Lewy Body Pathology in Normal Elderly Subjects

Abstract

Lewy body and Lewy neurite formation are the hallmark neuropathological findings in Parkinson’s disease (PD), Parkinson’s disease with dementia (PDD), dementia with Lewy bodies (DLB), and other alpha-synucleinopathies. They also have been described in the brains of normal older individuals and referred to as incidental Lewy body disease. The purpose of this study was to determine the prevalence of Lewy bodies and Lewy neurites (Lewy body pathology; LBP) in 139 autopsies from our normal volunteer control group of the University of Kentucky Alzheimer’s Disease Center. All subjects were followed longitudinally and were cognitively normal without any type of movement disorder, neuropsychiatric features, or other CNS findings. Thirty-three out of 139 normal subjects contained LBP in various brain regions. The most common regions involved were the medulla (26%), amygdala (24%), pons (20%), and midbrain (20%). No mean statistical differences were found between those with and without LBP on any demographic or cognitive variable, Braak stage, or neurofibrillary tangle and neuritic plaque quantitation. The high prevalence of LBP in our elderly, well educated group is not clear although it does not appear to be related to aging or the presence of AD pathology. Overall, our findings support the concept that incidental Lewy body disease most likely represents preclinical or presymptomatic PD, PDD or DLB.

INTRODUCTION

Lewy bodies (LB) and Lewy neurites (LN) are the histopathologic features of Parkinson’s disease (PD), Parkinson’s disease with dementia (PDD), dementia with Lewy bodies (DLB), and the mixed form of Alzheimer’s disease (AD) and DLB. They are also found in other neurodegenerative disorders such as progressive supranuclear palsy, corticobasal degeneration, multiple system atrophy, rare cases of neurodegeneration with brain iron accumulation type 1 (Hallervorden-Spatz disease), and several other disease entities (1). In the brainstem, LB are intracytoplasmic eosinophilic hyaline inclusions with a central core surrounded by a halo-like corona. In the cerebral cortex, LB do not have a corona, are sometimes irregular in shape, and some are smaller than those observed in the brainstem. Ultrastructurally, LB are composed of densely packed filaments associated with dense granular material in the core and radially arranged intermediate filaments in the corona (2, 3). The major component of LB and LN is aggregated alpha-synuclein that normally functions as a presynaptic protein.

Lewy bodies and LN also have been described in the brains of normal individuals (4-15) and these findings are often referred to as incidental Lewy body disease (16). Some of these studies used nonspecific LB markers (4-6) and antedated the use of alpha-synuclein immunohistochemistry, a more sensitive and specific method that is currently used to detect Lewy body pathology (LBP). The present study describes LBP in various brain regions in 33 of 139 asymptomatic, cognitively normal, longitudinally followed elderly control subjects.

MATERIALS AND METHODS

Participants

The subjects studied in this report were from the University of Kentucky Alzheimer’s Disease Center normal volunteer control group (17, 18). The present study represents 139 autopsies from 1990 to mid 2008, a period in which we performed 208 autopsies from the normal control group. Excluded from this study were normal subjects who developed DLB, mild cognitive impairment (MCI), AD, or other disorders as determined by clinical consensus conference diagnosis and re-reviewed in a neuropathology consensus conference that combined clinical and neuropathologic findings. Thus, the subjects in this study were cognitively normal elderly without any type of movement disorder, neuropsychiatric features, or other CNS findings. The details of the recruitment and inclusion criteria for the normal control group have been described previously (19). All subjects were contacted at 6-month intervals and had annual detailed cognitive function assessments and neurologic and physical examinations biannually or annually. Since 2000, all subjects underwent annual neurologic and physical examinations including the Unified Parkinson’s Disease Rating Scale. The details of the annual cognitive assessments of our subjects have been described previously (19). Initially the tests of cognition included the mental status procedures of the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) (20), some of the procedures used in the Washington University study of memory and aging project testing battery (21), those used by Eslinger et al. (22) supplemented by the Vocabulary and Digit Symbol subtests from the Weschler Intelligence Scale - revised (23) at baseline, and the Trail Making test from the Halstead-Reitan Neuropsychological Test Battery (24). Since 2005 we have used the standard test battery as required by the National Alzheimer’s Coordinating Center for all NIA-funded Alzheimer’s Disease Centers that includes most of these measures, plus other neuropsychological tests.

