Divalproex Sodium for the Treatment of Conduct Disorder: A Randomized Controlled Clinical Trial

Hans Steiner, M.D.; Maya L. Petersen, B.A.; Kirti Saxena, M.D.; Sekou Ford, M.D.; and Zakee Matthews, M.D.

Background: New treatments for conduct disorder are sorely needed. We aimed to test the efficacy of divalproex sodium for the treatment of conduct disorder.

Method: Seventy-one youths with conduct disorder according to DSM-IV criteria were enrolled in a randomized, controlled, 7-week clinical trial. Subjects were all adolescent males with at least 1 crime conviction. Subjects were randomized into high- and low-dose conditions and were openly managed by a clinical team. Subjects and independent outcome raters were blinded to condition. Clinical Global Impressions-Severity of Illness (CGI-S) and CGI-Improvement (CGI-I) ratings, Weinberger Adjustment Inventory ratings, and staff ratings of behavioral privilege were used to assess outcome.

Results: Intent-to-treat analyses showed significant associations between assignment to the high-dose condition and ratings on the CGI-S (p = .02) and CGI-I (p = .0008). Self-reported weekly impulse control was significantly better in the high-dose condition (p < .05), and association between improvement in self-restraint and treatment condition was of borderline statistical significance (p < .06). Parallel analyses comparing outcome by blood drug level achieved strengthened the results, as expected.

Conclusion: This preliminary study in a most difficult population suggests a role for divalproex sodium in the treatment of conduct disorder. Divalproex sodium improved self-reported impulse control and self-restraint, variables shown to be predictive of criminal recidivism. Studies are needed of longer-term impact and side-effect profiles. This is one of few controlled psychopharmacologic studies of conduct disorder.

(J Clin Psychiatry 2003;64:1183-1191)

Received May 31, 2001; accepted March 17, 2003. From the Division of Child Psychiatry and Child Development, Stanford University School of Medicine, Stanford, Calif.

This research was supported by grants from Abbott Pharmaceuticals, Chicago, Ill., the California Wellness Foundation, Los Angeles, Calif., and the California Youth Authority, Sacramento, Calif. (Dr. Steiner).

Dr. Steiner has received grant/research support and honoraria from and has served on the speakers or advisory board for Abbott.

Corresponding author and reprints: Hans Steiner, M.D., Stanford University School of Medicine, Division of Child Psychiatry and Child Development, 401 Quarry Road, Stanford, CA 94305-5719 (e-mail: steiner@stanford.edu).