- Email: [email protected]
Interactive behavior in bipolar manic and schizophrenic patients and its link to thought disorder
Interactive Behavior in Bipolar Manic and Schizophrenic Patients and Its Link to Thought Disorder L i n d a S. G r o s s m a n a n d M a r t i n H a r r o w
The research was designed to assess whether bipolar manic patients are unique in showing excessive interactive behavior and whether interactive behavior is linked to positive thought disorder in mania and other psychotic disorders. We compared the interactive behavior of manic patients versus schizophrenics and nonpsychotic patients. Degree of interactive behavior and severity of thought disorder were assessed in an inpatient sample of 140 patients, including 38 bipolar manic patients, and control samples of 40 acute schizophrenic patients, 32 nonpsychotic patients, and 30 chronically hospitalized schizophrenic patients. The assessment battery used standardized tests to assess interactive behavior and thought disorder. Manic patients displayed significantly more interactive behav-
ior on all three tests than the other three patient groups. Interactive behavior made a greater contribution to the thought disorder of manic patients than to the thought disorder of the other three diagnostic groups. The other two acute patient groups also showed relationships between interactive behavior and thought disorder. The chronic schizophrenics did not show this pattern. The data indicated that many types of acute patients show some interactive behavior, but bipolar manic patients demonstrate more excessive interactive behavior than do other patients. There is a closer link between excessive interactive behavior and thought disorder in manic patients.
HE C U R R E N T R E S E A R C H was designed to study the frequency of interactive behavior in patients with bipolar affective disorders and to provide information about the relationship between interactive behavior and thought disorder. A high frequency of intrusive or interactive behaviors is one of the more common clinical observations in mania and is not observed as often in other psychotic disorders. An important question that emerges is, if manic patients show excessively interactive behavior, is the thought disorder frequently reported in these patients linked to their interactive behavior and, if so, is this linkage specific to manic patients or is it found in other psychotic patients? Since interactiveness and thought disorder are frequently observed by psychiatrists treating manic patients, the occurrence of both and their potential linkage are important for clinical practice and for theory about bipolar disorders. Thought disorder and cognitive abnormality have historically been viewed as important symptoms of schizophrenia. 1 At one time, thought disorder was seen both as the central pathognomonic symptom of schizophrenia, and as rare or nonexistent in other psychiatric disorders. 2However, the past 20 years have produced a wealth of evidence indicating that many manic patients have thought disorder as severe as that found in schizophrenic patients during the period of acute ~lness. 3-8 Additional research has suggested that some manic patients show severe
thought disorder when evaluated years after the acute hospitalization phase. T M Research seeking to differentiate thought disorder in manic patients from thought disorder in schizophrenic patients has produced mixed results. Several recent studies have found positive results. 3-8A1-14 Various qualities of mania have been cited as being involved in manic thought disorder, such as racing thoughts, accelerated speech, overproductive ideation, excessive jocularity and wordplay, and overinclusive behavior. 4,8,11,15-19Many of these qualities have not been studied systematically, but rather are based on anecdotal evidence. A systematic study of some of these qualities that may be involved in manic thought disorder can provide clues as to the pathological factors involved in disordered thinking, and can also be helpful in furnishing important information about similarities and differences in the thought disorder found in manic and schizophrenic patients. This type of analysis can provide data concerning the nature of thought disorder in major psychiatric disorders and can
T
Copyright © 1996 by W.B. Saunders Company
From the Department of Psychiatry, University of Illinois College of Medicine. Supported in part by Grants No. MH-26341 and MH-30938 from the National Institute of Mental Health, and by research grants from the John D. and Catherine T. MacArthur Foundation and Michael Reese Hospital. Address reprint requests to Linda S. Grossman, Ph.D., Department of Psychiatry (M/C913), 912 S. Wood St, Chicago, IL 60612-7327. Copyright © 1996 by W.B. Saunders Company 0010-440X/96/3704-0003503. 00/0
ComprehensivePsychiatry, Vol. 37, No. 4 (July/August), 1996: pp 245-252
245
246
GROSSMAN AND HARROW
p r o d u c e useful distinctions b e t w e e n diagnostic categories. O n e quality of m a n i c p a t i e n t s that has b e e n r e p o r t e d f r e q u e n t l y is an increase in sociability a n d interactive behavior, t5-19 W h a t factors are involved in m a n i c p a t i e n t s ' excessive interactiveness? It is possible that features that may b e involved in m a n i c p a t i e n t s ' cognitive a b e r r a tions a n d t h o u g h t disorder, such as lack of inhibition, s e l f - m o n i t o r i n g problems, a n d o t h e r types of reality distortions, play a role in their p o t e n t i a l excessive interactive behavior. T h e q u e s t i o n also arises as to w h e t h e r the link b e t w e e n interactive behavior a n d t h o u g h t disorder is p a r t i c u l a r to m a n i c patients, as o p p o s e d to o t h e r types of acutely ill p a t i e n t s such as schizophrenics a n d p a t i e n t s with o t h e r forms of psychopathology. A n assessment of the relationship b e t w e e n t h o u g h t disorder a n d interactive b e h a v i o r in m a n i c p a t i e n t s a n d schizophrenic p a t i e n t s would be a first step toward investigating these issues. It is also possible that interactive behavior is n o t associated with t h o u g h t disorder, b u t r a t h e r is part of the overall m a n i c s y m p t o m p i c t u r e along with o t h e r n o t e d qualities of m a n i a such as jocularity or elation. T h e c u r r e n t report a t t e m p t s to provide inform a t i o n o n these issues by focusing o n the potential relationship b e t w e e n interactive behavior a n d t h o u g h t disorder in m a n i c p a t i e n t s as c o m p a r e d with schizophrenic a n d n o n p s y c h o t i c patients. Specifically, this r e p o r t addresses the following questions: (1) D o most m a n i c p a t i e n t s show interactive characteristics? (2) Is their interactive b e h a v i o r related to the severity of their t h o u g h t disorder? (3) Does the interactive quality occur in m a n i c p a t i e n t s who do n o t show t h o u g h t disorder? (4) Does the interactive behavior occur in o t h e r p a t i e n t s such as acutely ill schizophrenics, chronic schizophrenics, a n d n o n psychotic psychiatric p a t i e n t s and, if so, is it linked to t h o u g h t disorder in these o t h e r patients? METHOD
Patient Population The psychiatric population consisted of 140 currently hospitalized patients. All patients were diagnosed according to the Research Diagnostic Criteria,2° based on the Schedule for AffectiveDisorders and Schizophrenia,21the Present State Examination,22and the Schizophrenic State Inventory SSI. 23'24These standardized semistructured diagnostic inter-
views were conducted by trained raters who were blind to the results of tests used to assess interactiveness and thought disorder. Research diagnoses were assigned for each patient by senior clinicians who achieved an interrater reliability Kof .85.24Patients were screened to include only those with bipolar disorder, schizophrenia, and nonpsychotic psychiatric disorders, excluding patients with overt organic brain disorders. This sample included 38 manic patients, 40 acute schizophrenic patients, and 32 nonpsychotic patients admitted to the Illinois State Psychiatric Institute and Michael Reese Hospital. For comparison purposes, an additional 30 chronic schizophrenic patients who had been hospitalized for at least 8 of the past 10 years at a long-term state hospital were assessed. In the current report, we use the term "acute patients" to differentiate recently hospitalized manic, schizophrenic, and nonpsychotic patients from the sample of 30 chronically hospitalized patients. The mean age of the acute patients was 24.8 _+ 6.7 years (mean +__SD). This research was conducted as part of the Chicago Follow-up Study. 6,9,10,25-30 Forty-eight percent of the patients were women, 64% were white, and 36% were black. Patients had a mean educational level of 12.3 years. Patients in the three acute diagnostic groups achieved a mean age-corrected scale score of 10.5 on the information subtest of the Wechsler Adult Intelligence Scale,31 which provided an estimate of IQ scores. This mean score is within the average range of intellectual functioning in the general population. The chronically hospitalized schizophrenic patients had significantly lower scores on the information subtest than the other three groups, with a mean score of 5.9 (F = 22.5, df = 3,136, P < .001). To control for differences in IQ, we conducted a separate analysis of covariance with IQ as the covariate on each of the major variables studied. The results indicated that after IQ differences were controlled, there remained significantdiagnostic differences both for interactiveness (F = 22.18, df = 3,135, P < .001) and for thought disorder (F = 12.60,df = 3,128, P < .001). The mean number of previous hospitalizations for the acute samples was 1.5. The three nonchronic diagnostic groups (manic, acute schizophrenic, and nonpsychotic patients) did not differ in age, sex, education, or socioeconomic status, although nonpsychotic patients had a significantlyhigher proportion of whites to blacks than any of the other groups (×2 = 11.4, df = 2, P < .01). Nonpsychotic patients also had significantly fewer previous hospitalizations (F = 6.6, df= 3, P < .01). Chronically hospitalized schizophrenic patients were significantlyolder (F = 48.9, df = 3, P < .001) but did not differ from the other patients in sex, education, race, or number of previous hospitalizations (P > .20). Comparisons between patients from Michael Reese Hospital and the Illinois State Psychiatric Institute on interactive behavior and thought disorder showed that within each of the three acute diagnostic groups, there were no significant differences.