All subjects had normal cognition, intact activities of daily living, and normal neurologic examinations. None carried the diagnosis of PD, PDD, preclinical DLB, or other neurological diseases, and were not being treated with anti-Parkinson medications.

Tissue Sampling and Processing

The gross neuropathologic evaluation including surface structures, sections of left hemisphere, and right medial temporal lobe structures was carried out at the time of autopsy. Specimens for biochemical and molecular biological studies were taken from the left hemisphere along with adjacent sections for histologic evaluations that included 24 different brain regions. Evaluation of the right hemisphere, brainstem, and cerebellum was performed after fixation.

Sections were taken from the left cerebral hemisphere at the time of autopsy from the following regions and fixed in 4% formaldehyde: frontal pole (Brodmann area 10), middle frontal gyrus (area 9), gyrus rectus (area 11), temporal pole (area 38), superior and middle temporal gyri (areas 21, 22), inferior parietal lobule (areas 39, 40), occipital lobe (areas 17, 18), anterior cingulate gyrus (area 24), and posterior cingulate gyrus (area 23). Specimens were also taken from the hippocampus at the level of the lateral geniculate nucleus, entorhinal cortex, amygdala, ambient gyrus, basal ganglia, nucleus basalis of Meynert, thalamus, midbrain, and cerebellum. Similar sections were taken from the right hemisphere of most cases following fixation of the brain and additional sections were taken from the cerebellum, midbrain, pons, and medulla. All specimens were processed in the usual manner and sections were cut at 8-μm thickness. Sections were stained with hematoxylin and eosin and the modified Bielschowsky method. The Gallyas stain was used for sections of the entorhinal cortex, hippocampus, and amygdala. All sections of the cortex and ventromedial temporal lobe structures were stained with 10D-5 (Athena Neurosciences, San Francisco, CA) or the amyloid beta antibody (Novocastra, Newcastle, United Kingdom).

For assessment of LBP, the alpha-synuclein mouse monoclonal antibody (Novocastra, Newcastle, United Kingdom) immunohistochemistry was used. Immunohistochemistry for alpha-synuclein was performed on 10-μm thick sections that were pretreated with formic acid, blocked in 15% filtered horse serum in automation buffer, incubated with primary antibody for 1 hour, and developed with the avidin-biotin complex using Nova Red (Vector Laboratories, Burlingame, CA) as the chromagen.

Neurofibrillary tangles (NFT), diffuse plaques (DP), and neuritic plaques (NP) were counted using Bielschowsky-stained sections of frontal area 9, middle temporal gyrus, inferior parietal lobule, and occipital area 18. DP and NP were quantified in the amygdala, hippocampal CA1 and subiculum, and entorhinal cortex using the Bielschowsky-stained sections. Gallyas stained sections were used to quantify NFT in hippocampal CA1and subiculum, amygdala, and entorhinal cortex. Senile plaques, DP, and NP were counted as previously described (25). Semiquantitation of LBP in the different brain regions of varying size presented some difficulties. For example, the small locus ceruleus/subceruleus compared to the larger and less compact amygdala or neocortex or the elongated substantia nigra made it difficult to set specific LB or LN numbers. Thus, there was considerable subjectivity included in the scoring of severity. With this in mind, LBP was semiquantitated on a scale of 0 to 3 with 1+ (mild) being from 1 to 2 LB/20x field and no or a few LN in the region being studied (the rare cases with a single LB were not included); 2+ (moderate) being 3-6 LB/20x field, and scattered to mild LN per region; and 3+ (severe) being > 6 LB/20x field and moderate to severe LN per region (Figure A thru 1F).

