Experimentation with tobacco during adolescence as a factor influencing treatment of smoking in adulthood. A retrospective cohort

ABSTRACT BACKGROUND: There are still few studies on predictors of smoking cessation in Brazilian samples. Experimentation with tobacco during adolescence (ETA) may be one of the important predictors. OBJECTIVE: This study aimed, within the context of a treatment-seeking group of subjects, to test the hypothesis that ETA negatively affects the outcome of smoking cessation during adulthood. DESIGN AND SETTING: Retrospective (historic) cohort study conducted at a psychosocial care center in São Paulo, Brazil, between 2007 and 2010. METHODS: Data on sociodemographics, smoking and medical profiles were obtained through self-report questionnaires that were completed at the baseline and at any follow-up appointment. Logistic regression models were constructed to describe factors associated with the outcome of smoking cessation, measured according to the self-reported four-week success rate among 367 outpatient smokers. RESULTS: ETA was found to be associated with not quitting smoking through the treatment (odds ratio = 0.57; 95% confidence interval = 0.33-0.96; P