Journal of Neuropathology and Experimental Neurology
Copyright q 2003 by the American Association of Neuropathologists
Vol. 62, No. 11
November, 2003
pp. 1087 1095
Neuropathology of Cognitively Normal Elderly
D. S. KNOPMAN, MD, J. E. PARISI, MD, A. SALVIATI, MD, M. FLORIACH-ROBERT, MD, B. F. BOEVE, MD,
R. J. IVNIK, PHD, G. E. SMITH, PHD, D. W. DICKSON, MD, K. A. JOHNSON, BS, L. E. PETERSEN, BS,
W. C. MCDONALD, MD, H. BRAAK, MD, AND R. C. PETERSEN, PHD, MD
Abstract. Despite general agreement about the boundaries of Alzheimer disease (AD), establishing a maximum limit for
Alzheimer-type pathology in cognitively intact individuals might aid in defining more precisely the point at which Alzheimer
pathology becomes clinically relevant. In this study, we examined the neuropathological changes in the brains of 39 longitudinally followed, cognitively normal elderly individuals (24 women, 15 men; age range 74–95, median 85 years). Neuropathological changes of the Alzheimer type were quantified by determining neurofibrillary tangle (NFT) staging by the method
of Braak and Braak and by quantification of the abundance of diffuse, cored, and neuritic plaque burden using the scheme
developed by the Consortium to Establish a Registry for Alzheimer Disease (CERAD). Vascular, Lewy body, and argyrophilic
grain pathology were also assessed. We found 34 subjects (87%) with a Braak stage ,IV; 32 subjects (82%) with less than
moderate numbers of cored plaques and 37 subjects (95%) with less than moderate numbers of tau-positive neuritic plaques.
Many subjects had moderate or frequent diffuse plaques (n 5 19, 49%). By the National Institute on Aging-Reagan Institute
(NIA-RI) criteria, none of our cases met criteria for high ‘‘likelihood’’ of AD. Four met NIA-RI criteria for intermediate
‘‘likelihood.’’ Seven cases met CERAD criteria for possible AD. Nineteen met Khachaturian criteria for AD. Only 1 subject
had neocortical Lewy bodies. Small, old infarcts were common, but no subjects had more than 2 of these and none had a
single large infarction. Thus, the majority of individuals who are cognitively normal near the time of their death have minimal
amounts of tau-positive neuritic pathology (Braak stage ,IV and neuritic plaques ,6 per 3100 field in the most affected
neocortical region). The few subjects with more severe AD pathology can be expected based on incidence rates of AD in the
very elderly.
Key Words:
Alzheimer; Cognition; Normal aging.
INTRODUCTION
It has been challenging to define a minimum threshold
for the neuropathological diagnosis of Alzheimer disease
(AD). In patients with a high burden of AD pathology
there are firm grounds for claiming that AD caused the
dementia. In any series there are typically a few cases
with lower burdens of AD pathology that nonetheless are
attributed to AD in the absence of a better alternative (1–
4). When AD pathology of a lesser severity is present,
causal attribution of dementia to the pathology becomes
problematic. Defining a minimum threshold for AD pathology may be confounded by the presence of other
common known pathologies (5).
To understand the pathological threshold for the development of dementia by AD, it may be more feasible
conceptually to start with a definition of the maximum
limit in cognitively intact individuals rather than with a
From the Departments of Neurology (DSK, MF-R, AS, RCP, DFB,
JEP, KAJ, LEP), Laboratory Medicine & Pathology (JEP), Psychology
and Psychiatry (RJI, GES), Mayo Clinic and Mayo Foundation, Rochester, Minnesota; the Department of Neurosciences Mayo Jacksonville
(DWD), Jacksonville, Florida; and the Klinikum der Johann Wolfgang
Goethe-Universitat (HB), Frankfurt am Main, Germany.
Correspondence to: David Knopman, MD, Department of Neurology,
Mayo Clinic, 200 First Street SW, Rochester MN 55905. E-mail:
[email protected].
This work was supported in part by grants AG 06786 (Mayo Alzheimer’s Disease Patient Registry) and AG 16574 (Mayo Alzheimer’s
Disease Research Center) from the National Institute on Aging.
minimum limit in demented individuals. If cognitive integrity can be assured at a time point close to death, there
are few issues that can compromise the determination of
a maximum limit for AD (or other) pathology in asymptomatic individuals.
Because prospective cognitive assessment is required
to insure that so called normal subjects are cognitively
intact immediately prior to their terminal illness, nondemented subjects can be identified only through largescale, prospective longitudinal studies. Only a few series
that could verify the cognitive integrity of their ‘‘normal’’
subjects have been reported (5–11). While the majority
of cognitively intact individuals had little AD pathology,
several of the series contained examples with high burdens of AD pathology (7, 8). These outlier subjects
would appear to pose a major interpretive problem.
In light of the pivotal importance of studies of wellcharacterized nondemented individuals, as well as the
variability of AD pathology of normal subjects in the
recent studies (7, 8), we analyzed the neuropathological
features in prospectively recruited, prospectively diagnosed, cognitively normal individuals in the Mayo Alzheimer’s Disease Center and the Alzheimer’s Disease
Patient Registry projects in Rochester, Minnesota. We focused on the presence of AD pathology, but also evaluated the presence and the distribution of vascular lesions
and the presence of Lewy Bodies (LB). We then analyzed
associations between these pathological findings and demographics, cognitive status, and APOE genotyping.