Assessment Technique Batteries of psychologicaltests and interviewswere administered to the nonchronic patients during the first 1.5 weeks of their hospitalization. Tests for thought disorder included
INTERACTIVE BEHAVIOR AND THOUGHT DISORDER
the Proverbs T e s t y the Object Sorting T e s t y and the comprehension subtest of the Wechsler Adult Intelligence Scale. 31 During testing, all speech was recorded verbatim by the examiner. Any departures from standard test procedures were also recorded, as were any unusual behaviors on the part of the patients. These tests provide a wide sample of patients' verbalizations and behavior, affording a variety of measures of disturbed thought, speech, and behavior. The test battery has been used successfully in previous research.6.9,25 The Gorham Proverbs Test 32 contains 12 proverbs that the patient is asked to interpret. The comprehension subtest 3~ consists of 14 questions, most of which assess patients' social comprehension, including items that tap patients' knowledge of social customs and safety conventions (e.g., what one should do if one smells smoke in a crowded theater). The Object Sorting Test 33 consists of a collection of objects that patients can recognize from everyday life (e.g., a fork, a cigar, and a bicycle bell). One object is designated to be the starting object, which the examiner places in a box. The patients are asked to select which items from the remaining collection belong with the starting item. When patients are finished sorting, they are asked to explain the reasons for their sorting. The process is repeated with six other starting objects. These three tests were scored on an index assessing the degree to which patients behave in an interactive manner with the examiner. For this measure, a rating scale has been devised that assesses the number of interactive comments or actions a patient generates during a single test. This scale assesses the number of times during each test that patients (1) make side comments, (2) ask questions about the test or examiner, (3) request feedback from the examiner, (4) make jokes, (5) play with the objects, or (6) otherwise behave in ways that attempt to elicit a response from the examiner. Individual scales were used to assess each of these behaviors separately. Raters were blind to patients' diagnoses. Data were analyzed separately on an index of the amount of interactive behavior produced during each test, and a composite index was also computed that rated patients according to their most interactive behavior on any of the three tests. For this composite index, criteria were constructed to classify patients' behavior according to one of five levels of interactive behavior as follows: none, mild, moderate, extreme, or excessive. Since this index encompasses patients' behavior from all three tests, it provides a measure of a broad sample of behavior and is more sensitive to interactive behavior and more comprehensive than indices that use only one test. To assess the psychometric properties of the indices of interacti'veness, we first assessed interrater reliability. The raters achieved satisfactory interrater reliability on this measure (r = .85 to .98). In addition, assessment of correlations among the three individual tests of interactiveness indicated a significant relationship among the tests (P < .05 for all r values among the three tests). Regarding construct validity, we found that the diagnostic group expected to show the most interactiveness (i.e., manic patients) did so, and that the group expected to show the least interactive-
247
ness (i.e., chronic patients) also did. This will be discussed later in greater detail. An example of interactiveness occurred in a patient who answered the question, "Why does a train have an engine?", with a series of questions and comments as follows: "That's an easy one. Do you know the answer to it? So it can run, of course. Do you like your job? I like the sweater you are wearing." Patients were also rated for bizarre-idiosyncratic thinking or positive thought disorder, a construct that refers to the expression of strange, incoherent, inadequate, or incomprehensible thoughts and encompasses most of the qualities usually included by diverse theorists as positive thought disorder (e.g., loose associations, incoherent speech, private language such as is found in neologisms, illogical thinking, and autistic thinking). This measure of positive thought disorder was assessed from a composite index based on the patients' most pathological score on any of the three tests. The principles for scoring the tests for bizarre-idiosyncratic thinking have been described previously in detailed manuals, which provide numerous examples. 2a Satisfactory interrater reliability has been obtained for these measures, and the measures have been used previously with success.4,6,9.10,24,25 For this composite index, each patient was assigned to one of five categories or levels of severity of thought disorder ranging from no thought disorder (level 1) to very severe thought disorder (level 5). This composite score of positive thought disorder, based on patients' responses on three tests, encompasses a broad sample of patients' behavior to enhance its sensitivity to pathological thinking. Following is an example of a severely thought-disordered response that was verbalized by a patient from the current sample. In interpreting the proverb, "One swallow does not make a summer," the patient said, "Now listen to me carefully. That means don't eat sand with your sandals on." To provide data on a medication-free sample, efforts were made at the Illinois State Psychiatric Institute to put patients through a washout period lasting the first few weeks of the study. Seventeen manic patients were medicationfree when assessed for interactive behavior, as were 19 acute schizophrenic patients and 16 nonpsychotic patients. Of 21 medicated manic patients, three were taking lithium alone, eight lithium plus neuroleptics, six neuroleptics without lithium, and four anxiolytics and/or sleep medication. Of 20 medicated acute schizophrenic patients, none was taking lithium and 18 were taking neuroleptics either alone or in combination with other medications. Of 15 medicated nonpsychotic patients, none was taking lithium, two were taking neuroleptics, and the others were taking antidepressants, minor tranquilizers, and/or sleep medications. The significant relationship (reported later) found between interactive behavior and thought disorder occurred for both medicated and unmedicated manic patients.
RESULTS
Interactive Behavior Table 1 presents data on the differences in patients' behavior associated with their style of interacting based on the composite measure of
248
GROSSMAN AND HARROW
Table 1. Interactive Behavior as a Function of Diagnosis Level of Interactiveness
None to Mild Moderate Excessive Diagnostic Group
No.
%
No.
%
No.
%
Bipolar manic(n = 37) Acute schizophrenic (n = 31) Nonpsychotic (n = 31)
3 18 15
8 51 48
13 14 14
35 40 45
21 3 2
57 9 6
Chronic schizophrenic (n = 34)
23
77
7
23
0
NOTE. The data are based on the composite index of interactive behavior.
interactive behavior. These results indicate that 57% of manic patients showed excessive interactive behavior on the composite index, and another 35% showed moderate interactive behavior. In contrast, 9% of acute schizophrenics and 6% of nonpsychotic patients showed excessive interactive behavior. None of the chronic schizophrenic patients showed excessive interactive behavior. To determine whether the diagnostic groups differed on level of interactive behavior, we conducted a series of one-way analyses of variance ANOVAs) on indices of interactive behavior. These analyses showed significant differences on the composite index of interactive behavior (F = 23.94, df = 3,136, P < .0001). Post hoc Neuman-Keuls tests indicated that on the composite index, manic patients displayed significantly more interactive behavior than acute or chronic schizophrenic patients or nonpsychotic patients (P < .05). On this index, acute schizophrenic patients and nonpsychotic patients displayed significantly more interactive behavior than chronic schizophrenic patients (P < .05). The same pattern emerged when only medication-free patients were included in the analysis. Medication-free manic patients showed significantly more interactive behavior on this index than medication-free acute schizophrenic or nonpsychotic patients (F = 7.75, dr= 2,46, P < .01). Significant results were also found on each of the individual indices from the Proverbs Test (F = 12.26, df = 3,132, P < .0001), the comprehension subtest (F = 14.11, df = 3,136, P < .0001), and the Object Sorting Test ( F = 9.95, d f = 3 , 1 3 1 , P < .0001). Post hoc Neuman-Keuls tests indicated that on these indices, manic patients displayed significantly more interactive behavior than acute or chronic schizophrenic patients
or nonpsychotic patients (P < .05). Acute and chronic schizophrenic patients and nonpsychotic patients did not differ significantly from each other on these individual indices. We also used analysis of covariance to analyze the interactiveness data separately while simultaneously controlling for number of previous hospitalizations. Results of the analysis of covariance indicated that manic patients showed more interactiveness, even after controlling for number of previous hospitalizations (F = 24.48, df = 3,134, P < .001). We also examined the manic patients' interactive behavior to determine whether it was a function of their awareness of others or was more focused on themselves. The qualities of behavior in manic patients that contributed to their response's being rated as interactive were the number of times they verbalized questions and comments about the tests, and comments about their own life experiences. This was contrasted quantitatively with interactive behavior that was non-self-focused. Patients' non-selffocused interactive behavior was based on the number of times they offered comments about the examiner, offered sexual or jocular comments, or demonstrated playful or provocative behavior. Both types of interactive behavior were found, but the manic patients' interaction was more typically directed toward bringing the attention of others to themselves rather than toward demonstrating an interest in others.
