From operational diagnostic to dimensional-continuum concepts of psychotic and non-psychotic illness: Embracing catatonia across psychopathology and intrinsic movement disorder in neural network dysfunction

Psychiatry is currently negotiating several challenges that are typified by (but are not unique to) schizophrenia: do periodic refinements in operational diagnostic algorithms (a) resolve intricacies and subtleties within and between psychotic and non-psychotic disorders that are authentic and impactful, or (b) constitute arbitrary and porous boundaries that should be complemented, or even replaced, by dimensional-continuum concepts of abnormality and dysfunction. Critically, these issues relate not only to apparent boundaries between diagnoses but also to those between ‘health ’ and ‘illness ’ . This article considers catatonia within evolving dimensional-continuum approaches to the description of impairment and dysfunction among psychotic and non-psychotic disorders. It begins by considering the definition and assessment of catatonia vis-` a-vis other disorders, followed by its long-standing conjunction with schizophrenia, relationship with antipsychotic drug treatment, transdiagnostic perspectives and relationships, and pathobiological processes. These appear to involve dysfunction across elements in overlapping neural networks that result in a confluence of psychopathology and intrinsic hypo-and hyperkinetic motor dysfunction. It has been argued that while current diagnostic approaches can have utility in defining groups of cases that are closely related, contemporary evidence indicates categorical diagnoses to be arbitrary divisions of what is essentially a continuous landscape. Psychotic and non-psychotic diagnoses, including catatonia, may reflect arbitrary areas around points of intersection between orthogonal dimensions of psychopathology and intrinsic movement disorder in a poly-dimensional space that characterises this continuous landscape of mental health and dysfunction.


Psychiatric nosology and catatonia
From origins that appear to predate the 1800s (Berrios and Marková, 2018;Hirjak et al., 2022a), concepts of catatonia have evolved from (i) an initially autonomous disorder (Kahlbaum, 1874), through (ii) subsequent conjunction with schizophrenia (Kraepelin, 1899;Bleuler, 1911;Kendler, 2020), to (iii) recognition of conjunction with a broader range of other disorders (Chakrabarti, 2012;Solmi et al., 2018;Vaquerizo-Serrano et al., 2021;Luccarelli et al., 2022) and (iv) ultimate progression, full circle, towards autonomy in DSM-5 (Tandon et al., 2013) and ICD-11 (Reed et al., 2019).Nevertheless, this journey remains controversial (Fink et al., 2010;Ungvari et al., 2010;Heckers et al., 2010).Furthermore, on a background of incremental evolution in psychiatric nosology (Insel et al., 2010;Owen, 2014;Aftab and Ryzna, 2021), the field is currently negotiating several challenges that are typified by schizophrenia, but extend to bipolar disorder, major depressive disorder, autism spectrum disorder and beyond: do periodic refinements in operational diagnostic algorithms (a) resolve intricacies and subtleties within and between psychotic and non-psychotic disorders that are authentic and impactful , or (b) constitute arbitrary and porous boundaries that should be complemented, or even replaced, by dimensional-continuum concepts of abnormality and dysfunction.Critically, these issues relate not only to current boundaries between diagnoses but also to those boundaries that currently define the transition between 'health' and 'illness' (Owen, 2014; van Os and Reininghaus, 2016;Guloksuz and van Os, 2018;Waddington et al., 2022).
The historical epistemology of catatonia and the etymology of 'psychomotor' abnormality vis-à-vis psychopathology and intrinsic movement disorder have recently been reviewed, both in other articles in this Special Issue of Schizophrenia Research and elsewhere (Wijemanne and Jankovic, 2015;Walther et al., 2019;Kendler, 2020;Foucher et al., 2022;Hirjak et al., 2022b).This article considers catatonia within evolving concepts of dimensional-continuum approaches to the description of impairment and dysfunction among psychotic and nonpsychotic disorders.It begins by considering the definition and assessment of catatonia vis-à-vis other disorders, followed by its long-standing conjunction with schizophrenia, relationship with antipsychotic drug treatment, transdiagnostic perspectives and relationships, and pathobiological process.

Operational diagnosis of catatonia
Among scales and criteria for the assessment of catatonia (Bräunig et al., 2000;Sienaert et al., 2011), a diversity of signs is held to associate in reaching this diagnosis in the absence of any 'core', pathognomonic features.Table 1 summarises the two scales and diagnostic criteria for catatonia that have received the most extensive psychometric evaluation and usage [see Bush et al., 1996 for review and synthesis of 10 previous scales/criteria into the 23-item Bush-Francis Catatonia Rating Scale (BFCRS); see Northoff et al., 1999 for review and synthesis of four previous scales/criteria into the 40-item Northoff Catatonia Rating Scale (NCRS)], together with DSM-5 criteria for catatonia (Tandon et al., 2013; based on validation studies by Peralta et al., 2001aPeralta et al., , 2010)).In addition, Table 1 juxtaposes these scales and criteria for catatonia with criteria for those disorders with which catatonia has been most commonly associated, i.e. schizophrenia, bipolar disorder and major depressive disorder; more circumscribed and focused signs are held to associate in reaching these diagnoses, each of which must include one or more 'core', pathognomonic feature(s).These three diagnoses are among those being re-conceptualised within a dimensional-continuum model of psychotic and non-psychotic illness (Owen, 2014; van Os and Reininghaus, 2016;Guloksuz and van Os, 2018;Waddington et al., 2022), hence this absence of 'core' features in the diagnosis of catatonia would favour its re-conceptualisation within such a model.

