Abstract
Aims: This study assessed the validity of DSM-IV cannabis abuse and dependence criteria in an adolescent general population sample and evaluated the usefulness of additional cannabis use indicators was evaluated. Design/Setting: Data came from the 2008 Survey on Health and Consumption during the Day of Defense Preparation (ESCAPAD), a cross-sectional self-administered survey conducted in France. Participants: The analytical sample comprised 3,641 adolescents aged 17 to 19 years who reported cannabis use in the past 12 months. Measurements: To assess DSM-IV criteria of cannabis abuse and dependence, the Munich Composite International Diagnostic Interview (M-CIDI) was used. As additional cannabis use indicators, daily use, use when alone and use before midday were assessed. Confirmatory factor analyses and two-parameter logistic IRT models were run. Differential item functioning was assessed using the IRT log-likelihood ratio approach. Results: A one-factor model comprising both abuse and dependence criteria showed the best fit to the data. Abuse item legal problems showed the greatest severity whereas dependence items larger/longer and tolerance were found least severe. Discriminatory power was lowest for impaired control and legal problems. Additional cannabis use indicators increased the precision of the overall DSM-IV criterion set. Gender-based differential item functioning was observed for items tolerance, withdrawal and use before midday. Conclusion: The current DSM conceptualization with two distinct and graded diagnostic classes has limited validity among adolescents. In forthcoming revisions of the classification system, several existing criteria should be revised or dropped, new indicators of substance use disorders should be included and gender should be considered.