Psychother. Psychosom. 26: 12 19 (1975)
Aggression in General-Hospital Patients Ranan Rimón, Juhani Jussila and Olavi Katila
Abstract. The Buss-Durkee Inventory (BDI) rating for measuring aggression was per formed on 125 consecutive female patients admitted to a general hospital. The test scores of different diagnostic groups were all on a intermediate level, and with the exception of slightly lower scores in patients with blood disorders, only few differences reaching statisti cal significance were observed. The factor analysis of the test results revealed three factors: in addition to aggression factor (I) and hostility factor (II) a factor called remorse factor (III) was identified. The patient groups divided according to their medical diagnoses showed no differences with regard to these factors. The variance of factor scores could not be explained by such background variables as social class, marital status or domicile. There was, however, a highly significant negative correlation ( 0.33) between age and the aggression factor. Finally, the concept of suppressed aggression and hostility and the role of aggression dynamics in the development of organic illness is discussed.
Introduction
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Many psychosomatically oriented ecological investigations have shown that episodes of different organic illnesses are not distributed at random among the general population. Life experiences, psychosocial factors and personality characteristics have been shown to influence general susceptibility to illness (Hinkle and Wolff, 1958; Hinkle, 1961 \ Rahe et al., 1964; Rahe, 1972). The psychodynamic basis of the personality traits involved, especially concerning psychosomatic disorders, has been interpreted in different ways according to the psychiatric discipline to which the investigators belonged to, e.g. frustration in infancy, masochistic dependence upon the mother, distortion of psychosexual identity, and rejection of either masculine or feminine roles are concepts which the pertinent analytically oriented literature almost axiomatically uses. Again and again, the significance of repressed or inhibited aggression in some disease entities is emphasized (Bastiaans, 1969). Several authors maintain that patients with hypertension (Kaplan et a l, 1961), peptic ulcer (Kezur et al, 1951), ul cerative colitis (Engel, 1955), irritable colon syndrome (Engel, 1967), bronchial asthma (Stein and Sltiavi, 1967), rheumatoid arthritis (Johnson et al., 1947;
Rimón/Jussila/Katila
13
Cobb, 1959), and even patients with malignant diseases (Renneker and Cutler, 1952; Blumberg et ai, 1954) are unable to deal adequately with their aggressive impulses and hostile attitudes, and are inclined to inhibit the expression of their aggression. The prolonged buildup of undischarged hostile and aggressive ten dencies is believed to provoke an associated chronic physiological reaction which, in a certain life situation, may result in the development of a clinical disorder of a predisposed organ or organ system (Alexander, 1950). The concept of ‘aggression’ is by no means unambiguous but in any case it includes several, different components (Kaufman, 1965; Groen, 1972). Buss (1961) divides aggression into the outwardly directed instrumental response of aggression, the emotional reaction of anger, and the attitude of hostility. In addition, aggression can be divided into physical versus verbal, active versus passive, direct versus indirect, spontaneous versus provoked, and offensive versus defensive (Pitkánen, 1969). In many psychosomatic studies, aggression has been understood as a completely unambiguous idea, and it has been measured mainly by projective methods (e.g., Cleveland and Fisher, 1960; Geist, 1966). The aim of this investigation was to elucidate by means of an inventory technique to what extent the aggression dynamics of unselected general-hospital patients, including many with psychosomatic diseases, would exhibit significant psychological determinants relevant to the illness of the patients. Another aim of the study was to analyze the structure of the Finnish version of the Buss-Durkee Inventory (BDI) used (Buss and Durkee, 1957).
Subjects and Methods
/
Cardiovascular disorders coronary heart disease, congestive heart failure, peripheral vascular insufficiency
21
2
Endocrine diseases diabetes, hyperthyroidism
19
3
Connective tissue disturbances rheumatoid arthritis, collagénoses, disc degenerations
13
4
Neurological disorders organic brain syndromes, convulsive disorder, multiple sclerosis, parkinsonism
12
5
Gastro-intestinal diseases peptic ulcer, pancreatitis, liver and biliary diseases
17 Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Subjects The series consisted of 125 patients who were admitted to the Department of Medicine of the University Central Hospital in Turku. The mean age of the patients was 44.2 (SD = 15.6) years. The patients were classified according to their diagnoses as follows:
Rimón/Jussila/Katila
14
6
Blood and lymphatic disorders splenomegaly, leukemias, Hodgkin’s disease
12
7
Infections acute meningitis, tonsillitis, etc.
10
8
Respiratory diseases bronchial asthma, emphysema, bronchiectasis
12
9
Urinary tract disorders haematuria, nephrosis Total
9
125
The groups were comparable as to the age and social and marital status.
