Electroencephalography and clinical Neurophysiology, 84 (1992) 325-331
325
© 1992 Elsevier Scientific Publishers Ireland, Ltd. 0168-5597/92/$05.00
EVOPOT 91682
Serial visual evoked potential recordings in geriatric psychiatry Eileen P. Sloan and George W. Fenton Department of Psychiatry, Ninewells Hospital and Medical School, Dundee DD1 9SY, Scotland (U.K.)
(Accepted for publication: 6 March 1992)
Serial visual evoked potentials to flash and pattern reversal stimuli were recorded in elderly patients with senile dementia of the Alzheimer type (SDAT), multi-infarct dementia (MID) and functional psychiatric illness, and in a group of elderly control subjects. Recordings were made at 6 monthly intervals over a 2 year period. Latency and amplitude of the main componentswere measured and the flash P2-pattern reversal P100 latency difference value was calculated. In all groups significant changes over time did not occur for any parameters but in the SDAT group the regression coefficient for the latency of the flash P2 component and the flash P2-pattern reversal P100 latency difference was significant, reflecting a trend towards increasing flash P2 latency as time progressed. The flash P2-pattern reversal P100 latency difference was longer in the SDAT and MID groups than in the functional patients, confirming the findings of previous reports. The latency difference in the SDAT group only was significantlygreater than that in the control group. Summary
Key words: Senile dementia; Visual evoked potentials; Serial recordings
Over the past decade there has been considerable research interest in the diagnostic value of changes in the visual evoked potential (VEP) to flash and pattern reversal stimuli in elderly patients with dementia. As Harding et al. (1985) stress it is necessary to record the responses to both types of stimulus since an increased flash VEP latency alone could be due to peripheral rather than central pathology, while the recording of a normal pattern reversal response along with a delayed flash response rules out that possibility. Wright et al. (1984) report differential latency changes for the major positive components of the two responses in patients with dementia; flash P2 wave delay, pattern reversal P100 component normal. Indeed, they claim that the latency difference value is a robust diagnostic measure of dementia though it does not discriminate between senile dementia of the Alzheimer type (SDAT) and multi-infarct dementia (MID) patients (Wright and Furlong 1988). These differential VEP latency changes have been confirmed (Doggett et al. 1981; Damesi et al. 1985; Harding et al. 1985). However, the view that the pattern reversal VEP is not altered in dementia is challenged by the observa-
tions of Coben et al. (1983), Visser et al. (1985) and Pollock et al. (1989). Philpot et al. (1990) suggest that the flash VEP may be normal in the early stages of SDAT, abnormalities developing as the disease progresses. Hence, the extent to which latency changes over the course of time needs to be established. There have been few substantial serial VEP studies in old people, with or without dementia. The present investigation examines 4 groups of elderly people; the first two having a clinical diagnosis of dementia, one S D A T and the other MID, the third consisting of patients with functional (nonorganic) psychiatric illness and the fourth a sample of healthy controls. The subjects are examined at 6 monthly intervals over a 2 year period. Group differences and serial changes in latency and amplitude of the main components of the flash and pattern reversal VEPs are reported.
Method
Subjects
Correspondence to." Eileen P. Sloan, Ph.D., Department of Psychiatry, Ninewells Hospital, Dundee DD1 9SY, Scotland (U.K.). Tel.: 0382 60111, ext. 3182; Fax: 0382 645748.
* This work was supported by The Scottish Home and Health Department Grant Number K/MRS/50/C893.
Patients aged 65 years or over referred for the first time to the Dundee Geriatric Psychiatry service were recruited into the following 3 groups: primary degenerative dementia of the Alzheimer type (SDAT), multiinfarct dementia (MID) and functional non-organic psychiatric illness (Major Depressive Disorder), the diagnosis being rriade on clinical grounds by a consul-
326
E.P. SLOAN,G.W. FENTON
tant psychiatrist using DSM-III-R criteria. Healthy volunteer subjects aged 65 years or older formed a control group. The mean age of the subjects at the initial test session was: SDAT, 73.8 +_ 9.3 years; MID, 75.2 _+ 9.4 years; functional, 73.2 _+ 9.5 years; normal, 71.7 + 5.2 years. The respective sex distributions were: SDAT, 7 males, 33 females; MID, 10 males, 22 females; functional, 11 males, 22 females; normal, 9 males, 21 females. Subjects with eye disease, a history of epilepsy or alcohol abuse, and in the case of the SDAT and functional patients and the normal subjects, those with a history of neurological disease, were excluded from participation. On entry to the study each subject's premorbid IQ was calculated using the National Adult Reading Test (Nelson 1982). For the 3 patient groups the current IQ was measured using the Wechsler Adult Intelligence Scale (WAIS). This score was subtracted from the N A R T score to give an estimate of the degree of cognitive deterioration. The Mini-Mental State Exam (MMSE) was carried out on all subjects at each test session. The scores on the above variables at the initial test session are given in Table I. Differences between groups in premorbid and current IQ score are not significant. However, the degree of cognitive deterioration (premorbid IQ-current IQ) is significantly greater in the SDAT group than in the functional group ( P = 0.01). The MMSE score is similar in the two dementia groups and both groups in turn have significantly lower scores than the functional patients ( P < 0.002). The 3 patient groups have lower MMSE scores than the normal subjects ( P < 0.001). Subjects were seen at 6 monthly intervals for a period of 2 years if possible, giving a total of 5 test sessions. However, over the 2 year period the number of subjects in each group decreased as a result of death, illness or unwillingness to continue in the study. Table II shows subject attrition over the 2 year test period.
