Article Text

Download PDFPDF
Self-injurious behaviour in movement disorders: systematic review
  1. Jan-Frederik Fischer1,
  2. Tina Mainka1,2,
  3. Yulia Worbe3,
  4. Tamara Pringsheim4,
  5. Kailash Bhatia5,
  6. Christos Ganos1
  1. 1 Department of Neurology, Charité University Hospital Berlin, Berlin, Germany
  2. 2 Berlin Institute of Health, Berlin, Germany
  3. 3 Department of Neurophysiology, Saint-Antoine Hospital, Sorbonne Université, Paris, France
  4. 4 Department of Clinical Neurosciences, Psychiatry, Pediatrics and Community Health, University of Calgary, Calgary, Alberta, Canada
  5. 5 Department of Clinical and Movement Neurosciences, Queen Square Institute of Neurology, University College London, London, UK
  1. Correspondence to Dr Christos Ganos, Neurology, Charite University Hospital Berlin, Berlin 10117, Germany; cganos{at}


Self-injurious behaviours (SIBs) are defined as deliberate, repetitive and persistent behaviours that are directed towards the body and lead to physical injury and are not associated with sexual arousal and without suicidal intent. In movement disorders, SIBs are typically associated with tic disorders, most commonly Tourette syndrome, and neurometabolic conditions, such as classic Lesch-Nyhan syndrome. However, beyond these well-known aetiologies, a range of other movement disorder syndromes may also present with SIBs, even though this clinical association remains less well-known. Given the scarcity of comprehensive works on this topic, here we performed a systematic review of the literature to delineate the spectrum of movement disorder aetiologies associated with SIBs. We report distinct aetiologies, which are clustered in five different categorical domains, namely, neurodevelopmental, neurometabolic and neurodegenerative disorders, as well as disorders with characteristic structural brain changes and heterogeneous aetiologies (eg, autoimmune and drug-induced). We also provide insights in the pathophysiology of SIBs in these patients and discuss neurobiological key risk factors, which may facilitate their manifestation. Finally, we provide a list of treatments, including practical measures, such as protective devices, as well as behavioural interventions and pharmacological and neurosurgical therapies.

  • movement disorders
  • neuropsychiatry
  • tourette syndrome
View Full Text

Statistics from


The term self-injurious behaviour (SIB) encompasses deliberate, typically repetitive and persistent behaviours directed towards the body, leading to demonstrable physical injury in the absence of suicidal intent or sexual arousal.1 SIB is typically observed in neurodevelopmental and neuropsychiatric disorders, often with variable degrees of intellectual disability, but it may also occur after exposure to psychotropic substances2 3 and during psychotic experiences.4 Estimates of SIB prevalence vary, considering the broad range of aetiologies from 4% in the general population to more than 50% in different genetic syndromes.5 6 It may affect any body part and can manifest with marked variability in frequency, duration and intensity.1 Importantly, SIB may often be associated with serious medical, including life-threatening, complications.7

Within the field of movement disorders, SIB is most commonly associated with tic disorders, notably Tourette syndrome (TS), and is considered characteristic for patients with classic Lesch-Nyhan syndrome (LNS).8 9 However, beyond these well-known aetiologies, a range of other movement disorder syndromes may also present with SIB, even though this clinical association remains less well known. Importantly, in some of these patients, SIB may often outweigh the disability induced by the movement disorder, and hence early identification and management of SIB may constitute a top priority. There have been no papers to date to systematically address this topic, both on the range of conditions and their management. Therefore, we performed a systematic survey of the literature in order to provide a comprehensive list of movement disorder aetiologies associated with SIB, identify specific neurobiological risk factors in these disorders and discuss the available treatment strategies. We also complement these data with further cases from our own consultations and document characteristic examples covering a wide range of SIB in different movement disorders. The main goal of this review was to inform movement disorder specialists, general neurologists and neuropsychiatrists on this important issue and to guide diagnostic reasoning and treatment strategies when patients present both with movement disorders and SIB.


