Cycloid Psychosis: An Examination of the Validity of the Concept

Victor Peralta, MD, PhD, Manuel J. Cuesta, MD, PhD, and Maria Zandio, MD

Corresponding author Victor Peralta, MD, PhD Psychiatric Unit, Virgen del Camino Hospital, Irunlarrea 4, 31008 Pamplona, Spain. E-mail: victor.peralta.martin@cfnavarra.es Current Psychiatry Reports 2007, 9:184192 Current Medicine Group LLC ISSN 1523-3812 Copyright 2007 by Current Medicine Group LLC

Historical Perspective
The cycloid psychosis concept has a long tradition in European psychiatry, and its theoretical roots can be traced back to the work of Augustin Morel and to its concept of degeneration that was subsequently elaborated by Magnan. It was Magnan in the 1880s who first published a thorough description of a psychopathological disorder characterized by a sudden onset, polymorphous psychotic symptomatology, and recurrent course (bouffs dlirantes de les dgeneres). The notion of degeneration psychoses won acceptance in Germany, although it is no longer linked to the degeneration hypothesis, and laid the foundation for the concept of cycloid psychoses. The current conceptualization on the disorder stems from the Wernicke-Kleist-Leonhard school of psychiatry that was not prepared to accept Kraepelins very comprehensive definition of manic depressive illness and considered cycloid psychosis as a third psychosis in opposition to Kraepelins influential two-entity principle of endogenous psychoses [1,2]. The term zycloiden psychosen was first used by Kleist in 1926 to group together disorders that had been described up to that point and also many other conditions that he had previously classified under the headings delusional affective psychoses and affective psychoses. Kleist described two variants of the disorder, confusional insanity (characterized by contrasting phases of confused excitement and stupor) and motility psychosis (characterized by contrasting phases of hyperkinesis and hypo- or akinesis). Leonhard introduced a third variant, the anxiety-elation psychosis (characterized by contrasting phases of anxiety and elation that are related to paranoid and grandiose delusions, respectively), and set the modern conceptualization of the disorder [1,2]. The cycloid psychoses were considered by this author as a group of acute, recoverable psychoses that are neither manic depressive nor schizophrenic. More recently, Perris [3,4] produced a number of classical papers on the disorder and introduced the first operational criteria [5]. There is a broad agreement among authors that the main defining features of this disorder are acute onset, remitting course, benign outcome in the long run, and symptom polymorphism. These traits are also shared by alternative concepts, some of them national-based,

The diagnosis of cycloid psychosis has a long tradition in European psychiatry. However, it has been poorly assimilated within the DSM IV and ICD-10 diagnostic systems. Leonhard set the basis for the current conceptualization of the disorder, and Perris and Brockington developed the first operational diagnostic criteria. However, the two conceptualizations of the disorder are not the same and differ across a number of meaningful variables. Cycloid psychosis is a useful concept in that it possesses both clinical and predictive validity. Despite the high prevalence of mood symptoms and syndromes, cycloid psychosis does not equal schizoaffective disorder. Although a substantial body of evidence suggests that cycloid psychosis differs meaningfully from typical schizophrenia, it is less clear whether it differs from major mood disorders or represents an independent nosological entity. The existence of putative subtypes is also likely, and the differentiation between affective and nonaffective subtypes has received some support.

Introduction
This commentary is designed to review and summarize the literature on cycloid psychosis in order to show both the clinical validity and utility of this diagnosis and how it can illuminate the complex relationships between the prototypical psychotic disorders such as schizophrenia and affective psychoses. An attempt is made to examine available evidence on the validity of the diagnosis of cycloid psychosis, its relationship to the diagnoses of schizoaffective and brief or transient psychotic disorders, and its status regarding alternative nosological hypotheses.

Cycloid Psychosis: An Examination of the Validity of the Concept

Peralta et al.

