BACKGROUND Admission statistics for mania frequently show an increase in the summer. The present two-centre study was designed to test the hypothesis, in a representative sample of bipolar patients, that manic and depressive relapses show a seasonal pattern.
METHOD Two cohorts of bipolar I patients, one in London, England (n = 86), the other in Dunedin, New Zealand (n = 58), were tracked retrospectively during 1985-88 and prospectively during 1989-91, with the onset of all relapses being carefully dated.
RESULTS In the London cohort there were 221 episodes of mania and 76 of depression; in the Dunedin cohort there were 201 of mania and 61 of depression. No consistent seasonal pattern of mania was detected in either centre. There was an autumn preponderance of depressive episodes in both centres.
CONCLUSIONS Relapse of bipolar depression, but not of mania, appears to be determined in part by seasonal factors.