EPIDEMIOLOGY OF ALZHEIMER DISEASE IN A TOTAL POPULATION

阿尔茨海默病（AD）的流行病学研究基于大量的

An epidemiologic study of Alzheimer's disease (AD) based on a large population [that of the Saquenay-Lac-Saint-Jean (SLSJ) region of Quebec] with several unique and important features, including a computerized population database and one of the world's largest aluminum refineries, is proposed. The first phase of this study has 5 aims: (1) to validate a key informant ascertainment methodology for identifying cases of AD by means of a prevalence survey of 2 geographic regions in SLSJ; (2) to estimate age- and sex-specific prevalence rates of dementia and AD; (3) to follow prevalent cases in order to confirm diagnoses clinically and through autopsy; (4) to identify a fixed cohort of approximately 1,900 non-demented individuals that will be utilized in incidence, risk factor and survival studies; and (5) to perform a population-based case control study of AD focusing on family history od dementia, education and aluminum as risk factors. A 3-stage procedure will be utilized to identify cases. Stage 1 will provide initial identification of dementia suspects (by a network of key informants in year 1; by population screening in years 2 and 3). At stage 2, suspects will be assessed by specially-trained geriatricians and general practitioners to determine whether they meet DSM-IIIR criteria for dementia. At state 3, individuals considered to have dementia will receive a complete workup to determine the disease responsible for the dementia. Key informants will include both health professionals and lay persons. The prevalence survey will utilize a stratified random sample of 2,000 individuals over 64 identified from the socialized health care registry of Quebec. This sample will be screened, and those individuals who fail the screen will be evaluated in stages 2 and 3 for dementia. Non-demented subjects, who will comprise approximately 95% of the sample, will be re- evaluated and followed for development of dementia. Informants for cases of possible and probable AD and for 2 matched controls drawn from the non- demented population will participate in the case-control study. Other non- demented subjects will be administered the risk factor and family history questionnaires to study risk factors in relation to later development of AD.