Positive ThoughtDisorder A large percentage of manic patients had severe positive thought disorder (Table 2), which is in accordance with results from previous research. 2-4,6,9,t°,34 On the comprehensive measure of positive thought disorder, 67% (25 of 37) of these hospitalized manic patients showed severe positive thought disorder, and an additional 24% (nine of 37) showed clear evidence of abnormal thinking. Thus, approximately 90% (34 of 37) of the manic patients showed moderate to severe bizarre-idiosyncratic or abnormal thinking. Eighty percent (24 of 30) of the chronically hospitalized schizophrenic patients showed these levels of thought disorder, as did 75% (26 of 35) of the acute schizophrenic patients. When the data from this measure were subjected to a
249
INTERACTIVE BEHAVIOR AND THOUGHT DISORDER
Table 2. Relationship Between Interactive Behavior and Thought Disorder Level of Thought Disorder Level of Interactive Behavior Bipolar manics None or mild Moderate Extreme to excessive Acute schizophrenics None or mild Moderate Extreme to excessive Nonpsychotics None or mild Moderate Extreme to excessive Chronic schizophrenics None or mild Moderate Extreme to excessive
None or Severe/ Mild Moderate Very Severe
No. of Patients No.
%
No.
%
No.
%
0
0
1
100
15
3 20
6 40
6
40
21
0
18
86
3 37.5 6 25
2 12
25 50
3
100
35 8 24
Pearson• .40"
37 1
3
14
.38* 3 37.5 6 25
3
0
0
31 9
7 78
2 22
20
11 55
2
1 50
3O 16 14
5 31 1 7
0
0
.36* 3
15
0 6
30
0
1
50
0
11 12
69 86
NS
1
7 0
O
*P < ,05,
one-way ANOVA, the overall diagnostic difference was significant (F = 16.59, dr= 3,129, P < .001). Post hoc Neuman-Keuls tests indicated that manic patients and chronically hospitalized schizophrenic patients did not differ in severity of thought disorder. However, both groups showed significantly (P < .05) more thought disorder than acute schizophrenic patients, who in turn showed significantly (P < .05) more thought disorder than nonpsychotic patients.
Relationship Between Interactive Behavior and
Bizarre-IdiosyncraticThinking Two methods were used to study the relationship between interactive behavior and thought disorder in manic patients. One of the methods involved (1) separate ratings for interactive behaviors that occurred in thought-disordered or bizarre responses, and then (2) an assessment of whether such interactive behavior contributed to the bizarreness of the response. The other method involved studying the correlations between interactive behavior and thought disorder for each of the diagnostic groups analyzed
separately. Significant results were found for each of these methods, as outlined below. Table 3 presents the mean _ SD and one-way ANOVAs on indices of interactive behavior that were rated as contributing to the bizarre quality of thought disorder in responses to each of the three tests. One-way ANOVAs with post hoc Neuman-Keuls tests indicated that on all three tests, significantly more interactive responses occurred that contributed to the bizarreness of responses for manic patients (P < .001) than for the other three groups, who did not differ significantly from each other. The bipolar manic patients' interactive behavior contributed to their thought-disordered responses, but these patients also showed other interactive behavior that did not contribute directly to their thought disorder. We also analyzed the correlations between interactive behavior and thought disorder to provide further information about the relationship between these two factors (Table 2). To study the relationships in detail, this analysis was conducted using the indices of interactive behavior from each test and the comprehensive measure of thought disorder. The results indicated that manic patients' scores for interactive behavior, which previously differentiated manic patients from others, were significantly positively correlated with their scores on the composite index of thought disorder on the comprehension subtest (r = .32, dr= 37, P < .03), the Proverbs Test (r = .38, df = 35, P < .02), and the Object Sorting Test (r = .52, df= 35, P < .001). For the other two acute patient groups, this relationship was positive but not as uniform. Table 3, Mean Values (-+SD) and One-Way A N O V A on the Indices of Interactive Behavior That Contributed to Bizarre Thinking Interactive Behavior
Index
Manic
Acute Schizophrenic
Chronic SchizoNonphrenic psychotic
F
Comprehension Subtest 3.4 -+ 3.9 .90 -+ 1.5 .23 -+ .94 1,0 _+ 1.6 12.8" Proverbs Test 4.5 -+ 5.8 .54 -+ 1.2 .73 -+ 1.7 .81 - 1.6 12.4" Object Sorting 3.3-+4.2 1.0-+3.1 .07-+.25 .20-+.61 8.4* Test NOTE. Does not include interactive behavior that did not contribute to bizarre thinking. *P < .0001.
250
GROSSMAN AND HARROW
Acute schizophrenic patients' scores for interactive behavior were significantly positively correlated with their scores for thought disorder on the Proverbs Test (r = .35, df = 35, P < .02) and the Object Sorting Test (r = .34, df = 34, P < .03), but not on the comprehension subtest (r = .16, df= 35, P > .15). Nonpsychotic patients had significant positive correlations on the comprehension subtest (r = .39, df= 31, P < .02) and Proverbs Test (r = .33, df = 31, P < .05), but not on the Object Sorting Test (r = .15, df= 29, P > .20). In contrast, for chronic schizophrenic patients, interactive behavior and thought disorder scores did not show a significant relationship on any of these tests. DISCUSSION
Interactive Behavior in Manic Patients These data support views about the importance of interactive behavior in bipolar manic patients, indicating that many manic patients show excessive interactive behavior and that this behavior is related to other types of psychopathology they manifest. The research provided evidence that bipolar manics and schizophrenics differed in the extent of interactive behavior, but did not differ significantly in the extent of positive thought disorder. Some degree of interactive behavior can also be found in a number of other disturbed patients, although not to the same extent as in manic disorders. A moderate to high level of interactive behavior was found in a number of acute schizophrenics, as well as in some disturbed nonpsychotic patients, and in a small number of chronic schizophrenics. However, excessive interactive behavior was more frequent in manic patients than in patients with other types of disorders (P < .05).