1087
1088
KNOPMAN ET AL
MATERIALS AND METHODS
Subjects
A total of 192 subjects who were recruited to the Mayo Alzheimer’s Disease Patient Registry (12) and whose last clinical
diagnosis was ‘‘normal’’ had died at the time of data analysis
for the present study. All subjects had been residents of Rochester, Minnesota and were recruited from the Community Internal Medicine practice of the Mayo Clinic. Individuals with
other neurological diseases such as Parkinson disease or stroke
were excluded, even if they were cognitively normal. From this
group there were 39 subjects who had had brain autopsies, who
were originally enrolled as normal controls in the ADPR, and
who were diagnosed as normal on their last clinical evaluation
that occurred within 2 years of death.
All subjects were initially examined and interviewed by a
study physician. The neuropsychological evaluations at the
baseline included cognitive and behavioral scales, only 3 of
which will be referred to in this report: the Mini-Mental State
examination (13–15), Mattis Dementia Rating Scale (16–18),
and the Wechsler Memory Scale, Revised, Logical Memory
Subtest (WMS-LM) (19). A consensus committee of neurologists and neuropsychologists determined diagnoses and subjects
were enrolled in the present study only if their histories and
neuropsychometric assessments were consistent with normal
cognitive function (12, 20). APOE genotyping was performed
according to established methods (21) from a blood sample
drawn either at the baseline examination or during clinical follow-up. The subjects were reevaluated on an annual basis with
neuropsychological testing (12, 20). The examinations during
their follow-up were discussed at the consensus conference. At
their last clinical examination prior to death, they were considered by consensus conference to be cognitively intact/normal.
All cases were examined cognitively within 2 years of death.
Two cases were examined between 1.5 and 2 years, 9 within 1
and 1.5 years, and the remainder within 1 year. All subjects
gave informed consent for participation. The study was approved by the Mayo Institutional Review Board.
Neuropathological Assessment
Neuropathological examinations were performed according
to the recommendations of the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) (1). After removal, the
brain was divided into right and left hemi-brains. One hemibrain was fixed in 10% buffered formaldehyde for 10 to 14
days and sectioned. Routinely sampled brain areas included
middle frontal gyrus (Brodmann’s area, BA 9), inferior parietal
lobule (BA 39), superior temporal gyrus (BA 22), calcarine
cortex (BA 17), anterior cingulate gyrus (BA 24), hippocampus
at the level of the lateral geniculate body, amygdala, transentorhinal, and entorhinal cortices at the level of the mammillary
bodies, nucleus basalis, cerebellum, dorsomedial thalamus with
subthalamic nucleus, and midbrain with substantia nigra. Samples were processed in paraffin and stained with hematoxylin
and eosin (H&E). Selected sections were stained with modified
Bielschowsky silver and Gallyas and immunostained with antibodies to b-amyloid (clone 6F/3D, 1/10 dilution; Novocastra
Vector Labs, Burlingame, CA); tau (clone AT8, 1/1,000 dilution; Endogen, Woburn, MA); a-Synuclein (clone LB509, 1/
200 dilution; Zymed, South San Francisco, CA); ab Crystallin
J Neuropathol Exp Neurol, Vol 62, November, 2003
(Rabbit polyclonal, 1/5,000 dilution; Chemicon, Temecula,
CA); and ET3 (generously provided by Dr. Peter Davies, New
York, NY).
The regions in which the counting was performed were selected macroscopically according to Duyckaerts et al (22) as
the areas where the cortical ribbon was the thinnest (i.e. where
the angle of section was the closest to 908 relative to the cortical
pial surface). Pathological findings were recorded according to
Dickson et al (23) from 3200 magnification fields, excluding
fields at the crest of gyri or depth of sulci. Ten contiguous fields
were examined for frontal, parietal, temporal, and calcarine cortex; 5 contiguous fields for amygdala, transentorhinal, entorhinal, and anterior cingulate cortices, and substantia nigra; 4 contiguous fields for subiculum and nucleus basalis; and 3
contiguous fields for hippocampus.
Quantification and Identification of Plaques
Three types of plaques were identified and counted and are
described as follows: 1) Diffuse plaques: neuropil deposition of
finely granular material on Bielschowsky-stained sections. 2)
Dense core plaques: neuropil deposition of compact argyrophilic material, on Bielschowsky-stained sections. 3) Neuritic
plaques: identified by the presence of dystrophic neurites, arranged radially forming a discrete spherical lesion averaging
about 30 mm in diameter (23). Neuritic plaques were counted
on tau immunostained sections.
Plaque counts were normalized and expressed as number of
counts/mm2. The number of each type of plaque was also characterized according to the CERAD rating scheme of none,
sparse (1–5 per 3100 field), moderate (6–15 per 3100 field),
or frequent (.15 per 3100 field) (1). The score of the neocortical region with the highest count was used as the overall score
for each subject.
Quantification of Neurofibrillary Tangles
Neurofibrillary tangles (NFTs) were identified as flameshaped or globoid masses of intracellular argyrophilic fibers.
Extracellular ‘‘ghosts’’ NFT were also recorded. NFTs were
counted on both Bielschowsky and anti-tau stained slides. We
will report only the staging of NFT distribution according to
the scheme of Braak and Braak (24), which was completed
primarily with Bielschowsky stained slides but supplemented
by the anti-tau stained sections. Braak Stage I represents small
numbers of NFTs confined to the trans-entorhinal cortex; stage
II, moderate or larger numbers in transentorhinal cortex and few
in hippocampus; stage III, frequent NFTs in transentorhinal and
entorhinal cortices and moderate NFTs in hippocampus (CA1
and subiculum); stage IV, more severe hippocampal involvement and mild to moderate involvement of isocortex; stage V,
frequent NFTs in association isocortex; and stage VI, NFTs in
primary sensory cortices.
