- Email: [email protected]
An EEG power index (eyes open vs. eyes closed) to differentiate Alzheimer's from vascular dementia and healthy ageing
ELSEVIER
Archives of Gerontology and Geriatrics 22 (1996) 245-260
ARCHIVES OF GERONTOLOGY AND GERIATRICS
An EEG power index (eyes open vs. eyes closed) to differentiate Alzheimer's from vascular dementia and healthy ageing Mario Signorino a,*, Enrico Brizioli b, Loredana Amadio b, Natascia BelardineUP, Eugenio Pucci a, Franco Angeleri a "Institute for Nervous System Diseases, University of Ancona, Ospedale Regionale di Torrette di Ancona, Ancona, Italy bLN.R.C.A. - Italian National Research Centres on Ageing - Department of Research, Ancona, ltaly
Received 5 October 1995; revised 20 December 1995; accepted 23 December 1995
Abstract EEG differential power patterns between Alzheimer's (AD, 50 patients) and vascular (VaD, 37 patients) dementia and between these two and 36 healthy ageing subjects, were studied in the 6.5-12 Hz band of the ongoing EEG recorded during the rest eyes closed (REC) and eyes open (REO) conditions. From the EEGs (16 electrodes, 10-20 international system except for Fz, Cz, Pz), a 6.5-12 Hz band, wider than the alpha range (alpha-like), was chosen and processed to include the highest theta frequencies characterising the occipital dominant activity in dementia. A global and occipital EEG Power Index (PI) was calculated and used considering the absolute powers during REC and REO. The MANOVA was used to compare the figures. Beating in mind that the higher the PI value the greater the difference between the 6.5-12 Hz EEG band powers of REC vs. REO, the results were as follows: (i) in the patients with Alzheimer's and vascular dementia the global and occipital PIs were significantly lower than those in controls; (ii) in the patients with Alzheimer's dementia the same PIs were significantly lower that those of the patients with VaD; (iii) healthy elderly subjects showed significantly lower powers in the 6.5-12 Hz frequencies at T5 and O1 in REO as compared to dementia patients. The pathophysiological implications and the clinical applications of these results are discussed. Keywords: Alzheimer's disease; Vascular dementia; Ageing; Alpha-rhythms; EEG spectral
analysis
* Corresponding author, Tel.: + 39 71 888989; Fax: + 39 71 888989. 0167-4943/96/$15.00 © 1996 Elsevier Science Ireland Ltd. All fights reserved PII SO167-4943(96)00697-8
246
M. Signorino el al.
Arch. Gerontol. Geriatr. 22 (1996) 245 260
1. Introduction
Using both traditional EEG and its computerised quantitative analysis a reduction in alpha rhythm was found in patients with Alzheimer's disease (AD) (Letemendia and Pampiglione, 1958; Gordon and Sire, 1967; Penttil/i et al., 1985; Brenner et al., 1986; Giaquinto and Nolfe, 1986; Giannitrapani et al., 1991; Maurer and Dierks, 1992: Sloan and Fenton, 1993; Miyauchi et al., 1994). The disease also induces a decrease in alpha responsiveness to photic stimulation according to Politoff et al. (1992) and a fall in the desynchronizing effect on the dominant alpha-like ongoing activity normally induced by a static condition such as rest with eyes open (Pritchard et al., 1994; Signorino et al., 1995). These findings suggest that in AD the cortical mechanisms generating the alpha rhythm are altered, probably because of the lesions in the occipital neocortex where the alpha generator is thought to be located (Steriade et al., 1990). On the contrary in vascular dementia patients (VaD) the alpha rhythm is usually preserved although with a slower frequency than in normal elderly controls (Van der Drift and Kok, 1972). In conclusion, the alpha pattern and its 'reactivity' in AD, VaD and elderly normal subjects seem to be important for clarifying the different pathophysiological mechanisms of the ageing brain and the two types of possibly overlapping dementia (Friedland, 1993). On the other hand, for clinical application, we thought it would be interesting to characterise any EEG spectral patterns which differentiate AD from VaD and these types of dementia from healthy ageing. On the basis of this background, the aim of this study was as follows: (i) to identify the sub-alpha components, which often represent the dominant occipital rhythm not only in dementia patients (Giannitrapani, 1991) but also in healthy elderly subjects, according to some authors (review in Steriade et al., 1990) by analysing a wider band than the traditional alpha range (i.e. 8-12 Hz) towards the low frequencies (6.5-12 Hz); (ii) to compare the EEG ongoing power during the rest eyes closed and eyes open conditions not only at the occipital level but also at the other locations covering the cortical areas of the two hemispheres.
2. Materials and methods
2. I. Subjects Thirty-seven patients affected by vascular and 50 patients by Alzheimer's dementia were studied. The diagnoses were reached according to DSM-III-R (American Psychiatric Association, 1987), NINCDS-ADRDA (McKhann et al., 1984) and
50 31
AD VaD
*Mann-Whitney
n
Diagnosis
31 10
F
68.5 71.3
8.2 1.4
48-81 51-84
36 31.5
21.6 28.3
S.D.
Mean
Range
Mean
S.D.
Duration of the disease (months)
Age (years)
U-test P i 0.01.
19 27
M
Table 1 Demographic and clinical characteristics of the dementia patients
6-156 6-120
Range 13.3 17.6’
Mean
MMSE
6.1 5.1
S.D.
O-26 O-26
Range
4.6 4’
Mean
GDS
1 1
S.D.
3-6 3-6
Range
3.5 2.9:
Mean
GS
1.2 1
S.D.
2-l 2-5
Range
248
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
N I N D S - A I R E N (Roman et al., 1993). Clinical characteristics of the two groups of patients are summarised in Table 1. Thirty-six normal elderly subjects (NE) were also studied as age-matched controls (14 males, 22 females; mean age 64.4 years, S.D. 4.6; range 60-80 years). Oral informed consent was obtained from all subjects or from tutors of patients with severe dementia. All subjects except one were right-handed according to the Edinburgh Inventory (Oldfield, 1971). The severity of dementia was assessed according to the MMSE (Folstein et al., 1975), the Global Deterioration Scale (GDS; Reisberg et al., 1982) and Giannitrapani's scale (GS; Giannitrapani et al., 1991). Brain MRI was available for 59 patients and CT for the remaining patients. As the sample studied was the same as in a previous study, more details can be drawn from Signorino et al. (1995). 2.2. E E G recording and spectral analysis
Recording took place in a room with a low constant level of artificial lighting and environmental noise. The subjects were seated in a comfortable padded chair and asked to close their eyes when requested and to try to stay awake and relaxed. All EEGs were recorded between 9 a.m. and 11 a.m. after a normal night's sleep and the usual breakfast. Subjects underwent 5 min of E E G recordings at rest with eyes closed (REC) and then again with eyes open (REO). Special care was taken to prevent drowsiness and minimise artefact acquisition. The E E G was recorded by Ag/AgC1 cup electrodes at 16 monopolar locations arranged according to the international 10-20 system which explored the left scalp side (Fpl, F7, T3, T5, O1, F3, C3, P3) and the right scalp side (Fp2, F8, T4, T6, 02, F4, C4, P4) with linked earlobes as reference. Impedance was kept lower than 5 KOhm. EOG was monitored using two additional electrodes placed 1 cm below and 1 cm above the outer canthus of the eye. The bandpass filter was set in a range of 1 64 Hz. Q-notch filters at frequencies of 50 and 100 Hz were also used. Each channel was calibrated in microvolts and the DC level corrected using appropriate algorithms, including subtraction of the mean DC offset from each epoch selected. The signals were sampled at a rate of 128 samples/sec, with 12-bit analog-to-digital conversion. For each physiological condition, an expert neurologist selected 20 E E G artefact-free epochs of 2 sec duration. The Fast Fourier Transforms (FFT) of the selected epochs were calculated using the F F T algorithm, and the spectral density function W(f) was estimated as the average of these transforms (Bendat and Piersol, 1986). In order to reduce the bias error of the sample spectrum the E E G epochs were analysed using the Hanning data window. Absolute (AP) and relative (RP) powers were calculated for the 6.5-12 Hz band. A logarithmic transformation was applied to absolute powers (In x) and relative powers (ln Ix/(1 - x)]) for each band in each epoch to normalise the data, in accordance with John et al. (1980), Gasser et al. (1982), Oken and Chiappa (1988). A global spectrum was also calculated consisting of the mean of 16-channel spectra. For more technical details, see Angeleri and Signorino (1989) and Signorino et al. (1995).
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
249
For each subject and for each derivation, a Power Index (PI) was calculated as"
LAPREC - LAPREO x 100/LAPREC where LAPREC is the log of absolute power during REC and LAPREO is the log of absolute power during REO of the 6.5-12 Hz band. The PI was used in order to decrease the variability of power values, particularly absolute ones, in the subjects. Moreover, to decrease the number of variables to be analysed, PI values were pooled in two parameters: (i) the global PI calculated as the mean of PIs of all 16 derivations studied; (ii) the occipital PI, namely the mean of PI values at 01 and 02.
2.3. Statistical analysis The Mann-Whitney U test was performed to compare the clinical data. The Bonferroni-adjusted ~* = ot/k was employed when k simultaneous tests were necessary to evaluate differences between the experimental groups. To evaluate the differences in the variables in the brain electrical activity in the three groups under examination (AD, VaD, NE), under two different conditions (REC and REO) and for each of the electrodes, a three-way analysis of variance for repeated measures was used. The factors considered for each variable were: group/condition~electrode, using the MANOVA of the SPSS/PC package. The significant interactions were post hoc processed by the Tukey test. The significance level chosen was 5%. The Spearman rank order correlation (one-tailed; ~ = 0.01; CSS-2.1 software) was used to correlate GDS, GS and MMSE scores with power indices.
