Supplementary Table 1 | Studies reflecting the heterogeneity of neuropathological findings in elderly peopleover the age of 80 years, with or without dementia.

Study / Inclusion criteria, participants / Results / Comments
Crystal et al.
2000
Einstein Aging Study
(USA)1 / 35 clinically nondemented subjects (average age 86.9 years; SD 6.1),
28 demented subjects with known AD pathology (average age 89.1 years; SD 5.8), and
20 demented subjects who in pathological evaluation did not meet the diagnostic criteria for AD or DLB (labeled as DUE) / DUE accounted for 5% of all cases of dementia among patients in their 70s, 21% of patients in their 80s, and 48% of patients in their 90s.
90% of the subjects with DUE fit into one or more of three clinicopathological groups (hippocampal sclerosis, leukoencephalopathy, and multi-infarct dementia. Other pathological findings in DUE include cribiform changes, CVAs, and CAA.
30% of patients with DUE were given a clinical diagnosis of AD. / A significant percentage of dementia patients older than 80 years do not meet the pathological criteria for AD,DLB, or FTD; contribution of AD pathology to dementia in this patient population may be overestimated.
Incidence of non-AD pathology progressively increases beyond the age of 70 years and approaches 50% in nonagenarians.
The cause of dementia (i.e. deficits beyond memory loss) in hippocampal sclerosis remains unknown and may involve other as yet unidentified processes.
White et al.
2005
Honolulu–Asia Aging Study
(USA)2 / 363 Japanese American men (average age 85.5 years; SD 5.2) with prospective neuropsychological and postmortem neuropathological evaluation / Of 56% of subjects clinically diagnosed with AD, only 19% had plaques and tangles as the sole or dominant dementia-related lesions at autopsy. Other findings were 25% CVA/micro-infarcts, 40% mixed, and 14% unknown.
Analysis of variance in cognitive testing scores, with percentages attributed to each factor, were as follows:
Education 1.8%
Plaques and tangles 15.3%
Microinfarcts 4.8%
Cortical Lewy bodies 2.6%
Hippocampal sclerosis 2.8%
Atrophy 13.0%
Age at death 0.4%
Unexplained 59.3% / In this population of JapaneseAmerican men, a great deal of discrepancy existed between clinical diagnosis and pathological findings, especially for those diagnosed with AD.
Late-life cognitive impairment and dementia is often a combination of five neuropathological findings: AD, microvascular lesions, cortical Lewy bodies, hippocampal sclerosis, and diffuse atrophy/neuronal loss not associated with the above four pathologies (unknown).
Schneider et al.
2007
Rush Memory and Aging Project
(USA)3 / 141 consecutive autopsies from community-dwelling elderly who had complete neuropsychological evaluations, (average age 87.8 years; SD 5.6) / In those without dementia, 28.6% had no significant abnormalities, 24.2% had pure AD, and 17.6% had infarctions.
In those with dementia, 38% had AD and infarcts, 30% had pure AD, 12% had vascular dementia and AD with DLB. / After accounting for age, patients who had multiple pathologies were three times more likely to exhibit dementia than were those with only one pathology (OR 2.8; 95% CI 1.2–6.7)
Mixed brain pathologies accounted for most dementia cases in patients beyond the age of 80 years.
Sonnen et al.
2007
Adults Changes in Thought
(USA)4 / 221 consecutive autopsies from community-dwelling elderly, in three categories of normal (average age 84.7 years; SD 6.9; n=89), intermediate (average age 87.6 years; SD 7.0; n=47), and demented (average age 88.7 years; SD 5.9; n=75) / After adjusting for age, sex, education, and ApoE, independent correlates of dementia included:
Braak stage V/VI vs 0/I/II: RR 5.89 (95% CI 1.62–17.60; P<0.05),
More than two infarcts vs none: RR 4.80, 95% CI 1.91–10.26; P<0.001)
DLB vs none: RR 5.08;95% CI 1.37–18.96; P<0.05) / Estimates of adjusted population-attributable risk for Braak stage is 45%, for infarcts is 33%, and for Lewy bodies is 10%.
Interventions to reduce risk of infarcts can be helpful to partially prevent or delay the onset of dementia.
Haroutunian et al.
2008
Mount SinaiSchool of Medicine Department of Psychiatry Brain Bank
(USA)5 / Subjects with neuropsychological and postmortem neuropathological evaluations, in three age groups:
60–80 years (young old)
81–89 years (middle old)
90–107 years (oldest old) / Association between AD plaques and tangles and worsening
Clinical Dementia Rating (5 compared with 0) diminished with increasing age.
Plaques increased 22.84-fold in youngold, 13.79 in middleold, and 10.32 in oldest old.
Tangles increased 27-fold in the youngold, 18.19-fold in middleold, and 17.06-fold in oldestold. / Neuropathological features of dementia among individuals beyond the age of 80 or 90 years are different from those among individuals in their 70s.