Quantification of Lewy Bodies
The presence of Lewy bodies (LBs) (25) was assessed in
substantia nigra, amygdala, and cingulate cortex using a-synuclein-stained sections. In each region the total number of LBs
was counted and an LB score according to McKeith et al (26)
was recorded.
1089
NEUROPATHOLOGY OF COGNITIVELY NORMAL ELDERLY
TABLE 1
Clinical and Demographic Summary of Subjects*
Feature
Mean or
Proportion
Gender (% women)
Age in years 6 SD
Education (years) 6 SD
Last MMSE prior to death (years) 6 SD
Last DRS prior to death (years) 6 SD
Interval from last MMSE to death (months)
APOE e4 genotype (percentage)
61.5
85.4 6 5.4
13.1 6 3.1
28.0 6 1.6
134 6 5.5
7.8 6 5.8
34
Range
24 women; 15 men
74 to 95
8 to 20
24 to 30
122 to 143
0.1 to 23.2
10 e3/e4
1 e2/e4
2 e2/e3
19 e3/e3
7 missing
Median
85
13
28
135
9.4
* Abbreviations: SD, standard deviation; MMSE, Mini-Mental State Examination; DRS, Mattis Dementia Rating Scale; APOE,
apolipoprotein E.
Argyrophilic Grains
To compare quantitative and categorical aspects of the group
differences we used t-tests and chi-squared tests. Pearson or
Spearman correlations were used to assess the association between clinical and pathological features.
1.7), or percentage of women (61.5% vs 63%). An additional 58 subjects whose last diagnosis was ‘‘normal’’
had not been examined within 2 years of death. These
subjects were older (89.2 6 8.0 yr) than the 39 subjects
in the present report.
Causes of death based on the complete autopsy, or on
clinical impression if only a limited autopsy was performed, were obtained from the autopsy report. Cardiac
disease, usually acute myocardial infarct, was the most
frequent primary cause of death (17 subjects, 44%). Cancer-related illnesses caused 10 deaths (26%). Primary
pulmonary causes of death, including pulmonary embolism, pneumonia, or bronchitis accounted for most of the
remainder (13 subjects, 33%). Some patients had more
than 1 principal cause of death listed. Head trauma was
not part of the terminal illness in any subject. Recent
stroke was not listed as the principal cause of death in
any of the subjects, although acute and subacute cerebral
infarcts were observed pathologically in 10 cases.
RESULTS
NFT Pathology
Descriptive data on the 39 subjects is given in Table
1. Only 5 (13%) subjects were less than 80 years old,
while 9 subjects (23%) were 90 years or older at death.
The MMSE at the last clinical evaluation was $27 in 32
subjects; 1 subject had an MMSE score of 24; 2 had
scores of 25; and 4 had scores of 26. The DRS scores at
last clinical evaluation (mean 5 134.5) exceeded 130 (out
of a maximum score of 144) in 29 of 39 cases. The lowest score accepted among subjects as not cognitively impaired was 122.
The 39 subjects who underwent autopsy differed from
the nonautopsied cases who had been evaluated within 2
years of death (n 5 95) slightly in age (mean 85.4 6 5.4
yr vs 87 6 6.4 yr), but not in education (13.1 6 3.1 yr
vs 12.8 6 2.9 yr), final MMSE (28.0 6 1.6 vs 27.8 6
NFT distribution as described by Braak staging is given in Table 2 and Figure 1. Only 1 subject had no NFT
pathology. Twenty-two subjects (56%) had Braak stage I
or II and 11 (28%) had Braak stage III. Five (13%) subjects demonstrated Braak stage IV or higher, with 4 intermediate between Braak stages IV and V. One subject
had Braak stage V.
Two of the 5 subjects with Braak stage IV or higher
also had moderate numbers of tau-positive neuritic
plaques. One was an 88-year-old woman (case 38) who
had moderate numbers of neuritic plaques. Her last testing was 585 days prior to death, at which time her MMSE
was 28 and DRS was 140. She was said to have abruptly
declined cognitively in the last 4 months of her life by
family reports, but she was also found to have multiple
For the presence of argyrophilic grain disease, silver- and
tau-positive medial temporal (amygdaloid, entorhinal, and hippocampal) grains and tau-positive white matter coiled bodies
and ballooned amygdaloid neurons were required. We required
grains to be present in both amygdala and hippocampus. Ballooned, neurofilament, and ab-crystallin-positive neurons in
amygdala also were recorded (27). Detection of grains was confirmed by staining for ET3, a selective 4R tau antibody (28).
Vascular Pathology
The numbers and size of lacunar and large infarcts and the
presence of meningeal or parenchymal amyloid angiopathy
were recorded.
Statistical Methods
J Neuropathol Exp Neurol, Vol 62, November, 2003
1090
KNOPMAN ET AL
TABLE 2
Description of Neurofibrillary Tangle Distribution by Braak Stage and Plaque Densities
Case no.
