Self-injurious behaviour in people with intellectual disability

What is self-injurious behaviour?

Self-injurious behaviour can be seen in a number of different populations and in both children and adults. People who are depressed may attempt suicide and a very small minority of people experiencing psychosis may injure themselves in response to the auditory hallucinations 1. Selfinjury is also seen in penal institutions (see Box 'Self-injury in penal institutions') and delicate cutting with glass and blades is sometimes shown by young women in association with borderline personality disorder. self-injurious behaviour in people with intellectual disability tends to be a different than that shown by these populations.
Self-injury in penal institutions

In her book Gulag: A history,40 Anne Applebaum describes the practice of samorub (self-mutilation) in the notorious Soviet concentration camps.

 “Some of the methods were crude. Criminals in particular were famous for simply cutting off their three middle fingers with an axe, so they could no longer cut tress or hold a wheelbarrow in the mines. Others cut off a foot, or a hand, or rubbed acid into their eyes. Still others, upon departing for work, wrapped a wet rag around one foot: in the evening they returned with third degree frost-bite.”

It tends to be cruder in terms of the act itself (e.g. self-biting or head banging), does not normally involve the use of objects and is not an attempt at suicide. In the research literature a number of different terms have been used to describe these behaviours in people with intellectual disability including self-mutilation, automutilation, autoplexy, self-harm and self-abuse. The term selfinjurious behaviour is preferred because it can include behaviours which do not necessarily result in mutilation, for example face slapping, and is more specific than terms such as self-harm and self-abuse.

Inevitably there is some debate as to what might be included under the term self-injurious behaviour and consequently some definitions have been developed. Perhaps the most useful is that given by Tate and Baroff3: “Self-injurious behaviour does not imply an attempt to destroy, nor does it suggest aggression; it simply means a behaviour which produces physical injury to the individual’s own body… Common types of selfinjurious behaviour are forceful head-banging, face slapping, punching the face and head and scratching and biting one’s own body”. However, even when this definition is applied to exclude less severe behaviours it is clear that some behaviours that should be included, such as face slapping that leads to reddening of the face but not necessarily tissue damage, are excluded. However, the definition does have the benefit of not including more mild behaviours that are not really injurious.

Other terms that are associated with self-injurious behaviour tend to refer to different aspects of behaviour. The term “challenging behaviour” has replaced those of the “problem behaviour” and “behaviour disorder” and reminds us that these behaviours are a challenge to service providers. self-injurious behaviour is considered as one form of challenging behaviour along with other behaviours such as aggression and damaging the environment for example. Another term, stereotyped behaviour usually refers to meaningless, repetitive behaviours that have no immediate goal, such as rocking, hand-waving and spinning. Generally, these are different from self-injurious behaviour as they do not result in injury.

Finally, the term compulsive behaviour is used to describe behaviours that appear to be, to some extent, out of the control of the individual and the individual appears driven to show the behaviours. The reason for mentioning these two types of behaviour is that self-injurious behaviour is sometimes referred to as stereotyped or compulsive. This tends to mean that the self-injury is either repetitive, meaningless and not goal directed or that the self-injury appears to be out of the individual’s control and the person appears driven to show the behaviour. (Compulsive self-injurious behaviour is discussed in Chapter 5). It should be noted that these terms are often used carelessly and it is advisable to be cautious about inferring anything about the cause of self-injurious behaviour from their use.

How common is self-injurious behaviour?

Within the population of people with intellectual disability the prevalence of self-injurious behaviour varies depending on how the study was conducted. Differences in the definitions that have been used, the time period within which the behaviour should occur to be counted, whether the information was collected by questionnaire or observation and the population that is considered will all influence the final estimate. When the largest studies are considered and studies are only included if they have a similar definition, time period and methodology and they include people with intellectual disabilities in a given geographical area (rather than just those in hospitals for example) then the prevalence of self-injurious behaviour is estimated to be 4-10%4. However, it is very clear that the prevalence of self-injurious behaviour is related to individual characteristics and we have begun to think of these as risk markers for self-injurious behaviour.

Forms of self-injurious behaviour

Topographies of Self-Injurious Behaviour

There are many topographies of self-injurious behaviour and one individual may show several different topographies. The graph above shows topographies of self-injurious behaviour seen in three prevalence studies of groups of people with intellectual disabilities.

