Long-term observational comparison of risperidone and olanzapine in bipolar disorder

To compare long-term effectiveness and safety of risperidone versus olanzapine as adjunctive maintenance treatments of bipolar disorder. Retrospective observational chart review of 29 outpatients with bipolar or schizoaffective disorder (type I = 15, type II = 3, NOS = 5, schizoaffective = 6) who received risperidone or olanzapine added to lithium or valproate >3 months. Acute indications were depression (n = 8), manic/hypomanic/mixed states (n = 8), rapid cycling (n = 6), other indications (n = 6), and prophylaxis (n = 1). Logistic regression models adjusted for potential confounding factors (i.e., severity of illness, comorbid substance abuse, diagnostic subtype). Overall duration of follow-up was 65.9+/-70.1 weeks. Mild to moderate response was similar in the risperidone and olanzapine groups after adjusting for potential confounders (OR = 0.91, 95% CI [0.05, 16.17]). Somewhat greater adjusted moderate to marked response (OR >3.60, 95% CI [0.31, >42.00]) and longer duration of treatment (HR = 0.52, 95% CI [0.22, 1.22]) occurred in the risperidone group, but were still compatible with the null hypothesis. Weight gain occurred more frequently with olanzapine (57%) than risperidone (13%). EPS was similar, and tardive dyskinesia did not occur. Risperidone and olanzapine appeared to have similar real-world maintenance effectiveness for bipolar disorder, but differed somewhat in side effects.