Alterations in Cerebral Laterality During Acute Psychotic Illness Bruce E. Wexler, MD,

\s=b\ A

dichotic

George

R.

Heninger,

MD

listening test was used to assess cerebral lateral-

ity in 26 right-handed patients with schizophrenic, schizoaffective, or primary major depressive illness and in 23 controls. Clinical state was assessed by twice-daily nurses' ratings and patient self-ratings. Ratings of psychotic thought and behavior were lower during the week of highest laterality than during the week of lowest laterality (P < .01). Similarly, when most improved, patients had higher laterality than when most ill (P < .01). Changes in laterality were not specific to diagnostic group, were not present in control subjects, could not be related to direct drug effects, and were independent of changes in accuracy of performance. There were large, stable, interindividual differences in degree of lateralization, but no differences between patients and controls. These results are consistent with the hypothesis that in acute psychotic illness there is a breakdown in the interhemisphere inhibition that normally mediates

laterality. (Arch Gen Psychiatry 36:278-284, 1979)

cerebral

of lateral specialization of the human brain have disclosed that in most right-handed individ¬ uals the left cerebral hemisphere is specialized for the performance of verbal, analytic, linear, and logical tasks, while the right hemisphere appears to be specialized for spatial, geometric, and synthetic tasks.13 During normal function, it is thought that functional dominance shifts from hemisphere to hemisphere, depending on the type of task encountered. Although in the past many studies of hemispheric specialization have relied on the integration of information from neurosurgical and brain-damaged indi-

Investigations

Accepted

for publication June 18, 1978. From the Department of Psychiatry, Yale University, and the Connecticut Mental Health Center, New Haven. Reprint requests to Department of Psychiatry, Yale University School of Medicine, 34 Park St, New Haven, CT 06508 (Dr Wexler).

viduals, new techniques based on the selective presentation of test stimuli to a single hemisphere now make it possible to assess single-hemisphere function, patterns of function¬ al organization, and interhemisphere transfer of informa¬ tion in subjects with intact brains. These techniques can be employed to investigate hemispheric specialization in psychiatric patients with the goal of obtaining new infor¬

mation on the brain processes involved in the manifesta¬ tion of illness. A number of investigators, relying primarily on older techniques and indirect indices, have previously examined cerebral functional asymmetry in psychiatric patients. Looking at hand-eye dominance in psychiatric inpatients. Quinan4 found that patients with "constitutional psychiat¬ ric state" had an increased incidence of left-handedness and of mixed hand-eye dominance, and that schizophrenic patients had an incidence of mixed hand-eye dominance between that of these patients and that of normal subjects. Subsequent investigators also found an increased inci¬ dence of mixed dominance in psychotic patients,5·6 although other studies have found an increased incidence of left-handedness but not an increased incidence of mixed dominance in groups of schizophrenics.7 9 In endogenous depression, the variability of EEG amplitude recorded over the left hemisphere is one half of that recorded over the right. With treatment and symptom remission, the vari¬ ability of the left increases to equal that on the right. In addition, the greatest memory impairment was found in those patients with the greatest reduction of left hemi¬ sphere voltage-amplitude variability.1" The EEG amplitude over the left hemisphere in schizophrenic patients is reported to increase with chlorpromazine but not with

haloperidol.11 When skin conductance was used as a measure of activity of the ipsilateral cerebral hemisphere,

