Risk Factors for Antidepressant-Related Switch to Mania

Objective: Treatment of bipolar depression with antidepressants is strongly debated on the basis of the methodologically poor and insufficient data supporting their use and the widely held belief that antidepressants can induce new episodes of abnormal mood elevation or accelerate the rate of cycling. The present study aimed at identifying clinical risk factors for switch into hypomania, mania, or mixed states, within 8 weeks after introduction of an antidepressant or after increasing its dosage, in a prospective, longitudinal design.

Method: 221 consecutive DSM-IV-TR depressed bipolar I and II disorder patients were treated with antidepressants, which were added to previously prescribed mood stabilizers and/or atypical antipsychotics. No patient was on antidepressant monotherapy. The patients were enrolled from October 2005 through January 2010. The primary outcome was the assessment of switch to mania or hypomania within 8 weeks after the introduction or dose increase of an antidepressant. Both groups were compared with analysis of variance and χ2 procedures.

Results: Treatment-emergent affective switch was detected in 54 patients (24.4%) (switch group) while 167 patients (75.6%) (nonswitch group) did not experience a treatment-related switch. The main clinical differences significantly associated with the occurrence of an antidepressant-related switch, after performing logistic regression analysis, were higher rate of previous switches (P < .001) in the switch versus the nonswitch group, lower rate of responses to antidepressants (P < .001) in the switch versus the nonswitch group, and earlier age at onset (P = .026) in the switch versus the nonswitch group.

Discussion: Bipolar patients with an earlier age at onset and an illness course characterized by lower rate of response to antidepressants and higher rate of switches into mania or hypomania were found to be the ones with higher switch risk. Nevertheless, a greater number of previous antidepressant exposures was not associated with the occurrence of an antidepressant-associated switch.

Trial Registration: clinicaltrials.gov Identifier: NCT01503489

J Clin Psychiatry 2012;73(2):e271-e276

© Copyright 2012 Physicians Postgraduate Press, Inc.

Submitted: May 23, 2011; accepted July 11, 2011 (doi:10.4088/JCP.11m07166).

Corresponding author: Edward Vieta, MD, PhD, Director of Bipolar Disorders Program, Clinical Institute of Neuroscience, Hospital Clí­nic Barcelona, Villarroel 170, 08036, Barcelona, Catalonia, Spain (evieta@clinic.ub.es).