- Email: [email protected]
Learned control of slow potential interhemispheric asymmetry in schizophrenia
INTERNATIONAL JOURNAL OF PSYCHOPHYSIOLOGY ELSEVIER International Journal of Psychophysiology 34 (1999) 341-348
www.elsevier.com/locate/physcho
Learned control of slow potential interhemispheric asymmetry in schizophrenia John Gruzelier *, Elinor Hardman, Jennifer Wild, Rashid Zaman Department of Behavioural and Cognitive Sciences, Division ofNeuroscience and Psychological Medicine, Imperial College School ofMedicine at Charing Cross Hospital, University ofLondon, London, UK
Received 7 July 1998; received in revisedform 23 October 1998; accepted 5 January 1999
Abstract We report on the feasibility of teaching 16 (DSM-IV) schizophrenic patients, subdivided by syndrome, self-regulation of interhemispheric asymmetry having demonstrated efficient learning of interhemispheric control in normal subjects. Reversal of asymmetry may be important to treatment and recovery in schizophrenia for following improvement on neuroleptic drugs functional hemispheric asymmetries have reversed, with directions of reversal and pre-existing asymmetry dependent on syndrome. Asymmetry reversal in animals, manifested by spatial turning tendencies, has been used as a marker of neuroleptic action and involves striatal dopamine under reciprocal hemispheric control. We gave as feedback the left-right asymmetry in slow potential negativity recorded from the sensory motor strip (C3,4). Feedback took the form of a rocket on a screen which rose or fell with leftward or rightward shifts in negativity. Patients were able to learn control (P < 0.01). In those patients with lesser ability this was due to inability to sustain concentration throughout the session rather than slow initial learning. Active syndrome patients were better able to shift negativity rightward and withdrawn patients leftward, directions associated with drug reversal of functional asymmetry and symptom recovery for each syndrome. Accordingly our demonstration that many symptomatic schizophrenic patients are capable of learning control opens the door to electrocortical operant conditioning training in schizophrenia with therapeutic regimens. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Schizophrenia; EEG; Operant conditioning; Slow potentials; Laterality; Syndrome
* Corresponding author, Department of Behavioural and Cognitive Sciences, Imperial College School of Medicine, St Dunstan's Road, London W6 8RF. Tel.: +44-181-846-7386; fax: +44-181-846-1670. E-mail address:[email protected] (1. Gruzelier) 0167-8760/99/$ - see front matter © 1999 Elsevier Science B.Y. All rights reserved. PH: SOl 67- 8760 ( 9 9 ) 0 0 0 9 1 - 4
342
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
1. Introduction
There is an accumulation of evidence that hemispheric functional asymmetries in schizophrenia are dynamic and may reflect an imbalance of activity between the hemispheres rather than a unilateral anomaly (Gruzelier, 1999a). In a theoretical review we posited that underlying mechanisms were consistent with an imbalance in thalamo-cortical regulatory mechanisms having widespread neurophysiological and neurochemical influences throughout the hemispheres, not simply the more strongly innervated frontal thalamo-cortical regions (Gruzelier, 1999a). As cortical electroencephalographic activity (EEG) is under the control of thalamo-cortical mechanisms (Lopes da Silva, 1992), EEG measures are heuristic in measuring dynamic changes in functional imbalance. Changes in the direction of functional imbalance may be of therapeutic importance. Evidence has been reviewed showing that clinical improvement on neuroleptics has coincided with a reversal or shift in functional asymmetry (Gruzelier, 1999a), some of which were dose related (Hammond and Gruzelier, 1978; Mintz et a1., 1982; Harvey et a1., 1993). Measures showing drug reversal have included hemi-inattention (Hammond and Gruzelier, 1978; Posner et a1., 1988; Tomer and Flor-Henry, 1989; Maruff et a1., 1995), flash evoked potentials (Mintz et a1., 1982), lower motoneuron excitability (Tan and Gurgen, 1986), electrodermal orienting responses (Gruzelier et a1., 1981) and recognition memory (Gruzelier et a1., 1999a,b). Reversals of functional asymmetry with treatment in schizophrenia are consistent with animal research where a reversal of spatial turning tendencies was used as a marker of neuroleptic action (Ungerstedt, 1971; Pycock et a1., 1980; Glick and Ross, 1981). Glowinski et a1. (1984) have shown that the striatal dopamine pathways which are instrumental in these effects are under reciprocal hemispheric control, in keeping with neuroleptic influences on asymmetry reversa1. We reasoned that alteration of interhemispheric asymmetry through learned control may also benefit schizophrenic patients. In this feasi-
bility study we set out to explore whether schizophrenic patients could learn self-regulation of interhemispheric asymmetry through electrocortical conditioning biofeedback. The field of neurofeedback applications in psychopathology is developing as EEG mechanisms become better understood. Approaches have out grown the earlier work with alpha training with which the field was associated two decades ago, and which along with autonomic procedures was aimed at simply inducing relaxation (Evans and Barbanel, 1999). We chose to monitor cortical slow potential negativity encouraged by the results of Birbaumer and colleagues (Birbaumer et a1., 1991; Rockstroh et a1., 1993) who demonstrated that two-thirds of epileptic patients who were refractory to other forms of treatment benefited by learned control of cortical slow potentials and implemented self regulation when they felt a fit coming on. Slow cortical potentials reflect depolarisation of cortical pyramidal cells, in particular in layer IV, and index cortical excitability (Birbaumer et a1., 1990). In schizophrenia directions of functional hemispheric imbalance have been shown to be syndrome related (for review see Gruzelier, 1999a). An 'Active' syndrome was associated with a L > R asymmetry, the syndrome consisting of positive symptoms including overactivity, accelerated cognition or positive thought disorder, positive or labile affect, heightened self concepts and affective delusions. A 'Withdrawn' syndrome was associated with the opposite R > L asymmetry, consisting of affective blunting, social and emotional withdrawal and motor retardation. Here we subclassified patients by syndrome. The training programme was aimed at shifting the balance of cortical negativity interhemispherically. We had first developed training schedules with normal subjects and demonstrated that they could quickly learn interhemispheric control of frontal (F3,4) and central (C3,4) asymmetries (Hardman et a1., 1997). As larger effects were obtained with electrodes positioned centrally, central placements were adopted with schizophrenic patients. According to our model of hemispheric imbalance in schizophrenia it would be therapeutically advantageous if active syndrome patients with their L > R functional asymmetry
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
could shift activity rightward, away from their functionally dominant hemisphere and the direction associated with reversal of asymmetry with neuroleptic treatment. Conversely leftward shifts in negativity would be advantageous in withdrawn syndrome patients with their R > L asymmetry.
