Association of suicidal risk with ratings of affective temperaments.

BACKGROUND: Ratings of particular temperament-types with the TEMPS-A autoquestionnaire have been associated with suicidal risk, and combinations of such ratings may enhance the association. However, the predictive value of scores for individual temperaments and combinations remains to be quantified.

METHOD: We evaluated associations of TEMPS-A ratings for anxious (anx), cyclothymic (cyc), dysthymic (dys), hyperthymic (hyp) and irritable (irr) temperaments, with a history of suicidal acts or reported suicidal ideation in 882 patients with bipolar (BD; n = 509), major depressive (MDD; n = 268) or anxiety (ANX; n = 105) disorders.

RESULTS: With BD, scores for cyc and irr were highest and anx lowest; with MDD, dys scored highest, hyp lowest; anx was highest with ANX and MDD. Women (n = 497) had higher anx and cyc scores than men; scores for irr and hyp decreased with age. Scores for dys, anx, and cyc, were higher, and hyp lower, with greater HDRS21 depression ratings. Among 347 suicidal subjects (112 with attempts), cyc, dys, and irr scores were higher, hyp lower. Pooled score [cyc+ dys+ irr - hyp] best distinguished subjects with suicide attempts versus nonsuicidal subjects, including in Receiver-Operating Characteristic (ROC) analysis (AUC for acts = 70.1% [64.9- 75.3]). Multivariable modeling associated suicidal acts with TEMPS-A [cyc+ dys+ irr - hyp] composite-score, depression severity, BD or MDD diagnosis, and older at illness-onset.

CONCLUSIONS: Subjects with suicidal acts or ideation were best distinguished by composite TEMPS-A score [cyc+ dys+ irr - hyp]. These factors should help to identify those at suicidal risk.