Introduction
Although the Kraepelinian dichotomy of dementia praecox (later renamed schizophrenia, SZ) and manic depression (Kraepelin, Reference Kraepelin1896) may seem outdated, it remains central to current psychiatric classifications. Efforts to delineate ‘intermediate’ conditions have long been made in different psychiatric traditions, but the evidence drawn on proved inconclusive. This has caused difficulties in classification and nomenclature particularly for transient psychotic disorders that do not involve prominent affective features.
The ICD-10 Classification of Mental and Behavioural Disorders (WHO, 1992) subsumed earlier concepts such as bouffée délirante (Magnan, Reference Magnan1887), cycloid psychosis (Kleist, Reference Kleist1928; Leonhard, Reference Leonhard and Beckmann1999), reactive psychosis (Wimmer, Reference Wimmer1916) and schizophreniform psychosis (Langfeldt, Reference Langfeldt1939) under the heading of ‘acute and transient psychotic disorders’ (ATPDs). Influenced by findings of the World Health Organization (WHO) study on acute psychoses (Cooper et al. Reference Cooper, Jablensky, Sartorius, Stefanis, Rabavilas and Soldatos1999), ATPDs were characterized by: (a) acute onset (within 2 weeks), (b) polymorphic, schizophrenic or predominantly delusional symptoms and (c) association (or not) with psychological stress. Early remission within 1 or 3 months distinguishes the ATPD subtypes with schizophrenic symptoms from schizophrenia and those with polymorphic or delusional features from persistent delusional disorder. Although emotional changes and affective symptoms may occur, the diagnostic criteria for ATPDs exclude both depressive and manic episodes.
Developed in parallel with ICD-10, the Diagnostic and Statistical Manual of the American Psychiatric Association (APA) has, since its fourth edition (DSM-IV, APA, 1994), listed the category ‘brief psychotic disorder’ (BPD), featuring delusions, hallucinations, disorganized speech, and grossly disorganized or catatonic behaviour that lasts less than 1 month. DSM-IV also included ‘schizophreniform disorder’, that is typical SZ with duration intermediate between BPD and SZ. Because of differences in onset, duration and symptomatology, ATPDs overlap only partially with the DSM-IV categories BPD and schizophreniform disorder (Nugent et al. Reference Nugent, Paksarian and Mojtabai2011).
Case identification and follow-up studies may be difficult owing to the heterogeneity of the clinical features encompassed by ATPDs, and the varying rates of diagnosis change in subsequent episodes mainly to SZ or affective disorders (Jørgensen et al. Reference Jørgensen, Bennedsen, Christensen and Hyllested1997; Singh et al. Reference Singh, Burns, Amin, Jones and Harrison2004; Pillmann & Marneros, Reference Pillmann and Marneros2005; Aadamsoo et al. Reference Aadamsoo, Saluveer, Küünarpuu, Vasar and Maron2011; Salvatore et al. Reference Salvatore, Baldessarini, Tohen, Khalsa, Sanchez-Toledo, Zarate, Vieta and Maggini2011; Castagnini et al. Reference Castagnini, Foldager and Bertelsen2013). Preliminary findings also suggest that first-degree relatives (FDRs) of patients with ATPDs have a greater risk of ATPDs than family members of patients affected with SZ (Das et al. Reference Das, Malhotra and Basu1999), and that family predisposition may produce its effects by increasing sensitivity to life events (Das et al. Reference Das, Malhotra, Basu and Malhotra2001). Whether ATPDs belong to the low vulnerability end of the schizophrenic spectrum, link SZ and affective disorders, or constitute a distinct category deserves clarification. The aim of this study was to address the relationship between ATPDs and SZ and bipolar disorder (BD) by examining psychiatric morbidity in FDRs.