2. Methods
2.1. Subjects
Twenty-five medicated DSM-IV schizophrenic patients gave signed consent of whom 11 were male with an average age of 34.2 years. Psychiatrists who were independent of the laboratory measures rated the patients with Active, Withdrawn, or Mixed syndromes on the Schizophrenia Syndrome Scale (see Gruzelier et a1., 1999b). Four patients subsequently refused and four withdrew after 1-2 sessions: seven of the eight belonged to the Withdrawn syndrome. Sixteen cooperated in a training programme of up to 10 sessions. All completed six sessions: seven Active; four Withdrawn; and five Mixed syndrome; while 10 completed all 10 sessions. In two single case studies symptoms were monitored each session with the Positive and Negative Symptom Scale PANSS (Kay et a1., 1987). The average IQ of the patients was 104 and they had an average of 38 weeks total time in hospita1. 2.2. EEG conditioning procedures 2.2.1. Apparatus
Slow potentials were recorded with a grass polygraph from EEG Ag/AgCl electrodes placed at C3, C4 (10/20 system) and C4 linked mastoids and fastened with Elefix ge1. The electrooculogram (EOG) was recorded across the outer canthi and above and below the left eye. The biofeedback was contingent on the integrated C3-C4 signa1. Feedback took the form of an on-screen rocket ship, initially centrally placed, which rose as a function of an increase in left hemisphere negativity (relative to the right hemisphere) and fell to indicate a relative increase in right hemisphere negativity. The rocket was designed to
343
move in the vertical plane to avoid artefacts from lateral conjugate eye movements. Data were screened manually and on-line for eye and movement artefacts. 2.2.2. Procedure
Subjects were seated in a comfortable chair 140 em from the monitor on which the 10 x 6-cm rocket was displayed. Trials lasted 8 s with a 4-s interstimulus interva1. Artefact free trials were averaged for each subject according to trial type: 'A' instructing a leftward shift or 'B' a rightward shift. The mean DC shift between a 1-s pre-trial baseline and the last 6 s of the trial was calculated. Trials were presented in a pseudorandomised order. Patients were instructed to concentrate on contralateral body sensations to move the rocket. They were paid for participation and rewarded with lOp for each correct direction shift. The 60 trials were divided into blocks of 20, with rests between blocks and after every fifth trial when encouragement was given. There were up to 10 sessions on separate days. 2.3. Statistical analysis
The main statistical analyses consisted of two repeated measures analyses of variance with Greenhouse-Geisser conservative degrees of freedom. The first was with the 10 subjects who received all 10 sessions, the sessions were divided into two blocks of five sessions each - withinsubject factor Session/Block (2). Each session consisted of 60 trials divided into three blocks of 20 trials - within-subject factor Trial Block (3). Trials were subdivided according to whether leftward (A) or rightward (B) shifts in negativity were required within-subject trial type A/BLaterality (2). All patients who participated in training learned control, however, there were individual differences in performance. Accordingly the patients were divided into good and average performers based on consensual judgement between two independent raters through visual inspection of mean traces across blocks and individual sessions. This gave a between-subject factor - Performance (2). Interaction terms were elucidated by comparisons between means.
344
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
The second analysis examined syndrome relations in the larger group of 16 who had at least six sessions. In this analysis there was a betweensubject factor of Syndrome (Active/WithdrawnMixed), a within-subject factor of Session (6), but otherwise the same within-subject factors as in the first analysis.
3. Results 3.1. Training and performance
Means and standard deviations of good and average performers are shown in Fig. 1. There was clear evidence of the ability of schizophrenic patients to learn self regulation of interhemispheric negativity, shifting negativity to the left on A trials and to the right on B trials (Laterality: F = 28.22, d.f. = 1,8, P < 0.00l). It can also be seen that good performers had lateral shifts about twice as large as average performers (Group X Laterality: F = 4.48, d.f. = 1,8, P < 0.058). This effect was found to depend on session, for while there was evidence of a tendency towards improvement for the group as a whole across Ses-
sion/Block (F = 4.10, d.f. = 1,8, P < 0.077), there was a further interaction with Group (F = 10.23, d.f. = 1,8, P < 0.013). It was a falloff in performance over sessions in average performers, not slower initial learning, that differentiated good (Fig. 2) from average (Fig. 3) performers. There was also evidence common to both groups of improvement within sessions (Block x Laterality: F = 3.39, d.f. = 1.96, 15.26, P < 0.056), more so in the second Block of Sessions (F = 4.59, d.f. = 1.97, 15.78, P < 0.027). The interactions were examined further by analyses on each of the three blocks within sessions and, aside from comparing groups, comparing the effect of training by contrasting learning on the first five sessions with the remaining five. A tendency was shown for better performance with training in good performers in the first block (F = 4.98, d.f. = 1,8, P < 0.056), there were no group differences in the second block (F < 1.84), whereas in the third block there was a highly significant effect of training (F = 12.98, d.f. = 1,8, P < 0.007). The latter effect better characterised the good performers (F = 10.45, d.f. = 1,8, P < 0.012). From this may be inferred that reasons for reduced performance in average performers in-
Mean Leftward (a) and Rightward (b) Direction Shifts 6.00,----------------------------,
4.00
2.00
:;2-
]oo/--------r--Ul
c:
"'"
:;; -2.00 ~
"
...J
o Good Performers
-4.00
• Average Performers
-6.00 ' - - - - - - - - - - - - - (a) TRIALS
-------------' (b) TRIALS
Fig. 1. Mean leftward (Trial a) and rightward (Trial b) interhemispheric shifts in negativity for good and average performers showing significant (P < 0.001) regulation.
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348 Good Performers: Block I 15
~
1
\0
I
//~'-"-.'-.