Quality of Interactive Behavior in Manic Patients We studied the question of whether manic patients' responses during assessment are often directed socially toward the examiner or toward the patients' audience. The behavior of the current sample of manic patients seemed directed toward trying to elicit responses from the examiner or to engage the examiner in shared activity. In contrast, both acute and chronic schizophrenic patients differed from the manic patients, in that schizophrenic patients did not
show as extreme an interest in interacting with the external environment. The question arises as to whether the manic patients' interactive behavior was a function of their concern for other people and how other people feel or behave. Analysis of the type of interactive behavior shown by the manic patients suggests that the qualities of these acutely disordered manic patients' interactive behavior that caused their responses to be rated as interactive often were not comments about the examiner, but rather were questions and comments about their own test responses and life experiences. We have suggested elsewhere 16 that the interactive behavior of manic patients involves an intermingling into their behavior their own personal wishes and themes, and a lack of restraint in expressing them. The current data may also indicate that manic patients, even as they interact with others, focus more of their attention on themselves than on other people. Another aspect of the quality of manic interactive behavior that appeared should also be noted. Descriptions of some manic behavior emphasizing sexual and jocular behavior have often been cited in the literature. 9,16-19 The present sample consisted of patients, many of whom were psychotic, hospitalized for acute manic disorders. At this severe level of disorder, their predominant behavior was strange and bizarre. This type of disturbed and at times psychotic behavior with gross disturbances of reality testing, as opposed to playful behavior, has been observed by others at the acute phase in manic patients whose disorder is severe enough to result in hospitalization. 6,15-19
Factors Involved in Manic Interactive Behavior and Thought Disorder For manic patients, the data showing a significant relationship between interactive behavior and thought disorder could suggest that the underlying factors leading to excessive interactive behavior also influence or lead to thought disorder. In previous research, we have noted that thought disorder is usually not a product of one factor, but of several converging factors. 26'27 Some of these factors are evident in the thought disorder found in a variety of diagnostic groups, and include reality distortions involving impaired monitoring, poor perspective, and pos-
INTERACTIVE BEHAVIOR AND THOUGHT DISORDER
251
sible loss of inhibition. 26 In general, variables such as attention and distractibility may play a role in interactiveness. Other factors may be more specific to particular diagnostic groups. The present data, indicating a close link in manic patients between the tendency toward interactive behavior and thought disorder, suggest the importance of excessive interactive behavior in manic psychopathology. The less significant (but still positive) relationship between these two factors in acute schizophrenics suggests that there may also be a link between interactive behavior and thought disorder in these acute patients, but it is not as strong a relationship. Since almost all manics and acute schizophrenics with severe thought disorder (39 of 42, or >90%) showed at least a moderate level of interactive behavior, the association could fit in with our previous views on the importance of a high level of cognitive arousal and difficulty in monitoring one's own behavior as often involved in positive thought disorder and in the psychopathology found in acute patients. 26 However, other factors are also important in thought disorder in acute manic and schizophrenic patients. 16,24 Thus, severe excessive interactive behavior was more frequent in manic patients than in acute schizophrenics, and was more closely linked to factors associated with thought disorder in manic patients than in schizophrenic
patients. However, a number of acute schizophrenics showed interactive behavior that was also linked to thought disorder.
Chronic Schizophrenic Patients Recent theory about chronic schizophrenic patients has tended to emphasize their negative symptoms, and some believe that these patients do not experience many positive symptoms or severe levels of positive symptoms. The current data indicate a relatively low level of interactive behavior for these chronic schizophrenics, which is in accordance with formulations about the importance of negative symptoms and tendencies to withdraw by this type of chronic patient. However, the current data also indicate that schizophrenic patients hospitalized chronically and continuously (i.e., >5 years) show very severe positive thought disorder and reality distortions. For these chronic schizophrenics, severe positive thought disorder is prominent, contrary to some theoretical formulations. This thought disorder can become manifest on presentation of any of a variety of tests or stimuli, and it can occur with or without interactive behavior. Unlike the more acute manic and acute schizophrenic patients, correlations for the chronic schizophrenic patients between interactive behavior and positive thought disorder were not significant.
REFERENCES 1. Johnston MH, Holzman PS. Assessing Schizophrenic Thinking. San Francisco, CA: Jossey-Bass, 1979. 2. Bleuler E. Dementia praecox, or the group of schizophrenias (J. Zinkin, translator). New York, NY: International Press, 1950 (originally published 1911). 3. Andreasen NC. Thought, language and communication disorders. Diagnostic significance. Arch Gen Psychiatry 1979;36:1325-1330. 4. Andreasen NC, Powers P. Overinclusive thinking in mania and schizophrenia. Br J Psychiatry 1974;125:452-456. 5. Andreasen NC, Powers P. Creativity and psychosis: an examination of conceptual style. Arch Gen Psychiatry 1975; 342:70-73. 6. Harrow M, Grossman LS, Silverstein ML, Meltzer HY. Thought pathology in manic and schizophrenic patients: its occurrence at hospital admission and seven weeks later. Arch Gen Psychiatry 1982;39:665-671. 7. Harvey PD, Earle-Boyer EA, Levinson JC. Distractibility and discourse failure: their association in mania and schizophrenia. J Nerv Ment Dis 1986;174:274-279. 8. Solovay MR, Shenton ME, Holzman PS. Comparative
studies of thought disorders. I. Mania and schizophrenia. Arch Gen Psychiatry 1987;44:13-20. 9. Grossman LS, Harrow M, Sands JR. Features associated with thought disorder in manic patients at 2-4 year follow-up. Am J Psychiatry 1986;143:306-311. 10. Harrow M, Grossman LS, Silverstein S, Meltzer HY, Kettering RA. Longitudinal study of thought disorder in manic patients. Arch Gen Psychiatry 1986;43:781-785. 11. Andreasen NC, Grove WM. Thought, language and communication in schizophrenia: diagnosis and prognosis. Schizophr Bull 1986;12:348-359. 12. Hoffman RE, Stopek S, Andreasen NC. A comparative study of manic vs. schizophrenic speech disorganization. Arch Gen Psychiatry 1986;43:831-838. 13. Holzman PS, Shenton ME, Solovay MR. Quality of thought disorder in differential diagnosis. Schizophr Bull 1986;12:360-371. 14. Holzman PS, Solovay MR, Shenton ME. Thought disorder specificity in functional psychoses. In: Alpert M (ed): Controversies in Schizophrenia: Changes and Constancies. New York, NY: Guilford, 1984. 15. Carlson GA, Goodwin FK. The stages of mania: A
252
longitudinal analysis of the manic episode. Arch Gen Psychiatry 1973;28:221-228. 16. Grossman LS, Harrow M. Thought disorder and cognitive processes in mania. Annu Rev Psychopathol 1991;1:134-186. 17. Goodwin FK, Jamison KR. Manic-Depressive Illness. New York, NY: Oxford University Press, 1990. 18. Janowsky DS, Left M, Epstein RS. Playing the manic game. Arch Gen Psychiatry 1970;22:252-261. 19. Bunney WE, Murphy DL, Goodwin FK, Borge GF. The "switch process" in manic-depressive illness. I. A systematic study of sequential behavior changes. Arch Gen Psychiatry 1972;27:295-302. 20. Spitzer RL, Endicott J, Robins E. Research Diagnostic Criteria. Arch Gen Psychiatry 1978;35:773-782. 21. Endicott J, Spitzer R E A diagnostic interview: the Schedule for Affective Disorders and Schizophrenia. Arch Gen Psychiatry 1978;35:837-844. 22. Wing JK, Cooper JE, Sartorius N. The measurement and classification of psychiatric symptoms. London, UK: Cambridge University Press, 1974. 23. Schwartz D, Grinker RR, Harrow M, Holzman P. Six clinical features of schizophrenia. J Nerv Ment Dis 1978;166: 831-838. 24. Harrow M, Quinlan D. Disordered Thinking and Schizophrenic Psychopathology. New York, NY: Gardner, 1985. 25. Goldberg JF, Harrow M, Grossman LS. Course and outcome in bipolar and unipolar affective disorders: a
GROSSMAN AND HARROW
longitudinal follow-up study. Am J Psychiatry 1995;152:379384. 26. Harrow M, Lanin-Kettering I, Miller JF. Impaired perspective and thought pathology in schizophrenic and psychotic disorders. Schizophr Bull 1989;15:605-623. 27. Grinker RR, Harrow M (eds). A Multidimensional Approach to Clinical Research in Schizophrenia. Springfield, IL: Thomas, 1987. 28. Grossman LS, Harrow M, Fudala J, Meltzer HY. The longitudinal course of schizoaffective disorders: a prospective follow-up study. J Nerv Merit Dis 1984;172:140-149. 29. Harrow M, Goldberg JF, Grossman LS, Meltzer HY. Outcome in manic disorders: a naturalistic follow-up study. Arch Gen Psychiatry 1990;47:665-671. 30. Grossman LS, Harrow M, Goldberg JF, Fichtner CG. Follow-up of schizoaffective outcome using three successive assessments: comparisons with schizophrenia and affective disorders. Am J Psychiatry 1991;148:1359-1365. 31. Wechsler D. Wechsler Adult Intelligence Scale Manual. New York, NY: Psychological, 1955. 32. Gorham DR. A proverbs test for clinical and experimental use. Psychol Rep 1956;2:1-12. 33. Goldstein K, Sheerer M. Abstract and concrete behavior: an experimental study with special tests. Psychol Monogr 1941;52:1-239. 34. Earle-Boyer EA, Levinson JC, Grant R, Harvey PD. The consistency of thought disorder in mania and schizophrenia. II. An assessment of consecutive admissions. J Nerv Ment Dis 1986;174:443-447.
