870
BIOL PSYCHIATRY 1992;32:870-879
Topographical Changes in Alpha Power in Medicated and Unmedicated Schizophrenics During Digits Span Reverse Matching Test Yasuko Nagase, Yoshiro Okubo, Masato Matsuura, and
TakuyaKojima
The topographical distribution of alpha power reduction was compared in nine unmedicated schizophrenics (predominantly never-treated), 17 medicated schizophrenics, and 15 normal controls. The task involved four procedures: (1) listening to signal sound, (2) listening to digits for memorization, (3) after listening, and (4) listening to digits for recognition. The electroencephalograms (EEGs) during each procedure were analyzed with Fast Fourier Transformation and compared with EEGs at rest. While listening to the digits, medicated schizophrenics showed less alpha power reduction than normal controls and unmedicated schizophrenics. In addition, there were correlations found between the degree of alpha power reduction and medication dose, and score of chronic symptoms. These suggest that patients with different clinical backgrounds have differing cerebral activity.
Introduction There have been a number of electroencephalographic (EEG) studies performed on schizophrenic patients. Most of them have concentrated on EEGs at rest. Their major findings are as follows: schizophrenics have more power in lower frequency bands, less alpha power, and more beta power than normal subjects. When considering what kind of function is reflected in the differences between controls and schizophrenics, or considering focal functional differences, little can be seen from EEGs taken only at rest. EEG analysis during some kind of mental or motor task would thus seem to be useful. From this point of view, a number of studies on event-related potential (ERP) have been conducted. However, in order to investigate successive changes in cerebral activity over seconds, a study of background activity is more relevant. Guenther and Breitling (1985) reported that no significant power change occurred in the beta band of the left hemisphere in a group of schizophrenics during finger-thumb movements of the right hand, supposedly a task that activates the left hemisphere and
From the Department of Neuropsychiatry, Faculty of Medicine, Tokyo Medical and Dental University, Tokyo, Japan. Address reprint requests to Dr, Yasuko Nagase. Department of Neuropsychiatry, Faculty of Medicine, Tokyo Medical and Dental University, !-5-45, Yushima, Bunkyou.ku, Tokyo, ! 13, Japan. Received September 14, 1991; revised July 17, 1992.
© 1992 Society of Biological Psychiatry
0006-3223192/$05.00
Changes in Alpha Power in Schizophrenics
BIOL PSYCHIATRY 1992;32:870--879
871
Table I. Subjects and Clinical Background Unmedicated schizophrenics Medicated schizophrenics Normal controls
9 17 15
Male/female
Age
Duration of illness
BPRS
6/3 14/3 11/4
31.1 -4- 11.8 30.4.4- 6.9 27.6 - 4.6
2.3 years 8.6 years
38.9 48.5
that causes a power change in normal controls. Gruzelier et al (1990) reported on tasks involving the recognition of faces or words, and finger-thumb movements of the right and left hands. They found that there was a lack of reduction in beta-2 power during face recognition and finger-thumb movement of the left hand in schizophrenics, whereas there was a reduction in the normal control group. Alpha power reduction in EEGs is considered to reflect cortical activation (Gevins et al 1979; Ehrlichman and Wiener 1980; Pfurtscheller 1989). By investigating topographical changes during performance of a mental task that is thought to activate the left hemisphere, this study aims to elucidate the differences in cerebral function between schizophrenics and normal controls, and differences in cerebral function between schizophrenics with different clinical backgrounds.
Subjects and Methods Twenty-six schizophrenic patients were studied. There were 15 control subjects comprised mainly of hospital staff and medical students. Informed consent was obtained from each patient and normal subject after the nature of the study had been fully explained. All the schizophrenic patients satisfied DSM-III-R criteria for schizophrenia. Diagnoses were made by direct interview with the authors. Seventeen of the patients being treated with neuroleptics formed the medicated schizophrenic group. They were inpatients at the Department of Neuropsychiatry, Tokyo Medical and Dental University and at Asai Hospital. And nine patients who had taken no neuroleptics for at least 12 months formed the unmedicated schizophrenic group. They were outpatients at the Department of Neuropsychiatry, Tokyo Medical and Dental University. Three of the unmedicated schizophrenics had been treated in the past, but the other six had never been treated at all. The control subjects were healthy individuals with no history of psychiatric or neurological disease. The average ages, gender, duration of illness, and mean Brief Psychiatric Rating Scale (BPRS) scores for each group, are summarized in Table I. There was no significant difference among the ages of the three groups. The average total BPRS score was 48.5 in medicated schizophrenics and 38.9 in unmedicated schizophrenics. The former wes significantly higher than the latter (p < 0.05, t-test). The unmedicated schizophrenics had significandy higher scores for anxiety and suspicion than did the medicated schizophrenics. The medicated schizophrenics had significantly higher scores for emotional withdrawal, conceptual disorganization, mannerisms and posturing, motor retardation, uncool~erativeness, and blunted affect than did the unmedicated schizophrenics. The average medication do~e of neuroleptics in the medicated schizophrenic group was 790 mg chlorpromazine equivalent (Davis 1976) daily.
