Objectives: To examine hypomanic, cyclothymic and hostile personality traits in a large, euthymic, family-based group of individuals with bipolar disorder (BPD) and their affectively ill and healthy relatives. To test whether these traits follow a distribution with the most "pathological" scores in the bipolar disorder I (BPD I) group and the least "pathological" scores in the unaffected relatives.
Methods: Two-hundred and ninety-six individuals from 47 bipolar disorder families were administered a battery of personality questionnaires (Temperament Evaluation of Memphis, Pisa, Paris, and San Diego; Temperament and Character Inventory; Affective Neuroscience Personality Scale; Hypomanic Personality Scale; Borderline Traits Questionnaire) as well as a self-rating depression (Beck Depression Inventory) and mania (Altman Self-Rating Mania) scale. Out of the 296 participants, 57 were diagnosed with BPD I, 24 with bipolar disorder II (BPD II), 58 with recurrent major depression (MDE-R), 45 had one previous depressive episode (MDE-S), and 86 were unaffected. Twenty six individuals had another DSM-IV diagnosis.
Results: The BPD I group displayed elevated hypomanic, cyclothymic and hostile traits. These traits were also characteristic of the BPD II group but were less salient in the MDE-R group. The MDE-S group did not differ significantly from unaffected relatives. Hypomanic personality characteristics were clearly elevated in both BPD groups and differentiated BPD from major depressive disorder (MDD) individuals.
Conclusions: Our results provide preliminary support for the hypothesis that temperament is a genetically quantitative trait.