TABLE I Mean cognitive test scores and standard deviations (parentheses) at the initial test session. Test
SDAT
MID
Functional Normal
NART !Q
108.8 (7.0) 93.9 (15.1) 14.1 (12.2) 16.6 (5.8)
103.9 (8.9) 93.7 (15.9) 10.2 (15.9) 17.7 (7.4)
108.0 (8.8) 100.5 (11.2) 7.5 (10.3) 25.4 (4.0)
WAIS IQ NART-WAIS MMSE
107.7 (7.8) 28.4 (1.6)
TABLE II Number of subjects in each group at the initial test session and at the 6, 12, 18 and 24 month follow-up sessions. Group
1
6/12
12/12
18/12
24/12
SDAT MID Functional Normal
40 32 33 30
31 24 21 24
25 16 19 21
19 14 16 13
11 6 9 11
VEP recordings Silver-silver chloride electrodes were applied to the scalp according to the international 10-20 system at the following electrode sites: nasion, Fz, O1, 02, OPz, and Cz. Impedances were maintained at 3 k O or less. The subject was then seated in a comfortable chair in a darkened room. The VEP recordings were made from the following channels: nasion-Fz; O1-Fz; O2-Fz and OPz-Cz. For the pattern reversal VEP the stimulus consisted of a black and white checkerboard produced by a Medelec ST10 stimulator and presented on a television monitor placed 55 cm from the subject's face. Stimulation was binocular, presented full-field. The field was 27 ° 36' by 34 ° 36'. Each individual check subtended an angle of 56 min of arc. The luminance of the white squares was approximately 390 c d / m 2 and the black squares 6.5 c d / m 2. The flash stimulus was produced by a Grass PS22 stroboscope placed 55 cm from the subject's face. It was set at intensity 2 (0.26 lumen sec/m2). For the pattern reversal condition the subject, wearing corrective lenses if necessary, was asked to concentrate on a darker square in the centre of the screen. For the flash condition the subject was asked to look at the centre of the stroboscope. Stimulus presentation and data acquistion were carried out by a purpose built system analogue to digital conversion and data processing system controlled by an Apple IIe microcomputer (McAllister et al. 1983). The stimuli were presented at a rate of 1/sec. The E E G signal was amplified using an SLE electroencephalograph Type 100T, time constant 0.30 sec, high frequency filter 25 Hz and gain 100 ~ V / c m . In each condition the responses to 48 trials, with a sweep time of 500 msec and a sampling rate of 1000 samples/sec, were recorded and averaged. A 500 msec recording of prestimulus activity was also made. The averaged wave form was stored on floppy disk and plotted off-line. For the pattern reversal VEP the latency of components N1, P100 and N2 was measured and for the flash VEP the latency of components N1, P1, N2, P2 and N3 was measured on the V D U using a cursor under software control. The latency difference between the flash P2 and the pattern reversal P100 components was calculated. Baseline to peak amplitude was measured
SERIAL VEPs IN G E R I A T R I C PSYCHIATRY
327 ,.'.,
!
1
~"~ ....
m
~,,
~
:k ~.c.-~ ;.;,, 10 p.v
6/12
~u--.~'~"~---g~ ..... - ~'~'~"~~.~%, (I -4I
l{,~'
N-~..~--.~ ;~
_-..
~ ,
~ ;i/~ . . . . 4 ~
:if
K
, .F-%: .
.
.
.
I [V
Flash VEP Pattem Reversal VEP Fig. 1. Serial flash and pattern reversal VEPs in a female patient with SDAT (age 72 at initial session) over a 24 month period. Note the tendency for the flash P2 component latency to increase over time while the pattern reversal P100 shows no change in latency.
for the components of each wave form, the baseline correction being calculated on the basis of the amplitude of the prestimulus samples and applied to the
whole wave form. The measurements were made from the O1-Fz and O2-Fz derivations. Since the latency and amplitude measures from these 2 derivations were -
....
..L.
"
~ ...l..
lO v.v
6/12
12/12
18/12
Flash VEP
.~j~:~)..~
~- _ _ - . ~ C
~
.~-,
*'-.~
~
..
. ,...
j--.
,~: ~
Pattern Reversal VEP
Fig. 2. Serial flash and pattern reversal VEPs in a female patient with functional psychiatric illness (age 71 at initial session) over an 18 month period.
E.P. SLOAN, G.W. FENTON
328
not significantly different for each subject the results from the O1-Fz channel are reported.
Latency (msecs) 70
SDAT
Statistical analysis Because of the low numbers of subjects at the 24 month follow-up session, especially in the MID group, statistical analysis was carried out on the results up to 18 months only. However, the data at 24 months are presented in some tables and figures to indicate the trend at that test session. Differences in VEP measures between subject groups and across time were analysed for statistical significance using the SPSS-X3 A N O V A procedure, with Time and Diagnosis as the independent variables. The regression coefficient was calculated for each variable using the SPSS-X3 Regression procedure. In some cases independent t tests were carried out, using the SPSS-X3 t test procedure, to partial out significant differences between groups.
60
MID
50
40
30 F 20
10
Results
(1) Flash P2-pattern reversal PIO0 latency difference values Figs. 1 and 2 show serial flash and pattern reversal VEP recordings from a SDAT patient and a functional patient respectively. In Table III the mean flash P2-pattern reversal P100 latency difference values recorded at the initial and subsequent follow-up sessions are given. From Fig. 3 it can be seen that there is a clear trend for the latency difference values of the two dementia groups and, to a lesser extent the normal controls, to increase over time. Indeed, the regression coefficients calculated for each group show a significant slope for the values obtained for the SDAT patients ( P = 0.04). However, analysis of variance reveals no significant session (change over time) effect. There is a significant difference between diagnostic groups ( P < 0.001) but no session by group interaction. Since changes over time were not significant, the mean latency difference value across test session 1-18 months was calculated for each subject and used in independent samples t tests to determine differences
0
I
I
I
I
I
Initial
6
12
18
24
Test session (months) Fig. 3. Mean latency difference (msec) between the flash P2 and pattern reversal P100 components at each test session.