872
Y. Nagase et al
BIOL PSYCHIATRY 1992;32:870-879
Task procedure
signal sound
J, 1200Hz sound EEG analyzing epoch
2sec
listening for memorization
listening for recognition
reaction
J,
J.
t
"8,4,6"
"6,4,8"
2sec
2sec
2sec
Figure I. A set of mental tasks. At the beginning of a task, a signal sound was given. After 6--8 sec, subjects were asked to listen to three digits and to memorize them. Six to 8 sec later, subjects were asked to listen to another three digits and to respond if the latter ones were exactly the reverse of the former ones. EEG epochs analyzed are also shown.
The average educational level of the normal controls (17.2 years) was significantly higher than both the unmedicated schizophrenics (13.9 years) and the medicated schizophrenics (12.5 yt,ars) (p = 0.0001, t-test), but there was no significant difference between that of the unmedicated schizophrenics and medicated schizophrenics (p - 0.09, t-test).
Digits Span Reverse Matching Test As a mental task, a modified digits span reverse matching test was used. The subjects were asked to close their eyes and listen to an artificial sound and voice through headphones. The sound and voice were synthesized by a personal computer. The course of the task is summarized in Figure 1. A signal sound is presented. After 6-8 sec, three digits are presented (listening for memorization). After another 6-8 sec, another three digits are presented (listening for recognition). Subjects had been instructed to push a hand-held button to indicate "true" if the latter digits were in exactly the reverse order of the former ones. For example, if the first digits were "8,4,6" and the second ones "6,4,8", then the subjects were supposed to push the button. This task was repeated ten times for each subject with an EEG being recorded simultaneously. The mean interval between tasks was about 7 sec and tasks were presented randomly. About one third of the tasks were "true." To analyze the EEGs after recording, triggers were recorded at the beginning of each component of a task. Triggers and digits were supplied by a Nippon Electric Company (NEC) personal computer PC-6601.
Table 2. Significant Correlations between Clinical Data and the Alpha Power Reduction Index Alpha I Listening for memorization Conceptual disorganization Listening for recognition Conceptual disorganization Blunted affect Anxiety Dose of neuroleptics
Alpha 2
P4, 1"5
NS
T5 T5 F7 F4,C3,C4,Fz
NS NS NS NS
Changes in Alpha Power in Schizophrenics
BIOL PSYCHIATRY 1992;32:870-879
873
EEG Recording and Analysis For EEG recording, a Kouden Medical KMC-40IS was used. Data were recorded with a 0.3-sec time constant and a 120-Hz high frequency cut-off filter on an analogue tape from 16 electrodes according to the international 10-20 system. The following electrode sites were used: Fpl, Fp2, F3, F4, C3, C4, P3, P4, O1, 02, FT, FS, Fz, Pz, TS, and T6. All data were referenced to the linked ear. Prior to the presentation of a task, a resting EEG was recorded for several minutes with the subject's eyes closed to establish a baseline for each subject. As shown in Figure 1, Fast Fourier Transformation (FFT) spectra during a task were calculated during the following periods: 1. 2. 3. 4.
for for for for
2 2 2 2
sec sec sec sec
after listening to the signal sound (the preparation period), during listening to the digits for memorization, after listening to the digits (the period after listening), during listening to the digits for recognition.
