Clinical features of soft bipolarity in major depressive inpatients
Takeshi Utsumi, MD, PhD, Department of Psychiatry, Teikyo University School of Medicine, 2-11-1, Kaga, Itabashi-ku, Tokyo, 173-8605 Japan. Email: email@example.com
Abstract Because of the difficulties of ascertaining episode of hypomania by past history of the patients, it is of clinical value to find variables which predict the development of bipolar II disorder in depressive patients. Taking advantage of relatively long hospitalization, the authors tried to elucidate fine clinical features of the soft bipolarity. The subjects were 39 patients with Major Depressive Episode, diagnosed according to the 4th edition of the Diagnostic and Statistical Manual criteria. Among them, 15 patients were diagnosed as bipolar II disorder (BPII), whereas 24 patients were with unipolar depression (UP), using a structured clinical interview to assess the mood spectrum (SCI-MOODS). In addition to ordinary clinical and demographic variables, the authors studied fine symptomatology of depression, premorbid personality, and interpersonal relationship. Continuous variables were analyzed by t-test. Categorical variables were tested by χ2 analysis. In terms of premorbid personality, manic type (Zerssen) was found more frequently in BPII (UP 2/24, BPII 9/15, P < 0.05). Patients with BPII tended to show apparently quick disappearance of depressive symptoms (UP 2/24, BPII 9/15, P = 0.01). The most prominent result was a high prevalence of comorbidity of borderline personality disorder (BPD) among BPII (UP 0/24, BPII 6/15, P = 0.02). As Akiskal indicated that mood lability represents the most powerful predictor of hypomanias, patients with BPII showed quick response in mood to admission. The current subjects with BPII had high frequency of manic type of premorbid personality, indicating the usefulness of this variable for the prediction of hypomanias. Finally, the authors could observe development of BPD during hospitalization exclusively among BPII, to support the possibility of BPD as a state effect of BPII.
Given the therapeutic and prognostic importance of the unipolar–bipolar dichotomy, detecting what clinical features associate with the bipolarity in depressive patients is of paramount clinical significance. Much past research has focused on differences between depressive states appearing in unipolar depression and those in full-blown manic depressive or bipolar I disorder. Because of the difficulties of ascertaining episode of hypomania by past history of the patients,1–4 it is of clinical value to find variables which predict the development of bipolar II disorder in depressive patients. This disorder is characterized by severe depressions, typically recurrent, and hypomanias of at least a few days’ duration. Akiskal has termed ‘soft bipolarity’ in order to index more subtle psychopathology in bipolar II disorder and other bipolar spectrum conditions.5 Although bipolar II disorder was listed as one of authentic mood disorders in the 4th edition of the Diagnostic and Statistical Manual (DSM-IV),6 many cases are still being misdiagnosed as major depressive episode of unipolar disorder and subsequently receiving wrong medication. As a result, they often show miserable outcomes including suicide attempt, frequent recurrence and rapid mood swings.
In order to clarify fine psychopathological features of bipolar II disorder, the authors took advantage of relatively long hospitalization spans in the Japanese clinical setting. The mean duration of admission to psychiatric wards for patients with depressive disorder is more than 3 months in Japan on average, while that in Western countries is about 2 weeks. This might be explained in part by the Japanese medical insurance system enabling long-term treatment in hospital, which makes it possible for doctors to elucidate more detailed history, symptomatology, clinical course and the interpersonal relationship elicited in psychiatric wards. In this paper, the authors would like to characterize clinical features predicting the soft bipolarity.
The subjects are 39 patients with Major Depressive Episode, diagnosed on admission according to the DSM-IV6 criteria, with an age range of 17–57 years, who were consecutively admitted to the psychiatric ward of Teikyo University Hospital, Tokyo, Japan, during the period from July 2000 to June 2001. They were comprehensively evaluated for current and past history of hypomania, using a structured clinical interview to assess the mood spectrum (SCI-MOODS).7 As a result, among the 39 subjects, 15 patients were diagnosed as bipolar II disorder (BPII), whereas 24 patients were with unipolar depression (UP).
