Acta Neurol Scand 2000: 101: 5356 Printed in UK.

All rights reserved

Copyright # Munksgaard 2000

ACTA NEUROLOGICA SCANDINAVICA ISSN 0001-6314

Short communication

The electroencephalogram in dementia with Lewy bodies

Barber PA, Varma AR, Lloyd JJ, Haworth B, Snowden JS, Neary D. The electroencephalogram in dementia with Lewy bodies. Acta Neurol Scand 2000: 101: 5356. # Munksgaard 2000. Objectives Dementia with Lewy bodies (DLB) is the second commonest neurodegenerative cause of dementia. While there is consensus on the clinical diagnostic criteria for DLB, the use of EEG to increase the diagnostic sensitivity has not been substantiated. Material and methods We studied the resting EEG ndings in 18 consecutive patients with DLB and compared them with a control group of 20 patients with ``probable'' Alzheimer's disease (AD). We aimed to evaluate the use of EEG in a representative sample of patients with DLB. Results All patients with DLB fullled accepted clinical criteria for DLB. The DLB group had a more severe dementia than the AD group, as measured by the MiniMental State Examination (MMSE) test (DLB mean MMSE 9.4 and AD mean MMSE 17.2) despite a similar duration of overall severity of illness. The EEG was slow in both groups, predominantly in the 47 Hz range. Although there was no statistically signicant difference in the EEG ndings between the DLB and AD groups, there was a correlation between the EEG score and MMSE score (Spearman Rank correlation rs=x0.61, P<0.001). Conclusion These ndings suggest that although patients with DLB have a more aggressive course than AD, EEG abnormalities do not differ in the 2 groups. However, we believe the EEG provides important supporting diagnostic information in DLB.

P. A. Barber, A. R. Varma, J. J. Lloyd, B. Haworth, J. S. Snowden, D. Neary

Cerebral Function Unit, Manchester Royal Inrmary, Oxford Road, Manchester M13 9WL, United Kingdom

Key words: electroencephalogram; dementia; Lewy body Dr J. S. Snowden, Cerebral Function Unit, Manchester Royal Inrmary, Oxford Road, Manchester, M13 9WL, United Kingdom Accepted for publication June 29, 1999

Alzheimer's disease (AD) and Dementia with Lewy bodies (DLB) are the two most common causes of dementia in the elderly (1, 2). A correct diagnosis of DLB has important management implications. Neuroimaging (structural and functional) has not been shown to have any convincing differential diagnostic value between these two disorders (3, 4). The EEG is commonly used as an aid in the diagnosis of DLB and AD, and there is anecdotal evidence that suggests EEG may be helpful in the diagnosis of DLB (2, 3). More information is required regarding the use of EEG as a possible adjunct to already accepted clinical criteria. We, therefore, investigated the use of EEG in a more representative sample of individuals fullling currently accepted clinical criteria for DLB and compared the ndings with

a group of patients with a clinical diagnosis of AD to ascertain whether EEG can be useful in improving the diagnostic sensitivity of DLB.
Materials and methods
Dementia with Lewy bodies

The DLB study group consisted of 18 patients drawn from a larger group of 31 consecutive referrals to the Cerebral Function Unit, Department of Neurology at Manchester Royal Inrmary diagnosed as having DLB on clinical grounds (5), between 1991 and 1996. Nine patients were excluded from the study on the grounds that there was insufcient EEG data. Three patients were excluded because they were considered to have 53

Barber et al. Parkinson's disease with dementia (5). The remaining excluded patient had coexistent evidence of vascular disease and therefore did not fulll criteria. Twelve of the nal 18 patients have been reported in a previous study (4). In total 5 patients have subsequently had autopsy conrmation of their disease. Clinical onset was of cognitive and/or psychiatric breakdown in 16 patients and extrapyramidal features in the remaining 2 patients. Overall a total of 18 patients exhibited extrapyramidal features at the time of assessment. Initial symptoms of cognitive impairment consisted of memory loss, perceptuo-spatial disorientation and/or mental slowing. Psychiatric manifestations usually consisted of agitated behaviour, depression, delusions, illusions and/or hallucinations. Neuropsychological evaluation was carried out within 1 month of the EEG recording. The method of neuropsychological evaluation has been previously documented (6) and also included a quantitated Mini-Mental Score of cognitive function. Seventeen patients had uctuating cognitive impairment but none exhibited episodes of unconsciousness. The pattern of breakdown consisted of subcortical decit in all with/or without cortical decits in the domains of perception, visuospatial and linguistic function.
Alzheimer's disease group

without perceptuospatial and/or linguistic impairment in all patients. The age and sex characteristics are shown in Table 1. There was no signicant difference in relative proportions of males and females in the 2 groups (chi square test P=0.34). The DLB group were signicantly older than the AD group (t-test, P=0.0079).
The electroencephalogram (EEG)

