Predictors of electronic cigarette dependence among non-smoking electronic cigarette users: User behavior and device characteristics

Introduction: ECIGs differ in their ability to deliver nicotine to the user and, consequently, they may differ in their ability to produce dependence. This study examined individual device characteristics, device type, and user behaviors as predictors of ECIG dependence in a sample of never-smoking ECIG users.Methods: Participants (N = 134) completed an online survey that assessed demographics, ECIG use behavior, and ECIG dependence as measured via the Penn State Electronic Nicotine Dependence Index (PSECDI) and E-cigarette Dependence Scale (EDS-4). Participants uploaded a picture of their personal ECIG device/liquid, which was coded by raters to identify product features. Multivariable linear regressions examined device characteristics (e. g., adjustable power, nicotine concentration) and device type (e.g., vape pen, mod, pod, modern disposable) as predictors of dependence controlling for demographics and user behaviors (e.g., ECIG use duration and fre-quency, other tobacco use).Results: Longer durations of ECIG use and more use days/week were associated significantly with higher PSECDI (8's = 0.91 and 1.90, respectively; p's < 0.01) and EDS-4 scores (8's = 0.16 and 0.28, respectively; p's < 0.01). Higher nicotine concentrations were associated with higher PSECDI scores only (8 = 0.07, p =.011). Dependence scores did not differ as a function of ECIG device types after controlling for covariates.Conclusions: ECIG dependence was observed among the never-smoking ECIG users in this sample, regardless of their ECIG device/liquid features. Findings suggest that regulatory efforts aimed at reducing the dependence potential of ECIGs in never smokers should focus on overall nicotine emissions rather than product features.

Automated summary

This study examined the predictors of ECIG dependence among never-smoking ECIG users and found that longer durations of ECIG use and more use days per week were significantly associated with higher dependence scores. Higher nicotine concentrations were only associated with higher dependence scores on one of the measures. Device characteristics and types did not predict dependence after controlling for covariates.