EEG data were filtered with a 60-Hz high frequency cut-off and digitized by using an NEC-SANEI 7T18 topographic system at a sampling rate of 256 per epoch. One epoch consisted of 2 sec. EEG spectra were calculated using the FFI'. Only trials in which there were no artifacts caused by eye movements, eye blinks, or muscle activity were included in the averaging process. Artifact rejection was performed by off-line evaluation of the data. For each subject, 6-10 trials were averaged. F F r spectra during the resting state were analyzed separately. From 24-30 sec-long artifact-free epochs were selected from the baseline recording and averaged for each subject. In this study, alpha-I (8-10 Hz) and alpha-2 (10-13 Hz) bands were investigated. The absolute power and reduction index (RI) were calculated as follows: RI = (PR-PT)/PR PR: absolute power at rest. PT: absolute power during a task. This parameter reflects the alpha power reduction during a task.
Statistics Logarithmic transformation of absolute power was used to normalize the data. Data were analyzed using analysis of variance (ANOVA) and subsequent t-test. For ANOVA, 12 out of 16 electrodes were divided into four regions (left anterior: Fpl, F3, F7, right anterior: Fp2, F4, FS, left posterior: P3, O1, T5, right posterior: 1'4, 02, T6) and mean values of three electrodes in each region were used. To provide the basis for independent (comparison of resting states across groups) and dependent (during a task as compared to the resting state) t-test measurements, two-way ANOVA of a one-grouping factor and one within-subject factor design: 3 (diagnostic groups) x 4 (regions) were calculated for the resting states, and for the resting and task states, four-way ANOVA of a threegrouping factor and one within-subject factor design: 3 (diagnostic groups) x 41 (subjects nested within diagnostic groups) × 5 (states: rest, tasks) x 4 (regions) were calculated for each frequency band (alpha 1 and alpha 2). The Bonferroni adjustment was applied to the p value derived using the t-test. Sig-
874
Y. Nagase et al
BIOL PSYCHIATRY 1992;32:870-879
Normal Controls
,D,
..'iH~~H~". .', ~-i":~,~,*~:~:!-,,ii.,. ....
alpha1 8-10Hz
Unmedicated Schizophrenics
. .
.-". . . .
.....
~
.
.
. ~
"'.
.
Medicated Schizophrenics ;i~:
::'~i
.
" i:iii:
....:C~:.:~.::~. .
N '~''
O~DD~
alpha 2 i!
!
~.,,~ Q ~ !~,. I
Li;i~,?~]
iR
.
2
5
XIO 4 liv
Figure 2. Comparison of alpha power at rest. Absolute alpha power maps at rest are shown. Each image represents a schematic map of the head viewed from above with the nose uppermost, the occiput below, the left ear to the left, and the right ear to the right. The upper row presents alpha 1 (8-10 Hz) power, and the lower row alpha-2 (10-13 Hz) power. In controls and unmedicated schizophrenics, alpha power was predominant in the posterior area of the scalp. This tendency was not prominent in medicated schizophrenics.
nificant probability maps (SPMs) were created to investigate topographical changes in alpha power between the resting state and during the performance of a task.
Results
Comparison at Rest The result of the three-way ANOVA showed the interaction of diagnostic groups x regions was significant in the alpha 2 band (F = 2.50, df = 6, p = 0.0486, corrected by Huynh-Feld epsilon). Figure 2 shows the mean absolute alpha power for each group at rest. The t-test results indicate that there was no significant difference between the unmedicated schizophrenics and the normal controls in either alpha-I or alpha-2 power. There was also no significant difference between the medicated schizophrenics and the normal controls in alpha-1 power, but the medicated schizophrenics had significantly less alpha-2 power in OI, 02, T6 than normal controls (p < 0.05/2 = 0.025, Bonferroni adjustment).
EEG Changes during Task Performance All of the schizophrenics as well as the normal controls performed the tasks almost perfectly. Therefore, there was no difference among groups with regard to the performance of tasks. Results of the four-way ANOVA show that there was significant interaction of
Changes in Alpha Power in Schizophrenics
BIOL PSYCHIATRY
8"/5
1992;32:870-879
Normal Controls
Unmedicated Schizophrenics
~
.....
, .....
alpha 1 8-10Hz
,
.....
S... 21,, .,.~ ,':.
':. . • ....-~.,_..
Medicated Schizophrenics
~."
•.
"::;.. ,..LI~ '
. ....
. ..,...~,,
,'..'.
N
,...,' "7 ",. •.,
alpha 2 10.13Hz . . . .
?,,,,
,...,' "7 ',..., :
::o
..
•
~
•
:...,' "7 ',..:
~:;
::.
s
s
s
o::
...,-,
• ,,,,':
,
.....
," , ....
