Search OCA
Get Local!

Diagnosis of dementia: Clinicopathologic correlations

Neurology 1989;39:76-79

Francois Boller, MD, PhD; Oscar L. Lopez, MD; and John Moossy, MD

--Based on 54 demented patients consecutively autopsied at the University of Pittsburgh, we studied the accuracy of clinicians in predicting the pathologic diagnosis. Thirty-nine patients (72.2%) had Alzheimer's disease, while 15 (27.7%) had other CNS diseases (four multi-infarct dementia; three Creutzfeldt-Jakob disease; two thalamic and subcortical gliosis; three Parkinson's disease; one progressive supranuclear palsy; one Huntington's disease; and one unclassified). Two neurologists independently reviewed the clinical records of each patient without knowledge of the patient's identity or clinical or pathologic diagnoses; each clinician reached a clinical diagnosis based on criteria derived from those of the NINCDS/ADRDA. In 34 (63 %) cases both clinicians were correct, in nine (17%) one was correct, and in 11 (20%) neither was correct. These results show that in patients with a clinical diagnosis of dementia, the etiology cannot be accurately predicted during life.

Several recent papers and reports have addressed the problem of improving the clinician's ability to diagnose dementia. Notable among those reports are the diagnostic criteria for dementia of the American Psychiatric Association, known as DSM III,1 as well as the clinical and neuropathologic criteria for the diagnosis of Alzheimer's disease (AD).2,3 Other researchers have published guidelines for the differentiation of various types of dementia4 and for antemortem predictions about the neuropathologic findings of demented patients.5

Most studies on the accuracy of clinical diagnosis in patients with dementia, especially AD, have used clinicopathologic correlation,6-15 and have found a percentage of accuracy ranging from 43% to 87%. Two recent reports, however,16,17 have claimed an accuracy of 100%. These two reports are based on relatively small series and have consisted of very highly selected patient samples. In our own recent experience, several cases of dementia have yielded unexpected neuropathologic findings,18 and we hypothesized that, in larger series, there would be a significant number of discrepancies between clinical diagnoses and autopsy findings. The present paper reviews the neuropathologic diagnosis of 54 demented patients who were autopsied consecutively at the University of Pittsburgh over a 7-year period, and reports the ability of clinicians to predict autopsy findings.

Material and methods.
We independently reviewed the pathologic data and clinical records of 54 consecutive patients who had had an autopsy at the University of Pittsburgh (Presbyterian University Hospital [PUH] and the Pittsburgh (University Drive) Veterans Administration Medical Center [VAMC]), between 1980 and 1987.

The 54 cases included all those where dementia was diagnosed clinically but for which an obvious etiology, such as neoplasm, trauma, major vascular lesions, or clinically evident infection had not been found. The brains, evaluated by the Division of Neuropathology of the University of Pittsburgh, were obtained from patients cared for in different settings at their time of death.

On the basis of the amount of information available in each case, we divided the patients into three groups. Group 1 included 12 subjects who had been followed for a minimum of 1 year by the Alzheimer Disease Research Center (ADRC) of the University of Pittsburgh. ADRC evaluations include several visits and neurologic and neuropsychological testing as well as repeated laboratory tests, EEG, and CT.19,20

Group 2 included 28 patients who had been seen in the Neurology Service of PUH, of the VAMC, or in geriatric or psychiatric facilities of the University of Pittsburgh or at Western Psychiatric Institute and Clinic. All patients were personally evaluated by a neurologist and received a work-up to elucidate the etiology of their dementia.

Group 3 included 14 patients seen in other institutions; in most cases, they had also been seen by a neurologist and had had laboratory studies that included CT of the head. In three of the 14 cases, however, the information could be gathered only from the clinical summary found in the autopsy records.

Many of these subjects were referred for autopsy to the ADRC because of a public education campaign that encourages families to seek an autopsy for their relatives with dementia.