between the 4 groups. The latency difference value is longer in the SDAT groups than in the MID group, but the difference is not significant. Both the SDAT and MID groups have significantly higher mean latency difference values than the functional patients ( P = 0.001 and P = 0.01 respectively). The SDAT group also has a longer mean latency difference value than the normal subjects ( P = 0.03). Unexpected findings were a lack of significant differences between the MID patients and normal subjects and a significantly longer mean difference value for the controls compared to the functional patients. In Fig. 4 the mean latencies of the flash P2 and pattern reversal P100 components at each test session are displayed. It appears that the trend for
TABLE III Mean and standard deviation (parentheses) of flash P2-pattern reversal P100 latency difference (msec) at each test session. Group
1
6/12
12/12
18/12
24/12
SDAT
45.0 (25.4) 39.4 (20.8) 33.7 (21.1) 37.7 (23.3)
48.1 (24.2) 36.2 (24.2) 31.1 (20.0) 38.9 (20.4)
48.2 (24.6) 43.9 (26.3) 27.8 (19.4) 42.0 (22.1)
51.7 (32.8) 47.0 (35.6) 30.2 (20.7) 44.0 (26.0)
68.6 (25.3) 57.4 (30.8) 25.6 (25.8) 47.0 (24.3)
MID Functional Normal
SERIAL VEPs IN GERIATRIC PSYCHIATRY
329
TABLE IV Mean latency (msec) and standard deviations (parentheses) of the main flash VEP components at initial and follow-up (months) sessions. ANOVA results show that session effects are not significant (P > 0.05). Group effects are significant for components N1 (P = 0.02), N2 (P < 0.001), P2 (P < 0.001) and N3 (P = 0.007). Session × group interactions are not significant (P > 0.05). Session
SDAT
MID
Functional
Normal
N1
1 6 12 18 24
37.4 36.9 36.1 36.5 34.3
(8.4) (7.7) (6.6) (6.7) (5.4)
39.5 (8,6) 42.8(11,9) 41.4(15.6) 35.7 (4.5) 36.9 (4.4)
36.5 36.3 38.8 38.4 40.3
(4.7) (4.0) (6.2) (8.6) (8.3)
37.9 36.6 37.7 36.6 37.3
(4.3) (5.1) (5.5) (3.3) (4.0)
P1
1 6 12 18 24
71.6 72,7 72,2 73.4 74,2
(7.6) (6.3) (7.3) (6.2) (6.3)
75.3(10.3) 75.4(10.3) 75.1 (17.7) 72.3 (5.9) 73.1 (6.3)
73.3 73.2 72.8 73.4 72.5
(6.5) (7.5) (6.8) (6.4) (6.9)
72.8 71.6 74.0 73.8 73.8
(6.6) (7.5) (7.0) (5.5) (6.9)
N2
1 6 12 18 24
111.2 (12.3) 116,9 (15.0) 117,2 (15.8) 113,9 (10.7) 117,9 (15.6)
109.1 (12.0) 109.8 (12.3) 111.9 (17.1) 111.2 (15.3) 118.9.(15.8)
107.8 (12.6) 106.4 (10.6) 108.9 (11.3) 109.2 (12.4) 105.2 (12.3)
105.1 (10.0) 107.5 (10.3) 109.1 (8.2) 113.5 (13.0) 112.2 (10.9)
P2
1 6 12 18 24
146.8 (21.6) 152.1 (20.1) 153.1 (20.5) 157.0 (23.8) 164.5 (24.5)
143.3 (16.4) 144.4 (19.1) 150.9 (23.5) 151.5 (28.9) 164.0 (33.4)
138.0 (17.2) 136.8 (13.5) 134.1 (10.1) 136.5 (11.4) 134.1 (10.3)
138.8 (20.3) 141.3 (20.1) 143.4 (19.9) 143.5 (25.5) 145.7 (23.1)
N3
1 6 12 18 24
180.0 (26.3) 179.7 (20.0) 184.2 (28.3) 193.0 (34.5) 199.3 (40.5)
179.1 (22.9) 175.4 (19.8) 184.3 (30.4) 177.5 (43.3) 187.5 (46.7)
172.4 (23.7) 170.4 (21.1) 166.0 (15.7) 173.0 (23.0) 165.4 (7.3)
177.0 (23.5) 175.6 (27.6) 178.8 (16.1) 174.2 (13.6) 174.2 (10.3)
(2) Latency and amplitude values of the major flash and pattern reversal VEP components
p r o g r e s s i v e i n c r e a s e in t h e f l a s h P 2 - p a t t e r n r e v e r s a l P100 latency difference value relates to the increasing latency of the flash component.
M e a n l a t e n c i e s o f t h e m a i n c o m p o n e n t s o f t h e flash
TABLE V Mean latency (msec) and standard deviations (parentheses) of the main pattern reversal VEP components at initial and follow-up (months) sessions. ANOVA results show that session effects are not significant (P > 0.05). Group effects are noted for components NI (P = 0.015), P100 (P = 0.05) and N2 (P < 0.001). Session xgroup interactions are not significant (P > 0.05). Session
SDAT
MID
Functional
Normal
N1
1 6 12 18 24
60.6 (11.8) 61.3 (12.0) 61.6 (9.8) 62.4 (10.1) 58.4 (6.4)
63.0 (10.6) 65.5 (12.6) 65.4 (15.8) 58.6 (14.7) 61.0(14.7)
60.7 (8.8) 58.8 (9.6) 59.7 (10.1) 61.3 (10.2) 64.1 (13.0)
59.0 58.9 56.3 57.9 59.0
(6.8) (9.2) (7.5) (7.7) (6.9)
P100
1 6 12 18 24
101.7 (14.7) 104.0 (12.7) 104.9 (10.8) 106.6 (15.6) 95.1 (6.4)
103.9 (11.7) 108.1 (13.9) 107.0 (15.5) 105.8 (18.4) 106.6 (15.4)
104.3 (11.7) 105.7 (13.1) 106.3 (13.3) 106.3 (13.3) 108.5 (16.8)
100.5 102.5 101.3 99.5 98.7
(7.5) (7.0) (6.7) (5.4) (5.3)
N2
1 6 12 18 24
150.3 (19.1) 152.5 (16.8) 155.9 (18.2) 156.4 (21.3) 141.3 (11.3)
150.5 (16.3) 151.1 (19.0) 147.9 (8.1) 152.5 (19.7) 155.1 (19.4)
148.5 (15.8) 153.0 (18.7) 154.2 (21.1) 154.0 (18.4) 158.2 (27.7)
140.5 (13.8) 142.6 (12.0) 140.5 (13.3) 145.3 (10.4) 134.9 (10.2)
330
E.P. SLOAN, G.W. FENTON
~
Latency (msecs)
130
groups for a number of components. There are no significant session by group interactions. F o r VEP amplitudes there are no significant session effects. Significant group effects are noted for the following components: flash P2 ( P = 0.01); N3 ( P = 0.001); pattern reversal N1 ( P = 0 . 0 1 ) ; P100 ( P = 0.003); N2 ( P = 0.001). Again no significant session by group interactions are noted. The mean latency and amplitude values over sessions 1-18 months were compared for the 4 groups. The levels of significance for group differences in latency and amplitude are given in Table VI.
120
Discussion
160
SDAT MID
Flash P2 150 140
F
Pattern reversal P100
~
110 100
F MID N SDAT
I
I
Initial
6
I
12
I
I
18
24
Test session (months) F i g . 4. M e a n
latency (msec) of flash P2 c o m p o n e n t
and
pattern
reversal P100 c o m p o n e n t at each test session.
and pattern reversal VEPs recorded during the first and subsequent test sessions are given in Tables IV and V respectively. Analysis of variance reveals no significant session effect (change over time), though there are significant differences between the diagnostic
TABLE
In this study we were unable to demonstrate statistically significant lengthening of latency of the flash VEP P2 component over time. However, there was a consistent trend towards progressive increase in the two dementia groups during the 18 month follow-up period. This was most marked in the SDAT patients, where the regression coefficient of change over time was significant. In the single case study of Orwin et al. (1986) which reported serial increases in flash VEP P2 latency in a patient with SDAT, the changes only became prominent after 2.5 years. Using the criteria of Jorm (1988), the MMSE scores at the initial test session indicate that our two dementia patient groups were in the mild to moderate stage of the disorder. Serial cognitive testing revealed no significant deterioration over the following 18 month period. This relatively slow rate of decline in cognitive function and the fact that follow-up data to 18 months only could be used may account for the lack of significant latency change over time between the groups.