A systematic literature search in PubMed (MEDLINE 1966–2019) and Embase (1947–2019) was performed in May 2019 based on Medical Subject Headings terms (PubMed) and corresponding keywords (Embase): ((“self-injurious behaviour” OR “self-mutilation”) AND (“movement disorders” OR “dyskinesia” OR “myoclonus” OR “psychomotor disorders” OR “dystonic disorders” OR “tic disorder” OR “parkinsonian disorders” OR “hyperkinesis” OR “stereotypic movement disorder” OR “stereotyped behaviour” OR “ataxia” OR “tremor” OR “athetosis” OR “basal ganglia diseases” OR “motor skills disorder” OR “spasticity”)) NOT (“risk-taking” OR “dangerous behaviours” OR “sexual dysfunctions, psychological” OR “suicide” OR “suicide, attempted” OR “body modification, non-therapeutic” OR “suicidal ideation”). Preset filters were applied for the categories “English language” and “human”. Data extraction and study selection were performed by J-FF and TM. Studies were screened by title and abstract for clinical information combining basic patient characteristics like age, sex, number of patients and comorbidities and the presence of SIB, together with any kind of movement disorder. As recently proposed, SIB was defined as typically repetitive, ‘persistent, non-accidental and not-socially accepted behaviour resulting in demonstrable, self-inflicted physical injury, without intent of suicide or sexual arousal’.1 10

In the first screening step, clinical studies (eg, cross-sectional and case–control studies and controlled trials), case series and case reports were selected. Reviews were also included for cross-referencing. In vitro and animal research studies, as well as duplicates, comments and meeting protocols were excluded. All selected and available studies were then read in full text. Only clinical studies and cases were included, where a clear association between the presence of SIB and movement disorders was demonstrated. Cases in which signs of suicidality or sexual arousal were associated with SIB were excluded, as were reports with insufficient clinical information (eg, inconclusive clinical data, including age, clinical features and number of patients). Patients with SIB and functional movement disorders, including dystonia–causalgia syndrome, were not included in our study selection. The process of literature search, which was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement, can be seen in figure 1.11

Figure 1

Flow diagram of data acquisition.

From all selected articles, we extracted the following data: patient age, sex, method of SIB assessment, SIB onset, SIB forms and topography, presence of movement disorders and their aetiology, as well as neuropsychiatric comorbidities and different treatment strategies. Operational terms to cluster different SIB forms were used based on previous literature (see table 1).12 Repetitive amputational attempts, as well as the deliberate ingestion of dangerous liquids without suicidal intention, were also captured as SIB. Movement disorder aetiologies associated with SIB were grouped into five categories, as follows: ‘neurodevelopmental disorders’, ‘neurometabolic disorders’, ‘neurodegenerative disorders’, ‘disorders with characteristic structural brain changes’ and ‘others’, which includes cases in which their aetiologies did not clearly belong to any previous category, or where no specific aetiology was given. Treatments were captured and reported if they led to a significant reduction or cessation of SIB.

Table 1

Operationalised forms and corresponding definitions of SIB for standardised counting of different descriptions (exemplary SIB forms) registered in the literature


Aetiological overview

In total, 137 studies (individual studies and corresponding references provided as online supplementary table 1, list of supplementary references in online supplementary file) reporting 1407 patients with SIB within 48 aetiological entities were identified with our systematic literature review (see also figure 2). SIB assessment methodology was reported in only 28/137 reports (structured or semistructured interviews and direct observational data). Illustrated examples of SIB from our own clinics are shown in figure 3.

Supplementary data

Supplementary data

Figure 2

Comprehensive diagram of all captured movement disorder aetiologies associated with SIB. 1Associated with intellectual disability and/or autism spectrum disorder; 2associated with ST3GAL5 genetic mutation; 3associated with G8363A transfer RNALys genetic mutation; 4patients from our clinic; 5associated with B4GALNT1 genetic mutation. CBS, corticobasal syndrome FBXO7, F-box Protein 7, GLHS, Gomez-Lopez-Hernandez syndrome GNAO1, G Protein Subunit Alpha O1, H-ABC, hypomyelination with atrophy of the basal ganglia and cerebellum; HSP, hereditary spastic paraplegia; MLC, megalencephalic leucoencephalopathy; NBIA, neurodegeneration with brain iron accumulation; NMDA, N-methyl-D-aspartate; PKAN, pantothenate kinase-associated neurodegeneration; PKU, phenylketonuria; PSEN1, Presenilin 1; PTPS, 6-pyruvoyl-tetrahydropterin synthase; RTTN, Rotatin; SCA17, spinocerebellar ataxia type 17; SIB, self-injurious behaviour; TS, Tourette syndrome.