Treatment o f the Data and Results Means, standard deviations and internal consistency of reliability estimates (KR-20) of the BD1 scales are presented in table I. These scores are similar to the scores of the hospital patient sample for anonymous inventories reported by Buss (1961). The slight variations may be attributed to the only moderate reliability level of the scales. The correlations between items and the total score of the respective scales varied between + 0.27 and + 0.67. In table II the BDI aggression scores of the various diagnostic subgroups are presented. Generally, the total scores are all on intermediate level and comparable with each other. The total scores of groups 6 (blood disorders) and 9 (urinary tract diseases) are, however, slightly lower than those of the other groups, but the differences do not reach statistical significance (F-test and t-test). In group 6 the scores of all item categories except that of ‘irritability’ are below the mean, and when compared to corresponding scores of several other groups (e.g., groups 2 -4 , 7, 8), statistically significant differences (p < 0.05) can be observed as to the scales negativism’, ‘resentment’, ‘suspicion’ and ‘guilt’. In group 9, however, the only sta tistically significant differences (p < 0.05) are noted when the score of ‘negativism’ is compared to the corresponding test figure of group 1 (cardiovascular disorders), and when the score of ‘resentment’ is compared to that of group 4 (neurological disorders). In both instances this is, however, largely due to the fact that the scores of ‘negativism’ in group 1
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Procedure The BD1 for measuring aggression was used. It has been proved useful in numerous experiments with students, patients of general hospitals and out-patients of several clinics, and psychiatric patients (Buss, 1961; Buss et al, 1962: Rimon, 1969; Friedman, 1970; Persky et al., 1971; Olkinuora, 1972). This test attempts to measure several components of aggressiveness, consisting of eight scales, which together tap a variety of aggressive be haviours and hostile attitudes. Five scales, i.e. ‘assault’, ‘indirect’, ‘irritability’, ‘negativism’ and ‘verbal’, are presumed to represent the tendencies to the instrumental response of aggression and to anger. Two other scales, i.e. ‘resentment’ and ‘suspicion’, are intended to measure various hostile attitudes, and the final scale ‘guilt’ gives information on guilt feelings second ary to conscious or inhibited unconscious aggressive impulses. All BDIs were filled out anonymously. However, a hidden coding system revealed the connection between each patient and the corresponding inventory.
15
Aggression in General-Hospital Patients
Table I. Means, standard deviations and KR-20 estimates of reliability of the BD1 scales (N= 125) Scale
Number of items
Mean
SD
KR-20
1 Assault 2 Indirect 3 Irritability 4 Negativism 5 Resentment 6 Suspicion 7 Verbal 8 Guilt
10 9 11 5 8 10 13 9
3.5 4.4 6.2 2.4 25 4.4 5.4 5.3
2.0 2.0 2.3 1.4 1.8 2.4 2.8 2.3
0.54 0.60 0.60 0.52 0.55 0.66 0.65 0.69
and that of ‘resentment’ in group 4 are exceptionally high and well above the average. The test score of the item category ‘negativism’ in group 1 is also significantly (p < 0.01) higher that that of group 2 (endocrine diseases), and also higher than the corresponding figures of groups 5 (gastro-intestinal diseases) and 6 (p < 0.05). Similarly, the high score of resent ment’ in group 4 is significantly higher than that of group 6 (p < 0.05). The score of the scale ‘verbal’ in group 1 is, however, below the mean and significantly lower than that of group 5 (p < 0.05) for instance. On theoretical arguments and on the basis of earlier empirical analyses, the extraction of two factors was to be expected (Buss and Durkee, 1957; Persky et al., 1971). Because underfactoring can distort the factor structure more than overfactoring, we extracted 4 factors. These explained 48.1 % of the total variance. An orthogonal Varimax rotation was made using 2, 3 and 4 factors. Only factor loadings of 0.40 or above were included in the interpretation. The rotation solution of 2 and 3 factors were both interpretable. The rotation with 4 factors produced a specific factor IV; further treatment of this solution was omitted. The factor structure in the two-factor solution corresponded to the oblique rotation solution presented by Buss and Durkee (1957). The first factor could be identified as ‘aggression factor’ and the second as a ‘hostility factor’. ‘Guilt’ had no substantial loading on either factor. The rotation with three factors (table 111) indicated that ‘guilt’ forms a factor III together with ‘assault’ and ‘negativism’. This factor was named ‘remorse’ factor. The above factors were treated as new variables. Factor scores were estimated with the regression method. The multiple correlations of the scales with the factors were 0.79 (I), 0.75 (II), and 0.67 (III) indicating that the estimation was adequate. One-way analyses of variance and paired t-tests revealed no significant differences between patient groups with regard to these factors. The variance of factor scores could not either be explained by such background variables as social class, marital status or domicile (urban/rural). There was, how ever, a highly significant negative correlation (-0.33) between age and the aggression factor.