'Mutism' and 'grimacing' among cardinal features of catatonia
Given this diversity in features of catatonia, typically subdivided historically into 'hypokinetic' and 'hyperkinetic' variants (Wijemanne and Jankovic, 2015;Walther et al., 2019;Kendler, 2020;Foucher et al., 2022), this article continues by first juxtaposing one 'hypokinetic' with one 'hyperkinetic' sign.Then, for economy and coherence, it uses primarily hyperkinetic signs to exemplify relevant processes.Specifically, consider two typical signs of catatonia [i.e. they are noted historically (Kahlbaum, 1874) and included in each of the three syntheses of previous scales/criteria outlined above (BFCRS, NCRS, DSM-5)]: 'mutism' and 'grimacing'.When lack of spontaneity and flow of conversation/ alogia reaches the level of 'mutism : "no, or very little, verbal response"], does this then implicate the presence of catatonia as a distinct illness category and disease process, or does it reflect an arbitrary dichotomy at a point towards the upper end of the continuous dimension of the negative symptom of poverty of speech, with attendant impairment in ability to articulate positive symptoms?
In contrast, across history 'grimacing' has been conceptualized variously, from behavioural/pseudo-expressive, through psychomotor, to hyperkinetic/choreoathetoid/involuntary movements.Notably, the two most widely used scales for the general assessment of hyperkinetic movement disorder, the Abnormal Involuntary Movement Scale (AIMS; Guy, 1976) and the Rockland-Simpson Dyskinesia Rating Scale (RSDRS; Simpson et al., 1979), each include 'grimacing' as an item among choreoathetoid and other dyskinetic/involuntary movements.In the NCRS 'grimacing', 'choreoathetoid movements' and 'dyskinesias (abnormal involuntary movements)' are distinct items, each of which implicates catatonia, and both total and motor subscale scores on the NCRS are highly correlated with scores on each of the AIMS and the RSDRS (Northoff et al., 1999).However, in both the BFCRS and DSM-5 'grimacing' implicates catatonia, while neither choreoathetoid nor dyskinetic/involuntary movements implicates catatonia.
These contemporary differences between the NCRS vis-à-vis the BFCRS and DSM-5, together with the relationships between the NCRS and the AIMS and RSDRS, should be juxtaposed with the historical record.This indicates a number of conceptual trajectories for catatonia (Foucher et al., 2022), two of which bifurcate from the seminal observations of Kahlbaum (1874) and derive primarily from studies in schizophrenia during the pre-neuroleptic era.In one trajectory, the clinical descriptions of Kraepelin and Bleuler across the transition from the 19th to the 20th century are followed by emphasis on first rank symptoms that lead ultimately to DSM-5/ICD-11 diagnostic criteria; across this period through to the 1950s other authors also emphasised movement disorder, particularly hyperkinetic/choreoathetoid movements, as intrinsic to schizophrenia, though with less diagnostic impact (Waddington and Crow, 1988;Kendler, 2016).In the other trajectory, following the transition from the 19th to the 20th century through to the 1950s, hyperkinetic/choreoathetoid movements in schizophrenia were further elaborated by the Wernicke-Kleist-Leonhard school as 'parakinetic catatonia', with prevalence estimates of 5-8 % [see Waddington and Crow, 1988 for descriptions of these hyperkinetic/choreoathetoid and parakinetic catatonic features with tabulation of these estimates].
On this basis, it is not clear why the BFCRS and DSM-5, unlike the NCRS (see also Bräunig et al., 2000), include 'grimacing' but exclude other hyperkinetic/choreoathetoid/involuntary movements that historically also characterize catatonia.This may in part reflect the introduction of these scales during the era of widespread use of antipsychotic drugs, with resultant concern over confounding hyperkinetic movement disorder intrinsic to catatonia in schizophrenia with tardive dyskinesia associated with long-term exposure to antipsychotics.Nevertheless, in accordance with the historical record, there continues to be increasing rediscovery of such hyperkinetic (and hypokinetic) movement disorder as intrinsic features of psychotic illness (Waddington and Crow, 1988;Pappa and Dazzan, 2009;Whitty et al., 2009;Koning et al., 2010;Peralta et al., 2010;Walther et al., 2020;Hirjak et al., 2022b; see Section 3).Furthermore, studies have shown that both choreoathetoid movements and 'grimacing', together with posturing, are present to excess in infants and children who go on to receive a diagnosis of schizophrenia in adulthood (Walker et al., 1994;Jones et al., 1994;see Whitty et al., 2009).This raises a question conceptually similar to that posed in relation to 'mutism': does 'grimacing' implicate the presence of catatonia as a distinct illness category and disease process, or reflect an arbitrary dichotomy along a topographical continuum of hyperkinetic/ choreoathetoid/involuntary movement disorder?Such consideration of these two specific features of catatonia exemplifies several challenges that are accentuated by two clinical realities: (i) cross-sectional diversity in presenting features; and (ii) longitudinal diversity in presenting vis-à-vis other features that may emerge subsequently over what can vary between an acute episode and a persistent course of catatonic illness.Mechanisms that might subserve such diversity are considered below (see Section 6).

