Abstract
BACKGROUND: Despite the growing prevalence of cannabis use and associated consequences among U.S. young adults, relatively little is known about precisely what level of marijuana involvement may be problematic. METHOD: With this study we sought to identify empirically-derived cut-scores for the Marijuana Consequences Questionnaire (MACQ) that would distinguish among levels of cannabis risk in a sample of young adult college students (N = 496). We also examined how these levels of cannabis risk corresponded to a variety of indicators of cannabis involvement, including frequency of use, intoxication, other measures of cannabis consequences, and indicators of more severe cannabis involvement (e.g., physiological dependence, loss of control over use, cannabis use disorder). RESULTS: Receiver operating characteristic analyses yielded cutoffs that distinguished among three distinct levels of risk, "Low", "Moderate", and "High". These empirically derived cut scores showed strong overall differentiation among classifications, with good sensitivity and specificity. MACQ-based risk levels were validated across several indices of cannabis involvement. Cutoffs differed across genders. CONCLUSIONS: Findings offer a new application for the MACQ, allowing for the identification of those at greatest risk. As such, this measure may be used to facilitate appropriately targeted intervention.