The most striking finding of this study was no doubt the fact that very similar patterns of aggression dynamics were found in patients with various
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Discussion
16
Rimon/Jussila/Katila Table II. The mean scores obtained from the BDI in various diagnostic subgroups Scale
SD
mean
SD
mean
SD
mean
SD
3.2 4.2 6.0 3.2 2.6 4.4 4.4 5.5
1.4 2.3 2.1 1.2 2.0 2.9 2.5 2.6
3.6 4.7 6.2 2.2 2.3 4.8 5.6 5.1
2.2 1.9 2.3 1.1 1.7 2.2 2.9 2.3
3.5 3.6 5.8 2.5 2.5 4.1 5.9 5.9
2.0 1.9 2.5 1.4 1.5 2.1 2.8 2.0
3.6 4.2 6.5 2.7 3.4 4.2 6.3 6.0
1.4 1.6 2.8 1.5 1.7 2.1 2.9 2.6
33.5
12.1
34.5
11.8
33.8
12.0
36.9
10.8
II
e
mean
II
co
2, n= 19
c
Total
12
1, n = 21
CO
1 Assault 2 Indirect 3 Irritability 4 Negativism 5 Resentment 6 Suspicion 7 Verbal 8 Guilt
Diagnostic subgroups
Scale
I
11
III
h2
1 Assault 2 Indirect 3 Irritability 4 Negativism 5 Resentment 6 Suspicion 7 Verbal 8 Guilt
49 50 61 14 35 15 69 25
18 25 25 16 63 66 14 28
47 26 25 48 22 25 14 50
50 38 50 28 57 52 51 39
Eigenvalue Percentage
1.5 19.6
1.1 13.9
1.0 12.0
3.6 45.5
medical diseases. The low scores of several item categories, especially in those reflecting expression of resentment, hostility and guilt, in the patients with blood disorders is probably due to the fact that more than half of the cases (7/12) in this group had acute leukemia or Hodgkin’s disease. As Baltrusch (1969) and Achte et at. (1970) have emphasized, patients with malignancies frequently struggle with restraints on aggressiveness, suppressed guilt feelings and introverted hostility. The similarly low scores of patients with urinary tract diseases is, on the contrary, more surprising, and may depend on irrational but
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Table III. Varimax-rotated factor matrix (three factors)
Aggression in General-Hospital Patients
17
Total, n = 125
mean
SD
mean
8, n = 12
9, n = 9
SD
mean
SD
mean
SD
mean
SD
3.5 4.4 6.2 2.4 2.5 4.4 5.4 5.3
2.0 2.0 2.3 1.4 1.8 2.4 2.8 2.3
10
4.2 4.5 6.6 2.1 2.6 4.2 6.4 4.8
1.9 2.2 2.6 1.6 1.7 1.6 3.2 2.4
2.9 4.3 6.4 2.1 1.9 3.1 4.6 4.1
2.8 2.5 2.6 1.5 1.5 2.0 2.4 2.3
3.5 5.0 6.2 2.5 2.3 5.1 4.8 5.8
1.8 1.7 2.2 1.6 1.6 2.1 2.6 2.8
3.1 4.6 5.8 2.3 3.1 5.3 5.2 5.8
2.3 2.0 2.1 1.4 2.2 3.4 2.6 2.3
3.1 4.0 5.6 2.0 1.4 3.7 5.0 5.0
2.1 1.8 1.4 1.4 1.9 1.9 2.6 1.4
35.4
9.0
29.4
12.4
35.2
10.6
35.2
12.1
29.8
11.1
34.0 ± 1 1.2
dynamically significant concepts in relation to urinary and genital function and denial of guilt associated with the development of the illness. Another surprising finding was that patients with cardiovascular disorders (including 10 patients with coronary heart disease: CHD) had a high score on ‘negativism’ and a low score on the scale ‘verbal’. This is ostensibly at variance with many previous findings, mainly focussing on male patients, revealing an overtly aggressive personality profile in patients with CHD (Friedman and Rosenman, 1960; Riman and Rakkólainen, 1969). The relatively great number of patients (11) with cardiovascular disorders other than CHD in this group, and the weak reliability of the scale ‘negativism’ may, however, account for this unexpected observation. The high ‘resentment’ score in the group with neurological disorders, on the other hand, is quite easy to understand. This group consisted of 3 cases with convulsive disorder, 2 with multiple sclerosis and 2 with parkinsonism who had all encountered severe social and psychological problems in their work and con sequently had become embittered and resentful. The high total score in the group of respiratory diseases, including 9 cases of 12 with bronchial asthma, is at variance with the relatively common concept of inhibition and deficient integra tion of aggression dynamics in patients with bronchial asthma (Pierluot and van Roy, 1969). When the significance of the BDI score differences in various patient groups is discussed, it has to be emphasized that statistically significant differences may appear by chance in approximately 5 instances out of 100, when the significance level of 5 % is applied. In the present study, several paired t-tests were per-
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
C"
mean SD
II
6, n = 12
c
5, n = 17
Rimon/Jussila/Katila
18
formed, the groups were small, the reliability of the scales was not great, and the differences were after all relatively few. When the test results of the patients were investigated by factor analysis, i.e. on a more reliable level, no significant differences with regard to the medical diagnoses were observed. This indicates that the significant differences which occurred on scale level probably have arisen by chance. The factor analysis of the BDI revealed that, in addition to already de scribed aggression (I) and hostility (II) factors, ‘negativism’ can — even though its communality is relatively weak — form a factor III together with ‘guilt’ and ‘assault’. This factor III was named remorse factor. The interpretation of this factor is clear: when oppositional behaviour leads to physical violence against others, it happens in connection with guilt feelings. Aggression and hostility factors are quite distinct. The factorial validity among them highly points to the theoretical starting-point. Due to social and moral norms, free expression of aggression is not usually accepted. Subsequently, the repression, displacement and/or sublimation of it are important for an individual’s social adaptation. Because of the complex nature of human behaviour, the connection between provoking stimuli and re sponse is not directly predictable. Man’s cognitive qualifications for appraising a situation make his behaviour less dependent on psychological drives than is possible for the lower species. Man is able to inhibit or attenuate aggression to situational requirements. Thus various aggressive tendencies may become in wardly directed and turn against the individual himself. The question whether this, in certain subjects, may result in the development of physical illness or particularly in psychosomatic symptom formation still remains unanswered. The results of the present study do not point to a positive support of the hypothesis that suppressed aggression and a chronic buildup of undischarged hostile and aggressive tendencies would be significant determinants in the development of organic illness.
