Excess mortality of bipolar and unipolar manic-depressive patients.

To compare mortality in bipolar and unipolar manic-depressive patients, 2168 manic-depressive first admissions reported to the Danish Psychiatric Central Register were divided into a bipolar group (19%), i.e., patients with at least one admission for mania during an average observation period of six years, and a unipolar group. When compared with the general population, the total group had an increased mortality by suicide and accidents in both sexes and by non-violent causes in men. The bipolar group had a higher non-violent mortality than the unipolar group, but the violent mortality was not different. Statistical problems introduced by patients switching from the unipolar to the bipolar group during the period of observation are discussed.