We propose a study of the occurrence of Alzheimer's disease (AD) and other dementias in relation to genotype, age and other factors. Drawing on advances in both science and technology, the study seeks important new understanding of the causes and possible prevention of AD. Recent data from several groups show association between AD and allele epsilon4 of Apolipoprotein E (ApoE), a polymorphic lipid transporter. The allele frequency for epsilon4 in AD cases is 0.40 - 0.50 (cf. 0.014 in most populations). At equilibrium, 70% of prevalent AD cases would then bear at least one epsilon4 allele (cf. 26% of most populations.) New evidence suggests that epsilon4 is sufficient to provoke AD: in familial AD pedigrees all epsilon4 homozygotes develop AD by age 80; heterozygotes develop AD by age 90, and their disease expression appears to be maximal near age 80. We therefore propose to test two hypotheses: H1) epsilon4 is sufficient to provoke AD, not only within familial AD pedigrees but in all those bearing the allele. If epsilon4 accounts for 70% of AD cases, and disease expression in such cases is maximal near age 80, then: H2) the incidence of AD will be maximal before age 85 and decline thereafter. To test these hypotheses, we propose a prevalence and incidence analysis of AD in a remarkably long-lived and cooperative sample of 5650 elders who will have been genotyped for ApoE alleles. Simultaneously, we shall study other factors that influence disease risk. New co-twin control studies suggest that chronic exposure to anti- inflammatory drugs (AI's) is inversely associated with AD (o.r. 0.24, 95% c.i. 0.07 - 0.74). AI's may act by delaying the onset of AD, but the association of AD with AI's has been seen only in certain age or sex groups. We therefore propose to test three additional hypotheses: H3) AI's are associated with reduced incidence of AD before age 85; if the reduction in risk occurs because AI's delay the onset of AD, then (surprisingly) the incidence of AD among those with AI exposure will be increased after age 85; and H4) the relation of AI's to AD risk shows interaction with ApoE genotype, and with age or gender. Finally, we shall examine similar interactions of ApoE genotype, age and gender with other AD risk factors such as education, head injury, and family history of dementia. ApoE alleles and risk factors will be assessed in Years 1 and 2, as suspected prevalent causes of dementia are identified. In Years 2 and 3 a key informant network will identify incident cases of dementia who will be similarly evaluated. In Years 4 and 5 a new wave of screening and assessment will identify incident cases not detected by the key informants. Age-specific incidence rates will then be calculated, and each of the study's hypotheses will be tested.
Showing the most recent 10 out of 124 publications
