Previous psychiatric history as a risk factor for late-life dementia: a population-based case-control study.

OBJECTIVE To test the hypothesis that risk for dementia in late life is increased by a history of earlier psychiatric illness, and to examine the specificity of any such association. METHODS Frequency of earlier treated psychiatric illness was established by record searches and informant histories for all persons aged over 60 who were entered on the Camberwell dementia case register over a 2-year period and for an individually matched control group of the local elderly population, drawn from the files of the area Family Health Services Authority. RESULTS Of 559 persons with a clinical diagnosis of dementia, 70 (12.5%) had a history of psychiatric illness long preceding, and apparently unrelated to, the onset of dementia. When these patients were compared with a matched comparison group of dementia register patients who had no recorded psychiatric history, the two groups appeared broadly similar in type of dementia, severity and clinical features. Sixty-three of the earlier psychiatric histories could be identified from the available medical records alone. Applying the same procedure to the matched control group of Camberwell residents yielded from the 559 only 19 (3.4%) with a history of treated psychiatric illness before the age of 70. The difference in case frequency between the two groups was highly significant and corresponded to an estimated odds ratio of 3.6. The increase in risk, which was not restricted to dementia of Alzheimer type, appeared to be characteristic of major psychiatric illness, rather than specific for depression. CONCLUSION There is a positive association between previous psychiatric history and late-life dementia which is of predictive significance. The underlying causal links merit further research.