A randomized trial of risperidone, placebo, and haloperidol for behavioral symptoms of dementia

To compare effects of risperidone with placebo (efficacy and tolerability) and haloperidol (tolerability) for treating demented patients with aggression and other behavioral symptoms.

A 13-week double-blind study involving 344 patients with dementia randomly assigned to receive placebo or flexible doses (0.5 to 4 mg/d) of risperidone or haloperidol. Behavioral symptoms were assessed by the Behavior Pathology in Alzheimer's Disease Rating Scale (BEHAVE-AD), the Cohen-Mansfield Agitation Inventory (CMAI), and the Clinical Global Impression (CGI) scale. Tolerability assessments included the Extrapyramidal Symptom Rating Scale, sedation levels, Functional Assessment Staging, Mini-Mental State Examination, and incidence of adverse events.

The mean dose at endpoint was 1.1 mg/d of risperidone and 1.2 mg/d of haloperidol. Although not significant, a higher percentage of patients receiving risperidone than those receiving placebo showed clinical improvement (> or =30% reduction from baseline to endpoint in BEHAVE-AD total score) at endpoint and week 12. Reductions in the BEHAVE-AD total score were significantly greater with risperidone than with placebo at week 12. In a further analysis of aggression, the most dominant symptom in these patients, BEHAVE-AD and CMAI aggression cluster scores were significantly reduced compared with placebo at endpoint and week 12. CGI scores were also significantly reduced at endpoint and week 12. Severity of extrapyramidal symptoms with risperidone did not differ significantly from that of placebo and was less than that of haloperidol. A post hoc analysis showed significantly greater reductions in the BEHAVE-AD aggressiveness score with risperidone than haloperidol at week 12.

Low-dose risperidone (mean 1.1 mg/d) was well tolerated and associated with reductions in the severity and frequency of behavioral symptoms, particularly aggression, in elderly patients with dementia.