Effect of comorbid anxiety on treatment response in bipolar depression

Background: This secondary analysis from a randomized double-blind study of acute bipolar depression compared olanzapine monotherapy, olanzapine-fluoxetine combination (OFC) and placebo in patients with or without comorbid anxiety. Methods: Patients with bipolar disorder and a current depressive episode received olanzapine (5-20 mg/day), OFC (6/25, 6/50, or 12/50 mg/day), or placebo in an 8-week trial. Two populations were defined: comorbid (Hamilton Anxiety Rating Scale, HAM-A >= 18) or non-comorbid (HAM-A < 18) anxiety. Changes in Montgomery-angstrom sberg Depression Rating Scale (MADRS) and HA-M-A total scores, and rates of response (>= 50% decrease from baseline to endpoint) and remission (MADRS <= 12 or HAM-A <= 7) were analyzed. Results: Baseline MADRS and YMRS scores were significantly higher in patients with comorbid anxiety relative to those without. Patients without comorbid anxiety were more likely to achieve MADRS response and remission than those with comorbid anxiety (relative risk, RR: 1.21 and 1.29, respectively). Patients with comorbid anxiety had greater rates of response and remission with olanzapine and OFC relative to placebo (response RR: 1.45 and 2.14; remission RR: 1.96 and 2.32, respectively). Response and remission rates on the HAM-A scale were greater for OFC relative to placebo (RR: 2.00 and 3.20). Weight gain was greater for olanzapine (2.59 +/- 3.24 kg) and OFC (2.79 +/- 3.23 kg) relative to placebo, as were baseline to endpoint changes in cholesterol levels (6 +/- 31 and 10 +/- 67 mg/dL, respectively). Conclusions: Comorbid anxiety symptoms in patients with bipolar depression have a negative impact on treatment outcome. Olanzapine and, to a greater extent, olanzapine-fluoxetine combination were effective in reducing both depressive and anxiety symptoms in these patients. The significantly greater changes in weight, glucose and cholesterol parameters observed in the olanzapine and olanzapine-fluoxetine combination groups should be entered into the risk-benefit assessment in determining appropriate treatment options for these patients. (c) 2007 Elsevier B.V All rights reserved.