Consistent with the DSM-IV-TR diagnostic structure and coding, descriptive data for overall prevalence rates of comorbid cannabis abuse and dependence in the study sample were first evaluated. All subsequent study analyses utilized a composite variable representing the presence vs. absence of any CUD, to be consistent with the existing literature [6,7,19,20] and recent collapsing of substance abuse and dependence diagnoses into singular substance use disorders in DSM-5. Independent t-test and chi-square analyses were used to evaluate demographic and clinical differences between patients with a CUD and those without. For any significant differences identified in the univariate analyses, multivariate logistic regression analyses were used to evaluate which characteristics remained significantly associated with a comorbid CUD diagnosis, controlling for the presence of other SUDs. Within these multivariate models, the presence of AUDs and nicotine dependence were combined with non-cannabis drug use disorders in the coding of other comorbid SUDs, with the exception of when they were evaluated as separate independent variables in their respective multivariate models. Thus, multivariate analyses allowed for the evaluation of the unique associations between patient characteristics and CUDs,