Achte, K.A.; Vauhkonen, M.L., and Viitamdki, R.O.: Cancer and psyche. Monogr. No. 1. Psychiat. Clin. Helsinki Univ. Centr. Hosp. (1970). Alexander, F.: Psychosomatic medicine (New York 1950). Baltrusch, H.F.: Psychosomatische Beziehungen bei Krebskrankhciten. Z. psychosom. Med. 3: 196 (1969). Bastiaans, J.: The role of aggression in the genesis of psychosomatic disease. J. psychosom. Res. 13: 307 (1969). Blumberg, E.M.: West, P.M.. and Ellis, F.W.: A possible relationship between psychological factors and human cancer. Psychosom. Med. 16: 277 (1954). Buss. A.H.: The psychology of aggression (New York 1961). Bliss, A.H. and Durkee, A.: An inventory for assessing different kinds of hostility. J. cons. Psychol. 21: 343 (1957).
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
References
Aggression in General-Hospital Patients
19
Ranan Rimon, Karpantie 1 B, 70400 Kuopio 40 (Finland)
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Buss, A.H.; Fischer, H., and Simmons, A.J.: Aggression and hostility in psychiatric patients. J. cons. Psychol. 26: 84 (1962). Cleveland, S.E. and Fischer, S.: A comparison of psychological characteristics and physio logical reactivity in ulcer and rheumatoid arthritis groups. Psychosom. Med. 22: 283 (1960). Cobb, S.: Contained hostility in rheumatoid arthritis. Arthr. and Rheum. 2: 419 (1959). Engel, G.L.: Studies on ulcerative colitis. III. The nature of the psychological processes. Amer. J. Med. 19: 231 (1955). Engel, G.L.: Irritable cohen syndrome; in Friedman and Kaplan Comprehensive textbook of psychiatry, p. 1058 (Baltimore 1967). Friedman, A.S.: Hostility factors and clinical improvement in depressed patients. Arch. gen. Psychiat. 23: 524 (1970). Friedman, M. and Rosenman, R.H.: Overt behaviour pattern in coronary disease. J. amer. med. Ass. 173: 1320 (1960). Geist, H.: The psychological aspects of rheumatoid arthritis (Springfield 1966). Groen, J.J.: The study of human aggression. Psychother. Psychosom. 20: 312 (1972). Hinkle, L.E.: Ecological observations of the relation of physical illness, mental illness and the social environment. Psychosom. Med. 23: 289 (1961). Hinkle, L.E. and Wolff, H.G.: Ecologic investigations of the relationship between illness, life experiences, and the social environment. Ann. intern. Med. 49: 1373 (1958). Johnson, A.: Shapiro, L.B., and Alexander, F.: Preliminary report on a psychosomatic study of rheumatoid arthritis. Psychosom. Med. 9: 295 (1947). Kaplan, S.M.; Gottschalk, L.A.: Magliocco, E.B.; Rohovit, D.D., and Ross, W.D.: Hostility in verbal productions and hypnotic dreams of hypertensive patients. Psychosom. Med. 23: 312 (1961). Kezur, E.; Kapp, F., and Rosenbaum, 11.: Psychological factors in women with peptic ulcer. Amer. J. Psychiat. 108: 368 (1951). Kaufman, H.: Definitions and methodology in the study of aggression. Psychol. Bull. 64: 351 (1965). Olkinuora, M.: A factor analytical study of psychosocial background in bruxism. Proc. finn. dent. Soc.68: 184 (1972). Persky, H.; Smith, K.D., and Basu, G.K.: Relation of psychologic measures of aggression and hostility to testosterone production in man. Psychosom. Med. 33: 265 (1971). Pierloot, R.A. and Roy, J. van: Asthma and aggression. J. psychosom. Res. 13: 333 (1969). Pitkdnen, A descriptive model of aggression and nonaggression with applications to children’s behaviour; Diss. Jyvaskyla (1969). Rahe, R.H.; Meyer, M.; Smith, M.; Kjaer, G., and Holmes, T.H.: Social stress and illness onset. J. psychosom. Res. 8: 35 (1964). Rahe, R.H.: Subjects’ recent life changes and their nearfuture illness reports. Ann. clin. Res. 4: 250(1972). Renneker, R. and Cutler, M.: Psychological problems of adjustment to cancer of the breast. J. amer. med. Ass. 148: 833 (1952). Rimon, R.: A psychosomatic approach to rheumatoid arthritis. Acta rheum, scand., suppl. 13, p. 1 (1969). Rimon, R. ja Rakkolainen, V.: Koronaaritaudin ja sydaninfarktin psykosomaattisia nakokohtia. Suom. LaakLehti 23: 2446 (1969). Stein, M. and Shiavi, R.: Respiratory disorders; in Friedman and Kaplan Comprehensive textbook of psychiatry, p. 1068 (Baltimore 1967).
