Finally, binary logistic regression models showed that after controlling for age at interview and bipolar disorder type 1 or 2 diagnosis, the clinical history variables that significantly predicted the presence of moderate and severe risk of problem gambling over its absence were a history of rapid cycling (OR = 2.63, 95% CI 1.29–5.34, P = 0.008), a history of suicidal ideation or attempt (OR = 3.44, 95% CI 1.21–9.73, P = 0.02) and younger age at illness onset (OR = 0.94, 95% CI 0.90–0.98, P = 0.002).
