Psychological Functioning of Bipolar Manic-Depressives in Joy
R. MacVane,
MA; James
D.
Lange, MA; Walter
Armin
Remission
Brown, MD; Mary Zayat
\s=b\ Several dimensions of psychological functioning of bipolar manic-depressives in remission were investigated. The following hypotheses were examined: (1) During remission, these individuals exhibit psychological maladjustment or a lack of positive mental health. (2) Bipolar manic-depressives are characterized by strong external orientations. The Personal Orientation Inventory, Marlowe-Crowne Social Desirability Scale, Levenson's
Internal and Powerful Others Locus of Control Scales, and the Embedded Figures Test were administered to matched groups of 35 bipolar manic-depressives and 35 normal comparison subjects. Discriminant analysis disclosed no significant differences between the two groups on measures of positive mental health and on measures of external orientation. The relationship between these results and the etiology and treatment of manicdepressive disorder is briefly discussed, as well as the contradiction these findings pose to current views of this disorder pervasive among professional and lay populations. (Arch Gen Psychiatry 35:1351-1354, 1978)
manic-depressive disorder characterized by dramatic mood shifts from severe depression to episodes of mania has long been of interest. Psychoanalytic theorists1"1 contended that during remission, an abnormal dependent character formation is present in the manicdepressive. Arieti4 described the manic-depressive as tend¬ ing "to be a conformist, willing to accept what he is given by his surroundings and to rely less than the average person on his own interpretations and evaluations of the external world."1"111 Winokur et als studied remitted manic-depressives and found that only 61% of their subjects reported periods between psychotic episodes that were symptom free. A
Bipolar
...
Accepted for publication May 15, 1978. From the Department of Psychology, University of Rhode Island (Ms MacVane), the Veterans Administration Hospital (Mr Lange, Dr Brown, and Ms Zayat), and the Department of Psychiatry, Brown University (Dr Brown), Providence. Reprint requests to Veterans Administration Hospital/151, Psychiatric Services, Davis Park, Providence, RI 02908 (Dr Brown).
series of studies comparing hospitalized bipolar (recurrent mania and depression) manic-depressive subjects with hospitalized unipolar (depression only) subjects disclosed significant differences between the two groups. Hospital¬ ized bipolar subjects appeared to be generally less neurotic and more healthy than unipolar subjects."" Subjects with bipolar disorder are also reported as more impulsive than unipolar individuals.710 With the current widespread use of lithium carbonate therapy, many manic-depressives are spending much of their time in remission. Consequently, information about their psychological functioning during remission is of interest both in understanding the psychology of this disorder and in understanding how to treat the manicdepressive individual psychotherapeutically. The study reported here has investigated several dimen¬ sions of psychological functioning of bipolar manic-depres¬ sives in remission. The following hypotheses were exam¬ ined: (1) During remission, bipolar manic-depressive indi¬ viduals exhibit psychological maladjustment or a lack of positive mental health. (2) Bipolar manic-depressives are characterized by strong external orientations. The first hypothesis is consistent with the psychoanalytic view that manic-depressives are victims of ever-present psychopath¬ ology that occasionally flares up into a depressed or manic episode. This hypothesis has also been an outgrowth of the fact that prospective subjects being interviewed for our study frequently described recurrent mood swings as intensely damaging to all aspects of their lives. Regardless of etiology, the actual occurrence of repeated manic and depressed episodes, disruptive both intrapsychically and interpersonally, could impair psychological health even when the individual is in remission. The second hypothesis refers to the psychoanalytic description of bipolar manicdepressives as being excessively conforming and depen¬ dent on others. Cohen et al' reported that individuals with this disorder try to reach success or satisfaction through the success of others rather than by their own efforts and
Downloaded From: http://archpsyc.jamanetwork.com/ by a McMaster University User on 10/11/2015
they tend to deny their own originality and, conse¬ quently, are often stereotyped in attitudes and opinions. that
Subjects 35 remitted bipolar manic-depressives (12 subjects women and 23 men) and 35 healthy comparison subjects (14 women and 21 men). The subjects in each group were matched with regard to age, sex, and social position as estimated by Hollingshead's Two-Factor Index of Social Position." The mean age of the manic-depressive group was 49.9 years and that of the comparison group 45.5 years, with no significant difference between them (f 1.40; > .05). For social position, the mean of the manicdepressive group was 3.11 and that of the comparison group 3.03, with no significant difference between the two groups (r 0.35; were
=
=
>
.05).
