These analyses were repeated with sex, age, race, depression history, typical smoking rate, years of smoking, baseline FTND score added to the models. No significant sex difference was found and including these covariates in the models did not significantly alter the results.
