Predominant polarity in bipolar disorders: Further evidence for the role of affective temperaments

Background: Literature suggests bipolars may differ in several features according to predominant polarity, but the role of temperaments remains controversial. Methods: The EPIDEP study was designed to identify bipolar patients among a large sample of major depressives. Only bipolars were included in the current study. Patients were subtyped as predominantly depressive (PD) and predominantly manic and hypomanic (PM) according to a broad (more episodes of a given polarity) and a narrow (2/3 of episodes of one polarity over the other) definition, and compared on their characteristics. Results: Among 278 bipolars, 182 (79.8%) could be subtyped as PD and 46 (20.2%) as PM (broad definition); the respective proportions were of 111 (81.6%) and 25 (184%) using narrow definition. Expanding the definition added little in detecting differences between groups. Compared to PDs, PMs showed more psychosis, rapid cycling, stressors at onset family history of affective illness, and manic first episode polarity; they also received more antipsychotics and lithium. The PDs showed more chronic depression, comorbid anxiety, and received more antidepressants, anticonvulsants and benzodiazepines. The following independent variables were associated with manic/hypomanic predominant polarity: cyclothymic temperament, first hospitalization <= 25 years, hyperthymic temperament and alcohol use (only for broad definition). Limitation: Cross-sectional design, recall bias. Conclusions: Study findings are in accord with literature except for suicidality and mixicity which were related to predominant mania, and explained by higher levels of cyclothymic and hyperthymic temperaments. Temperaments may play a key role in the subtyping of bipolar patients according to predominant polarity, which warrants confirmation in prospective studies. (C) 2015 Elsevier B.V. All rights reserved