FIGURE.

All figures are alpha-synuclein immunostained. (Magnifications are the original magnifications.) Moderate (Grade 2)Lewy body and Lewy neurite formation in the dorsal motor glossopharyngeal/vagal complex of the medulla (40x). Inset: Two Lewy bodies (arrows) (A) (200x). Moderate (Grade 2)Lewy bodies in the locus ceruleus (arrows) (B) (400x). Moderate (Grade 2)Lewy bodies (arrows) and Lewy neurites in a portion of the substantia nigra (100x). Inset: Higher magnification of Lewy body (C) (500x). Severe (Grade 3)involvement of amygdala with many Lewy bodies and Lewy neurites (D) (200x). Moderate (Grade 2)Lewy body and Lewy neurite formation in the nucleus basalis of Meynert (E) (300x). Mild (Grade 1)Lewy body formation in the temporal pole cortex (F) (100x.)

Statistical Analyses

Comparisons of demographic, cognitive, and neuropathologic findings involved Chi-squared tests for frequency data and paired t-tests for mean differences using SYSTAT (version11.0). Analyses were adjusted for multiple comparisons.

RESULTS

No mean differences were found between the two groups on any of the demographic or cognitive variables (Table 1).

TABLE 1.

Demographics and Other Features

Variable	LBP	No LBP
Subjects	33	106
Sex (M:F)	18:15	45:61
Age at death (years)	85.58 ± 7.91	83.70 ± 7.87
Mean education (years)	16.06 ± 2.45	16.07 ± 2.41
Mean time from last evaluation to autopsy (months)	8.94 ± 6.12	8.60 ± 6.83
Mean postmortem interval (hours)	4.45 ± 4.28	4.95 ± 9.69

LBP, Lewy body pathology present; No LBP, Lewy body pathology not present.

The 139 normal subjects were cognitively intact on enrollment and remained so during follow-up. The control group was composed of 106 subjects (61 women, 45 men) without LBP. The group with LBP consisted of 33 subjects (15 women, 18 men). There was no significant difference in any of the CERAD battery of mental status testing between those with LBP and without LBP (Table 2). The fresh brain weights were determined at the time of autopsy. The mean brain weight was 1216.3 ± 126.9 g for non-LBP subjects and 1246.4 ± 1236 g for LBP subjects, which were not statistically different. Subjects had been followed for 1 to 15 years with an average of 7.00 ± 3.88 annual follow-up visits. The mean time between the last examination and death was 8.60 ± 6.83 months for those without LBP compared to 8.94 ± 6.12 months for those with LBP. The mean age at autopsy for those without LBP was 83.70 ± 7.87 years compared to 85.58 ± 7.91 years for those with LBP (overall mean age = 83.45 ± 7.85). The mean years of education were 16.07 ± 2.40 for non-LBP subjects, whereas those with LBP had an average of 16.06 ± 2.45 years. All but two of the cases were Caucasian. The most frequent causes of death were myocardial infarctions, pneumonia, congestive heart failure, chronic obstructive pulmonary disease, and cancer (none had cerebral metastases).

TABLE 2.

Mean (SD) CERAD Battery Cognitive Performance by Absence of LBP or Presence of LBP

Measure	No LBP	LBP
Mini-Mental State Exam	28.13 (2.0)	28.03 (1.5)
Blessed IMCT errors	2.96 (3.9)	2.36 (2.6)
Category Fluency (animals)	15.48 (5.2)	16.34 (5.8)
15-item Boston Naming Test	14.24 (1.0)	14.10 (1.4)
Trailmaking A (sec)	54.84 (34.9)	54.37 (22.7)
Trailmaking B (sec)	127.14 (55.6)	132.09 (55.8)
Word-list Learning (30 max)	19.87 (4.5)	19.54 (4.3)
Word-list Delayed Recall (10 max)	6.18 (2.2)	6.26 (2.0)

CERAD = Consortium to Establish a Registry in Alzheimer’s Disease; LBP = Lewy body pathology.

