Objective: Despite their widespread use in bipolar disorder, there is controversy surrounding the inclusion of antidepressant medications in the disorder's management. We sought to identify which demographic, socioeconomic, and clinical factors are associated with antidepressant exposure in bipolar disorder and which bipolar disorder patients are most likely to report a history of antidepressant-induced mania (AIM) when exposed to antidepressants.
Methods: Our study included subjects with bipolar I disorder (n = 309), bipolar II disorder (n = 66), and bipolar disorder not otherwise specified (n = 27) and schizoaffective disorder, bipolar type (n = 14), from a longitudinal, community-based study. Subjects were evaluated using the Diagnostic Interview for Genetic Studies, modified for DSM-IV criteria. We applied multivariate logistical regression modeling to investigate which factors contribute to antidepressant exposure in bipolar disorder patients. We also used a logistic regression modeling approach to determine which clinical factors in bipolar disorder patients are associated with a history of AIM. Data were gathered from February 2006 through December 2010.
Results: Our results suggest that the risk factors most strongly associated with antidepressant exposure are female sex (OR = 2.73, P = .005), older age (OR = 1.03, P = .04), greater chronicity of illness (OR = 2.29, P = .04), and, to a lesser extent, white race (OR = 0.44, P = .051). Factors associated with reduced antidepressant exposure include history of affective psychosis (OR = 0.36, P = .01) and a greater number of previous manic episodes (OR = 0.98, P = .03). In subjects who reported a history of AIM, regression analysis revealed that the only statistically significant factor associated with AIM history was female sex (OR = 3.74, P = .02).
Conclusions: These data suggest that there are certain identifiable factors associated with antidepressant exposure in bipolar disorder patients, and some of these, specifically female sex, are also associated with a history of AIM. These data may be useful in designing prospective trials to identify interventions that can reduce the risk of this adverse outcome.
© Copyright 2018 Physicians Postgraduate Press, Inc.