Method
This study drew its data from the Civil Registration System (CRS) and the Danish Psychiatric Central Register (DPCR). The CRS (Pedersen et al. Reference Perdersen, Gotzsche, Moller and Mortensen2006) records data from all those living in Denmark including date of birth (and death), sex, marital status, parents and siblings. The DPCR (Mors et al. Reference Mors, Perto and Mortensen2011) has stored information on all in-patients since 1969 and on out-patients since 1995. It provides data for national statistics, mental health planning, and research. Patient entries include information from case-notes and, since 1994, diagnosis according to ICD-10 Diagnostic Criteria for Research (DCR-10). The quality of psychiatric assessment is enhanced by uniformity of training throughout the country, including supervision and courses in ICD-10 and related diagnostic instruments. Because its differences from ICD-8 (WHO, 1971) were minor, ICD-9 was never used for official nosography. The Danish National Health Service provides free treatment for all residents, and no private psychiatric facilities exist.
A cohort of 2526352 persons born in Denmark between 1 January 1955 and 1 September 1993 was selected from the CRS. By linking them to the DPCR, we identified all patients, and their FDRs (parents and siblings), aged over 15 years who had ever been admitted to hospital or treated in out-patient services with a diagnosis of ATPDs, SZ or BD from January 1996 to September 2008.
Case register diagnoses
The diagnostic categories to be selected were the following: ATPDs, i.e. ICD-10 F23 or ICD-8 298 (other non-organic psychoses: 298.0 reactive depressive psychosis, 298.1 reactive excitation, 298.2 reactive confusion, 298.3 acute paranoid reaction and 298.8 reactive psychosis unspecified) and ICD-8 295.4 (acute schizophrenic episode); SZ, i.e. ICD-10 F20 or ICD-8 295 (except 295.4); BD, i.e. ICD-10 F30 (manic disorder) and F31 (bipolar disorder), or ICD-8 296.1 and 296.3 (manic depressive psychosis: manic and circular type); any ICD-8 and ICD-10 diagnosis.
The ATPD category comprises six different subtypes coded by a 4-letter/digit combination: F23.0 ‘acute polymorphic psychotic disorder without schizophrenic symptoms’ (APPD); F23.1 ‘acute polymorphic psychotic disorder with schizophrenic symptoms; F23.2 ‘acute schizophrenia-like psychotic disorder’; F23.3 ‘acute predominantly delusional disorder’; F23.8 ‘other’ and F23.9 ‘unspecified’ ATPDs for cases that cannot be classified elsewhere. Both subtypes with schizophrenic features have a temporal criterion of less than 1 month and were taken as akin to the same diagnostic group. Because ATPD diagnosis may sometimes be provisional, those patients who changed diagnosis in subsequent hospital admissions or out-patient treatments by the first year were withdrawn from this study.
Statistical analysis
The relative risk (RR) with 95% confidence intervals (CIs) of ATPDs, SZ and BD was calculated as the incidence rate ratio using Poisson regression with the logarithm of the person-years at risk as offset (Laird & Olivier, Reference Laird and Olivier1981), and adjusted by age, sex and calendar period. For each category, we determined the risk associated with FDRs having any ICD-8 and ICD-10 diagnosis and the additional risk if they were admitted with ATPDs, SZ or BD.
The attributable risk (Bruzzi et al. Reference Bruzzi, Green, Byar, Brinton and Schairer1985) was used to measure the fraction of cases with ATPDs, SZ or BD resulting from the effect of family psychiatric morbidity in the population.
Ethical issues
This study was approved by the Danish Data Protection Agency. Data comply with appropriate protection standards for anonymity.
Results
From a cohort of 2526352 persons at risk (a total of 26742239 person-years), there were 2537 cases with ATPDs, 10639 with SZ and 5292 with BD listed in the DPCR between 1996 and 2008.
The RR of ATPDs (1.93, 95% CI 1.76–2.11) was nearly twice as high in patients with ATPDs who had at least one FDR admitted with any mental disorder than patients without family psychiatric antecedents, both for males (RR 1.87, 95% CI 1.65–2.11) and for females (RR 2.00, 95% CI 1.74–2.29). An additional risk arose if they had FDRs admitted not only with ATPDs (RR 1.60, 95% CI 1.33–1.92) but also with BD (RR 1.55, 95% CI 1.23–1.96) or SZ (RR 2.06, 95% CI 1.70–2.50) (Table 1). The attributable risk due to family psychiatric morbidity was 20.3% of all cases of ATPDs, and 4.0% resulting from family history of ATPDs.