",:IS
~ ~
"
I0
-15
1.1
~,
I
3,1
9.\
71
10,1
345
the course of six sessions. There were no syndrome differences in the various training parameters as shown by Block and Session effects (F < 1.92). T test comparisons indicated there was no difference between the Performance groups and syndrome score or total score (t = 0.21-1.38). On average the Active syndrome group produced larger rightward than leftward shifts in negativity: rightward, 3.24; leftward, 1.58 (t = 3.10, d.f. = 5, P < 0.027), while the Withdrawn syn-
BLOCK1
Good Performers: Block 2 15
&:
BLOCKl
BLOCK2
Good Performers: Block 3
Average Performers: Block 2 cr.
lSI
&:
10 ~
>
51
-------~~~
-: ///~'.,.,///~~-~-
o----::co~-~--------'=~=_-r
-~
~
............. -,
~ ~
4,3
5.3
6}
7.3
8.3
~ ~51
<,
~-
9}
<,
10.3
r.z
BLOCK3
2.2
3,2
4.2
.r'
5,2
6,2
72
8.2
9.2
BLOCK2
Fig. 2. Mean leftward and rightward shifts for good performers across 10 sessions for the three within-session blocks of trials.
Average Performers: Block 3
fc
:2
eluded a loss of concentration towards the end of the sessions.
"',.,~-~-~,.
i
\01
I
~" ~
~~~-<~.~~ -, "
3.2. Syndrome
Both syndromes learned self control of interhemispheric negativity (F = 19.71, d.f. = 1,14, P < 0.001). A highly significant trial effect (Fl 14 = 19.71, P < 0.001) indicated that schizophrenic patients could learn interhemispheric control in
"
1.3
5.3
9.3
10J
BLOCK3
Fig. 3. Mean leftward and rightward shifts for average performers across 10 sessions for the three within-session blocks of trials.
346
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
drome produced larger leftward shifts: rightward, 1.84; leftward, 2.37 (t = 2.75, d.f. = 4, P < 0.07), as did the Mixed syndrome: rightward, 2.52; leftward, 3.69 (t = 3.196, d.f. = 4, P < 0.034); combining mixed/withdrawn (t = 0.47, d.f. = 5, P < 0.002). A median split on total symptom score (N = 12, seven sessions) did not show a clear relation between severity of symptoms and learned control (P < 0.139). Correlations were examined between each of the active, withdrawn and unreality syndrome scores and negativity shifts on leftward and rightward trials across the first five trials. There was a correlation (r = 0.60, P < 0.05) between Active syndrome rating score and leftward shift. The higher the Active rating the smaller the leftward shift, evidence of a possible ceiling effect in directional shift.
Mean A & B Shifts
I
./ I
2
)
4
5
6
s
7
9
'\..--.-
10
50
40
3.3. Case studies
~
..
POSIT1VE NEGATIVE GENERAL
In a patient who undertook 20 sessions of train-
ing interhemispheric shifts continued to grow in size after the 10th session, and an impressive ability to respond differentially with leftward or rightward shift trial type was sustained throughout the further 10 sessions (see Fig. 4). In other cases symptoms were monitored concurrently with the PANSS. General symptoms such as anxiety and depression appear to be more closely related to neurofeedback ability than specific psychotic symptoms. One patient first undertook the 10 training sessions with limited success. During this time General Psychopathology scores SINGLE CASE STUDY: Mean Leftward (A) and Rightward (B)
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
session
Fig. 5. A patient who resumed training after a 3-month absence following session 10. Upper figure leftward (A) and rightward (E) directional shifts in microvolts. Lower figure total positive, negative and general psychopathology ratings on the PANS scale.
were above the 50th percentile. Then after a break of 3 months, nine more sessions were undertaken. The General Psychopathology rating total had dropped by 50% (see Fig. 5 lower). Upon resuming training, a huge 18 f.LV differentiation was obtained on the first occasion (see Fig. 5 upper). This was then maintained for the further seven sessions.
Direction Shifts Across 20 Sessions
4. Discussion
'3 ......l
.10'-1_-,----,....--,_ _-,----_---,,-_-,----_,.---,._-,----,........, 2 3 4 5 6 7 s 9 10 11 12 13 14 15 16 17 III 19 20
Mean B
Session
Fig. 4. Mean leftward and rightward shifts in a patient undergoing 20 sessions.
Contrary to the belief that schizophrenic patients will be unable to learn self control of electrocortical activity due to attentional and motivational deficits, all patients who cooperated with training were able to learn to control interhemispheric negativity between lateralised central electrodes (C3/C4). This constituted two-thirds
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
of our symptomatic schizophrenic patients who initially agreed to participate and gave signed consent. Another study also involving operant conditioning of slow cortical potentials, but requiring the increasing and decreasing of negativity at Cz, had moderate success (Schneider et a1., 1992). Withdrawal was the feature that characterised those patients who did not cooperate in our study, though withdrawal itself did not preclude patients from cooperating. In medical students levels of withdrawal akin to introversion were found an advantage in learning interhemispheric control (Hardman et a1., 1997). At the other extreme, severity of positive symptoms also may not be a handicap to learning. It was nonspecific symptoms such as anxiety and distractibility that were more of a handicap to cooperation than positive symptoms. In medical students tiredness was disadvantageous to successful training while calmness was beneficia1. When examining individual differences in performance the greater difficulties in control were found with sustaining concentration towards the end of sessions rather than in initial learning. This suggests that shorter sessions and/or stronger reinforcers would facilitate training. The evidence that non-specific symptoms such as tension and anxiety may be counterproductive to neurofeedback training raises the question whether training first in stress reduction through peripheral biofeedback procedures such as EMG might facilitate the self regulation of electrocortical activity. It is common practice in other fields such as optimal performance training to precede neurofeedback with biofeedback stress reduction protocols and there are reports of successful stress reduction with biofeedback in schizophrenia (Schwartz, 1995; Pharr and Coursey, 1989). Currently there are growing applications of neurofeedback with a range of psychopathologies and applications for the enhancement of sporting and musical performance (Evans and Barbanel, 1999). There is no reason to suspect that success in demonstrating learned control of slow cortical potentials would not generalise to other more widely practised forms of training with EEG rhythms that have been successful in treating dysregulation of attentional, motivational and
347
arousal processes (Othmer et a1., 1999). As recently concluded in a review of early processing deficits and dysregulation of thalamo-cortical functions in schizophrenia (Gruzelier, 1999b; Gruzelier et a1., 1999a), there is a body of evidence including well documented abnormalities of slow to fast EEG rhythms, and which may extend to fast transients in the gamma range (Baldeweg et a1., 1998), which may respond to correction through operant conditioning procedures. Of theoretical and therapeutic importance was the demonstration that schizophrenic patients could learn to shift negativity away from their functionally dominant hemisphere: a rightward direction in the Active syndrome who have leftsided functional biases and a leftward direction in the Withdrawn syndrome who have right-sided functional biases. This indicates that the premise underlying the training of reversing functional biases in asymmetry through learning self-regulation is feasible in practice. The lack of literature on applying neurofeedback schedules to schizophrenia has represented the belief that schizophrenic patients would be unable to cooperate with training. The significance of the present study is that it removes this shibboleth, though patient motivation must be carefully monitored. Having demonstrated the feasibility of operant conditioning with schizophrenia, the way is clear to examine the efficacy of therapeutic intervention in schizophrenia, schizotypy and the schizophrenia spectrum.
