Vöhringer PA, Whitham EA, Thommi SB, Holtzman NS, Khrad H, Ghaemi SN. Affective temperaments in clinical practice: a validation study in mood disorders.
J Affect Disord 2012;
136:577-80. [PMID:
22100132 DOI:
10.1016/j.jad.2011.10.028]
[Citation(s) in RCA: 11] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 06/21/2010] [Revised: 07/06/2011] [Accepted: 10/20/2011] [Indexed: 10/15/2022]
Abstract
BACKGROUND
We sought to examine correlations between clinical validators and temperaments in clinical practice.
METHODS
We provided the self-report TEMPS-A (50 item long) to 123 consecutive patients seen in the Mood Disorders Program of Tufts Medical Center. Temperament was assessed as cyclothymia, dysthymia, irritable or hyperthymia. Cut-offs were tested using (50%) and (75%) thresholds of affirmative responses, as well as highest percent for dominant temperament. We reported no dominant temperament at 75% cut-off . Multivariate regression modeling was conducted to assess confounding bias.
RESULTS
Using clinical and demographic validators, cyclothymia was the most strongly validated temperament, followed by dysthymia and hyperthymia. Irritable temperament did not appear to be valid in this sample. A 75% item endorsement cut-off appeared to identify clinically important temperaments in slightly less than half of this sample. Those without any temperament at 75% cut-off had better prognostic features. 50% cut-off was highly nonspecific, and poorly correlated with diagnostic validators.
CONCLUSIONS
Affective temperaments correlate with clinical validators, most robustly for cyclothymia. 75% cut-off on the TEMPS may provide a useful categorical definition of abnormal affective temperaments in mood disorders. With that definition, slightly less than one-half of patients with mood disorders have affective temperaments. Those without abnormal affective temperaments have better prognostic features.
Collapse