Marital status, family history of major psychiatric illnesses, duration of hospitalization, Global Assessment of Functioning (GAF) score, the number of the depressive, manic and hypomanic episodes and admissions, history of attempted suicide, and response to antidepressants were investigated through the clinical charts. Comorbid psychiatric disorders were diagnosed according to DSM-IV. In addition to these demographic and clinical characteristics, the authors studied the characteristics of symptomatology of depression during the admission. The authors defined patients whose depressive symptoms disappeared (<10 points for Beck Depression score) within 1 week after admission as a ‘quick responder to admission’. As other variables, premorbid personality was studied, using the Zerssen’s criteria.8
Continuous variables were analyzed by t-test. Categorical variables were tested by χ2 analysis. The calculations were conducted using the SPSS (ver.10.0J) for Windows package (SPSS Inc., Chicago, IL, USA).
Participation in the study was voluntary. Each subject provided written informed consent after receiving a complete description of the study. The protocol was approved by the ethics committee of Teikyo University Hospital.
Age, gender, marital status and duration of hospitalization were not significantly different between the bipolar II (BPII) and unipolar (UP) groups (Table 1). Number of family members (UP 3.9 ± 1.0, BPII 2.4 ± 1.2, P < 0.05) was significantly different (Table 2).
Table 1. Comparison between unipolar depression and bipolar II disorder
|Mean age (years (SD))||34.3 (12.3)|| 31.9 (7.7)|
| Male|| 5 (20.8%)|| 8 (53.3%)|
| Female||19 (79.2%)|| 7 (46.7%)|
| Never married||10 (41.7%)|| 8 (53.3%)|
| Ever married||14 (58.3%)|| 7 (46.7%)|
|Duration of hospitalization (days [SD])||86.0 (65.7)||103.0 (54.5)|
Table 2. Comparison of premorbid variables between unipolar depression and bipolar II disorder
| (+)|| 6 (26.1%)|| 5 (33.3%)|| |
| (–)||17 (73.9%)||10 (66.7%)||ns|
|Number of family members|
| Mean (SD)|| 3.9 (1.0)|| 2.4 (1.2)||P < 0.05|
| Manic type|| 2 (9.1%)|| 9 (69.2%)||P < 0.05|
| Melancholic type||20 (90.9)|| 4 (30.8)||P < 0.05|
In terms of premorbid personality, manic type (Zerssen) was found more frequently in BPII (UP 2/24, BPII 9/15, P < 0.05; Table 2).
In terms of clinical features, ‘quick response to admission’ was more common in BPII (UP 3/24, BPII 8/15, P = 0.01; Table 3). The most prominent result in this study was high prevalence of comorbidity of borderline personality disorder (BPD) among BPII. While none of UP had this trait, 6 of 15 BPII subjects developed the trait of BPD during admission (P = 0.02; Table 3). These patients showed stable and lasting clinical features meeting the profile of BPD by DSM-IV,6 which were confirmed after the remission of the mood disorder.
Table 3. Comparison of clinical features between unipolar depression and bipolar II disorder
|Quick response to admission||3 (12.5%)||8 (53.3%)||0.01|
|Instability of episode||9 (37.5%)||7 (46.7%)||ns|
|Idea of reference||6 (25.0%)||9 (60.0%)||ns|
|Psychotic features||4 (17.4%)||1 (6.7%)||ns|
| Anxiety disorder||7 (30.4%)||6 (40.0%)||ns|
| Eating disorder||5 (21.7%)||2 (13.3%)||ns|
| BPD||0 (0.0%)||6 (40.0%)||0.02|
| Substance abuse||4 (17.4%)||7 (46.7%)||ns|
Diagnosis of bipolar II disorder is generally considered to be unreliable principally because hypomanic episodes can easily be missed in ordinary outpatient clinics, even when interviewed by the experts. Especially when the patient is the sole source of information, identification of episode tend to be more difficult, because hypomania has not been experienced as a morbid event but rather as a favorable ego syntonic episode. Then, it is a laborious task to detect a past hypomanic episode to establish reliable diagnosis of BPII.