"Where have all the catatonics gone"?
When Mahendra (1981) posed this question, he did so on the basis of opinion that across decades of the mid-20th century the incidence of catatonia, particularly 'catatonic schizophrenia', was in decline.Counter-arguments include that such decline was more apparent than real and reflected across these decades changes in diagnostic procedure and diagnostic criteria (van der Heijden et al., 2005), gaps in training (Wortzel et al., 2021) and, as espoused by Fink (2009), "the change in venue for psychiatric practice from asylum to office, the rejection of physical examination, and the dependence on item rating scales for diagnosis".However, another factor may critically influence the manifestation of catatonia.
A study in 1983 was able to assess 101 chronically ill schizophrenia subjects within a long-term psychiatric hospital ['asylum' as per Fink, 2009], who were initially admitted in the pre-neuroleptic era; thus, they remained antipsychotic-naïve for many years, before ultimately receiving treatment with antipsychotics.Among these, 22 % evidenced 'mutism' after having articulated delusions and/or hallucinations earlier in their illness and 42 % evidenced choreoathetoid movements, while among those subjects assessed more broadly, 9 % evidenced the grasp reflex [an additional sign of catatonia (Sienaert et al., 2011;Walther et al., 2019)]; furthermore, each of 'mutism' and grasp reflex was more common in subjects with choreoathetoid movements than in subjects without such movements (Waddington andYoussef, 1986a, 1986b;Youssef and Waddington, 1988).At this time, neither of the two most psychometrically evaluated and widely used rating scales for the assessment and diagnosis of catatonia (BFCRS, NCRS) had yet been described.Nevertheless, such enrichment for this triad of 'mutism', choreoathetoid movements and grasp reflex would likely attain or exceed the overall prevalence estimate for catatonia in schizophrenia as identified recently on meta-analysis across history  [9.8 % (95 % confidence interval 8.0-12.0%); Solmi et al., 2018].
In contrast, a study across 1995-2010 by the same investigators assessed subjects presenting with first episode psychosis (FEP) to the same mental health service, which had now evolved to provide primarily home-based/community outpatient care ['office' as per Fink, 2009].Among the 82 subjects with schizophrenia, presentations with 'mutism' were absent, with low scores for choreoathetoid movements and grasp reflex (Nkire et al., 2021a).All subjects received treatment with antipsychotics at FEP and among those subjects followed up at six years post-FEP (Kingston et al., 2018) this absence of 'mutism' endured.

Duration of untreated psychosis
On initial consideration, interpretation of a marked diminution in apparent prevalence of catatonia between these two populations would appear challenging due to differences in age/chronicity of illness, model of care and duration of antipsychotic drug treatment.However, one important factor would be overlooked: the chronically ill subjects with schizophrenia, who were typically first hospitalized in youth during the 1940s prior to being assessed in 1983, had, in the pre-neuroleptic era, experienced a mean duration of initially untreated psychosis (DUP) of 17.1 years, followed by a mean of 17.5 years since first treatment with an antipsychotic.In 1983, after controlling for age and durations of antipsychotic treatment and interpolated antipsychotic-free intervals, greater likelihood of 'mutism' was predicted by longer DUP (Waddington et al., 1995).
These findings, subsequently replicated in an independent sample (Meagher et al., 2001(Meagher et al., , 2004)), indicate that 'mutism' becomes more evident with increasing DUP.This would be similar to the manner in which greater psychopathology (especially negative symptoms, including poverty of speech), reduced functionality and poorer quality of life also become more evident with increasing DUP, both in these studies (Scully et al., 1997;Quinn et al., 2000;Nkire et al., 2021b) and on meta-analysis across studies (Howes et al., 2021).On this basis, in schizophrenia the absence of 'mutism' at FEP relative to its prominence in these more chronically ill subjects would reflect not their differing age/chronicities of illness or models of care but, rather, the eras in which they experienced FEP in relation to the introduction of antipsychotics: treatment with antipsychotics was timely [mean DUP 1.2 years (Nkire et al., 2021b)] in subjects who recently experienced FEP, but was delayed by decades [mean DUP 17.1 years (Waddington et al., 1995)] in subjects who experienced FEP in the pre-neuroleptic era.
Such an association between longer DUP and greater likelihood of 'mutism ' [i.e. between shorter DUP and reduced likelihood of 'mutism']: (i) may contribute to the perception that the incidence of catatonia has declined across decades in the mid-20th century (Mahendra, 1981) in temporal contiguity with the introduction of antipsychotic drugs; (ii) generalizes to catatonia an adverse, progressive biopsychosocial process associated with untreated psychosis that can be ameliorated by expeditious treatment with antipsychotics in a manner that endures across a period of at least 20 years (O'Keeffe et al., 2022); and (iii) appears similar to relationships between longer DUP and greater severity along several putative dimensions of psychotic illness (Howes et al., 2021;Nkire et al., 2021b;O'Keeffe et al., 2022).These findings further suggest that features of catatonia such as 'mutism' be conceptualized within a dimensional-continuum model.

