L. Bellodi,О. Gambini, and V. Brancato
Institute of Clinical Psychiatry, Milan Medical School, Ospedale S. Paolo, 20142 Milan, Italy
In spite of the many studies done in the two last decades, there is still confusion and c ontroversy about the relationship between anxiety and depression. Some people considered these quite distinct emotional reactions, anxiety being related to fear of danger, and depression to experience of loss, to be opposed. The symptoms of anxiety cover a broad range of psychopathology, whereas depression is a characteristic of a definite set of affective disorders. In theory, therefore, it would seem to make little sense for an individual to display such opposite emotional responses at the same time, at a single moment. It appears much more reasonable to consider affective disorders and anxiety in a psychiatric patient as separate entities. In reality, the situation is not as stated. There are many studies in the literature on this subject, based on different methods, that nourish the debate between those who think there is a continuum, a spectrum, of anxiety to depression (6) and those who consider them to be two distinct conditions (8,9). At the beginning, we studied the symptoms of depression and anxiety, strictly psychic and somatic components of the latter, following the short-term patterns of their development in a group of 15 patients diagnosed according to DSM-III (1) as having Generalized Anxiety Disorders (GAD) or Somatization Disorder (SD) This disorder is a particular expression of the anxiety state, comparable to GAD (4). The most widely accepted causal factor for SD is psychodynamic. Patients with certain predispositions respond to particular types of stress with conversion symptoms. The stress induces anxiety by awakening intrapsychic conflicts and the symptoms "bind" anxiety. We included some of those patients in our study, patients in whom the prevalent anxiety component is somatic, so that we could study in both types the disturbance of mood depression that is so frequently present, and even more so, to study the reciprocal modalities of development. For this latter approach, we took changes in mood into account as manifestations of depression, that is, a symptom, since the initial diagnostic procedure excluded these subjects having any Affective Disorders. The patients
were controls for a double-blind study of the therapeutic effectiveness of a short half-life benzodiazepine (clotiazepam). Their symptoms, after a wash-out period of any medication of at least two weeks, were rated on the Hamilton Anxiety Rating Scale (HARS) (3) at the start of the study, at the start of the fourth week, and at the start of the sixth week on placebo. Mean scores for GAD and SD were, respectively, at the start 26.5 + 6.0 and 35.5 + 2.1, at the fourth week 13.9 + 6.1 and 23.0 + 7.0, and at the sixth week 18.1 + 8.8 and 10.5 + 4.9. Since they were given placebo, the changes we saw were not just variations in time but spontaneous variations in time plus variations due to a placebo effect. However, our purpose was to see the changes in various symptoms at the same time, under equal trial conditions, and to measure the covariance with the time (10). The data were analyzed by univariate and multivariate analyses of variance (MANOVA) (Table 1). The between-design component was diagnosis. The within-design components were time, item, and time-item. We first analyzed the HARS scores for psychic anxiety (item 1) and the scores for somatic anxiety (items 7 to 12). Figure 1 shows the mean scores for both symptoms, that is, for psychic anxiety and for somatic anxiety, at the three different times and for the patients as a whole. Our results indicate that there is variability over a relatively short period of time, 6 weeks, in the anxiety state. The fluctuations in severity of the most strictly psychic components and the somatic components appear to follow the same pattern. On the other hand, when the patients are divided according to diagnosis, the fluctuations of severity of somatic anxiety had a different pattern, diluted over the time period of observation. This suggests that the somatic symptoms of patients with SD are only in part the somatic equivalents of anxiety, which is not the only factor in their psychogenesis: Some patients show autonomic symptoms and pain as a clinical picture of "masked depression" more than as an expression of somatic equivalents of anxiety (5). Next, we evaluated the changes in both the components of anxiety with respect
