Long-term naturalistic treatment of depressive symptoms in bipolar illness with divalproex vs. lithium in the setting of minimal antidepressant use

Objective: Risks have been associated with the long-term use of antidepressant in the treatment of bipolar disorder. We review our naturalistic experience with divalproex versus lithium in treating depressive symptoms of bipolar illness. Method: All patients with bipolar disorder treated with lithium or divalproex were identified in a university outpatient psychiatry clinic sample over one year (n=38 patients, 41 treatment trials). Treatment response was based on standard prospective symptom rating scales. Mean duration of follow-up was 90 weeks. Results: Lithium and divalproex were equally effective and tolerated in the total sample. Antidepressant effects were noted despite sparing use of standard antidepressant agents (19% received them). Lithium non-responders responded well to divalproex (50%), and vice versa (44%). Divalproex monotherapy (24%) was more common than lithium monotherapy (7%, P=0.07) and was notably effective in treating depressive symptoms, with a 7/10 response on the CGI-BP and improvement on the HDRS (14.8±9.2 to 7.6±7.8, P=0.003, duration of prospective follow up 26.7 weeks). Conclusions: Lithium and divalproex were equally effective and tolerated in this naturalistic sample, but responders may represent distinct subgroups. Both agents, but particularly divalproex, demonstrated long-term antidepressant effects, with limited adjunctive standard antidepressant use.