VI
S i g n i f i c a n t d i f f e r e n c e s b e t w e e n g r o u p s ( F = f u n c t i o n a l , N = n o r m a l c o n t r o l ) in t h e l a t e n c y a n d a m p l i t u d e o f t h e f l a s h a n d p a t t e r n r e v e r s a l V E P components. Flash
Pattern reversal
N1
P1
N2
P2
N3
N1
P100
N2
S D A T vs. M I D
0.02
-
0.03
.
S D A T vs. F
-
-
0.005
0.001
0.001
-
-
-
S D A T vs. N
-
-
0.005
0.001
-
0.02
-
0.001
M I D vs. F
0.04
-
0.001
0.03
0.05
-
0.001
M I D vs. N
0.03
.
0.002
0.001
0.001
Fvs. N
.
0.005
0.001
Latency
.
. .
.
.
.
.
.
.
.
.
.
Ampftude S D A T vs. M I D
0.03
0.05
0.03
-
-
S D A T vs. F
-
0.01
-
0.02
-
S D A T vs. N
-
0 . 0 0 1
-
0.01
0.001
M I D vs. F
-
-
-
0.05
-
M I D vs. N
0.01
-
-
0.03
0.001
Fvs. N
0.01
-
-
-
0.001
SERIAL VEPs IN GERIATRIC PSYCHIATRY
In a longitudinal study with elderly patients subject attrition, while unavoidable, introduces a degree of bias in data collected since those patients who remain in the study are perhaps healthier than those who drop out. In the present study, the lack of significant change in both VEP and cognitive measures could be due to the fact that patients who were in the more severe stages of dementia, and were therefore more likely to show significant increase in VEP latency, were lost. The decrease in pattern reversal P100 latency at the 24 month session in the SDAT patients supports this point, suggesting that those with prolonged latencies had dropped out by this stage, leaving those in whom latencies were within normal limits. Prolonged latency in the major flash and pattern VEP components in elderly patients may be have a useful role in predicting mortality and stage of dementia in elderly patients. The observation that the flash P2 latency tends to shorten over time in the functionally ill patients is intriguing since the trend is in the opposite direction to that seen in the other subject groups. This phenomenon may reflect an alteration in cortical functioning consequent to recovery from the abnormal mood state. If so, convergence towards the control data would have been expected. The effect of psychotropic medication on the VEP is short lived and amplitude rather than latency is influenced (Shagass 1983). Hence a drug action due to the small doses of antidepressants and neuroleptics that the functional patients were taking seems unlikely. The finding needs further investigation and replication by careful longitudinal study of a second series of functionally ill patients. The significantly greater VEP flash P2-pattern reversal P100 latency difference values in the SDAT group relative to the functional patients and the normal subjects support previous reports of differential VEP changes in dementia. The lack of significant difference between the SDAT and MID patients confirms that the latency difference phenomenon is not unique to SDAT (Wright and Furlong 1988). However, the SDAT patients did show a greater tendency towards flash VEP P2 latency increase over time. This may reflect a more severe degree of dementia in the SDAT group, a suggestion that is supported by the lack of significant difference between the MID and control subjects. Alternatively, the flash P2-pattern reversal P100 latency difference values may not discriminate between SDAT and MID in the early stages of the dementing process, but in established cases of dementia a normal flash P2-pattern P100 latency difference value favours a diagnosis of MID rather than SDAT. In this case serial recording of VEPs may be a useful aid to the differential diagnosis of these two conditions.
331
The significant latency and amplitude differences for other components of the flash and pattern reversal VEP reported by ourselves and others (Visser et al. 1976; Coben et al. 1983; Pollock et al. 1989) suggest that changes in these components merit further research attention since they may also be of diagnostic value. We thank Joan Knight and Carole Arrenberg who carried out the VEP recordings.
References Coben, L.A., Danziger, W. and Hughes, C. Visual evoked potentials in mild senile dementia of the Alzheimer type. Electroenceph. clin. Neurophysiol., 1983, 55: 121-130. Damesi, M.A., Huxley, P. and Murray, M. Flash and pattern visual evoked potentials in dementia. Electroenceph. clin. Neurophysiol., 1985, 61: S196. Doggett, C.E., Harding, G.F.A. and Orwin, A. Flash and pattern evoked potentials in presenile dementia (abstract). Electroenceph. clin. Neurophysiol., 1981, 52: 100P. Harding, G.F.A, Wright, C.E. and Orwin, A. Primary presenile dementia: the use of the visual evoked potential as a diagnostic indicator. Br. J. Psychiat., 1985, 147: 532-539. Jorm, A.F. Understanding Senile Dementia. Croom Helm, London, 1988. McAllister, H.G., Armstrong, G.A., McClelland, R.J. and Linggard, R. Data acquistion in neurophysiology: a flexible microcomputer system for recording neurophysiological data. Br. J. Audiol., 1983, 17: 275-277. Nelson, H.E. The National Adult Reading Test - Test Manual. NFER-Nelson, Windsor, 1982. Orwin, A., Wright, C.E., Harding, G.F.A., Rowan, D.C. and Rolfe, E.B. Serial evoked potential recordings in Alzheimer's disease. Br. Med. J., 1986, 293: 9-10. Philpot, M., Amin, D. and Levy, R. Visual evoked potentials in Alzheimer's disease: correlations with age and severity. Electroenceph, clin. Neurophysiol., 1990, 77: 323-329. Pollock, V.E., Schneider, L.S., Chui, H.C., Henderson, V., Zemansky, Y.M. and Sloane, R.B. Visual evoked potentials in dementia: a meta-analysis and empirical study of Alzheimer's disease patients. Biol. Psychiat., 1989, 25: 1003-1013. Shagass, C. Evoked potentials in adult psychiatry. In: J.R. Hughes and W.P. Wilson (Eds.), EEG and Evoked Potentials in Psychiatry and Behavioural Neurology. Butterworth, London, 1983: 169210. Visser, S.L., Stam, F.C., Van Tilburg, W., Jonker, C. and De Rijke, W. Visual evoked response in senile and presenile dementia. Electroenceph. clin. Neurophysiol., 1976, 40: 385-392. Visser, S.L., Van Tilburg, W., Hooijer, C., Op den Velde, W., Blom, J.L. and De Rijke, W. Visual evoked potentials (VEPs) in senile dementia (Alzheimer type) and in non-organic behavioural disorders in the elderly: comparison with EEG parameters. Electroenceph. clin. Neurophysiol., 1985, 60: 115-121. Wright, C.E. and Furlong, P. Visual evoked potentials in elderly patients with primary or multi-infarct dementia. Br. J. Psychiat., 1988, 152: 679-682. Wright, C.E., Harding, G.F.A. and Orwin, A. Presenile dementia the use of flash and pattern VEP in diagnosis. Electroenceph. clin. Neurophysiol., 1984, 57: 405-415.