Figure 3

Clinical images of self-inflicted injuries and SIB-related scars. (A) Scabs of the right ear canal after repetitive self-hitting, which could be prevented by bilateral arm splints (B), in a patient with stereotypic movement disorder and trisomy 21. (C,D) Scars and excoriation wounds caused by skin picking and self-scratching with stereotypic movement disorder (not otherwise specified) (C, left forearm; D, right lower leg). (E,F) Occipital loss of hair due to compulsive hair pulling (E) and multiple bulging scars at the right back of the hand (F) due to self-biting in a patient with serine deficiency. (G,H) Repetitive impacts on the nose and right forehead due to compulsive self-hitting with both fists caused a scarred bulge at nasal bridge and a swollen haematoma in the right frontal area in a patient with Tourette syndrome (G). Mechanical stress resulted in callused marks on both hands in the area of the second metacarpophalangeal joint (H, left hand shown). SIB, self-injurious behaviour.

Neurodevelopmental disorders

The majority of identified patients with SIB and movement disorders (n=1198) belonged to the group of neurodevelopmental disorders. Among those, most patients had a diagnosis of TS (n=718)7 13–18 [S1–S25], with SIB prevalence ranging between 14% and 66%15 [S5]. SIB manifestation followed the onset of tics with a lag of several years and an average age at onset around 10 years (range 4–18 years)14 16 18 [S6, S7, S10, S12, S14–S16, S20, S21, S24]. SIB in TS typically involved the head, including the eyes, lips, teeth and intraoral soft tissue. Most common SIB in TS involved skin picking and self-biting or self-hitting behaviours7 14 16–18 [S1, S6–S25]. Isolated cases presented with severe SIB, including amputation attempts, severe eye poking or head banging, self-strangulation, as well as drinking dangerous liquids7 [S1, S14, S18].

Fragile X syndrome (FXS, n=143 SIB cases; four studies) was the second most common aetiology in this category, typically presenting with stereotypic behaviour and SIB [S26–S29]. Where information was available, the onset of SIB was reported to occur in childhood [S28]. Self-hitting or self-biting in patients with FXS was the predominant SIB [S27, S28]. Motor stereotypies were also the main movement disorder in patients with Cri du chat syndrome (n=67 SIB cases) who also presented with SIB. There was marked variability regarding the type of SIB in these patients, including self-hitting, self-biting, banging against objects, forced vomiting, putting objects/fingers into body orifices or self-luxation of shoulder and patella [S30–S32]. Other aetiologies included Smith-Magenis syndrome (n=37 SIB cases) [S32–S34], Lowe syndrome (n=36 SIB cases) (online supplementary file 1), Cornelia de Lange syndrome (n=30 SIB cases),12 Rett syndrome (n=3) [S36, S37], G Protein Alpha Subunit O1 (GNAO1) encephalopathy (n=2) [S38], Prader-Willi syndrome (n=1) [S39], Angelman syndrome (n=1) [S32], salt-and-pepper syndrome (n=2) [S40], mitochondriopathy (n=1) [S42], primrose syndrome (n=1) [S43] and chromosomal aberrations like Tourettism associated with Xq25 microduplication (n=1) [S44], 17q23.2 duplication (n=1) [S45], 1q24-q25 microdeletion (n=1) [S46], 18q deletion syndrome (n=1) [S47], trisomy 13 (n=1) [S32] and trisomy 21 (n=1 case from our clinics; also see figure 3).

We also identified 149 cases with probable neurodevelopmental disorder in whom motor stereotypies and intellectual disability were characteristic, but further clinical details, beyond the presence of autistic features in some of them, were sparsely provided (mean age: 33.9 years)19 [S48–S50]. In these cases, the most commonly reported forms of SIB were self-hitting, self-biting, self-picking and banging behaviours [S48–50].