The research was designed to assess whether bipolar manic patients are unique in showing excessive interactive behavior and whether interactive behavior is linked to positive thought disorder in mania and other psychotic disorders. We compared the interactive behavior of manic patients versus schizophrenics and nonpsychotic patients. Degree of interactive behavior and severity of thought disorder were assessed in an inpatient sample of 140 patients, including 38 bipolar manic patients, and control samples of 40 acute schizophrenic patients, 32 nonpsychotic patients, and 30 chronically hospitalized schizophrenic patients. The assessment battery used standardized tests to assess interactive behavior and thought disorder. Manic patients displayed significantly more interactive behav-
ior on all three tests than the other three patient groups. Interactive behavior made a greater contribution to the thought disorder of manic patients than to the thought disorder of the other three diagnostic groups. The other two acute patient groups also showed relationships between interactive behavior and thought disorder. The chronic schizophrenics did not show this pattern. The data indicated that many types of acute patients show some interactive behavior, but bipolar manic patients demonstrate more excessive interactive behavior than do other patients. There is a closer link between excessive interactive behavior and thought disorder in manic patients.
HE C U R R E N T R E S E A R C H was designed to study the frequency of interactive behavior in patients with bipolar affective disorders and to provide information about the relationship between interactive behavior and thought disorder. A high frequency of intrusive or interactive behaviors is one of the more common clinical observations in mania and is not observed as often in other psychotic disorders. An important question that emerges is, if manic patients show excessively interactive behavior, is the thought disorder frequently reported in these patients linked to their interactive behavior and, if so, is this linkage specific to manic patients or is it found in other psychotic patients? Since interactiveness and thought disorder are frequently observed by psychiatrists treating manic patients, the occurrence of both and their potential linkage are important for clinical practice and for theory about bipolar disorders. Thought disorder and cognitive abnormality have historically been viewed as important symptoms of schizophrenia. 1 At one time, thought disorder was seen both as the central pathognomonic symptom of schizophrenia, and as rare or nonexistent in other psychiatric disorders. 2However, the past 20 years have produced a wealth of evidence indicating that many manic patients have thought disorder as severe as that found in schizophrenic patients during the period of acute ~lness. 3-8 Additional research has suggested that some manic patients show severe
thought disorder when evaluated years after the acute hospitalization phase. T M Research seeking to differentiate thought disorder in manic patients from thought disorder in schizophrenic patients has produced mixed results. Several recent studies have found positive results. 3-8A1-14 Various qualities of mania have been cited as being involved in manic thought disorder, such as racing thoughts, accelerated speech, overproductive ideation, excessive jocularity and wordplay, and overinclusive behavior. 4,8,11,15-19Many of these qualities have not been studied systematically, but rather are based on anecdotal evidence. A systematic study of some of these qualities that may be involved in manic thought disorder can provide clues as to the pathological factors involved in disordered thinking, and can also be helpful in furnishing important information about similarities and differences in the thought disorder found in manic and schizophrenic patients. This type of analysis can provide data concerning the nature of thought disorder in major psychiatric disorders and can
T
Copyright © 1996 by W.B. Saunders Company
From the Department of Psychiatry, University of Illinois College of Medicine. Supported in part by Grants No. MH-26341 and MH-30938 from the National Institute of Mental Health, and by research grants from the John D. and Catherine T. MacArthur Foundation and Michael Reese Hospital. Address reprint requests to Linda S. Grossman, Ph.D., Department of Psychiatry (M/C913), 912 S. Wood St, Chicago, IL 60612-7327. Copyright © 1996 by W.B. Saunders Company 0010-440X/96/3704-0003503. 00/0
ComprehensivePsychiatry, Vol. 37, No. 4 (July/August), 1996: pp 245-252
245
246
GROSSMAN AND HARROW
p r o d u c e useful distinctions b e t w e e n diagnostic categories. O n e quality of m a n i c p a t i e n t s that has b e e n r e p o r t e d f r e q u e n t l y is an increase in sociability a n d interactive behavior, t5-19 W h a t factors are involved in m a n i c p a t i e n t s ' excessive interactiveness? It is possible that features that may b e involved in m a n i c p a t i e n t s ' cognitive a b e r r a tions a n d t h o u g h t disorder, such as lack of inhibition, s e l f - m o n i t o r i n g problems, a n d o t h e r types of reality distortions, play a role in their p o t e n t i a l excessive interactive behavior. T h e q u e s t i o n also arises as to w h e t h e r the link b e t w e e n interactive behavior a n d t h o u g h t disorder is p a r t i c u l a r to m a n i c patients, as o p p o s e d to o t h e r types of acutely ill p a t i e n t s such as schizophrenics a n d p a t i e n t s with o t h e r forms of psychopathology. A n assessment of the relationship b e t w e e n t h o u g h t disorder a n d interactive b e h a v i o r in m a n i c p a t i e n t s a n d schizophrenic p a t i e n t s would be a first step toward investigating these issues. It is also possible that interactive behavior is n o t associated with t h o u g h t disorder, b u t r a t h e r is part of the overall m a n i c s y m p t o m p i c t u r e along with o t h e r n o t e d qualities of m a n i a such as jocularity or elation. T h e c u r r e n t report a t t e m p t s to provide inform a t i o n o n these issues by focusing o n the potential relationship b e t w e e n interactive behavior a n d t h o u g h t disorder in m a n i c p a t i e n t s as c o m p a r e d with schizophrenic a n d n o n p s y c h o t i c patients. Specifically, this r e p o r t addresses the following questions: (1) D o most m a n i c p a t i e n t s show interactive characteristics? (2) Is their interactive b e h a v i o r related to the severity of their t h o u g h t disorder? (3) Does the interactive quality occur in m a n i c p a t i e n t s who do n o t show t h o u g h t disorder? (4) Does the interactive behavior occur in o t h e r p a t i e n t s such as acutely ill schizophrenics, chronic schizophrenics, a n d n o n psychotic psychiatric p a t i e n t s and, if so, is it linked to t h o u g h t disorder in these o t h e r patients? METHOD
Patient Population The psychiatric population consisted of 140 currently hospitalized patients. All patients were diagnosed according to the Research Diagnostic Criteria,2° based on the Schedule for AffectiveDisorders and Schizophrenia,21the Present State Examination,22and the Schizophrenic State Inventory SSI. 23'24These standardized semistructured diagnostic inter-
views were conducted by trained raters who were blind to the results of tests used to assess interactiveness and thought disorder. Research diagnoses were assigned for each patient by senior clinicians who achieved an interrater reliability Kof .85.24Patients were screened to include only those with bipolar disorder, schizophrenia, and nonpsychotic psychiatric disorders, excluding patients with overt organic brain disorders. This sample included 38 manic patients, 40 acute schizophrenic patients, and 32 nonpsychotic patients admitted to the Illinois State Psychiatric Institute and Michael Reese Hospital. For comparison purposes, an additional 30 chronic schizophrenic patients who had been hospitalized for at least 8 of the past 10 years at a long-term state hospital were assessed. In the current report, we use the term "acute patients" to differentiate recently hospitalized manic, schizophrenic, and nonpsychotic patients from the sample of 30 chronically hospitalized patients. The mean age of the acute patients was 24.8 _+ 6.7 years (mean +__SD). This research was conducted as part of the Chicago Follow-up Study. 6,9,10,25-30 Forty-eight percent of the patients were women, 64% were white, and 36% were black. Patients had a mean educational level of 12.3 years. Patients in the three acute diagnostic groups achieved a mean age-corrected scale score of 10.5 on the information subtest of the Wechsler Adult Intelligence Scale,31 which provided an estimate of IQ scores. This mean score is within the average range of intellectual functioning in the general population. The chronically hospitalized schizophrenic patients had significantly lower scores on the information subtest than the other three groups, with a mean score of 5.9 (F = 22.5, df = 3,136, P < .001). To control for differences in IQ, we conducted a separate analysis of covariance with IQ as the covariate on each of the major variables studied. The results indicated that after IQ differences were controlled, there remained significantdiagnostic differences both for interactiveness (F = 22.18, df = 3,135, P < .001) and for thought disorder (F = 12.60,df = 3,128, P < .001). The mean number of previous hospitalizations for the acute samples was 1.5. The three nonchronic diagnostic groups (manic, acute schizophrenic, and nonpsychotic patients) did not differ in age, sex, education, or socioeconomic status, although nonpsychotic patients had a significantlyhigher proportion of whites to blacks than any of the other groups (×2 = 11.4, df = 2, P < .01). Nonpsychotic patients also had significantly fewer previous hospitalizations (F = 6.6, df= 3, P < .01). Chronically hospitalized schizophrenic patients were significantlyolder (F = 48.9, df = 3, P < .001) but did not differ from the other patients in sex, education, race, or number of previous hospitalizations (P > .20). Comparisons between patients from Michael Reese Hospital and the Illinois State Psychiatric Institute on interactive behavior and thought disorder showed that within each of the three acute diagnostic groups, there were no significant differences.