to changes in mood (item 6). Figure 2 shows the fluctuation for psychic anxiety and depression: In this case the only significant effects were time and item. We have found (Table 2) a significant variability in the psychopathological measurements during the period of observation as well as different degrees of severity for psychic anxiety and depression (greater for the former than for the latter). However, the analysis did not show significant differences in the patterns of changes in this pair of symptoms. Figure 3 shows the plot for the couple somatic anxiety and depression. In this case we also found (Table3) that time was a significant variable, where as item
was not, indicating that the initial mean values of somatic anxiety and depression did not differ in severity, and their variability with time was similar, but that when the patients were divided taking into account the diagnostic group, their variability had different patterns. Depressed mood, as an expression of one particular aspect of the affective sphere, is generally found in individual conditions of the anxious state, that is, in the context of a subjective view of being threatened, of immanent danger. As far as we have been able to determine, this is an affective condition whose severity varies with the fluctuating subjective feeling of anxiety, whether expressed in the specifically psychic components or in part in equivalent somatic components. Regarded in this way we might say that the depressed mood is an individual measurement of the recognition by the subject of how much this anxiety causes him or her to lose as an individual, in his or her interactions with the outside world and in his or her ability to function adequately. However much these findings may tend to emphasize the complexity of the psychogenetic relationships between anxiety and depression from the point of view of the dynamics of psychic processes, they contribute very little to a clear understanding of the reciprocal relationships between emotional and affective changes when present in a single individual in the form of a syndrome suggesting that one or more disorders alternate in time or at the same time. A reason for this persisting confusion may be suggested by the differences in methodology. Our study, like many others in the literature that have included factorial analysis of the anxiety-depression syndromes (7), was based on quantitative mea- surements of the severity of symptoms by rating
scales. There is an underlying assumption in this approach that should not be tacitly accepted, but should be considered in setting up the strategies of the studies. It is assumed that changes in intensity always have the same psychopathological weight. We know, however, that the innumerable factors that affect the existence, the severity, the duration, and the possibility of clinical detection, depending on the type of pathology and conditions of the study, often do not vary continuously. Therefore, small shifts on the rating scale do not have equal weight, but become important only when they produce values other than some critical preset value that defines a psycho-pathological threshold. For this reason, it is useful to pursue semiqualitative strategies. The most concrete expression of such a rationale is DSM-III, which does not formulate diagnoses simply from measuring symptoms and behaviors, as do the rating scales, but includes qualitative criteria such as presence or absence and applies quantitation only to the number of criteria satisfied. With regard to the problem of the mixed anxiety and depression syndrome, DSM-III provides rules for decision based on hierarchical models, with criteria that are not only symptomatological but also temporal for their duration to lead to a diagnosis with a reasonable degree of reliability, in which the diagnosis of the pathological condition of each patient is reached by exclusion of other diagnostic categories. Nevertheless, from a strictly psychopathological point of view and in terms of the diagnostic validation of the nosographic entities so far identified, there are still elements of doubt about whether this model is the best for the understanding and management of these mixed conditions, so often encountered in the clinician’s practice (2). In order to clarify the problem of comorbidity of anxious states and depressive syndromes, we attempted to verify in a series of patients with Generalized Anxiety Disorder and Major Affective Disorder which theoretical model was best for interpreting this phenomenon. Therefore, we studied 54 outpatients with Generalized Anxiety Disorder (n = 35) and with Major Depressive Disorder (n = 19), diagnosed according to DSM-III criteria, for at least 2 years. Irrespective of diagnosis, at the start of our study the symptoms that define criterion A for Generalized Anxiety Disorder (anxiety syndrome) and those that define criterion