The AD study group consisted of 20 consecutive patient referrals to the Cerebral Function Unit, Department of Neurology at Manchester Royal Inrmary who fullled clinical criteria for ``probable'' AD (7). Clinical onset was cognitive or neuropsychiatric in 17 patients. One patient presented with visual disorientation and another with aphasia. Two patients had mild extrapyramidal features with the remainder having normal neurological examinations. Neuropsychological evaluation was carried out, as previously documented, within 1 month of EEG recording and revealed amnesia, together with/

A 24 channel awake EEG was recorded on a Walter-Graphtek digital signal machine. Technicians were instructed to keep patients awake for greater inter-patient consistency. Hyperventilation and photic stimulation were not administered owing to the clinical condition of the population being investigated. Each EEG was interpreted by one neurologist (P.A.B.) and an EEG technician (B.H.) who were both aware of the possibility of dementia but were blind to the specic diagnosis. The EEG was examined only once in each patient within 1 month of presentation. The EEG ndings were classied as follows (8): grade 1, normal for age; grade 2, mild but denitely excessive theta activity and/or bitemporal spikes; grade 3, moderate excessive theta activity diffusely and/or bursts of bisynchronous, diffuse rhythmic 2to 3-Hz waves; grade 4, one or more of persistent diffuse delta, bisynchronous spikes or sharp waves, and triphasic waves. The EEG was also assessed for lateralizing features and the presence of periodicity. Grades 1 to 4 are mutually exclusive, but could coexist with lateralization and/or periodicity. None of the EEGs were uninterpretable.

Results

The duration of disease prior to presentation was similar in the 2 groups with an overall mean of 1.9 years. However, the DLB patients had lower MMSE (mean of 9.4) than the AD group (mean of 17.2; P=0.02; MannWhitney U-test). Table 2 shows the EEG results from the DLB and AD groups. There was not a signicant difference in the frequency of occurrence of grades between the 2

Table 1. DLB (n=18) Mean Age (SD) Sex (M:F) Duration of illness in years (SD) MMSE score, mean (SD) 67.5 (6.29)* 11:7 1.75 (0.77) 9.4 (6.08)* AD (n=20) 60.7 (7.32) 9:11 2.22 (0.98) 17.2 (6.76) Table 2. EEG grade 1 DLB AD 0 3 2 1 1 3 13 14 4 6 2 Lateralisation 5 10 Periodicity 1 1

* P<0.05. DLB: dementia with Lewy bodies; AD: Alzheimer's disease; SD: standard deviation.

DLB: dementia with Lewy bodies; AD: Alzheimer's disease.

54

The electroencephalogram in dementia with Lewy bodies groups (MannWhitney U; P=0.0724). In addition there was no signicant difference in focal EEG features in the 2 groups although more AD patients (n=10) had focal features than DLB (n=5). One patient in the DLB group showed bursts of diffuse delta activity, while 2 patients exhibited periodicity, one with DLB, the other with AD. There was an inverse correlation between the MMSE and the EEG grade (Spearman Rank correlation, rs=x0.61 P<0.001), i.e. a lower MMSE score was associated with a more severely impaired EEG. masked this trend. These observations are supported statistically as the EEG grade showed an inverse correlation with MMSE score. Although the ndings in this study suggest that the EEG is not diagnostic of DLB and cannot differentiate it with any certainty from AD, it can be used to support the diagnosis of DLB in a patient with a progressive dementing syndrome in combination with one of the following three core features: uctuation, motor parkinsonism, and visual hallucinations. Previous studies have suggested the usefulness of EEG in contributing to the diagnosis of DLB which may be abnormal early when the dementia is mild to moderate (1, 3). These reports consisted of small patient numbers and were primarily concerned with making neuropathological comparisons between DLB and AD. A recent study (11) investigated the standard EEG recording in 14 patients with post-mortem diagnosed DLB and compared them with the records from a group of patients with AD, also conrmed at post-mortem. While there are signicant differences in the designs of this and our study what can be concluded is that patients with DLB may have a more aggressive disease than AD and therefore the EEG exhibits a greater degree of slowing in these patients. However, what we have attempted to highlight is that EEG abnormalities when taken from a prospective sample cannot reliably differentiate the two conditions. The accurate clinical differentiation of DLB and AD is of importance with regard to patient management, long term prognosis and the evaluation of future therapeutic agents. Furthermore, the recent suggestion that patients with DLB may respond to cholinergic medications such as tacrine (9, 10), and therefore potentially to newer agents like donepezil too, emphasises the importance of accurate clinical diagnosis. In addition, patients with DLB are sensitive to neuroleptics and can have life threatening reactions inadvertently. The present clinical criteria for DLB allow the disorder to be distinguished from AD with a relatively high degree of sensitivity and specicity (5). The EEG has been considered by many to be one investigative method that may potentially assist in the differential diagnosis of DLB and AD. It is widely accepted in the literature that the EEG may be of use in DLB at an early stage of the disease to aid in its differentiation from AD. It can be concluded from this study that this assumption should be viewed with caution. In early dementia of DLB or AD an abnormal EEG may be useful to support the diagnosis in the specic clinical context. This study has also shown that the EEG correlates negatively with the MMSE and therefore may be useful in the prognostication of DLB. More work is needed to assess whether more 55