To summarise, a range of different neurodevelopmental conditions may present both with movement disorders and SIB. Within this category, the top three aetiologies include tic disorders and TS, FXS and Cri du chat syndrome. An association of movement disorders with intellectual disability is characteristic.

Neurometabolic disorders

In total, 168 patients with SIB and movement disorders were included in this category. LNS was the predominant aetiology (n=162)20–25 [S51–S84], including two female patients21 [S61]. SIB in LNS was characteristic but did not occur in all patients with LNS [S77]. The somatotopy of SIB in LNS was more pronounced for the mouth (n=63) and fingers/hands (n=57)21–24 [S51–S55, S57–S84], with characteristic self-biting of lips or fingers first manifesting around the age of 3 years (n=85)20 21 [S51, S53, S58–S65, S68–S73, S75, S76, S81–S83]. Only one study provided long-term follow-up data of patients with LNS, in whom a fluctuating course of SIB was observed, with a slight decrease around the age of 10 years [S72].

Four more cases from the literature were identified with the following aetiologies: 6-pyruvoyl-tetrahydropterin synthase deficiency [S85], late-treated phenylketonuria (PKU) [S86], Gaucher’s disease type II (n=1) [S41], and hepatolenticular degeneration [S87]. In the case of a man in his 70s with PKU, fluctuations in his serum phenylalanine levels correlated with the frequency and intensity of SIB (ie, banging different body parts against hard objects) [S86]. We also report two siblings from our clinics with serine deficiency syndrome and compulsive hair pulling, as well as self-biting (see also figure 3E).

Neurodegenerative disorders

Eleven patients (seven men) between 29 and 54 years with neuroacanthocytosis (NAC) were identified [S88–S97]. Patients with NAC typically exhibited self-biting behaviours, except one case with self-injurious repetitive head scratching [S88]. The mean onset of SIB was 35.7 years (reported in six cases) [S89–S92, S95]. Self-biting was mostly localised in the oral area (tongue/oral mucosa/cheeks/mouth/lip/teeth) (n=10/11 patients).

SIB was additionally documented in two patients with pantothenate kinase-associated neurodegeneration and one patient with neurodegeneration with brain iron accumulation (NBIA) (n=3, mean age: 9.3 years, 1 male; age of SIB onset: 11 years (n=1)) [S98–S100], two cases with hereditary spastic paraplegia due to mutations in the B4GALNT1 gene (n=2, adult age, not further specified) [S101], one patient with dementia and myoclonus diagnosed with Alzheimer’s disease due to Presenilin 1 (PSEN1) gene mutation (n=1, 35 years, male) [S102] and one patient with corticobasal syndrome associated with the presence of antiglycine receptor antibodies (n=1, age: 55 years, male) [S103]. SIB was also documented in a patient with juvenile parkinsonism due to an F-box Protein 7 (FBXO7) mutation syndrome under rasagiline therapy (1mg/d) (n=1, male, age: 16y) [S104]. More recently, two females with SCA17 and severe repetitive self-scratching were also reported [S105].

Characteristic structural brain changes

SIB was also described in seven syndromes with characteristic structural brain changes (10 patients, 6 male), including three patients diagnosed with Gomez-Lopez-Hernandez syndrome (n=3, age range 3–11 years (n=1, age not reported), 1 male) [S106–S108] and two patients with Joubert syndrome (mean age: 5 years, 1 male) [S109]. Further isolated cases included the following aetiologies: Rotatin (RTTN) missense mutation (5.8 years, male) [S110], hypomyelination with atrophy of the basal ganglia and cerebellum (n=1, 13 years, female) [S111], megalencephalic leucoencephalopathy (n=1, 10 years, male) [S112], rhombencephalosynapsis (n=1, 22 years, male) [S113] and Chiari malformation type II (n=1, age: 2.8 years, male) [S114]. Brain structure abnormalities in these disorders affected the cortical gyri, subcortical fibres, basal ganglia, corpus callosum, cerebellum, pons, brainstem, cisterna magna and fourth ventricle. SIB onset in these cases was variable from 3 to 19 years (reported in three studies) [S107, S110, S113] and included self-biting, compulsive self-touching or rubbing behaviour leading to skin injuries [S107, S112, S113]. Interestingly, in the case with Chiari malformation type II, SIB onset coincided with the occurrence of quadriplegia and an inability to communicate [S112].


