Proportion of Dementia Cases in the Oldest‐old Attributable to 8 Neuropathological Changes: The 90+ Study
María M. M. Corrada, Luohua Jiang, S. Ahmad Sajjadi, Thomas J. Montine, Syed A. Bukhari, Brenna Cholerton, Patricia A Boyle, Ron Brookmeyer, Claudia H. Kawas- Psychiatry and Mental health
- Cellular and Molecular Neuroscience
- Geriatrics and Gerontology
- Neurology (clinical)
- Developmental Neuroscience
- Health Policy
- Epidemiology
Abstract
Background
People aged 90 and older have the highest rates of dementia. At advanced ages, dementia is due to a variety of diseases but the relative contribution of each is unknown. We estimated the proportion of dementia cases attributable to 8 different neuropathological indices in a cohort of individuals aged 90 and older.
Methods
Participants (n = 425) are from The 90+ Autopsy Study, a community‐based longitudinal study of aging and cognition. Most participants (62.1%) were initially part of the Leisure World Cohort Study and the remaining were non‐cohort community volunteers (37.9%). The 8 neuropathologic features examined were Alzheimer’s disease neuropathologic change (ADNC), limbic predominant age‐related tdp‐43 encephalopathy (LATE‐NC), hippocampal sclerosis (HS), Lewy body disease (LBD), microvascular lesions (MVL), cerebral amyloid angiopathy (CAA), atherosclerosis, and arteriolosclerosis. We examined the association of the 8 neuropathologic features with all‐cause dementia using multivariable logistic regression adjusting for age, gender education, and cohort affiliation. We then estimated the proportion of dementia cases that could be attributed to each neuropathologic change.
Results
Participants were on average 97.6 years at time of death, most were women (68%), and about half (43%) had dementia at time of death (Table 1). ADNC, LBD, LATE‐NC, arteriolosclerosis and atherosclerosis were significantly associated with dementia, whereas HS, MVLs, and CAA trended toward a significant association with dementia (Table 2). For the individual neuropathologic features, ADNC (26%) and arteriolosclerosis (24%) had the highest and similar attributable fraction of dementia cases, followed by LATE‐NC (17%) and CAA (10%) (Figure). When combinations of neuropathologic features were considered, vascular pathologic changes had the highest attributable fraction of dementia cases (38%), non‐AD changes (LBD, LATE‐NC, HS) had an attributable fraction of 28%, and LATE‐NC and HS combined had an attributable fraction of 20% (Figure). Overall, 78% of dementia cases could be attributed to these 8 neuropathologic features.
Conclusions
Vascular pathologic changes, as a group, account for the largest fraction of dementia cases in this oldest‐old cohort. Furthermore, when considered together, non‐AD neuropathologic changes are as important contributors to dementia as ADNC. Therefore, interventions targeting a variety of diseases will be needed to significantly reduce the burden of dementia in the oldest‐old.