Diffuse amyloid
plaques*
Argyrophilic dense
core plaques*
Neuritic plaques
with tau positive
neurites*
Braak and Braak
stage
17
39
3
4
1
22
24
21
14
20
31
13
9
26
18
29
15
16
10
2
19
11
7
34
12
23
6
33
5
32
36
28
40
8
35
37
30
38
25
None
None
Sparse
Sparse
Moderate
Frequent
Frequent
Sparse
Sparse
Moderate
Frequent
Frequent
Frequent
Moderate
None
None
None
Sparse
Sparse
Sparse
Sparse
Frequent
Frequent
None
Frequent
None
Sparse
None
Sparse
Sparse
None
Moderate
Moderate
Frequent
Frequent
Frequent
Frequent
Frequent
Frequent
None
None
None
None
None
Sparse
Moderate
None
Sparse
Sparse
Moderate
Moderate
Sparse
Moderate
None
None
None
None
None
None
Sparse
Sparse
None
None
Sparse
None
None
None
None
Sparse
Sparse
Sparse
Sparse
Sparse
Sparse
Moderate
Frequent
Moderate
Sparse
None
None
None
None
None
None
None
Sparse
Sparse
Sparse
Sparse
Sparse
None
None
None
None
None
None
None
None
Sparse
None
None
None
Sparse
None
None
None
None
None
None
Sparse
Sparse
Sparse
Sparse
Sparse
Moderate
Moderate
Sparse
None
I
I
I
I
I
I
I
I
I
I
I
I-II
I-II
II
II
II
II
II
II atypical
II
II
II
II-III
II-III
III
III
III
III
III
III atypical
III atypical
III atypical
III atypical
IV-V atypical
IV-V
IV-V atypical
IV-V atypical
V
* The number of each type of plaque was characterized according to the CERAD rating scheme: none 5 1; sparse 5 1 to 5
per 3100 field; moderate 5 6 to 15 per 3100 field; or frequent 5 .15 per 3100 field. The rating of the neocortical area with
the highest plaque count was used.
new cerebral infarcts at that time. The other subject (case
30) with Braak stage IV–V and frequent numbers of neuritic plaques in the temporal lobe and moderate numbers
in other cortical regions was a 74-year-old woman last
examined 284 days prior to death. She scored 29 on the
MMSE and 134 on the DRS at the time. The other 3
subjects with higher Braak stages had only sparse neuritic
plaques. These 3 subjects included an 89-year-old woman
(case 25) who scored 128 on the DRS and recalled 2/16
on the WMS-LM at 329 days prior to death; an 82-yearold man (case 37) who scored 133 on the DRS 67 days
prior to death; and an 88-year-old man (case 35) who
J Neuropathol Exp Neurol, Vol 62, November, 2003
scored only 122 on the DRS, but who recalled 16/18 on
the WMS-LM at the same testing session 423 days prior
to death.
The majority of subjects had NFT distribution that fit
well into the Braak staging system, but there were 9 instances in which the distribution of NFT in 1 region was
discordant according to the Braak scheme, with the NFT
burden in another region. These discrepancies (Table 3)
usually involved difference between adjacent stages rather than more divergent stages. There was no correlation
between Braak stage and age (r 5 0.08), education (r 5
0.010), last MMSE (r 5 20.03), or last DRS (r 5 20.24).
NEUROPATHOLOGY OF COGNITIVELY NORMAL ELDERLY
Fig. 1. Histogram of Braak and Braak stages (24) based on
tau-immunostained sections in 39 cognitively intact individuals.
Braak Stage I represents mild numbers of NFT confined to the
trans-entorhinal cortex; stage II, moderate or larger numbers in
trans-entorhinal cortex; stage III, severe involvement in transentorhinal and entorhinal cortices; NFTs in hippocampus (CA1
and subiculum); stage IV, more severe hippocampal involvement; moderate involvement of amygdala; stage V, NFTs in
association isocortex in moderate or larger numbers; and stage
VI, NFTs in primary cortices.
Plaque Pathology
CERAD plaque count ratings for diffuse plaques, cored
plaques, and neuritic plaques are given in Table 2 and
Figure 2. Nineteen (49%) subjects had moderate or frequent diffuse plaques in at least 1 neocortical region, of
whom 9 had frequent diffuse plaques. Seven (18%) subjects had moderate or greater cored plaques, one of whom
had frequent cored plaques in at least 1 cortical region.
There were no subjects with frequent neuritic plaque densities, but there were 2 (5%) who had moderate numbers
of neuritic plaques (cases 30 and 38).
1091
Fig. 2. Histogram of CERAD plaque density rating scores
for diffuse plaques on Bielschowsky stained sections, cored
plaques on Bielschowsky stained sections, and neuritic plaques
on tau-immunostained sections in 39 cognitively intact individuals. The number of each type of plaque was characterized according to the CERAD rating scheme of none, sparse (1–5 per
3100 field), moderate (6–15 per 3100 field), or frequent (.15
per 3100 field) (1). The score of the neocortical region with
the highest count was used as the overall score for each subject.
The group of 7 subjects with moderate or greater numbers of cored plaques, including 2 with moderate neuritic
plaques, were comparable in age at death (mean 5 84.9
years), months between last evaluation and death (mean
5 9.5 months), MMSE score (mean 5 28.6), and DRS
score (mean 5 137) compared to the other 32 subjects.
The APOE e4 allele was present in only 1 (14%) of these
7 subjects, while it was present in 9 (28%) of the other
32 subjects (chi-square test, not significant). The group
with moderate cored plaque densities did not have a substantially higher Braak stage (mean 5 2.6) compared to
the others (mean 5 2.1) (t-test, not significant).