Assessment Technique Batteries of psychologicaltests and interviewswere administered to the nonchronic patients during the first 1.5 weeks of their hospitalization. Tests for thought disorder included
INTERACTIVE BEHAVIOR AND THOUGHT DISORDER
the Proverbs T e s t y the Object Sorting T e s t y and the comprehension subtest of the Wechsler Adult Intelligence Scale. 31 During testing, all speech was recorded verbatim by the examiner. Any departures from standard test procedures were also recorded, as were any unusual behaviors on the part of the patients. These tests provide a wide sample of patients' verbalizations and behavior, affording a variety of measures of disturbed thought, speech, and behavior. The test battery has been used successfully in previous research.6.9,25 The Gorham Proverbs Test 32 contains 12 proverbs that the patient is asked to interpret. The comprehension subtest 3~ consists of 14 questions, most of which assess patients' social comprehension, including items that tap patients' knowledge of social customs and safety conventions (e.g., what one should do if one smells smoke in a crowded theater). The Object Sorting Test 33 consists of a collection of objects that patients can recognize from everyday life (e.g., a fork, a cigar, and a bicycle bell). One object is designated to be the starting object, which the examiner places in a box. The patients are asked to select which items from the remaining collection belong with the starting item. When patients are finished sorting, they are asked to explain the reasons for their sorting. The process is repeated with six other starting objects. These three tests were scored on an index assessing the degree to which patients behave in an interactive manner with the examiner. For this measure, a rating scale has been devised that assesses the number of interactive comments or actions a patient generates during a single test. This scale assesses the number of times during each test that patients (1) make side comments, (2) ask questions about the test or examiner, (3) request feedback from the examiner, (4) make jokes, (5) play with the objects, or (6) otherwise behave in ways that attempt to elicit a response from the examiner. Individual scales were used to assess each of these behaviors separately. Raters were blind to patients' diagnoses. Data were analyzed separately on an index of the amount of interactive behavior produced during each test, and a composite index was also computed that rated patients according to their most interactive behavior on any of the three tests. For this composite index, criteria were constructed to classify patients' behavior according to one of five levels of interactive behavior as follows: none, mild, moderate, extreme, or excessive. Since this index encompasses patients' behavior from all three tests, it provides a measure of a broad sample of behavior and is more sensitive to interactive behavior and more comprehensive than indices that use only one test. To assess the psychometric properties of the indices of interacti'veness, we first assessed interrater reliability. The raters achieved satisfactory interrater reliability on this measure (r = .85 to .98). In addition, assessment of correlations among the three individual tests of interactiveness indicated a significant relationship among the tests (P < .05 for all r values among the three tests). Regarding construct validity, we found that the diagnostic group expected to show the most interactiveness (i.e., manic patients) did so, and that the group expected to show the least interactive-
247
ness (i.e., chronic patients) also did. This will be discussed later in greater detail. An example of interactiveness occurred in a patient who answered the question, "Why does a train have an engine?", with a series of questions and comments as follows: "That's an easy one. Do you know the answer to it? So it can run, of course. Do you like your job? I like the sweater you are wearing." Patients were also rated for bizarre-idiosyncratic thinking or positive thought disorder, a construct that refers to the expression of strange, incoherent, inadequate, or incomprehensible thoughts and encompasses most of the qualities usually included by diverse theorists as positive thought disorder (e.g., loose associations, incoherent speech, private language such as is found in neologisms, illogical thinking, and autistic thinking). This measure of positive thought disorder was assessed from a composite index based on the patients' most pathological score on any of the three tests. The principles for scoring the tests for bizarre-idiosyncratic thinking have been described previously in detailed manuals, which provide numerous examples. 2a Satisfactory interrater reliability has been obtained for these measures, and the measures have been used previously with success.4,6,9.10,24,25 For this composite index, each patient was assigned to one of five categories or levels of severity of thought disorder ranging from no thought disorder (level 1) to very severe thought disorder (level 5). This composite score of positive thought disorder, based on patients' responses on three tests, encompasses a broad sample of patients' behavior to enhance its sensitivity to pathological thinking. Following is an example of a severely thought-disordered response that was verbalized by a patient from the current sample. In interpreting the proverb, "One swallow does not make a summer," the patient said, "Now listen to me carefully. That means don't eat sand with your sandals on." To provide data on a medication-free sample, efforts were made at the Illinois State Psychiatric Institute to put patients through a washout period lasting the first few weeks of the study. Seventeen manic patients were medicationfree when assessed for interactive behavior, as were 19 acute schizophrenic patients and 16 nonpsychotic patients. Of 21 medicated manic patients, three were taking lithium alone, eight lithium plus neuroleptics, six neuroleptics without lithium, and four anxiolytics and/or sleep medication. Of 20 medicated acute schizophrenic patients, none was taking lithium and 18 were taking neuroleptics either alone or in combination with other medications. Of 15 medicated nonpsychotic patients, none was taking lithium, two were taking neuroleptics, and the others were taking antidepressants, minor tranquilizers, and/or sleep medications. The significant relationship (reported later) found between interactive behavior and thought disorder occurred for both medicated and unmedicated manic patients.
RESULTS
Interactive Behavior Table 1 presents data on the differences in patients' behavior associated with their style of interacting based on the composite measure of
248
GROSSMAN AND HARROW
Table 1. Interactive Behavior as a Function of Diagnosis Level of Interactiveness
None to Mild Moderate Excessive Diagnostic Group
No.
%
No.
%
No.
%
Bipolar manic(n = 37) Acute schizophrenic (n = 31) Nonpsychotic (n = 31)
3 18 15
8 51 48
13 14 14
35 40 45
21 3 2
57 9 6
Chronic schizophrenic (n = 34)
23
77
7
23
0
NOTE. The data are based on the composite index of interactive behavior.
interactive behavior. These results indicate that 57% of manic patients showed excessive interactive behavior on the composite index, and another 35% showed moderate interactive behavior. In contrast, 9% of acute schizophrenics and 6% of nonpsychotic patients showed excessive interactive behavior. None of the chronic schizophrenic patients showed excessive interactive behavior. To determine whether the diagnostic groups differed on level of interactive behavior, we conducted a series of one-way analyses of variance ANOVAs) on indices of interactive behavior. These analyses showed significant differences on the composite index of interactive behavior (F = 23.94, df = 3,136, P < .0001). Post hoc Neuman-Keuls tests indicated that on the composite index, manic patients displayed significantly more interactive behavior than acute or chronic schizophrenic patients or nonpsychotic patients (P < .05). On this index, acute schizophrenic patients and nonpsychotic patients displayed significantly more interactive behavior than chronic schizophrenic patients (P < .05). The same pattern emerged when only medication-free patients were included in the analysis. Medication-free manic patients showed significantly more interactive behavior on this index than medication-free acute schizophrenic or nonpsychotic patients (F = 7.75, dr= 2,46, P < .01). Significant results were also found on each of the individual indices from the Proverbs Test (F = 12.26, df = 3,132, P < .0001), the comprehension subtest (F = 14.11, df = 3,136, P < .0001), and the Object Sorting Test ( F = 9.95, d f = 3 , 1 3 1 , P < .0001). Post hoc Neuman-Keuls tests indicated that on these indices, manic patients displayed significantly more interactive behavior than acute or chronic schizophrenic patients
or nonpsychotic patients (P < .05). Acute and chronic schizophrenic patients and nonpsychotic patients did not differ significantly from each other on these individual indices. We also used analysis of covariance to analyze the interactiveness data separately while simultaneously controlling for number of previous hospitalizations. Results of the analysis of covariance indicated that manic patients showed more interactiveness, even after controlling for number of previous hospitalizations (F = 24.48, df = 3,134, P < .001). We also examined the manic patients' interactive behavior to determine whether it was a function of their awareness of others or was more focused on themselves. The qualities of behavior in manic patients that contributed to their response's being rated as interactive were the number of times they verbalized questions and comments about the tests, and comments about their own life experiences. This was contrasted quantitatively with interactive behavior that was non-self-focused. Patients' non-selffocused interactive behavior was based on the number of times they offered comments about the examiner, offered sexual or jocular comments, or demonstrated playful or provocative behavior. Both types of interactive behavior were found, but the manic patients' interaction was more typically directed toward bringing the attention of others to themselves rather than toward demonstrating an interest in others.