185

Biological triggering factors

Puerperal psychoses

Figure 1. Alternative concepts of cycloid psychosis and related disorders. Circles represent (from innermost to outermost) the following: bouffee delirante; DSM IV brief psychotic disorder; ICD-10 acute and transient psychotic disorders; cycloid psychosis, Perris and Brockington criteria; and cycloid psychoses, Leonhard criteria.
Psychotic mood disorder

Schizophreniform disorder

Schizophrenia

Major mood disorders

Reactive psychoses

Psychogenic triggering factors

such as acute schizoaffective psychosis [6], schizophreniform states [7], reactive psychosis [8], bouffe dlirante [9], atypical psychosis [10], or puerperal psychosis [11]. Each of these different conceptualizations of (probably) the same disorder reflects the degree to which different features are regarded as essential (Fig. 1). The cycloid psychosis concept seems to have prevailed over the rest on the basis of both its integrative character and the well-developed nosological system in which it is rooted. Cycloid psychosis has received much attention in the international literature during the past half-century, although there has been a steady decline in the research interest about the topic in the last 2 decades. This seems to be due to the uncertain nosological status of that diagnosis and the misleading idea that the concept has become assimilated by the current formal diagnostic systems under the different variants of nonaffective, remitting psychotic disorders.

Diagnostic Issues
According to Leonhard [1,2], cycloid psychoses (plural) are a group of acute and remitting psychotic disorders that can be reliably diagnosed on the basis of a cross-sectional assessment of the characteristic clinical picture. In the words of that author, When the diagnosis of cycloid psychosis is made, then it necessarily follows that a complete remission will occur and even if the illness recurs no defect will be left behind [1]. Although Leonhard [2] made a fine and thorough description of the clinical picture of cycloid psychosis and its subforms, he never developed operational diagnostic criteria. This task was first undertaken by Perris and Brockington [5], who set the first operational criteria for diagnosing the disorder, which are intended to capture the core features of Leonhards concept. These criteria were developed on the basis of the comprehensive study by Perris [3], and they

have become standard for diagnosing the disorder. More recently, Sigmund and Mundt [12] and Jabs et al. [13] have developed guidelines for diagnosing the disorder and its subtypes that are very close to Leonhards concept. However, the reliability and validity of these criteria remain to be examined. The Leonhard approach to the diagnosis of cycloid psychoses represents a downtop approach (from symptoms across basic axial syndromes and subtypes to the disorder), whereas the Perris and Brockington criteria represent a symptom collection approach without specific syndrome patterns and, therefore, subtypes of the disorder. In contrast to Leonhards classification in subtypes, Perris argued that an admixture of symptomatology was the rule rather than the exception, and disregarded the classification in subtypes. An important feature of the Perris and Brockington [5] criteria is that good outcome, a typical feature of cycloid psychoses, is not included in the definition (as is the case in the modern diagnostic systems); therefore, a tautological definition of the disorder is avoided. Subsequently, Perris [4] clearly stated that in cycloid psychosis, there is never a fully developed manic or depressive syndrome, which led many other authors to consider the presence of a major mood syndrome as incompatible with a diagnosis of cycloid psychosis. This is an important question that influences the results of any nosological or validation study. We believe that the coexistence of a full affective syndrome should not be considered an exclusion criterion for diagnosing cycloid psychosis, because 1) this specific criterion was not included in the original criteria, 2) cycloid psychosis is mainly defined by a specific pattern of symptoms irrespective of other psychopathological features, and 3) it allows us to examine the predictive validity of the concept in relation to other clinical syndromes on a purely descriptive basis without a priori assumptions about associations (or lack of them) among psychopathological syndromes [14].

186

Nonschizophrenic Psychotic Disorders

Unfortunately, the Leonhard and the Perris and Brockington concepts have not been empirically compared previously, and the extent to which they represent the same conceptualization of the disorder remains largely unknown. Here we present data from our group examining the differential characteristics between the two concepts [15]. The sample examined was made up of 660 psychotic inpatients, 137 of whom met either Leonhard (n = 120) or Perris and Brockington criteria (n = 69) for cycloid psychosis. The 137 patients fulfilling either of the criteria were classified as 1) fulfilling the Leonhard but not the Perris and Brockington criteria, 2) fulfilling the two sets of criteria, or 3) fulfilling the Perris and Brockington but not the Leonhard criteria. These groups were compared across a number of variables (Table 1). Summarizing these results, the two sets of criteria seem to represent somewhat different conceptualizations of the disorder in that they showed modest concordance ( = 0.43) and differed across a number of meaningful clinical variables. Compared with the Perris and Brockington criteria, the Leonhard criteria capture a group of patients with higher age at onset, less acute onset, better response to treatment, and fewer mood symptoms. According to the DSM IV classification, in the Leonhard cycloids, the diagnoses of schizophrenia and major mood disorder were underrepresented, whereas the diagnosis of brief psychotic disorder (BPD) was overrepresented.