No significant differences were found in any of the measures.

Because of the cumbersome nature of showing the difference between LB and LN in each case, they are presented together as LBP in Table 3.

TABLE 3.

Location of Lewy Body Pathology in Aged Normal Control Subjects

Sex	Age	Braak	CERAD	PMI (h)	Med	Pons	MB	Amy	OB	TP	EC	nbM	AC	Hippo	Fro	IT
F	78	I	no	2.00	1+
M	83	IV	probable	19.00	1+
M	77	I	possible	3.50	1+
M	87	II	no	2.00	1+
M	84	I	no	2.83	1+		1+		2+
F	91	I	possible	3.92	2+	1+
F	69	I	no	2.08	3+	3+	3+
F	101	V	probable	1.35	3+	3+	2+
M	89	III	no	1.58				1+
F	81	III	possible	3.00				1+
F	89	II	possible	2.00				1+
F	87	V	definite	2.33				1+	1+
M	86	III	probable	1.68			1+	1+
F	76	I	no	10.00	1+	1+		1+
M	90	III	no	2.50	1+			1+	1+
F	86	IV	probable	1.83			1+	2+	2+
M	79	I	probable	0.75	3+	3+	1+	1+
M	91	III	probable	3.52	1+	1+	2+	2+
F	65	I	no	3.12	2+	2+	1+	1+
M	80	II	no	2.25	3+	3+	1+	1+
M	80	IV	no	3.00	2+	1+	1+	1+
M	92	II	no	*	1+	1+	2+	1+	2+
M	91	II	probable	6.75	2+	1+		1+	2+
F	100	I	no	2.28	3+	2+	1+	1+	1+
F	91	II	definite	15.17	2+	2+	1+	1+	3+	1+
M	88	I	no	5.08	2+	1+	1+	1+		1+		2+
F	91	II	probable	1.25	3+	3+	3+	1+	2+		1+	2+
M	83	II	no	6.58	2+	2+	1+	2+	1+			1+
M	95	I	no	1.77	3+	2+	1+	2+		2+	2+	3+		1+
M	88	I	no	9.00	3+	3+	2+	2+	1+	1+	1+		1+	1+
F	95	0	definite	11.25	3+	3+	1+	2+	1+	1+	1+		1+	1+	1+	1+
F	76	I	no	4.42	3+	2+	1+	3+	2+	1+		3+
M	87	II	possible	*					2+

Braak = Staging for neurofibrillary pathology; CERAD = Consortium to Establish a Registry for Alzheimer’s Disease neuropathology criteria; PMI = postmortem interval; Med = medulla; MB = midbrain; Amy = amygdala; OB = olfactory bulb; TP = temporal pole; EC = entorhinal cortex; nbM = nucleus basalis of Meynert; AC = anterior cingulate; Hippo = hippocampus; Fro = frontal; IT = inferior temporal.

The lack of a rating means there was no LBP in that anatomical structure.

^*Time of death unknown

Of the 139 patients studied 33 (23.7%) had LBP in some region of the neocortex, allocortex, brainstem, or olfactory bulb and tract. The most common sites involved were the medulla (26%), amygdala (24%), pons (20%), and midbrain (20%).

In the 8 subjects with LBP limited to the brainstem, 4 cases were limited to the medulla and all of these were scored mild. Two had LBP in medulla, pons, and midbrain. One had medullary and pontine LBP only and one had midbrain LBP alone. Another had medulla, midbrain, and olfactory bulb LBP. Although neuron quantitation was not carried out, the substantia nigra did not show meaningful neuron loss in those graded 1+ and 2+, while several of those graded 3+ showed mild depigmentation and pigmentary incontinence. One participant had LBP limited to the olfactory bulb only.