3. Results
3.1. L A P and L R P values of the 6.5-12 Hz band." comparison for groups, conditions and electrodes The results of the comparison of absolute (LAP) and relative power (LRP) values in the groups (NE, VaD, AD) in the two experimental conditions (i.e., REO and REC) at each derivation are summarised in Fig. 1. As we can see, AD patients exhibit a trend, in both REC and REO conditions, which is quite similar to that of the controls during the REC condition. However controls have a selective REO decrease of activity in T5 and 01 in the left hemisphere, showing a hemispheric asymmetry not evident in the dementia patients (P < 0.05). As shown in Table 2 the results of MANOVA reveal statistically significant differences for group factors in all variables, including the second and third level interactions. The significant interaction condition x electrode may be mainly due
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
250
h i m e ~ l p oide
lee : o i I p slde
ViJliai S.S I
~-.
-'/~
/~"
j
/ /
/
4,/ '~t-
--,',. //
. Q. ~ _& x
"~
o
Lx
o
//x
//
\,~/t
/
. ~.
~\ ~___,,,___~"/*-~,,_~..~'~ "
,a___..-"~~~
"
0" -.
I,S
•
FT~
,e. ......... e
"
• ...........~
"
I
I
!
~
I
I
1:8
11"4
331
O~
1=4
C4
I
~
I
114 FPI
"
~...........,0
I
1=7
. 4P
"~
I
l
I
I
I
TI
T~
O1
F8
(~
~ m ~ e
I
I~1
~ m ~ e
mmkme -1.T -1.S ~
.
-1.$ -1.1
~,t
"'IL\
\ "
0.8
/
\x.
[email protected]
~ FI
""4 ~\
/-
J \.
I
I
4
I
I
~
TI
02
F4
CA
P4
---&--¥10-iqEO
=
~"
\
:,,
T4
ADJIC
\
..l
t --
FP1 ~I)4EO
• ~... N E - / I E C
I
I
Fir
T3 - --&-
\4,..
I T$
/
"
/
'
\\
t
t
O1
FI
1 CS
I
Pll
VaI~REC
..... 4,-.-NE-FIEO
Fig. 1. Log mean values of absolute (upper part) and relative powers (lower part) for groups, conditions and electrodes. The locations of recording electrodes are shown on the abscissa axis. They are grouped according to whether they belong to the left or right scalp side. On the ordinate axis the corresponding mean values of the log powers of the 6.5-12 Hz band. The geometric symbols and the lines refer to the groups and conditions: white squares and solid lines indicate the AD group in the eyes closed condition (AD-REC), black squares and solid lines indicate the AD group in the eyes open condition (AD-REO); white triangles and broken lines indicate the VaD group in the eyes closed condition (VaD-REC), and so on as shown in the figure. Note that the contour lines are quite symmetrical when comparing left and right sides; this occurs in all groups and conditions except those of normal subjects in eyes open condition (NE-REO). This line, in fact, shows lower values in the left side (Fpl, F7, T3, T5, O1, F3, C3) with respect to the right electrode positions. The difference is statistically significant in T5 and O l. Furthermore, we can see that the VaD group is characterised by the highest absolute power values in the eyes closed condition with respect to the other groups while in the same condition the relative powers of the VaD and NE groups tend to overlap.
251
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
Table 2 Repeated measures analysis of variance for the brain electrical activity Source of variation
Group Condition Electrodes Group x Electrodes Group × Condition Electrodes x Condition Group × Electrodes x Condition
LAP
LRP
F
D.f.
P value
F
D.f.
P value
11.32 108.65 57.09 1.83 22.39 32.50 3.28
2 1 15 30 2 15 30
<0.001 <0.001 <0.001 0.004 <0.001 <0.001 <0.001
14.74 144.59 112.81 1.85 14.37 7.34 1.75
2 1 15 30 2 15 30
<0.001 <0.001 <0.001 0.004 <0.001 <0.001 0.007
to the selective decrease o f L A P value between the R E C and the R E O in the controls. 3.2. L A P and L R P mean values o f the 6 . 5 - 1 2 H z band in R E C and R E O conditions 3.2.1. L A P (Fig. 2): the m e a n L A P values o f the elderly controls at all electrodes in the R E C condition is 4.46 + 0.51 (range = 3.21-5.61). In the R E O condition, this value decreases to 3.63 + 0.44 (range = 2.80-4.81). The difference is statistically significant (Tukey test = P < 0.05). V a D patients show statistically significant higher values in R E C . In fact the m e a n o f all the electrodes is 4.89 ___ 0.49 (range = 3.68-6.69) during R E C and 4.49 _ 0.46 (range = 3.63-6.17) during R E O . A D patients show significantly lower L A P values during R E C c o m p a r e d to the other groups, with a m e a n value o f 3.99 + 0.47 (range = 2.10-5.23) which is n o t statistically different f r o m the value during R E O with a m e a n value o f 3.86 _ 0.41; (range = 3.01-9.45). 3.2.2. L R P (Fig. 3): controls and V a D patients show similar values during R E C , while A D patients show a significantly lower value. I n the R E O condition, only the control g r o u p shows a statistically significant decrease in activity in respect to REC. 3.3. P o w e r indices
The results o f the c o m p a r i s o n o f PI values m a y be summarised as follows: in N E the global P I is 16.58 + 12.39 while the occipital PI value is greater (22.50
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
252
+ 14.38); in AD patients the global PI and the occipital PI are respectively 2.09 __+ 7.84 and 3.64 ___ 11.96; in VaD patients the global PI is 7.79 ___ 7.51 but the occipital PI is higher (12.38 + 12.34). The pooled PI values provide statistically significant differences in the 3 groups capable of discriminating not only dementia patients from normal subjects but also AD from VaD (Table 3). Neither the global nor the occipital PI correlates with MMSE, GDS and GS scores in either AD (Table 4) or VaD (Table 5).
LAP moon values 5.1 4.0 4.7 4.6
*'-.. "'...
4.8 4.1
"-.. |.It
"'""'--.... 8.7
°. -.$ I
|,S I
TulloylilvlilploCempmioonToot: N E . I m C
vii N E . f l E O p c 0 . 0 g
N D . N O C vs V m D . I m C p , c 0 . 0 S
Fig. 2. Graph of the differences in the mean values of the log of absolute powers (mean of all electrodes and of all subjects in each group) shown on the y axis, between eyes closed condition (left extremity of the x axis) and eyes open condition (right extremity of the x axis). The continuous line with squares refers to the AD group, the broken line with triangles refers to the VaD group, the small points line with rhombus refers to NE groups. The difference between eyes closed and open conditions is significant in the NE group. Furthermore, one can note a significant difference between the values of the VaD and AD groups in the eyes open condition.
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
253
I.ltP -1.2
.......41 -1
..4, . .....°'"'*"'
I 41.1
41,1
41.4
41.z
I
--.--v.°
TW~MullipleC,Onlmlm~Toal:
I
.... . .
NII-ImC va llE-PmOpcO.M AD.PmC vm WD-RnCpcO.06 AD-ImC vs N E . R ~ p c 0 . U
Fig. 3. Graph of the difference in the mean values of the log transformed relative powers (mean of all electrodes and of all subjects in each group) shown on the y axis, between eyes closed and open conditions (respectively left and fight extremity of the x axis). For further details see Fig. 2.
Table 3 Power Index (PI) Global PI
AD VaD NE
Occipital PI
Mean
S.D.
2.09 7.79 16.58
7.84 7.51 12.39
Range - 13.3 -7.2 -13.3
26.8 27.0 38.9
Mean
S.D.
3.64 12.38 22.50
11.96 12.34 14.38
Tukey multiple comparison test: all groups different. P < 0.05.
Range - 18.1 - 18.3 -11.8
38.3 43.5 51.8
M. Signorino et ell. : Arch. Gerontol. Geriatr. 22 11996) 245 260
254
Table 4 S p e a r m a n r a n k order correlations in A D (:~ = 0.01) R
t
P
Global PI vs. MMSE GDS GS
0.177 0.317 0.008
1.249 -2.319 - 0.054
0.218 0.025 0.957
0.009 0.088 11.079
0.060 0.612 1/.547
0.952 0.543 0.586
Occipital PI v,s. M M SE GDS GS
Table 5 S p e a r m a n r a n k order correlations in V a D (:~ = 1).01)
R
:
P
Global P I ~:s. MMSE GDS GS
Occipital PI
O.245 0.21)7 - 0.404
.495 - 1,249 2.616
0.144 0.220 0.013
0.169 0.170 0.363
1.016 1.022 2.304
0.316 0.314 0.027
v,~.
MMSE GDS GS
4. Discussion Before discussing the results, three points must be made. Firstly, the power indices (PIs) which were used to process the powers of the 6.5 12 Hz band in the two rest conditions cannot be completely identified with the traditional significance of EEG desynchronization; however, such indices evaluate two basic components of desynchronization, i.e. the occurrence and/or amplitude attenuation of the ongoing EEG frequencies in the 6.5 12 Hz band (Gotman, 1990). Secondly, this band, which is wider than the alpha frequency range, was considered better suited to studying the highest theta frequencies which, sometimes with the alpha rhythm, characterise the dominant activity in patients with dementia (review in Leuchter et al., 1993); briefly this band was defined as 'alpha-like rhythms'. Thirdly, the comparison of the two types of dementia and healthy ageing was carried out by the log transformation of both absolute and relative power values to obtain parametric statistics more reliably (John et al., 1980; Gasser et al., 1982; Oken and Chiappa, 1988; Leuchter et al., 1993; Schreiter-Gasser et al., 1993).
