Unweighted data were used to report the descriptive statistics (mean, SD for continuous variables and %, (n) for categorical variables) of smoking and smoking cessation attempt characteristics for drinkers at risk of alcohol dependence, drinkers not at risk and non-drinkers. Smoking and smoking cessation attempt characteristics were modelled as the outcome in a series of (linear, logistic or multinomial logistic, as appropriate) regression models with the full AUDIT score range among drinkers as the predictor, adjusted for survey year. As for RQ1, natural, cubic splines were applied to allow for non-linear trends.