Fish [16] described four putative subtypes (psychogenic reactions, atypical affective disorders, epileptoid psychosis [thus, cycloid psychosis is associated with electroencephalogram abnormalities], and true cycloid psychosis). More recently, Mojtabai [17], using latent class analysis of symptoms and familial liability, could differentiate between affective and nonaffective subgroups. Peralta and Cuesta [14] provided some support for such a distinction in that compared with the affective group, the nonaffective group had a predominance of females, better premorbid adjustment, lower recurrence rate, better treatment response, and more severe psychosocial stressors.

Cycloid Psychosis Within the DSM IV and ICD-10 Diagnostic Systems

The successive editions of the formal diagnostic systems have not incorporated criteria for diagnosing cycloid psychosis, and the degree to which these diagnostic systems have included cycloid features, other than the remitting course, varies highly among classifications. The third edition of the Diagnostic and Statistical Manual (DSM III) did not include specific cycloid features, whereas the revised edition (DSM III-R) included emotional turmoil and perplexity or confusion as defining features for the brief reactive psychosis. In the DSM IV, these symptoms are considered as associated but not defining features of the BPD. In contrast to DSM classifications, the ICD-10 includes many cycloid features under the diagnosis of acute and transient psychotic disorders (ATPDs) and more specifically under the subgroups of acute polymorphic psychotic disorder without and with symptoms of schizophrenia (F23.0 and F23.1, respectively). These features are acute onset, shifting symptomatology, emotional turmoil, perplexity, and changes in the psychomotor activity. This has led some authors to erroneously equate the concept of cycloid psychosis with the ATPD from the ICD-10 [18,19]. Despite the consideration of some cycloid features in the formal diagnostic systems, they capture the cycloid psychosis construct only partially. Studies examining the nosological status of cycloid psychosis in the modern classifications have consistently shown that cycloid psychosis does not correspond closely to any DSM III, DSM III-R, DSM IV, or ICD-10 category [14,2023]. In fact, most cases of cycloid psychosis are diagnosed by these systems as BPD, schizophreniform disorder, schizoaffective disorder, mood disorder with psychotic features, or psychotic disorder not otherwise specified. Studies examining the concordance between the ATPDs and cycloid psychosis have found that only 30% to 54% of patients with an ICD-10 diagnosis of ATPD met the criteria for cycloid psychosis [14,24,25]. Accordingly, identity of the two disorders cannot be taken for granted, and the attempts to include aspects of cycloid psychosis in the consensus classifications under

Subtypes of Cycloid Psychosis

According to Leonhard [1,2], cycloid psychoses need to be diagnosed on the basis of the presence of at least one of six basic bipolar axial syndromes, namely anxiety-happiness, excitation-inhibition-confusion, and hypermotility-hypomotility, which respectively conform the classical subtypes of anxiety-happiness, confusional, and motility psychoses. Each subform may present intrasyndromal bipolarity (ie, hypomotility and hypermotility features) and symptoms from other axial syndromes, all of which confer a typical polymorphous character to the whole clinical picture. Leonhard described these subforms as prototypal variants of a single disorder. However, in practice, pure forms corresponding to each subtype seldom occur, and most frequently, various cycloid syndromes may appear during the same episode and/or combine and merge together. Leonhard reported some differences in subtypes of cycloid psychoses regarding gender distribution (female preponderance in motility and confusional psychoses), age at onset (later in anxiety-happiness psychosis), number of episodes (higher recurrence in motility psychosis), and episode length (higher duration in anxiety-happiness psychosis). However, the subtypes did not show a distinctive familial pattern. No further attempts to validate this subtyping procedure have been done; thus, their validity remains uncertain.

