Functional neuroimaging in patients with catatonia: A systematic review

Abstract

BACKGROUND

Catatonia is a challenging and heterogeneous neuropsychiatric syndrome of motor, affective and behavioral dysregulation which has been associated with multiple disorders such as structural brain lesions, systemic diseases, and psychiatric disorders. This systematic review summarized and compared functional neuroimaging abnormalities in catatonia associated with psychiatric and medical conditions.

METHODS

Using PRISMA methods, we completed a systematic review of 6 databases from inception to February 7^th, 2024 of patients with catatonia that had functional neuroimaging performed.

RESULTS

A total of 309 studies were identified through the systematic search and 62 met the criteria for full-text review. A total of 15 studies reported patients with catatonia associated with a psychiatric disorder (n=241) and one study reported catatonia associated with another medical condition, involving patients with N-methyl-D-aspartate receptor antibody encephalitis (n=23). Findings varied across disorders, with hyperactivity observed in areas like the prefrontal cortex (PFC), the supplementary motor area (SMA) and the ventral pre-motor cortex in acute catatonia associated to a psychiatric disorder, hypoactivity in PFC, the parietal cortex, and the SMA in catatonia associated to a medical condition, and mixed metabolic activity in the study on catatonia linked to a medical condition.

CONCLUSION

Findings support the theory of dysfunction in cortico-striatal-thalamic, cortico-cerebellar, anterior cingulate-medial orbitofrontal, and lateral orbitofrontal networks in catatonia. However, the majority of the literature focuses on schizophrenia spectrum disorders, leaving the pathophysiologic characteristics of catatonia in other disorders less understood. This review highlights the need for further research to elucidate the pathophysiology of catatonia across various disorders.

INTRODUCTION

Catatonia is a heterogeneous neuropsychiatric syndrome that involves dysregulation of motor, affective, cognitive, and behavioral functions [1, 2]. Although catatonia is frequently attributed to endogenous psychiatric illness, numerous disorders have been linked to this syndrome, including structural brain lesions, developmental disorders, and several systemic diseases [3]. Catatonia is a serious condition that carries a high burden of morbidity and mortality. Prompt intervention, including correction of any underlying medical conditions and psychiatric care tailored to the underlying cause, is essential [4, 5].

Despite the increasing research interest in catatonia, most studies have focused on psychiatric conditions linked with this syndrome, with little research comparing catatonia due to a medical condition with catatonia due to psychiatric illness [6]. An exploratory study conducted at Mayo Clinic by Smith et al. with 95 participants examined the clinical characteristics of catatonia, comparing catatonia due to a general medical condition with catatonia due to psychiatric illness. The study found that the absence of a psychiatric history and a history of clinical seizures were potential covariates that could assist in identifying individuals with catatonia disorder due to a general medical condition [7]. Thus, showing potential differences between these subgroups.

While the underlying neurobiological basis for catatonia is not fully understood, researchers have begun to explore structural and functional abnormalities in catatonia through neuroimaging. Several grey and white matter characterization studies in patients with catatonia have identified noteworthy findings compared to psychiatric controls including diffuse cortical atrophy and localized abnormalities in frontal, temporal, occipital and cerebellar regions, multifocal lesions and encephalomalacia [7-9]. Functional neuroimaging has shown hypoperfusion in frontal, temporal and basal ganglia areas [10, 11]. However, findings have been inconsistent across studies [12-14]. The differences in findings between catatonia samples may be due to low sample size within the studies but also different disease-related mechanisms across patients.

Functional neuroimaging techniques, including functional magnetic resonance imaging (fMRI), positron emission tomography (PET), single photon emission computed tomography (SPECT) and functional near-infrared spectroscopy (fNIRS), have significantly advanced our ability to assess in vivo activation of specific brain regions across various psychiatric disorders. These modalities measure regional cerebral blood flow (rCBF), neuroreceptor density, and blood oxygen levels [15]. fNIRS specifically measures changes in hemoglobin oxygenation in the brain using near-infrared light [16]. SPECT and PET utilize radioactive tracers to quantify rCBF and neuroreceptor concentration, reflecting synaptic activity in distinct brain regions. Meanwhile, fMRI directly measures blood oxygen level–dependent (BOLD) signal, inferred from changes in water molecule density to characterize patterns of brain activity. Notably, fMRI offers superior spatial and temporal resolution compared to PET or SPECT [17]. While fNIRS has better temporal resolution than fMRI and PET, its spatial resolution is inferior to other techniques [16]. Both resting-state and task-based fMRI are employed to elucidate the pathophysiology of different neuropsychiatric conditions, with task-based fMRI often yielding better correlates of individual behavioral differences in several studies due to its more accurate functional connectivity model [18]. These different neuroimaging methods yield a wealth different structural and physiologic information about catatonia, providing useful information about the syndrome from several different perspectives.

A few reviews have examined the structural and functional abnormalities found on neuroimaging in patients with catatonia. However, the differences between catatonia due to a medical condition and due to psychiatric illness have not been reviewed to our knowledge, especially in light of emerging evidence suggesting distinct clinical phenotypes [1, 19, 20]. Moreover, the rapid advancements in functional neuroimaging necessitate an updated, comprehensive synthesis of evidence [21].

In this systematic review, we aimed to: (1) compare the functional neuroimaging findings in catatonia associated with a psychiatric disorder and associated with another medical conditions, (2) summarize the existing literature examining patients with catatonia using brain functional neuroimaging, and (3) correlate functional neuroimaging to current pathophysiological theories and symptoms of catatonia.

METHODS

We adhered to the guidelines and criteria outlined by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [22]. Additionally, a protocol has been registered and approved in the Open Science Framework database (registration DOI 10.17605/OSF.IO/E4528).