Pathologic data. All brains were removed by a neuropathologist as the first procedure of the autopsy at postmortem intervals of between 4 and 12 hours. The unfixed brain was weighed and the brainstem and cerebellum were separated by intercollicular section. The cerebral hemispheres were sectioned at 1-cm intervals and placed on a glass surface cooled by ice to prevent adhesion of the tissue to the cutting surface. The brainstem and cerebellum were sectioned in the transverse plane at 6-mm intervals. Brain sections were fixed in 10% buffered formalin. Selected tissue blocks for light microscopy were obtained from sections corresponding as exactly as possible to a set of predetermined areas used for processing brains for the ADRC protocol; additional details of the neuropathologic protocol have been previously published.18,21 Following standard tissue processing and paraffin embedding, 8-um-thick sections stained with hematoxylin and eosin and with the Bielschowsky ammoniacal silver nitrate impregnation were evaluted. Additional stains were used when indicated by the survey stains, including the Bielschowsky silver technique as previously reported.21

Clinical data. The medical history, as well as the results of examinations and laboratory tests, were obtained from the medical records libraries of the institutions where the patient had been followed and had died. We supplemented these data, when appropriate, with a personal or telephone interview with the relatives.

One neurologist (O.L.L.) recorded the information to be evaluated on two forms. The first form included sex, age, handedness, age at onset, age at death, course and duration of the disease, education, family history, EEG, CT, NMR, medical history, and physical examination as well as examination of blood and CSF for factors that could affect memory and other cognitive functions. The form also listed the results of neuropsychological assessment, and the characteristics and course of psychiatric and neurologic symptoms. The form provided details on the presence, nature, and course of cognitive deficits and neurologic signs. The second form was a 26-item checklist derived from the NINCDS-ADRDA Work Group Criteria for probable Alzheimer's disease.2 The forms did not include the patient's identity, the institution where they had been evaluated, the clinical diagnosis, or the pathologic findings.

Each form was reviewed independently by two other neurologists (F.B. and J.M.), who were asked to provide a clinical diagnosis. In cases of probable or possible AD, the two neurologists followed the diagnostic criteria of the NINCDS/ ADRDA work group.2

The results were tabulated on a summary sheet filled out after the two neurologists had provided their diagnosis on each case. The sheet included the diagnosis reached by the two neurologists and the diagnosis resulting from the autopsy.

The subjects included 26 women and 28 men who ranged in age from 30 to 91 years (mean, 72.2; SD, 10.7).

Autopsy findings. Table 1 shows that 39 (72.2%) of the 54 cases fulfilled histologic criteria for AD, with or without other histopathologic findings. The remaining 15 cases (27.7%) showed changes corresponding to other neurodegenerative disorders, cerebrovascular disease, or Creutzfeldt-Jakob disease (CJD). Seven cases met the histopathologic criteria for multi-infarct de=ADmentia (MID). Five cases (9.2%) showed changes associated with Parkinson's disease (PD).

Twenty-two of the 39 AD patients (56%) were age 65 or greater at the time of the onset of the disease. Seven of the 15 patients in the group with other diseases (47%) were age 65 or older at the time of disease onset.

Clinical diagnosis. There was a general adherence to the criteria specified by McKhann et al.2 However, the two clinicians in this study considered the diagnosis of probable AD when the probability of AD was strong even if a patient had another disease potentially associated with dementia that might or might not have made some contribution to the patient's clinical state (table 2).

Accuracy of the clinical diagnosis (table 3). Group 1 (N =3D 12). There were six men and six women. Ten cases (83.3%) met the histologic criteria for AD. In nine cases (75.0%), the diagnosis of both clinicians agreed with the pathologic findings; in the other case (8.3%), one clinical diagnosis agreed with the histologic findings. The remaining two cases (16.6%) had histopathologic diagnoses of CJD and progressive supranuclear palsy (PSP), respectively. Both cases were incorrectly diagnosed by both clinicians.

Group 2 (N =3D 28). There were 11 women and 17 men. Eighteen cases (64.2%= ) had the histopathologic features for AD with or without additional findings. Sixteen of these cases (57.1%) were correctly diagnosed by both clinicians, one case by one of them, and both incorrectly diagnosed one case. The remaining ten cases (35.7%) included two with CJD; two with subcortical gliosis (SG); two with PD, one of which was associated with MID; one case of Huntington's disease (HD); two cases with MID; and one unclassifed. Only one, the HD case (3.5%), was correctly diagnosed by both observers, and four cases (14.2%), two MID and two PD, one associated with MID, were correctly diagnosed by one clinician.