TABLE 3
Descriptions of Subjects with Atypical Braak Staging
Subject
14
Braak Stage
I
2
II atypical
8
III atypical
28
III atypical
36
40
38
35
III atypical
III atypical
IV-V atypical
IV-V atypical
30
IV-V atypical
Description
Very mild involvement in superficial transentorhinal (pre-a) but some tangles present in
temporal neocortex.
Deep layers (Pri-a and pre-b) were more involved than superficial one (pre-a) both in
transentorhinal and entorhinal region.
Subiculum and amygdala were more involved than transentorhinal and entorhinal region and hippocampus (CA1).
Subiculum and amygdala were more involved than transentorhinal and entorhinal region and hippocampus (CA1).
Subiculum and amygdala were more involved than hippocampus (CA1)
Subiculum and amygdala were more involved than hippocampus (CA1).
Hippocampus (CA1) was mildly involved.
Hippocampus (CA1) and subiculum were severely involved while isocortex was only
mildly affected.
Subiculum and temporal isocortex were severely involved while frontal and parietal
isocortex were only mildly affected.
J Neuropathol Exp Neurol, Vol 62, November, 2003
1092
KNOPMAN ET AL
strictly on quantitative terms, not for the likelihood of the
pathological findings ‘‘explaining’’ the clinical syndrome.
Seven cases met CERAD criteria for possible AD (1).
Nineteen met Khachaturian criteria for AD (30).
Vascular Pathology (Table 4)
Fig. 3. Histogram of plaque density ratings as a function
of Braak and Braak stage. For Braak and Braak stages 0–I (n
5 12), stage II (n 5 11), stage III (n 5 11), and stages IV–V
(n 5 5), the percent of subjects with moderate or frequent diffuse, cored and neuritic plaques is shown.
The subjects with moderate or frequent diffuse plaque
burden (n 5 19) were indistinguishable demographically
(age, time between last MMSE and death) and cognitively (MMSE and DRS) from those with lesser diffuse
plaque scores. Diffuse plaque burden was not associated
with a difference in Braak stage. There were no differences in plaque burden of any type between those subjects who died with ischemic heart disease or from other
causes. Among the 14 subjects with ischemic heart disease, 3 subjects (21%) had moderate or greater densities
of cored plaques, compared to 5 (21%) of 24 subjects
whose autopsy reports did not list ischemic heart disease.
Of the 23 subjects with Braak stage #II, only 4 had a
maximum of moderate numbers of cored plaques. Five
of these 23 had neither diffuse, cored, nor neuritic
plaques. Of the 16 subjects with Braak stages .II, 3 had
no plaque pathology of any type. Figure 3 shows the lack
of striking relationships between plaque density and
Braak stage.
Diagnostic Classification
By the National Institute on Aging–Reagan Institute
criteria (NIA-RI) (29), none of our cases met criteria for
high ‘‘likelihood’’ of AD. Four met NIA-RI criteria for
intermediate ‘‘likelihood.’’ Because our subjects were not
demented, the NIA-RI criteria should be interpreted
By virtue of enrollment criteria, no subject had clinical
strokes except possibly during their terminal illness.
Eighteen (46%) subjects had at least 1 small old infarct,
and only 9 had more than one. Over half of the small
infarcts were in the thalamus, putamen, or caudate (20/
39 infarcts). The remainder of infarcts were in neocortical
locations (11 infarcts in 6 subjects) or white matter (8
infarcts in 6 subjects). With one exception, none of the
infarcts were greater than 0.5 cm in their greatest linear
measurement and the majority were microscopic. The exception was Case 38, who in the 4 months prior to death
suffered from multiple cerebral infarcts probably associated with a coagulopathy and pancreatic adenocarcinoma.
Lewy Body Pathology
By virtue of enrollment criteria, no subject had clinical
Parkinsonism. Five (13%) subjects had Lewy bodies (Table 5). Four demonstrated Lewy bodies in substantia nigra. Only 1 subject (Case 38) had neocortical Lewy bodies. Two subjects, including case 38, had Lewy bodies in
cingulate gyrus. Four had Lewy bodies in amygdala (with
the fifth case missing sections from that region).
Argyrophilic Grain Pathology
Twelve (31%) subjects had the findings of argyrophilic
grains confirmed on Bielschowsky-stained, tau-immunostained, and ab-crystalline-immunostained sections. In a
subset of 30 cases, grain disease was confirmed in 10
cases (33%) using the specific 4R-tau antibody ET3. A
comparison of Gallyas and ET3-stained sections in these
30 cases showed that the 2 techniques were comparable
for identifying grains. Of 10 cases, 9 were also Gallyaspositive. Only in 6 (20%) of these cases were there large
numbers of grains.
DISCUSSION
The majority of our longitudinally followed, cognitively intact elderly individuals had low burdens of AD
pathology. Depending upon the histologic finding used to
TABLE 4
Summary of Vascular Pathology
Total number of lesions
Number of subjects with at least 1 infarction
J Neuropathol Exp Neurol, Vol 62, November, 2003
Cortical
infarctions
Lacunar
infarctions, basal
ganglia, thalamus
Lacunar
infarctions in
white matter
Totals
11
6
20
12
8
6
39
18
1093
NEUROPATHOLOGY OF COGNITIVELY NORMAL ELDERLY
TABLE 5
Lewy Body Pathology in Subjects with any Lewy Bodies*
Case
Substantia
nigra
26
16
5
37
38
1
2
1
0
2
Amygdala
1
NA
2 Frequent Lewy neurites
2 Frequent Lewy neurites
2 Frequent Lewy neurites
Cingulate gyrus
0
2 Frequent Lewy neurites
0
0
2
Neocortical
0
0
0
NA
1
Regional distribution
predominant
Brainstem-amygdala
Brainstem-cingulate
Brainstem-amygdala
Amygdala
Brainstem-amygdala-cingulate
* Rating scores: 0 5 zero Lewy bodies in region; 1 5 up to 5 Lewy bodies per region; 2 5 more than 5 Lewy bodies in
region; NA 5 missing. Rating scheme is from ref. 26.