Table 1. Differences between Leonhard and PB diagnostic criteria of CP*

P 0.14 0.199 0.722 0.889 0.723 0.68 4.75 14.1 19.33 1.34 0.19 4.74 0.508 0.01 0.001 0 0.51 0.508 0.01 B, C > A A>C A > B, C C, B > A Comparison among groups

CP according to Leonhard but CP according to both CP according to PB but not Leonhard For chi-squared not PB criteria (Group A, n = 69) criteria (Group B, n = 51) criteria (Group C, n = 17) (df = 2) 35.9 (15.7) 39 (56.5) 0.02 (0.55) 0.08 (0.69) 74.0 (19.3) 4.70 (2.74) 30.2 (12.5) 56 (81.2) 17 (24.6) 19 (27.5) 3.35 (1.35) 1.12 (0.47) 1.39 (0.66) 3.41 (1.33) 14 (27.5) 7 (41.2) 3.18 (1.23) 1.26 (0.58) 31 (60.8) 11 (64.7) 51 (100) 17 (100) 26.8 (8.44) 22.0 (3.82) 4.14 (2.79) 4.24 (2.10) 76.4 (20.4) 77.2 (15.7) 0.32 0.12 (0.64) 0.15 (0.65) 0.11 0.09 (0.36) 0.07 (0.38) 0.32 21 (41.2) 10 (58.8) 3.22 32.3 (9.0) 30.2 (9.2) 1.99

Demographics

Age, y

Male gender, n (%)

Familial liability to

Schizophrenia

Major mood disorders

Premorbid factors

Early familial dysfunction

Premorbid adjustment

Clinical variables

Age at onset, y

Acute onset (< 1 week), n (%)

Lifetime mood syndrome, n (%)

Comorbid drug abuse, n (%)

Psychosocial stressors

Treatment response (index episode) 2.92 (1.34) 2.06 (1.20) 1.21 (1.16) 0.79 (1.54) 0.90 (1.62) 1.62 (2.53) 1.60 (2.32) 1.86 (2.36) 1.98 (2.76) 1.77 (1.55) 2.20 (1.02) 3.30 (1.12)

Index episode syndromes 3.53 (1.06) 2.05 (1.37) 2.01 (1.57) 2.23 (2.53) 1.23 (2.01) 1.29 (2.66) 2.39 0.22 3.7 4.63 3.48 0.51 0.096 0.802 0.027 0.011 0.033 0.598 C>A C>A B>A

Reality distortion

Disorganization

Negative

Depressive

Mania

Cycloid Psychosis: An Examination of the Validity of the Concept

Catatonia

*All values are mean (SD) unless otherwise specied. CPcycloid psychosis; dfdegree of freedom; PBPerris and Brockington.

Peralta et al. 187

188

Table 1. Differences between Leonhard and PB diagnostic criteria of CP* (Continued)

P Comparison among groups C > B, A A < C, B C, B > A 0.004 0.458 A, B > C

CP according to Leonhard but CP according to both CP according to PB but not Leonhard For chi-squared not PB criteria (Group A, n = 69) criteria (Group B, n = 51) criteria (Group C, n = 17) (df = 2) 0 (0) 24 (34.8) 6 (8.7) 1 (1.4) 32 (46.4) 6 (8.7) 4 (7.8) 0 (0) 1.56 15 (29.4) 1 (5.9) 10.95 13 (25.5) 6 (35.3) 20.26 0 8 (15.7) 2 (11.8) 1.38 8 (15.7) 3 (17.6) 6.25 0.044 0.499 3 (5.9) 5 (29.4) 21.45 0

DSM IV diagnoses, n (%)

Schizophrenia

Schizophreniform disorder

Schizoaffective disorder

Major mood disorders

Brief psychotic disorder

Atypical psychotic disorder

Nonschizophrenic Psychotic Disorders

*All values are mean (SD) unless otherwise specied. CPcycloid psychosis; dfdegree of freedom; PBPerris and Brockington.

Cycloid Psychosis: An Examination of the Validity of the Concept

Peralta et al.

189

the diagnoses of BPD or ATPD may be considered as unsuccessful. In fact, the DSM IV and ICD-10 diagnoses of brief or transient psychotic disorders should be dismissed as proxy diagnoses for cycloid psychosis because of their conceptual bias and limited validity. Reasons for the poor concordance between BPD and ATPD on the one side and cycloid psychosis on the other are that formal diagnostic systems require the exclusion of major mood syndromes and a duration no longer than 1 to 3 months, depending on the specific disorder. The duration of a cycloid episode is on average 3 months and in a minority of patients may exceed several months [2]. In line with this, it has been reported that the modal duration of the acute remitting psychoses, when the duration criterion is relaxed, is 2 to 4 months [26]. Both the relatively long duration of some cycloid episodes and the ability of this diagnosis to predict a good outcome, this irrespective of its length, clearly suggest that cycloid psychosis does not equal with brief duration. Accordingly, if formal diagnostic systems have to incorporate the cycloid psychosis concept adequately, the duration criterion should be either relaxed (some authors [27] propose to extend the duration criterion to 6 months) or not included at all [28].

considered, and neither is the complex and varied temporal relationship between the schizophrenic and affective domains. As a result, the schizoaffective diagnosis as currently defined represents a rare condition within psychotic disorders with poor reliability and validity [30].