Positive ThoughtDisorder A large percentage of manic patients had severe positive thought disorder (Table 2), which is in accordance with results from previous research. 2-4,6,9,t°,34 On the comprehensive measure of positive thought disorder, 67% (25 of 37) of these hospitalized manic patients showed severe positive thought disorder, and an additional 24% (nine of 37) showed clear evidence of abnormal thinking. Thus, approximately 90% (34 of 37) of the manic patients showed moderate to severe bizarre-idiosyncratic or abnormal thinking. Eighty percent (24 of 30) of the chronically hospitalized schizophrenic patients showed these levels of thought disorder, as did 75% (26 of 35) of the acute schizophrenic patients. When the data from this measure were subjected to a
249
INTERACTIVE BEHAVIOR AND THOUGHT DISORDER
Table 2. Relationship Between Interactive Behavior and Thought Disorder Level of Thought Disorder Level of Interactive Behavior Bipolar manics None or mild Moderate Extreme to excessive Acute schizophrenics None or mild Moderate Extreme to excessive Nonpsychotics None or mild Moderate Extreme to excessive Chronic schizophrenics None or mild Moderate Extreme to excessive
None or Severe/ Mild Moderate Very Severe
No. of Patients No.
%
No.
%
No.
%
0
0
1
100
15
3 20
6 40
6
40
21
0
18
86
3 37.5 6 25
2 12
25 50
3
100
35 8 24
Pearson• .40"
37 1
3
14
.38* 3 37.5 6 25
3
0
0
31 9
7 78
2 22
20
11 55
2
1 50
3O 16 14
5 31 1 7
0
0
.36* 3
15
0 6
30
0
1
50
0
11 12
69 86
NS
1
7 0
O
*P < ,05,
one-way ANOVA, the overall diagnostic difference was significant (F = 16.59, dr= 3,129, P < .001). Post hoc Neuman-Keuls tests indicated that manic patients and chronically hospitalized schizophrenic patients did not differ in severity of thought disorder. However, both groups showed significantly (P < .05) more thought disorder than acute schizophrenic patients, who in turn showed significantly (P < .05) more thought disorder than nonpsychotic patients.
Relationship Between Interactive Behavior and
Bizarre-IdiosyncraticThinking Two methods were used to study the relationship between interactive behavior and thought disorder in manic patients. One of the methods involved (1) separate ratings for interactive behaviors that occurred in thought-disordered or bizarre responses, and then (2) an assessment of whether such interactive behavior contributed to the bizarreness of the response. The other method involved studying the correlations between interactive behavior and thought disorder for each of the diagnostic groups analyzed
separately. Significant results were found for each of these methods, as outlined below. Table 3 presents the mean _ SD and one-way ANOVAs on indices of interactive behavior that were rated as contributing to the bizarre quality of thought disorder in responses to each of the three tests. One-way ANOVAs with post hoc Neuman-Keuls tests indicated that on all three tests, significantly more interactive responses occurred that contributed to the bizarreness of responses for manic patients (P < .001) than for the other three groups, who did not differ significantly from each other. The bipolar manic patients' interactive behavior contributed to their thought-disordered responses, but these patients also showed other interactive behavior that did not contribute directly to their thought disorder. We also analyzed the correlations between interactive behavior and thought disorder to provide further information about the relationship between these two factors (Table 2). To study the relationships in detail, this analysis was conducted using the indices of interactive behavior from each test and the comprehensive measure of thought disorder. The results indicated that manic patients' scores for interactive behavior, which previously differentiated manic patients from others, were significantly positively correlated with their scores on the composite index of thought disorder on the comprehension subtest (r = .32, dr= 37, P < .03), the Proverbs Test (r = .38, df = 35, P < .02), and the Object Sorting Test (r = .52, df= 35, P < .001). For the other two acute patient groups, this relationship was positive but not as uniform. Table 3, Mean Values (-+SD) and One-Way A N O V A on the Indices of Interactive Behavior That Contributed to Bizarre Thinking Interactive Behavior
Index
Manic
Acute Schizophrenic
Chronic SchizoNonphrenic psychotic
F
Comprehension Subtest 3.4 -+ 3.9 .90 -+ 1.5 .23 -+ .94 1,0 _+ 1.6 12.8" Proverbs Test 4.5 -+ 5.8 .54 -+ 1.2 .73 -+ 1.7 .81 - 1.6 12.4" Object Sorting 3.3-+4.2 1.0-+3.1 .07-+.25 .20-+.61 8.4* Test NOTE. Does not include interactive behavior that did not contribute to bizarre thinking. *P < .0001.
250
GROSSMAN AND HARROW
Acute schizophrenic patients' scores for interactive behavior were significantly positively correlated with their scores for thought disorder on the Proverbs Test (r = .35, df = 35, P < .02) and the Object Sorting Test (r = .34, df = 34, P < .03), but not on the comprehension subtest (r = .16, df= 35, P > .15). Nonpsychotic patients had significant positive correlations on the comprehension subtest (r = .39, df= 31, P < .02) and Proverbs Test (r = .33, df = 31, P < .05), but not on the Object Sorting Test (r = .15, df= 29, P > .20). In contrast, for chronic schizophrenic patients, interactive behavior and thought disorder scores did not show a significant relationship on any of these tests. DISCUSSION
Interactive Behavior in Manic Patients These data support views about the importance of interactive behavior in bipolar manic patients, indicating that many manic patients show excessive interactive behavior and that this behavior is related to other types of psychopathology they manifest. The research provided evidence that bipolar manics and schizophrenics differed in the extent of interactive behavior, but did not differ significantly in the extent of positive thought disorder. Some degree of interactive behavior can also be found in a number of other disturbed patients, although not to the same extent as in manic disorders. A moderate to high level of interactive behavior was found in a number of acute schizophrenics, as well as in some disturbed nonpsychotic patients, and in a small number of chronic schizophrenics. However, excessive interactive behavior was more frequent in manic patients than in patients with other types of disorders (P < .05).