Acknowledgements
The research was supported by a Sr investigator award from NARSAD and the L.F. Saugstad fund. Presented to the 5 th Cognitive Neuroscience Society meeting, San Francisco, April, 1998. References Baldeweg, T., Spence, S., Hirsch, S.R. et aI., 1998. Gammaband electroencephalographic oscillations in a patient with somatic hallucinations. Lancet 352, 620-621.
348
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
Birbaumer, N., Elbert, T, Canavan, AG.M., Rockstroh, B., 1990. Slow potentials of the cerebral cortex and behaviour. Physiol. Rev. 70 (1), 1-30. Birbaumer, N., Elbert, T, Rockstroh, B., Daum, I., Woolf, P., Canavan, A, 1991. Clinical-psychological treatment of epileptic seizures: a controlled study. In: Ehlers, A et al.Gid.), Perspectives and Promises of Clinical Psychology. Plenum Press, New York. Evans, J.R., Barbanel, A (Eds.), 1999. Introduction to Quantitative EEG and Neurofeedback. Academic Press, San Diego. Glick, S.D., Ross, D.A, 1981. Lateralisation of function in the rat brain: basic mechanisms may be operative in humans. Trends Neurosci. 4, 196-199. Glowinski, J., Besson, M.l, Cheramy, A, 1984. Role of the thalamus in the bilateral regulation of dopaminergic and GABAergic neurons in the basal ganglia. In: Functions of the Basal Ganglia. London, Pitman. Ciba Foundation Symposium 107, 150-163. Gruzelier, lH., 1999a. Functional neuro-psychophysiological asymmetry in schizophrenia: a review and reorientation. Schizophr. Bull. 25 (1), 91-129. Gruzelier, J., 1999b. Implications of early sensory processing and subcortical involvement for cognitive dysfunction in schizophrenia. In: Tamminga, e.A (Ed.), Schizophrenia in a Molecular Age. American Psychiatric Press, Washington DC, pp. 29-69. Gruzelier, lH., Connolly, J.e., Eves, F.F. et aI., 1981. Effect of propranolol and phenothiazines on electrodermal orienting and habituation in schizophrenia. PsychoI. Med. 11,93-108. Gruzelier, J.H., Wilson, 1.., Liddiard, D., Peters, E., Pusavat, 1.., 1999a. Cognitive asymmetry patterns in schizophrenia: active and withdrawn syndromes and sex differences as moderators. Schizophr. Bull. 25 (2), 349-362. Gruzelier, J., Wilson, 1.., Richardson, A, 1999b. Cognitive asymmetry patterns in schizophrenia: retest reliability and modification with recovery. Int. J. Psychophysiol. 34(3), 323-331. Hammond, N.V., Gruzelier, J.H., 1978. Laterality, attention and rate effects in the auditory temporal discrimination of chronic schizophrenics: the effects of treatment with chlorpromazine. Q. J. Exp. Psychol. 30, 91-103. Hardman, E., Gruzelier, J., Cheesman, K. et aI., 1997. Frontal interhemispheric asymmetry: self regulation and individual differences in humans. Neurosci. Lett. 221, 117-120. Harvey, SA., Nelson, E., Haller, J.W., Early, TS., 1993. Lateralised attentional abnormality in schizophrenia is correlated with severity of symptoms. BioI. Psychiatry 33, 93-99.
Kay, S.R., Fiszbein, A, Opler, L.A, 1987. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr. Bull. 13, 261-276. Lopes da Silva, F., 1992. The rhythmic slow activity theta of the limbic cortex: an oscillation in search of a function. In: Baser, E, Bullock, TH. (Eds.), Induced Rhythms in the Brain. Maruff, P., Hay, D., Malone, V., Currie, J., 1995. Asymmetries in the covert orienting of visual spatial attention in schizophrenia. Neuropsychologia 33, 1205-1223. Mintz, M., Tomer, R., Myslobodsky, M., 1982. Neuroleptic-induced lateral asymmetry of visual evoked potentials in schizophrenia. BioI. Psychiatry 17 (24), 815-828. Othmer, S., Othmer, S.F., Kaiser, D.A, 1999. EEG biofeedback: an emerging model for its global efficacy. In: Evans, J.R., Barbanel, A (Eds.), Introduction to Quantitative EEG and Neurofeedback. Academic Press, San Diego, pp. 243-310. Pharr, O.M., Coursey, R.D., 1989. The use and utility of EMG biofeedback with chronic schizophrenic patients. Biofeedback Self-Regul. 14 (3), 229-245. Posner, M.E., Early, TS., Reiman, E., Pardo, P.J., Dhawan, M., 1988. Asymmetries in hemispheric control of attention in schizophrenia. Arch. Gen. Psychiatry 45, 814-821. Pycock, c.o., Kirwin, R.W., Carter, cr., 1980. The effect of lesion of cortical dopamine terminals on sub-cortical dopamine receptors in rats. Nature 286, 74-77. Rockstroh, B., Elbert, T, Birbaumer, N. et aI., 1993. Cortical self-regulation in patients with epilepsy. Epilepsy Res. 14, 63-72. Schneider, F., Rockstroh, B., Heimann, H., Lutzenberger, W., Birbaumer, N., 1992. Self-regulation of slow cortical potentials in psychiatric patients: schizophrenia. Biofeedback Self Regul. 17, 203-214. Schwartz, M.S., 1995. Biofeedback, 2nd ed. Guildford Press, New York. Tan, D., Gurgen, F., 1986. Modulation of spinal motor asymmetry by neuroleptic medication of schizophrenia patients. Int. J. Neurosci. 30, 165-172. Tomer, R., Flor-Henry, P., 1989. Neuroleptics reverse attention asymmetries in schizophrenic patients. BioI. Psychiatry 25, 852-860. Ungerstcdt, D., 1971. Stereotaxic mapping of the monoamine pathway in the rat brain. Acta Physiol. Scand. 82 (367), 1-48.