Other factors such as infarcts, DLB, hippocampal sclerosis, or factors as yet unidentified, may be contributing to the dementia in 80+ elderly.
Savva et al.
2009
Medical Research Council Cognitive Function and Aging Study
(UK)6 / 456 participants with full prospective neuropsychological testing and postmortem neuropathological evaluation (ages 63–103 years)
Five age groups:
<80 years (n=89), 80–84 years (n=79),85–89 years (n=125), 90–94 years (n=104), and >95 years (n=59) / The association between neocortical neuritic plaques and dementia was strong at age 75 years (OR 8.63; 95% CI 3.81–19.60),but diminished at age 95 years (OR 2.48; 95% CI 0.92–4.14).
Similarly, OR for hippocampal tangles and dementia was strong at age 75 years (8.61; 95% CI 3.66–20.27) and diminished at age 95 years (2.11; 95% CI, 1.05–4.25).
Neocortical cerebral atrophy maintained a strong relationship with dementia at age 75 years (OR 5.11; 95% CI 1.94–13.46) and at age 95 years (OR 6.10; 95% CI 2.80–13.28). / The association between neocortical and hippocampal atrophy was better than the related association with AD pathology in distinguishing the cohort with dementia from the cohort without dementia (in all age groups).
Therapeutic interventions that target AD pathology may prove to be effective for elderly in their 70s but not for those in their 80s and 90s.
White
2009 HonoluluAsia Aging Study
(USA)7 / 443 Japanese-American men with prospective cognitive and postmortem neuropathological evaluations; four subgroups based on age:
72–79 years
80–84 years
85–90 years
90+ years / With increasing age, the prevalence of severe cortical atrophy increased consistently, while the prevalence of severe AD pathology decreased in elderly beyond the age of 90 years.
Among patients with a clinical diagnosis of dementia, 18.6% had ‘pure’ AD pathology, 33.8% had infarcts, 10.9% had cortical Lewy bodies, 3.3% had hippocampal sclerosis, 14.2% had mixed pathologies, 7.1% had atrophy only, and 12.1% had only negligible findings.
Among those with no significant cognitive impairment, 11.6% had AD, 21.7% had infarcts, 5.8% had Lewy bodies, 7.9% had mixed pathology, 7% had atrophy only, and 46% had negligible findings. / Some of the lesion types (e.g. AD plus infarcts) appeared to be more closely associated with dementia than with individual pathological lesions.
Higher percentages of cases with dominant finding of infarcts may be related to the composition of the participants being only elderly men.
Schneider et al.
2009
Religious Orders Study and the Rush Memory and Aging Project
(USA)8 / Consecutive postmortem autopsies from 483 participants with full neuropsychological evaluation and a diagnosis of normal (average age 83.9 years; SD 6.6; n=170), MCI (average age 86.0; SD 6.3; n=134), or AD (average age 89.7 years; SD, 5.6; n=179) / Among 179 subjects with an AD clinical diagnosis, 87.7% had AD pathology, 39.7% had infarcts, 16.8% had DLB, 45.8% had mixed dementia, and 6.2% had no evidence of AD/infarcts/DLB.
Among 134 subjects with MCI, 54.5% had AD pathology, 32.8% had infarcts, 6.0% had DLB, 19.4% had mixed pathologies, and 26.1% did not have AD/infarcts/DLB.
Among 170 cognitively normal subjects, 38.2% had AD, 22.9% had infarcts, 2.9% had DLB, 9.4% had mixed pathologies, and 45.3% had no AD/infarcts/DLB. / The OR of clinically probable AD increases significantly when different neuropathologic lesions are combined:
4.7 (95% CI 3.2–6.9): AD pathology alone
1.6 (95% CI 1.1–2.4): Infarcts only
2.1 (95% CI 1.3–3.4): Lewy bodies
7.4 (95% CI 4.3–13.6): AD and infarcts
9.9 (95% CI 5.2–18.9): AD and Lewy bodies
16.2 (95% CI 7.4–35.7): AD, infarcts, and Lewy bodies
Clinically diagnosed AD represents a heterogeneous disorder, with most elderly exhibiting mixed pathologies.
Erten-Lyons et al.
2009
The Oregon Health & Science University Layton Aging and AD center
(USA)9 / Elderly participants were matched for comparable high burden of AD pathology; 24 had a clinical diagnosis of AD (average age 88.75 years; SD 10.25), and 12 had normal cognition (average age 91.27 years; SD 3.52) / Multiple regression analysis (after adjusting for sex, Braak and CERAD ( Consortium to Establish a Registry for AD) scores, presence of ischemic, hemorrhagic, or vascular pathology) showed a significant association between clinical AD (with cognitively intact as reference) and hippocampal volume(P=0.006), as well as total brain volume (P=0.009). / Larger hippocampal and total brain volume allows elderly to remain cognitively healthy despite having a high burden of AD pathology.
Factors that contribute to larger hippocampal and total brain volume have yet to be identified.