Data Sources and Search Strategies

A comprehensive search of several databases from their inception to February 7^th, 2024, in English and Spanish languages was conducted. The databases included Ovid MEDLINE(R) and Epub Ahead of Print, In-Process & Other Non-Indexed Citations, and Daily, Ovid EMBASE, Ovid PsycINFO, Ovid Cochrane Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, and Scopus. The search strategy was designed and conducted by an experienced librarian (LP) with input from the study’s principal investigator and lead authors (BS, LD). Controlled vocabulary supplemented with keywords was used to search for functional neuroimaging in catatonia for adult patients. The actual strategy listing all search terms used and how they were combined is available in Appendix table 1.

Study Selection

Participants, Interventions, Comparators, Outcomes, and Study (PICOS) design guidelines were used to establish specific inclusion/exclusion criteria [23]. The following inclusion criteria were used: (1) articles published in English or Spanish; (2) involving adults; (3) with a diagnosis of catatonia using any diagnostic criteria; (4) who underwent any type of functional neuroimaging. We allowed variation in prospective or retrospective nature of study, observational or interventional study design, and comparators (including studies with no comparison group). We included studies with at least ten patients with catatonia and allowed for studies with less than ten patients with catatonia if they had a comparison group. Case series of less than ten patients, case reports, methods papers, reports without outcome data, conference and poster presentations, reviews, meta-analyses, abstracts, and dissertations were excluded.

Procedures

Articles identified through searches were imported into Covidence systematic review software (Melbourne, Australia. Available at www.covidence.org). After removing duplicates, two study team members (LD, JPO) independently screened articles. A preliminary review of titles and abstracts was conducted to eliminate irrelevant articles. We then reviewed the full texts of the remaining articles for inclusion and exclusion criteria. Disagreements were resolved through consensus and consultation with the senior author (BS).

Study Quality and results

Two authors (LD, NAN) independently reviewed each article for quality and risk of bias using the adapted version of the Imaging Methodology Quality Assessment Checklist [24] (Appendix S2). Discrepancies in ratings were resolved by a third reviewer (JPO). This quality rating tool assesses the following areas: subjects, methods for image acquisition, analysis, results, and conclusions [24, 25] (please see S.2 for details).

Data Extraction

Data regarding the study population, imaging technique, characteristics of catatonia, findings, and other study details were extracted into a preformatted spreadsheet by two of the study team members (LD, NAN). In cases where the manuscript lacked sufficient or clear information, authors were contacted for clarification or additional details. Information from all included studies was then recorded and synthesized for the systematic review.

RESULTS

Study selection

A total of 310 studies were identified through the systematic search. Of these, one was a duplicate, 247 were excluded based solely on title and abstract, and 62 met the criteria for full-text review. No additional references were found in the reference lists of previous reviews. There was one article that could not be retrieved. Figure 1 displays the PRISMA flow diagram for the study inclusion process.

Figure 1. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow-chart.

Created using ShinyApp PRISMA flow chart creator [94]

Following the full-text review, 16 articles met the inclusion criteria. A total of 15 studies reported patients with catatonia associated with a psychiatric disorder (n=241) [10, 26-36], and one study reported catatonia associated with other medical conditions (n=23 patients with N-methyl-D-aspartate receptor [NMDAR] antibody encephalitis [37]). The detailed descriptions of the articles included in this review are presented in table 1.

Table 1.