www.elsevier.com/locate/physcho
Learned control of slow potential interhemispheric asymmetry in schizophrenia John Gruzelier *, Elinor Hardman, Jennifer Wild, Rashid Zaman Department of Behavioural and Cognitive Sciences, Division ofNeuroscience and Psychological Medicine, Imperial College School ofMedicine at Charing Cross Hospital, University ofLondon, London, UK
Received 7 July 1998; received in revisedform 23 October 1998; accepted 5 January 1999
Abstract We report on the feasibility of teaching 16 (DSM-IV) schizophrenic patients, subdivided by syndrome, self-regulation of interhemispheric asymmetry having demonstrated efficient learning of interhemispheric control in normal subjects. Reversal of asymmetry may be important to treatment and recovery in schizophrenia for following improvement on neuroleptic drugs functional hemispheric asymmetries have reversed, with directions of reversal and pre-existing asymmetry dependent on syndrome. Asymmetry reversal in animals, manifested by spatial turning tendencies, has been used as a marker of neuroleptic action and involves striatal dopamine under reciprocal hemispheric control. We gave as feedback the left-right asymmetry in slow potential negativity recorded from the sensory motor strip (C3,4). Feedback took the form of a rocket on a screen which rose or fell with leftward or rightward shifts in negativity. Patients were able to learn control (P < 0.01). In those patients with lesser ability this was due to inability to sustain concentration throughout the session rather than slow initial learning. Active syndrome patients were better able to shift negativity rightward and withdrawn patients leftward, directions associated with drug reversal of functional asymmetry and symptom recovery for each syndrome. Accordingly our demonstration that many symptomatic schizophrenic patients are capable of learning control opens the door to electrocortical operant conditioning training in schizophrenia with therapeutic regimens. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Schizophrenia; EEG; Operant conditioning; Slow potentials; Laterality; Syndrome
* Corresponding author, Department of Behavioural and Cognitive Sciences, Imperial College School of Medicine, St Dunstan's Road, London W6 8RF. Tel.: +44-181-846-7386; fax: +44-181-846-1670. E-mail address:[email protected] (1. Gruzelier) 0167-8760/99/$ - see front matter © 1999 Elsevier Science B.Y. All rights reserved. PH: SOl 67- 8760 ( 9 9 ) 0 0 0 9 1 - 4
342
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
1. Introduction
There is an accumulation of evidence that hemispheric functional asymmetries in schizophrenia are dynamic and may reflect an imbalance of activity between the hemispheres rather than a unilateral anomaly (Gruzelier, 1999a). In a theoretical review we posited that underlying mechanisms were consistent with an imbalance in thalamo-cortical regulatory mechanisms having widespread neurophysiological and neurochemical influences throughout the hemispheres, not simply the more strongly innervated frontal thalamo-cortical regions (Gruzelier, 1999a). As cortical electroencephalographic activity (EEG) is under the control of thalamo-cortical mechanisms (Lopes da Silva, 1992), EEG measures are heuristic in measuring dynamic changes in functional imbalance. Changes in the direction of functional imbalance may be of therapeutic importance. Evidence has been reviewed showing that clinical improvement on neuroleptics has coincided with a reversal or shift in functional asymmetry (Gruzelier, 1999a), some of which were dose related (Hammond and Gruzelier, 1978; Mintz et a1., 1982; Harvey et a1., 1993). Measures showing drug reversal have included hemi-inattention (Hammond and Gruzelier, 1978; Posner et a1., 1988; Tomer and Flor-Henry, 1989; Maruff et a1., 1995), flash evoked potentials (Mintz et a1., 1982), lower motoneuron excitability (Tan and Gurgen, 1986), electrodermal orienting responses (Gruzelier et a1., 1981) and recognition memory (Gruzelier et a1., 1999a,b). Reversals of functional asymmetry with treatment in schizophrenia are consistent with animal research where a reversal of spatial turning tendencies was used as a marker of neuroleptic action (Ungerstedt, 1971; Pycock et a1., 1980; Glick and Ross, 1981). Glowinski et a1. (1984) have shown that the striatal dopamine pathways which are instrumental in these effects are under reciprocal hemispheric control, in keeping with neuroleptic influences on asymmetry reversa1. We reasoned that alteration of interhemispheric asymmetry through learned control may also benefit schizophrenic patients. In this feasi-
bility study we set out to explore whether schizophrenic patients could learn self-regulation of interhemispheric asymmetry through electrocortical conditioning biofeedback. The field of neurofeedback applications in psychopathology is developing as EEG mechanisms become better understood. Approaches have out grown the earlier work with alpha training with which the field was associated two decades ago, and which along with autonomic procedures was aimed at simply inducing relaxation (Evans and Barbanel, 1999). We chose to monitor cortical slow potential negativity encouraged by the results of Birbaumer and colleagues (Birbaumer et a1., 1991; Rockstroh et a1., 1993) who demonstrated that two-thirds of epileptic patients who were refractory to other forms of treatment benefited by learned control of cortical slow potentials and implemented self regulation when they felt a fit coming on. Slow cortical potentials reflect depolarisation of cortical pyramidal cells, in particular in layer IV, and index cortical excitability (Birbaumer et a1., 1990). In schizophrenia directions of functional hemispheric imbalance have been shown to be syndrome related (for review see Gruzelier, 1999a). An 'Active' syndrome was associated with a L > R asymmetry, the syndrome consisting of positive symptoms including overactivity, accelerated cognition or positive thought disorder, positive or labile affect, heightened self concepts and affective delusions. A 'Withdrawn' syndrome was associated with the opposite R > L asymmetry, consisting of affective blunting, social and emotional withdrawal and motor retardation. Here we subclassified patients by syndrome. The training programme was aimed at shifting the balance of cortical negativity interhemispherically. We had first developed training schedules with normal subjects and demonstrated that they could quickly learn interhemispheric control of frontal (F3,4) and central (C3,4) asymmetries (Hardman et a1., 1997). As larger effects were obtained with electrodes positioned centrally, central placements were adopted with schizophrenic patients. According to our model of hemispheric imbalance in schizophrenia it would be therapeutically advantageous if active syndrome patients with their L > R functional asymmetry
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
could shift activity rightward, away from their functionally dominant hemisphere and the direction associated with reversal of asymmetry with neuroleptic treatment. Conversely leftward shifts in negativity would be advantageous in withdrawn syndrome patients with their R > L asymmetry.
