry were examined. The following covariates were tested: gender, age, cigarettes per day (in 4 levels: < 10, 11–20, 21–30, ≥31), and treatment (placebo vs. cNRT in the UW-TTURC trial; placebo vs. varenicline in the varenicline trial). Age did not differ across treatment arms and was not significantly associated with cessation outcomes, so it was not included in the final models.
