A double-blind, placebo-controlled study of valproate for aggression in youth with pervasive developmental disorders.

OBJECTIVE The aim of this study was to study valproate efficacy and safety for aggression in children and adolescents with pervasive developmental disorders (PDD). METHODS In this prospective double-blind, placebo-controlled study, 30 subjects (20 boys, 10 girls) 6-20 years of age with PDD and significant aggression were randomized and received treatment with valproate (VPA) or placebo (PBO) for 8 weeks as outpatients. Mean VPA trough blood levels were 75.5 mcg/mL at week 4 and 77.8 mcg/mL at week 8. RESULTS No treatment difference was observed statistically between VPA and PBO groups. The Aberrant Behavior Checklist--Community Scale (ABC-C) Irritability subscale was the primary outcome measure (p = 0.65), and CGI--Improvement (p = 0.16) and OAS (p = 0.96) were secondary outcome measures. Increased appetite and skin rash were significant side effects. Only 1 subject was dropped from the study owing to side effects, notably a spreading skin rash, which then resolved spontaneously. Two subjects receiving VPA developed increased serum ammonia levels, one with an associated parent report of slurred speech and mild cognitive slowing. Poststudy, of 16 VPA and PBO subjects receiving VPA, 10 subjects demonstrated sustained response, 4 of whom later attempted taper, with significant relapse of aggression. CONCLUSION The present negative findings cannot be viewed as conclusive, partly owing to the large placebo response, subject heterogeneity, and size of the groups. Larger studies are needed to expand upon these findings.