Identifying prospective subpopulations of combustible and electronic cigarette dual users in the United States via finite mixture modeling

To identify subpopulations of dual users of combustible and electronic cigarettes using current smoking and vaping behaviors.Secondary analysis of baseline data from a randomized controlled trial testing a smoking cessation intervention for dual users. Finite mixture modeling of frequency, quantity, and dependence on combustible and electronic cigarettes was used to identify classes. Demographics and additional smoking and vaping variables were used to further characterize the classes.United States.A total of 2896 adults who smoked weekly for the past year and vaped weekly for the past month.Self-report baseline measures assessed demographics and smoking and vaping behaviors and characteristics including days of use per week, frequency of use within a day, time to first use after waking, urges to smoke, smoking cessation motivation, self-efficacy to abstain from smoking, months since vaping initiation, reasons for initiating and maintaining vaping, and future plans to stop vaping.Eight probabilistic classes were identified and well-defined (relative entropy = 0.95, Lo-Mendell-Rubin adjusted likelihood ratio test P 0.0001; class probabilities 0.89-0.97). In general, classes crossed two levels of smoking with four levels of vaping. The largest class (31%) had relatively high levels of smoking (72% daily, 56% 11+ cigarettes per day [CPD], 96% within 30 minutes of waking) and vaping (74% daily, 100% 20+ electronic-CPD, 74% 30 minutes). The next largest class (27%) had relatively high levels of vaping (93% daily, 100% 20+ electronic-CPD, 82% 30 minutes) and very low levels of smoking (28% daily, 12% 11+ CPD, 0% 30 minutes). The six smaller classes (3%-13%) also had distinct smoking and vaping behaviors. All eight classes exhibited distinguishing characteristics beyond current smoking and vaping behaviors.Dual users of combustible and electronic cigarettes are not a homogeneous population, having eight well-defined prospective subpopulations.