Predictors of 10-year smoking abstinence in smokers abstinent for 1 year after treatment

Aims To identify factors predicting long‐term relapse to smoking in people attending smoking treatment services who have maintained at least 1 year abstinence. Design Observational, prospective study with multiple logistic regression used to model predictors of relapse between 1 and 10 years from cessation using variables measured pre‐cessation. Setting and participants Among smokers receiving behavioural support for cessation in a clinic in Spain, in some cases with nicotine patches or bupropion, 366 had remained abstinent after 1 year of follow‐up and were included into the study. Measurements Predictive measures (disease history, psychological disorder, age of starting smoking, years of smoking, cigarette dependence and smoking cessation treatment used) were obtained at the time of the quit attempt, and ‘failure’ (defined as reported smoking, loss to follow‐up, died or an expired air carbon monoxide reading of > 5 parts per million) was assessed 10 years later. Findings At follow‐up, abstinence status was confirmed in 50.5% (n = 185) of participants, while 21.0% (n = 77) reported that they had resumed smoking, and 28.5% (n = 104) were lost to follow‐up (also counted as having resumed smoking). In the multiple regression model, the main factor that predicted relapse had a psychological disorder (odds ratio = 1.85, 95% confidence interval = 1.13–3.05; P