2. Methods
2.1. Subjects
Twenty-five medicated DSM-IV schizophrenic patients gave signed consent of whom 11 were male with an average age of 34.2 years. Psychiatrists who were independent of the laboratory measures rated the patients with Active, Withdrawn, or Mixed syndromes on the Schizophrenia Syndrome Scale (see Gruzelier et a1., 1999b). Four patients subsequently refused and four withdrew after 1-2 sessions: seven of the eight belonged to the Withdrawn syndrome. Sixteen cooperated in a training programme of up to 10 sessions. All completed six sessions: seven Active; four Withdrawn; and five Mixed syndrome; while 10 completed all 10 sessions. In two single case studies symptoms were monitored each session with the Positive and Negative Symptom Scale PANSS (Kay et a1., 1987). The average IQ of the patients was 104 and they had an average of 38 weeks total time in hospita1. 2.2. EEG conditioning procedures 2.2.1. Apparatus
Slow potentials were recorded with a grass polygraph from EEG Ag/AgCl electrodes placed at C3, C4 (10/20 system) and C4 linked mastoids and fastened with Elefix ge1. The electrooculogram (EOG) was recorded across the outer canthi and above and below the left eye. The biofeedback was contingent on the integrated C3-C4 signa1. Feedback took the form of an on-screen rocket ship, initially centrally placed, which rose as a function of an increase in left hemisphere negativity (relative to the right hemisphere) and fell to indicate a relative increase in right hemisphere negativity. The rocket was designed to
343
move in the vertical plane to avoid artefacts from lateral conjugate eye movements. Data were screened manually and on-line for eye and movement artefacts. 2.2.2. Procedure
Subjects were seated in a comfortable chair 140 em from the monitor on which the 10 x 6-cm rocket was displayed. Trials lasted 8 s with a 4-s interstimulus interva1. Artefact free trials were averaged for each subject according to trial type: 'A' instructing a leftward shift or 'B' a rightward shift. The mean DC shift between a 1-s pre-trial baseline and the last 6 s of the trial was calculated. Trials were presented in a pseudorandomised order. Patients were instructed to concentrate on contralateral body sensations to move the rocket. They were paid for participation and rewarded with lOp for each correct direction shift. The 60 trials were divided into blocks of 20, with rests between blocks and after every fifth trial when encouragement was given. There were up to 10 sessions on separate days. 2.3. Statistical analysis
The main statistical analyses consisted of two repeated measures analyses of variance with Greenhouse-Geisser conservative degrees of freedom. The first was with the 10 subjects who received all 10 sessions, the sessions were divided into two blocks of five sessions each - withinsubject factor Session/Block (2). Each session consisted of 60 trials divided into three blocks of 20 trials - within-subject factor Trial Block (3). Trials were subdivided according to whether leftward (A) or rightward (B) shifts in negativity were required within-subject trial type A/BLaterality (2). All patients who participated in training learned control, however, there were individual differences in performance. Accordingly the patients were divided into good and average performers based on consensual judgement between two independent raters through visual inspection of mean traces across blocks and individual sessions. This gave a between-subject factor - Performance (2). Interaction terms were elucidated by comparisons between means.
344
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
The second analysis examined syndrome relations in the larger group of 16 who had at least six sessions. In this analysis there was a betweensubject factor of Syndrome (Active/WithdrawnMixed), a within-subject factor of Session (6), but otherwise the same within-subject factors as in the first analysis.
3. Results 3.1. Training and performance
Means and standard deviations of good and average performers are shown in Fig. 1. There was clear evidence of the ability of schizophrenic patients to learn self regulation of interhemispheric negativity, shifting negativity to the left on A trials and to the right on B trials (Laterality: F = 28.22, d.f. = 1,8, P < 0.00l). It can also be seen that good performers had lateral shifts about twice as large as average performers (Group X Laterality: F = 4.48, d.f. = 1,8, P < 0.058). This effect was found to depend on session, for while there was evidence of a tendency towards improvement for the group as a whole across Ses-
sion/Block (F = 4.10, d.f. = 1,8, P < 0.077), there was a further interaction with Group (F = 10.23, d.f. = 1,8, P < 0.013). It was a falloff in performance over sessions in average performers, not slower initial learning, that differentiated good (Fig. 2) from average (Fig. 3) performers. There was also evidence common to both groups of improvement within sessions (Block x Laterality: F = 3.39, d.f. = 1.96, 15.26, P < 0.056), more so in the second Block of Sessions (F = 4.59, d.f. = 1.97, 15.78, P < 0.027). The interactions were examined further by analyses on each of the three blocks within sessions and, aside from comparing groups, comparing the effect of training by contrasting learning on the first five sessions with the remaining five. A tendency was shown for better performance with training in good performers in the first block (F = 4.98, d.f. = 1,8, P < 0.056), there were no group differences in the second block (F < 1.84), whereas in the third block there was a highly significant effect of training (F = 12.98, d.f. = 1,8, P < 0.007). The latter effect better characterised the good performers (F = 10.45, d.f. = 1,8, P < 0.012). From this may be inferred that reasons for reduced performance in average performers in-
Mean Leftward (a) and Rightward (b) Direction Shifts 6.00,----------------------------,
4.00
2.00
:;2-
]oo/--------r--Ul
c:
"'"
:;; -2.00 ~
"
...J
o Good Performers
-4.00
• Average Performers
-6.00 ' - - - - - - - - - - - - - (a) TRIALS
-------------' (b) TRIALS
Fig. 1. Mean leftward (Trial a) and rightward (Trial b) interhemispheric shifts in negativity for good and average performers showing significant (P < 0.001) regulation.
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348 Good Performers: Block I 15
~
1
\0
I
//~'-"-.'-.
",:IS
~ ~
"
I0
-15
1.1
~,
I
3,1
9.\
71
10,1
345
the course of six sessions. There were no syndrome differences in the various training parameters as shown by Block and Session effects (F < 1.92). T test comparisons indicated there was no difference between the Performance groups and syndrome score or total score (t = 0.21-1.38). On average the Active syndrome group produced larger rightward than leftward shifts in negativity: rightward, 3.24; leftward, 1.58 (t = 3.10, d.f. = 5, P < 0.027), while the Withdrawn syn-
BLOCK1
Good Performers: Block 2 15
&:
BLOCKl
BLOCK2
Good Performers: Block 3
Average Performers: Block 2 cr.
lSI
&:
10 ~
>
51
-------~~~
-: ///~'.,.,///~~-~-
o----::co~-~--------'=~=_-r
-~
~
............. -,
~ ~
4,3
5.3
6}
7.3
8.3
~ ~51
<,
~-
9}
<,
10.3
r.z
BLOCK3
2.2
3,2
4.2
.r'
5,2
6,2
72
8.2
9.2
BLOCK2
Fig. 2. Mean leftward and rightward shifts for good performers across 10 sessions for the three within-session blocks of trials.
