Electronic nicotine delivery systems (ENDS) are the most commonly used tobacco product among young adults aged 18–25, raising concerns about nicotine dependence and health risks. While research has established that young adults with more ENDS-using peers are more likely to subsequently use these products, the mechanisms explaining this peer influence remain poorly understood. The Theory of Reasoned Action suggests that peer influence operates through changes in attitudes and subjective norms, yet prior research examining this pathway has produced mixed results. Specifically, studies of social enhancement outcome expectancies have found conflicting longitudinal associations, harm perceptions have shown inconsistent predictive relationships, and social normative beliefs have yielded varied findings for both dating willingness and social acceptability as predictors. Therefore, the current study aims to test whether a mediation model involving social enhancement outcome expectancies, harm perceptions, willingness to date ENDS users, and social acceptability beliefs can explain the association between peer ENDS use and subsequent individual ENDS use two years later among young adults.
The study conducted a survey among 1,721 young adults who had experience with ENDS use and participated in the final three waves of Project M-PACT. Data collection employed a longitudinal design that measured peer ENDS use and past 30-day ENDS use at baseline, assessed social enhancement outcome expectancies, harm perceptions, and social normative beliefs at one-year follow-up, and re-measured past 30-day ENDS use at two-year follow-up. Data analysis utilized structural equation modeling to test mediation effects, with 5,000 bootstrap resamples employed to assess the statistical significance of indirect effects while controlling for relevant covariates.
The study revealed that past 30-day ENDS use significantly increased from 15.7% at baseline to 19.1% at the two-year follow-up. While baseline peer ENDS use was not directly associated with individual past 30-day ENDS use at the two-year follow-up, peer ENDS use demonstrated significant associations with all four mediating variables at one-year follow-up (social enhancement outcome expectancies, harm perceptions, willingness to date ENDS users, and social acceptability beliefs). Among these mediators, only willingness to date someone who uses ENDS showed a significant association with past 30-day ENDS use at the final wave. Bootstrap analysis indicated that only the indirect effect through willingness to date someone who uses ENDS was statistically significant (b = 0.03, 95% CI: [0.02, 0.04]), whereas the indirect effects through social enhancement outcome expectancies, harm perceptions, and social acceptability beliefs were all non-significant. As a result, the influence of peer ENDS use on subsequent individual ENDS use was mediated through willingness to date ENDS users, a component of social normative beliefs, even after controlling for sociodemographic characteristics and baseline substance use.
Takeaway: Peer ENDS use influences subsequent individual ENDS use through increased willingness to date someone who uses ENDS.