Characteristics of Studies Included in Systematic Review

First author, Year	Study Sample	Etiology Classification	Schizophrenia /Schizoaffective Disorder (n)	Bipolar Disorder (n)	Depressive Disorder (n)	NMDARE (n)	Catatonia characteristics	Assessment of Catatonia	Severity of Catatonia	Functional neuroimaging method	Findings
Foucher, 2018 [27]	Catatonia, n = 20 Psychiatric control (Schizophrenia with cataphasia), n = 9 Healthy control, n = 27	A-PD	20	0	0	0	Acute Periodic Catatonia	Wernicke-Kleist-Leonhard definition / BFCRS	The mean BFCRS of PC was 4.7 ±3	fMRI cognitive task (rCBF)	Higher rCBF in left PrCG. No correlation in between rCBF and BFCRS.
Foucher, 2020 [26]	Catatonia, n = 9 Psychiatric control, n = 26 Healthy control, n = 37	A-PD	8	1	0	0	Acute Periodic Catatonia	Operationalized criteria for research of periodic Catatonia / BFCRS	The mean BFCRS of PC was 3±2	fMRI cognitive task (rCBF)	rCBF was higher in periodic Catatonia than non-periodic Catatonia and controls in l-SMA and l-LPM.
Hirjak,2020 [28]	Catatonia, n = 24 Psychiatric control (Schizophrenia without Catatonia), n = 87	A-PD	24	0	0	0	Catatonia due to schizophrenia	NCRS	NCRS total score was 6.88 ± 2.38	Resting fMRI and structural imaging (INA and GMV)	Reduced intrinsic network activity in FP and FT cortical motor networks. Linked INA and GMV alterations in CR and PF cortical motor networks were associated to behavioral Catatonia symptoms. Linked INA and GMV alterations in FT and FP networks were associated with affective Catatonia symptoms.
Kerik-Rotenberg, 2020 [37]	Catatonia, n = 23 NMDA-RE, n = 33	A-MC	0	0	0	23	Acute Catatonia due to NMDA-RE that received steroids for encephalitis and lorazepam for Catatonia	BFCRS	Not mentioned totally For one patient it was mentioned as 18	FDG-PET	Similar patterns between Catatonia and non-Catatonia but not formally differentiated. Focal bilateral hypermetabolism in CR, TL, and R-INS. Bilateral hypometabolism in the OL and PL. Focal hypometabolism in HES.
Lefebvre, 2023 [40]	Schizophrenia with psychomotor slowing, n = 60 (subgroup with Catatonia, n = 46) Psychiatric control (schizophrenia without psychomotor slowing), n=23 Healthy control, n = 40	A-PD	46	0	0	0	Patients from randomised, double-blind, four-arm, placebo-controlled trial of add-on TMS for psychomotor sloweing in schizophrenia spectrum disorders. Patients from inpatient and outpatient settings with catatonia defined as score of at least 1 in two of the first 14 items of the BFCRS. Unknown timeline of symptoms. Some patients on benzodiazepines	BFCRS	BFCRS mean 7.4 ± 4.5	Resting fMRI, structural MRI, single and double-pulse TMS effects from L PMC.	Reduced cortical excitability and lower cortical inhibition in catatonia compared to healthy controls found in TMS. The increased cortical inhibition was associated with increased L-M1/SMA and L-M1/R-CR resting state functional connectivity. This contrasted with the results of HC that demonstrated increased inhibition associated with decreased L-M1/R-PMC resting state functional connectivity.
Nakamura, 2021 [29]	Catatonia, n = 10	A-PD	4	1	5	0	Acute Catatonia over 50 years old that underwent ECT	BFCSI/BFCRS	Mean (SD) of BFCSI was 7.4 (1.5) and for BFCRS was 16.9 (1.7)	Resting fNIRS (oxyHb - RSFC)	Lower RSFC in PFC otherwise no changes in frontal and parietal cortices (more pronounced in subgroup of BD and MDD). Higher RSFC in PFC after resolution of Catatonia symptoms.
Northoff., 1999 [30]	Catatonia, n = 10 Psychiatric control, n = 10 Healthy control, n = 20	H-PD	3	6	1	0	Post acute akinetic Catatonia 1 week after single dose of IV lorazepam 2 mg	Lohr and Wisniewski / Rosebush / BFCRS / NCRS	Not mentioned	SPECT (r-CBF and binding of 123I)	Right-left alterations in the left sensorimotor cortex and reduced r-CBF and significant right-left alterations in the right lower PFC and PC. Relation of motor and affective symptoms in Catatonia with lFC and rPFC 123I binding as well as with rLP r-CBF.
Northoff, 2000 [31]	Catatonia, n = 10 Psychiatric control, n = 10 Healthy control, n = 20	H-PD	3	6	1	0	Post acute akinetic Catatonia 1 week after single dose of lorazepam 2 mg	Lohr and Wisniewski / Rosebush/ BFCRS / NCRS	On the NCS, patients 64.7 ± 12.1 on average. On the Rosebush scale on day 0, patients scored a mean of 10.8 out of a maximum possible score of 24. By day 1 after lorazepam, this decreased to a mean of 2.4, indicating response to treatment.	SPECT (r-CBF)	Lower PFC and PC r-CBF in Catatonia. Deficits in visual and spatial abilities with altered correlation pattern with rP r-CBF. Right-left alterations of PFC and PC r-CBF in Catatonia. Correlation of attention function with motor symptoms and rPC r-CBF only in Catatonia.
Northoff, 2002 [10]	Catatonia, n = 18 Psychiatric control, n = 69 Healthy control, n = 32	H-PD	7	7	4	0	Post-acute Catatonia that received IV lorazepam 2-4 mg/day for the first 24 hours	Rosebush	Not mentioned directly but the prevalence of symptoms was mentioned as below: Immobility (18 patients), staring (16), mutism (17), autism (18), posturing (18), rigidity (13), negativism (14), catalepsy (18), grimacing (14), echolalia/praxia (8), stereotypies (9) and verbigerations (7).	fMRI motor task with emotional stimulation (r-CBF correlating Personal constructs)	Decreased activation and increased deactivation of rOFC with negative emotion stimulation. Increased activation and decreased deactivation on rOFC with positive emotion stimulation. Dysfunctional activation patterns in OFC and PMC during negative emotional stimulation that correlated with affective, behavioral, and motor alterations in Catatonia as well as with dimensions of self-esteem, emotional arousal, and social contact.
Northoff, 2004 [32]	Catatonia, n = 10 Psychiatric control, n = 10 Healthy control, n = 10	H-PD	3	7	0	0	Post acute akinetic Catatonia 1 week after single dose of IV lorazepam 2 mg	Lohr and Wisniewski / Rosebush / BFCRS / NCRS	The NCS total was 64.7(12.1) on day 0 and 8.2(1.9) after treatment on day 1. BFCRS score was 26.8 (5.7)	fMRI motor task emotional stimulation (r-CBF and negative emotional stimulation)	Alterations in activation of OFC and in functional connectivity to PMC in negative and positive emotions. Behavioral and affective symptoms of Catatonia correlated with OFC activity, whereas catatonic motor symptoms were related to medial prefrontal activity.
Parekh, 2022 [38]	Catatonia, n = 15 Healthy control, N=15	A-PD	15	0	0	0	Acute retarded catatonia	DSM-5 criteria/BFCRS/NCRS	mean total score of 21.07 ± 5.69 on the BFCRS The mean motor subscore on the BFCRS was 8.00 ± 2.73.	fMRI and resting fMRI	Widespread increased ROI-to-ROI functional connectivity, especially between sensorimotor, salience, frontoparietal, temporal and cerebellar regions. Reduced sensorimotor network and vertex wise cortical complexity in right insular cortex. Lorazepam responders had higher ROI-ROI and seed to voxel functional hyperconnectivity.
Richter, 2010 [33]	Catatonia, n = 6	H-PD	3	3	0	0	Post-acute akinetic Catatonia in short-term responders of lorazepam separated into subjects that receive lorazepam and placebo	Lohr and Wisniewski / Rosebush / BFCRS / NCRS	The mean total score on the NCS was 69 ± 11.5. The mean score on the Rosebush scale was 10.67 ± 0.82. Scores on the scale range from 0-15	fMRI motor task emotional stimulation	Decreased signal in OFC with emotional pictures in Catatonia with no abnormal activity in Catatonia after lorazepam
Sambataro, 2021 [39]	Catatonia, n=30 Psychiatric control (non-catatonic) n=28	A-PD	30	0	0	0	Acute Catatonia	NCRS/DSM-IV-TR	The NCRS total score in the catatonia group ranged from 3 to 14, with a mean of 6.9 ± 2.6.	Resting state fMRI	Correlation between increased functional network connectivity (FNC) in corticostriatal state and high NCRS scores. Increased static FNC in cerebellar regions and reduced low-frequency oscillations in basal ganglia networks
Satoh, 1993 [34]	Catatonia, n = 6 Psychiatric control (Schizophrenia without Catatonia), n = 12 Healthy control, n = 7	H-PD	6	0	0	0	Post-acute Catatonia in schizophrenia patients (2-3 months after episode)	Presence of catatonic stupor or excitement, posturing, mannerisms and negativism	Mean BPRS (SD) Emotional withdrawal 3.5(0.5) Mannerism and posturing 1.7(1.3) Hallucinatory behavior 0 Unusual thought content 2.5(1.7)	SPECT (rCBF)	Low mean IMP perfusion in bilateral FL and PL when compared with TL and OL. Low CBF in both dorsal regions of PL in Catatonia.
Scheuerecker,2009 [35]	Catatonia, n = 12 Healthy control, n = 12	H-PD	12	0	0	0	Post-acute Catatonia (1 month to 5 years after episode)	Lohr and Wisniewski / Rosebush / BFCRS / NCRS	Rosebush catatonia scale (M ± SD) 5.17 ± 1.03 Northoff catatonia scale (M ± SD) 38.8 ± 1.03	fMRI-BOLD motor task (r-CBF)	Decreased activation of SMA, PFC and PC
Walther,2017 [36]	Catatonia, n = 15 Psychiatric control (Schizophrenia without Catatonia), n = 27 Healthy control, n = 41	A-PC	15	0	0	0	Acute Catatonia	BFCRS	BFCRS total was 8.2 (5.2)	fMRI resting (rCBF and GM density)	Increased rCBF in Catatonia in the SMA and vPMC (regions for self-initiated movements). SMA perfusion differed between excited and retarded Catatonia, with increased SMA activity in retarded. Catatonia severity associated with higher perfusion in bilateral SMA and left vPMC. Positive linear association of GM in CR and catatonia severity.

