Social influence and advocacy pathways during a web-based program for adolescent smoking prevention.

Introduction: Exposure to smokers has been identified as a predictor of adolescent tobacco use. Conversely, adolescents who tend to be advocates against smoking may become less likely to initiate smoking themselves. Several digital tobacco prevention programs have been developed to include social strategies. This study aimed to identify (1) whether programs can motivate adolescents to become advocates against smoking, and (2) if being an advocate against smoking and exposure to friends who smoke can predict smoking while controlling for a program's effect. Methods: We conducted a non-prespecified secondary analysis using data from a randomized controlled trial (RCT) with 18-month follow-up. High schools were randomized to either receive ASPIRE or a tobacco education booklet. We conducted a cross-lagged linear path model to allow for reciprocal associations, estimating a two-time-points, three-variable panel model with logistic regression. Results: Receiving ASPIRE was associated with a lower likelihood of smoking, but it did not predict becoming an advocate against smoking or changing adolescents' proportion of friends who smoke. After controlling for the effect of ASPIRE, the study shows that adolescents who were advocates against smoking had a decreased risk of smoking by follow-up, and smoking at baseline significantly predicted having a higher proportion of friends who smoke at follow-up. Discussion: Being an advocate against smoking can be a key predictor of lower odds of smoking, even when controlling for an individual-based intervention. Future research can study the mechanisms and long-term effects of advocacy and incorporate social strategies that can leverage social networks for tobacco prevention.