325
© 1992 Elsevier Scientific Publishers Ireland, Ltd. 0168-5597/92/$05.00
EVOPOT 91682
Serial visual evoked potential recordings in geriatric psychiatry Eileen P. Sloan and George W. Fenton Department of Psychiatry, Ninewells Hospital and Medical School, Dundee DD1 9SY, Scotland (U.K.)
(Accepted for publication: 6 March 1992)
Serial visual evoked potentials to flash and pattern reversal stimuli were recorded in elderly patients with senile dementia of the Alzheimer type (SDAT), multi-infarct dementia (MID) and functional psychiatric illness, and in a group of elderly control subjects. Recordings were made at 6 monthly intervals over a 2 year period. Latency and amplitude of the main componentswere measured and the flash P2-pattern reversal P100 latency difference value was calculated. In all groups significant changes over time did not occur for any parameters but in the SDAT group the regression coefficient for the latency of the flash P2 component and the flash P2-pattern reversal P100 latency difference was significant, reflecting a trend towards increasing flash P2 latency as time progressed. The flash P2-pattern reversal P100 latency difference was longer in the SDAT and MID groups than in the functional patients, confirming the findings of previous reports. The latency difference in the SDAT group only was significantlygreater than that in the control group. Summary
Key words: Senile dementia; Visual evoked potentials; Serial recordings
Over the past decade there has been considerable research interest in the diagnostic value of changes in the visual evoked potential (VEP) to flash and pattern reversal stimuli in elderly patients with dementia. As Harding et al. (1985) stress it is necessary to record the responses to both types of stimulus since an increased flash VEP latency alone could be due to peripheral rather than central pathology, while the recording of a normal pattern reversal response along with a delayed flash response rules out that possibility. Wright et al. (1984) report differential latency changes for the major positive components of the two responses in patients with dementia; flash P2 wave delay, pattern reversal P100 component normal. Indeed, they claim that the latency difference value is a robust diagnostic measure of dementia though it does not discriminate between senile dementia of the Alzheimer type (SDAT) and multi-infarct dementia (MID) patients (Wright and Furlong 1988). These differential VEP latency changes have been confirmed (Doggett et al. 1981; Damesi et al. 1985; Harding et al. 1985). However, the view that the pattern reversal VEP is not altered in dementia is challenged by the observa-
tions of Coben et al. (1983), Visser et al. (1985) and Pollock et al. (1989). Philpot et al. (1990) suggest that the flash VEP may be normal in the early stages of SDAT, abnormalities developing as the disease progresses. Hence, the extent to which latency changes over the course of time needs to be established. There have been few substantial serial VEP studies in old people, with or without dementia. The present investigation examines 4 groups of elderly people; the first two having a clinical diagnosis of dementia, one S D A T and the other MID, the third consisting of patients with functional (nonorganic) psychiatric illness and the fourth a sample of healthy controls. The subjects are examined at 6 monthly intervals over a 2 year period. Group differences and serial changes in latency and amplitude of the main components of the flash and pattern reversal VEPs are reported.
Method
Subjects
Correspondence to." Eileen P. Sloan, Ph.D., Department of Psychiatry, Ninewells Hospital, Dundee DD1 9SY, Scotland (U.K.). Tel.: 0382 60111, ext. 3182; Fax: 0382 645748.
* This work was supported by The Scottish Home and Health Department Grant Number K/MRS/50/C893.
Patients aged 65 years or over referred for the first time to the Dundee Geriatric Psychiatry service were recruited into the following 3 groups: primary degenerative dementia of the Alzheimer type (SDAT), multiinfarct dementia (MID) and functional non-organic psychiatric illness (Major Depressive Disorder), the diagnosis being rriade on clinical grounds by a consul-
326
E.P. SLOAN,G.W. FENTON
tant psychiatrist using DSM-III-R criteria. Healthy volunteer subjects aged 65 years or older formed a control group. The mean age of the subjects at the initial test session was: SDAT, 73.8 +_ 9.3 years; MID, 75.2 _+ 9.4 years; functional, 73.2 _+ 9.5 years; normal, 71.7 + 5.2 years. The respective sex distributions were: SDAT, 7 males, 33 females; MID, 10 males, 22 females; functional, 11 males, 22 females; normal, 9 males, 21 females. Subjects with eye disease, a history of epilepsy or alcohol abuse, and in the case of the SDAT and functional patients and the normal subjects, those with a history of neurological disease, were excluded from participation. On entry to the study each subject's premorbid IQ was calculated using the National Adult Reading Test (Nelson 1982). For the 3 patient groups the current IQ was measured using the Wechsler Adult Intelligence Scale (WAIS). This score was subtracted from the N A R T score to give an estimate of the degree of cognitive deterioration. The Mini-Mental State Exam (MMSE) was carried out on all subjects at each test session. The scores on the above variables at the initial test session are given in Table I. Differences between groups in premorbid and current IQ score are not significant. However, the degree of cognitive deterioration (premorbid IQ-current IQ) is significantly greater in the SDAT group than in the functional group ( P = 0.01). The MMSE score is similar in the two dementia groups and both groups in turn have significantly lower scores than the functional patients ( P < 0.002). The 3 patient groups have lower MMSE scores than the normal subjects ( P < 0.001). Subjects were seen at 6 monthly intervals for a period of 2 years if possible, giving a total of 5 test sessions. However, over the 2 year period the number of subjects in each group decreased as a result of death, illness or unwillingness to continue in the study. Table II shows subject attrition over the 2 year test period.
