Risk of Alzheimer disease with the epsilon4 allele for apolipoprotein E in a population-based study of men aged 62-73 years.Alzheimer Dis Assoc Disord 1998; 12(1):40-4AD
The epsilon4 allele at APOE, the polymorphic locus for apolipoprotein E, increases the risk of Alzheimer disease (AD), especially among those with the homozygous epsilon4/epsilon4 genotype. In family studies, epsilon4 homozygotes typically develop AD at 55-75 years, an age range when AD is otherwise relatively infrequent. Population-based studies of the AD risk associated with allele epsilon4 (and especially with genotype epsilon4/epsilon4) are limited in number, and most such studies have included few AD cases between the ages of 55 and 75 years. In a large population-based twin registry, the screening of 12,709 men who were 62-73 years old yielded 38 prevalent cases of AD whose onset age ranged from 54 to 73. Genotype at APOE was determined for 37 of these cases and independently, for a similarly aged probability sample of 344 men from the same registry. The epsilon4 allele frequencies among the AD cases and the population samples were 0.39 and 0.15, respectively. The odds ratios (ORs) for AD were 17.7 for genotype epsilon4/epsilon4 versus epsilon3/epsilon3 and 13.8 for epsilon4/epsilon4 versus all remaining genotypes. By contrast, the ORs with heterozygous epsilon4/epsilon3 were only 2.76 versus epsilon3/epsilon3 and 2.01 versus all genotypes other than epsilon4/epsilon3 (p for homozygote vs. heterozygote ORs=0.002). The estimated etiologic fraction for AD with homozygous epsilon4 among men in their mid-50s to mid 70s is therefore 0.20; for the much more common heterozygous genotype epsilon4/epsilon3, the fraction is 0.18. In combination with other studies that have adjusted statistically for age, these results suggest that the effect of the epsilon4 allele dose is neither linear nor homogeneous for age. Homozygous epsilon4/epsilon4 appears to confer an extreme risk of AD at the age when onset with this genotype is most likely. These results are consistent with the view that individual genotypes modify risk by predisposing to substantially different distributions of AD onsets.