Alzheimer's disease and its co-pathologies: Implications for hippocampal degeneration, cognitive decline, and the role of APOE ε4.

INTRODUCTION: In neurodegenerative dementias, the co-occurrence and interaction of amyloid β peptide (Aβ), tau pathology, and other pathological lesions confound their individual contributions to neurodegeneration and their modulation by risk factors. METHODS: We analyzed 480 post mortem human brains (ages 50-99) using regression and structural equation models to assess the relationships among Aβ, tau, limbic-predominant age-related TDP-43 encephalopathy neuropathological changes (LATE-NC), α-synuclein, other age-related lesions, and apolipoprotein E (APOE) ε4, as well as their effects on CA1 neuronal density, brain weight, and cognitive status. RESULTS: Aβ, tau, LATE-NC, and amygdala-predominant α-synuclein pathology were mutually interdependent. Tau was the strongest predictor of global neurodegeneration, while LATE-NC primarily, but not exclusively, affected hippocampal neuron loss. Small vessel disease correlated with both LATE-NC and α-synuclein, while APOE ε4 was mainly associated with extracellular parenchymal and capillary Aβ pathology. DISCUSSION: Although Alzheimer's disease pathology plays a central role in brain degeneration, coexisting pathologies can both exacerbate and independently contribute to it. These factors should be considered in patient stratification. HIGHLIGHTS: In aging individuals, amyloid β peptide (Aβ), tau pathology, limbic-predominant age-related TDP-43 encephalopathy neuropathological changes (LATE-NC), and amygdala-predominant α-synuclein pathology were interrelated but contributed independently to neurodegeneration. LATE-NC was the strongest driver of CA1 neuronal loss, while tau burden was the strongest predictor of global brain degeneration. Apolipoprotein E ε4 was associated with both extracellular and capillary Aβ deposits, but not with tau burden. Temporal lobe small vessel disease was associated with both LATE-NC and amygdala-predominant α-synuclein pathology. Neural network models can reliably identify hippocampal pyramidal neurons on hematoxylin-stained histological slides.