Outpatients with diagnoses of bipolar manic-depressive disor¬ der currently taking lithium carbonate were referred to the study from a Veterans Administration Hospital, a psychiatrist's private practice, a community mental health center, and a medical center's lithium carbonate clinic. Selection of bipolar manic-depressive subjects was made on the basis of information obtained from hospital and case records and an unstructured one-hour interview with each prospective subject focusing on past psychiatric prob¬ lems and treatment and details of their manic and depressive episodes. Decisions to include individuals as subjects were made independently by two researchers in accordance with the Feighner et al12 research criteria for bipolar manic-depressive disorder. The occurrence of at least one manic and one depressive episode was required, with each episode having to fulfill specific criteria. The
criteria excluded anyone who could also be classified into any other
diagnostic category (eg, alcoholism, schizophrenia, organic brain syndrome). Of the 53 individuals referred to the study and interviewed, 17 failed to meet the research criteria and were excluded. In selecting subjects,
no
attempt
was
made to
assess
how well
potential subjects were functioning while between affective episodes. The only requirements for manic-depressive subjects were that they (1) fulfilled the research criteria mentioned, (2) were currently taking lithium carbonate, and (3) were clearly between episodes. The presence of remission was assessed at the time of psychological testing. Brief interviews were conducted with each subject by two researchers followed by independent ratings on a modified version of the Brief Psychiatric Rating Scale (BPRS). Of the 36 manic-depressive subjects meeting the first two criteria, only one was later excluded from data analysis on the basis of the third criterion. This subject's BPRS ratings indicated
hypomanic symptomatology suggesting the possible onset of a manic episode. Subjects in the comparison group were individuals reporting no history of psychiatric disorder nor any treatment or medication for psychological problems or alcoholism in the past. These subjects were also independently rated on the BPRS at the time of testing by two researchers and the same exclusion criteria applied. None were excluded due to the presence of significant psychiatric symptoms. Comparison subjects received $5 for their participa¬ tion. All subjects gave informed consent after the been fully explained.
tendency to describe oneself favor¬
Internal and Powerful Others Locus of Control Scales.—These two are from a set of three locus of control orientation scales developed by Levenson and others." '" The scales represent two
scales
SUBJECTS AND METHODS The
instrument does measure this ably to win social approval.
procedures
had
Instruments Social Desirability Scale.-This scale attempts to measure the extent of an individual's tendency to describe himself in socially desirable terms to receive approval from others. Through a number of validity studies looking at the scale's behavioral correlates, Crowne and Marlowe1 ' have demonstrated that this
factors in that only items from the appropriate scale load on that one factor. A number of studies by Levenson and others" demonstrate acceptable validity of the scales. Personal Orientation Inventory.-The Personal Orientation Inven¬ tory (POI) attempts to measure self-actualization as defined by Maslow and also aims at measuring a person's mental health. Because this study focused on manic-depressives in terms of their psychological health, the POI was considered more appropriate than the Minnesota Multiphasic Personality Inventory (MMPI) and related tests that provide estimates of psychopathology rather than measures of positive mental health. Reviews by Shostrum™ and Knapp-'1 include a large number of studies demonstrating acceptable discriminant and predictive validity of this inventory as well as its correlation with other scales. For this study two scoring systems were used: the original scoring system of the 12 scales, and a modified scale form of the POI developed by Starrett.-'2 Embedded Figures Test.-The Embedded Figures Test (EFT) is a perceptual test that quantifies the subject's perceptual style on the dimension of field articulation. Individuals scoring in the fieldindependent range are believed to have an awareness of their needs, feelings, and personal attributes as being distinct from those of others. Field-dependent individuals, however, appear to have a less developed sense of separate identity and tend to rely more on external sources for definition of their attitudes and views of themselves.21 Brief Psychiatric Rating Scale.—The BPRS is a set of construct scales widely used as a rapid assessment technique for psychiatric patients." It consists of 18 symptom constructs and uses a seven-point scale of severity. A modified form of the BPRS was employed that was recently developed by a research group studying manic-depressive disorders at the National Institutes of Health (L.B. Bigelow, MD, D.L. Murphy, MD, oral and written communication, 1976). It consists of the original items plus four additional items, three of which are related to mania and one related to depression.
conceptually pure '"
Procedure
Subjects meeting the research criteria
were
tested within the
Immediately prior to subject for about 15 minutes and then rated him or her independently on the modified BPRS. Subjects took the tests in a varied order of presentation to offset any ordering effect. Using the SPSS-Subprogram discriminant, the direct method, three separate discriminant analyses were performed on the ten months following evaluation interviews. testing, two researchers interviewed each
data.