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
255
Focusing the analysis on the alpha-like band, a progressive decrease in the global and occipital PIs was found when comparing age-matched healthy subjects and patients with vascular and Alzheimer's dementia. In the occipital regions the decrease differences were the greatest thus suggesting that the reduction in the alpha-like reactivity in the two types of dementia has its core in the occipital regions where a consistent alpha generator component is thought to be located (reviews in Markand, 1990 and Steriade et al., 1990). The statistically significant PI decreases from healthy subjects to AD patients with the intermediate mean values of VaD gives more exhaustive evidence of previous observations in AD patients, e.g. the reduction of decrease in alpha-like powers during the REO condition that they show (Pritchard et al., 1994; Signorino et al., 1995). The same findings agree with the observations of several authors (Visser et al., 1976; Wright et al., 1984; Visser et al., 1985; Wright et al., 1986; Bajalan et al., 1986; Coburn et al., 1989) on the loss of the flash-VEP induced alpha desynchronization in AD patients and with the investigation by Politoff et al. (1992) of the lack of alpha cortical photic driving effect in the same patients. Furthermore, a minor negative effect on VaD of the alpha-like reactivity is established by the present study which found, however, statistically significant differences between the PI mean values of this group and those of the elderly subjects. The use of the 6.5-12 Hz alpha-like band in the EEG spectral analysis allowed the present study to investigate both the alpha and highest theta frequencies where recruiting and oscillating are like a single dominant activity. The results obtained by the spectral analysis suggest that the generator of alpha rhythm extends its recruiting oscillations towards the highest theta frequencies which in part or totally, replace alpha rhythm in dementia (review in Leuchter et al., 1993). Finally, the significant lower powers of the 6.5-12 Hz frequencies at the T5 and O1 derivations found in the elderly healthy subjects during the REO condition as compared with the analogous mean powers of the patients with dementia could suggest a loss of the EEG hemispheric asymmetry in the latter as observed several years ago by Angeleri et al. (1982). If the PI drop reflects, at least in part, the ongoing EEG desynchronization with its focal point in the occipital regions, it can be considered as an indicator of the efficiency of the neural networks involved in the mechanisms of vigilance and attention. Healthy ageing subjects, but even more so brain damaged dementia patients, can have alteration in these networks and their oscillations and therefore the functions of vigilance and attention. In AD the neural mechanisms ruling the alpha-like power patterns are definitely altered and the phenomenon can easily be attributed to the neurotransmitter deficit and neuronal lesions of the neocortex induced by the disease. In this context the neuropathological works that underline the involvement of the occipital areas must be quoted (Beach et al., 1989; Bell and Ball, 1990; Mizutani et al., 1990; Hof and Morrison, 1990). Also, the hypothesis which maintains that the deficit of the cortical projecting cholinergic system is associated with an intracortical disconnec-
256
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
tion seems worth mentioning (Soininen et al., 1989; Leuchter et al., 1992; Soininen and Riekkinen, 1992). The PI mean values in VaD, which are in the middle position of the values between AD and healthy ageing, can be explained by the characteristics of the vascular lesions and at the same time could justify the slower and different evolution of the mental deterioration occurring in these patients. Some authors consider, in fact, that the mental deficit in VaD is caused by an intracortical and cortical-subcortical deafferentation like that in AD patients, but induced by the typical vascular lesions of the hemispheric white matter instead of the drop in the cortical cholinergic system associated with the loss of a great quantity of pyramidal neurones which occurs in AD (review in Leuchter et al., 1992). The mental deterioration was more severe in AD than VaD patients, however, Spearman rank order correlation showed no statistically significant correlation between either the global or occipital PIs and AD or VaD dementia scores. These findings fit with those acquired by Giannitrapani et al. (1991) in dementia patients who showed no relation between the alpha loss and severity of cognitive impairment, but a positive correlation between the former and the degree of the cortical lesions. On the other hand, the previous works that report positive correlation between cognitive performances and changes in alpha activity are concerned with the rest eyes closed condition only (Penttil~i et al., 1985; Brenner et al., 1986; Primavera et al., 1990). The abnormal behaviour of the dominant activity in the 6.5-12 Hz band, as found in patients with dementia, may be explained on the basis of a model based on availability and facilitation of neural networks generating alpha activity in the ongoing EEG (Fig. 4). The model, initially proposed by Speckman and Caspers (1973), was recently developed by Pfurtscheller (1992). According to these authors, the more the vigilance increases, the more the thalamo-cortical facilitation increases and the neuronal availability decreases. Vice-versa, the more the vigilance decreases the more the availability of neural networks grows as long they are not involved in cognitive and sensorial processes. At the point of intersection of neuronal availability and facilitation, that is, in the condition of wakefulness with eyes closed, there is a greater equilibrium between the conditions of wakefulness and psychosensorial rest. This corresponds to the maximum expression of alpha rhythm as the dominant activity in the ongoing EEG. Following this model to interpret the changes induced by dementia a reduction, both in the availability and facilitation mechanisms, may be supposed and such a reduction is certainly greater in Alzheimer's disease. In the graph in Fig. 4 this reduction corresponds to a gentler curve in the neuronal availability and a flattening of the slope line of facilitation. The intersection of the two lines shows an enlargement and smoothing of the recruiting area of the background activity which is shifted towards the REO condition. The availability and facilitation changes induced by dementia probably, in their turn, follow the alterations in the strictly interrelated intracortical and cortical
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
257
sub-cortical n e u r o n a l circuits ruling the recruiting oscillation o f the alpha-like rhythms. I n this context the PIs m a y be considered a detector o f the n o r m a l a n d a b n o r m a l v a r i a t i o n s o f these n e u r a l functions. F o r diagnostic purposes it m a y be used to differentiate healthy ageing from A D a n d the latter from VaD. It m a y also help in the differential diagnosis o f single cases where the n o r m a l or pathological definition or diagnosis o f A D or V a D are u n c e r t a i n , m a i n l y in the initial, transi-
A "un~ " ~ 1 "
~,,~--,,,,;~ REC
~a
t "cone¢lo~" REC "coe~ous" REO
ca~l~tion
"conlc~oui" REO
amau~
"cor~¢io~" REC
¢eed~en
Fig. 4. Illustrative model of availability and facilitation of neural networks in dementia (adapted in part from Pfurtscheller, 1992). (A) Normal subjects. Neuronal availability, considered as a condition of inactivity and, consequently, of possible recruitment, gradually decreases from the unconscious state (lowest activity) through the rest eyes closed condition down to the eyes open conscious state; in the latter the lowest level of availability is reached because the majority of neuronal networks are recruited. On the contrary, facilitation, which can be thought of as an excitation state of the thalamo-cortical circuits during the arousal and attentive conditions, increases from the minimum in the unconscious state to the maximum in the eyes open conscious state. At the point of intersection of the two curves, that is, in the condition of wakefulness with eyes dosed, there is the greatest equilibrium between neuronal availability and facilitation to which the maximal alpha expression corresponds (see the scheme in the bottom left of the figure). Really, in normal physiology, it is the thalamo-cortical facilitation which brings about the neuronal recruitment in specific scanning rhythms, taking the neurones off the availability area. (B) Dementia patients. The gentler availability curve corresponds to the reduction in neuronal population and their interconnections in dementia. A decreased facilitation, resulting from an alteration (neurotransmissional and/or lesional) of the thaiamo-cortical circuits, means the surviving neuronal pool is hardly recruited and, consequently, stays longer in the availability area even during the eyes open condition. As a consequence of this, the profile of alpha expression in the different conditions changes as shown in the bottom right of the figure.
258
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245 -260
t i o n a l or m i x e d forms. H o w e v e r , a s t u d y o f the i n d i v i d u a l predictive v a l u e o f the PIs b y a s u i t a b l e statistical m e t h o d is still necessary.
Acknowledgements T h i s research has b e e n f o u n d e d with a g r a n t f r o m the C N R ( I t a l i a n N a t i o n a l Research Council), Progetto Finalizzato Invecchiamento, Contract 93.00.35PF40. T h e a u t h o r s wish to t h a n k Prof. D u i l i o G i a n n i t r a p a n i , V e t e r a n s A d m i n i s t r a tion, M e d i c a l C e n t e r , Perry P o i n t , M D , U S A for his c o n t r i b u t i o n to setting u p the research a n d S. Acciarri, S. Cercaci, V. D u r a z z i a n d E. Pistoli for the technical assistance. F i n a l l y a c k n o w l e d g e m e n t s are d u e to Dr. C r i s t i n a Sirolla for her h e l p in p r o c e s s i n g the statistical analysts.
