Across these types of studies, there are different rates of CU in bipolar disorder. While the current study had high rates of CUDs in bipolar I and II patients (any CUDs in 29.9% of bipolar II and 40.8% in bipolar I subjects). The study of Bahorik et al. [31] reported CU in 27%, Blanco et al. [32] 3.4% in non-bipolar individuals and van Rossum et al. [21] in 12.7% of their samples. However, in this study, compared to previous investigations, data from a highly affected and comorbid sample (CUD and bipolar disorders) are analyzed compared to cannabis use (CU and bipolar disorders) in the previous studies [21,30,31,32]. Thus, individuals in the current sample are affected by multiple substance use and mental comorbidities and have a high rate of comorbid CUD. Further, two previous studies [30,32] used epidemiological samples while clinical patients were enrolled in two other prospective investigations [21,31]. Also, assessment methods differed across prospective studies, using structured interviews and the same sample [30,32], general clinical assessments, self-reports and questionnaires [21] or self-reports, clinical interviews and psychopathology questionnaires [31]. The