Aggression in General-Hospital Patients Ranan Rimón, Juhani Jussila and Olavi Katila
Abstract. The Buss-Durkee Inventory (BDI) rating for measuring aggression was per formed on 125 consecutive female patients admitted to a general hospital. The test scores of different diagnostic groups were all on a intermediate level, and with the exception of slightly lower scores in patients with blood disorders, only few differences reaching statisti cal significance were observed. The factor analysis of the test results revealed three factors: in addition to aggression factor (I) and hostility factor (II) a factor called remorse factor (III) was identified. The patient groups divided according to their medical diagnoses showed no differences with regard to these factors. The variance of factor scores could not be explained by such background variables as social class, marital status or domicile. There was, however, a highly significant negative correlation ( 0.33) between age and the aggression factor. Finally, the concept of suppressed aggression and hostility and the role of aggression dynamics in the development of organic illness is discussed.
Introduction
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Many psychosomatically oriented ecological investigations have shown that episodes of different organic illnesses are not distributed at random among the general population. Life experiences, psychosocial factors and personality characteristics have been shown to influence general susceptibility to illness (Hinkle and Wolff, 1958; Hinkle, 1961 \ Rahe et al., 1964; Rahe, 1972). The psychodynamic basis of the personality traits involved, especially concerning psychosomatic disorders, has been interpreted in different ways according to the psychiatric discipline to which the investigators belonged to, e.g. frustration in infancy, masochistic dependence upon the mother, distortion of psychosexual identity, and rejection of either masculine or feminine roles are concepts which the pertinent analytically oriented literature almost axiomatically uses. Again and again, the significance of repressed or inhibited aggression in some disease entities is emphasized (Bastiaans, 1969). Several authors maintain that patients with hypertension (Kaplan et a l, 1961), peptic ulcer (Kezur et al, 1951), ul cerative colitis (Engel, 1955), irritable colon syndrome (Engel, 1967), bronchial asthma (Stein and Sltiavi, 1967), rheumatoid arthritis (Johnson et al., 1947;
Rimón/Jussila/Katila
13
Cobb, 1959), and even patients with malignant diseases (Renneker and Cutler, 1952; Blumberg et ai, 1954) are unable to deal adequately with their aggressive impulses and hostile attitudes, and are inclined to inhibit the expression of their aggression. The prolonged buildup of undischarged hostile and aggressive ten dencies is believed to provoke an associated chronic physiological reaction which, in a certain life situation, may result in the development of a clinical disorder of a predisposed organ or organ system (Alexander, 1950). The concept of ‘aggression’ is by no means unambiguous but in any case it includes several, different components (Kaufman, 1965; Groen, 1972). Buss (1961) divides aggression into the outwardly directed instrumental response of aggression, the emotional reaction of anger, and the attitude of hostility. In addition, aggression can be divided into physical versus verbal, active versus passive, direct versus indirect, spontaneous versus provoked, and offensive versus defensive (Pitkánen, 1969). In many psychosomatic studies, aggression has been understood as a completely unambiguous idea, and it has been measured mainly by projective methods (e.g., Cleveland and Fisher, 1960; Geist, 1966). The aim of this investigation was to elucidate by means of an inventory technique to what extent the aggression dynamics of unselected general-hospital patients, including many with psychosomatic diseases, would exhibit significant psychological determinants relevant to the illness of the patients. Another aim of the study was to analyze the structure of the Finnish version of the Buss-Durkee Inventory (BDI) used (Buss and Durkee, 1957).
Subjects and Methods
/
Cardiovascular disorders coronary heart disease, congestive heart failure, peripheral vascular insufficiency
21
2
Endocrine diseases diabetes, hyperthyroidism
19
3
Connective tissue disturbances rheumatoid arthritis, collagénoses, disc degenerations
13
4
Neurological disorders organic brain syndromes, convulsive disorder, multiple sclerosis, parkinsonism
12
5
Gastro-intestinal diseases peptic ulcer, pancreatitis, liver and biliary diseases
17 Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Subjects The series consisted of 125 patients who were admitted to the Department of Medicine of the University Central Hospital in Turku. The mean age of the patients was 44.2 (SD = 15.6) years. The patients were classified according to their diagnoses as follows:
Rimón/Jussila/Katila
14
6
Blood and lymphatic disorders splenomegaly, leukemias, Hodgkin’s disease
12
7
Infections acute meningitis, tonsillitis, etc.
10
8
Respiratory diseases bronchial asthma, emphysema, bronchiectasis
12
9
Urinary tract disorders haematuria, nephrosis Total
9
125
The groups were comparable as to the age and social and marital status.
