A randomized controlled trial of risperidone, lithium, or divalproex sodium for initial treatment of bipolar I disorder, manic or mixed phase, in children and adolescents

Abstract

Context: There was a paucity of comparative pharmacological research for initial treatment of bipolar I disorder, manic or mixed phase, in children and adolescents.

Objective: To investigate which medication to administer first to antimanic medication-naive subjects.

Design, setting, and participants: The Treatment of Early Age Mania (TEAM) study recruited 6- to 15-year-old children and adolescents with DSM-IV bipolar I disorder (manic or mixed phase) at 5 US sites from 2003 to 2008 into a controlled, randomized, no-patient-choice, 8-week protocol. Blinded, independent evaluators conducted all baseline and end-point assessments.

Interventions: Subjects received a titrated schedule of lithium, divalproex sodium, or risperidone. Medications were increased weekly only if there was inadequate response, and no dose-limiting adverse effects, to maximum doses of lithium carbonate (1.1-1.3 mEq/L), divalproex sodium (111-125 μg/mL), and risperidone (4-6 mg).

Main outcome measures: Primary outcome measures were the Clinical Global Impressions for Bipolar Illness Improvement-Mania and the Modified Side Effects Form for Children and Adolescents.

Results: There were 279 antimanic medication-naive subjects (mean [SD] age, 10.1 [2.8] years; 50.2% female) who had the following characteristics: 100% elated mood and/or grandiosity, 77.1% psychosis, 97.5% mixed mania, 99.3% daily rapid cycling, and mean (SD) mania duration of 4.9 (2.5) years. The mean (SD) titrated lithium level was 1.09 (0.34) mEq/L, and the mean (SD) divalproex sodium level was 113.6 (23.0) μg/mL. The mean (SD) titrated risperidone dose was 2.57 (1.21) mg. Higher response rates occurred with risperidone vs lithium (68.5% vs 35.6%; χ(2)(1) = 16.9, P < .001) and vs divalproex sodium (68.5% vs 24.0%; χ(2)(1) = 28.3, P < .001). Response to lithium vs divalproex sodium did not differ. The discontinuation rate was higher for lithium than for risperidone (χ(2)(1) = 6.4, P = .011). Increased weight gain, body mass index, and prolactin level occurred with risperidone vs lithium (F(1,212) = 45.5, P < .001; F(1,212) = 39.1, P < .001; and F(1,213) = 191.4, P < .001, respectively) and vs divalproex sodium (F(1,212) = 34.7, P < .001; F(1,212) = 45.3, P < .001; and F(1,213) = 209.4, P < .001, respectively). The thyrotropin level increased in subjects taking lithium (t(62) = 11.3, P < .001).

Conclusions: Risperidone was more efficacious than lithium or divalproex sodium for the initial treatment of childhood mania but had potentially serious metabolic effects.

Trial registration: clinicaltrials.gov Identifier: NCT00057681

Figures

Figure 1

Flowchart of subjects (6-15 years of age) in the Treatment of Early Age Mania (TEAM) study that investigated which medication to administer first to antimanic medication-naive subjects. BMI indicates body mass index (calculated as weight in kilograms divided by height in meters squared); PDD, pervasive developmental disorder.

Figure 2

Comparisons of end-point Clinical Global Impressions for Bipolar Illness Improvement-Mania (CGI-BP-IM) response rates by medication. The P values, odds ratios (ORs), and 95% CIs were determined from logistic regression of CGI-BP-IM response, controlling for age, sex, socioeconomic status, baseline psychosis, baseline CGI-BP mania severity score, stimulant medication use, allergy/asthma medication use, and site. The error bars illustrate the 95% CIs for the proportions in the risperidone group (95% CI, 58.9%-78.2%), the lithium group (95% CI, 25.7%-45.5%), and the divalproex sodium group (95% CI, 15.6%-32.4%). For the comparison between the risperidone and lithium groups, the OR is 5.0 (95% CI, 2.3-10.9); for the comparison between the risperidone and divalproex sodium groups, the OR is 8.3 (95% CI, 3.8-18.0); and for the comparison between the lithium and divalproex sodium groups, the OR is 1.6 (95% CI, 0.8-3.5).