Background. Comparatively little is known about the long-term natural history of depressive disorders in the elderly living in the community. This is a follow-up of a subsample of the Continuing Health in the Community study random sample of the elderly population living in Liverpool.

Methods. The investigators followed up 120 cases of depression identified by a semi-structured interview schedule (GMS) for a period of 5 years. A similar number of other subjects defined as subcases of depression, other cases of mental illness and a random selection of non-cases were also included.

Results. The 5-year outcome for the cases of depression was worse than the outcome of the non-cases (relative mortality risk of 2·1, 95% confidence interval 1·1 to 3·9). Thirty-four per cent of the cases of depression died and 28% had dropped out during the follow-up. Of the 46 cases of depression who had a complete follow-up, 22% recovered from their symptoms, 30% were found to be AGECAT cases at one of the three follow-up waves, 24% were AGECAT cases at two of the three follow-up waves and the remaining 24% were AGECAT cases at each follow-up wave. Fifteen per cent of the surviving cases of depression were organic cases at the follow-up. Their anxiety co-morbid state and depression score were identified as predictors of poor outcome.

Conclusion. The findings of this study indicate that depressive disorders (most of which were untreated) found in the elderly community have a poor prognosis.