Treatment o f the Data and Results Means, standard deviations and internal consistency of reliability estimates (KR-20) of the BD1 scales are presented in table I. These scores are similar to the scores of the hospital patient sample for anonymous inventories reported by Buss (1961). The slight variations may be attributed to the only moderate reliability level of the scales. The correlations between items and the total score of the respective scales varied between + 0.27 and + 0.67. In table II the BDI aggression scores of the various diagnostic subgroups are presented. Generally, the total scores are all on intermediate level and comparable with each other. The total scores of groups 6 (blood disorders) and 9 (urinary tract diseases) are, however, slightly lower than those of the other groups, but the differences do not reach statistical significance (F-test and t-test). In group 6 the scores of all item categories except that of ‘irritability’ are below the mean, and when compared to corresponding scores of several other groups (e.g., groups 2 -4 , 7, 8), statistically significant differences (p < 0.05) can be observed as to the scales negativism’, ‘resentment’, ‘suspicion’ and ‘guilt’. In group 9, however, the only sta tistically significant differences (p < 0.05) are noted when the score of ‘negativism’ is compared to the corresponding test figure of group 1 (cardiovascular disorders), and when the score of ‘resentment’ is compared to that of group 4 (neurological disorders). In both instances this is, however, largely due to the fact that the scores of ‘negativism’ in group 1
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Procedure The BD1 for measuring aggression was used. It has been proved useful in numerous experiments with students, patients of general hospitals and out-patients of several clinics, and psychiatric patients (Buss, 1961; Buss et al, 1962: Rimon, 1969; Friedman, 1970; Persky et al., 1971; Olkinuora, 1972). This test attempts to measure several components of aggressiveness, consisting of eight scales, which together tap a variety of aggressive be haviours and hostile attitudes. Five scales, i.e. ‘assault’, ‘indirect’, ‘irritability’, ‘negativism’ and ‘verbal’, are presumed to represent the tendencies to the instrumental response of aggression and to anger. Two other scales, i.e. ‘resentment’ and ‘suspicion’, are intended to measure various hostile attitudes, and the final scale ‘guilt’ gives information on guilt feelings second ary to conscious or inhibited unconscious aggressive impulses. All BDIs were filled out anonymously. However, a hidden coding system revealed the connection between each patient and the corresponding inventory.
15
Aggression in General-Hospital Patients
Table I. Means, standard deviations and KR-20 estimates of reliability of the BD1 scales (N= 125) Scale
Number of items
Mean
SD
KR-20
1 Assault 2 Indirect 3 Irritability 4 Negativism 5 Resentment 6 Suspicion 7 Verbal 8 Guilt
10 9 11 5 8 10 13 9
3.5 4.4 6.2 2.4 25 4.4 5.4 5.3
2.0 2.0 2.3 1.4 1.8 2.4 2.8 2.3
0.54 0.60 0.60 0.52 0.55 0.66 0.65 0.69
and that of ‘resentment’ in group 4 are exceptionally high and well above the average. The test score of the item category ‘negativism’ in group 1 is also significantly (p < 0.01) higher that that of group 2 (endocrine diseases), and also higher than the corresponding figures of groups 5 (gastro-intestinal diseases) and 6 (p < 0.05). Similarly, the high score of resent ment’ in group 4 is significantly higher than that of group 6 (p < 0.05). The score of the scale ‘verbal’ in group 1 is, however, below the mean and significantly lower than that of group 5 (p < 0.05) for instance. On theoretical arguments and on the basis of earlier empirical analyses, the extraction of two factors was to be expected (Buss and Durkee, 1957; Persky et al., 1971). Because underfactoring can distort the factor structure more than overfactoring, we extracted 4 factors. These explained 48.1 % of the total variance. An orthogonal Varimax rotation was made using 2, 3 and 4 factors. Only factor loadings of 0.40 or above were included in the interpretation. The rotation solution of 2 and 3 factors were both interpretable. The rotation with 4 factors produced a specific factor IV; further treatment of this solution was omitted. The factor structure in the two-factor solution corresponded to the oblique rotation solution presented by Buss and Durkee (1957). The first factor could be identified as ‘aggression factor’ and the second as a ‘hostility factor’. ‘Guilt’ had no substantial loading on either factor. The rotation with three factors (table 111) indicated that ‘guilt’ forms a factor III together with ‘assault’ and ‘negativism’. This factor was named ‘remorse’ factor. The above factors were treated as new variables. Factor scores were estimated with the regression method. The multiple correlations of the scales with the factors were 0.79 (I), 0.75 (II), and 0.67 (III) indicating that the estimation was adequate. One-way analyses of variance and paired t-tests revealed no significant differences between patient groups with regard to these factors. The variance of factor scores could not either be explained by such background variables as social class, marital status or domicile (urban/rural). There was, how ever, a highly significant negative correlation (-0.33) between age and the aggression factor.