SIB was further reported in anti-N-methyl-D-aspartate receptor encephalitis (n=1, age: 34 years, male) [S115], secondary to meningitis in infancy (n=1, age: 14.2 years, female) [S116], secondary to cerebral damage due to a traffic accident (n=1, age 12 years, male) [S117], following perinatal brain damage (cerebral palsy: n=1, age: 1.9 years; hypoxia at birth: n=1; age: 2.5 years) [S32] and in the context of hydrocephalus (n=1, age: 16 years, male) [S118]. Drug-related SIB was also reported in two cases (increase of ropirinole dosage in Parkinson’s disease: n=1; age: 50 years, male; subsequent to intramuscular flupentixol injection: n=1, age: 21 years, male)26 [S119]. One case with oromandibular dystonia (n=1, age: 1.4 years, female) [S120] and one case with stereotypic movement disorder in the form of repetitive head banging in the absence of intellectual disability, and no further clinical specifiers were also identified.27

Pathophysiological insights and risk factors

Although the exact pathophysiology of SIB remains elusive and a range of intrapersonal and interpersonal factors should be considered in the scientific analysis of why an individual engages in SIB, the pattern of results we report supports the presence of three key risk neurobiological factors for the manifestation of SIB in patients with movement disorders (see also figure 4). First, for the majority of syndromes presenting with SIB (25/48 aetiological categories), including TS, FXS, Smith-Magenis syndrome, primrose syndrome or RTTN mutation, restricted behavioural patterns were characteristic. This included not only the movement disorder itself (eg, tics or stereotypies), but also other clinical features ranging from mental rumination, to obsessive–compulsive behaviours or even full-blown obsessive–compulsive disorder (OCD). For example, the diagnosis of the neurodevelopmental Cornelia de Lange syndrome in itself was not associated with SIB; however, the occurrence of restricted behaviours like stereotypies or compulsions in these patients increased the risk of SIB.12 In TS, not only the severity and complexity of tics but also the presence of OCD was associated with SIB,13 and indeed, in the majority of individually documented patients with TS and SIB we identified, OCD coexisted with tics. Although not all studies in TS could replicate this association,28 obsessive thoughts may include pathological mental rumination related to non-purposeful SIB. Obsessional worries about self-injury and the subsequent feeling of relief after the execution of SIB may, therefore, support the maintenance of a vicious reinforcement circle.14 In general, restricted patterns of behavioural output, for example, the continuous presence of stereotypic movements, coupled with behavioural inflexibility (eg, the need to repeat certain actions), is a relevant risk factor for the presence of SIB in movement disorders.

Figure 4

Key risk factors for the manifestation of SIB in movement disorders. Behavioural inflexibility, intellectual disability and pathological reinforcement learning constitute key risk factors leading to increased susceptibility for SIB in patients with movement disorders. *Other neurotransmitter systems, for example, the opioid or serotonergic system, might also be affected. SIB, self-injurious behaviour.

Restricted behavioural patterns are closely linked with SIB and overall highly prevalent in people with neurodevelopmental disorders and particularly with intellectual disability.29 30 Specifically, SIB is overall common in populations with intellectual disability.31 Indeed, in about half of the aetiologies we identified, as, for example, FXS, Cri du chat syndrome and Rett syndrome, the presence of intellectual disability was characteristic. There is a large body of literature on this topic, and several theories regarding pathophysiological implications have been proposed.32 A clear link between the level of intellectual disability (defined as severe/profound vs mild/moderate intellectual disability) and the manifestation of SIB has been previously reported.33 Factors including both the processing of sensory input, including pain thresholds and nociceptive signalling, as well as the capacity to communicate and express these experiences, might promote SIB, at least in a subgroup of these patients.34 The additional presence of autistic features, which further limit expressional capacity, was further associated with SIB.33

Another hypothesis refers to the control of noxious stimuli, involving the endogenous opioid system. For example, acute pain following self-inflicted injuries could result in activation of inhibitory pathways for noxious stimuli and alleviate chronic pain.35 Finally, the role of impulsivity should also be mentioned in this context, as indeed several studies have documented its significance as a predictor in the manifestation of SIB in different clinical populations.13 36

The formation and habitual maintenance of SIB, despite its detrimental effects on the body, also point towards aberrant reinforcement of pathological behavioural output.

