(1) Background: Bipolar disorder (BD) is divided into type I (BD-I) and type II (BD-II). Polarity at onset (PO) is a proposal to specify the clinical course of BD, based on the type of the first episode at disorder onset-depressive (D-PO) or manic (M-PO). At the same time, affective temperaments represent preexisting variants of the spectrum of affective disorders. Our objectives were to investigate the hypothesis that temperament may exert an influence on PO, and that this factor can serve as an indicator of the forthcoming course of the disorder, carrying significant therapeutic implications. (2) Methods: We included 191 patients with BD and examined clinical variables and temperament; the latter was assessed using the short version of the Temperament Evaluation of Memphis, Pisa, Paris, and San Diego-Auto-questionnaire (TEMPS-A-39-SV). We tested the associations between these variables and PO using standard univariate/bivariate methods followed by multivariate logistic regression models. (3) Results: 52.9% of the sample had D-PO and 47.1% had M-PO. D-PO and M-PO patients scored higher for dysthymic and hyperthymic temperaments, respectively (p < 0.001). Also, they differed in BD subtypes, age at first affective episode, illness duration, number of depressive episodes, seasonality, suicide risk, substance use, lithium, and benzodiazepine use (p < 0.05). Only BD-II and age at first depressive episode were predictors of D-PO, whereas BD-I, age at first manic/hypomanic episode, and hyperthymic temperament were predictors of M-PO (p < 0.01). (4) Conclusions: Our findings point to the importance of carefully assessing temperament and PO in patients with BD, to better predict the clinical course and tailor therapeutic interventions to individual patients' needs.
Keywords: affective temperaments; bipolar disorder; depressive polarity; manic/hypomanic polarity; polarity at onset.