,,,,,
II p
BIOL PSYCHIATRY 1992;32:870-879
Topographical Changes in Alpha Power in Medicated and Unmedicated Schizophrenics During Digits Span Reverse Matching Test Yasuko Nagase, Yoshiro Okubo, Masato Matsuura, and
TakuyaKojima
The topographical distribution of alpha power reduction was compared in nine unmedicated schizophrenics (predominantly never-treated), 17 medicated schizophrenics, and 15 normal controls. The task involved four procedures: (1) listening to signal sound, (2) listening to digits for memorization, (3) after listening, and (4) listening to digits for recognition. The electroencephalograms (EEGs) during each procedure were analyzed with Fast Fourier Transformation and compared with EEGs at rest. While listening to the digits, medicated schizophrenics showed less alpha power reduction than normal controls and unmedicated schizophrenics. In addition, there were correlations found between the degree of alpha power reduction and medication dose, and score of chronic symptoms. These suggest that patients with different clinical backgrounds have differing cerebral activity.
Introduction There have been a number of electroencephalographic (EEG) studies performed on schizophrenic patients. Most of them have concentrated on EEGs at rest. Their major findings are as follows: schizophrenics have more power in lower frequency bands, less alpha power, and more beta power than normal subjects. When considering what kind of function is reflected in the differences between controls and schizophrenics, or considering focal functional differences, little can be seen from EEGs taken only at rest. EEG analysis during some kind of mental or motor task would thus seem to be useful. From this point of view, a number of studies on event-related potential (ERP) have been conducted. However, in order to investigate successive changes in cerebral activity over seconds, a study of background activity is more relevant. Guenther and Breitling (1985) reported that no significant power change occurred in the beta band of the left hemisphere in a group of schizophrenics during finger-thumb movements of the right hand, supposedly a task that activates the left hemisphere and
From the Department of Neuropsychiatry, Faculty of Medicine, Tokyo Medical and Dental University, Tokyo, Japan. Address reprint requests to Dr, Yasuko Nagase. Department of Neuropsychiatry, Faculty of Medicine, Tokyo Medical and Dental University, !-5-45, Yushima, Bunkyou.ku, Tokyo, ! 13, Japan. Received September 14, 1991; revised July 17, 1992.
© 1992 Society of Biological Psychiatry
0006-3223192/$05.00
Changes in Alpha Power in Schizophrenics
BIOL PSYCHIATRY 1992;32:870--879
871
Table I. Subjects and Clinical Background Unmedicated schizophrenics Medicated schizophrenics Normal controls
9 17 15
Male/female
Age
Duration of illness
BPRS
6/3 14/3 11/4
31.1 -4- 11.8 30.4.4- 6.9 27.6 - 4.6
2.3 years 8.6 years
38.9 48.5
that causes a power change in normal controls. Gruzelier et al (1990) reported on tasks involving the recognition of faces or words, and finger-thumb movements of the right and left hands. They found that there was a lack of reduction in beta-2 power during face recognition and finger-thumb movement of the left hand in schizophrenics, whereas there was a reduction in the normal control group. Alpha power reduction in EEGs is considered to reflect cortical activation (Gevins et al 1979; Ehrlichman and Wiener 1980; Pfurtscheller 1989). By investigating topographical changes during performance of a mental task that is thought to activate the left hemisphere, this study aims to elucidate the differences in cerebral function between schizophrenics and normal controls, and differences in cerebral function between schizophrenics with different clinical backgrounds.
Subjects and Methods Twenty-six schizophrenic patients were studied. There were 15 control subjects comprised mainly of hospital staff and medical students. Informed consent was obtained from each patient and normal subject after the nature of the study had been fully explained. All the schizophrenic patients satisfied DSM-III-R criteria for schizophrenia. Diagnoses were made by direct interview with the authors. Seventeen of the patients being treated with neuroleptics formed the medicated schizophrenic group. They were inpatients at the Department of Neuropsychiatry, Tokyo Medical and Dental University and at Asai Hospital. And nine patients who had taken no neuroleptics for at least 12 months formed the unmedicated schizophrenic group. They were outpatients at the Department of Neuropsychiatry, Tokyo Medical and Dental University. Three of the unmedicated schizophrenics had been treated in the past, but the other six had never been treated at all. The control subjects were healthy individuals with no history of psychiatric or neurological disease. The average ages, gender, duration of illness, and mean Brief Psychiatric Rating Scale (BPRS) scores for each group, are summarized in Table I. There was no significant difference among the ages of the three groups. The average total BPRS score was 48.5 in medicated schizophrenics and 38.9 in unmedicated schizophrenics. The former wes significantly higher than the latter (p < 0.05, t-test). The unmedicated schizophrenics had significandy higher scores for anxiety and suspicion than did the medicated schizophrenics. The medicated schizophrenics had significantly higher scores for emotional withdrawal, conceptual disorganization, mannerisms and posturing, motor retardation, uncool~erativeness, and blunted affect than did the unmedicated schizophrenics. The average medication do~e of neuroleptics in the medicated schizophrenic group was 790 mg chlorpromazine equivalent (Davis 1976) daily.