define the AD pathologic burden, 5% (moderate numbers
of neocortical neuritic plaques), 13% (Braak stages IV or
higher) to 18% (moderate cored plaques) of the subjects
showed greater AD pathology. Our findings add to the
small number of cases that have been published previously (5–11), demonstrating the modal pattern and the
outliers. We will argue that the outliers do not actually
present an interpretive problem.
Studies across the age spectrum suggest that AD neuropathology precedes clinical symptomatology by several
years or decades (31). Therefore, to derive a threshold
for clinically silent AD pathology, the presence of a few
cases with high burdens of AD pathology should be expected because there must be instances of incident AD
in a cohort with a mean age of 85 years. In 80- to 84-,
85- to 89-, and 90- to 94-year-old cohorts, the annual
incidence rates of AD in Rochester Minnesota were 1.9,
3.2, and 4.6 per 100 persons per year, respectively (32).
Furthermore, because there may be a multi-year preclinical phase of AD that does not lead to scoring at an impaired level on cognitive testing (33–38), the numbers of
subjects with higher grade AD pathology may reflect
those who would have become incident cases not just 1
year later, but perhaps 2, 3, or even 4 years in the future.
Thus, it would have actually been worrisome if a few of
our subjects, given their ages, did not have higher burdens of AD pathology.
Taking into account the admixture in our series of individuals with what has been called preclinical AD (10),
the cut-off points for clinical relevance of AD-type pathology derived from the present analysis are Braak stage
$IV or numbers of neuritic plaques $ moderate (i.e. ,6
per 3100 field in any neocortical region). Lower levels
of AD pathology, even including frequent diffuse
plaques, should not be sufficient to cause dementia. Correlational studies in demented patients have also shown
the importance of increasing burdens of neuritic plaques
(39) and neocortical NFTs (39, 40). An alternative hypothesis to the claim that there is a generalized threshold
for AD pathology to produce dementia would be that
there is inherent person-to-person variability in susceptibility to AD pathology. If this were the case, the implication is that one can never be certain whether AD
pathology is clinically relevant or not. However, our results do not support this alternative explanation. The data
show that most cognitively intact individuals have little
AD pathology.
There was no relationship between the amount of AD
pathology and any cognitive measure in our normal subjects. The lack of such a relationship could be due to
inadequate power with only 39 subjects or to the insensitivity of the clinical instruments. With these caveats noted, the conclusion to be drawn would be that the burden
of AD pathology on cognition is a discontinuous process.
Until a certain threshold is reached, AD pathology appears to exert no dramatic influence on cognition. In addition, our findings do not support an association reported
once before (41) between ischemic heart disease and
greater burdens of AD plaque pathology.
The major difficulty of testing AD pathological thresholds in dementia patients is the presence of other pathologies—particularly vascular and synucleinopathic—that
might act synergistically with AD pathology (42). These
other pathologies include both what can be measured by
standard evaluations, for example the CERAD (1) or
NIA-RI (29) approaches, but also what must be sought
by considerably more arduous means, such as microvascular disease (43). Thus, in a dementia patient with AD
pathology of, for example, Braak stage III and only
sparse neuritic plaques, we assert that AD pathology
should not be considered to be the sole etiologic agent.
Interpretation of cerebrovascular pathology is more
difficult because of the multifocal nature of stroke. In
addition, individuals with clinical strokes were excluded
from the study group. A large number of our cases had
some infarcts, but most were of the lacunar type, that is,
very small in terms of volume. There were no instances
of major vessel distribution infarcts. In the British neuropathology study, 8% of nondemented subjects had infarcts or lacunes, while 33% had infarcts, lacunes, and
small vessel disease ‘‘consisting of diffuse pallor of myelin staining in white matter associated with hyaline degeneration of subcortical arteries and arterioles, microinfarcts or a combination of these features’’ (5).
Compared to dementia patients in our population series
J Neuropathol Exp Neurol, Vol 62, November, 2003
1094
KNOPMAN ET AL
with vascular dementia clinically (44), the abundance of
cerebrovascular pathology in the normal subjects in the
present series was substantially less.
Approximately 13% of our cases had Lewy bodies,
consistent with the work of others. Age-specific prevalence rates of incidental Lewy body disease have shown
rates of 12.5% for 70- to 79-year-old individuals and
18.2% for 80- to 89-year-old individuals (25). Our results
confirm that nigral or amygdala Lewy bodies do not necessarily impair cognition.
A rather large proportion of our cases had argyrophilic
grain disease. The relevance of these grains to cognitive
function is uncertain, particularly in light of the high
numbers observed in cognitively intact subjects. The high
number of cases in our series compared to others (27)
may be due to reasons of technique as well as our heightened awareness of this entity.
