Identifying and Understanding Use Trajectories of Cigarettes and E-Cigarettes

The rising use of e-cigarettes, particularly dual use with cigarettes, necessitates understanding usage patterns and identifying participant and product characteristics that influence behaviors to guide tobacco regulatory decisions. Analysis included adults (≥18 years) from the Population Assessment of Tobacco and Health (PATH) study, a nationally representative, longitudinal cohort. Participants were classified into six e-cigarette/cigarette category at each of waves 1-5 (2013-2019). Latent class analysis (LCA) was performed to identify distinct trajectories of cigarette and e-cigarette use patterns over time in Mplus. Random forest was used to examine the importance of baseline participant and product characteristics in predicting the distinct groups of use trajectories. Within each cigarette class, odds ratios and 95% CIs of the top 15 predictors of e-cigarette use group were calculated in multinomial logistic regression controlling for age, race/ethnicity, and education. Four cigarette-use trajectories were identified: "Cigarette never smokers" (31.3%), "Past experimental cigarette smokers" (23.8%), "Current experimental cigarette smokers" (18.8%), and "Current cigarette smokers" (26.0%). E-cigarette-use trajectories included "Never users" (53.6%), "Past progressors" (32.3%), and "Current progressors" (14.3%). Random forest highlighted common and unique predictors across cigarette-use groups. For example, perceptions of e-cigarette harm relative to cigarettes strongly predicted e-cigarette progression among current smokers, whereas younger age and higher social media use were more relevant among never and past smokers. Distinct cigarette and e-cigarette use trajectories were identified, with predictors varying by cigarette-use group. These findings underscore the importance of targeted regulatory strategies based on subgroup-specific behaviors and characteristics.