872
Y. Nagase et al
BIOL PSYCHIATRY 1992;32:870-879
Task procedure
signal sound
J, 1200Hz sound EEG analyzing epoch
2sec
listening for memorization
listening for recognition
reaction
J,
J.
t
"8,4,6"
"6,4,8"
2sec
2sec
2sec
Figure I. A set of mental tasks. At the beginning of a task, a signal sound was given. After 6--8 sec, subjects were asked to listen to three digits and to memorize them. Six to 8 sec later, subjects were asked to listen to another three digits and to respond if the latter ones were exactly the reverse of the former ones. EEG epochs analyzed are also shown.
The average educational level of the normal controls (17.2 years) was significantly higher than both the unmedicated schizophrenics (13.9 years) and the medicated schizophrenics (12.5 yt,ars) (p = 0.0001, t-test), but there was no significant difference between that of the unmedicated schizophrenics and medicated schizophrenics (p - 0.09, t-test).
Digits Span Reverse Matching Test As a mental task, a modified digits span reverse matching test was used. The subjects were asked to close their eyes and listen to an artificial sound and voice through headphones. The sound and voice were synthesized by a personal computer. The course of the task is summarized in Figure 1. A signal sound is presented. After 6-8 sec, three digits are presented (listening for memorization). After another 6-8 sec, another three digits are presented (listening for recognition). Subjects had been instructed to push a hand-held button to indicate "true" if the latter digits were in exactly the reverse order of the former ones. For example, if the first digits were "8,4,6" and the second ones "6,4,8", then the subjects were supposed to push the button. This task was repeated ten times for each subject with an EEG being recorded simultaneously. The mean interval between tasks was about 7 sec and tasks were presented randomly. About one third of the tasks were "true." To analyze the EEGs after recording, triggers were recorded at the beginning of each component of a task. Triggers and digits were supplied by a Nippon Electric Company (NEC) personal computer PC-6601.
Table 2. Significant Correlations between Clinical Data and the Alpha Power Reduction Index Alpha I Listening for memorization Conceptual disorganization Listening for recognition Conceptual disorganization Blunted affect Anxiety Dose of neuroleptics
Alpha 2
P4, 1"5
NS
T5 T5 F7 F4,C3,C4,Fz
NS NS NS NS
Changes in Alpha Power in Schizophrenics
BIOL PSYCHIATRY 1992;32:870-879
873
EEG Recording and Analysis For EEG recording, a Kouden Medical KMC-40IS was used. Data were recorded with a 0.3-sec time constant and a 120-Hz high frequency cut-off filter on an analogue tape from 16 electrodes according to the international 10-20 system. The following electrode sites were used: Fpl, Fp2, F3, F4, C3, C4, P3, P4, O1, 02, FT, FS, Fz, Pz, TS, and T6. All data were referenced to the linked ear. Prior to the presentation of a task, a resting EEG was recorded for several minutes with the subject's eyes closed to establish a baseline for each subject. As shown in Figure 1, Fast Fourier Transformation (FFT) spectra during a task were calculated during the following periods: 1. 2. 3. 4.
for for for for
2 2 2 2
sec sec sec sec
after listening to the signal sound (the preparation period), during listening to the digits for memorization, after listening to the digits (the period after listening), during listening to the digits for recognition.