Table 1. Relative risk (RR)a of acute and transient psychotic disorders (ATPDs), bipolar disorder (BD) and schizophrenia (SZ) associated with family psychiatric morbidity
CI, Confidence interval.
a Adjusted by age, calendar period, unknown mother or father, and family history of ATPDs, BD and SZ. Seventy-one cases with ATPDs, 180 with BD and 127 with SZ had unknown family psychiatric history.
Analysis of psychiatric morbidity by family members revealed that the greatest risk of ATPDs resulted from BD in the father (RR 2.15, 95% CI 1.50–3.08) and SZ in the mother (RR 2.06, 95% CI 1.48–2.87) and/or father (RR 2.72, 95% CI 1.86–3.98) (Table 2).
Table 2. Relative riska (95% confidence interval) of acute and transient psychotic disorders (ATPDs), bipolar disorder (BD) and schizophrenia (SZ) associated with psychiatric morbidity by family member
a Adjusted by age, calendar period, unknown mother or father, and family history of ATPDs, BD and SZ.
With regard to the ATPD subtypes, the risk of APPD was high if patients had FDRs admitted with either SZ (RR 1.83, 95% CI 1.09–3.06) or ATPDs (RR 1.58, 95% CI 1.00–2.49); SZ in family members had the greatest effect on those featuring schizophrenia-like symptoms (RR 2.52, 95% CI 1.62–3.94), and BD did not increase significantly the risk for any specific subtype (Table 3).
Table 3. Relative riska (95% confidence interval) of acute and transient psychotic disorders (ATPDs F23) associated with family psychiatric morbidity by subtype
a Adjusted by age, calendar period, unknown mother or father, and family history of ATPDs, bipolar disorder (BD) and schizophrenia (SZ).
F23.0 Acute polymorphic psychotic disorder without symptoms of SZ; F23.1 acute polymorphic psychotic disorder with symptoms of SZ; F23.2 acute SZ-like psychotic disorder; F23.3 acute predominantly delusional disorder; F23.8 other and F23.9 unspecified acute and transient psychotic disorders.
The risk of SZ and BD was significantly higher in patients with SZ and BD having FDRs admitted with any mental disorder than those without family psychiatric antecedents (Table 1). Despite some overlap, the additional risk of SZ (RR 2.80, 95% CI 2.58–3.04) and BD (RR 3.68, 95% CI 3.29–4.12) was markedly greater if patients with SZ and BD had FDRs admitted with that condition. By contrast, ATPDs in FDRs of patients with SZ and BD effected only a moderate increase of risk for both disorders.
Furthermore, SZ and BD in both parents and siblings were the principal risk factors for the same disorder (Table 2). The attributable risk was 27.0% of all cases with SZ and 25.2% of those with BD; SZ accounted for 5.6% and BD for 6.5%.
Discussion
To our knowledge, this is the first large-scale, population-based study dealing with the family psychiatric morbidity of ATPDs, and it marks a clear advance from earlier research in several ways. First, it is representative of the ATPD category, including polymorphic, schizophrenia-like and delusional subtypes. Second, comparing ATPDs with SZ and BD makes it possible to identify specific risk factors for these conditions. Third, it enhances the understanding of ATPDs and contributes to building a case for their revision in future psychiatric classifications.
The findings of the present study suggest that the risk of ATPDs is significantly high in patients with ATPDs having FDRs admitted with any mental disorder or with ATPDs, but the main additional effect on ATPDs resulted from a family history of SZ. This is probably because ATPDs constitute a composite category comprising subtypes with schizophrenic features, which are set apart from SZ only by temporal criteria. Available evidence indicates that these subtypes not only have increased rates of SZ in family members but also are characterized by a male preponderance, an earlier age of onset and a tendency to evolve into SZ and related disorders (Singh et al. Reference Singh, Burns, Amin, Jones and Harrison2004; Chang et al. Reference Chang, Pang, Chung and Chan2009; Aadamsoo et al. Reference Aadamsoo, Saluveer, Küünarpuu, Vasar and Maron2011; Salvatore et al. Reference Salvatore, Baldessarini, Tohen, Khalsa, Sanchez-Toledo, Zarate, Vieta and Maggini2011; Castagnini et al. Reference Castagnini, Foldager and Bertelsen2013).
