Objective: The objective of this study was to evaluate efficacy and safety of low-dose risperidone for treating psychosis of Alzheimer disease (AD).
Method: The authors conducted a randomized, eight-week, double-blind, placebo-controlled, multicenter trial involving nursing home residents diagnosed with AD and psychosis. Four hundred seventy-three patients were randomly assigned to placebo (N = 238) or 1.0 to 1.5 mg risperidone per day (N = 235). Coprimary efficacy end points were: changes in scores on the Behavioral pathology in Alzheimer's Disease (BEHAVE-AD) Psychosis subscale and Clinical Global Impression of Change (CGI-C). Protocol-specified subgroup analyses were performed by demographics and dementia severity.
Results: Efficacy analysis included 416 patients. Both groups improved significantly on the BEHAVE-AD Psychosis subscale and CGI-C with no significant difference between groups. In the subgroups analyses, a statistically significant treatment by Mini-Mental Status Examination (MMSE) interaction on the CGI-C (F([2,381]) = 3.90, p = 0.021) was observed with patients with more severe dementia (MMSE <10) showing significant differences at end point favoring risperidone treatment (chi(2) ([1]) = 5.11, p = 0.024). Mean risperidone dose was 1.03 +/- 0.24 mg per day. All-cause discontinuation rates were 25% for both risperidone and placebo. Treatment-emergent adverse events occurred in 74% risperidone versus 64% placebo patients, with somnolence occurring significantly more frequently with risperidone (16.2% versus 4.6%). Nine (3.8%) risperidone- and six (2.5%) placebo patients died during or within 30 days after treatment.
Conclusion: This trial did not confirm earlier findings in this population.