The boundary between mixed and manic episodes in the ICD-10 classification

Objective: To investigate the boundary between ICD‐10 mixed and manic episodes, which has apparently remained understudied. Method: In‐patients with ICD‐10 mixed (n=36) and manic episodes (n=145) were compared in terms of demographic, clinical, therapeutical and outcome variables. Results: Of in‐patients with manic episode, 26 (18%) had several depressive symptoms at admission. These patients (dysphoric manic patients) were very similar to patients with ICD‐10 mixed episode in terms of current symptomatic presentations and several clinical and therapeutic variables, which were significantly different from those in patients with pure mania. Conclusion: The ICD‐10 boundary between mixed and manic episodes is unlikely to be effective although experienced clinicians made the diagnoses. The system may have a high probability of diagnosing dysphoric manic patients as having manic episode, despite their great similarities to patients with mixed episode in terms of current psychopathological presentations as well as clinically important variables.

Keywords: bipolar disorder; mood disorders; mixed states

Early authors such as Kraepelin ([1]) and Weygandt ([2]) first described mixed mania as a psychiatric condition, where both prominent manic and depressive syndromes simultaneously co‐occur for a certain period. Recent researchers in the field of mood disorders have renewed their interest in mixed mania by studying its symptomatic presentations ([3][4][5][6][7][8][9][10][11][12][13][14][15][16][17]), demographic variables ([3][4][5], [18]), specific pharmacological responses ([19][20][21][22]), specific biological substrates ([23][24][25][26]), and natural courses ([27][28][29][30][31][32][33]). Studies reveal that patients with mixed mania may be characterized by a higher proportion of female subjects, a higher incidence of suicidal ideation, a poorer response to lithium carbonate, a shorter interepisode stability and a higher likelihood to have had a past mixed episode, compared with patients with pure mania. The majority of these studies have concentrated on the boundary between mixed and pure manias, as defined by North American diagnostic systems such as DSM‐III‐R and DSM‐IV: these diagnostic systems define mixed mania as the meeting criteria for both major depressive and manic episodes at least for 1 week. Recently, McElroy and colleagues ([3]) proposed the Cincinnati criteria for mixed mania, which require for the diagnosis a full manic syndrome plus three or more depressive symptoms. Several studies report that the Cincinnati criteria may provide a clinically more meaningful classification of mixed mania than do the DSM‐IV criteria ([7], [14][15][16], [20], [31]). Little is known, however, as to the boundary of mixed states, as defined by the ICD‐10 system, a diagnostic classification that is widely used in countries other than North America.

According to the ICD‐10 system ([34], [35]), a diagnosis of mixed episode is made, when both depressive and manic symptoms are, to a similar degree, co‐occurring at least for 2 weeks. Compared with the DSM‐IV classification and the Cincinnati criteria, the ICD‐10 definition of mixed episode seems somewhat vague, because precise rules regarding the severity of admixed depressive and manic symptoms required for the diagnosis are lacking. This indistinctness would not directly implicate the inferiority of the ICD‐10 diagnosis of mixed episode, as many researchers pointed out ([1][2][3][4]), the phenomenology of mixed episode is complex and very labile, and reasonable boundaries between mixed and depressive episodes and between mixed and manic episodes still remain unclear. However, the lack of rules regarding the symptom severities may cause limitations in reliability of the ICD‐10 concept of mixed episode.

The aim of the present study was to examine whether or not the ICD‐10 boundary between mixed and manic episodes is effective.

A total of 40 variables, which were shown by previous studies to play a role in distinguishing between pure manic and mixed episodes, were compared among three groups, patients with mixed episode, and manic patients with and without depressive symptoms.

In our hospital, clinical data including clinical diagnoses, sociodemographic variables, psychopathological variables assessed by using standardized instruments (see below), and pharmacotherapeutic variables before and during hospitalization are systematically rated and databased. The majority of data shown in the present study were obtained from this database, which may minimize possible contamination from several bias sources usual in a retrospective study. All patients, who were hospitalized at the Psychiatric Hospital, Ludwig‐Maximilian University, Munich during the period of 1995–2000, and were diagnosed as currently having an ICD‐10 mixed (F31.6 and F38.0) or manic episode (F30.1, F30.2, F31.1, and F31.2), were considered as subjects of the present study. Clinical diagnoses for the patients were made by means of a consensus among experienced psychiatrists including at least one person with a professor designation. These diagnoses were made solely on the basis of ICD‐10 diagnostic criteria. Forty patients currently having a mixed episode and 145 patients currently having a manic episode were identified. The ICD‐10 system includes into mixed episode a rare condition, called 'unstable mixed states' ([36]) or 'ultra‐rapid cycling', which is characterized by a very rapid alternation between manic and depressive symptoms. Patients with less severe manic symptoms (hypomanic level, as defined by the ICD‐10) may also be included into mixed episode when using the ICD‐10. Based on systematic psychiatric ratings at admission (see below) and chart reviews, two patients each had 'ultra‐rapid cycling' and manic symptoms in a hypomanic level, respectively. These four patients in total were too small for statistical analyses, and were therefore excluded for the following analyses, leaving 36 mixed patients.