The most striking finding of this study was no doubt the fact that very similar patterns of aggression dynamics were found in patients with various
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Discussion
16
Rimon/Jussila/Katila Table II. The mean scores obtained from the BDI in various diagnostic subgroups Scale
SD
mean
SD
mean
SD
mean
SD
3.2 4.2 6.0 3.2 2.6 4.4 4.4 5.5
1.4 2.3 2.1 1.2 2.0 2.9 2.5 2.6
3.6 4.7 6.2 2.2 2.3 4.8 5.6 5.1
2.2 1.9 2.3 1.1 1.7 2.2 2.9 2.3
3.5 3.6 5.8 2.5 2.5 4.1 5.9 5.9
2.0 1.9 2.5 1.4 1.5 2.1 2.8 2.0
3.6 4.2 6.5 2.7 3.4 4.2 6.3 6.0
1.4 1.6 2.8 1.5 1.7 2.1 2.9 2.6
33.5
12.1
34.5
11.8
33.8
12.0
36.9
10.8
II
e
mean
II
co
2, n= 19
c
Total
12
1, n = 21
CO
1 Assault 2 Indirect 3 Irritability 4 Negativism 5 Resentment 6 Suspicion 7 Verbal 8 Guilt
Diagnostic subgroups
Scale
I
11
III
h2
1 Assault 2 Indirect 3 Irritability 4 Negativism 5 Resentment 6 Suspicion 7 Verbal 8 Guilt
49 50 61 14 35 15 69 25
18 25 25 16 63 66 14 28
47 26 25 48 22 25 14 50
50 38 50 28 57 52 51 39
Eigenvalue Percentage
1.5 19.6
1.1 13.9
1.0 12.0
3.6 45.5
medical diseases. The low scores of several item categories, especially in those reflecting expression of resentment, hostility and guilt, in the patients with blood disorders is probably due to the fact that more than half of the cases (7/12) in this group had acute leukemia or Hodgkin’s disease. As Baltrusch (1969) and Achte et at. (1970) have emphasized, patients with malignancies frequently struggle with restraints on aggressiveness, suppressed guilt feelings and introverted hostility. The similarly low scores of patients with urinary tract diseases is, on the contrary, more surprising, and may depend on irrational but
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Table III. Varimax-rotated factor matrix (three factors)
Aggression in General-Hospital Patients
17
Total, n = 125
mean
SD
mean
8, n = 12
9, n = 9
SD
mean
SD
mean
SD
mean
SD
3.5 4.4 6.2 2.4 2.5 4.4 5.4 5.3
2.0 2.0 2.3 1.4 1.8 2.4 2.8 2.3
10
4.2 4.5 6.6 2.1 2.6 4.2 6.4 4.8
1.9 2.2 2.6 1.6 1.7 1.6 3.2 2.4
2.9 4.3 6.4 2.1 1.9 3.1 4.6 4.1
2.8 2.5 2.6 1.5 1.5 2.0 2.4 2.3
3.5 5.0 6.2 2.5 2.3 5.1 4.8 5.8
1.8 1.7 2.2 1.6 1.6 2.1 2.6 2.8
3.1 4.6 5.8 2.3 3.1 5.3 5.2 5.8
2.3 2.0 2.1 1.4 2.2 3.4 2.6 2.3
3.1 4.0 5.6 2.0 1.4 3.7 5.0 5.0
2.1 1.8 1.4 1.4 1.9 1.9 2.6 1.4
35.4
9.0
29.4
12.4
35.2
10.6
35.2
12.1
29.8
11.1
34.0 ± 1 1.2
dynamically significant concepts in relation to urinary and genital function and denial of guilt associated with the development of the illness. Another surprising finding was that patients with cardiovascular disorders (including 10 patients with coronary heart disease: CHD) had a high score on ‘negativism’ and a low score on the scale ‘verbal’. This is ostensibly at variance with many previous findings, mainly focussing on male patients, revealing an overtly aggressive personality profile in patients with CHD (Friedman and Rosenman, 1960; Riman and Rakkólainen, 1969). The relatively great number of patients (11) with cardiovascular disorders other than CHD in this group, and the weak reliability of the scale ‘negativism’ may, however, account for this unexpected observation. The high ‘resentment’ score in the group with neurological disorders, on the other hand, is quite easy to understand. This group consisted of 3 cases with convulsive disorder, 2 with multiple sclerosis and 2 with parkinsonism who had all encountered severe social and psychological problems in their work and con sequently had become embittered and resentful. The high total score in the group of respiratory diseases, including 9 cases of 12 with bronchial asthma, is at variance with the relatively common concept of inhibition and deficient integra tion of aggression dynamics in patients with bronchial asthma (Pierluot and van Roy, 1969). When the significance of the BDI score differences in various patient groups is discussed, it has to be emphasized that statistically significant differences may appear by chance in approximately 5 instances out of 100, when the significance level of 5 % is applied. In the present study, several paired t-tests were per-
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
C"
mean SD
II
6, n = 12
c
5, n = 17
Rimon/Jussila/Katila
18
formed, the groups were small, the reliability of the scales was not great, and the differences were after all relatively few. When the test results of the patients were investigated by factor analysis, i.e. on a more reliable level, no significant differences with regard to the medical diagnoses were observed. This indicates that the significant differences which occurred on scale level probably have arisen by chance. The factor analysis of the BDI revealed that, in addition to already de scribed aggression (I) and hostility (II) factors, ‘negativism’ can — even though its communality is relatively weak — form a factor III together with ‘guilt’ and ‘assault’. This factor III was named remorse factor. The interpretation of this factor is clear: when oppositional behaviour leads to physical violence against others, it happens in connection with guilt feelings. Aggression and hostility factors are quite distinct. The factorial validity among them highly points to the theoretical starting-point. Due to social and moral norms, free expression of aggression is not usually accepted. Subsequently, the repression, displacement and/or sublimation of it are important for an individual’s social adaptation. Because of the complex nature of human behaviour, the connection between provoking stimuli and re sponse is not directly predictable. Man’s cognitive qualifications for appraising a situation make his behaviour less dependent on psychological drives than is possible for the lower species. Man is able to inhibit or attenuate aggression to situational requirements. Thus various aggressive tendencies may become in wardly directed and turn against the individual himself. The question whether this, in certain subjects, may result in the development of physical illness or particularly in psychosomatic symptom formation still remains unanswered. The results of the present study do not point to a positive support of the hypothesis that suppressed aggression and a chronic buildup of undischarged hostile and aggressive tendencies would be significant determinants in the development of organic illness.