Less clear is the relationship between ATPDs and BD. Previously, clinical and epidemiological differences from SZ and changes of diagnosis to BD in subsequent episodes have been interpreted as indicating that the APPD subtype has a closer kinship to affective disorders (Jørgensen et al. Reference Jørgensen, Bennedsen, Christensen and Hyllested1997; Sajith et al. 2002; Pillmann & Marneros, Reference Pillmann and Marneros2005; Linden et al. Reference Linden, Harris, Whitaker and Healy2009; Salvatore et al. Reference Salvatore, Baldessarini, Tohen, Khalsa, Sanchez-Toledo, Zarate, Vieta and Maggini2011). Reminiscent of the French concept of bouffée délirante (Magnan, Reference Magnan1887) and cycloid psychosis (Kleist, Reference Kleist1928; Leonhard, Reference Leonhard and Beckmann1999), APPD features typically varied delusions, hallucinations, perceptual changes, agitation, perplexity and emotional turmoil shifting from day to day or even from hour to hour. However, our findings fail to support any relationship between the two categories, as BD did not significantly increase the risk for any specific ATPD subtype. The risk of APPD was similar in patients who had FDRs admitted with either ATPDs or SZ, and it seems more likely that APPD has an intermediate position between the subtypes with schizophrenic symptoms and acute predominantly delusional disorder; the latter being associated only with a family history of ATPDs.
There are few studies with which to make meaningful comparisons. Das et al. (Reference Das, Malhotra and Basu1999) reported that FDRs of patients with ATPDs have higher rates of ATPDs than family members of schizophrenic patients; yet the risk of SZ is significantly high in the family of patients with SZ. Evidence also suggests that ATPDs are associated neither with brain changes nor with pre-morbid dysfunctions, and that they may be distinct from SZ in course and outcome, although their diagnostic stability is low (Jørgensen et al. Reference Jørgensen, Bennedsen, Christensen and Hyllested1997; Singh et al. Reference Singh, Burns, Amin, Jones and Harrison2004; Pillmann & Marneros, Reference Pillmann and Marneros2005; Chang et al. Reference Chang, Pang, Chung and Chan2009; Linden et al. Reference Linden, Harris, Whitaker and Healy2009; Salvatore et al. Reference Salvatore, Baldessarini, Tohen, Khalsa, Sanchez-Toledo, Zarate, Vieta and Maggini2011; Castagnini et al. Reference Castagnini, Foldager and Bertelsen2013). Related conditions such as the ‘non-affective acute remitting psychoses’ have shown a characteristic geographic distribution, an association with stress, fever or systemic infection, and low rates of relapse particularly in developing countries, but seldom fulfil the temporal criteria for ATPDs (Susser & Wanderling, Reference Susser and Wanderling1994; Mojtabai et al. Reference Mojtabai, Varma and Susser2000).
In addition, while the cognitive impairment involved in insidious-onset schizophrenia would be associated with negative symptoms and poor outcome, short-lived psychotic disorders seem to be mediated by emotion-driven pathways (van Os et al. Reference van Os, Kenis and Rutten2010). This raises the question of whether environmental factors play a role in the causation of ATPDs. In keeping with the stress–vulnerability model, it would seem that family predisposition manifests itself with an increased susceptibility that renders patients with ATPDs less likely to cope with adverse events (Das et al. Reference Das, Malhotra, Basu and Malhotra2001). It will be rewarding for future research to focus on the interaction between predisposing factors and environmental exposures (i.e. early loss, social adversities, etc.), and how they influence emotional reactivity.