Discussion

The aim of this study was to formally investigate EEG ndings in a large, prospective and representative group of consecutive patients with DLB and compare them with a group of patients with ``probable'' AD. Previous studies have suggested that EEG may be of use when DLB is mild to moderate in severity (2, 3). All 18 patients also fullled McKeith et al.'s (5) criteria for DLB. Five of our patients have been conrmed to suffer from the disorder at autopsy. These newer (5) criteria do not insist on uctuating cognitive impairment as an essential feature for the diagnosis of DLB. However, only 1 of our patients did not exhibit uctuations, thereby improving the clinical diagnostic sensitivity and specicity of DLB (5). The DLB group had a lower mean MMSE than the AD group indicating more severe cognitive impairment. Both groups had similar duration of illness, and therefore are representative of the referral pattern of patients with ``dementia'' to a tertiary centre. Hence, the 2 groups were not matched for MMSE. DLB appears to be a more aggressive condition than AD, progressing rapidly to an end stage of profound dementia and parkinsonism, although in other cases the course often resembles that of AD. The DLB group in this study were also signicantly older than the AD group as is usually the case and this factor should also be taken into consideration when discussing differences in disease severity (7). Nevertheless, despite older age at onset and more severe disease in the DLB group no EEG differences were found in the degree of slowing. This study has most importantly shown that the EEG grade in the DLB and AD groups is entirely independent of disease. Focal features were seen in both groups but were in fact more common in the AD group, although this did not reach statistical signicance. However there appears to be a denite trend towards the EEG showing a greater degree of background slowing in the DLB group. The use of ordinal data in grading the EEG may well have

Barber et al. careful extraction of parameters from the EEG may give more sensitive indicators of disease. Further studies should be encouraged with the use of serial EEG and spectral analysis in an attempt to provide better quantitative and qualitative data in respect to patients with DLB and AD.
References
1. LENNOX G, LOWE J, LANDON M, BYRNE EJ, MAYER RJ, GODWIN-AUSTIN RB. Diffuse Lewy body disease: correlation using anti-ubiquitin immunocytochemistry. J Neurol Neurosurg Psychiatry 1989;52:123647. 2. CRYSTAL HA, DICKSON DW, LIZARDI JE, DAVIES P, WOLFSON LI. Antemortem diagnosis of diffuse Lewy body disease. Neurology 1990;40:15238. 3. DICKSON DW, RYAN D, CRYSTAL H et al. Hippocampal degeneration differentiates diffuse Lewy body disease (DLBD) from Alzheimer's disease: light and electron microscopic immunocytochemistry of CA23 neurites specic to DLBD. Neurology 1992;41:14029. 4. VARMA AR, TALBOT PR, SNOWDEN JS, LLOYD JJ, TESTA HJ, NEARY D. A 99mTc-HMPAO single-photon emission computed tomography study of Lewy body disease. J Neurol 1997;244:3499. 5. MCKEITH IG, GALASKO D, KOSAKA K, PERRY EK, DICKSON DW, HANSEN LA. Consensus guidelines for the clinical and pathologic diagnosis of dementia with Lewy bodies (DLB): Report of the consortium on DLB international workshop. Neurology 1996;47:111324. 6. NEARY D, SNOWDEN JS, BOWEN DM et al. Neuropsychological syndromes in presenile dementia due to cerebral atrophy. J Neurol Neurosurg Psychiatry 1986;49:16374. 7. MCKHANN G, DRACHMANN D, FOLSTEIN M, KATZNAN 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 disease. Neurology 1984;34:93945. 8. ROBINSON DJ, MASKEY H, BROWNE WT, FRY R, WILLIAMSON PC, HACKINSKI VC. Electroencephalophy as an aid in the exclusion of Alzheimer's disease. Arch Neurol 1994;51: 2804. 9. LEVY R, EAGER S, GRIFFITHS M et al. Lewy bodies and response to tacrine in Alzheimer's disease. Lancet 1994;343: 176. 10. WILCOCK LI, SCOTT MI. Tacrine for senile dementia of the Alzheimer's or Lewy body type. Lancet 1994;344:544. 11. BRIEL RC, MCKEITH IG, BARKER WA et al. EEG ndings in dementia with Lewy bodies and Alzheimer's disease. J Neurol Neurosurg Psychiatry 1999;66:4013.

56