The role of microglial cells in the pathogenesis of Alzheimer's disease (AD) neurodegeneration is unknown. Although several works suggest that chronic neuroinflammation caused by activated microglia contributes to neurofibrillary degeneration, anti-inflammatory drugs do not prevent or reverse neuronal tau pathology. This raises the question if indeed microglial activation occurs in the human brain at sites of neurofibrillary degeneration. In view of the recent work demonstrating presence of dystrophic (senescent) microglia in aged human brain, the purpose of this study was to investigate microglial cells in situ and at high resolution in the immediate vicinity of tau-positive structures in order to determine conclusively whether degenerating neuronal structures are associated with activated or with dystrophic microglia. We used a newly optimized immunohistochemical method for visualizing microglial cells in human archival brain together with Braak staging of neurofibrillary pathology to ascertain the morphology of microglia in the vicinity of tau-positive structures. We now report histopathological findings from 19 humans covering the spectrum from none to severe AD pathology, including patients with Down's syndrome, showing that degenerating neuronal structures positive for tau (neuropil threads, neurofibrillary tangles, neuritic plaques) are invariably colocalized with severely dystrophic (fragmented) rather than with activated microglial cells. Using Braak staging of Alzheimer neuropathology we demonstrate that microglial dystrophy precedes the spread of tau pathology. Deposits of amyloid-beta protein (Abeta) devoid of tau-positive structures were found to be colocalized with non-activated, ramified microglia, suggesting that Abeta does not trigger microglial activation. Our findings also indicate that when microglial activation does occur in the absence of an identifiable acute central nervous system insult, it is likely to be the result of systemic infectious disease. The findings reported here strongly argue against the hypothesis that neuroinflammatory changes contribute to AD dementia. Instead, they offer an alternative hypothesis of AD pathogenesis that takes into consideration: (1) the notion that microglia are neuron-supporting cells and neuroprotective; (2) the fact that development of non-familial, sporadic AD is inextricably linked to aging. They support the idea that progressive, aging-related microglial degeneration and loss of microglial neuroprotection rather than induction of microglial activation contributes to the onset of sporadic Alzheimer's disease. The results have far-reaching implications in terms of reevaluating current treatment approaches towards AD.