Abbreviations: AD, Alzheimer disease; CVA, cerebrovascular accident; CAA, cerebral amyloid angiopathy; DLB, dementia with Lewy bodies; DUE, dementia of unknown etiology; FTD, frontotemporal dementia; OR, odds ration; CDR, clinical dementia rating; MCI, mild cognitive impairment; ApoE, apolipoprotein E; RR, relative risk; 80+ elderly, elderly people >80 years of age.

1.Crystal, H. A. et al. The relative frequency of “dementia of unknown etiology” increases with age and is nearly 50% in nonagenarians. Arch. Neurol.57, 713–719 (2000).

2. White, L. et al. Recent clinical–pathologic research on the causes of dementia in late life: update from the Honolulu–Asia Aging Study. J. Geriatr. Psychiatry Neurol.18, 224–227 (2005).

3.Schneider, J. A., Arvanitakis, Z., Bang, W. & Bennett, D. A. Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology69, 2197–2204 (2007).

4.Sonnen, J.A. et al. Pathological correlates of dementia in a longitudinal, population-based sample of aging. Ann Neurol62, 406-13 (2007).

5.Haroutunian, V. et al. Role of the neuropathology of Alzheimer disease in dementia in the oldest-old. Arch. Neurol.65, 1211–1217 (2008).

6.Savva, G. M. et al. Age, neuropathology, and dementia. N. Engl. J. Med.360, 2302–2309 (2009).

7.White, L. Brain lesions at autopsy in older Japanese-American men as related to cognitive impairment and dementia in final years of life: a summary report from the Honolulu–Asia Aging Study. J. Alzheimers Dis. doi:10.3233/JAD-2009-1178

8.Schneider, J. A., Arvanitakis, Z., Leurgans, S. E. & Bennett, D. A. The neuropathology of probable Alzheimer disease and mild cognitive impairment. Ann. Neurol.66, 200–208 (2009).

9.Erten-Lyons, D. et al. Factors associated with resistance to dementia despite high Alzheimer disease pathology. Neurology72, 354–360 (2009).