Prevalence, incidence and correlates of depression in the oldest old: The OCTO study

Abstract

While there has been extensive research on depression in later life, few studies have focused on the prevalence of symptoms and patterns of change over time among the oldest old. This paper reports on a longitudinal study of a population-based sample of older people initially aged 84 to 90. Participants were interviewed at two-year intervals over a six-year period of time. From an initial population base of 400, 324 people participated in the initial survey, 254 of whom provided information on depressive symptoms. Using a short form of the CES-D we found prevalence of significant depressive symptoms to range between 43 and 50% for the four times of measurement. Incidence of clinically significant symptoms ranged between 9 and 16% between measurement periods, while some people who were previously depressed showed improvement. A multivariate model indicated that depressive symptoms were related to functional impairments and disabilities but these effects were mediated by mastery beliefs. These findings indicate that depressive symptoms affect a somewhat higher proportion of the population in very late life than among the young old, but there is evidence of recovery and that psychological resources mediate the impact of impairment and disability on mood. Treatment strategies which enhance psychological resources such as mastery may reduce the impact of disability on mood, and may even have a preventive role.