Forms of self-injurious behaviour

Self-injurious behaviour in people with intellectual disability can take a variety of forms. The most common are scratching or picking, biting and head hitting or banging (see Box 3.2) and there is enormous variability in the severity of self-injurious behaviour. Behaviours can be very mild in that the act does not incur immediate damage e.g. a soft face slap, or very severe such that the single act can cause significant injury e.g. a head bang to the sharp corner of the table. How frequently the behaviour occurs is also extremely variable, from a small number of incidents in a month to behaviours which can occur many times in an hour. (see Box 3.3) Additionally the temporal pattern of self-injurious behaviour can vary. Some self-injurious behaviour occurs in discreet bursts i.e. a bout of headbanging may be limited to 10 or 20 headbangs in a day, all occurring within the space of a minute or two. Alternatively, the behaviour may occur at a low level throughout a day, every few minutes.

How frequent is self-injurious behaviour


Frequency of Self-Injurious Behaviour

Oliver et al. (1987)18 carried out a total population survey of self-injurious behaviour in individuals with an intellectual disability in one health region of the UK. Six hundred and sixteen adults and children were found to have engaged in self-injurious behaviour sufficient to have caused tissue damage in the previous four months. 596 of these were screened and the frequency of self-injurious behaviour recorded in this group is shown in the graph above.

Another way in which the pattern may vary is that self-injurious behaviour may be problematic for weeks or months and then disappear or occur at a much lower level for a similar period. This pattern may be repeated over time. Whilst the importance of these different patterns is at present unclear, it is possible to speculate on what these may tell us about the reasons for self-injurious behaviour and this is considered in more detail in Chapter 6.

Risk markers for self-injurious behaviour

The strongest risk marker for self-injurious behaviour is the degree of intellectual disability and this finding has been replicated across numerous studies. It is now very clear that the greater the degree of intellectual disability then the higher the prevalence of self-injurious behaviour (see Box 3.4). 

How is the prevalence of self-injurious behaviour influenced by intellectual disability?

Several studies have examined the effect of degree of learning disability on the prevalence of selfinjurious behaviour. The studies outlined in the table show that the prevalence of self injury increases as the degree of learning disability becomes more severe.

We have conducted a recent analysis of all prevalence studies and this indicates that the probability of showing self-injurious behaviour in children and adults who have severe or profound intellectual disability is more than four times higher than for those who have a moderate or mild intellectual disability5. Another important risk marker is the presence of a genetic syndrome.  Children and adults who have Lesch-Nyhan Syndrome almost always show self-injurious behaviour, primarily by biting their fingers and lips6. This one-to-one relationship between a genetic syndrome and selfinjurious behaviour is not found in any other genetic syndrome. However, it is clear that the prevalence of self-injurious behaviour is much higher than chance in Prader-Willi7, Cri du Chat 8 and Smith-Magenis9 syndromes for example (see Box 3.5). For some of these syndromes it is clear that it may be the risk factor of severe or profound intellectual disability that contributes to the higher prevalence e.g. Cri du Chat Syndrome. However, for others this risk factor is not present and an alternative explanation needs to be sought e.g. Prader-Willi and Lesch-Nyhan Syndromes. The extent to which there is an association between self-injurious behaviour and Cornelia de Lange Syndrome has been the subject of some debate. Our recent research has focused on this question and this is discussed in more detail in Chapter 4.

Prevalence of self-injurious behaviour in genetic syndromes

Self-injurious behaviour is common in several genetic syndromes, including Cornelia de Lange. Data from various studies are shown above. Each datum point represents the result of one study of a syndrome. As can be seen, the prevalence rate varies across studies. The prevalence rate of self injury in individuals with intellectual disability who do not have a syndrome is 4-10%. In all cases above the prevalence is much higher, up to eight times higher in Lesch-Nyhan syndrome and two to four times higher in Cornelia de Lange syndrome (see Chapter 4).