The historical record across the pre-neuroleptic era
Notably, Kendler (2016) reminds us that in 18 textbook or review articles describing schizophrenia and published across the preneuroleptic era 1899-1956, essentially all included what would now be called the positive and negative symptoms that define schizophrenia in each of six operational psychiatric diagnostic criteria across the contemporary, post-neuroleptic era 1972-2013.However, among these 18 articles 1899-1956, while essentially all included classic motor signs in their description of the catatonic subtype of schizophrenia, 16 (89 %) also included movement disorder (most commonly hyperkinetic features) in their general descriptions of schizophrenia.In contrast, such movement disorder was not included in any of the six contemporary operational psychiatric diagnostic criteria 1972-2013.Why should the definition of psychotic illness by psychopathology have endured essentially unaltered across 1899-2013, while movement disorder intrinsic to psychotic illness has been rediscovered yet has lost its status as a defining characteristic of psychosis?While this may in part again reflect the introduction of these operational criteria during the era of widespread use of antipsychotic drugs, with resultant concern over confounding movement disorder intrinsic to schizophrenia with that induced by antipsychotic drugs, their interpretation as two distinct, confounding processes appears to be incomplete.

Studies across the contemporary era of antipsychotic drugs
In reality, evidence continues to elaborate antipsychotic drugs as precipitating covert movement disorder in those with sub-threshold, illness-related predisposition, or exacerbating overt movement disorder intrinsic to psychotic illness (Whitty et al., 2009).Three sets of studies relating to hyperkinetic/choreoathetoid/involuntary movement disorder illustrate such relationships.
In 1982, Owens and colleagues examined a rare and critical population of 47 chronically ill subjects with schizophrenia who had remained untreated with antipsychotics for decades due to a dynamic, predominantly family-oriented approach to management in their hospital units.Using both the AIMS and the RSDRS, they reported hyperkinetic, choreoathetoid movements to have a prevalence, severity and topography [buccal-lingual-masticatory (BLM, including 'grimacing') > peripheral limb-truncal (PLT)] substantially indistinguishable from those of 364 otherwise similar subjects, in units of the same hospital, who had received conventional, long-term treatment with antipsychotics (Owens et al., 1982).Importantly, the group that had remained untreated with antipsychotics were slightly older that those who had received antipsychotic treatment; on controlling for this difference in age, hyperkinetic movements were found to be more prominent in the group that had received antipsychotic treatment than in the group who had not received such treatment (Crow et al., 1982).This would be consistent with a previously unappreciated extent of hyperkinetic movements intrinsic to the disease process of schizophrenia that was exacerbated by long-term treatment with antipsychotics.
In 1987, further data from the study of chronically ill subjects with schizophrenia who had experienced FEP during the pre-neuroleptic era, and only subsequently received treatment with antipsychotics, were analysed.A hypothesis was that this population would show hyperkinetic movements that were intrinsic to the disease process of schizophrenia and exacerbated by long-term antipsychotic treatment; thus, the question posed was whether the presence and severity of hyperkinetic/ choreoathetoid/involuntary movements, including 'grimacing', would be more associated (i) with the clinical features of psychotic illness for which antipsychotics were prescribed, or (ii) with the duration, dosage and continuity of antipsychotic treatment.Findings (Waddington et al., 1987) were decisive: each of the presence and severity of such movements was associated primarily with 'mutism', flattened affect and cognitive impairment, and these relationships were topographically specific for orofacial (BLM) rather than non-orofacial (PLT) movements.Furthermore, modest associations between hyperkinetic movements and both average dose of antipsychotics across total period of exposure and current dose of antipsychotics were negative rather than positive; these findings would be consistent with 'mutism', flattened affect and cognitive impairment resulting in a conservative rather than aggressive approach to sustained antipsychotic treatment, and lower antipsychotic doses on the day of assessment unmasking otherwise suppressed hyperkinetic movements.
Nevertheless, while important, such studies identify associations but cannot fully resolve causal relationships; these would require studies over decades in subjects randomized to treatment vs non-treatment with antipsychotics, which are both unethical and impractical.To address this issue, a model was used in which rats are treated with antipsychotics for substantial periods of their adult lives and periodically assessed for 'vacuous jaw/chewing movements' (VCM) as a putative index of orofacial (BLM) dyskinesia (Waddington et al., 1983).In young rats age 3.5 months at initiation, antipsychotic treatment for six months induced a progressive increase in VCM above a more slowly increasing baseline level of VCM following vehicle treatment.This excess of antipsychoticinduced VCM in antipsychotic-treated rats across youth through to middle age reached a level similar to that evident spontaneously in vehicle-treated rats across middle through to old age (i.e.22 months); prolonged antipsychotic treatment across age 22-25 months did not materially induce any further increase in VCM (Waddington et al., 1986;Waddington, 1990).
While brain changes associated with progression from young through to old age in rats differ from those associated with acute and chronic schizophrenia in humans, these studies elaborate conceptually related processes associated with long-term antipsychotic treatment: (i) in young adulthood such treatment appears not to induce hyperkinetic BLM movements de novo but, rather, to progressively precipitate such sub-threshold and/or exacerbate spontaneous/disease-related movements; and (ii) across middle through to old age, this difference between antipsychotic-treated and non-treated subjects diminishes, due primarily to a greater increase in spontaneous/disease-related relative to antipsychotic drug-precipitated movements.In schizophrenia, intrinsic, hyperkinetic BLM movements (including 'grimacing') appear to be: (i) exacerbated by long-term treatment with antipsychotic drugs; (ii) associated with 'mutism' as an arbitrary dichotomy at a point towards the upper end of a continuous dimension of poverty of speech; (iii) associated similarly with the dimensions of flattening of affect and cognitive impairment; and (iv) associated only weakly, and negatively, with length and vigour of antipsychotic treatment.This suggests that hyperkinetic features of catatonia involve dimensions of movement disorder topography and its non-motor features involve dimensions of psychopathology.