In our hospital, a broad range of 196 psychiatric and related somatic symptoms were, as part of the routine documentation at the hospital, systematically evaluated for all patients at both admission and discharge by using a standardized instrument (the AMDP‐system, Association for Methodology and Documentation in Psychiatry) ([37]). The AMDP‐system is a comprehensive rating instrument, developed on the basis of German traditional descriptive psychopathology on functional psychoses: it is commonly used in most psychiatric institutes in German‐speaking countries. Each psychiatric symptom in the AMDP‐system is scored from 0 (absent) to 3 (severe) with defined anchor statements by using a semistructured interview method. Several studies indicated moderate to high interrater agreements for most included symptoms ([38], [39]). Assessments by using the AMDP‐system usually summarize a patient's psychiatric states for a preceding 1‐week period ([37]). On the basis of the AMDP‐scorings, the presence of several psychiatric symptoms, which compose major international diagnostic criteria for manic and (major) depressive episodes ([34], [35], [40]), were explored in the present study. The presence of each psychiatric symptom was defined in the present study as being scored 1 or higher on the corresponding AMDP item.

On the basis of the AMDP‐ratings, several summary scales for depressive, manic and paranoid–hallucinatory syndromes (the AMDP scores for depressive, manic, and paranoid–hallucinatory syndromes), which have been validated in large psychiatric samples ([41][42][43]), can be calculated by summing up the scores on 13 items (rumination, loss of feeling, loss of vitality, depressed mood, hopelessness, feeling of inadequacy, feeling of guilt, inhibition of drive, worse in the morning, interrupted sleep, shortened sleep, early waking, decreased appetite) for depressive syndrome, seven items (flight of idea, euphoria, exaggerated self‐esteem, increased drive, motor restlessness, logorrhea, excessive social contact) for manic syndrome, and 13 items (delusional mood, delusional perceptions, sudden delusional ideas, delusional ideas, systematized delusions, delusional dynamics, delusions of reference, delusions of persecution, verbal hallucinations, bodily hallucinations, depersonalization, thought withdrawal, other feelings of alien influence) for paranoid–hallucinatory syndrome. The AMDP scale for depressive syndrome includes several items related to sleep and appetite disturbances, which may also appear in pure manic phase. To appropriately assess the depression severity in patients with manic and mixed episodes, we constructed another AMDP depression subscale (AMDP scale for depressive syndrome without sleep or appetite disturbance) by summing up the scores on nine AMDP items (rumination, loss of feeling, loss of vitality, depressed mood, hopelessness, feeling of inadequacy, feeling of guilt, inhibition of drive, worse in the morning). Cronbach's alpha coefficients of the new scale, as calculated for assessments at admission and discharge in the subjects, were 0.84 and 0.73, respectively, indicating an acceptable internal consistency. The GAF score was also assessed for all subjects at both admission and discharge. The CGI (Clinical Global Impression) score was assessed in a similar manner at discharge. Data regarding family history of affective disorders were based on comprehensive information from the patients themselves and their relatives, but were collected non‐blindly to clinical variables. All subjects gave informed consent to be assessed by using the instruments. Well‐trained psychiatrists, who were one of the experienced psychiatrists engaged in clinical diagnosis of each patient, administered all the instruments. The patients were treated with medications as clinically appropriate during the index hospitalization in a naturalistic manner.

