Paul T. Enlow

Date of Graduation


Document Type


Degree Type



Eberly College of Arts and Sciences



Committee Chair

Christina L Duncan

Committee Co-Chair

Melissa Blank

Committee Member

Geri Dino

Committee Member

Amy Herschell

Committee Member

Nicholas Turiano


Use of electronic cigarettes (e-cigs, electronic vapor products) rose dramatically in the United States over the past decade and a half, most notably among high school students. This increase in use, coupled with concerns about renormalization of smoking and potential health risks, has raised concern that e-cigs represent a growing public health risk. Unfortunately, there is limited research on risk factors for e-cig use among teens, and many studies have methodological flaws that are unaddressed. The primary aim of the current study is to apply an empirically tested theory of health-risk behaviors in adolescents, Problem Behavior Theory (PBT), to lifetime- and current-use of electronic cigarettes. The second aim was to identify individual risk factors for different levels of e-cig use. A total of 519 high school students between the age of 13 and 19 years (M age = 15.99; 57.9% female) were recruited from four schools and one adolescent medicine clinic. Students completed a packet of questionnaires during their regularly scheduled classes or while attending a clinic appointment. Structural Equation Modeling and logistic regressions were used to test the primary and secondary aims, respectively. Results partially supported the hypotheses of the first aim. The PBT structural model demonstrated good fit, but not all 5 domains were associated with e-cig use. The Biology/Genetics, Perceived Environment, Personality, and Behavior latent variables were significant predictors of lifetime e-cig use. The Personality and Behavior latent variables predicted current e-cig use; The Social Environment construct was not associated with either outcome variable. Results from binomial and multinomial logistic regressions identified risk- (cigarette, alcohol, and marijuana use; modeling of smoking; and extraversion) and protective- (self-efficacy to resist using e-cigs, perceived costs of using e-cigs) factors of lifetime-only e-cig use. Results from logistic regressions also identified risk (marijuana use, modeling of smoking) and protective factors (perceived costs, smoking self-efficacy) for past-30-day e-cig use relative to lifetime-only use. Overall, results suggest that personality and behavioral variables were the strongest predictors of e-cig use. These results can be used to inform substance use screening and help develop educational prevention efforts.