Explaining the Decoupling of Ecologic Trends in Adolescent Marijuana Use and Perceived Harm of Marijuana Use: Evidence from Washington State

Introduction: Over the past 40 years, Monitoring the Future data show that prevalence of adolescent marijuana use and perceived harmfulness of marijuana use mirror one another. Recently, during a period of increasing liberalization of marijuana laws, trends in answers to marijuana use and perceived harm questions have diverged, with prevalence of adolescent marijuana use dropping slightly since 2010 while the percentage of youth reporting that marijuana use is harmful dropped monotonically since 2006, including sizable decreases since 2010. We investigated two potential explanations: 1) the strength of association between perceived harm and marijuana use has weakened over time and 2) changes in other marijuana risk factors or in alcohol use or smoking offset expected increases due to low perceived harm.

Methods and Results: Data were from surveys of 30,365 10^thgrade students in Washington State. Surveys were administered biennially and spanned 2000-2014. We used logistic regression models to examine changes in associations between marijuana use and predictors of marijuana use over time and test whether other predictors of marijuana use could account for decoupling in trends for perceived harm and marijuana use. All models included gender and race/ethnicity as covariates and adjusted standard errors for clustering of students within schools. The strength of associations between marijuana use and marijuana-use predictors were positive and statistically significant and varied little across time; if anything, the association between perceived harm and marijuana use grew stronger in recent years. Findings suggest spillover effects from decreases in alcohol and cigarette use counteracted the influence of increased marijuana risk factors and can account for decoupling in trends for use and perceived harm.

Conclusions: The findings caution against misinterpreting the recent decoupling in trends for marijuana risk factors and marijuana use. This decoupling does not mean we should abandon prevention efforts to inform adolescents about the potential harms of marijuana use. Results also suggest that reductions in adolescent alcohol use and smoking can have beneficial spillover effects on marijuana use, perhaps mitigating the effects of liberalization of marijuana laws.