Most available information on behavioral symptoms in Alzheimer's disease has come from studies of persons identified through medical care facilities, samples which are susceptible to significant biases and lacking in appropriate comparison groups. The proposed study will evaluate the application of standardized measures of the behavioral symptoms of Alzheimer's disease in persons from a geographically-defined community population. informant-based measures of behavioral symptoms will be administered to a stratified, random sample of 760 residents, age 65 and older, from the community. The basic hypothesis is that the prevalence of hallucinations, delusions, depressive symptoms, anxiety symptoms, and aggressive behavior will be substantially higher in persons with Alzheimer's disease relative to comparable unaffected persons. This study will also estimate participation rate, measurement error associated with informant ratings, and prevalence of 5 types of behavioral symptoms in persons with Alzheimer's disease for the purposes of designing subsequent longitudinal studies. ADMINISTRATIVE NOTES: The Committee judged the application to be appropriate for inclusion in priority category of more experienced investigators for testing new methods or techniques.