Achte, K.A.; Vauhkonen, M.L., and Viitamdki, R.O.: Cancer and psyche. Monogr. No. 1. Psychiat. Clin. Helsinki Univ. Centr. Hosp. (1970). Alexander, F.: Psychosomatic medicine (New York 1950). Baltrusch, H.F.: Psychosomatische Beziehungen bei Krebskrankhciten. Z. psychosom. Med. 3: 196 (1969). Bastiaans, J.: The role of aggression in the genesis of psychosomatic disease. J. psychosom. Res. 13: 307 (1969). Blumberg, E.M.: West, P.M.. and Ellis, F.W.: A possible relationship between psychological factors and human cancer. Psychosom. Med. 16: 277 (1954). Buss. A.H.: The psychology of aggression (New York 1961). Bliss, A.H. and Durkee, A.: An inventory for assessing different kinds of hostility. J. cons. Psychol. 21: 343 (1957).
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
References
Aggression in General-Hospital Patients
19
Ranan Rimon, Karpantie 1 B, 70400 Kuopio 40 (Finland)
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/7/2018 4:48:44 PM
Buss, A.H.; Fischer, H., and Simmons, A.J.: Aggression and hostility in psychiatric patients. J. cons. Psychol. 26: 84 (1962). Cleveland, S.E. and Fischer, S.: A comparison of psychological characteristics and physio logical reactivity in ulcer and rheumatoid arthritis groups. Psychosom. Med. 22: 283 (1960). Cobb, S.: Contained hostility in rheumatoid arthritis. Arthr. and Rheum. 2: 419 (1959). Engel, G.L.: Studies on ulcerative colitis. III. The nature of the psychological processes. Amer. J. Med. 19: 231 (1955). Engel, G.L.: Irritable cohen syndrome; in Friedman and Kaplan Comprehensive textbook of psychiatry, p. 1058 (Baltimore 1967). Friedman, A.S.: Hostility factors and clinical improvement in depressed patients. Arch. gen. Psychiat. 23: 524 (1970). Friedman, M. and Rosenman, R.H.: Overt behaviour pattern in coronary disease. J. amer. med. Ass. 173: 1320 (1960). Geist, H.: The psychological aspects of rheumatoid arthritis (Springfield 1966). Groen, J.J.: The study of human aggression. Psychother. Psychosom. 20: 312 (1972). Hinkle, L.E.: Ecological observations of the relation of physical illness, mental illness and the social environment. Psychosom. Med. 23: 289 (1961). Hinkle, L.E. and Wolff, H.G.: Ecologic investigations of the relationship between illness, life experiences, and the social environment. Ann. intern. Med. 49: 1373 (1958). Johnson, A.: Shapiro, L.B., and Alexander, F.: Preliminary report on a psychosomatic study of rheumatoid arthritis. Psychosom. Med. 9: 295 (1947). Kaplan, S.M.; Gottschalk, L.A.: Magliocco, E.B.; Rohovit, D.D., and Ross, W.D.: Hostility in verbal productions and hypnotic dreams of hypertensive patients. Psychosom. Med. 23: 312 (1961). Kezur, E.; Kapp, F., and Rosenbaum, 11.: Psychological factors in women with peptic ulcer. Amer. J. Psychiat. 108: 368 (1951). Kaufman, H.: Definitions and methodology in the study of aggression. Psychol. Bull. 64: 351 (1965). Olkinuora, M.: A factor analytical study of psychosocial background in bruxism. Proc. finn. dent. Soc.68: 184 (1972). Persky, H.; Smith, K.D., and Basu, G.K.: Relation of psychologic measures of aggression and hostility to testosterone production in man. Psychosom. Med. 33: 265 (1971). Pierloot, R.A. and Roy, J. van: Asthma and aggression. J. psychosom. Res. 13: 333 (1969). Pitkdnen, A descriptive model of aggression and nonaggression with applications to children’s behaviour; Diss. Jyvaskyla (1969). Rahe, R.H.; Meyer, M.; Smith, M.; Kjaer, G., and Holmes, T.H.: Social stress and illness onset. J. psychosom. Res. 8: 35 (1964). Rahe, R.H.: Subjects’ recent life changes and their nearfuture illness reports. Ann. clin. Res. 4: 250(1972). Renneker, R. and Cutler, M.: Psychological problems of adjustment to cancer of the breast. J. amer. med. Ass. 148: 833 (1952). Rimon, R.: A psychosomatic approach to rheumatoid arthritis. Acta rheum, scand., suppl. 13, p. 1 (1969). Rimon, R. ja Rakkolainen, V.: Koronaaritaudin ja sydaninfarktin psykosomaattisia nakokohtia. Suom. LaakLehti 23: 2446 (1969). Stein, M. and Shiavi, R.: Respiratory disorders; in Friedman and Kaplan Comprehensive textbook of psychiatry, p. 1068 (Baltimore 1967).