Average Performers: Block 3
fc
:2
eluded a loss of concentration towards the end of the sessions.
"',.,~-~-~,.
i
\01
I
~" ~
~~~-<~.~~ -, "
3.2. Syndrome
Both syndromes learned self control of interhemispheric negativity (F = 19.71, d.f. = 1,14, P < 0.001). A highly significant trial effect (Fl 14 = 19.71, P < 0.001) indicated that schizophrenic patients could learn interhemispheric control in
"
1.3
5.3
9.3
10J
BLOCK3
Fig. 3. Mean leftward and rightward shifts for average performers across 10 sessions for the three within-session blocks of trials.
346
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
drome produced larger leftward shifts: rightward, 1.84; leftward, 2.37 (t = 2.75, d.f. = 4, P < 0.07), as did the Mixed syndrome: rightward, 2.52; leftward, 3.69 (t = 3.196, d.f. = 4, P < 0.034); combining mixed/withdrawn (t = 0.47, d.f. = 5, P < 0.002). A median split on total symptom score (N = 12, seven sessions) did not show a clear relation between severity of symptoms and learned control (P < 0.139). Correlations were examined between each of the active, withdrawn and unreality syndrome scores and negativity shifts on leftward and rightward trials across the first five trials. There was a correlation (r = 0.60, P < 0.05) between Active syndrome rating score and leftward shift. The higher the Active rating the smaller the leftward shift, evidence of a possible ceiling effect in directional shift.
Mean A & B Shifts
I
./ I
2
)
4
5
6
s
7
9
'\..--.-
10
50
40
3.3. Case studies
~
..
POSIT1VE NEGATIVE GENERAL
In a patient who undertook 20 sessions of train-
ing interhemispheric shifts continued to grow in size after the 10th session, and an impressive ability to respond differentially with leftward or rightward shift trial type was sustained throughout the further 10 sessions (see Fig. 4). In other cases symptoms were monitored concurrently with the PANSS. General symptoms such as anxiety and depression appear to be more closely related to neurofeedback ability than specific psychotic symptoms. One patient first undertook the 10 training sessions with limited success. During this time General Psychopathology scores SINGLE CASE STUDY: Mean Leftward (A) and Rightward (B)
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
session
Fig. 5. A patient who resumed training after a 3-month absence following session 10. Upper figure leftward (A) and rightward (E) directional shifts in microvolts. Lower figure total positive, negative and general psychopathology ratings on the PANS scale.
were above the 50th percentile. Then after a break of 3 months, nine more sessions were undertaken. The General Psychopathology rating total had dropped by 50% (see Fig. 5 lower). Upon resuming training, a huge 18 f.LV differentiation was obtained on the first occasion (see Fig. 5 upper). This was then maintained for the further seven sessions.
Direction Shifts Across 20 Sessions
4. Discussion
'3 ......l
.10'-1_-,----,....--,_ _-,----_---,,-_-,----_,.---,._-,----,........, 2 3 4 5 6 7 s 9 10 11 12 13 14 15 16 17 III 19 20
Mean B
Session
Fig. 4. Mean leftward and rightward shifts in a patient undergoing 20 sessions.
Contrary to the belief that schizophrenic patients will be unable to learn self control of electrocortical activity due to attentional and motivational deficits, all patients who cooperated with training were able to learn to control interhemispheric negativity between lateralised central electrodes (C3/C4). This constituted two-thirds
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
of our symptomatic schizophrenic patients who initially agreed to participate and gave signed consent. Another study also involving operant conditioning of slow cortical potentials, but requiring the increasing and decreasing of negativity at Cz, had moderate success (Schneider et a1., 1992). Withdrawal was the feature that characterised those patients who did not cooperate in our study, though withdrawal itself did not preclude patients from cooperating. In medical students levels of withdrawal akin to introversion were found an advantage in learning interhemispheric control (Hardman et a1., 1997). At the other extreme, severity of positive symptoms also may not be a handicap to learning. It was nonspecific symptoms such as anxiety and distractibility that were more of a handicap to cooperation than positive symptoms. In medical students tiredness was disadvantageous to successful training while calmness was beneficia1. When examining individual differences in performance the greater difficulties in control were found with sustaining concentration towards the end of sessions rather than in initial learning. This suggests that shorter sessions and/or stronger reinforcers would facilitate training. The evidence that non-specific symptoms such as tension and anxiety may be counterproductive to neurofeedback training raises the question whether training first in stress reduction through peripheral biofeedback procedures such as EMG might facilitate the self regulation of electrocortical activity. It is common practice in other fields such as optimal performance training to precede neurofeedback with biofeedback stress reduction protocols and there are reports of successful stress reduction with biofeedback in schizophrenia (Schwartz, 1995; Pharr and Coursey, 1989). Currently there are growing applications of neurofeedback with a range of psychopathologies and applications for the enhancement of sporting and musical performance (Evans and Barbanel, 1999). There is no reason to suspect that success in demonstrating learned control of slow cortical potentials would not generalise to other more widely practised forms of training with EEG rhythms that have been successful in treating dysregulation of attentional, motivational and
347
arousal processes (Othmer et a1., 1999). As recently concluded in a review of early processing deficits and dysregulation of thalamo-cortical functions in schizophrenia (Gruzelier, 1999b; Gruzelier et a1., 1999a), there is a body of evidence including well documented abnormalities of slow to fast EEG rhythms, and which may extend to fast transients in the gamma range (Baldeweg et a1., 1998), which may respond to correction through operant conditioning procedures. Of theoretical and therapeutic importance was the demonstration that schizophrenic patients could learn to shift negativity away from their functionally dominant hemisphere: a rightward direction in the Active syndrome who have leftsided functional biases and a leftward direction in the Withdrawn syndrome who have right-sided functional biases. This indicates that the premise underlying the training of reversing functional biases in asymmetry through learning self-regulation is feasible in practice. The lack of literature on applying neurofeedback schedules to schizophrenia has represented the belief that schizophrenic patients would be unable to cooperate with training. The significance of the present study is that it removes this shibboleth, though patient motivation must be carefully monitored. Having demonstrated the feasibility of operant conditioning with schizophrenia, the way is clear to examine the efficacy of therapeutic intervention in schizophrenia, schizotypy and the schizophrenia spectrum.
