BACKGROUND:

Little information exists from controlled studies about clinical characteristics that predict treatment response in mania. The presence of depressive symptoms during manic episodes may be associated with poor response to psychopharmacological treatments. This is an investigation of the relation between depressive symptoms and treatment response in acute manic episodes.

METHODS AND DESIGN:

In a parallel-group, double-blind study, 179 patients hospitalized for acute manic episodes were randomized to receive divalproex sodium, lithium carbonate, or placebo (ratio, 2:1:2). The study was carried out at 9 academic medical centers. Patients had comprehensive evaluations of behavior and symptoms before and during 3 weeks of treatment. The primary outcome measure, change in mania factor scores derived from the Schedule for Affective Disorders and Schizophrenia: Change Version, was compared in patients with and without depressive symptoms at baseline according to nurse- or physician-rated scales.

RESULTS:

Depressive symptoms were associated with poor antimanic response to lithium and with better response to divalproex. This was not due to differences in overall severity of illness, substance abuse, gender, age, or history.

CONCLUSIONS:

These data suggest that even a modest level of pretreatment depression-related symptoms is a robust predictor of lithium nonresponse, and is associated with better response to divalproex. Although their overall efficacy in acute mania is similar, lithium and divalproex may be most effective in clinically and biologically distinct groups of patients.