The strengths of this study lie in the number of cases
studied, the prospective nature of the clinical evaluations
and diagnoses, and the uniformity of the neuropathological evaluations. Perhaps the main weakness of our study
was the possibility that we failed to recognize cognitive
decline in some of our subjects due to the insensitivity
of the cognitive instruments. Other potential weaknesses
include the possibility that the study subjects were not
representative of a community sample, and that as a result, there were biases in who was recruited. Any autopsy
study is susceptible to some biases because of the low
rate of autopsy. We were able to define the universe of
subjects from which the current sample was obtained and
found that the only difference between autopsied and not
autopsied subjects was that autopsied patients were
slightly younger. The age of our subjects is advanced and
clearly comparable to the peak of dementia, but it would
have been more informative had we had a greater representation of individuals under age 80, and especially
under age 70 years. Another weakness is that we included
subjects whose interval from last testing to death extended beyond 1 year. One of the 2 subjects in our series,
Case 38, with neuropathological measurements in the
range consistent with AD, was last examined 1.6 years
prior to death. Had she consented to evaluation 6 months
or less prior to death, she might have declined from her
rather high DRS score of 140 that she achieved 1.6 years
prior to death. Another potential criticism could be that
we have reported ranges of pathology rather than actual
counts of NFTs or plaques in cortical regions. We performed quantitative measurements of NFTs and plaques,
but semiquantitative counts have been shown to improve
reliability and to increase comparability to other studies
(1).
ACKNOWLEDGMENTS
This work is dedicated to the memory of Emre Kokmen, MD, one
of the founders of the Mayo Alzheimer’s Disease Patient Registry. Dr.
J Neuropathol Exp Neurol, Vol 62, November, 2003
Kokmen personally examined a number of the subjects in this report.
We would also like to thank the participants of the Registry for their
willingness to be part of this research project. We also thank Dr. Peter
Davies and Marisol Espinoza for the ET3 antibody. A preliminary version of this work was presented at the 73rd Annual Meeting of the
American Association of Neuropathologists, Pittsburgh, PA, June 11–
15, 1997, and the 50th Annual Meeting of the American Academy of
Neurology, Minneapolis, MN, April 1998.
REFERENCES
1. Mirra SS, Heyman A, McKeel D, et al. The Consortium to Establish
a Registry for Alzheimer’s Disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer’s disease.
Neurology 1991;41:479–86
2. Crystal HA, Dickson DW, Sliwinski MJ, et al. Pathological markers
associated with normal aging and dementia in the elderly. Ann
Neurol 1993;34:566–73
3. Tierney MC, Fisher RH, Lewis AJ, et al. The NINCDS-ADRDA
Work Group criteria for the clinical diagnosis of probable Alzheimer’s disease: A clinicopathologic study of 57 cases. Neurology
1988;38:359–64
4. Riley KP, Snowdon DA, Markesbery WR. Alzheimer’s neurofibrillary pathology and the spectrum of cognitive function: Findings
from the Nun Study. Ann Neurol 2002;51:567–77
5. Neuropathology Group of the Medical Research Council Cognitive
Function and Ageing Study. Pathological correlates of late-onset
dementia in a multicentre, community-based population in England
and Wales. Lancet 2001;357:169–75
6. Tomlinson BE, Blessed G, Roth M. Observations on the brains of
non-demented old people. J Neurol Sci 1968;7:331–56
7. Davis DG, Schmitt FA, Wekstein DR, Markesbery WR. Alzheimer
neuropathologic alterations in aged cognitively normal subjects. J
Neuropathol Exp Neurol 1999;58:376–88
8. Hulette CM, Welsh-Bohmer KA, Murray MG, et al. Neuropathological and neuropsychological changes in ‘‘normal’’ aging: Evidence for preclinical Alzheimer disease in cognitively normal individuals. J Neuropathol Exp Neurol 1998;57:1168–74
9. Morris JC, Storandt M, McKeel DW, Jr., et al. Cerebral amyloid
deposition and diffuse plaques in ‘‘normal’’ aging: Evidence for
presymptomatic and very mild Alzheimer’s disease. Neurology
1996;46:707–19
10. Price JL, Morris JC. Tangles and plaques in nondemented aging and
‘‘preclinical’’ Alzheimer’s disease. Ann Neurol 1999;45:358–68
11. Berg L, McKeel DW, Jr., Miller JP, et al. Clinicopathologic studies
in cognitively healthy aging and Alzheimer’s disease: Relation of
histologic markers to dementia severity, age, sex, and apolipoprotein E genotype. Arch Neurol 1998;55:326–35
12. Smith GE, Bohac DL, Waring SC, et al. Apolipoprotein E genotype
influences cognitive ‘phenotype’ in patients with Alzheimer’s disease
but not in healthy control subjects. Neurology 1998;50:355–62
13. Folstein MF, Folstein SE, McHugh PR. ‘‘Mini-mental state’’. A
practical method for grading the cognitive state of patients for the
clinician. J Psychiatr Res 1975;12:189–98
14. Tangalos EG, Smith GE, Ivnik RJ, et al. The Mini-Mental State
