Mutual Pathways between Peer and Own E-Cigarette Use among Youth in The United States: A Cross-Lagged Model

AbstractBackground Electronic cigarettes (e-cigarettes) have become the most common tobacco product used among adolescents in the United States (US). Prior research has shown that peer e-cigarette use was associated with increased risk of own e-cigarette use. Nonetheless, there is little empirical evidence on the directionality of these associations - if peer use predicts own use (peer influence) or if own use predicts peer use (peer selection). Methods We estimated the association between peer and own e-cigarette use among US adolescents 12–17 years of age. We used the fixed-effects model to hold constant observed and unobserved time-invariant variables and the cross-lagged model to investigate the mutual relationship between peer and own e-cigarette use over time using data from a population-based longitudinal study, Population Assessment of Tobacco and Health. Stratified analyses were conducted by sex and age subgroups. Results Having best friends who use e-cigarettes was strongly associated with the use of e-cigarettes among adolescents (OR = 17.5; 95% CI = 15.1 to 20.3 for 12–17-year-olds). This association weakened but remained statistically significant when time-invariant variables were controlled for using fixed-effect model (OR = 4.8; 95% CI = 4.0 to 5.6). Results from cross-lagged model showed a statistically significant predicting path leading from peer use at the prior time point to own use at the following time point, but not vice versa. Conclusions We found strong relationships between peer e-cigarette use and own e-cigarette use at both between- and within-individual levels. Peer influence paths were more robust than peer selection paths for e-cigarette use. Incorporating peers into prevention and intervention programs may help enhance these strategies.