TABLE I Mean cognitive test scores and standard deviations (parentheses) at the initial test session. Test
SDAT
MID
Functional Normal
NART !Q
108.8 (7.0) 93.9 (15.1) 14.1 (12.2) 16.6 (5.8)
103.9 (8.9) 93.7 (15.9) 10.2 (15.9) 17.7 (7.4)
108.0 (8.8) 100.5 (11.2) 7.5 (10.3) 25.4 (4.0)
WAIS IQ NART-WAIS MMSE
107.7 (7.8) 28.4 (1.6)
TABLE II Number of subjects in each group at the initial test session and at the 6, 12, 18 and 24 month follow-up sessions. Group
1
6/12
12/12
18/12
24/12
SDAT MID Functional Normal
40 32 33 30
31 24 21 24
25 16 19 21
19 14 16 13
11 6 9 11
VEP recordings Silver-silver chloride electrodes were applied to the scalp according to the international 10-20 system at the following electrode sites: nasion, Fz, O1, 02, OPz, and Cz. Impedances were maintained at 3 k O or less. The subject was then seated in a comfortable chair in a darkened room. The VEP recordings were made from the following channels: nasion-Fz; O1-Fz; O2-Fz and OPz-Cz. For the pattern reversal VEP the stimulus consisted of a black and white checkerboard produced by a Medelec ST10 stimulator and presented on a television monitor placed 55 cm from the subject's face. Stimulation was binocular, presented full-field. The field was 27 ° 36' by 34 ° 36'. Each individual check subtended an angle of 56 min of arc. The luminance of the white squares was approximately 390 c d / m 2 and the black squares 6.5 c d / m 2. The flash stimulus was produced by a Grass PS22 stroboscope placed 55 cm from the subject's face. It was set at intensity 2 (0.26 lumen sec/m2). For the pattern reversal condition the subject, wearing corrective lenses if necessary, was asked to concentrate on a darker square in the centre of the screen. For the flash condition the subject was asked to look at the centre of the stroboscope. Stimulus presentation and data acquistion were carried out by a purpose built system analogue to digital conversion and data processing system controlled by an Apple IIe microcomputer (McAllister et al. 1983). The stimuli were presented at a rate of 1/sec. The E E G signal was amplified using an SLE electroencephalograph Type 100T, time constant 0.30 sec, high frequency filter 25 Hz and gain 100 ~ V / c m . In each condition the responses to 48 trials, with a sweep time of 500 msec and a sampling rate of 1000 samples/sec, were recorded and averaged. A 500 msec recording of prestimulus activity was also made. The averaged wave form was stored on floppy disk and plotted off-line. For the pattern reversal VEP the latency of components N1, P100 and N2 was measured and for the flash VEP the latency of components N1, P1, N2, P2 and N3 was measured on the V D U using a cursor under software control. The latency difference between the flash P2 and the pattern reversal P100 components was calculated. Baseline to peak amplitude was measured
SERIAL VEPs IN G E R I A T R I C PSYCHIATRY
327 ,.'.,
!
1
~"~ ....
m
~,,
~
:k ~.c.-~ ;.;,, 10 p.v
6/12
~u--.~'~"~---g~ ..... - ~'~'~"~~.~%, (I -4I
l{,~'
N-~..~--.~ ;~
_-..
~ ,
~ ;i/~ . . . . 4 ~
:if
K
, .F-%: .
.
.
.
I [V
Flash VEP Pattem Reversal VEP Fig. 1. Serial flash and pattern reversal VEPs in a female patient with SDAT (age 72 at initial session) over a 24 month period. Note the tendency for the flash P2 component latency to increase over time while the pattern reversal P100 shows no change in latency.
for the components of each wave form, the baseline correction being calculated on the basis of the amplitude of the prestimulus samples and applied to the
whole wave form. The measurements were made from the O1-Fz and O2-Fz derivations. Since the latency and amplitude measures from these 2 derivations were -
....
..L.
"
~ ...l..
lO v.v
6/12
12/12
18/12
Flash VEP
.~j~:~)..~
~- _ _ - . ~ C
~
.~-,
*'-.~
~
..
. ,...
j--.
,~: ~
Pattern Reversal VEP
Fig. 2. Serial flash and pattern reversal VEPs in a female patient with functional psychiatric illness (age 71 at initial session) over an 18 month period.
E.P. SLOAN, G.W. FENTON
328
not significantly different for each subject the results from the O1-Fz channel are reported.
Latency (msecs) 70
SDAT
Statistical analysis Because of the low numbers of subjects at the 24 month follow-up session, especially in the MID group, statistical analysis was carried out on the results up to 18 months only. However, the data at 24 months are presented in some tables and figures to indicate the trend at that test session. Differences in VEP measures between subject groups and across time were analysed for statistical significance using the SPSS-X3 A N O V A procedure, with Time and Diagnosis as the independent variables. The regression coefficient was calculated for each variable using the SPSS-X3 Regression procedure. In some cases independent t tests were carried out, using the SPSS-X3 t test procedure, to partial out significant differences between groups.
60
MID
50
40
30 F 20
10
Results
(1) Flash P2-pattern reversal PIO0 latency difference values Figs. 1 and 2 show serial flash and pattern reversal VEP recordings from a SDAT patient and a functional patient respectively. In Table III the mean flash P2-pattern reversal P100 latency difference values recorded at the initial and subsequent follow-up sessions are given. From Fig. 3 it can be seen that there is a clear trend for the latency difference values of the two dementia groups and, to a lesser extent the normal controls, to increase over time. Indeed, the regression coefficients calculated for each group show a significant slope for the values obtained for the SDAT patients ( P = 0.04). However, analysis of variance reveals no significant session (change over time) effect. There is a significant difference between diagnostic groups ( P < 0.001) but no session by group interaction. Since changes over time were not significant, the mean latency difference value across test session 1-18 months was calculated for each subject and used in independent samples t tests to determine differences
0
I
I
I
I
I
Initial
6
12
18
24
Test session (months) Fig. 3. Mean latency difference (msec) between the flash P2 and pattern reversal P100 components at each test session.
