Sodium valproate in the treatment of aggressive behavior in patients with dementia--a randomized placebo controlled clinical trial.

The efficacy and tolerability of sodium valproate 2 x 240 mg compared to placebo were investigated in aggressive behavior in dementia. A randomized, placebo controlled, double-blind cross-over design. The trial included a baseline period (one week); a placebo period (three weeks); a wash-out period with placebo (one week); and a treatment period with sodium valproate (three weeks). A psychogeriatric short-stay ward at a psychiatric teaching hospital. Demented patients who met Patel's criteria for aggressive behavior and had a score of > or =3 on at least one of the items of the Social Dysfunction and Aggression scale-9 (SDAS-9). A fixed dose of sodium valproate 2 x 6 ml of a 40 mg/ml suspension (daily defined dose of 480 mg) was compared to placebo. Primary outcome variables were changes of the score of SDAS-9 and Clinical Global Impression scale (CGI) performed at the last week of each treatment period. Data of 42 patients (F=25 and M=17; age 80.4+/-6.8 years) were analyzed. Treatment with sodium valproate showed no differences compared to placebo on aggressive behavior. The mean plasma level of sodium valproate was 40.9+/-10.8 microg/ml. Regression analysis showed a trend for improvement between the plasma levels of sodium valproate and the SDAS-9 and the CGI scores. Adverse events were not related to the plasma levels of sodium valproate. Secondary outcome measurements showed significant improvement on restless, melancholic and anxious behavior; a trend for improvement was found on suspicious and dependent behavior. Possible limitations of this study are the low dose of sodium valproate, the relatively short treatment period (three weeks), and the absence of statistical corrections for multiple comparisons. This study showed no effect of sodium valproate 2 x 240 mg over placebo on aggressive behavior in dementia. [ABSTRACT FROM AUTHOR]