The Nosological Status of Cycloid Psychosis

The intriguing character of cycloid psychosis, a disorder comparable to schizophrenia in terms of psychotic symptoms and to mood disorders in terms of course and outcome, together with the broad array of illness-related etiological and pathophysiologic factors [30] raises the question about its very true nature. More specifically, its nosological status within the broad field of psychotic disorders, and particularly in relation to mood disorders and schizophrenia, continues to be a matter of debate. Studies addressing this question have supported the view of cycloid psychosis as an independent nosological entity [1,3,20,31], as an atypical variant of affective psychosis [32], as an atypical form of schizophrenia [33], and as a heterogeneous condition [14,16,17]. Finally, other studies have yielded inconclusive results [34,35]. From a nosological perspective, the key question regarding the validity of the cycloid psychosis diagnosis is whether it defines a syndrome that can be meaningfully differentiated from schizophrenia on the one hand and atypical affective illness on the other. Accordingly, five nosological hypotheses can be articulated that can be usefully evaluated from the existing data. These hypotheses consider cycloid psychosis to be the following: 1) a form of schizophrenia, 2) a form of affective illness, 3) a heterogeneous condition, 4) a distinct nosological entity, or 5) an intermediate form between schizophrenia and mood disorders (schizoaffective disorder). Elsewhere [28], we have comprehensively reviewed the existing data on sociodemographic, etiological, clinical, psychopathological, and course features of cycloid psychosis. Here, we put them in relation to the five nosological hypotheses mentioned previously, taking into account the degree of the evidence supporting each hypothesis (Table 2). It is evident from inspecting Table 2 both that there are no compelling data supporting a particular hypothesis and that the most validated hypotheses are the independent and mood disorder ones. Evidence for the mood disorder hypothesis comes from the high prevalence of mood symptoms and syndromes and the recurring-remitting pattern. Familial data also support this hypothesis, as most but not all [36] studies resulted in a familial liability pattern that is closer to mood disorders than to schizophrenia. The main support for the independent hypothesis comes from the distinctive symptom pattern. Additional evidence is genetic, namely the relatively low heritability regarding schizophrenia and mood disorders and homotypia. Given that each hypothesis has received some support from the existing

The Relationship Between Cycloid and Schizoaffective Psychoses

Kleist [29] already noted that approximately one half of the patients with motility psychosis may present with a full affective syndrome within or between episodes and considered that the disorder would correspond to the circular cases of the manic depressive illness. On the other hand, there exists a notorious phenomenological similarity between the anxiety-happiness psychosis and manic depressive illness. Interestingly, the term schizoaffective psychosis was coined by Kasanin [6] in 1933 to describe a clinical picture very similar to that of cycloid psychosis in which affective symptoms were present, although few of his patients would have met modern criteria for schizoaffective disorder. Given that more than two thirds of cycloid patients may present with lifetime mood symptoms or syndromes [14], it could be reasoned that a condition with such a high prevalence of mood symptoms would be closely related to the current concept of schizoaffective disorder. However, all the studies examining the relationship between schizoaffective and cycloid disorders revealed a poor concordance between the two diagnoses [14,20,21,23]. The central feature of schizoaffective disorder as defined in DSM IV and ICD-10 is the coexistence of schizophrenic symptoms and a major mood syndrome. This represents a very restrictive view of affectivity in that in other affective domains such as anxious and ecstatic mood, rapid and brief changes between affective states and mood swings not fulfilling intensity and duration criteria for a major mood syndrome are not

190

Nonschizophrenic Psychotic Disorders

Table 2. Characteristics of cycloid psychosis as supporting 5 alternative nosological hypotheses