RESULTS
The BPRS scores of the manic-depressive group were clearly indicative of remission. Their mean scores on the 22 items ranged from 1.0 to 1.3 (1, symptom not present; 2, very mild) with the exception of items "somatic concern" and "tension," with means of 1.5 and 1.8, respectively. Comparison group mean scores ranged from 1.0 to 1.2, excluding "tension" with a mean of 1.5. Commonly used statistical procedures for computing interrater reliability were inappropriate for use with our BPRS ratings due to the limited range of BPRS scores obtained by the subjects. Most of the ratings subjects
Downloaded From: http://archpsyc.jamanetwork.com/ by a McMaster University User on 10/11/2015
Table 1.—Means and Standard Deviations for the Manie-Depressive and Comparison Groups
Comparison (N 35) =
Scale
Mean Personal Orientation Inventory Modified scale form 10.14 1 5.09
Original
SD
10
Mean
12 Embedded Figures Test Social Desirability Scale Internal Locus of Control Scale Powerful Others Locus of Control Scale
=
35)
Discriminant
SD
Function
2
Weights
df
Assorted scales 3.25
9.89 5.51 14.63 7.17 15.37
2.78 1.50 3.29 1.42
13.37 7.80 16.57
4.33 1.39 3.23
17.66 83.66 20.57
3.20
16.86
3.15
11.1£ 2.54
84.00
10.64 2.66
19.91
4.59
15.37 12.29 12.91 15.97 11.94 7.03
3.24 2.53
15.69 17.97
=
=
4.06
scales
1
Table 2.—Discriminant Analysis Results for Three Separate Analyses (n, 35; n,,
Manie-Depressive 35) (
1.84 3.83 1.47
3.20
20.40 20.49 15.66 12.40 12.11 16.17 11.83 7.06 15.77
Embedded Figures Test Social Desirability Scale Internal Locus of Control Scale
5
>.05
12
>.05
1_-0.15916
2
3. Personal Orientation
0.48458 0.62450 0.8797 8.396 -0.59118 -0.27558
Inventory
1__0.61626 2_1.05040 3
-0.19540
0.76246 0.15070 0.8531 9.851 0.77985 -0.77985 0.24821
52.00
5.78
34.71
4.97
34.03
5.35
10
0.01575 0.12194
23.14
6.17
24.43
7.18
12
-1.43104
42.35
>05
-0.21840
17.60 97.90 16.00
90.50 17.23
4
Powerful Others Locus of Control Scale 0.48574 2. Personal Orientation Inventory
3.29
2.91 2.71 2.67 3.58 1.84 1.16
0.63351 -0.86483 0.9587 2.784
3.38 5.87
-0.73975
received fell between 0 and 3. To compute interrater reliability, we first looked at each of the 22 BPRS items separately, and totaled the scores given by each rater across all 70 subjects on that particular item. Next, these two sums from raters 1 and 2 were compared to determine how large a discrepancy existed between their ratings. To do this, a "difference score" for each BPRS item was computed by subtracting the total from rater 1 on that particular item from the total from rater 2. Therefore, a low difference score for an item would indicate a high degree of similarity (high interrater reliability) between its two sets of ratings across the 70 subjects. The mean of the 22 difference scores was 2.27, indicating a very high degree of similarity between the two sets of ratings across all the BPRS items. Means and standard deviations for the two groups on all measures are presented in Table 1. As an additional check, t tests were performed on the EFT, the Social Desirability Scale, and the two Levenson scales. No significant differ¬ ences between groups were found. The t test was not performed on POI scales because this instrument was designed to yield a profile of scores, making it statistically inappropriate to interpret the scales individually. Three separate discriminant analyses were performed on the following measures on the two groups: (1) the 12 original POI scales (hypothesis 1), (2) Starrett's modified scale form of the POI (hypothesis 1), and (3) the Embedded Figures Test, the Marlowe-Crowne Social Desirability Scale, and the Levenson Internal and Powerful Others scales (hypothesis 2). Results of the three analyses are
presented in Table 2. These analyses showed no significant differences between the manic-depressive and comparison
on any of the measures. In addition, none of these three combinations of measures successfully discriminated between the two groups to the extent of being able to then classify a significant number of the 70 subjects into their correct groups. None of the three separate discriminant functions reached the significance level.
groups
COMMENT
The results of this study indicate that remitted manicdepressives do not differ from individuals with no history of psychiatric problems in terms of positive mental health as measured by the POI and also on a number of dimen¬ sions regarding external vs internal orientation. In this last respect, the two groups did not differ significantly on the field dependence/independence dimension, the desire to win social approval, or the extent to which they see events in their lives as being under their control. While not nullifying these findings, the higher male-female ratio in the sample does not reflect the male-female ratio in the manic-depressive population.
The data do not offer any evidence to contradict the null that no difference exists between the two groups on the measures. This is further supported by the fact that strict research criteria for manic-depressive groups were strictly adhered to; also, both the manicdepressive and comparison groups were well defined. Furthermore, the means of the two groups on the EFT, the Marlowe-Crowne Social Desirability Scale, the two Leven-
hypothesis
Downloaded From: http://archpsyc.jamanetwork.com/ by a McMaster University User on 10/11/2015
scales, and the scales from both forms of the POI showed no discernible trends in the hypothesized direc¬ tions. The validity data available indicate that these instru¬ ments are sufficiently sensitive to disclose differences that may exist between individuals along the measured dimen¬ sions and to discriminate between clinical and nonclinical groups. However, like most current psychometric methods, these instruments may be somewhat limited in their sensitivity to subtle deviations from "normal." The findings cast serious doubts on the view pervasive among both the professional and lay populations that holds that remitted individuals with this disorder are consistent¬ ly less healthy than the general population, and casts doubt as well on the psychodynamic view that these individuals tend to be dependent on others and the environment to the son
point of being extremely externally oriented and otherdirected in behavior, attitudes, and values. The results reported in this study clearly differ from such observations and reports, and there are two likely explanations for this discrepancy. Unlike subjects in most past research and psychoanalytic case studies, manic-depressives in this study were all taking lithium carbonate, medication that has been demonstrated to reduce the frequency and inten¬ sity of manic and depressed episodes. Second, this study also differed in that it employed a matched comparison group. Research by Winokur et al"' included no control group measuring the incidence of symptomatology in normal subjects over the two-year observation period. Reports by psychoanalytic theorists were essentially descriptive case studies also lacking systematic comparison of manic-depressives with the general population. In general, this study's results definitely lend empirical support to the widespread clinical impression that, when stabilized on lithium carbonate, manic-depressive individu¬ als frequently do not seem to want or need ongoing psychotherapy. The findings of this study seem consistent with those of Donnelly et al"1 in their comparison of bipolar and unipolar manic-depressives. Unlike unipolar subjects, bipolar manic-depressives showed relatively normal MMPI profiles during both an acute depressive phase and on remission. However a direct comparison of their results with those of the present study cannot be made due to the use of different instruments. The results of this study should also be examined in light of the subjective experience reported by many manicdepressive individuals during the initial screening inter¬ views. When describing their prelithium carbonate years, many related past feelings of helplessness and alarm associated with lack of control over the occurrence of severe mood swings. Many also spoke of the pervasive disruptive effects they thought these episodes had had on their lives in terms of employment, family, and social relationships, and on their self-concept. Yet this studyindicates that as a group these individuals do not differ on several important dimensions from those with no history of psychiatric problems. These findings give added cred¬ ence to the view that severe mood swings are related to a
recurrent alteration in brain neurotransmitters. Mood swings do not appear to be associated with major persis¬ tent psychological deficits in persons with this disorder,
begging for approval from "normal" society bereft of power and control over their totally feeling lives. These findings possibly also attest to the resiliency of human personality. nor
leave them
or
Allan Berman PhD, University of Rhode Island, assisted with this Louis Hafken, MD, cooperated with this study.