Sources of aberrant reinforcement might be traced back to a specific individual’s environment (interpersonal variables, eg, social attention, escape from task demands or specific stressors) or to intrapersonal conditions (eg, need for sensory stimulation) [S121, S122]. Although the interplay between environment and neurobiology is complex, the main neurotransmitter underlying repetitive behaviours is dopamine. It is, therefore, not surprising that many of the movement disorder aetiologies we identified have documented abnormalities of dopaminergic neurotransmission, which we believe constitute another risk factor for SIB [S123]. For example, pathological hyperactivity of tonic and phasic dopaminergic expression in TS is suggested to be a key element in the formation of tic behaviours.37 Importantly, in TS, there is evidence of an association between SIB and increased dopamine transporter capacity in brain areas, including the striatum.15 In LNS, there is evidence of selective dysfunction of dopaminergic pathways. Although in these cases a reduction, rather than an increase of dopamine in the basal ganglia, is characteristic,38 aberrant activation of dopaminergic receptors may lead to SIB. Neonate rats depleted of striatal dopamine following administration of 6-hydroxydopamine do not exhibit profound impairment of motor function compared with rats lesioned during adulthood but become hyperactive on introducing a D1-receptor agonist and characteristically develop SIB when treated with levodopa.39 Not surprisingly, therefore, SIB was exceedingly rare in adult patients with neurodegenerative parkinsonism, and we were able to identify only one case of a patient in his 50s with Parkinson’s disease who developed skin-picking behaviour following dose increase of ropinirole, which dissipated on discontinuation of medication.26 In line with this are also the aetiologies with characteristic basal ganglia pathology, such as NBIA [S99, S100], NAC [S91, S97], hepatolenticular degeneration [S87], as well as primrose syndrome associated with calcifications in the basal ganglia [S43]. Moreover, invasive neuromodulation, such as deep brain stimulation (DBS) for the treatment of SIB, targeted the basal ganglia, including the internal pallidum (GPi) and the nucleus accumbens [S19, S75]. Of note, abnormalities of other neurotransmitters, such as the opioid system and other monoamines, such as serotonin and norepinephrine, have also been suggested to contribute to the formation and recurrence of SIB. For example, analysis of cerebrospinal fluid metabolites in patients with LNS with SIB yielded additional changes in serotonin and norepinephrine turnover.22 The effects of pharmacological targeting suggest also the involvement of the opioid system. Both a step-up of a beforehand reduced oxycodone therapy and antiopioid effects of naloxone or naltrexone were reported to reduce SIB.16 17 27


Treatment of SIB in movement disorders is complex and requires different, often multiple and complementary approaches based on the type of SIB and the underlying aetiology (see also box 1). Crucially, data on the efficacy of the different therapeutic approaches are limited, and controlled trials remain an aspiration. Indeed, most of the information we gathered stems from case reports and case series, and only 15 reports documented the usage of standardised tools to monitor treatment response. Further, we identified only one randomised controlled trial in patients with LNS for ecopipam, a selective D1 receptor antagonist.23

Box 1

Exemplified treatment options for self-injurious behaviour in movement disorders.

Protective restraints

Mouth guards

Lip bumpers

Bite splints


Arm splint


Behavioural/psychosocial therapy


Differential reinforcement of other behaviours

Noncontingent reinforcement


Opioid agonists/antagonists (eg, oxycodone and naloxone)

Antipsychotics (eg, haloperidol, quetiapine and risperidone)

Botulinum toxin


Anticonvulsants (eg, gabapentin and pregabalin)

SSRI (eg, escitalopram)



Interventional/surgical therapy

Tooth extraction




  • DBS, deep brain stimulation; ECT, electroconvulsive therapy; SAM, S-adenosylmethionine; SIB, self-injurious behaviour; SSRI, selective serotonin reuptake inhibitor.

