Comparative analysis of affective temperament in patients with difficult-to-treat and easy-to-treat major depression and bipolar disorder: Possible application in clinical settings.

BACKGROUND: Difficult-to-treat major depressive disorder (MDD-DT), which involves antidepressant refractoriness or antidepressant-related adverse psychiatric effects, is bipolar in nature; therefore, it may share common temperamental features with bipolar disorder. To examine this hypothesis, affective temperament was compared between MDD-DT, easy-to-treat major depressive disorder (MDD-ET), and bipolar disorder.

METHODS: Affective temperament was measured in 320 patients (69, 56, and 195 with MDD-ET, MDD-DT, and bipolar disorder, respectively) using the self-rated questionnaire version of the Temperament Evaluation of Memphis, Pisa, Paris, and San Diego (TEMPS-A), with between-group differences examined using multiple logistic regression analysis controlling for confounders. Optimal cut-off points for TEMPS-A scores to discriminate between diagnostic groups were determined using receiver-operating characteristic analysis.

RESULTS: Of the five temperamental domains, the mode for cyclothymic temperament score was highest, followed by those of bipolar disorder, MDD-DT, and MDD-ET. The cyclothymic temperament score discriminated significantly between bipolar disorder and MDD-DT (odds ratio [OR]: 1.12, 95% confidence interval [CI]: 1.04-1.20, p=0.0022), MDD-DT and MDD-ET (OR: 1.15, 95% CI: 1.01-1.31, p=0.0334), and bipolar and major depressive disorders (OR: 1.17, 95% CI: 1.07-1.28, p=0.0003). Optimal cut-off points for the cyclothymic temperament scores to discriminate between bipolar disorder and major depressive disorder and MDD-DT and MDD-ET were 9 (sensitivity: 64.6%, specificity: 76.0%) and 6 (66.1%, 62.3%), respectively.

CONCLUSIONS: MDD-DT has a quantitatively stronger bipolar temperamental feature, cyclothymic temperament, relative to that of MDD-ET. Cut-off points determined in this study could be clinically helpful. Because of our study design, longitudinal changes in temperamental scores during treatment cannot be fully excluded.