Quality of Interactive Behavior in Manic Patients We studied the question of whether manic patients' responses during assessment are often directed socially toward the examiner or toward the patients' audience. The behavior of the current sample of manic patients seemed directed toward trying to elicit responses from the examiner or to engage the examiner in shared activity. In contrast, both acute and chronic schizophrenic patients differed from the manic patients, in that schizophrenic patients did not
show as extreme an interest in interacting with the external environment. The question arises as to whether the manic patients' interactive behavior was a function of their concern for other people and how other people feel or behave. Analysis of the type of interactive behavior shown by the manic patients suggests that the qualities of these acutely disordered manic patients' interactive behavior that caused their responses to be rated as interactive often were not comments about the examiner, but rather were questions and comments about their own test responses and life experiences. We have suggested elsewhere 16 that the interactive behavior of manic patients involves an intermingling into their behavior their own personal wishes and themes, and a lack of restraint in expressing them. The current data may also indicate that manic patients, even as they interact with others, focus more of their attention on themselves than on other people. Another aspect of the quality of manic interactive behavior that appeared should also be noted. Descriptions of some manic behavior emphasizing sexual and jocular behavior have often been cited in the literature. 9,16-19 The present sample consisted of patients, many of whom were psychotic, hospitalized for acute manic disorders. At this severe level of disorder, their predominant behavior was strange and bizarre. This type of disturbed and at times psychotic behavior with gross disturbances of reality testing, as opposed to playful behavior, has been observed by others at the acute phase in manic patients whose disorder is severe enough to result in hospitalization. 6,15-19
Factors Involved in Manic Interactive Behavior and Thought Disorder For manic patients, the data showing a significant relationship between interactive behavior and thought disorder could suggest that the underlying factors leading to excessive interactive behavior also influence or lead to thought disorder. In previous research, we have noted that thought disorder is usually not a product of one factor, but of several converging factors. 26'27 Some of these factors are evident in the thought disorder found in a variety of diagnostic groups, and include reality distortions involving impaired monitoring, poor perspective, and pos-
INTERACTIVE BEHAVIOR AND THOUGHT DISORDER
251
sible loss of inhibition. 26 In general, variables such as attention and distractibility may play a role in interactiveness. Other factors may be more specific to particular diagnostic groups. The present data, indicating a close link in manic patients between the tendency toward interactive behavior and thought disorder, suggest the importance of excessive interactive behavior in manic psychopathology. The less significant (but still positive) relationship between these two factors in acute schizophrenics suggests that there may also be a link between interactive behavior and thought disorder in these acute patients, but it is not as strong a relationship. Since almost all manics and acute schizophrenics with severe thought disorder (39 of 42, or >90%) showed at least a moderate level of interactive behavior, the association could fit in with our previous views on the importance of a high level of cognitive arousal and difficulty in monitoring one's own behavior as often involved in positive thought disorder and in the psychopathology found in acute patients. 26 However, other factors are also important in thought disorder in acute manic and schizophrenic patients. 16,24 Thus, severe excessive interactive behavior was more frequent in manic patients than in acute schizophrenics, and was more closely linked to factors associated with thought disorder in manic patients than in schizophrenic
patients. However, a number of acute schizophrenics showed interactive behavior that was also linked to thought disorder.
Chronic Schizophrenic Patients Recent theory about chronic schizophrenic patients has tended to emphasize their negative symptoms, and some believe that these patients do not experience many positive symptoms or severe levels of positive symptoms. The current data indicate a relatively low level of interactive behavior for these chronic schizophrenics, which is in accordance with formulations about the importance of negative symptoms and tendencies to withdraw by this type of chronic patient. However, the current data also indicate that schizophrenic patients hospitalized chronically and continuously (i.e., >5 years) show very severe positive thought disorder and reality distortions. For these chronic schizophrenics, severe positive thought disorder is prominent, contrary to some theoretical formulations. This thought disorder can become manifest on presentation of any of a variety of tests or stimuli, and it can occur with or without interactive behavior. Unlike the more acute manic and acute schizophrenic patients, correlations for the chronic schizophrenic patients between interactive behavior and positive thought disorder were not significant.
REFERENCES 1. Johnston MH, Holzman PS. Assessing Schizophrenic Thinking. San Francisco, CA: Jossey-Bass, 1979. 2. Bleuler E. Dementia praecox, or the group of schizophrenias (J. Zinkin, translator). New York, NY: International Press, 1950 (originally published 1911). 3. Andreasen NC. Thought, language and communication disorders. Diagnostic significance. Arch Gen Psychiatry 1979;36:1325-1330. 4. Andreasen NC, Powers P. Overinclusive thinking in mania and schizophrenia. Br J Psychiatry 1974;125:452-456. 5. Andreasen NC, Powers P. Creativity and psychosis: an examination of conceptual style. Arch Gen Psychiatry 1975; 342:70-73. 6. Harrow M, Grossman LS, Silverstein ML, Meltzer HY. Thought pathology in manic and schizophrenic patients: its occurrence at hospital admission and seven weeks later. Arch Gen Psychiatry 1982;39:665-671. 7. Harvey PD, Earle-Boyer EA, Levinson JC. Distractibility and discourse failure: their association in mania and schizophrenia. J Nerv Ment Dis 1986;174:274-279. 8. Solovay MR, Shenton ME, Holzman PS. Comparative
studies of thought disorders. I. Mania and schizophrenia. Arch Gen Psychiatry 1987;44:13-20. 9. Grossman LS, Harrow M, Sands JR. Features associated with thought disorder in manic patients at 2-4 year follow-up. Am J Psychiatry 1986;143:306-311. 10. Harrow M, Grossman LS, Silverstein S, Meltzer HY, Kettering RA. Longitudinal study of thought disorder in manic patients. Arch Gen Psychiatry 1986;43:781-785. 11. Andreasen NC, Grove WM. Thought, language and communication in schizophrenia: diagnosis and prognosis. Schizophr Bull 1986;12:348-359. 12. Hoffman RE, Stopek S, Andreasen NC. A comparative study of manic vs. schizophrenic speech disorganization. Arch Gen Psychiatry 1986;43:831-838. 13. Holzman PS, Shenton ME, Solovay MR. Quality of thought disorder in differential diagnosis. Schizophr Bull 1986;12:360-371. 14. Holzman PS, Solovay MR, Shenton ME. Thought disorder specificity in functional psychoses. In: Alpert M (ed): Controversies in Schizophrenia: Changes and Constancies. New York, NY: Guilford, 1984. 15. Carlson GA, Goodwin FK. The stages of mania: A
252
longitudinal analysis of the manic episode. Arch Gen Psychiatry 1973;28:221-228. 16. Grossman LS, Harrow M. Thought disorder and cognitive processes in mania. Annu Rev Psychopathol 1991;1:134-186. 17. Goodwin FK, Jamison KR. Manic-Depressive Illness. New York, NY: Oxford University Press, 1990. 18. Janowsky DS, Left M, Epstein RS. Playing the manic game. Arch Gen Psychiatry 1970;22:252-261. 19. Bunney WE, Murphy DL, Goodwin FK, Borge GF. The "switch process" in manic-depressive illness. I. A systematic study of sequential behavior changes. Arch Gen Psychiatry 1972;27:295-302. 20. Spitzer RL, Endicott J, Robins E. Research Diagnostic Criteria. Arch Gen Psychiatry 1978;35:773-782. 21. Endicott J, Spitzer R E A diagnostic interview: the Schedule for Affective Disorders and Schizophrenia. Arch Gen Psychiatry 1978;35:837-844. 22. Wing JK, Cooper JE, Sartorius N. The measurement and classification of psychiatric symptoms. London, UK: Cambridge University Press, 1974. 23. Schwartz D, Grinker RR, Harrow M, Holzman P. Six clinical features of schizophrenia. J Nerv Ment Dis 1978;166: 831-838. 24. Harrow M, Quinlan D. Disordered Thinking and Schizophrenic Psychopathology. New York, NY: Gardner, 1985. 25. Goldberg JF, Harrow M, Grossman LS. Course and outcome in bipolar and unipolar affective disorders: a
GROSSMAN AND HARROW
longitudinal follow-up study. Am J Psychiatry 1995;152:379384. 26. Harrow M, Lanin-Kettering I, Miller JF. Impaired perspective and thought pathology in schizophrenic and psychotic disorders. Schizophr Bull 1989;15:605-623. 27. Grinker RR, Harrow M (eds). A Multidimensional Approach to Clinical Research in Schizophrenia. Springfield, IL: Thomas, 1987. 28. Grossman LS, Harrow M, Fudala J, Meltzer HY. The longitudinal course of schizoaffective disorders: a prospective follow-up study. J Nerv Merit Dis 1984;172:140-149. 29. Harrow M, Goldberg JF, Grossman LS, Meltzer HY. Outcome in manic disorders: a naturalistic follow-up study. Arch Gen Psychiatry 1990;47:665-671. 30. Grossman LS, Harrow M, Goldberg JF, Fichtner CG. Follow-up of schizoaffective outcome using three successive assessments: comparisons with schizophrenia and affective disorders. Am J Psychiatry 1991;148:1359-1365. 31. Wechsler D. Wechsler Adult Intelligence Scale Manual. New York, NY: Psychological, 1955. 32. Gorham DR. A proverbs test for clinical and experimental use. Psychol Rep 1956;2:1-12. 33. Goldstein K, Sheerer M. Abstract and concrete behavior: an experimental study with special tests. Psychol Monogr 1941;52:1-239. 34. Earle-Boyer EA, Levinson JC, Grant R, Harvey PD. The consistency of thought disorder in mania and schizophrenia. II. An assessment of consecutive admissions. J Nerv Ment Dis 1986;174:443-447.