As early as 1913, Kraepelin had already included mild forms, fluently transitive to norm, in his rubric ‘manic-depressive insanity’ and in 1921, Kretschmer coined the term ‘cycloid personality’. In substantial studies, Hoffmann (1921) searched for mild depressive disorders and personality variations (pertaining to the manic-depressive context) among the relatives of manic-depressed patients. Our own studies of unipolar patients’ familial characteristics indicate that mild ‘reactive’ or ‘neurotic’ atypical depression prevails in the same kins in which a high occurrence of affective psychoses is also found (Angst et al. 1980). No demarcation of depressive or cycloid personality traits from mild or early manifestations of depressive or manic-depressive disorders has yet been reached. Kraepelin explicitly raised the hypothesis that this so-called ‘premorbid personality’ may be an initial manifestation of illness, an opinion which has been raised again by recent investigations (von Zerssen 1977; Hirschfeld and Klerman 1979; Akiskal et al. 1979; Akiskal 1983).
‘Residual’ depressive disorders which do not satisfy the diagnostic criteria are yielded as a by-product as more severe depressive disorders are operationalized. They may, for instance, be too short, too poor in symptoms, or without social consequences. On the other hand, quite an important part of these mild depressive syndromes may appear to be socially relevant, because they keep recurring. They have therefore been labeled as ‘chronic minor’, ‘intermittent’ (Spit- zer et al. 1978), ‘neurotic’ , ‘characterologic’ (Akiskal et al. 1978; Klerman et al. 1979) and more recently as ‘dysthymic’ or ‘atypical’ disorders (DSM-III 1980). Akiskal and associates (1978) are correct in stressing that the ‘dysthymic disorders’ are a heterogeneous group of depressive states. In his latest paper (1983), Akiskal suggests dividing such chronic depressions into three groups: (1) primary atypical depressions with residual chronicity, (2) chronic secondary dysphorias following nonaffective disorders, and (3) characterologic atypical depressions. The last mentioned group represents a potential mild type of unipolar or bipolar illness. Akiskal emphasizes that classic depressive episodes can conjoin, and in this context speaks of a ‘chronic subsyndromal’ and ‘chronic intermittent’ course. The subsequent occurrence of minor and major atypical depression has recently been called ‘double atypical depression’ (Keller et al. 1983).
The pertinent methodology has been described in a previous paper (Angst et al. 1984d; additional information in Angst and Dobler-Mikola 1984b and c). To briefly summarize: the results were obtained from a longitudinal study of 591 men and women who had been investigated initially at the age of 19-20, and for a fourth time at the age of 23-24. The data given in the present publication refer to this fourth investigation, a protracted semistructured interview (SPIKE) and selfreporting by means of the SCL-90 (Derogatis 1977).
Depressive mood was assessed with the following seven symptoms: joyless, depressive, sad, sick of life, loss of interest, loss of efficiency, feelings of inferiority.
To define a minor or major depressive ‘case’ we used different thresholds of definition (Table 1): the mildest considering social impairment (SOC), a more severe one (WORK) taking into account subjective and/or objective work impairment, and the most rigorous one considering simultaneous presence of a predetermined minimum number of criteria symptoms (SYM). The same criteria are applied for classifying longer lasting ‘extensive depressive episodes’ (EDE), too (Angst and Dobler-Mikola 1984c).
‘Social impairment’ stands for role dysfunction (work, interpersonal relations, pastime activities), ‘work impairment’ implies subjectively perceived loss of efficiency, conflicts resulting therefrom, absence from work, loss of job. As criteria symptoms for defining atypical depression, those 8 listed in the DSM-III and the RDC served; however, in contrast to North American practice, we used 3 of the 8 symptoms as a threshold of case-definition for males, and 5 out of 8 symptoms for females.
Independent of the assessment of psychopathology, a 1-h sociological interview (Angst et al. 1984d) was used to elicit detailed information on sociodemographic characteristics, social adjustment, coping resources (mastery and self-esteem) and distressing life events.
Recurrent and Nonrecurrent Brief Depressive Episodes
We term depressive manifestations lasting less than 2 weeks ‘brief depressive episodes’ (BDE). They reoccur frequently, and we have therefore tried to compute their cumulative duration over 1 year (Angst and Dobler-Mikola 1984b). This approach turned out to be somewhat problematic, because the computations are complicated and the indications provided by the probands about the duration of recurrent episodes are not very reliable. In addition, the results of our analysis of familial characteristics lead to the presumption that the cumulative duration of depressive manifestations over 1 year is not of
Table 1 gives the 1-year prevalence rates of the two BDE subgroups versus those of the EDE (minimum duration 2 weeks). Of the 22 to 23-year-old population 31.7% indicate an occurrence of BDE in the course of 1 year. The majority—60%— belong to the NRBD, whose prevalence of 18.8% is quite high but who are not, in aggregate, socially impaired by their depressive mood. Accordingly, the prevalence rates for NRBD with social impairment (3.4%), with work impairment (2.5%), and with pertinent symptoms (1.1%) are quite low.