Downloaded From: http://archpsyc.jamanetwork.com/ by a Thomas Jefferson University User on 08/01/2017

schizophrenic patients were reported as showing less responsiveness on the left than on the right, while the opposite was true for depressed patients.12 Patients with less language lateralization have more severe depressions than patients with strong language lateralization, as measured by intracarotid amobarbital injection.13 A recent study of interhemispheric relationships with somatic ther¬ apy suggested that electroconvulsive therapy (ECT) was associated with improvement in nonverbal task perform¬

and with deterioration of tasks of verbal skills. Shifts in interhemispheric relationships on the test measures used were not found with phenothiazine treatment. Inter¬ actions between EEG slowing and ECT were observed, and there was a persistant increase in the EEG energy content over the dominant hemisphere when unilateral ECT was given to either side of the head.14 One study has been reported using the more direct techniques of unilateral sensory stimulation to investigate lateral brain function in psychiatric patients. Beaumont and Dimond13 presented visual stimuli selectively to each hemisphere. They found that schizophrenic patients did as well as normal or psychiatric controls in simple identifica¬ tion tasks, but did worse when matching letters presented to the left hemisphere, and in matching digits and shapes presented to the right. When required to match across the hemispheres, their performance fell further below that of the control groups for both letters and shapes. These studies suggest that disorders of cerebral lateral¬ ity may be associated with psychiatric illness. Some results indicate dysfunction in a single hemisphere, while others suggest abnormal patterns of functional organization or impaired interhemisphere integration. All of these find¬ ings need replication. In addition, the laterality variables need to be studied more thoroughly by relating them to

Eight were diagnosed as suffering from schizo¬ phrenia (five men, three women; median age, 28 years; range, 19 to 41 years); six, schizo-affective disorder (three men, three women; median age, 30 years; range, 27 to 51 years); and 12, primary major depressive disorder (four men, eight women; median age, 35 years; range, 18 to 60 years). The median age of the controls was 25 years, age of 32 years.

with a range of 17 to 38 years. There were nine men among the control subjects. The median length of hospital stay for the patients was 79 days (range, 13 to 234 days).

ance

improved diagnostic classification, quantified measures of acute symptomology, drug treatment, and changes in clinical state. In an initial study designed to approach these problems, we have repeatedly administered a dichotic listening test of cerebral laterality to patients and healthy controls. In the patients, this was done both prior to and during periods of pharmacologie treatment and prior to and during periods of symptom change. This made it possible to assess differences in cerebral laterality between periods of relative illness and health. By comparing the results to clinical symptom ratings, it was possible to look for relationships between changes in laterality and changes in specific symptom dimensions. SUBJECTS AND METHODS

Subjects Twenty-six consecutive patients admitted to the Research Ward of the Connecticut Mental Health Center, New Haven, Conn, who met research criteria for the diagnosis of schizophrenia, schizoaffective illness, or primary major depressive disorder'" and who right-handed (determined as described later) were studied. Twenty-three right-handed staff members and college students who had no history of psychiatric illness or psychiatric hospitaliza¬ tion served as controls. After a description of the procedures, the subject and, in the case of patients, an immediate family member gave informed written consent for participation in the studies. The patients ranged in age from 18 to 60 years, with a median were

Research

Design

Following admission, most patients had an initial 7- to 14-day drug-free evaluation period prior to starting psychotropic medica¬ tion. Patients then began a course of chlorpromazine, haloperidol, thioridazine, lithium carbonate, amitriptyline, imipramine, or desipramine. For various clinical reasons, some patients had a shorter drug-free period, were treated with more than one medi¬ cation, or received no medication. Within the first 14 days of hospitalization, usually before drug treatment, patients began participating in the dichotic listening task, which was generally administered on two successive days once a week throughout the hospital stay. The median number of test sessions administered to patients was 16 (range, 5 to 43). Eleven of the control subjects had a similar twice-weekly testing session in each of three to five different weeks over a nine-week period. The remaining 12 control subjects were tested only two to four times in a one-week period. Behavioral Measures Once during the 7 am to 3 pm shift and once during the 3 to 11 pm shift, patients were rated by nursing staff on a previously described 14-item clinical rating scale (0, absent; 8, severe).17 At 9 am and 7 pm, patients rated themselves from 0 to 4 on a 29-item modified Profile of Mood States (0, not at all; 4, extremely).18 At 7 pm, patients rated themselves from 0 to 4 on a 15-item symptom and side effect checklist (0, not at all; 4, very much). These ratings continued from the time of admission throughout the hospitaliza¬ tion. Each of the 29 self-rating items was added to other similar items to generate a score on 7 different highly interrelated factors generally described as depression, anxiety, anger, fatigue, increased activity, confusion, and friendliness. The sum of all 15

symptoms and side effect items side effect

was

used for the symptom and

score.

