Effectiveness of lithium in children and adolescents with conduct disorder: a retrospective naturalistic study

The most severe forms of conduct disorder (CD) are disabling conditions, often resistant to treatment and likely to evolve into antisocial behaviours. Mood stabilizers and atypical antipsychotics are often used to treat severe cases of CD, as are antidepressants and psychostimulants less frequently, despite a relative lack of efficacy data. Use of lithium in hospitalized children and adolescents with CD has been evaluated in a small number of studies.

To explore the efficacy and tolerability of lithium (administered either as monotherapy or in association with atypical antipsychotics) in children and adolescents with CD and to identify variables associated with positive or negative responses to such treatment.

This retrospective study included 60 consecutive patients (46 males and 14 females; range 8-17 years; mean age 14.2 +/- 2.4 years) who were treated with lithium for CD diagnosed on the basis of the Kiddie-Schedule for Affective Disorders and Schizophrenia-Present and Lifetime (K-SADS-PL) clinical interview and the DSM-IV criteria for CD. The sample consisted of 44 inpatients (who remained in hospital during the first 2 or 3 weeks of treatment and were then assessed as outpatients) and 16 outpatients; the follow-up period was 6-12 months (mean 8.4 +/- 2.2 months). While all patients were initially treated with lithium, an atypical antipsychotic could be added if necessary to achieve satisfactory control of symptoms. Outcome measures included the Modified Overt Aggression Scale (MOAS), the Clinical Global Impression-Severity (CGI-S) and Clinical Global Impression-Improvement (CGI-I) scales, and the Aggression Questionnaire (which assessed the type of aggression, i.e. predatory vs affective). Patients were considered responders to pharmacological treatment at the end of the follow-up period if they satisfied all of the following criteria: >or=50% decrease in MOAS score, CGI-I score of 1 or 2 ('very much improved' or 'much improved') and CGI-S score of 0.8 indicates good treatment efficacy).

At the end of follow-up, 29 of 60 patients (48.3%) were classified as responders (10 receiving lithium monotherapy and 19 receiving lithium plus atypical antipsychotic therapy). For the sample as a whole, mean total MOAS score improved significantly (p < 0.001) and the effect size (pre- vs post-treatment) was 1.03. Mean MOAS verbal, and physical towards objects and other persons aggression scores, both in patients taking lithium as monotherapy and those also taking an add-on atypical antipsychotic, also improved significantly (p < 0.001) and the effect size was >or=0.80 for all items. Improvement in mean MOAS self-aggression score (p < 0.001) and an effect size of 0.59 was found when an atypical antipsychotic was added to lithium therapy. Predictors of a positive response to treatment were less severe disease at baseline, lower MOAS aggression scores and an impulsive (affective, nonpredatory) type of aggression. Gastrointestinal adverse effects, polydipsia and increased urinary frequency, tremor and increased thyroid stimulating hormone levels were the most frequently reported adverse effects. Two patients discontinued treatment because of adverse effects (vomiting and thyroid dysfunction).

Lithium alone or in combination with an atypical antipsychotic may reduce aggressive behaviours in children and adolescents with CD. The adverse effects of such therapy are relatively common but rarely severe.