On the other hand, the recurrent subtype, representing 40% of the BDE, is more frequently seen to have social consequences. The 1-year prevalence is 11% for social impairment, 7.6% for work impairment, and 4.4% for the presence of diagnostically required symptomatology.
The male/female ratio is not as marked in the BDE prevalence rates (Table 2) as might have been expected from studies of major atypical depression. The nonrecurrent manifestations are however found to be more common in men and the recurrent ones in women. RBD are seen in 7.4% of the men and in 9.8% of the women, representing a sex ratio of 1:1.3, which is clearly below the usual one of 1:2 in major atypical depression. An analogous sex ratio is also found in the RBD (SYM) group. This is however due to the application of different thresholds for the minimum number of symptoms required in men and women respectively; otherwise, the ratio would shift to the usual 1:2.
Earlier evidence of men’s underreporting of depressive mood, and especially of specific symptoms (Angst and Dobler-Mikola 1984a) was based on the following:
— men state significantly fewer depressive episodes that date back more than 3 months;
— on average, men report two symptoms less than women do, despite equal social impairment.
We consider the data here presented with the hypothesis of an equal frequency of BDE and the two BDE subgroups of recurrent and nonrecurrent manifestations in men and women as most probable. The equally high prevalence rate of the BDE, 31.2% for men and 32.1% for women, which largely supports our hypothesis is very impressive.
The RBD (SYM) groups differs from the EDE (SYM) only in the shorter duration of depressive episodes. The 1-year prevalence happens to be the same (4.4%) for both groups, and very similar results are obtained on the threshold of work and social impairment. We can deduce therefore that among the corresponding age group of the general population, recurrent short (less than 2 weeks) lasting disturbances are as frequent as longer lasting depressive episodes. To be more informative, and to facilitate comparisons with other studies, Table 2 also contains the 3-month and 4-week prevalence rates.
Clinical Characteristics of the RBD
Following the operational definition of RBD with a 1-year prevalence rate of 4.4%, the question of course ensues as to what their clinical relevance and further characteristics may be. What exactly are they, minor atypical depression—somewhere on a continuum between physiological disturbances and major atypical depression? Personality disorders? Or are they mainly precipitated by external factors, or caused by distressing life events? A further paper will examine the question of a possible heterogeneity of the entire RBD group. Presently, we wish to focus on the more detailed specification of the RBD, based on the clinical and the extensive sociological interview.
Age of Onset, Positive Family History, and Bipolarity
The age of onset is a very useful criterion to establish whether the RBD, as a subtype of minor atypical depression, are on a continuum to major atypical depression. If the age of onset is the same at both ends this will support the theory of a continuum. If on the other hand it differs, this will rather point to heterogeneity, in that a concentration of under 15-year-olds among the early onset group would be traced to personality disorders or neurotic development. A family history of atypical depression and/or occurrence of bipolarity (i.e. hypomanic episodes of at least 2 weeks duration) on the other hand would rather be presumed to be affiliated with major affective disorders. The respective findings are presented in Table 3.
About 42%-47% of the RBD evidence an age of onset under 15 years, as compared to major atypical depression (EDE and EDE (SYM)) with 29%-32%. The various groups do not differ significantly, with the exception of the NRBD group with an inferior share of age at onset under 15.
All of the diagnostic groups show significantly more positive family histories than the control group, but there is no difference between the individual depressive subgroups. There is also a tendency for the RBD to have a higher frequency than the EDE. The family history was evaluated asking the proband about the occurrence of a depressive syndrome among parents and siblings.
Hypomania, on the other hand, is underrepresented in the RBD as compared to the EDE cases. The cells are however too small for statistical computations. The underrepresentation could be a consequence of definition. The defining criterion for a hypomanic case was a minimum duration of 2 weeks. Considering the fact that the RBD are characterized by short depressive mood swings in particular, it might well be that in this group the hypomanic episodes are also generally shorter than 2 weeks and thus do not reach the threshold of definition, although their frequency may be equal to that in the EDE cases. However, our findings do not confirm this presumption.
Symptoms and Syndromes
Criterial Symptoms of Depression. The eight DSM-III symptoms for the definition of major atypical depression have likewise been adopted for the minor depressive group. The mean values are given in Table 4. There is practically no difference between the EDE and RBD groups, which means that the short-lasting depressive manifestations have the same number of classical symptoms as the longer-lasting ones. Hence, the hypothesis of a higher symptom frequency among major atypical depression (EDE) is not confirmed. Indeed, one out of eight symptoms, ‘loss of interest’, is significantly more frequent in RBD cases.