Examination in general medical practice: Clinical utility and acceptance. Mayo Clin Proc 1996;71:829–37
15. Crum RM, Anthony JC, Bassett SS, Folstein MF. Population-based
norms for the Mini-Mental State Examination by age and educational level. JAMA 1993;269:2386–91
16. Mattis S. Mental status examination for organic mental syndromes
in the elderly patient, in geriatric psychiatry. A handbook for psychiatrists and primary care physicians. In: Bellak L, Karasu TE,
eds. New York: Grune and Stratton, 1976:77–101
17. Lucas JA, Ivnik RJ, Smith GE, et al. Normative data for the Mattis
Dementia Rating Scale. J Clin Exp Neuropsychol 1998;20:536–47
NEUROPATHOLOGY OF COGNITIVELY NORMAL ELDERLY
18. Smith GE, Ivnik RJ, Malec JF, et al. Psychometric Properties of the
Mattis Dementia Rating Scale. Assessment 1994;1:123–32
19. Wechsler DA. Wechsler Memory Scale-Revised. New York: Psychological Corporation, 1987
20. Petersen RC, Kokmen E, Tangalos E, Ivnik RJ, Kurland LT. Mayo
Clinic Alzheimer’s Disease Patient Registry. Aging (Milano) 1990;
2:408–15
21. Petersen RC, Smith GE, Ivnik RJ, et al. Apolipoprotein E status as
a predictor of the development of Alzheimer’s disease in memoryimpaired individuals. JAMA 1995;273:1274–78
22. Duyckaerts C, Bennecib M, Grignon Y, et al. Modeling the relation
between neurofibrillary tangles and intellectual status. Neurobiol
Aging 1997;18:267–73
23. Dickson DW, Crystal HA, Mattiace LA, et al. Identification of normal and pathological aging in prospectively studied nondemented
elderly humans. Neurobiol Aging 1992;13:179–89
24. Braak H, Braak E. Neuropathological staging of Alzheimer-related
changes. Acta Neuropathol (Berl) 1991;82:239–59
25. Gibb WR, Lees AJ. The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson’s disease. J Neurol Neurosurg Psychiatry 1988;51:745–52
26. McKeith IG, Galasko D, Kosaka K, et al. Consensus guidelines for
the clinical and pathologic diagnosis of dementia with Lewy bodies
(DLB): Report of the consortium on DLB international workshop.
Neurology 1996;47:1113–24
27. Braak H, Braak E. Argyrophilic grain disease: Frequency of occurrence in different age categories and neuropathological diagnostic criteria. J Neural Transm 1998;105:801–19
28. Togo T, Sahara N, Yen SH, et al. Argyrophilic grain disease is a
sporadic 4-repeat tauopathy. J Neuropathol Exp Neurol 2002;61:
547–56
29. Hyman BT, Trojanowski JQ. Consensus recommendations for the
postmortem diagnosis of Alzheimer disease from the National Institute on Aging and the Reagan Institute Working Group on diagnostic criteria for the neuropathological assessment of Alzheimer
disease. J Neuropathol Exp Neurol 1997;56:1095–97
30. Khachaturian ZS. Diagnosis of Alzheimer’s disease. Arch Neurol
1985;42:1097–1105
31. Braak H, Braak E. Frequency of stages of Alzheimer-related lesions
in different age categories. Neurobiol Aging 1997;18:351–57
1095
32. Edland SD, Rocca WA, Petersen RC, Cha RH, Kokmen E. The
incidence of Alzheimer’s disease does not vary by gender in Rochester, Minnesota. Arch Neurol 2002;59:1589–93
33. Tierney MC, Szalai JP, Snow WG, Fisher RH. The prediction of
Alzheimer disease. The role of patient and informant perceptions
of cognitive deficits. Arch Neurol 1996;53:423–27
34. Masur DM, Sliwinski M, Lipton RB, Blau AD, Crystal HA. Neuropsychological prediction of dementia and the absence of dementia
in healthy elderly persons. Neurology 1994;44:1427–32
35. Jacobs DM, Sano M, Dooneief G, et al. Neuropsychological detection and characterization of preclinical Alzheimer’s disease. Neurology 1995;45:957–62
36. Katzman R, Aronson M, Fuld P, et al. Development of dementing
illnesses in an 80-year-old volunteer cohort. Ann Neurol 1989;25:
317–24
37. Howieson DB, Dame A, Camicioli R, et al. Cognitive markers preceding Alzheimer’s dementia in the healthy oldest old. J Am Geriatr
Soc 1997;45:584–89
38. Elias MF, Beiser A, Wolf PA, et al. The preclinical phase of Alzheimer disease: A 22-year prospective study of the Framingham
cohort. Arch Neurol 2000;57:808–13
39. Nagy Z, Esiri MM, Jobst KA, et al. Relative roles of plaques and
tangles in the dementia of Alzheimer’s disease: Correlations using
three sets of neuropathological criteria. Dementia 1995;6:21–31
40. Arriagada PV, Growdon JH, Hedley-Whyte ET, Hyman BT. Neurofibrillary tangles but not senile plaques parallel duration and severity of Alzheimer’s disease. Neurology 1992;42:631–39
41. Sparks DL, Hunsaker JC 3rd,Scheff SW, et al. Cortical senile
plaques in coronary artery disease, aging and Alzheimer’s disease.
Neurobiol Aging 1990;11:601–7
42. Nagy Z, Esiri MM, Jobst KA, et al. The effects of additional pathology on the cognitive deficit in Alzheimer disease. J Neuropathol
Exp Neurol 1997;56:165–70
43. White L, Petrovitch H, Hardman J, et al. Cerebrovascular pathology
and dementia in autopsied Honolulu-Asia aging study participants.
Ann N Y Acad Sci 2002;977:9–23
44. Knopman D, Parisi JE, Rocca WA, et al. Vascular dementia in a
community-based autopsy study. Arch Neurol 2003;60:569–76
Received May 14, 2003
Revision received July 7, 2003
Accepted July 15, 2003
J Neuropathol Exp Neurol, Vol 62, November, 2003