References American Psychiatric Association (1987): Diagnostic and Statistical Manual of Mental Disorders (DSM-III), 3rd Edition, revised, American Psychiatric Press, Washington, DC. Angeleri, F., Provinciali, L., Scarpino O. and Signorino, M. (1982): EEG spectral analysis, P300 and CNV in bilateral parietotemporal areas during single and contemporary intermittent verbal task: a clinical approach. Res. Commun. Psychol. Psychiatr. Behav., 7, 85 96. Angeleri, F. and Signorino, M. (1989): EEG computerizzato, brain mapping e attivit',i mentale. In: Neurologia e Scienze di Base. Scritti in Onore di Giorgio Macchi. Vita e Pensiero, Milano. Editors: G. Gainotti, G., Bentivoglio, M., Bergonzi, P. and F. Ferro Milone. (in Italian). Bajalan, A.A.A., Wright, C.E. and Van Der Vliet, V.J. (1986): Changes in the human visual evoked potential caused by the anticholinergic agent hyoscine hydrobromide: comparison with results in Alzheimer's disease. J. Neurol. Neurosurg. Psychiatr., 49, 175-182. Beach, T.G., Walker, R. and McGeer, E.G. (1989): Lamina-selective A68 immunoreactivity in primary visual cortex of Alzheimer's disease patients. Brain Res., 501, 171-174. Bell, M.A. and Ball, M~J. (1990): Neuritic plaques and vessels of visual cortex in aging and Alzheimer's dementia. Neurobiol. Aging, 11, 359 370. Bendat, J.S. and Piersol, A.P. (1986): Random Data. Analysis and Measurement Procedures. 2nd Edn. Wiley Interscience Press, New York. Brenner, R.P., Ulrich, R., Spiker, D.G., Sclabassi, R.J., Reynolds, C.F., Marin, R.S. and Boller, F. (1986): Computerized EEG spectral analysis in elderly normal, demented and depressed subjects. Electroenceph. Clin. Neurophysiol., 64, 483 492. Coburn, K.L., Ashford, J.W. and Moreno, M.A. (1989): VEP latency changes in Alzheimer's disease. In: I. Raz (Editor) Neurological and psychiatric applications on topographic brain mapping of EEG and evoked potentials, pp. : 1 7. Bio-Logic, Chicago. Folstein, F.M., Folstein, S.E. and McHugh, P.R. (1975): Mini-Mental State. A practical method for grading the cognitive state of patients for the clinician. J. Psychiatr. Res., 12, 189-198. Friedland, R.P. (1993): Alzheimer's disease: clinical features and differential diagnosis. Neurology, 43 (suppl. 4), $45-$51. Gasser, T., Bacher, P. and Mocks, J. (1982): Translbrmation towards the normal distribution of broad band spectral parameters of EEG. Electroenceph. Clin. Neurophysiol., 53, 119-124. Giannitrapani, D. (1991): In: F. Angeleri (Editor) Analisi spettrale delI'EEG e funzioni mentali. Monduzzi, Bologna.
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
259
Giannitrapani, D., Collins, J. and Vassiliadis, D. (1991): The EEG spectra of Alzheimer's disease. Int. J. Psychophysiol., 10, 259-269. Giaquinto, S. and Nolfe, G. (1986): The EEG in the normal elderly: a contribution to the interpretation of aging and dementia. Electroenceph. Clin. Neurophysiol. 63, 540-546. Gordon, E.B. and Sim, M. (1967): The EEG in presenile dementia. J. Neurol. Neurosurg. Psychiatr., 30, 285-291. Gotman, J. (1990): The Use of Computers in Analysis and Display of EEG and Evoked Potentials. In: Current Practice of Clinical Electroencephalography. Second Edition, pp. 51-83. Editors: D.D. Daly and T.A. Pedley. Raven Press, New York. Hof, P.R. and Morrison, J.H. (1990): Quantitative Analysis of a Vulnerable Subset of Piramidal Neurons in Alzheimer's Disease: II. Primary and Secondary Visual Cortex. J. Comp. Neurol., 301, 55-64. John, E.R., Ahn, H., Prichep, L., Trepetin, M., Brown, D. and Kaye, H. (1980): Developmental equations for the electroencephalogram. Science, 210, 1255-1258. Letemendia, F. and Pampiglione, G. (1958): Clinical and electroencephalographic observations in Alzheimer's disease. J. NeuroL Neurosurg. Psychiatr., 21, 167-172. Leuchter, A.F., Newton, T.F., Cook, I.A., Walter, D.O., Rosenberg-Thompson, S. and Lachenbruch, P.A. (1992): Changes in brain functional connectivity in Alzheimer's-type and multi-infarct dementia. Brain, 115, 1543-1561. Leuchter, A.F., Cook, I.A., Newton, T.F., Dunkin, J., Walter, D.O., Rosenberg-Thompson, S., Lachenbruch, P.A. and Weiner, H. (1993): Regional differences in brain electrical activity in dementia: use of spectral power and spectral ratio measures. Electroenceph. Clin. Neurophysiol., 87, 385-393. Markand, O.N. (1990): Alpha rhythms. J. Clin. Neurophysiol., 7, 163-189. Maurer, K. and Dierks, T. (1992): Functional imaging procedures in dementia: mapping of EEG and evoked potentials. Acta Neurol. Scand., suppl. 139, 40-46. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D. and Standlan, E.M. (1984): 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, 34, 939-944. Miyauchi, T., Hagimoto, H., Ishii, M., Endo, S., Tanaka, K., Kajiwara, S., Endo, K., Kajiwara, A. and Kosaka, K. (1994): Quantitative EEG in patients with pre-senile and senile dementia of the Alzheimer type. Acta Neurol. Scand., 89, 56-64. Mizutani, T., Amano, N., Sasaki, H., Morimatsu, Y., Mori, H., Yoshimura, M., Yamanouchi, H., Hayakawa, K. and Shimada, H. (1990): Senile dementia of Alzheimer type characterised by laminar neuronal loss exclusively in the hippocampus, parahippocampus and medial occipitotemporal cortex. Acta Neuropathol., 80, 575-580. Oken, B.S. and Chiappa, K.H. (1988): Short-term variability in EEG frequency analysis. Electroenceph. Clin. Neurophysiol., 69, 191-198. Old_field, R.C. (1971): The assessment and analysis of handness: the Edimburgh Inventory. Neuropsychologia, 9, 97-113. Penttil~i, M., Partanen, V,J., Soininen, H. and Riekkinen, P.J. (1985): Quantitative analysis of occipital EEG in different stages of Alzheimer's disease. Electroenceph. Clin. Neurophysiol., 60, 1-6. Pfurtscheller, G. (1992): Event-related synchronization (ERS): an electrophysiological correlate of cortical areas at rest Electroenceph. Clin. Neurophysiol., 83, 62-69. Politoff, A.L., Monson, N., Hass, P. and Stadter, R. (1992): Decreased alpha bandwidth responsiveness to photic driving in Alzheimer disease. Electroenceph. Clin. Neurophysiol., 82, 45-52. Primavera, A., NoveUo, P., Finocchi, C., Canevari, E. and Corsello, L. (1990): Correlation between Mini-Mental State Examination and quantitative electroencephalography in senile dementia of Alzheimer type. Neuropsychobiology, 23, 74-78. Pritchard, W.S., Duke, W.D., Coburn, K.L., Moore, N.C., Tucker, K.A., Jann, M.W. and Hostetler, R.M. (1994): EEG-based, neural-net predictive classification of Alzheimer's disease versus control subjects is augmented by non linear EEG measures. Electroenceph. Clin. Neurophysiol., 91, 118-130.
260
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245 260
Reisberg, B., Ferris, S.H., De Leon, M.J. and Crook, T. (1982): The Global Deterioration Scale for Assessment of Primary Degenerative Dementia. Am. J. Psych., 139, 1136 1139. Roman, C.G., Tatemichi, T.K., Erkinjutti, T., Cummings, J.L., Masdeu, J.C., Garcia, J.H., Amaducci, L., Orgogozo, J.M. Brun, A., Hofman, A., Moody, D.M., O'Brien, M.D., Yamaguchi, T., Grafman, J., Drayer, B.P., Bennett, D.A., Fisher, M., Ogata, J., Kokmen, E., Bermejo, F., Wolf, P.A., Gorelick, P.B., Bick, K.L., Pajeau, A.K., Bell, M.A., Phil, D., De Carli, C., Gulebras, A., Korczyn, A.D., Bogousslavsky, J., Hartmann, A. and Scheinberg P. (1993): Vascular Dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology, 43, 250 260. Schreiter-Gasser, U., Gasser, T. and Ziegler, P. (1993): Quantitative EEG analysis in early onset Alzheimer's disease: a controlled study. Electroenceph. Clin. Neurophysiol., 86, 15-22. Signorino, M., Pucci, E., Belardinelli, N., Nolfe, G. and Angeleri, F. (1995): EEG Spectral Analysis in Vascular and Alzheimer Dementia. Electroenceph. Clin. Neurophysiol., 94, 313-325. Sloan, E.P. and Fenton, G.W. (1993): EEG power spectra and cognitive change in geriatric psychiatry: a longitudinal study. Electroenceph. Clin. Neurophysiol., 86, 361-367. Soininen, H., Partanen, J., Laulumaa, V., Helkala, E.L., Laakso, M. and Riekkinen, P.J. (1989): Longitudinal EEG spectral analysis in early stage of Alzheimer's disease. Electroenceph. Clin. Neurophysiol., 72, 290-297. Soininen, H. and Riekkinen, P.J. (1992): EEG in diagnostic and follow-up of Alzheimer's disease. Acta Neurol. Scand., 139, 36-39. Speckman, E.J. and Caspers, H. (1973): Neurophysiologische Grundlagen der Provokationsmethoden in der Elektroenezphalographie. Z. EEG-EMG, 4, 157-167. Steriade, M., Gloor, P., Llinas, R.R., Lopes da Silva, F.H. and Mesulam, M.M. (1990): Basic mechanism of cerebral rhythmic activities. Electroenceph. Clin. Neurophysiol., 76, 481-508. Van der Drift, J.H.A. and Kok, N.K.D. (1972): The EEG in cerebrovascular disorders in relation to pathology. In: A. Remond (Editor). The Handbook of Electroencephalography and Clinical Neurophysiology. Vol.14, pp. 12-64. Elsevier, Amsterdam. Visser, S.L., Stare, F.C., Vantilburg, W., Opden Velde, W., Blom, J.L. and De Rijke, W. (1976): Visual evoked response in senile and presenile dementia. Electroenceph. Clin. Neurophysiol., 40, 385-392. Visser, S.L., Van Tilburg, W., Hoijer, C., Jonker, C. and De Rijke, W. (1985): 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., 60, 115 121. Wright, C.E., Harding, G.F.A. and Orwin, A. (1984): Presenile dementia - the use of the flash and pattern VEP in diagnosis. Electroenceph. Clin. Neurophysiol., 57, 405-415. Wright, C.E., Harding, G.F.A. and Orwin, A. (1986): The flash and pattern VEP as a diagnostic indicator of dementia. Documenta Ophthalmologica, 62, 89-96.