Addressing the enigma of neuroleptic malignant syndrome
When signs of catatonia are marked and extend to hyperthermia and autonomic instability, these rare and potentially dangerous events have been referred to as 'malignant' catatonia (Wijemanne and Jankovic, 2015;Walther et al., 2019).When in the course of antipsychotic treatment signs of catatonia occur in their most severe form with prominent hyperthermia and marked autonomic instability, these constitute extremely rare but potentially fatal neuroleptic malignant syndrome (NMS; Strawn et al., 2007;Guinart et al., 2021).Thus, the above spectrum of overlapping features has prompted the question "Are NMS and catatonia separate syndromes or is NMS a type of catatonia?"(Fink, 1996).
This question should be evaluated on the following background: (i) Antipsychotic-naïve subjects with schizophrenia show intrinsic hypokinetic and hyperkinetic motor features (Waddington and Crow, 1988;Pappa and Dazzan, 2009;Whitty et al., 2009;Koning et al., 2010;Peralta et al., 2010;Walther et al., 2020;Hirjak et al., 2022b) that can be modulated by acute and long-term treatment with antipsychotic drugs; (ii) Antipsychotic-naïve subjects with schizophrenia show dysregulation of baseline body temperature, with lability in response to heat and cold stress, together with some evidence that in schizophrenia baseline body temperature may be elevated in those subjects from whom antipsychotics have been withdrawn but less so in those subjects still receiving antipsychotics (Chong and Castle, 2004;Shiloh et al., 2005); (iii) In his initial description of dementia praecox, Kraepelin (1899) noted altered autonomic dysfunction, including increased heart rate, sweating and salivation, and altered pupillary function, which suggest increased sympathetic and decreased parasympathetic output; and (iv) More contemporary studies in schizophrenia variably suggest increases in heart rate, respiration rate and blood pressure, and decreased baroreflex sensitivity, though any influences of antipsychotic and other medications can be unclear (Bär, 2015).
An early study adopted a signalled escape paradigm that involved temporal warning of the presentation of an aversive tone stimulus to which an escape response was to be made; the characteristic heart rate deceleration shown immediately following the warning by control subjects was markedly attenuated in the schizophrenia group (Waddington et al., 1978).Thus, in schizophrenia: (i) there may be a general deficit in the ability to modulate heart rate in response to the assessment situation despite correct cognitive evaluation; or (ii) cognitive evaluation of the assessment situation may be impaired and fails to produce an appropriate efferent signal for cardiovascular change.Subsequently, the most widely investigated index of autonomic dysregulation in contemporary studies remains heart rate variability, which is reliably reported to be reduced in schizophrenia (Montaquila et al., 2015;Alvares et al., 2016;Stogios et al., 2021).However, these processes have received only limited study in antipsychotic-naïve subjects with schizophrenia and many antipsychotic drugs interact with receptors known to be involved in such processes (Stogios et al., 2021).
Thus, all of the defining characteristics of NMS, i.e. severity of catatonia with prominent hyperthermia and marked autonomic instability, appear present to a sub-threshold extent as intrinsic features of schizophrenia, each of which is (or has a viable substrate to be) influenced by treatment with antipsychotics.However, while these profiles would be consistent with NMS being an antipsychotic-exacerbated extreme of catatonia (Fink, 1996), its exceptional rarity and potentially fatal intensity suggest one or more individual vulnerability factor (s) of comparably rarity (Hirjak et al., 2021).On this basis, just as 'mutism' would appear to constitute an arbitrary dichotomy at a point towards the upper end of a continuous dimension of severity of poverty of speech, so the presence of NMS may constitute an arbitrary dichotomy at a point towards the upper end of a continuous dimension of severity of catatonia and its exacerbation by antipsychotic drugs.