Abbreviations: 123I: Iodine-123, A-PD: Acute Catatonia associated to a Psychiatric Disorder, A-MC: Acute Catatonia associated to a Medical Condition, AM: Amygdala, BD: Bipolar Disorder, BFCRS: Bush Francis Catatonia Rating Scale, BOLD: Blood oxygenation level dependent, C: Cortex, CR: Cerebellum, Dep: Depression, DLPFC: Dorsolateral Prefrontal Cortex, DSM-IV: Diagnostic and Statistical Manual of Mental Disorders fourth edition, DSM-5: Diagnostic and Statistical Manual of Mental Disorders fifth edition, ECT: Electroconvulsive therapy, F: Frontal, FDG-PET: fluorodeoxyglucose -positron emission tomography, fMRI: Functional magnetic resonance imaging, fNIRS: Functional Near-infrared spectroscopy, GM: gray matter, GMV: gray matter volume, H-PD: History of Catatonia associated to a Psychiatric Disorder, HES: Heschl gyrus, IMP: 123-I Iodine amphetamine, INA: intrinsic neural activity, INS: Insula, IPL: Inferior Parietal Lobe, l: Left, L: Lobe, LPM: lateral premotor cortex, NCRS: Northoff catatonia rating scale, NMDA-RE: NMDA receptor encephalitis, O: Occipital, OFC: Orbitofrontal cortex, P: Parietal, PMC: Premotor cortex, PrCG: Pre-central gyrus, r: Right, rCBF: Regional cerebral blood flow, RSFC: Resting state functional connectivity, SMA: Supplemental Motor area, SPECT: Single photon emission computed tomography, T: Temporal, TMS: Transcranial Magnetic Stimulation, TP: Temporal Pole, vPMC: ventral premotor cortex.

Studies Sample

In total, 16 studies examined the outcomes of 264 patients diagnosed with catatonia, with eight of these studies having a sample size equal to or greater than 15 patients [10, 27, 28, 36–40]. Data available indicate a weighted average age of 37.30 years, with 50.93% of patients being female. Various imaging techniques were used across studies, with 3 studies using SPECT [30, 31, 34], 11 studies using fMRI [10, 26–28, 32, 33, 35, 36, 38–40], 1 study using fNIRS [29], and 1 study using 2-deoxy-2-[18F]fluoro-D-glucose positron emission tomography [18-FDG-PET] [37].

In the case of catatonia due to a psychiatric condition, fMRI has been the most commonly used functional imaging technique [10, 26-28, 32, 33, 35, 36, 38-40]. Among the fMRI studies, three used a motor task with emotional stimulation [10, 32, 33], five were conducted in resting state [28, 36, 38-40], two involved a cognitive task [26, 27], and one study used exclusively a motor task [35]. Additionally, four of the fMRI studies were done in post-acute catatonia patients [10, 32, 33, 35].

Three studies used SPECT to evaluate patients with catatonia associated with psychiatric disorders all focusing on post-acute catatonia [30, 31, 34]. Furthermore, resting fNIRS was used in one study to evaluate patients with acute catatonia due to a psychiatric disorder [29], and FDG-PET was used in the only study found of acute catatonia due to a medical condition [37].

We summarized the included studies into the following categories: (1) acute catatonia associated with a psychiatric disorder, (2) history of catatonia associated with a psychiatric disorder and, (3) acute catatonia associated with a medical condition. The brain regions showing abnormalities in the reviewed neuroimaging studies are summarized in figure 2.