Acknowledgements
The research was supported by a Sr investigator award from NARSAD and the L.F. Saugstad fund. Presented to the 5 th Cognitive Neuroscience Society meeting, San Francisco, April, 1998. References Baldeweg, T., Spence, S., Hirsch, S.R. et aI., 1998. Gammaband electroencephalographic oscillations in a patient with somatic hallucinations. Lancet 352, 620-621.
348
1. Gruzelieret al. / International Journal ofPsychophysiology 34 (1999) 341-348
Birbaumer, N., Elbert, T, Canavan, AG.M., Rockstroh, B., 1990. Slow potentials of the cerebral cortex and behaviour. Physiol. Rev. 70 (1), 1-30. Birbaumer, N., Elbert, T, Rockstroh, B., Daum, I., Woolf, P., Canavan, A, 1991. Clinical-psychological treatment of epileptic seizures: a controlled study. In: Ehlers, A et al.Gid.), Perspectives and Promises of Clinical Psychology. Plenum Press, New York. Evans, J.R., Barbanel, A (Eds.), 1999. Introduction to Quantitative EEG and Neurofeedback. Academic Press, San Diego. Glick, S.D., Ross, D.A, 1981. Lateralisation of function in the rat brain: basic mechanisms may be operative in humans. Trends Neurosci. 4, 196-199. Glowinski, J., Besson, M.l, Cheramy, A, 1984. Role of the thalamus in the bilateral regulation of dopaminergic and GABAergic neurons in the basal ganglia. In: Functions of the Basal Ganglia. London, Pitman. Ciba Foundation Symposium 107, 150-163. Gruzelier, lH., 1999a. Functional neuro-psychophysiological asymmetry in schizophrenia: a review and reorientation. Schizophr. Bull. 25 (1), 91-129. Gruzelier, J., 1999b. Implications of early sensory processing and subcortical involvement for cognitive dysfunction in schizophrenia. In: Tamminga, e.A (Ed.), Schizophrenia in a Molecular Age. American Psychiatric Press, Washington DC, pp. 29-69. Gruzelier, lH., Connolly, J.e., Eves, F.F. et aI., 1981. Effect of propranolol and phenothiazines on electrodermal orienting and habituation in schizophrenia. PsychoI. Med. 11,93-108. Gruzelier, J.H., Wilson, 1.., Liddiard, D., Peters, E., Pusavat, 1.., 1999a. Cognitive asymmetry patterns in schizophrenia: active and withdrawn syndromes and sex differences as moderators. Schizophr. Bull. 25 (2), 349-362. Gruzelier, J., Wilson, 1.., Richardson, A, 1999b. Cognitive asymmetry patterns in schizophrenia: retest reliability and modification with recovery. Int. J. Psychophysiol. 34(3), 323-331. Hammond, N.V., Gruzelier, J.H., 1978. Laterality, attention and rate effects in the auditory temporal discrimination of chronic schizophrenics: the effects of treatment with chlorpromazine. Q. J. Exp. Psychol. 30, 91-103. Hardman, E., Gruzelier, J., Cheesman, K. et aI., 1997. Frontal interhemispheric asymmetry: self regulation and individual differences in humans. Neurosci. Lett. 221, 117-120. Harvey, SA., Nelson, E., Haller, J.W., Early, TS., 1993. Lateralised attentional abnormality in schizophrenia is correlated with severity of symptoms. BioI. Psychiatry 33, 93-99.
Kay, S.R., Fiszbein, A, Opler, L.A, 1987. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr. Bull. 13, 261-276. Lopes da Silva, F., 1992. The rhythmic slow activity theta of the limbic cortex: an oscillation in search of a function. In: Baser, E, Bullock, TH. (Eds.), Induced Rhythms in the Brain. Maruff, P., Hay, D., Malone, V., Currie, J., 1995. Asymmetries in the covert orienting of visual spatial attention in schizophrenia. Neuropsychologia 33, 1205-1223. Mintz, M., Tomer, R., Myslobodsky, M., 1982. Neuroleptic-induced lateral asymmetry of visual evoked potentials in schizophrenia. BioI. Psychiatry 17 (24), 815-828. Othmer, S., Othmer, S.F., Kaiser, D.A, 1999. EEG biofeedback: an emerging model for its global efficacy. In: Evans, J.R., Barbanel, A (Eds.), Introduction to Quantitative EEG and Neurofeedback. Academic Press, San Diego, pp. 243-310. Pharr, O.M., Coursey, R.D., 1989. The use and utility of EMG biofeedback with chronic schizophrenic patients. Biofeedback Self-Regul. 14 (3), 229-245. Posner, M.E., Early, TS., Reiman, E., Pardo, P.J., Dhawan, M., 1988. Asymmetries in hemispheric control of attention in schizophrenia. Arch. Gen. Psychiatry 45, 814-821. Pycock, c.o., Kirwin, R.W., Carter, cr., 1980. The effect of lesion of cortical dopamine terminals on sub-cortical dopamine receptors in rats. Nature 286, 74-77. Rockstroh, B., Elbert, T, Birbaumer, N. et aI., 1993. Cortical self-regulation in patients with epilepsy. Epilepsy Res. 14, 63-72. Schneider, F., Rockstroh, B., Heimann, H., Lutzenberger, W., Birbaumer, N., 1992. Self-regulation of slow cortical potentials in psychiatric patients: schizophrenia. Biofeedback Self Regul. 17, 203-214. Schwartz, M.S., 1995. Biofeedback, 2nd ed. Guildford Press, New York. Tan, D., Gurgen, F., 1986. Modulation of spinal motor asymmetry by neuroleptic medication of schizophrenia patients. Int. J. Neurosci. 30, 165-172. Tomer, R., Flor-Henry, P., 1989. Neuroleptics reverse attention asymmetries in schizophrenic patients. BioI. Psychiatry 25, 852-860. Ungerstcdt, D., 1971. Stereotaxic mapping of the monoamine pathway in the rat brain. Acta Physiol. Scand. 82 (367), 1-48.