Commonalities across disorders in which catatonia is encountered
Despite its long-standing conjunction with schizophrenia, catatonia has also long been recognized to occur in conjunction with nonpsychotic disorders, the most common being bipolar disorder, major depressive disorder and also autism spectrum disorder (Fink et al., 2010;Chakrabarti, 2012;Wijemanne and Jankovic, 2015;Solmi et al., 2018;Walther et al., 2019;Ghaziuddin et al., 2021;Vaquerizo-Serrano et al., 2021).That major depressive disorder and autism spectrum disorder are not commonly treated with antipsychotic drugs would elaborate the historical record across the pre-neuroleptic era that catatonia can be manifested both independent of such treatment and independent of conjunction with schizophrenia.Yet the general commonality of the clinical features of catatonia across these non-schizophrenia diagnoses raises the issue of whether there may be some overlap of disease processes across these disorders.
Though schizophrenia, bipolar disorder, major depressive disorder and autism spectrum disorder are classically considered to be distinct conditions, contemporary evidence indicates increasing overlap between them at multiple levels.For example, certain dimensions of psychopathology are evident not only in schizophrenia, bipolar disorder and major depressive disorder but also in autism spectrum disorder (Lambert et al., 2018;Ghaziuddin and Ghaziuddin, 2020;Guineau et al., 2022;Jutla et al., 2022).Also, the association between longer DUP and increasing severity of psychopathology (see Section 2.2) generalizes from schizophrenia to bipolar disorder and major depressive disorder with psychotic features (Howes et al., 2021;Nkire et al., 2021b), though it remains of unclear relevance to autism spectrum disorder (Ghaziuddin and Ghaziuddin, 2020).Furthermore, several recent structural and functional neuroimaging studies have identified common pathobiological processes across schizophrenia, bipolar disorder and major depressive disorder (Goodkind et al., 2015;Sheffield et al., 2017;Baker et al., 2018;Ma et al., 2019;Tu et al., 2019;Huang et al., 2020) that overlap with structural and functional neuroimaging abnormalities in autism spectrum disorder (Nair et al., 2020;Jutla et al., 2022;Jalbrzikowski, 2021;Lam et al., 2022).
Molecular genetic studies now indicate considerable overlap among genes associated with risk for eight psychiatric disorders that is mediated via 109 pleiotropic loci (i.e.having multiple phenotypic expressions), among which: 83 % related to schizophrenia, 72 % bipolar disorder, 48 % major depressive disorder, 36 % autism spectrum disorder, 16 % attention-deficit/hyperactivity disorder, 14 % Tourette syndrome, 11 % obsessive-compulsive disorder and 7 % anorexia nervosa (Brainstorm Consortium, 2018;Cross-Disorder Group of the Psychiatric Genomics Consortium, 2019).Notably, the four disorders showing the greatest conjunction with catatonia (schizophrenia, bipolar disorder, major depressive disorder and autism spectrum disorder) are the same four that share the greatest extent of overlap in dimensions of psychopathology, aspects of pathobiology and pleiotropic loci.This leads to the testable prediction that catatonia may be less likely in conjunction with attention-deficit/hyperactivity disorder, Tourette syndrome, obsessive-compulsive disorder and anorexia nervosa.However, current datasets do not yet allow this prediction to be tested empirically (Solmi et al., 2018).
Catatonia has also been described in 22q11.2deletion syndrome [velocardiofacial syndrome (VCFS)] (Faedda et al., 2015), which also involves a pleiotropic locus in terms of increased risk not only for schizophrenia, bipolar disorder and major depressive disorder but also for autism spectrum disorder (Radoeva et al., 2014;Jalbrzikowski, 2021).That catatonia has been variably described in association with other disorders across the spectrum of psychiatric, neurological and general medical illness (Oldham, 2018;Solmi et al., 2018;Anand et al., 2019;Walther et al., 2019;Luccarelli et al., 2022), including Covid-19 (Dawood et al., 2022), suggests that varying extents of these pleiotropic loci may be associated with varying risk for catatonia beyond those disorders in which they have been systematically evaluated.Collectively, the above studies on diagnoses in which catatonia is most commonly encountered associate catatonia with a concept of porous boundaries across diagnoses that are characterized by overlap in each of dimensions of psychopathology and continua of pathobiology and genetic risk.

Commonalities across disorders in which neuroleptic malignant syndrome is encountered
Though extremely rare, NMS can be encountered in essentially any disorder for which an antipsychotic is prescribed: it is encountered most commonly not only in schizophrenia, bipolar disorder and major depressive disorder but also in autism spectrum disorder (Berloffa et al., 2021;Guinart et al., 2021), i.e. in those disorders showing the greatest overlap in psychopathology, pathobiology and pleiotropic loci.NMS is more likely to prove fatal in those subjects with pre-existing catatonic features and early signs of autonomic instability (Guinart et al., 2021).Notably, heart rate variability, the most widely investigated index of autonomic dysregulation, is also reduced not just in schizophrenia (Montaquila et al., 2015;Alvares et al., 2016;Stogios et al., 2021) but also in bipolar disorder, major depressive disorder and autism spectrum disorder (Kemp et al., 2010;Bassett, 2016;Faurholt-Jepsen et al., 2017;Cheng et al., 2020).Collectively, these findings would support the proposition that NMS is an idiosyncratic, antipsychotic-induced exacerbation of features intrinsic to catatonia (see Section 3.3) that is manifested trans-diagnostically.