Nosological hypotheses Form of SZ Demographics Age at onset Male/female ratio Genetic factors Familial risk of MD and SZ Homotypia Heritability (relatively low) Premorbid features Premorbid functioning (good) Personality (histrionic) Triggers Hormonal Psychosocial Symptoms First-rank symptoms Symptom polymorphism Mood symptoms/syndromes Type of treatment Mood stabilizers Antipsychotics Benzodiazepines Electroconvulsive therapy Course Diagnostic stability Recurrence Outcome (good) X NE NE NE XX XX NE NE NE XX NE XX NE NE NE NE XX NE NE XX NE NE XX NE NE NE X NE NE X XX XX XX NE X NE XX NE X XX NE NE XX NE XXX NE XX X XX NE NE XX XX XX XX XX XX X X NE NE XX NE NE X X X NE NE NE NE NE XX X X NE X XX XX NE NE NE NE NE XX X X X XX X X X Form of MD Heterogeneous disorder Independent disorder Schizoaffective (third psychosis) disorder

MDmood disorder; NEno evidence; SZschizophrenia; Xsome but insufcient evidence; XXsome and consistent evidence; XXXcompelling evidence.

data, none of them should be fully disregarded. In fact, all the hypotheses are compatible with the consideration of cycloid psychosis on a psychotic continuum [37,38] between the prototypical disorders of schizophrenia and major mood disorders, although it must be acknowledged that in many respects, cycloid psychosis is closer to the affective pole than to the schizophrenic one.

Clinical Validity
A key issue in the validating process of any psychiatric disorder is to acknowledge that different validators (ie, genetic, clinical, outcome) may define different populations of patients or, alternatively, that a specific syndrome may be related to a set of validators but not to others [39]. That means that clinical and etiologi-

cal validity may not converge. For example, DSM IV schizophrenia is particularly useful for predicting outcome (clinical validity), largely because some degree of chronicity is in-built. But a much broader definition embracing all nonaffective psychotic disorders is more useful for defining a syndrome with high heritability (etiological validity). These considerations may help to explain the fact that the validity of cycloid psychosis may vary as a function of the validators considered and that no single nosological hypothesis has received compelling support. The cycloid psychosis construct appears to have historical, face, and empirical validity in that it can be easily differentiated from neighboring syndromes on psychopathological and outcome grounds; thus, the cycloid psychosis diagnosis appears to have clinical

Cycloid Psychosis: An Examination of the Validity of the Concept

Peralta et al.

191

validity and utility [1,4,13,14,40]. Although available data do not support a specific etiology and pathophysiology for cycloid psychosis, the clinical and heuristic value of the concept is undeniable, as it defines a discrete clinical syndrome with a relatively characteristic symptom pattern predicting good interepisode recovery in the vast majority of the cases. The main reason for recognizing the cycloid psychosis concept is that patients presenting with cycloid symptoms may be erroneously diagnosed with schizophrenia, which may convey dramatic therapeutic and prognostic implications. All these features reflect well the clinical and research relevance of a concept that deserves to be included, on its own, in future formal diagnostic criteria of psychotic disorders, at least as a provisional or alternative diagnosis [40], in order to promote a more complete examination of its clinical and biological underpinnings.

Conclusions
Cycloid psychosis is a diagnostic concept that has evolved gradually in the last 120 years in Europe. It embraces a group of acute, recurrent, and benign psychotic disorders whose symptomatology is neither typically schizophrenic nor typically affective, but rather polymorphic. The disorder does not fit with the current conceptualization of schizoaffective disorder. Despite the inclusion of some cycloid features in the DSM IV and ICD-10 systems under the diagnoses of brief or transient psychotic disorders, these diagnoses should be dismissed as proxy concepts of cycloid psychosis because of both conceptual bias and limited validity. Whereas Leonhard set the basis for the current conceptualization of the disorder, Perris and Brockington developed the first operational criteria for diagnosing the disorder, which are the most used to date. However, the two conceptualizations are not the same and differ across a number of meaningful variables. From a nosological perspective and depending on the type of validators considered, available data support the consideration of cycloid psychosis either as an atypical form of mood disorders or as an independent entity. Although the clinical, prognostic, and heuristic value of the concept of cycloid psychosis is undeniable, the neurobiological basis of the disorder remains to be established.

