Do predictors of abstinence change in the medium- and long-term follow-up of smokers who have quit smoking? A prospective cohort study.

INTRODUCTION: We hypothesize that the predictors of smoking cessation in the medium-term are not the same as in the long-term of follow-up. The aim of this study was to identify predictors for smoking cessation (continuous abstinence) and determine if these are maintained over time. METHODS: This is an observational longitudinal (prospective cohort) multicenter study conducted in daily clinical practice in Spain. Patients were consecutively enrolled as they attended consultations, and all patients followed for 12 months. To identify predictors of smoking cessation (at 24 and 52 weeks post-cessation) we have collected sociodemographic and clinical data, smoking consumption characteristics, and psychological and physical dependence variables. Multivariate logistic regression models were fitted. The analysis was by intention to treat. RESULTS: A total of 337 participants were considered for the study. Predictors of smoking cessation at 24 weeks were baseline weight (AOR=1.02; 95% CI: 1-1.03), not having made a previous quit attempt (AOR=2.72; 95% CI: 1.44-5.15), lower sedation levels on the psychological dependence test (AOR=1.78; 95% CI: 1.06-2.97), and adherence to treatment (AOR=8.03; 95% CI: 3.85-16.73). At 52 weeks, predictors of smoking cessation were being male (AOR=2.38; 95% CI: 1.35-4.18), low self-efficacy (AOR=2.60; 95% CI: 1.36-5.00), not having made a previous quit attempt (AOR=5.06; 95% CI: 2.20-11.66), lower sedation levels on the psychological dependence test (AOR=1.96; 95% CI: 1.13-3.40), and adherence to treatment (AOR=12.03; 95% CI: 4.14-34.94). These last three predictors were those that were maintained between 24 and 52 weeks of follow-up. CONCLUSIONS: Not having previous attempts to quit smoking, lower sedation levels in the psychological dependence test, and having greater adherence to treatment have been maintained as predictors of quitting over time.