LAUSR

Abstract Objectives Cocaine use is on the rise and it is comorbid with marijuana use. We examined the association between lifetime cocaine + marijuana polysubstance use (CM PSU) versus cocaine only and lifetime cocaine use disorder (CocUD) and examined the potential mediation by cocaine use patterns. Methods: A total of 2,968 lifetime cocaine users were identified from the National Epidemiologic Survey on Alcohol and Related Conditions-III. Mediation analysis was utilized to examine cocaine use quantity, frequency, and duration as potential mediators in the association between CM PSU and CocUD. A parallel multiple mediator model and a structural equation model were used, respectively, to examine: (1) the individual contribution of cocaine use quantity, frequency, and duration, and (2) combined contribution as a set specified by a latent variable. Results: Cocaine users were divided into 2,782 (93.7%) CM polysubstance users and 186 (6.3%) cocaine only users. CM PSU was associated with decreased risk of CocUD, but after including the mediators, the association was no longer significant. Examined separately, only quantity was found to be a significant mediator over and above frequency and duration, while the latent variable with three cocaine use pattern indicators explained 56.6% of the total association between CM PSU and CocUD. Conclusions: Compared to cocaine only users, CM polysubstance users were less likely to use cocaine heavily; this lower intensity of cocaine use was in turn associated with decreased risk of CocUD. Future research is warranted to determine the nature of the association between CM PSU and reduced CocUD.