Archives of Gerontology and Geriatrics 22 (1996) 245-260
ARCHIVES OF GERONTOLOGY AND GERIATRICS
An EEG power index (eyes open vs. eyes closed) to differentiate Alzheimer's from vascular dementia and healthy ageing Mario Signorino a,*, Enrico Brizioli b, Loredana Amadio b, Natascia BelardineUP, Eugenio Pucci a, Franco Angeleri a "Institute for Nervous System Diseases, University of Ancona, Ospedale Regionale di Torrette di Ancona, Ancona, Italy bLN.R.C.A. - Italian National Research Centres on Ageing - Department of Research, Ancona, ltaly
Received 5 October 1995; revised 20 December 1995; accepted 23 December 1995
Abstract EEG differential power patterns between Alzheimer's (AD, 50 patients) and vascular (VaD, 37 patients) dementia and between these two and 36 healthy ageing subjects, were studied in the 6.5-12 Hz band of the ongoing EEG recorded during the rest eyes closed (REC) and eyes open (REO) conditions. From the EEGs (16 electrodes, 10-20 international system except for Fz, Cz, Pz), a 6.5-12 Hz band, wider than the alpha range (alpha-like), was chosen and processed to include the highest theta frequencies characterising the occipital dominant activity in dementia. A global and occipital EEG Power Index (PI) was calculated and used considering the absolute powers during REC and REO. The MANOVA was used to compare the figures. Beating in mind that the higher the PI value the greater the difference between the 6.5-12 Hz EEG band powers of REC vs. REO, the results were as follows: (i) in the patients with Alzheimer's and vascular dementia the global and occipital PIs were significantly lower than those in controls; (ii) in the patients with Alzheimer's dementia the same PIs were significantly lower that those of the patients with VaD; (iii) healthy elderly subjects showed significantly lower powers in the 6.5-12 Hz frequencies at T5 and O1 in REO as compared to dementia patients. The pathophysiological implications and the clinical applications of these results are discussed. Keywords: Alzheimer's disease; Vascular dementia; Ageing; Alpha-rhythms; EEG spectral
analysis
* Corresponding author, Tel.: + 39 71 888989; Fax: + 39 71 888989. 0167-4943/96/$15.00 © 1996 Elsevier Science Ireland Ltd. All fights reserved PII SO167-4943(96)00697-8
246
M. Signorino el al.
Arch. Gerontol. Geriatr. 22 (1996) 245 260
1. Introduction
Using both traditional EEG and its computerised quantitative analysis a reduction in alpha rhythm was found in patients with Alzheimer's disease (AD) (Letemendia and Pampiglione, 1958; Gordon and Sire, 1967; Penttil/i et al., 1985; Brenner et al., 1986; Giaquinto and Nolfe, 1986; Giannitrapani et al., 1991; Maurer and Dierks, 1992: Sloan and Fenton, 1993; Miyauchi et al., 1994). The disease also induces a decrease in alpha responsiveness to photic stimulation according to Politoff et al. (1992) and a fall in the desynchronizing effect on the dominant alpha-like ongoing activity normally induced by a static condition such as rest with eyes open (Pritchard et al., 1994; Signorino et al., 1995). These findings suggest that in AD the cortical mechanisms generating the alpha rhythm are altered, probably because of the lesions in the occipital neocortex where the alpha generator is thought to be located (Steriade et al., 1990). On the contrary in vascular dementia patients (VaD) the alpha rhythm is usually preserved although with a slower frequency than in normal elderly controls (Van der Drift and Kok, 1972). In conclusion, the alpha pattern and its 'reactivity' in AD, VaD and elderly normal subjects seem to be important for clarifying the different pathophysiological mechanisms of the ageing brain and the two types of possibly overlapping dementia (Friedland, 1993). On the other hand, for clinical application, we thought it would be interesting to characterise any EEG spectral patterns which differentiate AD from VaD and these types of dementia from healthy ageing. On the basis of this background, the aim of this study was as follows: (i) to identify the sub-alpha components, which often represent the dominant occipital rhythm not only in dementia patients (Giannitrapani, 1991) but also in healthy elderly subjects, according to some authors (review in Steriade et al., 1990) by analysing a wider band than the traditional alpha range (i.e. 8-12 Hz) towards the low frequencies (6.5-12 Hz); (ii) to compare the EEG ongoing power during the rest eyes closed and eyes open conditions not only at the occipital level but also at the other locations covering the cortical areas of the two hemispheres.
2. Materials and methods
2. I. Subjects Thirty-seven patients affected by vascular and 50 patients by Alzheimer's dementia were studied. The diagnoses were reached according to DSM-III-R (American Psychiatric Association, 1987), NINCDS-ADRDA (McKhann et al., 1984) and
50 31
AD VaD
*Mann-Whitney
n
Diagnosis
31 10
F
68.5 71.3
8.2 1.4
48-81 51-84
36 31.5
21.6 28.3
S.D.
Mean
Range
Mean
S.D.
Duration of the disease (months)
Age (years)
U-test P i 0.01.
19 27
M
Table 1 Demographic and clinical characteristics of the dementia patients
6-156 6-120
Range 13.3 17.6’
Mean
MMSE
6.1 5.1
S.D.
O-26 O-26
Range
4.6 4’
Mean
GDS
1 1
S.D.
3-6 3-6
Range
3.5 2.9:
Mean
GS
1.2 1
S.D.
2-l 2-5
Range
248
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
N I N D S - A I R E N (Roman et al., 1993). Clinical characteristics of the two groups of patients are summarised in Table 1. Thirty-six normal elderly subjects (NE) were also studied as age-matched controls (14 males, 22 females; mean age 64.4 years, S.D. 4.6; range 60-80 years). Oral informed consent was obtained from all subjects or from tutors of patients with severe dementia. All subjects except one were right-handed according to the Edinburgh Inventory (Oldfield, 1971). The severity of dementia was assessed according to the MMSE (Folstein et al., 1975), the Global Deterioration Scale (GDS; Reisberg et al., 1982) and Giannitrapani's scale (GS; Giannitrapani et al., 1991). Brain MRI was available for 59 patients and CT for the remaining patients. As the sample studied was the same as in a previous study, more details can be drawn from Signorino et al. (1995). 2.2. E E G recording and spectral analysis
Recording took place in a room with a low constant level of artificial lighting and environmental noise. The subjects were seated in a comfortable padded chair and asked to close their eyes when requested and to try to stay awake and relaxed. All EEGs were recorded between 9 a.m. and 11 a.m. after a normal night's sleep and the usual breakfast. Subjects underwent 5 min of E E G recordings at rest with eyes closed (REC) and then again with eyes open (REO). Special care was taken to prevent drowsiness and minimise artefact acquisition. The E E G was recorded by Ag/AgC1 cup electrodes at 16 monopolar locations arranged according to the international 10-20 system which explored the left scalp side (Fpl, F7, T3, T5, O1, F3, C3, P3) and the right scalp side (Fp2, F8, T4, T6, 02, F4, C4, P4) with linked earlobes as reference. Impedance was kept lower than 5 KOhm. EOG was monitored using two additional electrodes placed 1 cm below and 1 cm above the outer canthus of the eye. The bandpass filter was set in a range of 1 64 Hz. Q-notch filters at frequencies of 50 and 100 Hz were also used. Each channel was calibrated in microvolts and the DC level corrected using appropriate algorithms, including subtraction of the mean DC offset from each epoch selected. The signals were sampled at a rate of 128 samples/sec, with 12-bit analog-to-digital conversion. For each physiological condition, an expert neurologist selected 20 E E G artefact-free epochs of 2 sec duration. The Fast Fourier Transforms (FFT) of the selected epochs were calculated using the F F T algorithm, and the spectral density function W(f) was estimated as the average of these transforms (Bendat and Piersol, 1986). In order to reduce the bias error of the sample spectrum the E E G epochs were analysed using the Hanning data window. Absolute (AP) and relative (RP) powers were calculated for the 6.5-12 Hz band. A logarithmic transformation was applied to absolute powers (In x) and relative powers (ln Ix/(1 - x)]) for each band in each epoch to normalise the data, in accordance with John et al. (1980), Gasser et al. (1982), Oken and Chiappa (1988). A global spectrum was also calculated consisting of the mean of 16-channel spectra. For more technical details, see Angeleri and Signorino (1989) and Signorino et al. (1995).
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
249
For each subject and for each derivation, a Power Index (PI) was calculated as"
LAPREC - LAPREO x 100/LAPREC where LAPREC is the log of absolute power during REC and LAPREO is the log of absolute power during REO of the 6.5-12 Hz band. The PI was used in order to decrease the variability of power values, particularly absolute ones, in the subjects. Moreover, to decrease the number of variables to be analysed, PI values were pooled in two parameters: (i) the global PI calculated as the mean of PIs of all 16 derivations studied; (ii) the occipital PI, namely the mean of PI values at 01 and 02.
2.3. Statistical analysis The Mann-Whitney U test was performed to compare the clinical data. The Bonferroni-adjusted ~* = ot/k was employed when k simultaneous tests were necessary to evaluate differences between the experimental groups. To evaluate the differences in the variables in the brain electrical activity in the three groups under examination (AD, VaD, NE), under two different conditions (REC and REO) and for each of the electrodes, a three-way analysis of variance for repeated measures was used. The factors considered for each variable were: group/condition~electrode, using the MANOVA of the SPSS/PC package. The significant interactions were post hoc processed by the Tukey test. The significance level chosen was 5%. The Spearman rank order correlation (one-tailed; ~ = 0.01; CSS-2.1 software) was used to correlate GDS, GS and MMSE scores with power indices.