There is some evidence that autism is associated with self-injurious behaviour10. However, some of the criteria for diagnosing autism do include the presence of self-injurious behaviour. It is not clear therefore, whether the diagnosis of autism has been given because an individual shows self-injurious behaviour or because there is genuinely an association between autism and self-injury. Autism is characterised by the socalled “Triad of Impairments” namely, impairments of socialisation, communication and imagination accompanied by the presence of repetitive and restricted behaviour11 (see Box 2.7). The potential significance of an association between autism and self-injurious behaviour may lie with neurotransmitter disturbance that is assumed to occur in autism or the commonly associated intellectual disability. Additionally, it has been argued that individuals who have autism may show repetitive behaviours because there is an impairment of executive function12 and thus limited control of the initiation of behaviours and the termination of behaviours (see Box 3.6). This is put forward as an explanation for the repetitive behaviours that are observed in autism such as stereotyped, compulsive and ritualistic behaviours and, for some people self-injurious behaviour. There may be another significance to the association between autism and self-injurious behaviour in that stereotyped behaviours, which are common in autism, may precede the development of self-injurious behaviour (see Chapter 5)

What is executive function?

 Executive functions are mental processes that help us to control our behaviour. There are several features of executive functions, for example: planning, holding a mental representation in short term memory, and being able to inhibit an inappropriate response. This latter feature has led some authors to link problems with executive function-executive dysfunction to attention-deficit-hyperactivitydisorder (ADHD) as individuals with ADHD find it difficult to inhibit ongoing responses and have difficulty maintaining attention. It has also been recently proposed that executive dysfunction may play a role in some of the deficits of functioning seen in children and adults with autism. This may in part explain the repetitive and inflexible behaviours shown by many individuals with autism.

It has been argued that an expressive communication deficit is a risk marker for self-injurious behaviour14. More specifically, it is suggested that if expressive communication is significantly poorer than receptive communication then the risk is increased. However, it is extremely difficult to disentangle poor expressive communication from the degree of intellectual disability because many people with profound and severe intellect disability will have significant expressive communication problems. Consequently, much research in this area is confounded. However, it may well be that poor expressive communication is a critical contributor to the development of self-injurious behaviour and the reason for this is discussed in Chapter 5.

There is some limited evidence that both physical disability and sensory disability are risk factors for self-injurious behaviour15, 16. However, the studies that have reported this association have never really been replicated and, similar to the expressive communication risk factor, there is nearly always overlap with degree of intellectual disability. However, there does appear to be an association between vision impairment and selfinjurious behaviour that is directed toward the eye17. This is further discussed in Chapter 5. The final risk marker is age. From our past research, and similar research in Europe and the United States, it is clear that the prevalence of self-injurious behaviour rises with age up until the mid-twenties (see Box 3.7)18. It does appear that the development of self-injurious behaviour primarily occurs between the ages of approximately seven and mid-teens. However, this conclusion does depend on studies that have defined self-injurious behaviour by the resultant tissue damage. It is possible, and there is some research to support this, that the self-injurious behaviour does occur at a very young age19 but it is not recognised as self-injury because there is no tissue damage due to the children being too small to incur injury.
In addition to these individual characteristics there are some behaviours which appear to be associated with self-injurious behaviour. There is evidence that stereotyped, compulsive behaviours and movement disorders are more common when self-injury is present20, 21. The significance of this is at present unclear. It maybe that stereotyped behaviour can evolve into self-injurious behaviour over time (see Chapter 5) or that the presence of these types of behaviours and movement disorders indicate a fundamental problem with control of movements22 (this is thought to be located in the basal ganglia see Box 3.8) and this is associated with self-injurious behaviour.

In combination these risk markers enable us to identify children who may be at higher risk of developing self-injurious behaviour. We would argue that children with a greater degree of intellectual disability, poor expressive communication and stereotyped or compulsive behaviours are at higher risk for developing self-injurious behaviour. Additionally, some syndromes indicate a higher risk for the development of self-injurious behaviour and for these syndromes these other risk markers may be less relevant. At present there is no reason to think that these risk factors are not relevant to self-injurious behaviour in children with Cornelia de Lange Syndrome, consequently children who have these risk factors may be considered to be at greater risk for developing self-injurious behaviour.

The prevalence of self-injurious behaviour across the lifespan.