study.
References 1. Freud S: Mourning and melancholia, in Collected Papers. London, Hogarth Press, 1956, vol 4. 2. Abraham K: Notes on the psychoanalytical investigation and treatment of manic-depressive insanity and allied conditions, in Selected Papers on Psychoanalysis. New York, Basic Books Inc, 1953. 3. Cohen M, Baker G, Fromm-Reichmann F, et al: An intensive study of 12 cases of manic-depressive psychosis. Psychiatry 17:103-137, 1954. 4. Arieti S: Manic-depressive psychosis, in Arieti S (ed): American Handbook of Psychiatry. New York, Basic Books Inc, 1957. 5. Winokur G, Clayton PJ, Reich T: Manic-Depressive Illness. St Louis, CV Mosby Co, 1969. 6. Donnelly E, Murphy D: Primary affective disorder: MMPI differences between unipolar and bipolar depressed subjects.J Clin Psychol 29:303-306,
1973. 7. Donnelly E, Murphy D: Primary affective disorder: Bender-Gestalt sequence of placement as an indicator of impulse control. Percept Mot Skills 38:1079-1082, 1974. 8. Donnelly E, Murphy D: Social desirability and bipolar affective disorder. J Consult Clin Psychol 41:469, 1975. 9. Donnelly E, Murphy D, Scott W: Perception and cognition in patients with bipolar and unipolar depressive disorders. Arch Gen Psychiatry 32:1128-1131, 1975. 10. Donnelly E, Murphy D, Goodwin F: Cross-sectional and longitudinal comparisons of bipolar and unipolar depressed groups on the MMPI. J Consult Clin Psychol 44:233-237, 1976. 11. Hollingshead AB, Redlich FG: Social Class and Mental Illness. New York, John Wiley & Sons Inc, 1958. 12. Feighner J, Robins E, Guze S, et al: Diagnostic criteria for use in psychiatric research. Arch Gen Psychiatry 26:57-63, 1972. 13. Crowne D, Marlowe D: The Approval Motive. New York, John Wiley & Sons Inc, 1964. 14. Levenson H: Distinctions within the concept of internal-external control: Development of a new scale, in Proceedings of the 80th Annual Convention of the American Psychological Association. Washington, DC, American Psychological Association Inc, 1972, vol 7, pp 261-262. 15. Levenson H: Perceived parental antecedents of internal, powerful others and chance locus of control orientations. Dev Psychol 9:268-274, 1973. 16. Levenson H: Multidimensional locus of control in psychiatric patients. J Consult Clin Psychol 4:397-404, 1973. 17. Levenson H: Activism and powerful others: Distinctions within the concepts of internal-external control. J Pers Assess 38:377-383, 1974. 18. Levenson H, Miller J: Multidimensional locus of control in sociopolitical activists of conservative and liberal ideologies. J Pers Soc Psychol 33:199-208, 1976. 19. Prociuk T, Breen L: Defensive externality and its relation to academic performance. J Pers Soc Psychol 31:549-556, 1975. 20. Shostrum E: EITS Manual for the Personal Orientation Inventory. San Diego, Educational and Industrial Testing Service, 1966 (supplemented
1968).
21. Knapp R: The Measurement ofSelf-Actualization and Its Theoretical Implications. San Diego, Educational and Industrial Testing Service,
1971.
Inventory for Psychol 44:676-677,
22. Starrett RH: Modifications of the Personal Orientation
improved specific 1976. 23. Witkin
trait
prediction.
J
Consult Clin
H, Oltman P, Raskin E, et al: A Manual for the Embedded Figures Tests. Palo Alto, Calif, Consulting Psychologists Press, 1971. 24. Overall J, Gorham D: The Brief Psychiatric Rating Scale. Psychol Rep 10:799-812,
1962.