Group 3 (N =3D 14). In this group there were nine women and five men. Eleven cases (78.5%) met the histopathologic criteria for AD with or without additional findings. Eight of these cases (57.1%) were correctly diagnosed by both clinicians, two cases by one of them, while both were incorrect in one case. Of the remaining three cases (21.4%), only one was correctly diagnosed (7.1%) by one clinician. Both missed the two other cases of MID.

There was no statistically significant difference in diagnostic agreement across patient groups in which the amount of clinical information was different (X2 =3D 1.19; p > 0.05).

Our results indicate that in a population of patients with dementias of varied etiology, the diagnosis could be correctly inferred by at least one of two clinicians in approximately 80% of cases. For one observer, the sensitivity of clinical diagnosis for AD was 85% and the specificity was 13%, and for the other, it was 95% and 33% respectively.

In the cases with a discrepancy between the clinical diagnosis and the neuropathologic findings, the great majority of patients had atypical clinical courses and findings. The three cases with autopsy findings of CJD had a much longer course than is usually seen with that condition and failed to show the usual EEG abnormalities. The patient with autopsy findings of PSP did not show the disorder in the extraocular movements usually associated with that condition. An atypical course was also present for two AD cases and two MID cases that did not have any feature suggestive of vascular disease. In one MID case, the CT did not show any focal lesions, while in the other it was not available. With regard to the two patients with SG, the pathologic diagnosis is so unusual and so infrequently recorded that clear clinical correlates are not evident.18 The third category of possible error is the patient listed as unclassified, for whom no specific neuropathologic diagnosis could be reached.22

The small number of neuropathologic diagnoses of Parkinson's disease reflects that, for the purpose of this series, the diagnosis of PD was made only when there were both a clear-cut clinical history and the neuropathologic findings characteristic of the disease, such as Lewy bodies, neuronal loss, globose neurofibrillary tangles, astrocytosis, and extraneuronal melanin pigment in substantia nigra and locus ceruleus.

Are these results derived from a sample of 54 patients representative of disease patterns in the community? Generally, the diagnosis of patients reported from major medical centers tend to be biased since the more complicated cases are referred there. In this study, however, this bias may be less important. Due to the major public education campaign about dementia and AD sponsored by the ADRC, there is a widespread awareness in Pittsburgh and in the surrounding regions of Western Pennsylvania of the value of an autopsy for a definitive diagnosis. Therefore, the great majority of cases were referred to us because the family wanted to know the precise etiology of a case of dementia.

The significant improvement in the clinical diagnosis of AD is a recent phenomenon. Due to the publicity and the advances in communication of scientific investigations, most physicians are more likely to consider AD as the main cause of dementia. The current risk of overdiagnosing AD reminds one of what occurred during the 1960s with the diagnosis of "atherosclerotic dementia."6 The high sensitivity and low specificity for AD shown in our study may reflect that possibility.

Because of the varying criteria for "other dementias" in many publications, we chose to analyze the accuracy of clinical diagnosis in terms of the diagnosis of AD alone or AD plus other neuropathologic findings. Several retrospective studies have attempted to point out reliable clinical and pathologic features for diagnosing the dementias, especially AD. The study of Tomlinson et al6 is not included in table 4 because there was no attempt to validate the clinical diagnosis with pathologic findings. The reports surveyed vary considerably in size and methodology. Sample size, for example, ranges from 26 subjects9 to 776 subjects.7 Some studies base the diagnosis on limited clinical information,7'9'14'15 others use widely accepted diagnostic criteria such as those specified in DSM III,13 and one group uses a standardized clinical assessment of patients enrolled in a longitudinal study.12 The reported accuracy of the clinical diagnosis of AD ranges from 43%7 to 87%.15

Recent prospective studies that adhere to strict clinical criteria,10'11'17 those in DSM III8 or those proposed by McKhann et al,16 indicate improved accuracy of clinical diagnosis of the most common causes of dementia, especially AD. In sample sizes ranging from 11 subjects16 to 58 subjects,l0 the accuracy of clinical diagnosis is reported as ranging from 71%10 to 100%16'17' Only two series, both based on small samples, report a 100% accuracy. We consider it unlikely that such accuracy could be confirmed in large series because of some inevitable imprecision in clinical diagnoses and the variability of clinical pictures. Furthermore, although researchers generally agree on the application of uniform criteria in clinical diagnosis of dementia, opinions still differ about specific diagnostic criteria, as well as about the pathologic characterization of dementia. Except for those small series, the results summarized in table 4(7-15) is are remarkably consistent with ours.

