Beta-amyloid deposition and other measures of neuropathology predict cognitive status in Alzheimer's disease

The relationship between progressive cognitive decline and underlying neuropathology associated with Alzheimer s disease (AD) is a key issue in defining the mechanisms responsible for functional loss. This has been a subject of much controversy, with separate studies comparing various clinical and neuropathological indices in AD. Further, it is difficult to compare studies with differences in histochemical staining protocols, brain regions examined, and data quantification criteria. There are many difficulties in designing a clinical-pathological correlative study involving AD patients. It is necessary to control for several key parameters. For example, a broad range of cognitively impaired subjects is needed, as well as short postmortem delays, brief intervals between cognitive testing and death, and the most sensitive detection and quantification techniques. In this study, we carefully controlled for each of these parameters to determine if there is a relationship between global cognitive dysfunction and multiple neuropathological indices. We selected 20 individuals representing a broad range of cognitive ability from normal to severely impaired based on the MMSE, Blessed IMC, and CDR. We counted plaque number, NFT number, dystrophic neurite number, and the relative extent of thioflavine positive plaques and neuritic involvement within plaques. We also quantified cortical area occupied by beta-amyloid immunoreactivity (A beta Load) and PHF-1 positive neuropil threads and tangles (PHF Load) using computer-based image analysis. Interestingly, we found that most pathologic measures correlated highly with the severity of dementia. However, the strongest predictor of premortem cognitive dysfunction on all three cognitive measures was the relative area of entorhinal cortex occupied by beta-amyloid deposition. In conclusion, our data show that in a carefully controlled correlative study, a variety of neuropathological variables are strongly correlated with cognitive impairment. Plaque related variables may be as strongly related to cognitive dysfunction as other established measures, including synapse loss, cell death and tau hyperphosphorylation, although no correlative study can demonstrate causality.