Background: Previous studies of the natural course of bipolar disorder have suggested that relapse and psychosocial impairment are more common than many would expect during treatment under routine conditions. The present research sought to identify patterns of consistency in longitudinal recovery after an index manic or depressive episode in patients formerly hospitalized for major affective disorders. The data extend prior findings from the Chicago Follow-up Study involving four successive assessments of course and outcome for bipolar and unipolar patients over a 10-year follow-up interval.
Methods: Thirty-four RDC bipolar I manic, 17 psychotic unipolar depressed, and 72 nonpsychotic unipolar depressed patients were assessed at index hospitalization and prospectively followed-up at 2, 4.5, 7.5 and 10 years. Psychosocial functioning, rehospitalization, and overall outcome were rated by standardized instruments. Patterns of consistent remission or impairment in functioning were compared across follow-up periods.
Results: Consistently good functioning was observed more often among patients with unipolar nonpsychotic depression than bipolar disorder or unipolar psychotic depression. Psychosis at index hospitalization was not associated with poorer outcomes for either the bipolar or unipolar groups, although it significantly predicted psychosis at follow-up more robustly for unipolar than bipolar patients. Less than half of the bipolar cohort had good work performance at each follow-up, while unipolar nonpsychotic depressed patients had consistently better work functioning. Rehospitalization was more common for the bipolar than unipolar patients at the 4.5- and 7.5-year assessments. Sustained remission across follow-ups was associated with remission at subsequent follow-ups regardless of diagnostic polarity at index hospitalization.
Limitations: The observational design, small sample size, and naturalistic treatment of subjects in this protocol prompts caution when interpreting treatment outcome findings. The availability of a bipolar cohort with relatively low attrition over a 10-year period may limit the generalizability of the current findings to patients who remain engaged in sustained contact with a long-term research program.
Conclusions: Over a 10-year follow-up period, about half of bipolar patients show sustained remissions or patterns of improvement, while 30-40% experience some functional decline. The degree of consistency in remission patterns over time may hold greater prognostic significance than the polarity of an index affective episode in anticipating subsequent levels of psychosocial adjustment in severe mood disorders.