Relationships across rating scales
As Kendler (2016) reminds us, descriptions across the pre-neuroleptic era 1899-1956 consistently included delusions, hallucinations, thought disorder, emotional blunting, abnormalities of volition and disorders of movement and posture as primary features of schizophrenia; however, operational psychiatric diagnostic criteria across the contemporary, post-neuroleptic era 1972-2013 fail to include disorders of movement and posture.It is on this background that the dimensionalcontinuum approach to (ab)normality and (dys)function is posited to transcend conventional diagnostic boundaries (Owen, 2014;van Os and Reininghaus, 2016;Guloksuz and van Os, 2018;Waddington et al., 2022).While DSM-5 does not embrace the dimensional-continuum approach in its main Section II Diagnostic Criteria and Codes, it considers in Section III Emerging Measures and Models eight emerging dimensions of psychotic illness for assessment (hallucinations, delusions, disorganized speech, negative symptoms, impaired cognition, depression, mania and abnormal psychomotor behavior), each as a single item.Inclusion of a psychomotor dimension now reflects both the historical record (Kendler, 2016) and contemporary studies (Waddington and Crow, 1988;Pappa and Dazzan, 2009;Whitty et al., 2009;Koning et al., 2010;Peralta et al., 2010;Walther et al., 2020;Hirjak et al., 2022b), but is not further defined beyond "abnormal or bizarre motor behavior or catatonia" (Barch et al., 2013;Heckers et al., 2013).
However, on application of the most comprehensive instruments available for assessing each of these dimensions, the situation becomes more complex and revealing.For example: in relation to positive and negative symptoms the Schedule for the Assessment of Positive Symptoms (SAPS; Andreasen, 1984) and the Schedule for the Assessment of Negative Symptoms (SANS; Andreasen, 1983) include 34 and 25 individual items, respectively; in relation to psychomotor behavior, the NCRS and the RSDRS include 40 and 43 individual items, respectively.On juxtaposing these scales (Table 2), it is notable that 11 items in the NCRS overlap with 1 item in the SAPS, 5 items in the SANS, and 4 specific items and multiple generic items for dyskinesia in the RSDRS; items for rigidity and akinesia in the NCRS also overlap with numerous scales for their assessment as extrapyramidal side effects of antipsychotic drugs (see van Strien et al., 2015) and in Parkinson's disease (see Tolosa et al., 2021).Such extents of overlap between the most comprehensive scale for assessing catatonia and those developed for assessing psychopathology and movement disorder in schizophrenia suggest transdiagnostic intersections along dimensions of abnormality.

Relationships across diagnoses
By way of example, the common manifestation of both psychosis and depression may lead to a variety of diagnoses.Consider three clinical scenarios: (a) where positive psychotic symptoms are manifested both concurrently with a major depressive episode and in the absence of prominent mood symptoms, the DSM-5 diagnosis may be Schizoaffective disorder; (b) where positive psychotic symptoms are manifested concurrently with a major depressive episode but are not evident in the absence of prominent mood symptoms, the diagnosis may be Major depressive disorder with psychotic features; and (c) where positive psychotic symptoms are manifested concurrently with a major depressive episode following a manic episode but are not evident in the absence of prominent mood symptoms, the diagnosis may be bipolar disorder, depressed, with psychotic features.Do (a)-(c) reflect incisive diagnoses that resolve distinct illness categories and disease processes, or arbitrary boundaries that are better explained as varying points of intersection along continua within a dimensional model of psychotic and affective psychopathology (Guloksuz and van Os, 2018;Waddington and Russell, 2019)?
Similarly, the common manifestation of both psychosis and hyperkinetic movement disorder in antipsychotic-naïve subjects may lead to a variety of diagnoses.Consider three clinical scenarios (Table 3): (1) where positive psychotic symptoms are articulated while manifesting the negative symptom of flat affect, together with hyperkinetic, choreoathetoid movements but not 'grimacing', the diagnosis may be Schizophrenia [with dyskinesia but not catatonia] by both DSM-5 and NCRS; (2) where positive psychotic symptoms are articulated while manifesting the negative symptom of flat affect, together with hyperkinetic, choreoathetoid movements and 'grimacing', the diagnosis may be Schizophrenia [with dyskinesia but not catatonia] by DSM-5 but Schizophrenia with catatonia by NCRS; and (3) where positive psychotic symptoms are no longer articulated due to extreme poverty of speech (i.e. 'mutism') with flat affect, together with hyperkinetic, stereotyped movements and 'grimacing', the diagnosis may be Schizophrenia with catatonia by both DSM-5 and NCRS.Do (1)-(3) reflect incisive diagnoses that resolve distinct illness categories and disease processes, or arbitrary boundaries that are better explained as varying points of intersection along continua within a dimensional model of psychopathology and intrinsic movement disorder?Furthermore, when these same scenarios relating to psychopathology and hyperkinetic movement disorder are considered in antipsychotic-treated subjects, they reveal yet greater transdiagnostic b EPS, rating scales for assessment of extrapyramidal side effects of antipsychotic drugs (see van Strien et al., 2015).c PD, rating scales for assessment of Parkinson's disease (see Tolosa et al., 2021).
J.L. Waddington complexity due to the possible contribution of such medication: intrinsic dyskinesia vs antipsychotic-precipitated/exacerbated dyskinesia vs tardive dyskinesia; intrinsic catatonia vs antipsychotic-precipitated/exacerbated catatonia vs [potentially, in its most severe form] NMS.

Relationships across developmental course
It is now well recognized that: (i) psychotic-like experiences (PLEs) appear to be distributed across the general population and are associated with increased risk for evolution to psychotic illness ( van Os and Reininghaus, 2016;Healy et al., 2019); and (ii) the earlier these prodromal psychological anomalies occur in childhood and adolescence, such increase in risk generalizes beyond psychosis to a broader range of psychiatric disorders (McGorry et al., 2018;Healy et al., 2019).Similarly, it is now apparent that prodromal abnormalities of motor function are also associated with increased risk for, and poorer outcome in, psychotic illness; these involve a variety of hypo-and hyperkinetic motor features (Whitty et al., 2009;van Harten et al., 2017;Hirjak et al., 2018;Pieters et al., 2022) that include manifestation in childhood of choreoathetoid movements, posturing (Walker et al., 1994) and 'grimacing' (Jones et al., 1994) that are characteristic of adult catatonia.This prodromal confluence between psychopathological features, which are increasingly interpreted in dimensional terms, and motor features further suggests that such movement disorder may also be appropriately interpreted dimensionally.