Future Directions
Despite the long tradition of cycloid psychosis diagnosis, many questions remain about its validity. First, given that the Perris and Brockington criteria for cycloid psychosis do not fit completely with the original Leonhard conceptualization, there is a need for studies addressing the comparative validity of the two concepts. Whereas it can be reasonably hypothesized that cycloid psychosis is a primary neurotransmitter disorder with relapsing and remitting disturbances [16], the neurobiological substrate is virtually unknown, as the few existing studies have yielded contradictory findings [28]. Future studies should pay particular attention to examining the putative subtypes (ie, Leonhards classical subtypes and the affective-nonaffective distinction) in relation to clinical, etiological, and pathophysiologic variables. The few studies that have examined the stability of the diagnosis across episodes of the illness have found it to be of adequate standard; however, these studies are subjected to a number of limitations, including their mainly retrospective character. Given the importance of this aspect of the validity, the stability of the diagnosis should be addressed using a first-episode, prospective approach to the diagnosis. Treatment of cycloid psychosis with antipsychotics, mood stabilizers, benzodiazepines, or electroconvulsive therapy has an empirical basis, but no single randomized trial comparing efficacy among interventions or against placebo has been published to date. The comparison against placebo should be particularly insightful given that an undetermined number of patients may have self-remitting episodes. Antipsychotic drugs are a mainstay of therapy, and although atypical drugs appear to exert a particularly beneficial effect in patients with cycloid psychosis [41], more studies are warranted.

Acknowledgments
None of the authors has a possible conflict of interest, financial or otherwise.

References and Recommended Reading

Papers of particular interest, published recently, have been highlighted as: Of importance Of major importance
1. 2. 3. 4. 5. Leonhard K: Cycloid psychoses-endogenous psychoses which are neither schizophrenic nor manic-depressive. J Ment Sci 1961, 107: 632648. Leonhard K: The Classification of Endogenous Psychoses, edn 5. New York: Irvington Publishers; 1979. Perris C: A study of cycloid psychoses. Acta Psychiatr Scand Suppl 1974, 253:177. Perris C: The concept of cycloid psychotic disorder. Psychiatr Dev 1988, 6: 3756. Perris C, Brockington IF: Cycloid psychoses and their relation to major psychosis. In Biological Psychiatry. Edited by Perris C, Struwe G, Jansson B. Amsterdam: Elsevier; 1981:447450. Kasanin J: The acute schizo-affective psychoses. Am J Psychiatry 1933, 13:97126. Langfeldt G: The Schizophreniform States. Copenhagen: E. Munksgaard; 1939. McCabe MS: Reactive psychoses, a clinical and genetic investigation. Acta Psychiatr Scand Suppl 1975, 259:1133. Pichot P: The diagnosis and classification of mental disorders in French-speaking countries: background, current views and comparison with other nomenclatures. Psychol Med 1982, 12:475492.

6. 7. 8. 9.

192
10.

Nonschizophrenic Psychotic Disorders

Mojtabai R, Susser ER, Bromet EJ: Clinical characteristics, 4-year course, and DSM-IV classification of patients with nonaffective acute remitting psychosis. Am J Psychiatry 2003, 160: 21082115. 28. Peralta V, Cuesta MJ: Cycloid psychosis. Int J Psychiatry 2005, 17:5362. A general review of etiology, psychopathology, clinical features, and nosology of cycloid psychosis. 29. Kleist K: On the clinical position of motility psychoses. In Catatonia: An Anthology of Classical Contributions. Edited by Ungvari GS. Hong Kong: Scientific Communications International; 2006:7680. 30. Maier W: Do schizoaffective disorder exists at all? Acta Psychiatr Scand 2006, 13:369371. This is an informative essay about the limited reliability and validity of schizoaffective disorder as defined in DSM IV and ICD-10. 31. Maj M: Cycloid psychotic disorder: validation of the concept by means of follow-up and family study. Psychopathology 1990, 23:196204. 32. Cutting JC: Relationship between cycloid psychosis and typical affective psychosis. Psychopathology 1990, 23: 212219. 33. Vaillant GE: An historical view of remitting schizophrenias. J Nerv Ment Dis 1964, 138:4856. 34. Cutting JC, Clare AW, Mann A: Cycloid psychosis: an investigation of the concept. Psychol Med 1978, 8: 637648. 35. Brockington IF, Perris C, Kendell RE, et al.: The course and outcome of cycloid psychosis. Psychol Med 1982, 12:97105. 36. Pfuhlmann B, Jabs B, Althaus G, et al.: Cycloid psychoses are not part of a bipolar affective spectrum: results of a controlled family study. J Affect Disord 2004, 83:1119. This is a controlled family study that supports the nosological independence of cycloid psychosis from bipolar disorder. 37. Crow TJ: The continuum of psychosis and its implication for the structure of the gene. Br J Psychiatry 1986, 149: 419429. 38. Peralta V, Cuesta MJ: The underlying structure of diagnostic systems of schizophrenia: a comprehensive polydiagnostic approach. Schizophr Res 2005, 79: 217229. This study shows how psychotic disorders extend along a continuum of risk factors, clinical features, and outcome. 39. Kendell RE: The choice of diagnostic criteria for biological research. Arch Gen Psychiatry 1982, 39:13341339. 40. Brockington IF, Perris C, Meltzer HY: Cycloid psychoses. Diagnosis and heuristic value. J Nerv Ment Dis 1982, 170: 651656. 41. Ehlis AC, Zielasek J, Herrmann MJ, et al.: Beneficial effect of atypical antipsychotics on prefrontal brain function in acute psychotic disorders. Eur Arch Psychiatry Clin Neurosci 2005, 255: 299307. This study reported a strong association between dose of atypical antipsychotics (but not of typical antipsychotic drugs) and a measure of prefrontal response control. The authors conclude that in cycloid psychosis, atypical antipsychotics may exert a beneficial effect on prefrontal brain function. 27.

