Objective: The authors compared efficacy of olanzapine versus placebo and risperidone as measured by the Neuropsychiatric Inventory and Clinical Global Impression-Severity of Psychosis scale in patients with dementia-related psychosis. Methods: Patients with moderate-to-severe psychotic symptoms associated with dementia were recruited from outpatient or residential settings and randomly assigned to 10-week, double-blind, flexible-dose treatment with olanzapine (N = 204; 2.5 mg-10 mg/day; mean: 5.2 mg/ day), risperidone (N = 196; 0.5 mg-2 mg/day; mean: 1.0 mg/day) or placebo (N = 94). Results: Most measures of neuropsychiatric functioning improved in all treatment groups, including the placebo group, and no significant treatment differences occurred. Overall discontinuation was lowest in the placebo group, and the olanzapine group had a significantly higher incidence of discontinuation due to adverse events (16.2%) relative to placebo (3.2%) and risperidone (8.7%) groups. Treatment-emergent extra-pyramidal symptoms were more numerous for risperidone- than placebo- or olanzapine-treated patients. Abnormally high prolactin levels occurred in 78.0% of risperidone patients, compared with 16.7% for olanzapine and 5.0% for placebo. The incidence of weight gain greater than 7% from baseline was higher in the olanzapine group relative to risperidone, but neither active-treatment group showed a statistical difference from placebo (1.1%). No other statistically significant and clinically relevant differences were seen for any other vital sign, electrocardiographic measure, or laboratory hematology and chemistry, including glucose, except for cholesterol, which decreased from baseline to endpoint in both active-treatment groups. Conclusions: Patients' neuropsychiatric functioning improved with olanzapine, risperidone, and placebo treatment. There was a substantial response in the placebo group, and no significant differences emerged among treatments.
Bibliographical noteFunding Information:
This work was sponsored by Eli Lilly and Company. Special thanks for helping in the organization of this trial go to Jennifer Wilkie.