Citation

Oquendo MA, Galfalvy HC, Sher L, Sullivan GM, Grunebaum MF, Goldberg PH, Parsey RV, Mann JJ. Psychiatr. Danub. 2006; 18(Suppl 1): 46-47.

Affiliation

Columbia University Department of Neuroscience, 1051 Riverside Drive, 10032 NYNew York, USA. mao4@columbia.edu.

Copyright

(Copyright © 2006, Facultas Universitatis Studiorum Zagrabiensis - Danube Symposion of Psychiatry)

DOI

unavailable

PMID

16963916

Abstract

Prospective studies of predictors of suicidal acts in bipolar samples are rare. Among depressed bipolar attempters (n=57) randomized to lithium or valproate over 2.5 years, we identified predictors for "suicide event" defined as a suicidal act or ideation with a plan requiring intervention. 29 suicide events (13 attempts) occurred in 24 (42%) patients, mostly within 6 months (21/29). Baseline depression predicted events (p=0.009). Later onset was protective (p=0.056) and impulsivity (p=0.065) and number of previous attempts (p=0.09) increased risk for events. Ongoing depression (H.R.=1.09, p=0.0024) and ongoing suicidal ideation (HR=1.09, p=0.054) both increased risk for events. For attempts, age was protective (HR=0.92, p=0.035) and baseline number of attempts (HR=1.5, p=0.019) and baseline ideation (HR=1.1, p=0.015) were risk factors. Ongoing aggression tended to predict attempts (HR=1.3, p=0.095). There were no differences in suicidal acts between patients on lithium or valproate. However, subjects aged 40 or more were robustly protected from suicidal behavior if on lithium. Thus, our reported predictors of suicidal behavior in independent samples of mostly unipolar depression also predict suicidal acts in bipolar subjects. Protective effects of lithium were not evident among the younger portion of the sample.

Language: en