To test the hypothesis that the APOE epsilon4 allele is associated with the clinical manifestations of AD through an association with the pathologic hallmarks of disease.
Participants were older Catholic nuns, priests, and brothers who agreed to annual neurologic and neuropsychological evaluation for AD and other common neurologic conditions and brain autopsy at the time of death. There were 77 persons without dementia and 51 with probable AD; 38 participants had one or more epsilon4 alleles.
In logistic regression analyses, controlling for age, sex, and education, the epsilon4 allele was strongly associated with the likelihood of clinical AD (odds = 3.46, 95% CI = 1.44 to 8.33). However, controlling for the effect of AD pathology, the association of the epsilon allele with clinical AD was reduced by >50% and was no longer significant (odds = 1.58, 95% CI = 0.56 to 4.43). Similarly, in linear regression analyses, controlling for age, sex, and education, the epsilon4 allele was strongly associated with level of cognitive function proximate to death (regression coefficient = -0.477, p = 0.005). However, after controlling for the effect of AD pathology, the association of the epsilon4 allele with level of cognition was reduced by >80% and was no longer significant (regression coefficient = -0.093). Similar results were found in analyses using separate measures of neuritic plaques, diffuse plaques, and neurofibrillary tangles, and in analyses of five different cognitive systems (episodic memory, semantic memory, working memory, perceptual speed, and visuospatial ability).
The APOE epsilon4 allele appears to be associated with the clinical manifestations of AD through an association with the pathologic hallmarks of AD rather than another mechanism.