Figure 2. Regions with Altered Activity Found in Functional Neuroimaging Studies of Catatonia.

Neuroimaging findings in patients with catatonia. The results are divided into a) acute catatonia due to a psychiatric disorder, b) history of catatonia due to a psychiatric disorder and (c acute catatonia due to a medical condition. Red indicates hyperactivity, blue sign indicates hypoactivity, and yellow indicates normalization of hyperactivity after lorazepam. Figure 2 was partly generated using Servier Medical Art, provided by Servier, licensed under a Creative Commons Attribution 3.0 unported license.

Abbreviations: ACC: Anterior cingulate cortex, AM: Amygdala, CP: Cerebellar peduncle, CR: Cerebellum, DLPFC: Dorsolateral Prefrontal Cortex, HES: Heschl gyrus, INS: Insula, IPL: Inferior Parietal Lobe, OFC: Orbitofrontal cortex, OL: Occipital lobe, PFC: Prefrontal Cortex, PC: Parietal cortex, PCC: Posterior cingulate cortex, PL: Parietal lobe, SMA: Supplemental Motor area, TP: Temporal Pole, vPMC: ventral premotor cortex

Acute catatonia associated with a psychiatric disorder [26–29, 36, 38–40]

A total of eight studies evaluated patients with acute catatonia associated with a psychiatric disorder (n=169). The majority of patients (93.49%) had a diagnosis of schizophrenia spectrum illness, while 4.14% had unipolar depression, and 2.37% had bipolar disorder. Nakamura et al. examined functional imaging in an elderly cohort of patients before and after ECT treatment using fNIRS. They reported lower left and right prefrontal cortex (PFC) functional connectivity during the episode of catatonia particularly in patients with bipolar disorder and major depressive disorder (p=0.028 and p=0.046 respectively) with increased mean functional connectivity in PFC after resolution of symptoms of catatonia [29].

While examining patients with schizophrenia, two studies of Foucher et al. showed hyperactivity of the left PFC, anterior to medial cingulate cortex, left supplemental motor area and lateral premotor area when performing a cognitive task. Both of these studies used fMRI and assessed acute periodic catatonia [26, 27]. Similarly, Walther et al. evaluated patients with schizophrenia and showed hyperperfusion in the Supplementary Motor Area (SMA) and left ventral premotor cortex (V-PMC), with more pronounced hyperperfusion observed in individuals with severe and retarded forms of catatonia. Hirjak et al., compared schizophrenia spectrum illness patients with and without catatonia. They reported reduced intrinsic network activity in the frontoparietal and frontotemporal cortical motor networks in patients with catatonia, alongside increased aberrant intrinsic neural activity in cerebellar and parietal frontal motor networks associated with behavioral symptoms, and increased aberrant intrinsic neural activity in frontotemporal and frontoparietal networks linked to affective symptoms of catatonia [28, 36].

Three other studies investigated functional network connectivity in catatonia patients. Parekh et al. showed a widespread increase in functional connectivity between regions of interest (ROI), particularly between sensorimotor, salience, frontoparietal, temporal, and cerebellar ROIs (Hedges’ g between 1.11 and 1.73.) Compared to other studies, patients in this study reported the highest severity of disease, with an average Bush-Francis Catatonia Rating Scale (BFCRS) score of 21.07 (SD 5.69). Sambataro et al., found an increased static functional connectivity (sFNC) in the cerebellum within the sensory-motor network, as well as low-frequency oscillations affecting the basal ganglia, visual, salience, default mode, and executive networks. Finally, Lefebvre et al. used data from transcranial magnetic stimulation (TMS) to measure cortical excitability and found that the increase in cortical inhibition was associated with increased sFNC in the left primary motor and SMA network and left primary motor and right celebellar network. [38-40].

History of catatonia associated with a psychiatric disorder [10, 30-35]

A total of seven studies evaluated patients with history of catatonia associated with a psychiatric disorder (n=72). The majority of patients (51.39%) had a diagnosis of schizophrenia spectrum illness, while 40.28% had bipolar disorder, and 8.33% had unipolar depression. Remission of symptoms of catatonia when functional neuroimaging was performed ranged from 1 week to 5 years.

Five studies examined the short-term effects of lorazepam treatment on brain activation patterns [10, 30-33]. The majority of these studies found reduction in activation or normalization in the orbitofrontal cortex (OFC) following lorazepam administration. Results by Northoff et al. 2002 and 2004 demonstrated dysfunctional activation with decreased activation and increased deactivation patterns in the OFC and PMC during negative emotional stimulation. These altered activation patterns correlated with affective, behavioral, and motor changes. In two of these studies in 1999 and 2000, Northoff et al. explored the impact of lorazepam on catatonic syndrome using SPECT imaging, unlike other studies that utilized fMRI. In the 1999 study, following a single dose of IV lorazepam, significant right-left differences were observed in both the left sensorimotor cortex and the right lower PFC and parietal cortex. The motor and affective symptoms of catatonia correlated with left upper frontal, right lower prefrontal and right lower parietal cortical binding. Similarly, in the 2000 study, SPECT imaging revealed decreased regional cerebral blood flow (r-CBF) in the PFC and parietal cortex of catatonic patients after lorazepam administration [10, 30-33].

Two other studies examined patients in post-acute catatonia from 1-month up to 5 years since their last episode [34, 35]. Low perfusion was observed in bilateral frontal lobe and parietal lobe when compared with temporal lobe and occipital lobe. Additionally, other findings revealed hypoactivity in parietal cortex, supplemental motor area, and PFC [34, 35].