Mitsuda H: The concept of atypical psychoses-from the aspect of clinical genetics. Acta Psychiatr Scand 1965, 41:372377. 11. McNeil TF: A prospective study of postpartum psychoses in a high-risk group. 1. Clinical characteristics of the current postpartum episodes. Acta Psychiatr Scand 1986, 74: 205216. 12. Sigmund D, Mundt C: The cycloid type and its differentiation from core schizophrenia: a phenomenological approach. Compr Psychiatry 1999, 40:418. 13. Jabs BE, Puhlmann B, Bartsch AJ, et al.: Cycloid psychoses: from clinical concepts to biological foundations. J Neural Transm 2003, 109:907919. 14. Peralta V, Cuesta MJ: Cycloid psychosis: a clinical and nosological study. Psychol Med 2003, 33:443453. 15. Peralta V, Cuesta MJ, Zandio M, et al.: The cycloid psychoses: a comparison between Leonhard and Perris diagnostic systems. Paper presented at the 2nd International Congress of the World Federation of Societies of Biological Psychiatry. Santiago de Chile, Chile; April 1721, 2007. 16. Fish F: The cycloid psychoses. Compr Psychiatry 1964, 5:155169. 17. Mojtabai R: Heterogeneity of cycloid psychosis: a latent class analysis. Psychol Med 2000, 30:721726. 18. Susser E, Varma VK, Malhotra S, et al.: Delineation of acute and transient psychotic disorders in a developing country setting. Br J Psychiatry 1995, 167: 216219. 19. Das SK, Malhotra S, Basu D: Family study of acute and transient psychotic disorders: comparison with schizophrenia. Soc Psychiatry Psychiatr Epidemiol 1999, 34:328332. 20. Maj M: Clinical course and outcome of cycloid psychotic disorder: a three-year prospective study. Acta Psychiatr Scand 1988, 78:182187. 21. Brockington IF, Roper A, Buckley M, et al.: Bipolar disorder, cycloid psychosis and schizophrenia: a study using lifetime psychopathology ratings, factor analysis and canonical variate analysis. Eur Psychiatry 1991, 6: 223236. 22. Beckmann H, Fritze J, Lanczik M: Prognostic validity of the cycloid psychoses. Psychopathology 1990, 23: 205211. 23. McGorry PD, Singh B, Connell S, et al.: Diagnostic concordance in functional psychosis revisited: a study of inter-relationships between alternative concepts of psychotic disorder. Psychol Med 1992, 22:367378. 24. Pillmann F, Haring A, Balzuweit S, et al.: Concordance of acute and transient psychoses and cycloid psychoses. Psychopathology 2001, 34:305311. 25. Marneros A, Pillmann F: Acute and Transient Psychoses. Cambridge: Cambridge University Press; 2004. This book contains an extensive clinical study on acute remitting psychoses, together with the most comprehensive review on the topic. The authors conclude that ATPDs and their core group of brief, polymorphous psychoses are not a separate clinical entity. 26. Mojtabai R, Varma VK, Susser E: Duration of remitting psychoses with acute onset. Br J Psychiatry 2000, 176:576580.