Affordable Access

Hippocampal connectivity and Alzheimer's dementia: effects of synapse loss and tangle frequency in a two-component model.

Authors
Type
Published Article
Journal
Neurology
Publication Date
Volume
44
Issue
11
Pages
2081–2088
Identifiers
PMID: 7969963
Source
Medline
License
Unknown

Abstract

Our prior research on patients with Alzheimer's disease (AD) found a high correspondence between premortem dementia and accumulation of neurofibrillary tangles (NFTs) with concurrent loss of synapse density in several brain regions. In the present study, we examined these same clinicopathologic relationships in the context of seven subregions of the hippocampal formation using a sample of 16 AD patients who had been administered three well-known mental status tests antemortem. We found NFT counts to be most strongly correlated with degree of dementia when they were seen in CA1, the subiculum, and CA4; NFTs in these subregions appeared significantly clustered on factor analysis. Synapse loss was most strongly correlated with dementia when it occurred in the molecular layers of the dentate fasciculus and stratum lacunosum, CA2/3, and CA4; synapse loss in these subregions appeared significantly clustered on factor analysis. In general, these results were compatible with a two-component model of hippocampal connectivity and function in the context of AD. The first component consists of subregions preceding CA1 in a hypothesized input-processing sequence intrinsic to the hippocampus that summates neuronal excitation and that influences cognition primarily through synapse density. The second component consists of an "output module," mainly CA1 and the subiculum, that receives the processed signal, passes it on to extrahippocampal cortical and subcortical targets, and affects cognition primarily by NFT accumulation in output neurons. A "net pathology" score combining standardized z-scores for synapse density and NFTs was significantly correlated with all three mental status measures in all hippocampal subregions except the entorhinal cortex, and stepwise regressions on these data found net pathology in CA4 to be the most independent significant predictor of premortem dementia.

Statistics

Seen <100 times