In hospital practice, thorough screening of past episodes could be done by taking history repeatedly and using other sources of information, to establish a precise diagnosis for the type of mood disorders. Although there existed the limitation of sample size, the current study has advantage on extraordinary accuracy of diagnosis for BPII. Moreover, hospital treatment enabled the authors to observe the patients in situ. One could directly note various facets of patients through hospital life, including their behavioral pattern, personality traits, and interpersonal relationship.
Differential diagnosis of depressive states
In terms of psychopathology, it is very important clinically to describe the feature of depression predicting the bipolarity. Many studies illustrated the difference in depressive states between unipolar depression and bipolar I disorder. For example, Marchand reviewed presentation of a depressive patient who may be bipolar; hyperphagia, hypersomnia, melancholic features, severe anhedonia, seasonal mood change, psychomotor slowing, psychotic features, history of poor response to antidepressants, history of recurrent but brief depressive episodes, history of antidepressant-induced mania or hypomania, family history of bipolar illness in first-degree relatives, early age of onset, and postpartum onset.9
For BPII, Akiskal et al. demonstrated factors predicting the switch into hypomanias by a comprehensive prospective study.10 Akiskal et al. have shown that mood lability represents the most powerful predictor of which depressive patients would develop frank hypomania.
In the current study, many BPII patients showed improvement of depressive state in the short-term. In 9 of 15 patients, depressive symptoms disappeared within 1 week. This suggests that, on admission, they showed abrupt switching of mood from precedent depressive state seen in the outpatient setting, although the authors could not discriminate clear hypomania or mixed state in these patients. However, this disappearance of depressive symptoms did not necessarily mean the remission of the disease, as indicated in the result showing no difference in the duration of hospitalization between UP and BPII. It is necessary to keep in mind that the admission itself could be a major risk factor for evoking hypomania by releasing the patients from their suffering in ordinary life.11
The subjects in the present study showed a clear contrast in their premorbid personality. Many BPII patients had the manic type, whereas most UP patients had melancholic type (Tellenbach12). The manic type was characterized by traits as follows: inconstant, independent, unconventional, broadminded, imaginative, generous and daring. It was abstracted by Zerrsen as a premorbid character of the unipolar mania, a rare type of mood disorder which manifests exclusively manic episode. This premorbid character is not so often found among the patients of bipolar I disorder. Rather, some clinical reports referred to the association between the manic type and the bipolar spectrum, especially bipolar II disorder. As some authors pointed out, the manic type is not clearly distinguished from disease, but it forms a kind of continuum with bipolar disorder.13,14 In BPII patients, the state and the trait tend to interweave each other, so one could find often the manic type of character among them. The present result supported these clinical findings. Besides, significantly fewer numbers of family members among BPII may reflect the life full of ups and downs often seen in the manic type.
Comorbidity of borderline personality disorder
In addition to the premorbid character, the present study found the association between BPII and borderline personality disorder. The most outstanding finding is that the authors could observe not the coexistence but the development of BPD among four of six patients in the clinical course, taking advantage of the relatively long hospitalization. This finding was reflected in the observation of patients’ interpersonal relationship. The inpatients with UP tended to accept their sick role and to behave as a follower. In contrast, the patients with BPII proved to be far unstable. A considerable part of them appeared to be a leader or a wire-puller. Some showed the psychopathology of manipulation and alternation between idealization and devaluation, which suggests the possibility of the transition between BPII and BPD.
The relationship between mood disorder and BPD has been an important controversial issue.15,16 Recently, many studies support that BPD could be the state effect of BPII.17–20 The present study showed that BPII could develop clinical features compatible to BPD.
Therefore, the authors tried to delineate the characteristics referring to soft bipolarity. Although useful screening scales have been developed to detect hypomanic episodes, such as the SCI-MOODS7 and the Mood Clinic Data Questionnaire (MCDQ),21 clinicians should pay attention to psychopathology other than mood.