they produce values other than some critical preset value that defines a psycho-pathological threshold. For this reason, it is useful to pursue semiqualitative strategies. The most concrete expression of such a rationale is DSM-III, which does not formulate diagnoses simply from measuring symptoms and behaviors, as do the rating scales, but includes qualitative criteria such as presence or absence and applies quantitation only to the number of criteria satisfied. With regard to the problem of the mixed anxiety and depression syndrome, DSM-III provides rules for decision based on hierarchical models, with criteria that are not only symptomatological but also temporal for their duration to lead to a diagnosis with a reasonable degree of reliability, in which the diagnosis of the pathological condition of each patient is reached by exclusion of other diagnostic categories. Nevertheless, from a strictly psychopathological point of view and in terms of the diagnostic validation of the nosographic entities so far identified, there are still elements of doubt about whether this model is the best for the understanding and management of these mixed conditions, so often encountered in the clinician’s practice (2). In order to clarify the problem of comorbidity of anxious states and depressive syndromes, we attempted to verify in a series of patients with Generalized Anxiety Disorder and Major Affective Disorder which theoretical model was best for interpreting this phenomenon. Therefore, we studied 54 outpatients with Generalized Anxiety Disorder (« = 35) and with Major Depressive Disorder (n = 19), diagnosed according to DSM-III criteria, for at least 2 years. Irrespective of diagnosis, at the start of our study the symptoms that define criterion A for Generalized Anxiety Disorder (anxiety syndrome) and those that define criterion
Affective Disorder can be found among the group with Anxiety Disorder, and vice versa. In other words, 2.9% of the anxious patients and 5.3% of the depressive patients were misclassified initially. These results lead to some conclusions. First, there are fluctuations in the syndrome even in patients whose diagnoses had been made according to rigid criteria as either Affective or Anxiety Disorders. In most cases, however, these fluctuations are not large enough to alter assignment to the group of initial diagnosis when the variables useful for individuation of a discriminant function are examined. Second, there are, nonetheless, a small number of subjects in whom the shifts of psychopathological aspects, either from anxiety to depression or from depression to anxiety, contribute significantly to a change in diagnosis. This indicates that there is an underlying predisposition that can be expressed at different times on one or the other side of the clinical picture. Such patients were a small minority in our patient population, probably because we started out by seeking those with maximal dichotomy between anxiety and depression, but in a less highly selected population there might well be more of them. In conclusion, we think that the problem still remains unsettled and that no pure psychopathological approach can give a definitive answer, even if this approach were extended to include such areas of interest as clarification of possible correlations between concomitant personality disorders, thus providing better understanding and definitions of the psychodynamic aspects underlying certain syndromes, or even more precise diagnostic definitions of longitudinal changes in these syndromes. On the contrary, we believe that a psychopathological approach could provide the basis for validation studies the results of which could complete psychopathological and descriptive analysis. Those new acquisitions should be later verified in a dynamic process to reach a better understanding of the phenomena observed.
REFERENCES
1. American Psychiatric Association. (1980): Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. American Psychiatric Association, Washington, D.C. 2. Goldberg, D., and Simon, N. (1986): In: Drug Treatment of Neurotic Disorders: Focus on Alprazolam, edited by M. H. Lader and H. С Davies, pp. 76-82. Churchill-Livingstone, New York. 3 Hamilton, M. (1959): Br. J. Med. Psychol, 32:50. 4 Hyler, S. E., and Sussman, N. (1984): Somatoform disorders: Before and after DSM III. Hosp. Community Psychiatry, 35(5):469-478. 5 Ibor, J. J. I. (1972): Br. J. Psychiatry, 120:245-258. g’ Johnston, E., Owens, C, Firth, C. D., McPherson, H., Davie, C, Riley, G., and Gold, A. (1980): ‘ Psychol. Med, 10:321-328. 7 Mullaney, J. A. (1984): J. Affective Disord, 7:139-148. % Prusoff, B„ and Klerman, G. L. (1974): Arch. Gen. Psychiatry, 30:302-309.
9 Roth, M., Gurney, G., Garside, R. F., and Kerr, T. A. (1972): Br. J. Psychiatry, 121:147-161. 10 Russel, G. F. M., and De Silva, P. (1983): Br. J. Clin. Pharmacol., 15:1475-1535.
Anxious Depression: Assessment and Treatment, edited by G. Racagni and E. Smeraldi. Raven Press, New York © 1987.
The Prevalence and Outcome of Anxious