3. Results
3.1. L A P and L R P values of the 6.5-12 Hz band." comparison for groups, conditions and electrodes The results of the comparison of absolute (LAP) and relative power (LRP) values in the groups (NE, VaD, AD) in the two experimental conditions (i.e., REO and REC) at each derivation are summarised in Fig. 1. As we can see, AD patients exhibit a trend, in both REC and REO conditions, which is quite similar to that of the controls during the REC condition. However controls have a selective REO decrease of activity in T5 and 01 in the left hemisphere, showing a hemispheric asymmetry not evident in the dementia patients (P < 0.05). As shown in Table 2 the results of MANOVA reveal statistically significant differences for group factors in all variables, including the second and third level interactions. The significant interaction condition x electrode may be mainly due
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
250
h i m e ~ l p oide
lee : o i I p slde
ViJliai S.S I
~-.
-'/~
/~"
j
/ /
/
4,/ '~t-
--,',. //
. Q. ~ _& x
"~
o
Lx
o
//x
//
\,~/t
/
. ~.
~\ ~___,,,___~"/*-~,,_~..~'~ "
,a___..-"~~~
"
0" -.
I,S
•
FT~
,e. ......... e
"
• ...........~
"
I
I
!
~
I
I
1:8
11"4
331
O~
1=4
C4
I
~
I
114 FPI
"
~...........,0
I
1=7
. 4P
"~
I
l
I
I
I
TI
T~
O1
F8
(~
~ m ~ e
I
I~1
~ m ~ e
mmkme -1.T -1.S ~
.
-1.$ -1.1
~,t
"'IL\
\ "
0.8
/
\x.
[email protected]
~ FI
""4 ~\
/-
J \.
I
I
4
I
I
~
TI
02
F4
CA
P4
---&--¥10-iqEO
=
~"
\
:,,
T4
ADJIC
\
..l
t --
FP1 ~I)4EO
• ~... N E - / I E C
I
I
Fir
T3 - --&-
\4,..
I T$
/
"
/
'
\\
t
t
O1
FI
1 CS
I
Pll
VaI~REC
..... 4,-.-NE-FIEO
Fig. 1. Log mean values of absolute (upper part) and relative powers (lower part) for groups, conditions and electrodes. The locations of recording electrodes are shown on the abscissa axis. They are grouped according to whether they belong to the left or right scalp side. On the ordinate axis the corresponding mean values of the log powers of the 6.5-12 Hz band. The geometric symbols and the lines refer to the groups and conditions: white squares and solid lines indicate the AD group in the eyes closed condition (AD-REC), black squares and solid lines indicate the AD group in the eyes open condition (AD-REO); white triangles and broken lines indicate the VaD group in the eyes closed condition (VaD-REC), and so on as shown in the figure. Note that the contour lines are quite symmetrical when comparing left and right sides; this occurs in all groups and conditions except those of normal subjects in eyes open condition (NE-REO). This line, in fact, shows lower values in the left side (Fpl, F7, T3, T5, O1, F3, C3) with respect to the right electrode positions. The difference is statistically significant in T5 and O l. Furthermore, we can see that the VaD group is characterised by the highest absolute power values in the eyes closed condition with respect to the other groups while in the same condition the relative powers of the VaD and NE groups tend to overlap.
251
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
Table 2 Repeated measures analysis of variance for the brain electrical activity Source of variation
Group Condition Electrodes Group x Electrodes Group × Condition Electrodes x Condition Group × Electrodes x Condition
LAP
LRP
F
D.f.
P value
F
D.f.
P value
11.32 108.65 57.09 1.83 22.39 32.50 3.28
2 1 15 30 2 15 30
<0.001 <0.001 <0.001 0.004 <0.001 <0.001 <0.001
14.74 144.59 112.81 1.85 14.37 7.34 1.75
2 1 15 30 2 15 30
<0.001 <0.001 <0.001 0.004 <0.001 <0.001 0.007
to the selective decrease o f L A P value between the R E C and the R E O in the controls. 3.2. L A P and L R P mean values o f the 6 . 5 - 1 2 H z band in R E C and R E O conditions 3.2.1. L A P (Fig. 2): the m e a n L A P values o f the elderly controls at all electrodes in the R E C condition is 4.46 + 0.51 (range = 3.21-5.61). In the R E O condition, this value decreases to 3.63 + 0.44 (range = 2.80-4.81). The difference is statistically significant (Tukey test = P < 0.05). V a D patients show statistically significant higher values in R E C . In fact the m e a n o f all the electrodes is 4.89 ___ 0.49 (range = 3.68-6.69) during R E C and 4.49 _ 0.46 (range = 3.63-6.17) during R E O . A D patients show significantly lower L A P values during R E C c o m p a r e d to the other groups, with a m e a n value o f 3.99 + 0.47 (range = 2.10-5.23) which is n o t statistically different f r o m the value during R E O with a m e a n value o f 3.86 _ 0.41; (range = 3.01-9.45). 3.2.2. L R P (Fig. 3): controls and V a D patients show similar values during R E C , while A D patients show a significantly lower value. I n the R E O condition, only the control g r o u p shows a statistically significant decrease in activity in respect to REC. 3.3. P o w e r indices
The results o f the c o m p a r i s o n o f PI values m a y be summarised as follows: in N E the global P I is 16.58 + 12.39 while the occipital PI value is greater (22.50
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
252
+ 14.38); in AD patients the global PI and the occipital PI are respectively 2.09 __+ 7.84 and 3.64 ___ 11.96; in VaD patients the global PI is 7.79 ___ 7.51 but the occipital PI is higher (12.38 + 12.34). The pooled PI values provide statistically significant differences in the 3 groups capable of discriminating not only dementia patients from normal subjects but also AD from VaD (Table 3). Neither the global nor the occipital PI correlates with MMSE, GDS and GS scores in either AD (Table 4) or VaD (Table 5).
LAP moon values 5.1 4.0 4.7 4.6
*'-.. "'...
4.8 4.1
"-.. |.It
"'""'--.... 8.7
°. -.$ I
|,S I
TulloylilvlilploCempmioonToot: N E . I m C
vii N E . f l E O p c 0 . 0 g
N D . N O C vs V m D . I m C p , c 0 . 0 S
Fig. 2. Graph of the differences in the mean values of the log of absolute powers (mean of all electrodes and of all subjects in each group) shown on the y axis, between eyes closed condition (left extremity of the x axis) and eyes open condition (right extremity of the x axis). The continuous line with squares refers to the AD group, the broken line with triangles refers to the VaD group, the small points line with rhombus refers to NE groups. The difference between eyes closed and open conditions is significant in the NE group. Furthermore, one can note a significant difference between the values of the VaD and AD groups in the eyes open condition.
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
253
I.ltP -1.2
.......41 -1
..4, . .....°'"'*"'
I 41.1
41,1
41.4
41.z
I
--.--v.°
TW~MullipleC,Onlmlm~Toal:
I
.... . .
NII-ImC va llE-PmOpcO.M AD.PmC vm WD-RnCpcO.06 AD-ImC vs N E . R ~ p c 0 . U
Fig. 3. Graph of the difference in the mean values of the log transformed relative powers (mean of all electrodes and of all subjects in each group) shown on the y axis, between eyes closed and open conditions (respectively left and fight extremity of the x axis). For further details see Fig. 2.
Table 3 Power Index (PI) Global PI
AD VaD NE
Occipital PI
Mean
S.D.
2.09 7.79 16.58
7.84 7.51 12.39
Range - 13.3 -7.2 -13.3
26.8 27.0 38.9
Mean
S.D.
3.64 12.38 22.50
11.96 12.34 14.38
Tukey multiple comparison test: all groups different. P < 0.05.
Range - 18.1 - 18.3 -11.8
38.3 43.5 51.8
M. Signorino et ell. : Arch. Gerontol. Geriatr. 22 11996) 245 260
254
Table 4 S p e a r m a n r a n k order correlations in A D (:~ = 0.01) R
t
P
Global PI vs. MMSE GDS GS
0.177 0.317 0.008
1.249 -2.319 - 0.054
0.218 0.025 0.957
0.009 0.088 11.079
0.060 0.612 1/.547
0.952 0.543 0.586
Occipital PI v,s. M M SE GDS GS
Table 5 S p e a r m a n r a n k order correlations in V a D (:~ = 1).01)
R
:
P
Global P I ~:s. MMSE GDS GS
Occipital PI
O.245 0.21)7 - 0.404
.495 - 1,249 2.616
0.144 0.220 0.013
0.169 0.170 0.363
1.016 1.022 2.304
0.316 0.314 0.027
v,~.
MMSE GDS GS
4. Discussion Before discussing the results, three points must be made. Firstly, the power indices (PIs) which were used to process the powers of the 6.5 12 Hz band in the two rest conditions cannot be completely identified with the traditional significance of EEG desynchronization; however, such indices evaluate two basic components of desynchronization, i.e. the occurrence and/or amplitude attenuation of the ongoing EEG frequencies in the 6.5 12 Hz band (Gotman, 1990). Secondly, this band, which is wider than the alpha frequency range, was considered better suited to studying the highest theta frequencies which, sometimes with the alpha rhythm, characterise the dominant activity in patients with dementia (review in Leuchter et al., 1993); briefly this band was defined as 'alpha-like rhythms'. Thirdly, the comparison of the two types of dementia and healthy ageing was carried out by the log transformation of both absolute and relative power values to obtain parametric statistics more reliably (John et al., 1980; Gasser et al., 1982; Oken and Chiappa, 1988; Leuchter et al., 1993; Schreiter-Gasser et al., 1993).
