Alzheimer disease (AD) appears to result from a combination of genetic and environmental determinants, but the nature of these etiologic factors is obscure. To elucidate the causes of AD and their effects, we propose a longitudinal study of this disease in 10,000 twin pairs of the National Academy of Sciences Registry of aging twin veterans. Some 375 cases in 270 twin pairs will be ascertained over 5 years by initial telephone screening of the Registry, with subsequent formal diagnostic assessment and annual follow-up of screen-positive subjects and their co-twins. We will assess influence of genetic factors in AD by comparing age- specific concordance rates in monozygotic (MZ) and dizygotic (DZ) pairs, as well as pairs of unrelated individuals. We will characterize the distributions of onsets in MZ and DZ co-twins after onset in the first twin. We will estimate the uncensored lifetime risks of AD among MZ or DZ twins of affected individuals; if (as some studies suggest) AD is a Mendelian dominant trait with complete but age-dependent penetrance, these risks should approach 100% and 50% respectively. We will also investigate variability of onset due to heterogeneity of genetic determinants, and variability due to varying environmental influences, by contrasting within-pair and among-pair variation of AD onset times in MZ pairs, and by application of methods relying on age-specific concordance data in both MZ and DZ pairs. We will examine effects of genetic background on timing of onset by contrasting onset variability among concordant DZ vs. MZ pairs. We will investigate possible variations in genetic loading for AD among different ethnic groups by comparing incidence and prevalence among several ethnic groups in the Registry. Even in MZ twin pairs, age at onset of AD can vary substantially. This fact suggests that non-genetic factors may influence onset, and we will therefore examine other host or environmental risk factors for AD by comparing risk factor exposures in discordant MZ pairs whose genotype, an otherwise strong risk factor, is identical. We will further examine the relationship of such risk factors to onset differences in MZ pairs, and test whether risk factors operate by accelerating onset of disease in susceptible individuals. Finally, we will examine whether the above genetic and/or environmental factors may influence variations in course and progression of disease, as they may act on onset. This study will complement the current molecular search for genes that predispose to AD by examining evidence to support or refute such genetic influences, by investigating the age-dependent expression of AD (and hence its predisposing genes, if such exist), and by studying the possible influence of non-genetic host or environmental factors on disease expression. Even if susceptibility to AD proves ultimately to be genetically determined, this study holds potential for the identification of two important strategies for prevention...