DESCRIPTION: (Adapted from Applicant's Abstract) A longitudinal study of a cohort of 350 community-dwelling persons with clinically diagnosed Alzheimer's disease followed for an average period of four years is proposed. The descriptive aims of the study will be to characterize change in cognitive function using sequential, annual neuropsychological performance tests and to describe quantitatively, in this cohort, entry into institutional long-term care, hospitalization, and the occurrence of behavioral abnormalities and impaired physical function. The study will test two sets of hypotheses. The first is, that among those with Alzheimer's disease, the presence of each of the following factors predicts greater decline in cognitive function: lower level of formal education, lower occupational prestige, occupation with a lower educational development category and lower family income. The second is that entry into institutional long-term care and hospitalization are more strongly predicted by the presence of behavioral abnormalities and by the presence of physical impairment than by level of cognitive impairment. Because behavioral abnormalities and impairment of physical function may be more susceptible to intervention than the underlying cognitive deficit itself, study results have the potential to serve as a basis for design of trials that may lead to reduction in disease morbidity and cost.