between the 4 groups. The latency difference value is longer in the SDAT groups than in the MID group, but the difference is not significant. Both the SDAT and MID groups have significantly higher mean latency difference values than the functional patients ( P = 0.001 and P = 0.01 respectively). The SDAT group also has a longer mean latency difference value than the normal subjects ( P = 0.03). Unexpected findings were a lack of significant differences between the MID patients and normal subjects and a significantly longer mean difference value for the controls compared to the functional patients. In Fig. 4 the mean latencies of the flash P2 and pattern reversal P100 components at each test session are displayed. It appears that the trend for
TABLE III Mean and standard deviation (parentheses) of flash P2-pattern reversal P100 latency difference (msec) at each test session. Group
1
6/12
12/12
18/12
24/12
SDAT
45.0 (25.4) 39.4 (20.8) 33.7 (21.1) 37.7 (23.3)
48.1 (24.2) 36.2 (24.2) 31.1 (20.0) 38.9 (20.4)
48.2 (24.6) 43.9 (26.3) 27.8 (19.4) 42.0 (22.1)
51.7 (32.8) 47.0 (35.6) 30.2 (20.7) 44.0 (26.0)
68.6 (25.3) 57.4 (30.8) 25.6 (25.8) 47.0 (24.3)
MID Functional Normal
SERIAL VEPs IN GERIATRIC PSYCHIATRY
329
TABLE IV Mean latency (msec) and standard deviations (parentheses) of the main flash VEP components at initial and follow-up (months) sessions. ANOVA results show that session effects are not significant (P > 0.05). Group effects are significant for components N1 (P = 0.02), N2 (P < 0.001), P2 (P < 0.001) and N3 (P = 0.007). Session × group interactions are not significant (P > 0.05). Session
SDAT
MID
Functional
Normal
N1
1 6 12 18 24
37.4 36.9 36.1 36.5 34.3
(8.4) (7.7) (6.6) (6.7) (5.4)
39.5 (8,6) 42.8(11,9) 41.4(15.6) 35.7 (4.5) 36.9 (4.4)
36.5 36.3 38.8 38.4 40.3
(4.7) (4.0) (6.2) (8.6) (8.3)
37.9 36.6 37.7 36.6 37.3
(4.3) (5.1) (5.5) (3.3) (4.0)
P1
1 6 12 18 24
71.6 72,7 72,2 73.4 74,2
(7.6) (6.3) (7.3) (6.2) (6.3)
75.3(10.3) 75.4(10.3) 75.1 (17.7) 72.3 (5.9) 73.1 (6.3)
73.3 73.2 72.8 73.4 72.5
(6.5) (7.5) (6.8) (6.4) (6.9)
72.8 71.6 74.0 73.8 73.8
(6.6) (7.5) (7.0) (5.5) (6.9)
N2
1 6 12 18 24
111.2 (12.3) 116,9 (15.0) 117,2 (15.8) 113,9 (10.7) 117,9 (15.6)
109.1 (12.0) 109.8 (12.3) 111.9 (17.1) 111.2 (15.3) 118.9.(15.8)
107.8 (12.6) 106.4 (10.6) 108.9 (11.3) 109.2 (12.4) 105.2 (12.3)
105.1 (10.0) 107.5 (10.3) 109.1 (8.2) 113.5 (13.0) 112.2 (10.9)
P2
1 6 12 18 24
146.8 (21.6) 152.1 (20.1) 153.1 (20.5) 157.0 (23.8) 164.5 (24.5)
143.3 (16.4) 144.4 (19.1) 150.9 (23.5) 151.5 (28.9) 164.0 (33.4)
138.0 (17.2) 136.8 (13.5) 134.1 (10.1) 136.5 (11.4) 134.1 (10.3)
138.8 (20.3) 141.3 (20.1) 143.4 (19.9) 143.5 (25.5) 145.7 (23.1)
N3
1 6 12 18 24
180.0 (26.3) 179.7 (20.0) 184.2 (28.3) 193.0 (34.5) 199.3 (40.5)
179.1 (22.9) 175.4 (19.8) 184.3 (30.4) 177.5 (43.3) 187.5 (46.7)
172.4 (23.7) 170.4 (21.1) 166.0 (15.7) 173.0 (23.0) 165.4 (7.3)
177.0 (23.5) 175.6 (27.6) 178.8 (16.1) 174.2 (13.6) 174.2 (10.3)
(2) Latency and amplitude values of the major flash and pattern reversal VEP components
p r o g r e s s i v e i n c r e a s e in t h e f l a s h P 2 - p a t t e r n r e v e r s a l P100 latency difference value relates to the increasing latency of the flash component.
M e a n l a t e n c i e s o f t h e m a i n c o m p o n e n t s o f t h e flash
TABLE V Mean latency (msec) and standard deviations (parentheses) of the main pattern reversal VEP components at initial and follow-up (months) sessions. ANOVA results show that session effects are not significant (P > 0.05). Group effects are noted for components NI (P = 0.015), P100 (P = 0.05) and N2 (P < 0.001). Session xgroup interactions are not significant (P > 0.05). Session
SDAT
MID
Functional
Normal
N1
1 6 12 18 24
60.6 (11.8) 61.3 (12.0) 61.6 (9.8) 62.4 (10.1) 58.4 (6.4)
63.0 (10.6) 65.5 (12.6) 65.4 (15.8) 58.6 (14.7) 61.0(14.7)
60.7 (8.8) 58.8 (9.6) 59.7 (10.1) 61.3 (10.2) 64.1 (13.0)
59.0 58.9 56.3 57.9 59.0
(6.8) (9.2) (7.5) (7.7) (6.9)
P100
1 6 12 18 24
101.7 (14.7) 104.0 (12.7) 104.9 (10.8) 106.6 (15.6) 95.1 (6.4)
103.9 (11.7) 108.1 (13.9) 107.0 (15.5) 105.8 (18.4) 106.6 (15.4)
104.3 (11.7) 105.7 (13.1) 106.3 (13.3) 106.3 (13.3) 108.5 (16.8)
100.5 102.5 101.3 99.5 98.7
(7.5) (7.0) (6.7) (5.4) (5.3)
N2
1 6 12 18 24
150.3 (19.1) 152.5 (16.8) 155.9 (18.2) 156.4 (21.3) 141.3 (11.3)
150.5 (16.3) 151.1 (19.0) 147.9 (8.1) 152.5 (19.7) 155.1 (19.4)
148.5 (15.8) 153.0 (18.7) 154.2 (21.1) 154.0 (18.4) 158.2 (27.7)
140.5 (13.8) 142.6 (12.0) 140.5 (13.3) 145.3 (10.4) 134.9 (10.2)
330
E.P. SLOAN, G.W. FENTON
~
Latency (msecs)
130
groups for a number of components. There are no significant session by group interactions. F o r VEP amplitudes there are no significant session effects. Significant group effects are noted for the following components: flash P2 ( P = 0.01); N3 ( P = 0.001); pattern reversal N1 ( P = 0 . 0 1 ) ; P100 ( P = 0.003); N2 ( P = 0.001). Again no significant session by group interactions are noted. The mean latency and amplitude values over sessions 1-18 months were compared for the 4 groups. The levels of significance for group differences in latency and amplitude are given in Table VI.
120
Discussion
160
SDAT MID
Flash P2 150 140
F
Pattern reversal P100
~
110 100
F MID N SDAT
I
I
Initial
6
I
12
I
I
18
24
Test session (months) F i g . 4. M e a n
latency (msec) of flash P2 c o m p o n e n t
and
pattern
reversal P100 c o m p o n e n t at each test session.
and pattern reversal VEPs recorded during the first and subsequent test sessions are given in Tables IV and V respectively. Analysis of variance reveals no significant session effect (change over time), though there are significant differences between the diagnostic
TABLE
In this study we were unable to demonstrate statistically significant lengthening of latency of the flash VEP P2 component over time. However, there was a consistent trend towards progressive increase in the two dementia groups during the 18 month follow-up period. This was most marked in the SDAT patients, where the regression coefficient of change over time was significant. In the single case study of Orwin et al. (1986) which reported serial increases in flash VEP P2 latency in a patient with SDAT, the changes only became prominent after 2.5 years. Using the criteria of Jorm (1988), the MMSE scores at the initial test session indicate that our two dementia patient groups were in the mild to moderate stage of the disorder. Serial cognitive testing revealed no significant deterioration over the following 18 month period. This relatively slow rate of decline in cognitive function and the fact that follow-up data to 18 months only could be used may account for the lack of significant latency change over time between the groups.