The findings of this study also indicate that SZ and BD are more likely than ATPDs to involve a family predisposition, at least partly attributable to the effect of each disorder. This lends support to previous surveys showing that FDRs of patients with SZ and BD have an increased risk of both conditions, an effect that may result from common genetic factors (Laursen et al. Reference Laursen, Labouriau, Licht, Bertelsen, Munk-Olsen and Mortensen2005; Lichtenstein et al. Reference Lichtenstein, Yip, Björk, Pawitan, Cannon, Sullivan and Hultman2009). Further evidence, challenging a sharp demarcation between the two categories, has come from research in offspring of dual matings, that is couples with one parent affected with SZ and the other with BD (Gottesman et al. Reference Gottesman, Laursen, Bertelsen and Mortensen2010), and molecular genetics studies. Although meta-analysis of genome-wide association data reveals a weak effect, which accounts just for part of the genetic risk of SZ and/or BD (Gershon et al. Reference Gershon, Alliey-Rodriguez and Liu2011), putative genomic alterations not only overlap between SZ and BD but also with those for childhood-onset neurodevelopmental disorders such as autism and attention-deficit hyperactivity disorder, challenging the basis of current classification (Owen et al. Reference Owen, O'Donovan, Thapar and Craddock2011).
Clinical and nosological implications
Acuteness of onset, shifting polymorphic symptoms and early remission are reported as characteristic features, but it proved difficult to differentiate ATPDs from longer-lasting psychotic or affective disorders because of the lack of specificity of their symptomatology (Castagnini & Berrios, Reference Castagnini and Berrios2011). Likewise, little evidence supports the subdivision of acute polymorphic psychotic disorders into those ‘with’ or ‘without’ schizophrenic symptoms on the basis of 1 or 3 months’ duration; and the vaguely defined acute predominantly delusional disorder is a diagnosis by exclusion from APPD, persistent delusional disorder and SZ (Sartorius et al. Reference Sartorius, Üstun, Korten, Cooper and van Drimmelen1995). Only the ‘negative symptoms’ can distinguish SZ from ATPDs, as the Schneiderian first-rank symptoms of SZ have been observed in both clinical groups (Jäger et al. Reference Jäger, Bottlender, Strauss and Möller2003; Pillmann & Marneros, Reference Pillmann and Marneros2005; Salvatore et al. Reference Salvatore, Baldessarini, Tohen, Khalsa, Sanchez-Toledo, Zarate, Vieta and Maggini2011).
The proposed revision of ATPDs in ICD-11 (Gaebel, Reference Gaebel2012) recommends that polymorphic psychotic disorder without schizophrenic features (i.e. APPD) should be regarded as the typical syndrome with a duration of less than 3 months, whereas the subtypes with schizophrenic and predominantly delusional symptoms should be reclassified into other categories of the newly renamed F2 section ‘Schizophrenia spectrum and other primary psychotic disorders’. Also under discussion is the proposal to add ‘qualifiers’ for symptoms, course, cognition and functional impairment. Although the changes proposed would bring ATPD closer to the DSM categories of BPD and schizophreniform disorder, it seems questionable whether this category will prove useful in clinical practice and research given its uncertain validity.
Methodological limitations
The limitations of this study are those inherent to genetic epidemiological research, which uses information routinely collected from case psychiatric registers. Although the clinical categories are assumed to have characteristic symptoms, and patients given the same diagnosis are likely to have a similar prognosis, response to treatment and increased rates of illness in family members, they often show variability. In addition, while the diagnoses of SZ and BD were validated (Kessing, Reference Kessing1998; Jakobsen et al. Reference Jakobsen, Frederiksen, Hansen, Jansson, Parnas and Werge2005), no assessment of ATPDs has yet been conducted using operational symptom checklists, and we selected only those patients who did not develop another diagnosis in the 12 months following initial admission. Because most diagnostic changes occur early (Castagnini et al. Reference Castagnini, Foldager and Bertelsen2013), misdiagnosis of ATPDs would hardly affect the findings of the study.
Another possible methodological issue hinges on the fact that the three diagnostic categories were taken as akin to time-dependent variables and hence not mutually exclusive. However, the results did not change when comparison included only cases admitted for the first time with ATPDs.
Lastly, the attributable risk needs to be viewed cautiously because of variance in the frequency of mental disorders in the population. High rates of ATPDs may be related to the fact that this category subsumes the concept of reactive psychosis, a traditional diagnosis in Denmark before the introduction of ICD-10 (Castagnini, Reference Castagnini2010), and to an unwillingness to diagnose SZ in the first episode of psychosis.
Conclusions
These findings reflect the heterogeneity of the clinical features subsumed under ATPDs and suggest that their differentiation from SZ and BD based on family psychiatric predisposition remains uncertain.
Declaration of Interest
None.