EEG data were filtered with a 60-Hz high frequency cut-off and digitized by using an NEC-SANEI 7T18 topographic system at a sampling rate of 256 per epoch. One epoch consisted of 2 sec. EEG spectra were calculated using the FFI'. Only trials in which there were no artifacts caused by eye movements, eye blinks, or muscle activity were included in the averaging process. Artifact rejection was performed by off-line evaluation of the data. For each subject, 6-10 trials were averaged. F F r spectra during the resting state were analyzed separately. From 24-30 sec-long artifact-free epochs were selected from the baseline recording and averaged for each subject. In this study, alpha-I (8-10 Hz) and alpha-2 (10-13 Hz) bands were investigated. The absolute power and reduction index (RI) were calculated as follows: RI = (PR-PT)/PR PR: absolute power at rest. PT: absolute power during a task. This parameter reflects the alpha power reduction during a task.
Statistics Logarithmic transformation of absolute power was used to normalize the data. Data were analyzed using analysis of variance (ANOVA) and subsequent t-test. For ANOVA, 12 out of 16 electrodes were divided into four regions (left anterior: Fpl, F3, F7, right anterior: Fp2, F4, FS, left posterior: P3, O1, T5, right posterior: 1'4, 02, T6) and mean values of three electrodes in each region were used. To provide the basis for independent (comparison of resting states across groups) and dependent (during a task as compared to the resting state) t-test measurements, two-way ANOVA of a one-grouping factor and one within-subject factor design: 3 (diagnostic groups) x 4 (regions) were calculated for the resting states, and for the resting and task states, four-way ANOVA of a threegrouping factor and one within-subject factor design: 3 (diagnostic groups) x 41 (subjects nested within diagnostic groups) × 5 (states: rest, tasks) x 4 (regions) were calculated for each frequency band (alpha 1 and alpha 2). The Bonferroni adjustment was applied to the p value derived using the t-test. Sig-
874
Y. Nagase et al
BIOL PSYCHIATRY 1992;32:870-879
Normal Controls
,D,
..'iH~~H~". .', ~-i":~,~,*~:~:!-,,ii.,. ....
alpha1 8-10Hz
Unmedicated Schizophrenics
. .
.-". . . .
.....
~
.
.
. ~
"'.
.
Medicated Schizophrenics ;i~:
::'~i
.
" i:iii:
....:C~:.:~.::~. .
N '~''
O~DD~
alpha 2 i!
!
~.,,~ Q ~ !~,. I
Li;i~,?~]
iR
.
2
5
XIO 4 liv
Figure 2. Comparison of alpha power at rest. Absolute alpha power maps at rest are shown. Each image represents a schematic map of the head viewed from above with the nose uppermost, the occiput below, the left ear to the left, and the right ear to the right. The upper row presents alpha 1 (8-10 Hz) power, and the lower row alpha-2 (10-13 Hz) power. In controls and unmedicated schizophrenics, alpha power was predominant in the posterior area of the scalp. This tendency was not prominent in medicated schizophrenics.
nificant probability maps (SPMs) were created to investigate topographical changes in alpha power between the resting state and during the performance of a task.
Results
Comparison at Rest The result of the three-way ANOVA showed the interaction of diagnostic groups x regions was significant in the alpha 2 band (F = 2.50, df = 6, p = 0.0486, corrected by Huynh-Feld epsilon). Figure 2 shows the mean absolute alpha power for each group at rest. The t-test results indicate that there was no significant difference between the unmedicated schizophrenics and the normal controls in either alpha-I or alpha-2 power. There was also no significant difference between the medicated schizophrenics and the normal controls in alpha-1 power, but the medicated schizophrenics had significantly less alpha-2 power in OI, 02, T6 than normal controls (p < 0.05/2 = 0.025, Bonferroni adjustment).
EEG Changes during Task Performance All of the schizophrenics as well as the normal controls performed the tasks almost perfectly. Therefore, there was no difference among groups with regard to the performance of tasks. Results of the four-way ANOVA show that there was significant interaction of
Changes in Alpha Power in Schizophrenics
BIOL PSYCHIATRY
8"/5
1992;32:870-879
Normal Controls
Unmedicated Schizophrenics
~
.....
, .....
alpha 1 8-10Hz
,
.....
S... 21,, .,.~ ,':.
':. . • ....-~.,_..
Medicated Schizophrenics
~."
•.
"::;.. ,..LI~ '
. ....
. ..,...~,,
,'..'.
N
,...,' "7 ",. •.,
alpha 2 10.13Hz . . . .
?,,,,
,...,' "7 ',..., :
::o
..
•
~
•
:...,' "7 ',..:
~:;
::.
s
s
s
o::
...,-,
• ,,,,':
,
.....
," , ....
,,,,,
II p