Pathobiology of catatonia: neural networks and systems biology
At the forefront of contemporary neuroscience, traditional notions relating a given neuropsychiatric disorder to abnormalities in a particular brain region have been largely replaced by concepts of neural networks/systems biology whereby abnormalities in elements of any one or more of several interconnected regions in a given network can produce qualitatively and/or quantitatively similar dysfunction (Geschwind and Konopka, 2009;Pulvermüller et al., 2021;Prasad et al., 2022aPrasad et al., , 2022b)).These systems biology concepts have been extended to psychotic illness and associated movement disorder (Whitty et al., 2009).Following initial neuroimaging studies on catatonia across brain regions (Northoff, 2000), subsequent structural and functional neuroimaging studies on catatonia, primarily in schizophrenia, have implicated dysfunction in elements of a neural network that interconnects cortical (pre)motor areas [ventromedial prefrontal/orbitofrontal cortex, cingulate, presupplementary and supplementary motor areas and inferior parietal lobe)], striatum [caudate, putamen and globus pallidus], thalamus (with associated cerebellar connectivity) and primary motor cortex (Walther et al., 2019;Haroche et al., 2020;Hirjak et al., 2020;Sambataro et al., 2021).

Synthesis: a dimensional-continuum concept embracing catatonia
In his formulation of schizophrenia, Bleuler considered catatonia to represent increased severity of disease rather than a distinct syndrome (Rogers, 1991).More recently, others have argued that the structure of catatonia as a diagnostic category remains to be discovered (Wilson et al., 2015).The above Sections 1 to 6 document the numerous areas in which a dimensional perspective, which derives primarily from consideration of psychopathology, can be applied also to movement disorder in general and, in particular, to its co-occurrence with psychopathology in catatonia across psychotic and non-psychotic diagnoses.Like psychotic and non-psychotic illness (Waddington et al., 2022), movement disorder such as in catatonia also appears to be characterised by dimensional continuities that cannot be fully captured by operational diagnostic algorithms that are in considerable part predicated on discontinuities (i.e. points of rarity) between them.
In contrast, from a dimensional perspective within a neural network model, psychopathology and movement disorder (both of which are present in catatonia) blend inexorably.Such systems pathobiology captures and integrates dimensions of abnormality to describe dysfunction in a manner that dilutes any need for 'psychomotor' as an epistemological conjunction between 'psychological' and 'motor' processes in the genesis of catatonia and would overlap with Research Domain Criteria (Insel et al., 2010).Others have also proffered variant dimensional perspectives of catatonia (Peralta and Cuesta, 2001;Ungvari et al., 2010;Hirjak et al., 2019;Dell'Osso et al., 2022); as a specific example, orbitofrontal-prefrontal dysfunction in hypokinetic catatonia was evident similarly across subjects with either schizophrenia or bipolar disorder (Northoff et al., 2004).This is not to impugn the status of catatonia as an entity.Rather, its purpose is to include catatonia within the dimensional construct that is currently being evaluated in relation to schizophrenia, bipolar disorder, major depressive disorder, autism spectrum disorder and beyond; this construct involves dimensions across the contemporary breadth of psychotic, affective, cognitive and other domains of psychopathology, hypo-and hyperkinetic movement disorder and other aspects of phenomenology, together with their underlying structural and functional pathobiology.
In an influential article, Owen (2014) argues that while current diagnostic approaches have some utility in defining groups of cases that are more closely related than by chance, contemporary evidence indicates categorical diagnoses to be arbitrary divisions of what is essentially a continuous landscape.He suggests that we are dealing with traits that are best conceptualized as continuous rather than categorical at both the phenotypic and genotypic levels, such that there should be an increasing focus on dimensional measures of both psychopathology and underlying brain dysfunction.More specifically, he posits that current psychiatric syndromes occupy a gradient determined by decreasing relative contribution of genetically and/or environmentally induced neurodevelopmental impairment: for example, some disorders like intellectual disability are evident from birth, others like autism spectrum disorder and attention deficit-hyperactivity disorder present typically in childhood, and yet others like schizophrenia, bipolar disorder and major depressive disorder present most commonly in adolescence or adulthood.
In elaboration of Owen's perspective, psychotic and non-psychotic diagnoses appear to reflect arbitrary areas around points of intersection between orthogonal dimensions of psychopathology in a polydimensional space that characterises mental health through to illness; these arbitrary areas reside within a continuous landscape of outcomes in brain development that, from conception, is sculpted by geneenvironment interactions that determine cerebral function through to dysfunction (Waddington et al., 2022).Importantly, catatonia exemplifies how such concepts appear to generalize beyond dimensions of psychopathology to embrace also dimensions of intrinsic movement disorder across psychotic and non-psychotic illness within Owen's (2014) continuous landscape.

Role of the funding source
There was no specific funding for this manuscript.

CRediT author contribution statement
JLW is the sole author of the manuscript.

Declaration of competing interest
None.

Table 1
Threshold criteria for diagnosis of psychiatric disorders by DSM-5, BFCRS and NCRS.

Table 2
Overlap between NCRS, psychopathology and movement disorder scale items.
a RSDRS, multiple items across this scale.

Table 3
DSM-5 and NCRS criteria for catatonia applied to three clinical scenarios.Northoff Catatonia Rating Scale; -, absent.For further details on clinical scenarios, see Section 5.2.