Downloaded From: http://archpsyc.jamanetwork.com/ by a McMaster University User on 10/11/2015
R. MacVane,
MA; James
D.
Lange, MA; Walter
Armin
Remission
Brown, MD; Mary Zayat
\s=b\ Several dimensions of psychological functioning of bipolar manic-depressives in remission were investigated. The following hypotheses were examined: (1) During remission, these individuals exhibit psychological maladjustment or a lack of positive mental health. (2) Bipolar manic-depressives are characterized by strong external orientations. The Personal Orientation Inventory, Marlowe-Crowne Social Desirability Scale, Levenson's
Internal and Powerful Others Locus of Control Scales, and the Embedded Figures Test were administered to matched groups of 35 bipolar manic-depressives and 35 normal comparison subjects. Discriminant analysis disclosed no significant differences between the two groups on measures of positive mental health and on measures of external orientation. The relationship between these results and the etiology and treatment of manicdepressive disorder is briefly discussed, as well as the contradiction these findings pose to current views of this disorder pervasive among professional and lay populations. (Arch Gen Psychiatry 35:1351-1354, 1978)
manic-depressive disorder characterized by dramatic mood shifts from severe depression to episodes of mania has long been of interest. Psychoanalytic theorists1"1 contended that during remission, an abnormal dependent character formation is present in the manicdepressive. Arieti4 described the manic-depressive as tend¬ ing "to be a conformist, willing to accept what he is given by his surroundings and to rely less than the average person on his own interpretations and evaluations of the external world."1"111 Winokur et als studied remitted manic-depressives and found that only 61% of their subjects reported periods between psychotic episodes that were symptom free. A
Bipolar
...
Accepted for publication May 15, 1978. From the Department of Psychology, University of Rhode Island (Ms MacVane), the Veterans Administration Hospital (Mr Lange, Dr Brown, and Ms Zayat), and the Department of Psychiatry, Brown University (Dr Brown), Providence. Reprint requests to Veterans Administration Hospital/151, Psychiatric Services, Davis Park, Providence, RI 02908 (Dr Brown).
series of studies comparing hospitalized bipolar (recurrent mania and depression) manic-depressive subjects with hospitalized unipolar (depression only) subjects disclosed significant differences between the two groups. Hospital¬ ized bipolar subjects appeared to be generally less neurotic and more healthy than unipolar subjects."" Subjects with bipolar disorder are also reported as more impulsive than unipolar individuals.710 With the current widespread use of lithium carbonate therapy, many manic-depressives are spending much of their time in remission. Consequently, information about their psychological functioning during remission is of interest both in understanding the psychology of this disorder and in understanding how to treat the manicdepressive individual psychotherapeutically. The study reported here has investigated several dimen¬ sions of psychological functioning of bipolar manic-depres¬ sives in remission. The following hypotheses were exam¬ ined: (1) During remission, bipolar manic-depressive indi¬ viduals exhibit psychological maladjustment or a lack of positive mental health. (2) Bipolar manic-depressives are characterized by strong external orientations. The first hypothesis is consistent with the psychoanalytic view that manic-depressives are victims of ever-present psychopath¬ ology that occasionally flares up into a depressed or manic episode. This hypothesis has also been an outgrowth of the fact that prospective subjects being interviewed for our study frequently described recurrent mood swings as intensely damaging to all aspects of their lives. Regardless of etiology, the actual occurrence of repeated manic and depressed episodes, disruptive both intrapsychically and interpersonally, could impair psychological health even when the individual is in remission. The second hypothesis refers to the psychoanalytic description of bipolar manicdepressives as being excessively conforming and depen¬ dent on others. Cohen et al' reported that individuals with this disorder try to reach success or satisfaction through the success of others rather than by their own efforts and
Downloaded From: http://archpsyc.jamanetwork.com/ by a McMaster University User on 10/11/2015
they tend to deny their own originality and, conse¬ quently, are often stereotyped in attitudes and opinions. that
Subjects 35 remitted bipolar manic-depressives (12 subjects women and 23 men) and 35 healthy comparison subjects (14 women and 21 men). The subjects in each group were matched with regard to age, sex, and social position as estimated by Hollingshead's Two-Factor Index of Social Position." The mean age of the manic-depressive group was 49.9 years and that of the comparison group 45.5 years, with no significant difference between them (f 1.40; > .05). For social position, the mean of the manicdepressive group was 3.11 and that of the comparison group 3.03, with no significant difference between the two groups (r 0.35; were
=
=
>
.05).