In table 3, although there was no statistical difference (p > 0.05) in diagnostic agreement across patient groups, there is a trend toward a lower percentage of diagnostic errors for the patients who had been followed most intensely (16% in group 1 compared with 21% in groups 2 and 3). The difference is not great, and it is, in fact, surprising to find out that in the patients about whom relatively little was known (group 3) the percentage of diagnostic error was the same as among patients seen by neurologists and for whom much more data were available (group 2). These paradoxical findings probably indicate that both clinicians learned to extract essential diagnostic criteria2 in spite of the variations in the amount of information available for consideration. It may well be that clinical, radiographic, and laboratory assessment of patients with dementia is burdened with information that is excessive and unessential for purely diagnostic purposes.


We thank Dr. A. Julio Martinez and Dr. Gutti Rao from the Division of Neuropathology for autopsy data. Mrs. Margaret Forbes, Ms. Annette Grechen, and Mrs. Paula Gent helped in the preparation of the manuscript.


1. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. Organic Dementia Disorders, 3rd ed. Washington DC, APA, 1983:101-161.

2. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan E. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA work group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Dis=ADease. Neurology 1984;34:939-944.

3. Khachaturian Z. Diagnosis of Alzheimer's disease. Arch Neurol 1985;42:1097-1105.

4. Cummings J, Benson F. Dementia: a clinical approach, 1st ed. Boston: Butterworths, 1983.

5. Rosen WG, Terry R, Fuld P, Katzman R, Peck A. Pathological verification of ischemic score in differentiation of dementias. Ann Neurol 1980;7:486-488.

6. Tomlinson BE, Blessed G, Roth M. Observations on the brains of demented old people. J Neurol Sci 1970;11.205-242.

7. Todorov A, Go R, Constantinidis J, Elston R. Specificity of the clinical diagnosis of dementia. J Neurol Sci 1975;26:81-98.

8. Sulkava R, Haltia M, Paetau A, Wikstrom J, Palo J. Accuracy of clinical diagnosis in primary degenerative dementia: correlation with neuropathological findings. J Neurol Neurosurg Psychiatry 1983;46:9-13.

9. Perl D, Pendlebury W, Bird E. Detailed neuropathologic evaluation o f banked brain specimens submitted with clinical diagnosis of Alzheimer's disease. In: Wirtman R, Corkin S, Growdon J, eds. Alzheimer's disease: advances in basic research and therapies. Proceedings of the Fourth Meeting of International Study Group on the Treatment of Memory Disorders Associated with Aging. Zurich, January 1984. Cambridge, MA: CBSM, 1984:463. Molsa PK, Paljarvi L, Rinne JO, Rinne UK, Sako E. Validity of clinical diagnosis in dementia: a prospective clinicopathological study. J Neurol Neurosurg Psychiatry 1985;48:1085-1090.

11. Neary D, Snowden JS, Bowen D, et al. Neuropsychological syn=ADdromes in presenile dementia due to cerebral atrophy. J Neurol Neurosurg Psychiatry 1986;49:163-174.

12. Wade J, Mirsen T, Hachinski V, Fismm~ M, Lau C, Merskey H. The clinical diagnosis of Alzheimer disease. Arch Neurol 1987;44:24-29.

13. Alafuzoff I, Igbal K, Friden H, Adolfsson R, Winblad B. Histopathological criteria for progressive dementia disorders: clinicalpathological correlation and classification by multivariate data analysis. Acta Neuropathol (Berl) 1987,74:209-225.

14. Kokmen E, Offord K, Okazaki H. A clinical and autopsy study of dementia in Olmsted County, Minnesota, 1980-1981. Neurology 1987;37:426-430.