Laterality Measures language-related dichotic listening task was used to assess laterality. The task consisted of reporting which of six possible nonsense syllables was heard following the simultaneous stimulation of each ear with different stimuli. The six syllables were formed from the six English stop consonants b, d, g, p, t, and k preceded and followed by the vowel a. The stimuli were hybrid utterances composed of synthetic vocalic portions and natural aspirated portions and modified on the Haskins laboratory pulse code modulation system. Each stimulus pair was preceded by the orienting statement, "Now write..." (eg, "Now write aba"). There was a two-second pause between stimulus pairs. Each of the six syllables was paired with every other to make 15 pairs. Each pair was presented twice within a 30-pair sequence, the second time with each syllable to the ear opposite that of the first presentation. The order of the presentation was randomized and adjusted to eliminate the presence of three or more identical stimuli in succession. Two 30-pair sequences were combined to A

cerebral

make

one

60-item tape. Three 60-item tapes with different

Downloaded From: http://archpsyc.jamanetwork.com/ by a Thomas Jefferson University User on 08/01/2017

random sequences were utilized in the study so that they could be rotated from day to day to reduce the possibility that the subject would memorize the sequence. Tapes were played on a profession¬ al-quality, two-track stereo tape machine through a pair of stereo headphones matched for auditory response characteristics. In each session, the test tape was played twice to present a total of 120 stimulus pairs. The earphones were switched after 30 and again after 90 presentations. This balanced for any differences between the two channels in the audio equipment and ensured that neither ear received the same sequence of stimuli twice in one test day. Subjects were instructed to report which of the syllables they heard on each trial. A prepared answer sheet had space for only one response even though the subjects heard two stimuli. They were told that if it sounded as if they heard two different syllables, they were to mark the second syllable to the side of the first (no subject gave two responses to more than one third of the trials and most did not give any). Only first answers were used in scoring. An error was scored when the subject did not report either of the two stimuli delivered on that trial. A laterality score was calculated by subtracting the left ear responses from the right ear responses, dividing the difference by the sum of the right ear plus the left ear responses, and multiplying the quotient by 100.19·2" Before the initial test session, one track of the stereo tape machine was turned off and each ear tested monaurally with 15 stimuli. This reduced any confusion about the test and eliminated mistakes in identification of the stimuli, which could be pointed out by the examiner. Specific misidentified syllables were played for the seven patients who made several errors on the same syllable in the monaural test. Those patients who made several different errors were then given 15 more practice identifications presented to each ear singly. At the completion of this session, all but six patients were more than 90% correct in monaural identifi¬ cation. Auditory acuity was tested in each ear in all subjects by the method of ascending and descending limits, using tones of 250, 500, 1,000, 2,000, 4,000 cps. No subject had a difference in auditory acuity between the ears of more than 10 db on more than two of the five frequencies tested. When the patient was most improved, handedness was assessed by observing which hand was used in eight tasks: writing,

throwing a ball, dealing cards, opening a jar, threading a needle, brushing teeth, cutting with scissors, and striking a match. Subjects who wrote and did at least five of the other seven tasks with the

same

hand

were

classified to be

Data

so

handed.

Analysis

reliability,

the test scores for the two successive administered each week were averaged. The behavioral measures were also averaged for the corresponding time period. Independent ' tests were used for statistical evalua¬ tion of mean differences between groups, paired t tests to evaluate change within subjects over time, and Pearson correlation coeffi¬ cients to evaluate relationships among the variables. Changes in variables with abnormal distributions were evaluated with nonTo increase

days the

test

was

parametric tests.

