Six-year survival of depressed elderly Finns: a community study.

Background: The 6-year survival of depressed elderly (N = 290) Finns was assessed and compared to that of a non-depressed population (N = 982), and the factors related to high mortality were analysed.
Method: Depression was determined according to the DSM-III criteria, and a majority of the depressed population suffered from dysthymic disorder. Survival distributions were represented as Kaplan-Meier survival curves. The Cox proportional hazards model was used to analyse the simultaneous relationships between mortality and certain variables.
Results: The survival curves showed an increased mortality of the depressed elderly in both men and women. However, when the simultaneous relationships of age, marital status, education, smoking, functional abilities, somatic illnesses and depression were taken into account, depression did not predict mortality. The predictors of mortality were high age, a high number of medicines, smoking, disability, male sex and occurrence of somatic illnesses.
Conclusions: Evidence of once measured depression is not predictive of increased mortality in an unselected elderly population (60+) when the other factors known to influence survival probability are taken into account.