daily smoking within 20 years since initiation were considered as long‐term survivors. In order to account for this, we first adopted and implemented a mixture cure model to analyze this transition in the context of survival framework (Yu and Peng 2008). The results, however, showed an inflated false positive error rate (data not shown) probably due to the potential inaccuracy of the variance estimation as shown in previous simulation studies (Yu and Peng 2008). We therefore addressed this issue by dichotomizing the survival time variable and classified those becoming a daily smoker within 1 year after smoking the first cigarette as rapid progression (N = 450), and those becoming a daily smoker after > 1 year as slow progression (N = 1079). We then conducted the association analysis on this binary variable with the logistic linear mixed effects model implemented in the glmmML R package (Broström and Holmberg 2011).