  • *In cases with Lesch-Nyhan syndrome.

Restraints or other protective devices are used to physically prohibit the patient from SIB. Examples of such devices and restraints are gloves or hand wrappings to prevent hand-biting or to protect from severe lesions by self-hitting18 [S74, S80], as well as splints to restrain excessive movements (see figure 3B) [S64]. Oral devices include tongue protectors, mouthguards/biteguards or lip bumpers [S41, S55, S60, S98]. Although some restraints may not limit the SIB itself, they may reduce its impact and avert more serious injuries, as in the case of helmets, which are used in patients with repetitive head banging.18 Limiting the impact of SIB through restraints may also support wound healing, and in some cases may even lead to complete cessation of SIB [S54, S114]. Interestingly, patients are not typically averse to the use of restraints or protective devices and may report feeling safer.14 For example, more than 90% of patients with LNS preferred using restraints.20

Behavioural therapy for SIB in movement disorders has mostly been reported in combination with other treatment modalities14 [S16]. There are two key elements in behavioural therapy of SIB. One is the recognition of triggers and environmental reinforcers of SIB [S122]. For example, pain or stressful situations are postulated as setting situations for self-injury.40 SIB in some cases is also understood as a form of communication (eg, to get attention).20 Once this occurs, SIB can become maintained socially motivated, for example, by attention contingent to the act of self-injury.40 The second element relates to the differential reinforcement of alternative responses, such as non-harmful behaviours. For example, parents of patients with LNS reported that talking to the patient in situations without SIB (‘when he is good’) were very helpful intervention techniques, while punitive measures were the least effective methods.20 In TS, behavioural therapy led to improved SIB, especially in patients with comorbid compulsive behaviours, for example, in squeezing a stress ball instead of self-scratching.14 Of note, some forms of SIB might persist even after behavioural therapy, even though with reduced frequency and/or intensity14 [S7, S16].

Pharmacological approaches for SIB are diverse. In the absence of randomised pharmacological controlled trials in patients with movement disorders and SIB, in most reports, we found the usage of a specific agent was guided by clinical judgement. These included antipsychotics (eg, quetiapine, haloperidol and risperidone) [S49, S52, S87, S96], anticonvulsants (eg, gabapentin and pregabalin) [S11, S69], SSRI (eg, escitalopram) [S87] and drugs acting on the opioid system. Interestingly, both opioid agonists (eg, oxycodone) and antagonists (eg, naloxone and naltrexone) have shown efficacy in treating patients with SIB.16 19 27 In some cases, as, for example, in dystonia-related SIB (eg, severe self-biting), local treatment with botulinum toxin (eg, in the masseter muscles to limit the impact of self-biting) has been reported to reduce the injurious impact of repetitive behaviours [S58, S62, S112, S120]. Baclofen successfully reduced both severity of dystonia and the frequency of SIB in three patients diagnosed with LNS.24

Should previous treatments be insufficient, for example, in patients with severe SIB (eg, directly life-threatening SIB or leading to serious medical complications), interventional approaches, such as DBS, might be required. In TS, DBS was reported as effective either as single treatment or combined with anterior capsulotomy (n=1) to reduce or completely ameliorate SIB (n=4; DBS targets: nucleus accumbens, GPi and thalamic centromedian–parafascicular complex)7 18 [S12, S19]. One case series reported five patients with TS and SIB who underwent thalamic (n=3) and pallidal (n=2) DBS for the unwanted behaviours with significant improvement of tics; however, the exact effects on SIB were not explicitly reported [S25]. Similarly, symptoms of patients with LNS, including SIB, improved after DBS (six patients; DBS targets: GPi (n=5), pallidal stimulation (n=1)) [S51, S57, S59, S75, S76, S83], NAC (one patient; DBS target: GPi) [S93]. Psychosurgery approaches such as cingulotomy and limbic leucotomy have also been conducted and led to reduction of SIB in three case reports (one report in combination with DBS)18 [S6, S20]. Electroconvulsive therapy (ECT) in two adults with TS improved both tics and behavioural abnormalities, including SIB, after several sessions (7–23 ECT sessions) [S14, S18]. Of note, further practical interventions, such as tooth extraction surgery, have been documented in patients with severe self-biting, as, for example, in LNS [S67, S78].

