The epidemiology of functional psychoses of late onset.

For the functional psychoses of late life, epidemiological information comes from two sources: studies of persons who have reached psychiatric services; and surveys of elderly persons sampled from the general population. A conspectus of published data from both sources leads to the following conclusions: States phenomenologically similar to those found in clinics do occur in the community in non-trivial numbers. There is no notable divergence in the information obtained from clinical series and from population-based surveys. These states are more common in women, they become more common with increasing age and are sometimes associated with decline in cognitive performance or with degenerative changes in the brain revealed by neuroimaging. Genetic factors appear to be less important than in early-onset psychoses but remain ill-defined, and the roles of social isolation and disorders of personality have not yet been sufficiently elucidated. Both clinical and community-based studies have found an association with sensory impairment. The community-based data suggest that paranoid symptoms may be detectable at subclinical level, and an association between them and cognitive impairment is demonstrable in individuals who are not diagnosable cases either of psychosis or of dementia. Differences exist between late-onset paranoid psychoses and affective psychoses in symptomatology and response to treatment. These observations confirm the importance of the late-onset psychoses for research directed towards uncovering the origins of psychotic symptoms in any age group.