In the medulla, the earliest changes were limited to the dorsal glossopharyngeal/vagus motor nuclear complex. In more advanced cases, the LBP spread to the gigantocellularis reticular nucleus and other reticular nuclei. In the pons, early changes involved the locus ceruleus and, with progression, involved the locus ceruleus/subceruleus complex and other neurons of the pontine tegmentum. In the substantia nigra, the early changes were in the pigmented neurons of the pars compacta and with more severe involvement, it was more diffusely scattered in pigmented neurons. In general, there was often mild asymmetry in the LBP between left and right in the brainstem.

The amygdala showed LBP in 24 out of 33 cases but most had only mild involvement. Only three cases had LBP in the amygdala only. All but 4 subjects had 2 or more regions of the brainstem involved and had LBP in the amygdala. All cases that contained LBP in the neocortex, allocortex or nbM had amygdala involvement.

All 6 subjects with neocortical involvement had abundant LBP in medulla, pons, midbrain, and amygdala. Of the 6 subjects with neocortical involvement, 2 met the initial neuropathological criteria for the diagnosis of DLB (26) or recent refinement of the criteria (27) with LBP in 2 or more cortical areas plus amygdala and entorhinal cortex.

The mean Braak stage for neurofibrillary pathology was 1.80 ± 1.26 in subjects without LBP and 2.09 ± 1.33 with LBP. Frequencies of Braak staging did not differ by groups, (k² = 3.76, p = 0.58). No significant differences were found in NFT and NP counts in all 4 lobes of the neocortex or in hippocampal CA1 and subiculum, entorhinal cortex, and amygdala between those with and without LBP. The 2 groups did not differ in age at autopsy. Lewy body pathology appeared to be more common in the eighth decade (Fig. 2).

DISCUSSION

To our knowledge, this represents the largest study of LBP in longitudinally evaluated normal control subjects to date. The major finding in this study is that 23% of elderly subjects with normal cognition and no clinical evidence of PD, DLB, or other CNS disorders contained alpha-synuclein positive LBP in one or more regions of the central nervous system. While others have described LBP in normal elderly subjects, none have used longitudinally evaluated elderly normal controls and focused only on LBP in normal aging. In the pre-immunocytochemistry era, Lipkin (4) reported LB in 4.9% of control subjects and Forno (5) found them in 4.7% of routine autopsies. Gibb and Lees (6) found LBP increased from 3.8% to 12.8% between the sixth and ninth decade in 273 patients dying of diseases other than PD. Many of them had other neurological diseases. Since the beginning of alpha-synuclein immunohistochemistry use, Tsuboi et al (7) found LBP in 9.2% of normal elderly subjects, Parkkinen et al found LBP in 11% in one study (8) and 13% in another (12), and Wakisaka et al (13) found them in 15% (13). Jellinger (14) showed that 31% of aged controls had alpha-synuclein positive LB in the basal nucleus and in medulla, pons, or midbrain. In a series of 904 autopsied subjects, Parkkinen et al. (12) found LBP in the brainstem or basal forebrain nucleus of 106 subjects (13%), but only 32 (30%) did not have any neurologic impairment and were considered normal. In a community-based Japanese population (the Hisayama Study), LBP was present in 15% (11 / 73) of nondemented subjects without PD (13). Sengoku (28) found LBP in 31.9% of consecutive autopsied patients from a general geriatric hospital. In the only two studies of small numbers of longitudinally followed cognitively normal elderly individuals, Knopman et al (11) described LBP in 13% (5/39) and Mikolaenko et al (15) found LBP in 8.3% (3/36).

In a study of 30 incidental LBP cases from 413 autopsies, Del Tredici et al (9) found that LBP begins in the dorsal glossopharyngeal-vagal complex, locus ceruleus/subceruleus complex, caudal raphe nuclei, gigantocellularis reticular nucleus, olfactory bulb, tract and anterior olfactory nucleus but not in the substantia nigra. These subjects were not followed longitudinally but did not have PD, AD, or other neurological or psychiatric diseases. Of our 7 cases limited to the brainstem, four involved medullary nuclei alone supporting the findings of Del Tredici et al (9). Nine other cases had brainstem and amygdala only involvement.

