Predicting Substance-Specific Impaired Driving in Emerging Adults

Driving while alcohol-, marijuana-, and/or other-drug- impaired (DWI) significantly increases risk of a motor vehicle crash, particularly among youth. While previous research has reported on the prevalence of DWI and risk factors influencing DWI in college students, little is known about DWI risk in those not attending a university after high school (e.g. community school, work force).  Moreover, it is unclear the extent to which environmental/social influences predict substance-specific DWI risk during the transition from high school to post high school. This study examined the prevalence and predictors of alcohol, marijuana, and poly-substance-based DWI in post-high school emerging adults.

The NEXT Generation Study is a nationally-representative cohort study starting with U.S. 10th graders in the 09-10 school year (Wave1, W1). The data (N=2052) from W3 (high school seniors) and W4 (first year post high school) were analyzed. In W3, DWI was assessed with a single overall DWI item which asked participants how many days (ds) they drove after drinking alcohol or using illicit drugs in the past 30 ds (recoded as a dichotomous variable: DWI ≥ 1 d vs. no DWI in the past 30 ds). In W4, three substance-specific DWI items were collected to individually capture driving after alcohol, marijuana, or illicit drug use in the past 30 ds (a 4-group categorical variable was created: marijuana-, alcohol-, poly-substance DWI ≥ 1 d, vs no DWI in the past 30 ds as the outcome variable). Multinomial logistic regressions estimated the association between W4 DWI and W3 covariates (peer/parent influence, father/mother’s monitoring knowledge, drinking/binge drinking, marijuana and other illicit drug use), and W4 environmental/social-context variables (work status, school status, and residence), adjusting for W3 DWI, demographic and design variables.

Proportions of participants reporting DWI slightly increased from W3 (14%) to W4 (15%). Of those not reporting DWI in W3 (N=891), 11% reported DWI in W4; 50.3% of those reporting DWI in W3 reported DWI in W4, indicating a strong propensity for repeat offense.  Results from the multinomial logistic regressions show that W3 marijuana use (Odds Ratio [OR]=10.02, P<.001) and having drunk friends (OR=1.92, P<.001) were associated with W4 marijuana-based DWI. W3 binge drinking (OR=2.16, P<.05) was associated with W4 poly-substance DWI. None of social context variables were significantly associated with W4 DWI in the adjusted models. 

In a nationally-representative sample of emerging adults, previous marijuana use and having drunk friends were predictors of marijuana-based DWI; previous binge drinking was a predictor of poly-substance based DWI. Results suggest that substance use but not social context predicted risk of DWI in youth.