VI
S i g n i f i c a n t d i f f e r e n c e s b e t w e e n g r o u p s ( F = f u n c t i o n a l , N = n o r m a l c o n t r o l ) in t h e l a t e n c y a n d a m p l i t u d e o f t h e f l a s h a n d p a t t e r n r e v e r s a l V E P components. Flash
Pattern reversal
N1
P1
N2
P2
N3
N1
P100
N2
S D A T vs. M I D
0.02
-
0.03
.
S D A T vs. F
-
-
0.005
0.001
0.001
-
-
-
S D A T vs. N
-
-
0.005
0.001
-
0.02
-
0.001
M I D vs. F
0.04
-
0.001
0.03
0.05
-
0.001
M I D vs. N
0.03
.
0.002
0.001
0.001
Fvs. N
.
0.005
0.001
Latency
.
. .
.
.
.
.
.
.
.
.
.
Ampftude S D A T vs. M I D
0.03
0.05
0.03
-
-
S D A T vs. F
-
0.01
-
0.02
-
S D A T vs. N
-
0 . 0 0 1
-
0.01
0.001
M I D vs. F
-
-
-
0.05
-
M I D vs. N
0.01
-
-
0.03
0.001
Fvs. N
0.01
-
-
-
0.001
SERIAL VEPs IN GERIATRIC PSYCHIATRY
In a longitudinal study with elderly patients subject attrition, while unavoidable, introduces a degree of bias in data collected since those patients who remain in the study are perhaps healthier than those who drop out. In the present study, the lack of significant change in both VEP and cognitive measures could be due to the fact that patients who were in the more severe stages of dementia, and were therefore more likely to show significant increase in VEP latency, were lost. The decrease in pattern reversal P100 latency at the 24 month session in the SDAT patients supports this point, suggesting that those with prolonged latencies had dropped out by this stage, leaving those in whom latencies were within normal limits. Prolonged latency in the major flash and pattern VEP components in elderly patients may be have a useful role in predicting mortality and stage of dementia in elderly patients. The observation that the flash P2 latency tends to shorten over time in the functionally ill patients is intriguing since the trend is in the opposite direction to that seen in the other subject groups. This phenomenon may reflect an alteration in cortical functioning consequent to recovery from the abnormal mood state. If so, convergence towards the control data would have been expected. The effect of psychotropic medication on the VEP is short lived and amplitude rather than latency is influenced (Shagass 1983). Hence a drug action due to the small doses of antidepressants and neuroleptics that the functional patients were taking seems unlikely. The finding needs further investigation and replication by careful longitudinal study of a second series of functionally ill patients. The significantly greater VEP flash P2-pattern reversal P100 latency difference values in the SDAT group relative to the functional patients and the normal subjects support previous reports of differential VEP changes in dementia. The lack of significant difference between the SDAT and MID patients confirms that the latency difference phenomenon is not unique to SDAT (Wright and Furlong 1988). However, the SDAT patients did show a greater tendency towards flash VEP P2 latency increase over time. This may reflect a more severe degree of dementia in the SDAT group, a suggestion that is supported by the lack of significant difference between the MID and control subjects. Alternatively, the flash P2-pattern reversal P100 latency difference values may not discriminate between SDAT and MID in the early stages of the dementing process, but in established cases of dementia a normal flash P2-pattern P100 latency difference value favours a diagnosis of MID rather than SDAT. In this case serial recording of VEPs may be a useful aid to the differential diagnosis of these two conditions.
331
The significant latency and amplitude differences for other components of the flash and pattern reversal VEP reported by ourselves and others (Visser et al. 1976; Coben et al. 1983; Pollock et al. 1989) suggest that changes in these components merit further research attention since they may also be of diagnostic value. We thank Joan Knight and Carole Arrenberg who carried out the VEP recordings.
References Coben, L.A., Danziger, W. and Hughes, C. Visual evoked potentials in mild senile dementia of the Alzheimer type. Electroenceph. clin. Neurophysiol., 1983, 55: 121-130. Damesi, M.A., Huxley, P. and Murray, M. Flash and pattern visual evoked potentials in dementia. Electroenceph. clin. Neurophysiol., 1985, 61: S196. Doggett, C.E., Harding, G.F.A. and Orwin, A. Flash and pattern evoked potentials in presenile dementia (abstract). Electroenceph. clin. Neurophysiol., 1981, 52: 100P. Harding, G.F.A, Wright, C.E. and Orwin, A. Primary presenile dementia: the use of the visual evoked potential as a diagnostic indicator. Br. J. Psychiat., 1985, 147: 532-539. Jorm, A.F. Understanding Senile Dementia. Croom Helm, London, 1988. McAllister, H.G., Armstrong, G.A., McClelland, R.J. and Linggard, R. Data acquistion in neurophysiology: a flexible microcomputer system for recording neurophysiological data. Br. J. Audiol., 1983, 17: 275-277. Nelson, H.E. The National Adult Reading Test - Test Manual. NFER-Nelson, Windsor, 1982. Orwin, A., Wright, C.E., Harding, G.F.A., Rowan, D.C. and Rolfe, E.B. Serial evoked potential recordings in Alzheimer's disease. Br. Med. J., 1986, 293: 9-10. Philpot, M., Amin, D. and Levy, R. Visual evoked potentials in Alzheimer's disease: correlations with age and severity. Electroenceph, clin. Neurophysiol., 1990, 77: 323-329. Pollock, V.E., Schneider, L.S., Chui, H.C., Henderson, V., Zemansky, Y.M. and Sloane, R.B. Visual evoked potentials in dementia: a meta-analysis and empirical study of Alzheimer's disease patients. Biol. Psychiat., 1989, 25: 1003-1013. Shagass, C. Evoked potentials in adult psychiatry. In: J.R. Hughes and W.P. Wilson (Eds.), EEG and Evoked Potentials in Psychiatry and Behavioural Neurology. Butterworth, London, 1983: 169210. Visser, S.L., Stam, F.C., Van Tilburg, W., Jonker, C. and De Rijke, W. Visual evoked response in senile and presenile dementia. Electroenceph. clin. Neurophysiol., 1976, 40: 385-392. Visser, S.L., Van Tilburg, W., Hooijer, C., Op den Velde, W., Blom, J.L. and De Rijke, W. Visual evoked potentials (VEPs) in senile dementia (Alzheimer type) and in non-organic behavioural disorders in the elderly: comparison with EEG parameters. Electroenceph. clin. Neurophysiol., 1985, 60: 115-121. Wright, C.E. and Furlong, P. Visual evoked potentials in elderly patients with primary or multi-infarct dementia. Br. J. Psychiat., 1988, 152: 679-682. Wright, C.E., Harding, G.F.A. and Orwin, A. Presenile dementia the use of flash and pattern VEP in diagnosis. Electroenceph. clin. Neurophysiol., 1984, 57: 405-415.