Functional Syndromes. The comparison of the various depressive groups among each other and with the control group respectively regarding the presence of other functional syndromes is of great interest (Table 5) as it detected marked differences. The NRBD differ significantly from the control group only in reduced appetite and sexual symptoms, while the RBD evidence significantly more functional auxiliaries: as many as 13 out of 21 functional syndromes are significantly more frequent in the RBD than in the control group. ‘Intestinal troubles’ head the somatic syndromes, and anxiety, exhaustion, sleep disturbances the neurotic and vegetative syndromes. Out of 21 functional syndromes, 2 (anxiety states and premenstrual symptoms) have higher frequency in group RBD than in group EDE (Table 6).
5.3 Self-Assessment of Symptoms
The self-assessment by means of the SCL-90 symptom inventory yields several findings that could be expected (Table 7). The control group has the lowest scale values, the NRBD lies between the controls and the RBD. Here again, we observe no significant difference between the RBD and the EDE (Table 8).
A quantitative and qualitative parallel of symptomatology in both RBD and EDE appears. The mid-position of the NRBD is in line and compatible with the hypothesis of a continuum.
Illness behavior is studied under the following angles:
- Seeking professional treatment of atypical depression
- Discussing one’s atypical depression with close persons
- Extent of subjective worrying about one’s atypical depression
Naturally, these criteria do not apply to the controls. And to economize time, these questions were not investigated within the NRBD group below a certain threshold value. We therefore restrict the analysis to the comparison of the systematically investigated groups RBD and EDE (Table 9).
In the course of the past 12 months, one-fifth to one-third of the probands who had been classified as depressives sought professional care because of their depressive syndromes. It is interesting to note that the treatment frequency of recurrent short atypical depression and that of longer-lasting atypical depression differ in no way. One-third of each the RBD (SYM) and the EDE (SYM) groups—i.e., under the more stringent case-definition—have been treated. Self-medication is very similar: it lies between 4% and 11%, and there are also no differences between the groups.
Surprisingly, the majority of the probands discussed their depressive states with close persons. No differences were found between the groups.
The same can be said of the extent to which one worries about one’s atypical depression and which was assessed with an analog- scale ranging from 1 to 100 (THERMO). Subjects with short- lasting atypical depression suffer from these manifestations to an equal extent as those having longer-lasting depressive episodes.
Consequently, measured by illness behavior, RBD and EDE engender equal impairment and treatment quota.
Environmental Factors and Onset of Depression
Environmental factors that might be involved in triggering off atypical depression were assessed on different levels within the frame of our interviews:
- Precipitation. Persons who had suffered any depressive manifestations in the course of the past 12 months were asked directly whether they knew of any specific cause of their sad mood.
- Life events were assessed with a life-event-inventory based on Tennant and Andrews (1976). Following the data obtained by a Swiss normative study, we made up a ‘life- stress-scale’ (Bischofberger and Thomaier 1982).
- Social adjustment. A modified version (Angst et al. 1984d) of the SAS (Weissman and Paykel 1974) was applied. To assess distress and conflict experienced in interpersonal relationships during the last 3 months we established two
Precipitation. It is interesting to note that persons with longer- lasting depressive episodes (EDE groups) (65% and 57%) indicate a specific reason for a depressive manifestation significantly more often than persons of the RBD (44% and 33%) or NRBD (38%) groups (Tables 10 and 11). It therefore appears that the shorter-lasting depressive manifestations which keep recurring throughout a year are subjectively much less felt as having been induced by environmental factors.
Life Events. Distressing life events are distributed equally across all depressive groups, and—surprisingly—none of the latter differ from the control group. Hence, distressing life events cannot be considered as a trigger to depressive manifestations. This is quite startling, because both life events and depressive manifestations fall within the same span of 12 months. A higher global life event score is therefore not only no significant trigger of BDE, but also does not occur simultaneously with the latter.
Social Integration. Table 10 indicates the extent to which the probands of the RBD groups are exposed to chronic distress and conflict in their interpersonal relationships. The RBD and RBD (SYM) cases judge their interpersonal relationships significantly more often to be pregnant with distress and conflict than the control probands. On the other hand, the NRBD do not differ significantly from the control group. Here again,
we have no difference between the RBD and EDE groups: people with longer (more than 2 weeks) lasting depressive episodes do not report more conflict and distress in their interpersonal relationships than those having RBD episodes (Table 11).
The fact that persons having atypical depression often experience conflict and distress in their everyday life leads to the question whether and what differences can be found in the various social roles. Table 12 shows the association of atypical depression with conflict and distress in the social role areas ‘family’, ‘friends’, ‘work’ and ‘partners’.