It is difficult to compare our results with the other studies because of case selection and study design. None of the other studies used longitudinally followed systematically evaluated controls except as noted above (11, 15). Some studies calculated the percentage of alpha-synuclein positive normals in all alpha-synuclein positive cases rather than the whole autopsy series. Some were designed to compare dementia with nondementia controls without regard for movement disorders. Other studies used only specific brain areas and not a broad spectrum of brain regions. However, the percentage of normal elderly controls with LBP in our study (23%) is higher than most other studies except for the report by Jellinger (14) who found LBP in 31% of controls and Sengoku et al (28) who found it in 31.9%.

The reason for the high prevalence of LBP in our study is not clear. Several authors have suggested that aging and severe AD pathology have a major effect in the evolution of LBP (8, 10, 13). Two of these studies showed an age-related increase in LBP between the eighth and tenth decade (8, 13). Although our subjects with LBP were old (mean age > 80 yrs), no significant difference in mean age of those with and without LBP was present. However, in the LBP group, LBP was slightly increased in the eighth and ninth decades compared to the seventh decade.

Our study did not show an association with AD pathology as indicated by statistically insignificant mean Braak scores of 2.1 in subjects with LBP compared to 1.80 in controls without LBP. In addition, NP counts between controls with and without LBP were not significantly different. Thus, it would appear from our study that AD pathology does not have a strong effect on the evolution of LBP as suggested by others (13, 14), and synergism between LBP and tau or Aβ lesions is not present as indicated by Mikolaenko et al (15). In addition, the presence of LBP does not indicate a transition to the mixed form of AD/DLB in our study. Those with neocortical and allocortical LBP had Braak stage scores (for neurofibrillary pathology) from 0 to II. None of our LBP subjects showed clinical manifestations of DLB or a decline in cognitive function, even at a late stage of life, suggesting that they were not developing DLB yet. Synergy between AD and DLB pathology may be restricted to subjects experiencing progressive clinical symptoms and cognitive decline rather than the asymptomatic, cognitively normal subjects presented here.

Lewy bodies were initially described in the amygdala in AD using α-synuclein immunohistochemistry by Lippa et al (29) and have been reported to occur in this site in 43% to 60% of AD patients (30, 31). The amygdala was a common site of LBP in our controls, occurring in 24 out of 33 cases with LBP. Interestingly, almost two-thirds of those with brainstem LBP also had amygdala involvement. Although LBP was prominent in the amygdala, none had cognitive impairment, which would suggest that LBP in amygdala is not always related to cognitive impairment as suggested by Parkkinen et al (12).

Jellinger (10) suggested that the burden of alpha-synuclein pathology was significantly greater in demented patients versus nondemented subjects Most neuropathologists would agree with that conclusion. However, several of our cases had neocortical and allocortical LBP and met the original neuropathological criteria for DLB (26) but were not demented. Six of our cognitively normal control subjects had developed Parkinson signs or early neuropsychiatric manifestations of DLB and were excluded from this study. Each of these met the latest neuropathologic criteria for DLB (32) and probably had preclinical or early DLB. The details of these subjects are presented in a separate publication (33).

