Epidemiologic studies of disease fall into two large categories: descr
iptive and analytic. Both approaches have been used in studies of deme
ntia and Alzheimer's disease (AD). The prevalence of dementia in perso
ns age 65 and older in North America is approximately 6 to 10%, with A
D accounting for two-thirds of these cases. If milder cases are includ
ed, the prevalence rates double. Only age and family history of dement
ia are consistently associated with AD in all studies, but putative pr
otective agents such as estrogen, NSAIDs, and vitamin E are undergoing
clinical trials.