Methods: Repeated cross-sectional National College Health Assessment-II surveys were conducted at colleges and universities from 2008 to 2016. Undergraduates ages 18-26 years from two large public Oregon institutions (n = 7,412) and 123 non-RML institutions (n = 274,340) reported on drunk driving and indicators of violence and risky sexual behavior. Students’ perceptions of the effect of alcohol and other drugs on their academic performance were also assessed. Mixed-effects regressions controlled for clustering of participants within institutions and accounted for individual-, context-, and institution-level factors, as well as secular changes in each outcome from 2008 to 2016.
Results: RML was significantly associated with decreases in all three indicators of violence, e.g., recent perpetration of physical aggression, β (SE) = -.33 (.15), p = .03, Odds Ratio (OR) = .72. RML was negatively associated with two of four indicators of risky sexual behavior (e.g., unintentional pregnancy in the past year for females, β (SE) = -.33 (.11), p = .003, OR = .72). RML was not significantly associated with drunk driving (β (SE) = .10 (.09), p > .05). RML was negatively associated with self-report of alcohol (β (SE) = -.30 (.14), p = .04, OR = .74) interfering with grades, but not other drugs (β (SE) = .01 (.13), p > .05).
Conclusions: Substance use increases college students’ risks for violence and negative sexual health outcomes. The finding that RML is associated with decreases in such outcomes may inform policy decisions about marijuana legalization. Future research should investigate whether changes in marijuana and/or alcohol use mediate the observed effects; whether such trends are replicated in other Oregon samples and in other RML states; and how health risk outcomes linked to marijuana (e.g., driving after using marijuana) may have changed after RML.