Bipolar disorder begins during adolescence but often escapes diagnosis at this time because episodes are misinterpreted with other psychiatric disorders. Lithium, carbamazepine, valproate, and other drugs are used for the treatment of acute episodes and maintain treatment of bipolar disorders.

To compare the efficacy of Valproate vs. Lithium in the long-term treatment of patients with subtypes of bipolar disorders.

120 patients with more than two episodes of BD (according to DSM-IV) in a longitudinal, comparative and randomized clinical trial, for 2 years (104 weeks) divided in two equal parallel-groups with open label pre-randomized phase. Primary outcome measure was time to relapse/recurrence of any mood episodes. Survival analyses (Kaplan-Mayer and Cox Proportional Hazard) were used for statistical analysis.

Cumulative survival for valproate's group was S(104)=0.3570, (35.70% and lithium's group S(104)=0.3136, (31,36%). Valporate is superior to lithium for the treatment of non classic BD I (Log Rank 0.0309, p=0.0100). Valproate prolongs median survival time with 11±9 and mean survival time to 7±4 weeks more than lithium. Treatment with lithium has 20.4% higher relative risk for relapse/recurrence than treatment with valproate Vp (β = 0.186, p = 0.434). Valproate is superior for the treatment of BD with psychiatric comorbidity (Log Rank =0.0007. p=0.0026).

We found that valproate is significantly more effective than lithium in prophylactic treatment of bipolar I disorders in non-classic subtype and in bipolar disorder with psychiatric comorbidity.