School Social Capital and Tobacco Experimentation Among Adolescents: Evidence From a Cross-Classified Multilevel, Longitudinal Analysis

Abstract

PurposeSchool social capital incorporates the intangible prosocial resources from social networks, including expectations and social norms, found in a school environment. School social capital may influence health behaviors such as smoking. This study examined the association of school social capital with smoking behaviors from childhood into adolescence. MethodsWe used a cohort sampled from three U.S. cities for the Healthy Passages Longitudinal Study of Adolescent Health. The primary outcome was cigarette smoking at grade 10 (Wave 3). The primary predictor of interest was school social capital at grade 5 (Wave 1). We included potential covariates at the individual, school, and neighborhood levels at Wave 1. To account for simultaneous clustering in schools and neighborhoods, cross-classified multilevel models were used. ResultsAfter exclusions and imputations for missing variables, our final sample contained 3,968 students as constituents of 118 schools and 479 neighborhoods. With adjustment for the covariates, school social capital for grade 5 was negatively associated with cigarette smoking in grade 10. We estimated that a 1 standard deviation increase in the school average social capital for grade 5 is associated with an odds ratio of .86 (95% credible interval: .75–.98) for school-level smoking in grade 10. ConclusionsThis study suggests that school social capital in late elementary years is associated with reduced smoking behaviors among adolescents in the U.S. Influencing school social capital through enrichment of positive social norms and parent/teacher expectations may be a useful strategy to reduce adolescent smoking, with long-term implications for adult health.