Outpatients with diagnoses of bipolar manic-depressive disor¬ der currently taking lithium carbonate were referred to the study from a Veterans Administration Hospital, a psychiatrist's private practice, a community mental health center, and a medical center's lithium carbonate clinic. Selection of bipolar manic-depressive subjects was made on the basis of information obtained from hospital and case records and an unstructured one-hour interview with each prospective subject focusing on past psychiatric prob¬ lems and treatment and details of their manic and depressive episodes. Decisions to include individuals as subjects were made independently by two researchers in accordance with the Feighner et al12 research criteria for bipolar manic-depressive disorder. The occurrence of at least one manic and one depressive episode was required, with each episode having to fulfill specific criteria. The
criteria excluded anyone who could also be classified into any other
diagnostic category (eg, alcoholism, schizophrenia, organic brain syndrome). Of the 53 individuals referred to the study and interviewed, 17 failed to meet the research criteria and were excluded. In selecting subjects,
no
attempt
was
made to
assess
how well
potential subjects were functioning while between affective episodes. The only requirements for manic-depressive subjects were that they (1) fulfilled the research criteria mentioned, (2) were currently taking lithium carbonate, and (3) were clearly between episodes. The presence of remission was assessed at the time of psychological testing. Brief interviews were conducted with each subject by two researchers followed by independent ratings on a modified version of the Brief Psychiatric Rating Scale (BPRS). Of the 36 manic-depressive subjects meeting the first two criteria, only one was later excluded from data analysis on the basis of the third criterion. This subject's BPRS ratings indicated
hypomanic symptomatology suggesting the possible onset of a manic episode. Subjects in the comparison group were individuals reporting no history of psychiatric disorder nor any treatment or medication for psychological problems or alcoholism in the past. These subjects were also independently rated on the BPRS at the time of testing by two researchers and the same exclusion criteria applied. None were excluded due to the presence of significant psychiatric symptoms. Comparison subjects received $5 for their participa¬ tion. All subjects gave informed consent after the been fully explained.
tendency to describe oneself favor¬
Internal and Powerful Others Locus of Control Scales.—These two are from a set of three locus of control orientation scales developed by Levenson and others." '" The scales represent two
scales
SUBJECTS AND METHODS The
instrument does measure this ably to win social approval.
procedures
had
Instruments Social Desirability Scale.-This scale attempts to measure the extent of an individual's tendency to describe himself in socially desirable terms to receive approval from others. Through a number of validity studies looking at the scale's behavioral correlates, Crowne and Marlowe1 ' have demonstrated that this
factors in that only items from the appropriate scale load on that one factor. A number of studies by Levenson and others" demonstrate acceptable validity of the scales. Personal Orientation Inventory.-The Personal Orientation Inven¬ tory (POI) attempts to measure self-actualization as defined by Maslow and also aims at measuring a person's mental health. Because this study focused on manic-depressives in terms of their psychological health, the POI was considered more appropriate than the Minnesota Multiphasic Personality Inventory (MMPI) and related tests that provide estimates of psychopathology rather than measures of positive mental health. Reviews by Shostrum™ and Knapp-'1 include a large number of studies demonstrating acceptable discriminant and predictive validity of this inventory as well as its correlation with other scales. For this study two scoring systems were used: the original scoring system of the 12 scales, and a modified scale form of the POI developed by Starrett.-'2 Embedded Figures Test.-The Embedded Figures Test (EFT) is a perceptual test that quantifies the subject's perceptual style on the dimension of field articulation. Individuals scoring in the fieldindependent range are believed to have an awareness of their needs, feelings, and personal attributes as being distinct from those of others. Field-dependent individuals, however, appear to have a less developed sense of separate identity and tend to rely more on external sources for definition of their attitudes and views of themselves.21 Brief Psychiatric Rating Scale.—The BPRS is a set of construct scales widely used as a rapid assessment technique for psychiatric patients." It consists of 18 symptom constructs and uses a seven-point scale of severity. A modified form of the BPRS was employed that was recently developed by a research group studying manic-depressive disorders at the National Institutes of Health (L.B. Bigelow, MD, D.L. Murphy, MD, oral and written communication, 1976). It consists of the original items plus four additional items, three of which are related to mania and one related to depression.
conceptually pure '"
Procedure
Subjects meeting the research criteria
were
tested within the
Immediately prior to subject for about 15 minutes and then rated him or her independently on the modified BPRS. Subjects took the tests in a varied order of presentation to offset any ordering effect. Using the SPSS-Subprogram discriminant, the direct method, three separate discriminant analyses were performed on the ten months following evaluation interviews. testing, two researchers interviewed each
data.