Braak and Del Tredici (34) developed a staging procedure for LBP associated with PD that begins in the medulla and olfactory structures and continues in a topographical predictable sequence in 6 stages progressively involving olfactory, autonomic, limbic, and somatosensory systems. One of our cases involved the olfactory bulb alone. Braak and Del Tredici (34) indicate that the earliest LBP may develop more or less simultaneously in the medulla and olfactory bulb (Stage 1). Perhaps our single case represents one instance where the change occurs first in the olfactory bulb and would have been followed by LBP in the medulla if the patient had lived longer. In the cases with brainstem only LBP, 4 had medulla involvement alone and the pathological changes were mild and predominantly limited to the dorsal glossopharyngeal/vagal nuclear complex and the nucleus gigantocellularis. One case involved medulla and the locus ceruleus, the former more extensively than the latter. These cases conformed to Braak stages 1 and 2 (35). Three cases had medulla, locus ceruleus/subceruleus and substantia nigra LBP and conform to Braak stage 3 except that the amygdala was not involved. Eight cases had LBP in the amygdala, medulla, locus ceruleus, and substantia nigra fulfilling complete Braak stage 3 criteria. It would be expected that some of these cases may have clinical findings such as motor symptoms, autonomic, or cognitive decline but they were not detected. Perhaps they had subtle undetected clinical findings. Another explanation is that possibly the burden of LBP and neuron loss was not large enough to cause symptoms. Regardless of whether they were in the presymptomatic or undetected symptomatic phase of PD, it seems clear that these individuals had a relatively uniform pathological process advancing in the predictable manner described by Braak et al (35). However, 2 subjects had amygdala and substantia nigra involvement and one had amygdala and medulla only and do not fit the criteria well. It would be expected that not every case would fit precisely into the Braak staging scheme and variation may occur in a small percentage of cases.

All 6 cases with allocortical or neocortical involvement had LBP in the medulla, pons, midbrain, and amygdala. All had involvement of the temporal neocortex and two had involvement of the anterior cingulate gyrus. Because of mild LBP scores they would not have met the Braak stage 5 criteria for PD (35). Two would have met the initial criteria for pathological diagnosis of DLB (26) or the recently amended criteria (27). None of these cases showed significant cognitive decline, neuropsychiatric, or Parkinsonism signs by history or neurological examination.

A limitation of our study is that it is a sample of convenience. All subjects were white except two, of advanced age, and most were well educated. Many had professional, educational, or business occupations and most were of middle class socio-economic status. Thus, our study is not applicable to the general population. However, it is a relatively uniform sample whose subjects showed excellent compliance to our longitudinal evaluation process. We purposely excluded those with other neurological disease, especially PD and cognitive decline or major psychiatric features to gain insight into alterations in the aging brain.

Our study shows that the presence of LBP in elderly control subjects is not always accompanied by neurologic clinical manifestations. There is no clear explanation for this, but it is possible that the burden of LBP was not large enough to cause symptoms in some cases or that our cases had strong “brain reserve” to withstand the effects of LBP. Perhaps the high level of education in our subjects (mean = 16 years) plays a role in this regard. Thus, in evaluating elderly, well educated subjects with no cognitive decline at autopsy, one should consider that the threshold of LBP causing clinical symptoms may be higher than in the average population.

It is also possible that some of these subjects developed clinical or cognitive changes between the last evaluation and autopsy, although this seems unlikely for the majority because of the short interval between the last exam and death of the subjects with LBP (8.94 ± 6.12 months).

A more likely explanation for our findings is based on the recognition that diseases such as PD, AD, and DLB have pathological manifestations long before the classic clinical findings appear. In some PD cases, autonomic, olfactory, and sleep disorders (36-38) antedate motor symptoms and signs by many years. Del Tredici et al (9) and others (6, 39) suggest that LBP in the brainstem represents the precursor of PD. It is unlikely that our findings represent nonspecific alterations found in the aging brain. DelleDonne et al (16) showed that subjects with incidental Lewy body disease had diminished immunoreactivity for tyrosine hydroxylase and vascular monoamine transporter 2 in the putamen compared to normal controls. This decline in dopaminergic immunoreactivity correlated inversely with substantia nigra neuron loss and PD stages. Their findings suggest that incidental Lewy body disease most likely represents preclinical PD. It is possible that most, if not all of our subjects, were in the presymptomatic phase of PD, PDD, or DLB and, if they had lived longer, they would have demonstrated clinical neurological impairment. Our study also suggests that a large number of normal elderly are probably at risk for developing these disorders.