All included mixed patients (n=36) had three or more depressive symptoms, listed in the Cincinnati criteria for mixed mania. On the other hand, 26 (23%) of 145 manic patients included in the study had three or more Cincinnati criteria for mixed mania. The scorings on the AMDP‐system were mainly used to know whether or not a patient met each symptom of the Cincinnati criteria. McElroy and colleagues ([3]) proposed the criteria to appropriately assess depressive symptoms that are possibly related to mixed mania. The criteria were reported by several studies to be successful for establishing a reasonable boundary between manic and mixed episodes ([7], [14][15][16], [20], [31]). The criteria include nine depressive symptoms (depressed mood, decreased interest or pleasure, increase in appetite, hypersomnia, psychomotor retardation, fatigue or loss of energy, feelings of worthlessness or excessive guilt, hopelessness, suicidal ideation; the requirement for diagnosing mixed mania is the presence of three or more symptoms). The AMDP‐system includes the majority of the symptoms listed in the Cincinnati criteria: but one depressive symptom in the criteria, hypersomnia, is not covered by the AMDP‐system. This symptom was considered as present, when the scorings on the AMDP‐system did not provide evidence of insomnia. Medical records of these patients without insomnia were then checked to know if these patients had hypersomnia. The 26 manic patients with depressive symptoms and the remaining pure manic patients (n=119) were called in the present study as patients with dysphoric mania and patients with pure mania, respectively.

The established three groups, 36 patients with mixed episode, 26 patients with dysphoric mania, and 119 patients with pure mania, were compared in terms of the variables, which were suggested by previous studies to be related to the boundary between mixed and pure manic patients. For overall comparisons among the three groups, we used chi‐squared tests for categorical variables, and analyses of variance (if the statistical assumptions for this analysis were met) or Kruskal–Wallis tests for continuous variables. The significant level for each overall comparison was adjusted at P<0.0013 (0.05 divided by 40 tests) by using Bonferroni's method. A significant overall difference, produced by a chi‐squared test or Kruskal–Wallis test, was followed by multiple comparisons using Dunn's method (a non‐parametric method for multiple comparisons) ([44]). Turkey's multiple comparisons followed a significant F‐value. All statistical analyses were performed by using an SPSS software ([45]). All statistical statements in the study were two‐tailed.

The results are demonstrated in Tables 1–tblr rid="t3">3. Of demographic and clinical variables (Table 1), the period since the last admission and suicide attempts during the index episode were significantly different among the three groups, although the three groups did not significantly differ in age, age at first onset of affective episode, age at first hospitalization, family history, or lifetime history of alcohol or drug abuse. Both patients with mixed and dysphoric episodes had a shorter period since the last admission, and a higher incidence of suicide attempts, compared to patients with pure manic episode. The proportion of female subjects was higher in both mixed and dysphoric patients than in pure manic patients; but the difference was not significant after Bonferroni's correction (χ2=9.03, df=2, P=0.012).

Most psychopathological variables were significantly different among the three groups (Table 2). The current manic syndrome, as assessed by the AMDP manic scale, was significantly higher in pure manic patients than in mixed and dysphoric patients, while the latter two groups had a greater severity in current depressive syndrome, as measured by the two AMDP depressive scales, compared with the former group. Pure manic patients had higher frequencies of several manic symptoms (euphoria, grandiosity, and increased drive) than did the other two groups; but irritability, pressured speech and distractivity were equally frequent in the three groups. All depressive symptoms explored were significantly more frequent in mixed and dysphoric patients than in pure manic patients. The current severity of paranoid–hallucinatory syndrome and the GAF score at admission were not different among the three groups.

Table 3 summarizes pharmacological treatments during the index hospitalization and outcome variables. Both mixed and dysphoric patients more frequently required antidepressant treatments, as compared with pure manic patients. Although the duration of the index hospitalization was not different among the three groups, patients with mixed and dysphoric episodes had a significantly greater severity of depressive syndrome and a significantly higher CGI score at discharge, compared with patients with pure manic episode, suggesting a worse outcome in the former two groups.

All tests for groups comparisons, performed in the study, showed that both patients with mixed and dysphoric episodes significantly differed from patients with pure mania, while no significant difference was found between mixed and dysphoric manic episodes in these analyses.

Compared with DSM‐IV criteria, the ICD‐10 concept of mixed episode is composed of several different psychiatric conditions. In addition to a condition, where a full manic syndrome is accompanied by a similarly severe depressive syndrome (this is a similar condition to mixed mania, as defined by DSM‐IV), it may include two possibly different clinical entities, 'ultra‐rapid cycling' ('unstable mixed states'), and a condition, where a hypomanic syndrome is accompanied in a similar severity by a (hypo‐) depressive syndrome. In our sample of in‐patients, the latter two conditions were very rare (two patients each had these conditions), so that the present study explored the boundary of the first condition only. Despite this limitation, this is, to our knowledge, the first study that examines the boundary between ICD‐10 mixed and manic episodes.

