E-cigarette awareness, perceived harmfulness, and ever use among U.S. adults

Abstract

The overarching aims of this study are to (a) estimate and update knowledge on rates and predictors of awareness, perceived harmfulness, and ever use of e-cigarettes among U.S. adults and (b) to utilize that information to identify risk-factor profiles associated with ever use. Data were collected from the 2015 Health Information National Trends Survey (N=3738). Logistic regression was used to explore relationships between sociodemographics (gender, age, race/ethnicity, sexual orientation, educational attainment, income, and census region), current use of other tobacco products (conventional cigarettes, cigars, and smokeless tobacco), ever use of alternative products (hookah, pipes, roll-your-own cigarettes, and snus) and e-cigarette awareness, perceived harm, and ever use. Classification and regression tree (CART) modeling was used to examine risk-factor profiles of e-cigarette ever use. Results showed that most respondents were aware of e-cigarettes (83.6%) and perceived them to be not at all or moderately harmful (54.7%). Prevalence of e-cigarette ever use was 22.4%. Current cigarette smoking and ever use of alternative tobacco products were powerful predictors of use. Other predictors of use of e-cigarettes were age, race/ethnicity, and educational attainment. Awareness and perceived harm were significant predictors among particular smoker subgroups. Fifteen risk profiles were identified across which prevalence of e-cigarette use varied from 6 to 94%. These results underscore the need to continue monitoring patterns of e-cigarette use. They also provide new knowledge regarding risk-profiles associated with striking differences in prevalence of e-cigarette use that have the potential to be helpful when considering the need for or impact of e-cigarette regulatory policies.

Keywords: Awareness; Classification and regression tree (CART); Electronic cigarettes; Ever use; Perceived harm; Prevalence.

Figure 1

A pruned, weighted classification and regression tree (CART) model of associations between ever use of e-cigarettes and the following risk factors in the U.S. adult population: cigarette smoking status, ever use of hookah, age, educational attainment, race/ethnicity, ever use of snus, ever use of roll-your-own cigarettes, annual household income, awareness of e-cigarettes, census region, cigar smoking status, perceived harm of e-cigarettes, smokeless tobacco use, and ever use of pipe filled with tobacco. Squares (nodes) represent rates of ever use of e-cigarettes among for the entire population (top-most node) or population subgroups (all other nodes). Nodes also list the proportion of the adult population represented. Using the parent node (top-most node) as an example, 79% of the population had never used e-cigarettes and 21% had used them at least once, and this node represent 100% of the US non-institutionalized adult population. Lines bellow the nodes represent the binary “yes”-“no” branching around a particular risk factor and risk factor levels, with subgroups in whom the risk factor/level is absent moving leftward and downward and those whom it is present moving rightward and downward for further potential partitioning. The bottom row comprises terminal nodes (i.e., final partitioning for a particular subgroup).