Neuropathology of nondemented aging: presumptive evidence for preclinical Alzheimer disease

Abstract

Objective: To determine the frequency and possible cognitive effect of histological Alzheimer's disease (AD) in autopsied older nondemented individuals.

Design: Senile plaques (SPs) and neurofibrillary tangles (NFTs) were assessed quantitatively in 97 cases from 7 Alzheimer's Disease Centers (ADCs). Neuropathological diagnoses of AD (npAD) were also made with four sets of criteria. Adjusted linear mixed models tested differences between participants with and without npAD on the quantitative neuropathology measures and psychometric test scores prior to death. Spearman rank-order correlations between AD lesions and psychometric scores at last assessment were calculated for cases with pathology in particular regions.

Setting: Washington University Alzheimer's Disease Research Center.

Participants: Ninety-seven nondemented participants who were age 60 years or older at death (mean=84 years).

Results: About 40% of nondemented individuals met at least some level of criteria for npAD; when strict criteria were used, about 20% of cases had npAD. Substantial overlap of Braak neurofibrillary stages occurred between npAD and no-npAD cases. Although there was no measurable cognitive impairment prior to death for either the no-npAD or npAD groups, cognitive function in nondemented aging appears to be degraded by the presence of NFTs and SPs.

Conclusions: Neuropathological processes related to AD in persons without dementia appear to be associated with subtle cognitive dysfunction and may represent a preclinical stage of the illness. By age 80-85 years, many nondemented older adults have substantial AD pathology.

Figure 1

Photomicrographs of neuritic (A, double arrow) and diffuse plaques (B, single arrows), from adjacent areas of the temporal cortex of a case that was rated npAD with the Khachaturian and Washington University criteria, probable AD with CERAD criteria, and intermediate probability of AD with NIA-Reagan criteria. Stained with a double immunohistochemical procedure in which Aβ is stained red and paired helical filaments are stained black.

Figure 2

Classification of cases by four neuropathological criteria (Khachaturian, Washington University, CERAD, NIA-Reagan), N=97.

Figure 3

A. Number of npAD and no-npAD participants (defined by the Washington University criteria) at each Braak neurofibrillary stage. B. & C. Number of cases rated at each level of the CERAD (B), and NIA-Reagan (C) criteria, for each Braak neurofibrillary stage. No cases were identified at Braak neurofibrillary stage VI.

Figure 4

The correlation of age with tangle density in limbic structures (A) and with plaque burden in neocortical areas (B). There is a significant increase in tangle density with age, but no age-related increase in plaque burden.

Figure 5

There is no relation between neocortical plaque burden and limbic NFT density for 97 nondemented cases, sorted by increasing NFT density (A) and by increasing plaque burden (B). Once present in the brain, both NFT density and plaque burden vary continuously from low to high values.