The present study appears to indicate that the ICD‐10 mixed episode has some validity when compared with manic episode without depressive symptoms (pure mania). However, a considerable proportion (18%) of patients with an ICD‐10 diagnosis of manic episode had three or more depressive symptoms, as listed in the Cincinnati criteria for mixed mania. These patients (dysphoric manic patients) had quite similar symptomatic presentations to patients with an ICD‐10 diagnosis of mixed episode, and the two groups were significantly different from patients with pure mania in the current depression severity, the current mania severity, and frequencies of several manic symptoms and all depressive symptoms explored. The dysphoric manic patients were also similar to patients with mixed episode, and these two groups were significantly different from pure manic patients in several demographic, clinical, therapeutical and outcome variables: these variables were reported by previous validation studies to be related with mixed mania ([3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27]). Both patients with mixed and dysphoric manic episodes tended to have a somewhat higher proportion of female subjects (P=0.012), had a shorter period since the last admission, suggesting a lower inter‐episode stability, and a higher incidence of suicide attempts during the index episode, as compared with pure manic patients. The two groups more frequently required antidepressant treatments during the index hospitalization, but their depressive symptoms were less likely to be well resolved even after the treatments, thus resulting in worse outcome at discharge, as measured by the CGI score. These findings indicate that the boundary between mixed and manic episodes, as established by the ICD‐10 system, is unlikely to be effective although the diagnoses are made by experienced clinicians; and that this diagnostic system may have a high probability of diagnosing dysphoric manic patients as having manic episode, despite their large similarities to patients with mixed episode in terms of current psychopathological presentations as well as several other clinically important variables.

That the ICD‐10 system may have inaccuracy with regard to the boundary between mixed and manic episodes is probably due to the vague definition of mixed episode in the diagnostic system. Our results suggest that a clear definition regarding the number of admixed depressive symptoms is required for establishing a clinically meaningful boundary between mixed and manic episodes in the system. The present study used an operational criterion, the presence of three or more depressive symptoms among the Cincinnati criteria ([3]), for settling a provisional boundary between mixed and manic episodes, and this procedure appeared somewhat successful in our data. This indicates that the Cincinnati criteria may provide a reasonable boundary between mixed and manic episodes in the ICD‐10 system.

A further possible explanation of a high probability of the misdiagnoses may be that the ICD‐10 requires for a diagnosis of mixed episode a simultaneous presence of both depressive and manic symptoms for at least 2 weeks. Given the findings in this study that the presence of depressive symptoms at admission effectively distinguished dysphoric mania from pure mania in several clinical, therapeutical and outcome variables, the ICD‐10 criterion regarding the duration of a co‐occurrence of manic and depressive symptoms may be too long. The assessments using the AMDP system, used in the present study, summarize a patient's psychiatric states during 7 days preceding his admission. This suggests that a 1‐week duration of the simultaneous presence of both depressive and manic symptoms possibly results in a reasonable classification. A study with a precise and intensive observation of daily symptomatic alternations from the onset of an episode may further clarify the duration necessary for a reasonable diagnosis of mixed episode: such a study is, however, very difficult to conduct, and no study in this nature has been conducted for any kind of diagnostic criteria for mixed episode.

The definition of mixed episode also requires a reasonable definition of the boundary between depressive and mixed episodes: the research on this boundary is, however, apparently behind in general, as compared with the boundary between manic and mixed episodes: only few studies have ever attempted to establish a reasonable boundary between depressive and mixed episodes ([4], [31], [46], [47]). The majority of these studies investigated the boundary in the DSM‐IV (or III‐R) system. It still remains unclear whether or not the ICD‐10 definition regarding the boundary between depressive and mixed episodes is valid, which requires further study.

Several cautions should be exercised in interpreting the results of the present study. First, data on family history were collected non‐blindly to clinical status. A large family study with a sophisticated methodology should be carried out to confirm our results. Secondly, this study focused only on data obtained during a hospitalization, although a broad range of variables was systematically assessed by using standardized instruments. Results in the study should therefore be endorsed by a long‐term prospective study, which attempts to clarify the natural course of mixed episode. Studies of this kind are apparently lacking, and are definitely required to identify the clinical meaningfulness of the concept of mixed episode.

Tetsuya Sato is supported by the Alexander von Humboldt Foundation.