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
255
Focusing the analysis on the alpha-like band, a progressive decrease in the global and occipital PIs was found when comparing age-matched healthy subjects and patients with vascular and Alzheimer's dementia. In the occipital regions the decrease differences were the greatest thus suggesting that the reduction in the alpha-like reactivity in the two types of dementia has its core in the occipital regions where a consistent alpha generator component is thought to be located (reviews in Markand, 1990 and Steriade et al., 1990). The statistically significant PI decreases from healthy subjects to AD patients with the intermediate mean values of VaD gives more exhaustive evidence of previous observations in AD patients, e.g. the reduction of decrease in alpha-like powers during the REO condition that they show (Pritchard et al., 1994; Signorino et al., 1995). The same findings agree with the observations of several authors (Visser et al., 1976; Wright et al., 1984; Visser et al., 1985; Wright et al., 1986; Bajalan et al., 1986; Coburn et al., 1989) on the loss of the flash-VEP induced alpha desynchronization in AD patients and with the investigation by Politoff et al. (1992) of the lack of alpha cortical photic driving effect in the same patients. Furthermore, a minor negative effect on VaD of the alpha-like reactivity is established by the present study which found, however, statistically significant differences between the PI mean values of this group and those of the elderly subjects. The use of the 6.5-12 Hz alpha-like band in the EEG spectral analysis allowed the present study to investigate both the alpha and highest theta frequencies where recruiting and oscillating are like a single dominant activity. The results obtained by the spectral analysis suggest that the generator of alpha rhythm extends its recruiting oscillations towards the highest theta frequencies which in part or totally, replace alpha rhythm in dementia (review in Leuchter et al., 1993). Finally, the significant lower powers of the 6.5-12 Hz frequencies at the T5 and O1 derivations found in the elderly healthy subjects during the REO condition as compared with the analogous mean powers of the patients with dementia could suggest a loss of the EEG hemispheric asymmetry in the latter as observed several years ago by Angeleri et al. (1982). If the PI drop reflects, at least in part, the ongoing EEG desynchronization with its focal point in the occipital regions, it can be considered as an indicator of the efficiency of the neural networks involved in the mechanisms of vigilance and attention. Healthy ageing subjects, but even more so brain damaged dementia patients, can have alteration in these networks and their oscillations and therefore the functions of vigilance and attention. In AD the neural mechanisms ruling the alpha-like power patterns are definitely altered and the phenomenon can easily be attributed to the neurotransmitter deficit and neuronal lesions of the neocortex induced by the disease. In this context the neuropathological works that underline the involvement of the occipital areas must be quoted (Beach et al., 1989; Bell and Ball, 1990; Mizutani et al., 1990; Hof and Morrison, 1990). Also, the hypothesis which maintains that the deficit of the cortical projecting cholinergic system is associated with an intracortical disconnec-
256
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
tion seems worth mentioning (Soininen et al., 1989; Leuchter et al., 1992; Soininen and Riekkinen, 1992). The PI mean values in VaD, which are in the middle position of the values between AD and healthy ageing, can be explained by the characteristics of the vascular lesions and at the same time could justify the slower and different evolution of the mental deterioration occurring in these patients. Some authors consider, in fact, that the mental deficit in VaD is caused by an intracortical and cortical-subcortical deafferentation like that in AD patients, but induced by the typical vascular lesions of the hemispheric white matter instead of the drop in the cortical cholinergic system associated with the loss of a great quantity of pyramidal neurones which occurs in AD (review in Leuchter et al., 1992). The mental deterioration was more severe in AD than VaD patients, however, Spearman rank order correlation showed no statistically significant correlation between either the global or occipital PIs and AD or VaD dementia scores. These findings fit with those acquired by Giannitrapani et al. (1991) in dementia patients who showed no relation between the alpha loss and severity of cognitive impairment, but a positive correlation between the former and the degree of the cortical lesions. On the other hand, the previous works that report positive correlation between cognitive performances and changes in alpha activity are concerned with the rest eyes closed condition only (Penttil~i et al., 1985; Brenner et al., 1986; Primavera et al., 1990). The abnormal behaviour of the dominant activity in the 6.5-12 Hz band, as found in patients with dementia, may be explained on the basis of a model based on availability and facilitation of neural networks generating alpha activity in the ongoing EEG (Fig. 4). The model, initially proposed by Speckman and Caspers (1973), was recently developed by Pfurtscheller (1992). According to these authors, the more the vigilance increases, the more the thalamo-cortical facilitation increases and the neuronal availability decreases. Vice-versa, the more the vigilance decreases the more the availability of neural networks grows as long they are not involved in cognitive and sensorial processes. At the point of intersection of neuronal availability and facilitation, that is, in the condition of wakefulness with eyes closed, there is a greater equilibrium between the conditions of wakefulness and psychosensorial rest. This corresponds to the maximum expression of alpha rhythm as the dominant activity in the ongoing EEG. Following this model to interpret the changes induced by dementia a reduction, both in the availability and facilitation mechanisms, may be supposed and such a reduction is certainly greater in Alzheimer's disease. In the graph in Fig. 4 this reduction corresponds to a gentler curve in the neuronal availability and a flattening of the slope line of facilitation. The intersection of the two lines shows an enlargement and smoothing of the recruiting area of the background activity which is shifted towards the REO condition. The availability and facilitation changes induced by dementia probably, in their turn, follow the alterations in the strictly interrelated intracortical and cortical
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
257
sub-cortical n e u r o n a l circuits ruling the recruiting oscillation o f the alpha-like rhythms. I n this context the PIs m a y be considered a detector o f the n o r m a l a n d a b n o r m a l v a r i a t i o n s o f these n e u r a l functions. F o r diagnostic purposes it m a y be used to differentiate healthy ageing from A D a n d the latter from VaD. It m a y also help in the differential diagnosis o f single cases where the n o r m a l or pathological definition or diagnosis o f A D or V a D are u n c e r t a i n , m a i n l y in the initial, transi-
A "un~ " ~ 1 "
~,,~--,,,,;~ REC
~a
t "cone¢lo~" REC "coe~ous" REO
ca~l~tion
"conlc~oui" REO
amau~
"cor~¢io~" REC
¢eed~en
Fig. 4. Illustrative model of availability and facilitation of neural networks in dementia (adapted in part from Pfurtscheller, 1992). (A) Normal subjects. Neuronal availability, considered as a condition of inactivity and, consequently, of possible recruitment, gradually decreases from the unconscious state (lowest activity) through the rest eyes closed condition down to the eyes open conscious state; in the latter the lowest level of availability is reached because the majority of neuronal networks are recruited. On the contrary, facilitation, which can be thought of as an excitation state of the thalamo-cortical circuits during the arousal and attentive conditions, increases from the minimum in the unconscious state to the maximum in the eyes open conscious state. At the point of intersection of the two curves, that is, in the condition of wakefulness with eyes dosed, there is the greatest equilibrium between neuronal availability and facilitation to which the maximal alpha expression corresponds (see the scheme in the bottom left of the figure). Really, in normal physiology, it is the thalamo-cortical facilitation which brings about the neuronal recruitment in specific scanning rhythms, taking the neurones off the availability area. (B) Dementia patients. The gentler availability curve corresponds to the reduction in neuronal population and their interconnections in dementia. A decreased facilitation, resulting from an alteration (neurotransmissional and/or lesional) of the thaiamo-cortical circuits, means the surviving neuronal pool is hardly recruited and, consequently, stays longer in the availability area even during the eyes open condition. As a consequence of this, the profile of alpha expression in the different conditions changes as shown in the bottom right of the figure.
258
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245 -260
t i o n a l or m i x e d forms. H o w e v e r , a s t u d y o f the i n d i v i d u a l predictive v a l u e o f the PIs b y a s u i t a b l e statistical m e t h o d is still necessary.
Acknowledgements T h i s research has b e e n f o u n d e d with a g r a n t f r o m the C N R ( I t a l i a n N a t i o n a l Research Council), Progetto Finalizzato Invecchiamento, Contract 93.00.35PF40. T h e a u t h o r s wish to t h a n k Prof. D u i l i o G i a n n i t r a p a n i , V e t e r a n s A d m i n i s t r a tion, M e d i c a l C e n t e r , Perry P o i n t , M D , U S A for his c o n t r i b u t i o n to setting u p the research a n d S. Acciarri, S. Cercaci, V. D u r a z z i a n d E. Pistoli for the technical assistance. F i n a l l y a c k n o w l e d g e m e n t s are d u e to Dr. C r i s t i n a Sirolla for her h e l p in p r o c e s s i n g the statistical analysts.
References American Psychiatric Association (1987): Diagnostic and Statistical Manual of Mental Disorders (DSM-III), 3rd Edition, revised, American Psychiatric Press, Washington, DC. Angeleri, F., Provinciali, L., Scarpino O. and Signorino, M. (1982): EEG spectral analysis, P300 and CNV in bilateral parietotemporal areas during single and contemporary intermittent verbal task: a clinical approach. Res. Commun. Psychol. Psychiatr. Behav., 7, 85 96. Angeleri, F. and Signorino, M. (1989): EEG computerizzato, brain mapping e attivit',i mentale. In: Neurologia e Scienze di Base. Scritti in Onore di Giorgio Macchi. Vita e Pensiero, Milano. Editors: G. Gainotti, G., Bentivoglio, M., Bergonzi, P. and F. Ferro Milone. (in Italian). Bajalan, A.A.A., Wright, C.E. and Van Der Vliet, V.J. (1986): Changes in the human visual evoked potential caused by the anticholinergic agent hyoscine hydrobromide: comparison with results in Alzheimer's disease. J. Neurol. Neurosurg. Psychiatr., 49, 175-182. Beach, T.G., Walker, R. and McGeer, E.G. (1989): Lamina-selective A68 immunoreactivity in primary visual cortex of Alzheimer's disease patients. Brain Res., 501, 171-174. Bell, M.A. and Ball, M~J. (1990): Neuritic plaques and vessels of visual cortex in aging and Alzheimer's dementia. Neurobiol. Aging, 11, 359 370. Bendat, J.S. and Piersol, A.P. (1986): Random Data. Analysis and Measurement Procedures. 2nd Edn. Wiley Interscience Press, New York. Brenner, R.P., Ulrich, R., Spiker, D.G., Sclabassi, R.J., Reynolds, C.F., Marin, R.S. and Boller, F. (1986): Computerized EEG spectral analysis in elderly normal, demented and depressed subjects. Electroenceph. Clin. Neurophysiol., 64, 483 492. Coburn, K.L., Ashford, J.W. and Moreno, M.A. (1989): VEP latency changes in Alzheimer's disease. In: I. Raz (Editor) Neurological and psychiatric applications on topographic brain mapping of EEG and evoked potentials, pp. : 1 7. Bio-Logic, Chicago. Folstein, F.M., Folstein, S.E. and McHugh, P.R. (1975): Mini-Mental State. A practical method for grading the cognitive state of patients for the clinician. J. Psychiatr. Res., 12, 189-198. Friedland, R.P. (1993): Alzheimer's disease: clinical features and differential diagnosis. Neurology, 43 (suppl. 4), $45-$51. Gasser, T., Bacher, P. and Mocks, J. (1982): Translbrmation towards the normal distribution of broad band spectral parameters of EEG. Electroenceph. Clin. Neurophysiol., 53, 119-124. Giannitrapani, D. (1991): In: F. Angeleri (Editor) Analisi spettrale delI'EEG e funzioni mentali. Monduzzi, Bologna.