Finally, in some of the disorders we identified, treatments addressing the underlying aetiology may also improve SIB. For example, in neurometabolic disorders, such as PKU, the dietary regulation of phenylalanine intake was shown to improve SIB [S86]. Similarly, in patients with LNS, oral intake of S-adenosylmethionine, a dietary supplement which potentially restocks endogenous nucleotides, but not allopurinol, led to a remarkable reduction of SIB.21 25 Interestingly, some effectiveness in patients with LNS was also reported for probenecid, an uricosuric agent [S70].


Here we document the clinical association of SIB with different movement disorder aetiologies, which we present across distinct categorical domains, including neurodevelopmental, neurometabolic and neurodegenerative conditions, as well as disorders with characteristic structural changes, and those with heterogeneous aetiologies. Tic disorders and TS, FXS, Lesch-Nyhan syndrome and Cri du Chat syndrome were overall the most commonly reported aetiologies, with the predominant movement disorders being stereotypies and tics, followed by dystonia and/or chorea.

We acknowledge that our literature review may have not captured all possible movement disorders aetiologies and that reporting bias might have influenced the prevalence of SIB identified. At this time, the majority of the literature on SIB in movement disorders, particularly beyond tic disorders or LNS, consists mostly of case series and case reports with methodological heterogeneity, owing largely to the surprisingly rare usage of standardised assessment tools. We therefore were not able to establish specific patterns of SIB for the different aetiologies. We also could not identify population-based studies documenting the prevalence of SIB or prospective long-term cohort studies of risk factors associated with the development and clinical features of SIB over time. Similarly, beyond a single randomised control trial on the usage of ecopipam in LNS, there were no further head-to-head trials testing different therapeutic approaches that would allow differentiating existing treatments by classes of evidence.

However, our extensive review showed that in almost all disorders, the likelihood of SIB occurrence was high in the presence of behavioural inflexibility with restricted and repetitive behavioural patterns, intellectual disability and aberrant reward processing, potentially related to abnormal dopaminergic neurotransmission. We therefore encourage clinicians to use standardised instruments and actively assess for the presence of SIB in patients with movement disorders, as treatment implications are important, even though the quality of existing evidence for the different management practices in all these disorders is unfortunately low. Pragmatic approaches, as restraints and protective devices, as well as behavioural therapies, should be first-line considerations. Pharmacological interventions may be helpful in some patients, including treating the underlying aetiology, particularly in certain neurometabolic disorders. Invasive treatments, such as DBS, may further be of use in cases with severe and treatment refractory SIB.


View Abstract


  • Twitter @YuliaWorbe, @ChristosGanos

  • Correction notice This article has been corrected since it appeared Online First. Minor text changes have been made, and Figure 2 replaced.

  • Contributors Study design: CG, J-FF, TM and TP. Study supervision: CG. Study selection: J-FF, TM and TP. Data extraction: J-FF and TM. Initial draft: CG and J-FF. Revision of the draft for important intellectual content: J-FF, TM, YW, TP, KB and CG. Interpretation of results: J-FF, TM, YW, TP, KB and CG.

  • Funding This study was funded by Volkswagen Foundation.

  • Competing interests TM was supported by the BIH-Charité Clinician Scientist Program of the Charité-Universitätsmedizin Berlin and the Berlin Institute of Health. YW was supported by an Agence National de Recherche grant, Dystonia Medical Research Foundation and a travel grant from Merz. KB received grant support form EU 2020 horizon, National Institute of Health Research/Research for Patient Benefit, Wellcome/MRC and PD UK, and received honoraria/financial support to speak/attend meetings from Ipsen, Merz, Sun Pharma, Allergan, Teva Lundbeck and UCB. KB received royalties from Oxford University Press and a stipend for Movement Disorders Clinical Practice editorship. CG received research grants from the VolkswagenStiftung (Freigeist Fellowship) and the German Parkinson Society and was also supported by the Deutsche Forschungsgemeinschaft (GA2031/1-1 and GA2031/1-2).

  • Patient consent for publication Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.