Catatonia associated with another medical condition [37]

Only one study evaluated acute catatonia associated with a medical condition. Kerik-Rotenberg et al. used 18-FDG-PET to examine patterns of abnormal regional brain metabolism in patients with NMDAR encephalitis and compared them with healthy controls. Thirty-three patients with NMDAR encephalitis were enrolled, of which 23 (69%) had experienced catatonia during their disease course. In a voxel-based statistical analysis, patients with NMDAR encephalitis had extensive bilateral hypometabolism throughout the occipital lobe (calcarine, cuneus, and lingual regions) and parietal lobe (postcentral gyrus, post cingulum gyrus, precuneus, angular gyrus) and extending into Heschl’s gyrus. Additionally, they exhibited foci of hypermetabolism in the right temporal pole, amygdala, and insula and bilaterally in cerebellar peduncle and cerebellum [37].

DISCUSSION

In this systematic review, we reviewed 15 neuroimaging studies (n=218) of patients with catatonia associated with psychiatric disorders and associated with medical conditions. The majority of functional neuroimaging research in this area has primarily focused on catatonia within the context of schizophrenia spectrum disorders. This review identifies a significant research gap, particularly in exploring catatonia related to other medical conditions. The need for further research in this area is underscored by the fact that only one study was found that examined acute catatonia in patients with NMDA-R encephalitis.

In studies of acute catatonia associated with a psychiatric disorder, findings often revealed hyperactivity in brain regions, such as the PFC, the SMA, and the V-PMC. The degree of hyperactivity corresponded with severe symptoms and retarded presentation [26-29, 36, 38, 39]. Conversely, studies on history of catatonia associated with a psychiatric disorder indicated hypoactivity in several brain regions, including the PFC, the parietal cortex, and the SMA. Notably, decreased aberrant activation of the OFC was also observed following the administration of lorazepam [10, 30-35]. Finally, in terms of acute catatonia associated with a medical condition, we found a single study with patients with NMDAR encephalitis that reported hypometabolism in the occipital and parietal lobes as well as hypermetabolism in the temporal pole, insula, amygdala, and cerebellum [37].

Networks dysfunction

As described within the literature, catatonia is believed to manifest from the dysfunction within the top-down cortico-striatal-thalamic loop system and cortico-cerebellar circuits modulated by dopaminergic transmission that is triggered by dysregulation from the horizontal modulation of the prefrontal-parietal modulated by GABAergic and glutamatergic transmission [1, 41-44]. The clinical presentation seems to depend on the precise regions within these circuits related to motor, behavioral and emotional processing that are primarily impacted [42, 45]. The limited availability of studies across different etiologies restricts our ability to clearly discern distinct patterns of brain dysfunction for different etiologies.

In our systematic review, all network connectivity studies focused on acute catatonia associated with a psychiatric disorder [28, 29, 38-40]. These studies align with the current theoretical model of catatonia’s pathophysiology, revealing reduced resting state functional connectivity, particularly within the sensorimotor network and the cerebellum. The resting state functional connectivity was restored alongside symptom resolution [28, 29, 38]. Furthermore, there was notable evidence of diffuse ROI to ROI hyperconnectivity, especially in frontoparietal and cerebellar regions. Enhanced connectivity was also observed in sensorimotor, salience, and temporal regions [38-40]. Interestingly, the severity of catatonic symptoms correlated with increased functional connectivity in the corticostriatal network [28, 39]. The findings suggest that specific network disruptions may underlie varying symptom presentations. For instance, the frontotemporal and frontoparietal networks dysfunction were linked with affective symptoms, while the cerebellum, PFC, and frontal cortical motor networks dysfunction correlated with behavioral symptoms [28].

Frontal/Prefrontal regions abnormalities

The SMA is essential in orchestrating motor behavior, encompassing the planning, initiation, sequencing, and execution of movements, as well as integrating these processes with cognitive functions. This area is pivotal in synthesizing sequences into comprehensive, higher-order representations, a key process in facilitating advanced motor activities [46-49]. In the context of catatonia, there has been a link between dysfunction in the SMA and catatonic symptoms such as immobility, muscular rigidity, and abnormal posturing. Interestingly, this link also extends to the excited subtype of catatonia, indicating the SMA’s broader role in various manifestations of this condition [36, 50, 51]. Hyperperfusion of the SMA observed in functional neuroimaging studies of acute catatonia associated with a psychiatric disorder is thought to be a compensatory response. This increased activity may represent an attempt to overcome reduced or inhibitory output from the basal ganglia, thereby countering extensive motor inhibition [40, 52]. Conversely, decreased activation of the SMA observed in studies on history of catatonia associated with a psychiatric disorder could indicate ongoing dysfunction in this region [35]. We did not find any information regarding prefrontal or frontal lobe abnormalities in the single study on acute catatonia associated with a medical condition.

Similarly to the SMA, the premotor cortex (PMC) is integral to motor movement preparation, playing a significant role in higher-order cognitive processes as well [53]. Furthermore, it exhibits mirror neuron activity, further underlining its complexity and functionality [54, 55]. In the context of acute catatonia associated with a psychiatric disorder, hyperperfusion in the left v-PMC has been observed [36]. This phenomenon could be interpreted as analogous to the dysfunction seen in the SMA. It could represent a compensatory mechanism to counteract motor inhibition originating from the basal ganglia or the cerebellum, similar to the hypothesized role of the SMA in such conditions, while it may also play a role in echophenomena.

There is a strong correlation between negative emotional states and increased activity in two frontal cortex regions, the OFC and the ventromedial prefrontal cortex (VM-PFC), whereas the dorsolateral prefrontal cortex (DL-PFC) usually shows decreased activity [56]. The VM-PFC, in particular, plays a pivotal role in regulating emotions and influencing behavioral responses to environmental factors [57, 58]. A key factor in the modulation of these activation and deactivation patterns in these regions is the GABAergic system [30, 44]. The dorsomedial prefrontal cortex (DM-PFC) is crucial for understanding and predicting other people’s mental states. Conversely, the rostral medial PFC is responsible for self-awareness of mental state and the processing of information related to self [59]. In patients with acute catatonia associated with a psychiatric disorder, hypermetabolism in the DL-PFC may be associated with symptoms like immobility, impulsivity, and negativism [60]. Similarly, DM-PFC dysfunction could be linked to motor abnormalities related to akinesia [61]. Additionally, reduced functionality in the VM-PFC may contribute to fear-related symptoms in catatonia [62-64].