15. Joachim CL, Morris JH, Selkoe D. Clinically diagnosed Alzheimer's disease: autopsy neuropathological results in 150 cases. Ann Neurol 1988;24:50-56.

16. Martin EM, Wilson RS, Penn RD, Fox JH, Clasen RA, Savoy SM. Cortical biopsy results in Alzheimer's disease: correlation with cognitive deficits. Neurology 1987;37:1201-1204.

17. Morris JC, Berg L, Fulling K, Torack RM, McKeel DW. Validation of clinical diagnostic criteria in senile dementia of the Alzheimer type. Ann Neurol 1987;22:122.

18. Moossy J, Martinaz J, Hanin I, Rao G, Yonas H, Boiler F. Thalamic and subcortical gliosis with dementia. Arch Neurol 1987;44:510-513.

19. Huff J, Becker J, Belle S, Nebes R, Holland A, Boller F. Cognitive deficits and clinical diagnosis of Alzheimer's disease. Neurology 1987;37:1119-1124.

20. Huff J, Boiler F, Lucchelli F, Querriera R, Beyer J, Belle S. The neurological examination in patients with probable Alzheimer's disease. Arch Neurol 1987;44:929-932.

21. Moossy J, Zubenko G, Martinez AJ, Rao G. Bilateral symmetry of morphologic lesions in Alzheimer's disease. Arch Neurol 1988;45:251-254.

22. Heilig CW, Knopman DS, Mastri AR, Frey W II. Dementia without Alzheimer pathology. Neurology 1985;35:762-765.

From the Departments of Neurology (Drs. Boller, Lopez, and Moossy), Psychiatry (Dr. Boller), Pittsburgh (University Drive) Veterans Administration Medical Center (Dr. Boller), Department of Pathology (Division of Neuropathology) (Dr. Moossy), and the Pittsburgh Alzheimer Disease Research Center (Drs. Boller, Lopez, and Moossy), University of Pittsburgh Medical School, Pittsburgh, PA.

Supported in part by NIH Grants nos. AG05133 and AG03705, NIMH Grant no. MH30915, by funds from the Veterans Admin., and by the Pathology Education and Research Foundation (PERF) of the Department of Pathology, University of Pittsburgh.

Presented in part at the fortieth annual meeting of the American Academy of Neurology, Cincinnati. OH, April 1988.

Received April 7, 1988. Accepted for publication in final form July 20, 1988.

Address correspondence and reprint requests to Dr. Boller, Department of Neurology, 322 Scaife Hall, University of Pittsburgh Medical School, Pittsburgh, PA 15261.

January 1989 NEUROLOGY 39 79

Home | News | Organics | GE Food | Health | Environment | Food Safety | Fair Trade | Peace | Farm Issues | Politics
Forum | Español | Campaigns | Buying Guide | Press | Search | Volunteer | Donate | About Us | Contact Us | Email This Page

Organic Consumers Association - 6771 South Silver Hill Drive, Finland MN 55603
E-mail: Staff · Activist or Media Inquiries: 218-226-4164 · Fax: 218-353-7652
Please support our work. Send a tax-deductible donation to the OCA

Fair Use Notice: The material on this site is provided for educational and informational purposes. It may contain copyrighted material the use of which has not always been specifically authorized by the copyright owner. It is being made available in an effort to advance the understanding of scientific, environmental, economic, social justice and human rights issues etc. It is believed that this constitutes a 'fair use' of any such copyrighted material as provided for in section 107 of the US Copyright Law. In accordance with Title 17 U.S.C. Section 107, the material on this site is distributed without profit to those who have an interest in using the included information for research and educational purposes. If you wish to use copyrighted material from this site for purposes of your own that go beyond 'fair use', you must obtain permission from the copyright owner. The information on this site does not constitute legal or technical advice.
Please Support Our Sponsors!

Organic Valley


Dr. Bronner's Magic Soaps

Dr. Bronner's
Magic Soaps

Botani Organic


Aloha Bay

Aloha Bay

Eden Organics

Eden Foods

Frey Vineyards


Intelligent Nutrients