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245-260
259
Giannitrapani, D., Collins, J. and Vassiliadis, D. (1991): The EEG spectra of Alzheimer's disease. Int. J. Psychophysiol., 10, 259-269. Giaquinto, S. and Nolfe, G. (1986): The EEG in the normal elderly: a contribution to the interpretation of aging and dementia. Electroenceph. Clin. Neurophysiol. 63, 540-546. Gordon, E.B. and Sim, M. (1967): The EEG in presenile dementia. J. Neurol. Neurosurg. Psychiatr., 30, 285-291. Gotman, J. (1990): The Use of Computers in Analysis and Display of EEG and Evoked Potentials. In: Current Practice of Clinical Electroencephalography. Second Edition, pp. 51-83. Editors: D.D. Daly and T.A. Pedley. Raven Press, New York. Hof, P.R. and Morrison, J.H. (1990): Quantitative Analysis of a Vulnerable Subset of Piramidal Neurons in Alzheimer's Disease: II. Primary and Secondary Visual Cortex. J. Comp. Neurol., 301, 55-64. John, E.R., Ahn, H., Prichep, L., Trepetin, M., Brown, D. and Kaye, H. (1980): Developmental equations for the electroencephalogram. Science, 210, 1255-1258. Letemendia, F. and Pampiglione, G. (1958): Clinical and electroencephalographic observations in Alzheimer's disease. J. NeuroL Neurosurg. Psychiatr., 21, 167-172. Leuchter, A.F., Newton, T.F., Cook, I.A., Walter, D.O., Rosenberg-Thompson, S. and Lachenbruch, P.A. (1992): Changes in brain functional connectivity in Alzheimer's-type and multi-infarct dementia. Brain, 115, 1543-1561. Leuchter, A.F., Cook, I.A., Newton, T.F., Dunkin, J., Walter, D.O., Rosenberg-Thompson, S., Lachenbruch, P.A. and Weiner, H. (1993): Regional differences in brain electrical activity in dementia: use of spectral power and spectral ratio measures. Electroenceph. Clin. Neurophysiol., 87, 385-393. Markand, O.N. (1990): Alpha rhythms. J. Clin. Neurophysiol., 7, 163-189. Maurer, K. and Dierks, T. (1992): Functional imaging procedures in dementia: mapping of EEG and evoked potentials. Acta Neurol. Scand., suppl. 139, 40-46. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D. and Standlan, E.M. (1984): 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, 34, 939-944. Miyauchi, T., Hagimoto, H., Ishii, M., Endo, S., Tanaka, K., Kajiwara, S., Endo, K., Kajiwara, A. and Kosaka, K. (1994): Quantitative EEG in patients with pre-senile and senile dementia of the Alzheimer type. Acta Neurol. Scand., 89, 56-64. Mizutani, T., Amano, N., Sasaki, H., Morimatsu, Y., Mori, H., Yoshimura, M., Yamanouchi, H., Hayakawa, K. and Shimada, H. (1990): Senile dementia of Alzheimer type characterised by laminar neuronal loss exclusively in the hippocampus, parahippocampus and medial occipitotemporal cortex. Acta Neuropathol., 80, 575-580. Oken, B.S. and Chiappa, K.H. (1988): Short-term variability in EEG frequency analysis. Electroenceph. Clin. Neurophysiol., 69, 191-198. Old_field, R.C. (1971): The assessment and analysis of handness: the Edimburgh Inventory. Neuropsychologia, 9, 97-113. Penttil~i, M., Partanen, V,J., Soininen, H. and Riekkinen, P.J. (1985): Quantitative analysis of occipital EEG in different stages of Alzheimer's disease. Electroenceph. Clin. Neurophysiol., 60, 1-6. Pfurtscheller, G. (1992): Event-related synchronization (ERS): an electrophysiological correlate of cortical areas at rest Electroenceph. Clin. Neurophysiol., 83, 62-69. Politoff, A.L., Monson, N., Hass, P. and Stadter, R. (1992): Decreased alpha bandwidth responsiveness to photic driving in Alzheimer disease. Electroenceph. Clin. Neurophysiol., 82, 45-52. Primavera, A., NoveUo, P., Finocchi, C., Canevari, E. and Corsello, L. (1990): Correlation between Mini-Mental State Examination and quantitative electroencephalography in senile dementia of Alzheimer type. Neuropsychobiology, 23, 74-78. Pritchard, W.S., Duke, W.D., Coburn, K.L., Moore, N.C., Tucker, K.A., Jann, M.W. and Hostetler, R.M. (1994): EEG-based, neural-net predictive classification of Alzheimer's disease versus control subjects is augmented by non linear EEG measures. Electroenceph. Clin. Neurophysiol., 91, 118-130.
260
M. Signorino et al. / Arch. Gerontol. Geriatr. 22 (1996) 245 260
Reisberg, B., Ferris, S.H., De Leon, M.J. and Crook, T. (1982): The Global Deterioration Scale for Assessment of Primary Degenerative Dementia. Am. J. Psych., 139, 1136 1139. Roman, C.G., Tatemichi, T.K., Erkinjutti, T., Cummings, J.L., Masdeu, J.C., Garcia, J.H., Amaducci, L., Orgogozo, J.M. Brun, A., Hofman, A., Moody, D.M., O'Brien, M.D., Yamaguchi, T., Grafman, J., Drayer, B.P., Bennett, D.A., Fisher, M., Ogata, J., Kokmen, E., Bermejo, F., Wolf, P.A., Gorelick, P.B., Bick, K.L., Pajeau, A.K., Bell, M.A., Phil, D., De Carli, C., Gulebras, A., Korczyn, A.D., Bogousslavsky, J., Hartmann, A. and Scheinberg P. (1993): Vascular Dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology, 43, 250 260. Schreiter-Gasser, U., Gasser, T. and Ziegler, P. (1993): Quantitative EEG analysis in early onset Alzheimer's disease: a controlled study. Electroenceph. Clin. Neurophysiol., 86, 15-22. Signorino, M., Pucci, E., Belardinelli, N., Nolfe, G. and Angeleri, F. (1995): EEG Spectral Analysis in Vascular and Alzheimer Dementia. Electroenceph. Clin. Neurophysiol., 94, 313-325. Sloan, E.P. and Fenton, G.W. (1993): EEG power spectra and cognitive change in geriatric psychiatry: a longitudinal study. Electroenceph. Clin. Neurophysiol., 86, 361-367. Soininen, H., Partanen, J., Laulumaa, V., Helkala, E.L., Laakso, M. and Riekkinen, P.J. (1989): Longitudinal EEG spectral analysis in early stage of Alzheimer's disease. Electroenceph. Clin. Neurophysiol., 72, 290-297. Soininen, H. and Riekkinen, P.J. (1992): EEG in diagnostic and follow-up of Alzheimer's disease. Acta Neurol. Scand., 139, 36-39. Speckman, E.J. and Caspers, H. (1973): Neurophysiologische Grundlagen der Provokationsmethoden in der Elektroenezphalographie. Z. EEG-EMG, 4, 157-167. Steriade, M., Gloor, P., Llinas, R.R., Lopes da Silva, F.H. and Mesulam, M.M. (1990): Basic mechanism of cerebral rhythmic activities. Electroenceph. Clin. Neurophysiol., 76, 481-508. Van der Drift, J.H.A. and Kok, N.K.D. (1972): The EEG in cerebrovascular disorders in relation to pathology. In: A. Remond (Editor). The Handbook of Electroencephalography and Clinical Neurophysiology. Vol.14, pp. 12-64. Elsevier, Amsterdam. Visser, S.L., Stare, F.C., Vantilburg, W., Opden Velde, W., Blom, J.L. and De Rijke, W. (1976): Visual evoked response in senile and presenile dementia. Electroenceph. Clin. Neurophysiol., 40, 385-392. Visser, S.L., Van Tilburg, W., Hoijer, C., Jonker, C. and De Rijke, W. (1985): 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., 60, 115 121. Wright, C.E., Harding, G.F.A. and Orwin, A. (1984): Presenile dementia - the use of the flash and pattern VEP in diagnosis. Electroenceph. Clin. Neurophysiol., 57, 405-415. Wright, C.E., Harding, G.F.A. and Orwin, A. (1986): The flash and pattern VEP as a diagnostic indicator of dementia. Documenta Ophthalmologica, 62, 89-96.