The OFC plays a vital role in interpreting emotional and environmental situations while making context-dependent decisions. It projects to the amygdala, which activates different cortical areas. A decreased activation in the OFC can cause the hyperactivation of the amygdala, resulting in a fear response [65, 66]. In two studies of patients with history of catatonia associated with a psychiatric disorder by Northoff et al., there was dysfunctional activation patterns in OFC during negative emotional stimulation which correlated with affective and behavioral alterations in catatonia [10, 32]. Echo-phenomena in catatonia are also thought to arise when the OFC is not functioning properly, as mirror neurons are disinhibited [67]. In patients with history of catatonia associated with a psychiatric disorder, the dysfunction of the OFC shows improvement with lorazepam, which suggests that the use of benzodiazepines in treatment can partially restore the balance between excitation and inhibition in these affected regions [33]. This possibly explains the observed hypoactivity in frontoparietal regions in individuals who have undergone treatment and highlights the potential therapeutic significance of targeting the OFC to address neural dysregulations in catatonia associated with psychiatric conditions [10, 30-35, 44, 67, 68].

Other Cortical regions abnormalities

Studies on acute catatonia associated with a psychiatric disorder found increased parietal lobe activity; particularly, in one study by Walther et al., greater symptom severity correlated with higher perfusion in the left anterior region [36]. In contrast, in studies of patients with history of catatonia associated with a psychiatric disorder, there was typically decreased activation in dorsal parietal regions, including alterations in left and right parietal activation [30, 31, 34]. Notably, Northoff et al. found a correlation between attention function, motor symptoms, and right parietal r-CBF in history of catatonia associated with a psychiatric disorder patients, suggesting that these abnormalities may be a trait marker, potentially predisposing individuals to the manifestation of catatonia [30]. A dysfunctional parietal cortex, specifically on the right, has been associated with motor behaviors like catalepsy and posturing in patients with history of catatonia associated with a psychiatric disorder [9, 69, 70].

Additionally, parietal lobe involvement in both acute catatonia associated with a psychiatric disorder and history of catatonia associated with a psychiatric disorder has been linked to deficits in visuospatial function [71]. This could explain the frequently reported intense fear in catatonia patients, which seems disconnected from their lack of movement [72]. This dissociation might result from a fronto-parietal dysfunction, where attention is shifted towards emotional processing at the expense of movement awareness [31, 71]. Regarding the study of acute catatonia associated with NMDAR encephalitis, bilateral hypometabolism was observed in several parietal lobe areas, including the precuneus, angular, and postcentral regions [37]. This pattern of hypometabolism contrasts with the hypermetabolism seen in acute catatonia associated with a psychiatric disorder, pointing to potential differences in the underlying dysfunction between these catatonia subtypes.

In the study of acute catatonia associated with NMDAR encephalitis, other distinct neuroimaging patterns were observed, including right hypermetabolism in the temporal lobe, including the temporal pole [37]. This hypermetabolism could correlate with impulsive behaviors seen in catatonia [73, 74]. Alongside this, the study found bilateral hypometabolism in the occipital lobe, encompassing the superior, middle, and inferior occipital lobes, the calcarine, cuneus, and lingual areas, and also in Heschl’s gyrus. The activation of Heschl’s gyrus is known to play a role in auditory perception, and alterations in this area have been linked to behavioral changes in language processing, as observed in conditions like schizophrenia [75].

Interestingly, the observed occipital hypometabolism in the study of acute catatonia associated with a medical condition has been suggested as a potential biomarker for NMDA-R encephalitis [76]. Given the limited number of studies conducted involving other medical conditions, it remains to be seen whether this pattern is unique to NMDA-R encephalitis or can be generalized to other medical conditions associated with catatonia.

The anterior cingulate cortex (ACC) is involved in error monitoring, conflict detection, and emotional regulation, while the medial cingulate cortex (MCC) is often considered the cognitive division of the ACC [77-79]. Additionally, the ACC associates potential actions with the value of their outcomes and prediction errors connecting directly to the hippocampus, PMC and the MCC; the latter of which has also extensive connections with motor-related areas (PMC and primary motor) [80, 81]. In studies on acute catatonia associated with a psychiatric disorder, increased rCBF was found in the MCC, and more severe catatonia cases were linked with higher perfusion in the bilateral ACC [27, 36]. In the study of acute catatonia associated with NMDAR encephalitis, hypometabolism was observed in the bilateral posterior cingulate cortex (PCC), suggesting a potential difference in the neural mechanisms underlying these subtypes [37]. In contrast, no results were reported in these areas in history of catatonia associated with a psychiatric disorder. Dysfunction in the ACC/MCC circuits may contribute to the symptoms of catatonia, such as perseveration and withdrawal as a conflict monitoring base for error, and may contribute to akinesia and mutism, particularly given their significant role in decision-making [52, 80, 82–84].

The right insula, involved in emotional processing, social cognition, behavior, sensorimotor integration, and awareness, showed focal hypermetabolism in acute catatonia associated with a medical condition and the right amygdala [37]. This hyperactivity could be linked to mannerisms, gegenhalten, and aberrant motor performance, illustrating the insula’s role in complex emotional and cognitive processes [52, 85]. Additionally, gray matter loss in the anterior insula and dorsal anterior cingulate across diagnoses of major psychiatric disorders has been found [86].