It appears that the associations in the two RBD groups are generally stronger than those in the EDE groups, even though the differences between short and longer-lasting atypical depressions are not significant. The RBD and RBD (SYM) probands seem to experience intimate close relationships as often being distressing. Conflict and distress in relationships with parents, siblings or partners are particularly associated with RBD. Further analyses will have to examine whether this presence of conflict and distress in RBD is synonymous with personality disorder, and whether this group therefore could be heterogeneous, for instance consisting of a ‘neurotic’ conflictual atypical depression and primary atypical depression.
The accumulation of interpersonal conflict and distress in depressive people might originate from a lack of coping resources. Persons with poor coping resources seem more prone to atypical depression. Our interview assessed two dimensions of coping—mastery and self-esteem (Pearlin and Schooler 1978). Table 13 gives the distribution over the quartiles of the two scales for the control and the depressive groups.
The RBD and RBD (SYM) subjects have markedly fewer coping resources than the controls on both dimensions ‘mastery’ and ‘self-esteem’. As compared to the controls the NRBD group scores lower in ‘mastery’ but not in ‘selfesteem’. Between the RBD group and the EDE group there are no differences (Table 14). Summarized, people suffering from RBD or EDE often feel helpless and inferior to their environment.
Our findings support the dichotomy of BDE lasting less than 2 weeks into recurrent (R) and nonrecurrent (NR) subtypes:
Recurrence means, in this context, monthly occurrence of depressive mood changes over 1 year. By applying additional diagnostic criteria such as ‘work impairment’ and presence of a minimum symptom number, a harder RBD (SYM) group can be obtained which was frequently in need of treatment. The NRBD subgroup has practically never been treated (1%), the RBD in 18%, and the RBD (SYM) in 33%. The latter should be considered as ‘cases’ as much from a social as a medical point of view.
The 1-year prevalence rate of RBD (SYM) cases is 4.4% at the age of 23 to 24 years, 3.9% in males and 4.9% in females. The male/female ratio therefore is 1:1.3. In point of fact, it is rather thought to approach 1:1, because men tend to underreport depressive mood changes (Angst and Dobler-Mikola 1984a).
Subjects with minor atypical depression (RBD (SYM)) and major atypical depression (EDE, EDE(SYM)), are equally ill, impaired, and in need of treatment. This is documented by several criteria (frequency of medical treatment, number of depressive symptoms, SCL-scale values, social impairment, positive family history). A syndrome of anxiety assessed by the interview was significantly more frequent in minor atypical depression (RBD) with 50% than in major atypical depression (EDE) with 29%. The RBD also tend to score a little higher than the EDE on the respective SCL-scale ‘anxiety’ with 2.0 versus 1.8 (not significant). So it is possible that among minor depressed subjects a premorbid trait of anxiety is frequently present as suggested by the study of Van Valkenburg et al. (1983). It could well be that subjects with RBD differ in personality from EDE being more unstable, ‘neurotic’ and hostile. Unfortunately the critical SCL-scale values (somatization, hostility, anxiety) do not differ significantly, but there still remains a trend to higher scores (more complaints) in the RBD(SYM) group.
The striking resemblance of RBD and EDE reconfirms the hypothesis of a continuum between short and protracted depressive manifestations. Akiskal’s statement (1983) that actual major affective disorders can be superimposed on ‘subaffective disorders’ and that the latter genetically represent attenuated forms of the major disorders is certainly legitimate. Progress has been made in diagnostics by the diagnosis of a ‘double atypical depression’ (Keller et al. 1983), i.e., the presence of minor and major atypical depression in one and the same person. Our own data cannot give evidence in this respect. However, it might be possible to find out prospectively about the risk of a RBD illness switching to major atypical depression with time.
Our diagnosis of an RBD is not identical with the DSM-III concept of ‘dysthymic disorder’. The latter requires chroni- city, as persistent or intermittent depressive manifestations of at least 2 years’ duration, whereas our diagnosis requires only 1 year of intermittent, nonchronic course. The second criterion of the DSM-III, subsyndromal intensity, is only satisfied with respect to time, i.e., a duration of less than 2 weeks. We have adopted the same strict conditions for the symptomatic threshold as in major atypical depression, and even additionally required work impairment, to define a case. The exclusion criterion ‘affective delusions and hallucinations’ also applies for our case definition, but not that of an ‘absence of hypomanic or manic swings’. The latter diagnostic requirement appears to be inconsequent and would not prevent an RBD diagnosis, but rather necessitate the supplement ‘bipolar subtype’.
We believe our RBD (SYM) definition to be operationally more strict than the DSM-III concept of dysthymic disorder or atypical atypical depression. Obviously, it results in a somewhat smaller group of depressive syndromes, which—as we hope to have proved—is nevertheless of clinical relevance. Further investigations will have to clarify the relationship between RBD (SYM) and personality disorders and other types of secondary atypical depression. In the search for heterogeneity we shall break down the RBD (SYM) by age of onset, precipitation and conflict in social roles (conflictual atypical depression).