RESULTS
The BPRS scores of the manic-depressive group were clearly indicative of remission. Their mean scores on the 22 items ranged from 1.0 to 1.3 (1, symptom not present; 2, very mild) with the exception of items "somatic concern" and "tension," with means of 1.5 and 1.8, respectively. Comparison group mean scores ranged from 1.0 to 1.2, excluding "tension" with a mean of 1.5. Commonly used statistical procedures for computing interrater reliability were inappropriate for use with our BPRS ratings due to the limited range of BPRS scores obtained by the subjects. Most of the ratings subjects
Downloaded From: http://archpsyc.jamanetwork.com/ by a McMaster University User on 10/11/2015
Table 1.—Means and Standard Deviations for the Manie-Depressive and Comparison Groups
Comparison (N 35) =
Scale
Mean Personal Orientation Inventory Modified scale form 10.14 1 5.09
Original
SD
10
Mean
12 Embedded Figures Test Social Desirability Scale Internal Locus of Control Scale Powerful Others Locus of Control Scale
=
35)
Discriminant
SD
Function
2
Weights
df
Assorted scales 3.25
9.89 5.51 14.63 7.17 15.37
2.78 1.50 3.29 1.42
13.37 7.80 16.57
4.33 1.39 3.23
17.66 83.66 20.57
3.20
16.86
3.15
11.1£ 2.54
84.00
10.64 2.66
19.91
4.59
15.37 12.29 12.91 15.97 11.94 7.03
3.24 2.53
15.69 17.97
=
=
4.06
scales
1
Table 2.—Discriminant Analysis Results for Three Separate Analyses (n, 35; n,,
Manie-Depressive 35) (
1.84 3.83 1.47
3.20
20.40 20.49 15.66 12.40 12.11 16.17 11.83 7.06 15.77
Embedded Figures Test Social Desirability Scale Internal Locus of Control Scale
5
>.05
12
>.05
1_-0.15916
2
3. Personal Orientation
0.48458 0.62450 0.8797 8.396 -0.59118 -0.27558
Inventory
1__0.61626 2_1.05040 3
-0.19540
0.76246 0.15070 0.8531 9.851 0.77985 -0.77985 0.24821
52.00
5.78
34.71
4.97
34.03
5.35
10
0.01575 0.12194
23.14
6.17
24.43
7.18
12
-1.43104
42.35
>05
-0.21840
17.60 97.90 16.00
90.50 17.23
4
Powerful Others Locus of Control Scale 0.48574 2. Personal Orientation Inventory
3.29
2.91 2.71 2.67 3.58 1.84 1.16
0.63351 -0.86483 0.9587 2.784
3.38 5.87
-0.73975
received fell between 0 and 3. To compute interrater reliability, we first looked at each of the 22 BPRS items separately, and totaled the scores given by each rater across all 70 subjects on that particular item. Next, these two sums from raters 1 and 2 were compared to determine how large a discrepancy existed between their ratings. To do this, a "difference score" for each BPRS item was computed by subtracting the total from rater 1 on that particular item from the total from rater 2. Therefore, a low difference score for an item would indicate a high degree of similarity (high interrater reliability) between its two sets of ratings across the 70 subjects. The mean of the 22 difference scores was 2.27, indicating a very high degree of similarity between the two sets of ratings across all the BPRS items. Means and standard deviations for the two groups on all measures are presented in Table 1. As an additional check, t tests were performed on the EFT, the Social Desirability Scale, and the two Levenson scales. No significant differ¬ ences between groups were found. The t test was not performed on POI scales because this instrument was designed to yield a profile of scores, making it statistically inappropriate to interpret the scales individually. Three separate discriminant analyses were performed on the following measures on the two groups: (1) the 12 original POI scales (hypothesis 1), (2) Starrett's modified scale form of the POI (hypothesis 1), and (3) the Embedded Figures Test, the Marlowe-Crowne Social Desirability Scale, and the Levenson Internal and Powerful Others scales (hypothesis 2). Results of the three analyses are
presented in Table 2. These analyses showed no significant differences between the manic-depressive and comparison
on any of the measures. In addition, none of these three combinations of measures successfully discriminated between the two groups to the extent of being able to then classify a significant number of the 70 subjects into their correct groups. None of the three separate discriminant functions reached the significance level.
groups
COMMENT
The results of this study indicate that remitted manicdepressives do not differ from individuals with no history of psychiatric problems in terms of positive mental health as measured by the POI and also on a number of dimen¬ sions regarding external vs internal orientation. In this last respect, the two groups did not differ significantly on the field dependence/independence dimension, the desire to win social approval, or the extent to which they see events in their lives as being under their control. While not nullifying these findings, the higher male-female ratio in the sample does not reflect the male-female ratio in the manic-depressive population.
The data do not offer any evidence to contradict the null that no difference exists between the two groups on the measures. This is further supported by the fact that strict research criteria for manic-depressive groups were strictly adhered to; also, both the manicdepressive and comparison groups were well defined. Furthermore, the means of the two groups on the EFT, the Marlowe-Crowne Social Desirability Scale, the two Leven-
hypothesis
Downloaded From: http://archpsyc.jamanetwork.com/ by a McMaster University User on 10/11/2015
scales, and the scales from both forms of the POI showed no discernible trends in the hypothesized direc¬ tions. The validity data available indicate that these instru¬ ments are sufficiently sensitive to disclose differences that may exist between individuals along the measured dimen¬ sions and to discriminate between clinical and nonclinical groups. However, like most current psychometric methods, these instruments may be somewhat limited in their sensitivity to subtle deviations from "normal." The findings cast serious doubts on the view pervasive among both the professional and lay populations that holds that remitted individuals with this disorder are consistent¬ ly less healthy than the general population, and casts doubt as well on the psychodynamic view that these individuals tend to be dependent on others and the environment to the son
point of being extremely externally oriented and otherdirected in behavior, attitudes, and values. The results reported in this study clearly differ from such observations and reports, and there are two likely explanations for this discrepancy. Unlike subjects in most past research and psychoanalytic case studies, manic-depressives in this study were all taking lithium carbonate, medication that has been demonstrated to reduce the frequency and inten¬ sity of manic and depressed episodes. Second, this study also differed in that it employed a matched comparison group. Research by Winokur et al"' included no control group measuring the incidence of symptomatology in normal subjects over the two-year observation period. Reports by psychoanalytic theorists were essentially descriptive case studies also lacking systematic comparison of manic-depressives with the general population. In general, this study's results definitely lend empirical support to the widespread clinical impression that, when stabilized on lithium carbonate, manic-depressive individu¬ als frequently do not seem to want or need ongoing psychotherapy. The findings of this study seem consistent with those of Donnelly et al"1 in their comparison of bipolar and unipolar manic-depressives. Unlike unipolar subjects, bipolar manic-depressives showed relatively normal MMPI profiles during both an acute depressive phase and on remission. However a direct comparison of their results with those of the present study cannot be made due to the use of different instruments. The results of this study should also be examined in light of the subjective experience reported by many manicdepressive individuals during the initial screening inter¬ views. When describing their prelithium carbonate years, many related past feelings of helplessness and alarm associated with lack of control over the occurrence of severe mood swings. Many also spoke of the pervasive disruptive effects they thought these episodes had had on their lives in terms of employment, family, and social relationships, and on their self-concept. Yet this studyindicates that as a group these individuals do not differ on several important dimensions from those with no history of psychiatric problems. These findings give added cred¬ ence to the view that severe mood swings are related to a
recurrent alteration in brain neurotransmitters. Mood swings do not appear to be associated with major persis¬ tent psychological deficits in persons with this disorder,
begging for approval from "normal" society bereft of power and control over their totally feeling lives. These findings possibly also attest to the resiliency of human personality. nor
leave them
or
Allan Berman PhD, University of Rhode Island, assisted with this Louis Hafken, MD, cooperated with this study.