Cerebellar abnormalities

In relation to the cerebellum, in the acute catatonia associated with NMDAR encephalitis study by Kerik-Rotenberg et al., showed bilateral hypermetabolism in the cerebellum [37]. Congruently, Lefebvre at al. found that increased cortical inhibition was associated with increased resting state functional connectivity between the left motor cortex and the right cerebellum [40]. Hirjak et al. found that behavioral symptoms of catatonia were associated with aberrant intrinsic neural activity in cerebellar and prefrontal/cortical motor networks. The authors raise the concern for the possibility of overlap between catatonia in schizophrenia spectrum disorders and cerebellar cognitive affective syndrome [28]. The cerebellum’s role extends beyond motor coordination to include affective and cognitive functioning, influencing the ACC/MCC through the thalamic and cerebellar-ventral tegmental area (VTA) pathways [87-90]. Neurotransmitter imbalances in the cerebellum might lead to multidomain dysmetria, contributing to the psychomotor dysfunction observed in catatonia [91, 92].

In summary, the functional neuroimaging findings presented in this review highlight the involvement of multiple neural circuits and neurotransmitter systems in catatonia across varying clinical contexts. While certain brain regions appear to be commonly impacted, the specific patterns of dysfunction may differ based on the underlying etiology triggering the catatonic state. For example, acute catatonia associated with NMDAR encephalitis exhibited greater occipital and parietal lobe hypometabolism compared to psychiatric cases, which showed more frontostriatal abnormalities [37]. Patients with history of catatonia associated with a psychiatric disorder demonstrated OFC hypoactivity after lorazepam treatment [10, 30-32], whereas this was not a predominant finding in acute catatonia associated with a psychiatric disorder. The first example warrants careful consideration due to the limited sample size of the study, which fails to adequately represent the diverse spectrum of catatonia associated to a medical condition.

Our systematic review, when compared with the recent findings of Haroche et al., shows consistent results in functional neuroimaging of catatonia, including hyperactivity in motor regions like the SMA and hypoactivity in frontal areas such as the OFC. Both reviews also highlight frontal, temporal, and basal ganglia hypometabolism [19]. However, our review provided more nuanced insights by categorizing patients into acute catatonia associated with a psychiatric disorder, history of catatonia associated with a psychiatric disorder and acute catatonia associated with a medical condition. This categorized approach revealed more specific regional effects related to motor, cognitive, and affective symptoms within each subgroup. Differentiating catatonia based on medical versus psychiatric origins, while initially seeming reductionistic, is a crucial step towards comprehending this complex syndrome. Identifying distinct patterns of circuit and neurotransmitter dysfunction relative to the primary etiology can not only improve characterization but also aid in future diagnostics and prognostics. Understanding the intricate interplay between underlying illnesses and neural correlates can lead to more targeted and personalized treatment approaches, offering new avenues for managing catatonia.

Limitations and Future Directions

The current review has several key limitations. First, most studies have focused on patients with schizophrenia spectrum disorders, which may not generalize to catatonia caused by other underlying conditions. Furthermore, some of the studies on schizophrenia spectrum disorders have used the same cohort of patients. Second, even in studies of acute catatonia, many patients were already partially treated, which likely affected the results. Third, the inclusion of observational studies with varying methods precluded quantitative analysis, and the heterogeneity across studies limits clear conclusions. Fourth, the small number of available studies is insufficient to determine causal relationships between neural changes and specific catatonic symptoms or etiologies.

Additional limitations include the fact that catatonia often goes unrecognized in medical settings, so studies likely underestimate catatonia due to medical conditions [93]. Also, it remains unclear whether there are core neural deficits common to all catatonia, or if observed changes simply reflect abnormalities associated with each specific underlying disorder. Further research using standardized methods and assessments is needed to elucidate the relationships between symptoms, etiology, and neural correlates. In particular, longitudinal studies tracking changes over illness course could help delineate primary pathophysiological mechanisms from secondary effects.

Future catatonia research should prioritize standardizing neuroimaging and analysis methods to improve cross-study comparability and accommodate different analytical approaches. Furthermore, dividing patients into more homogeneous subgroups, expanding into other etiologies including other psychiatric disorders and medical conditions can help us better understand the neural underpinnings of catatonia and their relevance to clinical manifestations. Larger sample sizes are required to improve statistical power for detecting neuroimaging findings and their associations with symptoms, which can be accomplished through collaborative multicenter studies. Furthermore, looking into correlations between neural changes, clinical progression, and treatment responses has the potential to reveal fundamental neurobiological mechanisms in catatonia.

Conclusion

This systematic review underscores the complexity of catatonia, a multifaceted neuropsychiatric syndrome with diverse underlying etiologies. The included studies emphasize the convergence of dysfunctions in key neural circuits, including the cortico-striatal-thalamic loop system and prefrontal-parietal cortices, contributing to the range of motor, behavioral, and autonomic symptoms observed in catatonia. We also identified distinct patterns of circuit dysregulation in catatonia, offering potential avenues for personalized treatment approaches based on the underlying pathophysiology. Catatonia associated with medical conditions and psychiatric disorders may represent distinctive phenotypes, may require specific therapeutic and diagnostic interventions and further research. While further research is needed to solidify these insights and address limitations, such as the predominance of schizophrenia spectrum illness in the literature, this study summarizes and organizes catatonia’s neurobiological underpinnings, emphasizing the need for standardized methods, larger sample sizes, and longitudinal studies.

Highlights.

• Catatonia is associated with multiple illnesses beyond psychiatric disorders.

• Functional neuroimaging findings indicate multiple circuits dysfunction.

• Most of functional neuroimaging research in catatonia has focused on schizophrenia.

• Catatonia in medical and psychiatric disorders may show distinct dysfunction patterns.

Data availability

The details of the search strategy and the studies that were included can be found in Appendix Table 1 and Table 1, respectively.