Association for Public Policy Analysis & Management

Introduction: Electronic cigarettes (e-cigarettes) were originally introduced to the U.S. in 2007 and marketed as a cessation device for adult smokers, however they continue to grow in popularity among both youth and adults as an alternative tobacco product. Policy decisions about regulating alternative tobacco products require evidence about their role and variation in use among different populations, particularly adult smokers, who are at risk for smoking-related morbidity and mortality.

Research Objective: The purpose of this study was to investigate whether nicotine dependence was associated with 1) variations in heavy smoking over time among dual users of conventional and e-cigarettes and 2) whether nicotine dependence was associated with use of e-cigarettes among adult smokers.

Study Design: This secondary data analysis used data from Wave 1 (October 2013 - December 2014) and Wave 2 (October 2014 - October 2015) of the PATH Study, a longitudinal, nationally-representative cohort study of over 46,000 U.S. youth and adults. The study population for this analysis was 2,232 current, established, CC smokers who remained established CC-only users across waves, 414 "established dual users" who used both CC and e-cigarettes during both waves, and 277 “new dual users” who remained CC users but began regularly using e-cigarettes between Waves 1 and 2. Weighted logistic regression models offered adjusted odds-ratios (AOR) assessing the association between nicotine dependence in Wave 1 and variation in heavy smoking behavior in Wave 2. Weighted, multinomial logistic regression models offered adjusted relative risk ratios (ARRR) assessing statistically significant differences in group membership (CC-only use, established dual use, and new dual use), adjusted for nicotine dependence in Wave 1 (binary measure based on a high/low split of total score), plans to quit smoking within the next 12 months (binary measure based on Wave 1 quit intent), sex, age, race, income, and educational attainment.

Results: Established dual use was not significantly associated with variation in heavy smoking in Wave 2, whereas new dual use was associated with decreased odds of heavy smoking in Wave 2 (AOR=0.73, p<0.05), after controlling for covariates. High nicotine dependence in Wave 1 and Wave 1 heavy smoking were both associated with higher likelihood of heavy smoking in Wave 2 (AOR=1.67, 3.24 respectively, p<0.01. There were also significant differences among different demographic groups in modeling variation in Wave 2 heavy smoking.

Compared to CC-only users, planning to quit smoking (ARRR=1.64, p<0.05) and higher nicotine dependence scores (ARRR=1.93, p<0.01) were associated with higher likelihood of current dual use. There was no association found between nicotine dependence or intent to quit and group membership for new dual use (p>0.05).

Conclusions: This study provided evidence nicotine dependence is significantly associated with both smoking behavior and alternative tobacco use over time. Results such as these illustrate the complicated nature of the relationship between alternative tobacco products and conventional cigarettes for smokers, i.e. that these products can act as substitutes for each other over time, therefore policymakers need to consider comprehensive tobacco policy strategies rather than focusing on either combustible or alternative tobacco products in policy interventions.