Person‐specific contribution of neuropathologies to cognitive loss in old age
Progressive loss of cognitive function is common in old age, and prevention of age‐related cognitive decline is a critical public health priority. Efforts to mitigate late life cognitive decline have met with limited success, however, in part due to the complexity of factors that contribute to cognitive aging. Whereas it is widely recognized that Alzheimer disease (AD), cerebrovascular disease, and Lewy body disease are important drivers of late life cognitive decline and dementia, recent work has shown that several additional age‐related neuropathologies (ie, TDP‐43, hippocampal sclerosis, cerebral amyloid angiopathy, atherosclerosis, arteriolosclerosis) also are common and relatively independently associated with decline.1 Mixed neuropathologies are the most common cause of the clinical syndrome of dementia, including AD dementia.9 Mixed neuropathologies also are common among persons with mild or no cognitive impairment and are related to decline.14 Importantly, however, individuals have different combinations of neuropathologies, yet prior studies have not examined the degree to which specific combinations of neuropathologies contribute to cognitive loss at an individual level. Thus, our understanding of the relative effect of any given neuropathology and the potential impact of a therapeutic agent that targets a specific neuropathology in an individual is extremely limited.
We previously reported that common age‐related neuropathologies account for less than half of the between‐person variability in late life cognitive decline.4 In this study, we examined patterns of neuropathologic comorbidity and quantified the contribution of 9 age‐related neuropathologies to pathologic cognitive loss at a person‐specific level. Participants were >1,000 deceased individuals from 2 longitudinal epidemiologic, clinical–pathologic studies of aging, who completed at least 2 cognitive evaluations (maximum = 22), died, and underwent uniform neuropathologic examinations. In analyses, we used linear mixed models to first examine the associations of the neuropathologies with cognitive decline. We then quantified each person's cognitive loss attributable to neuropathology over the entire study period and determined the proportion of cognitive loss accounted for by each relevant neuropathology at a person‐specific level.