study.
References 1. Freud S: Mourning and melancholia, in Collected Papers. London, Hogarth Press, 1956, vol 4. 2. Abraham K: Notes on the psychoanalytical investigation and treatment of manic-depressive insanity and allied conditions, in Selected Papers on Psychoanalysis. New York, Basic Books Inc, 1953. 3. Cohen M, Baker G, Fromm-Reichmann F, et al: An intensive study of 12 cases of manic-depressive psychosis. Psychiatry 17:103-137, 1954. 4. Arieti S: Manic-depressive psychosis, in Arieti S (ed): American Handbook of Psychiatry. New York, Basic Books Inc, 1957. 5. Winokur G, Clayton PJ, Reich T: Manic-Depressive Illness. St Louis, CV Mosby Co, 1969. 6. Donnelly E, Murphy D: Primary affective disorder: MMPI differences between unipolar and bipolar depressed subjects.J Clin Psychol 29:303-306,
1973. 7. Donnelly E, Murphy D: Primary affective disorder: Bender-Gestalt sequence of placement as an indicator of impulse control. Percept Mot Skills 38:1079-1082, 1974. 8. Donnelly E, Murphy D: Social desirability and bipolar affective disorder. J Consult Clin Psychol 41:469, 1975. 9. Donnelly E, Murphy D, Scott W: Perception and cognition in patients with bipolar and unipolar depressive disorders. Arch Gen Psychiatry 32:1128-1131, 1975. 10. Donnelly E, Murphy D, Goodwin F: Cross-sectional and longitudinal comparisons of bipolar and unipolar depressed groups on the MMPI. J Consult Clin Psychol 44:233-237, 1976. 11. Hollingshead AB, Redlich FG: Social Class and Mental Illness. New York, John Wiley & Sons Inc, 1958. 12. Feighner J, Robins E, Guze S, et al: Diagnostic criteria for use in psychiatric research. Arch Gen Psychiatry 26:57-63, 1972. 13. Crowne D, Marlowe D: The Approval Motive. New York, John Wiley & Sons Inc, 1964. 14. Levenson H: Distinctions within the concept of internal-external control: Development of a new scale, in Proceedings of the 80th Annual Convention of the American Psychological Association. Washington, DC, American Psychological Association Inc, 1972, vol 7, pp 261-262. 15. Levenson H: Perceived parental antecedents of internal, powerful others and chance locus of control orientations. Dev Psychol 9:268-274, 1973. 16. Levenson H: Multidimensional locus of control in psychiatric patients. J Consult Clin Psychol 4:397-404, 1973. 17. Levenson H: Activism and powerful others: Distinctions within the concepts of internal-external control. J Pers Assess 38:377-383, 1974. 18. Levenson H, Miller J: Multidimensional locus of control in sociopolitical activists of conservative and liberal ideologies. J Pers Soc Psychol 33:199-208, 1976. 19. Prociuk T, Breen L: Defensive externality and its relation to academic performance. J Pers Soc Psychol 31:549-556, 1975. 20. Shostrum E: EITS Manual for the Personal Orientation Inventory. San Diego, Educational and Industrial Testing Service, 1966 (supplemented
1968).
21. Knapp R: The Measurement ofSelf-Actualization and Its Theoretical Implications. San Diego, Educational and Industrial Testing Service,
1971.
Inventory for Psychol 44:676-677,
22. Starrett RH: Modifications of the Personal Orientation
improved specific 1976. 23. Witkin
trait
prediction.
J
Consult Clin
H, Oltman P, Raskin E, et al: A Manual for the Embedded Figures Tests. Palo Alto, Calif, Consulting Psychologists Press, 1971. 24. Overall J, Gorham D: The Brief Psychiatric Rating Scale. Psychol Rep 10:799-812,
1962.
Downloaded From: http://archpsyc.jamanetwork.com/ by a McMaster University User on 10/11/2015
