Funding: An investigator-initiated trial funded by H. Lundbeck AS.

Study design: Prospective, randomised, placebo-controlled parallel-group multicenter study.

Aim: To investigate efficacy and side effects (especially mood switches) of escitalopram,a selective serotonin reuptake inhibitor, in the acute and maintenance treatment of bipolar depression.

Hypotheses:

1. Escitalopram, given in addition to mood stabilising medications, is significantly more efficacious, measured by response and remission rates than placebo in bipolar depression (the acute phase study).
2. Continuation therapy with escitalopram gives significantly longer mean time to depressive relapse and fewer depressive relapses compared to placebo (the continuation study).
3. The incidence of "mood switching" (defined as development of mixed episodes, mania, or hypomania according to DSM-IV criteria) do not differ significantly between escitalopram and placebo in either the acute or the continuation phases.

Patients: In- and outpatients receiving care in the specialised psychiatric services of Western Norway. The population is intended to be representative of the patients treated for bipolar depression in ordinary specialist care. Patients must have a MADRS score of at least 20 at baseline. Patients with ongoing substance abuse or dependence, organic mental illness, and non-affective psychotic symptoms are excluded.

Medication: Escitalopram 10-20 mg daily or placebo in addition to mood stabilisers. The dose of mood stabilisers must have been constant for the last six weeks prior to randomisation.

Method: Phase 1 is a eight-week acute treatment trial with six clinical assessments. Patients treated with escitalopram who have not responded after eight weeks (defined by at least 50% reduction of MADRS score compared to baseline) leave the study. Placebo non-responders are treated openly with escitalopram and repeat phase 1. Responders are re-randomised to 32 weeks of maintenance treatment (phase 2). Phase 2 has nine clinical assessments. Patients who develop hypomania, mania or depressive episodes (defined as episodes meeting DSM-IV criteria for Major Depressive Episode with MADRS scores of at least 20 points) leave the study in this phase. Patients leaving the study prematurely will be offered alternative treatment.