The average temperament scores and the correlation structure between the TEMPS-A subscales observed in our sample is consistent with previous reports in other samples of healthy controls and/or subjects with BD and major depressive disorder (23,39–40). The cyclothymic and irritable temperaments showed the highest degree of correlation (r=0.68) in our data, and there has been some debate whether irritable is its own temperament or a subtype of the cyclothymic temperament (27,41). The high degree of correlation (r=0.61) between the dysthymic and anxious temperaments in our sample is also consistent with the observation of overlap between these temperaments, largely in terms of cognitive attributes (42).