Reliability

Distribution of controls.

scores

in

hospitalized patients and

though this covered a period of nine weeks. These data indicate that the laterality scores are highly reliable, that there is no significant first-time effect for either patients or controls, and that in controls there is no long-term trend over time. Five patients had negative laterality scores throughout the hospitalization. They could be considered left ear dominant even though they were right-handed, since during the first two weeks of testing their average score was more negative than would be expected by chance at the < .05 level ( 2). An increase in their degree of lateralization would result in their laterality scores becoming more negative. For this to be considered an increase in laterality in the statistical analyses, and for it not to obscure the magnitude of the positive increases recorded in the right eardominant patients, the data were analyzed aisecond time, using revised laterality scores for these five patients that were arrived at by simply assigning a positive sign to them. even

RESULTS

and First-Time Effects

The reliability was assessed for the laterality scores the first and second days the test was given. The test-retest correlations for the 26 patients and the 23 controls were .89 and .91, respective¬ ly. The mean laterality score on the first and the second test days were 9.3 and 7.4, respectively, for the patients and 13.3 and 14.3, respectively, for the controls, both nonsignificant changes. In the 11 control subjects who took the test repeatedly, there was no mean difference from the first to the last time the test was taken

laterality

Comparison of Patients

to Controls

The laterality scores for the first week the test was given illustrated in the Figure for the 23 controls and the 26 patients. There was no significant difference between the mean laterality scores (including the revised score) of the controls and patients. In addition, there were no signifi¬ cant differences in laterality among the diagnostic subgroups of schizophrenia, schizo-affective disorder, and are

Downloaded From: http://archpsyc.jamanetwork.com/ by a Thomas Jefferson University User on 08/01/2017

Table

Table 1.—Behavioral Ratings and Test Errors During Week of Lowest and Highest Laterality

Nurses'

retardation

Physical complaints Hyperactivity elation

Psychomotor agitation Verbal anxiety Expressed anger

Time of

Highest Laterality (Mean ± SD)

Highest Difference*

Illness

Time of Lowest Illness

(Mean ± SD)

Rating (Mean + SD)

Rating (Mean ± SD)

7.7 ±20.1

10.7 ±22.9

3.0

±

7.6t

13.6 ±17.6

17.6 ±18.2

4.0

±

7.1*

45.8

34.7 ±17.5

—11.1

±18.2$

1.1 3.7: 3.1 ± 1.0 0.7: 0.9

-2.1

0.8§

or

4.5 : 3.9:

1.5 1.5

2.9: 1. 1.8 ± 1.5

-0.9: -0.7:

0.6:

1.3

0.6:

1.0

0.0:

1.0

1.9:

1.4 1.5

-0.5: -0.6:

2.0

3.3

1.4t

1.0

-1.0: -1.0: -0.7:

1.8 1.5

-1.0: -1.1

2.3Î

1.1

-0.2

1.0

-0.5: -0.2 :

3.0: 2.6+

1.1

1.6 1.4

:

1.1

1.7

±

2.0: 1.6: 0.4:

1.7

Confusion

1.2: 1.5: 1.0:

1.0 0.8 0.9 0.8 0.9

1.0: 1.2: 1.0 0.8+ 0.7 1.5 ± 1.0 1.0

9.8

17.8 ±10.3

42.2 ±20.4 1.2 ±22.6

37.7 ±15.4 16.6 ±25.2

Anxiety Anger Increased

activity

1.3

±

-0.1 1.4 —0.1+ 1.

physical symptoms (0, ab¬ sent; 60, severe) 20.8 Laterality measures Test errors (per

±

Friendliness

:

1.1

1.1Í

pairs) Laterality

score

highest laterality

minus time of lowest

fP < .01 (paired f test, two-tailed). JP< .05.

-3.0 ±10.1

-4.5

±

15.4.

19.3

9.9II

laterality.

§P