Figure 1: Kebbon & Windahl (1986)
Kebbon and Windahl (1986)47 and Oliver et al. (1987)18 both reported on the prevalence of self-injurious behaviours across various age groups. As can be seen above, self-injurious behaviour increases steadily with age until adolescence (Oliver et al.) and then falls off in adulthood (Kebbon and Windahl).
Figure 2: Oliver et al (1987)

Overview of the causes of self-injurious behaviour

The theories of the causes of self-injurious behaviour can be broken down into two broad areas. Biological theories emphasise fundamental neurological and medical factors as influential in the development of self-injurious behaviour. Conversely, psychological theories consider the environment to be important, particularly with regard to the development of self-injurious behaviour. Whilst these theories tend to be clearly separated in the research literature there is no evidence that they are mutually exclusive and, in genetic syndromes such as Cornelia de Lange, interactions between aspects of the theories might help towards a better understanding of self-injurious behaviour.

The basal ganglia and self-injurious behaviour

The term basal ganglia refers to a collection of structures of the brain including the striatum and globus pallidus. Historically, research around this area of the brain has been restricted to examining its role in movement disorders such as Parkinson’s disease and Tourette’s syndrome. In both Parkinson’s disease and Tourette’s Syndrome movement control is disordered. However, the high prevalence of selfinjurious behaviour in individuals with Tourette’s syndrome and the association of self-injurious behaviour with other repetitive movements or movement disorders suggests that there may be a common brain pathway involving the basal ganglia. Damage to the basal ganglia during development may therefore lead to motor stereotypies, self-injurious behaviour or movement disorder. This theory is supported by the increased levels of selfinjurious behaviour in individuals with intellectual disability and specific genetic disorders that affect early brain development.

Neurotransmitter or neuromodulator dysfunction

Neurotransmitters provide a chemical link between the ends of the individual nerves that make up our brain and the rest of our nervous system. Different neurotransmitters tend to be associated with different functions and three neurotransmitters have been implicated as relevant to self-injurious behaviour on the basis of animal research, pharmacological intervention and, more recently, MRI studies

Biological research into self-injurious behaviour

Animal studies: For many years scientists have tried to develop an animal model of self-injurious behaviour in an attempt to both understand the behaviour and evaluate treatments for it. It has been found that injecting rats with certain chemicals, such as amphetamines, can induce self-injurious behaviour.

Pharmacological intervention: If the theory that (at least in some individuals) self-injurious behaviour is caused by neurotransmitter disturbance is true, then a drug treatment would be the most appropriate solution. Several studies have been carried out using various different drugs to treat self-injurious behaviour but results have been inconclusive as yet.

Magnetic Resonance Imaging studies: MRI scanners give high resolution images of the brain (a bit like an x-ray). Scientists have recently used these techniques to look at the brains of individuals with Lesch-Nyhan syndrome to see if there are any anatomical or functional differences that might be an indication of what causes the behaviour. It has been found recently that certain parts of the brain in individuals with Lesch-Nyhan syndrome (nearly all who have this syndrome show self-injurious behaviour) were significantly smaller than in individuals who do not have Lesch- Nyhan syndrome.

Dopamine is clearly implicated in movement and the effect of dysfunction of dopamine is most clearly seen in Parkinson’s disease. There is some evidence that dopaminergic dysfunction is related to self-injurious behaviour and stereotyped behaviours25. However, at present medication that targets dopaminergic dysfunction has not been reliably demonstrated to influence self-injurious behaviour 26.
Serotonin dysfunction is implicated in both depression and compulsive behaviours and there is some evidence that the use of serotonergic agents, and more specifically SSRI’s (selective serotonin reuptake inhibitors) influence self-injurious behaviour, interestingly by both decreasing and increasing the behaviour. Similar to the dopamine story at present there is no reliable evidence that the use of serotonergic agents can influence self-injurious behaviour.
Endorphins are the body’s natural painkillers and are similar in chemical structure to morphine and are released at times of great stress and when pain is experienced (see Box 3.10). It has been suggested that in people who shows self-injurious behaviour endorphin production might be disturbed such that two things happen. First, it has been suggested that individuals are releasing too much endorphin and consequently they are not experiencing pain when they self-injure. Second, it has been suggested that people may become addicted to their own endorphins and consequently self-injure in order to both gain a euphoric effect (similar to that caused by opiate use) and avoid the negative consequences of withdrawal from their own endorphins. The evidence for both of these theories is rather weak and tends to rely on the observation that giving medication that effectively blocks endorphins can lead to a reduction in self-injurious behaviour in some people (see Box 3.11). However, it can be argued that this medication works by simply increasing the pain that is experienced when self-injury occurs and consequently the behaviour decreases (this is related to the idea of response efficiency, see Chapter 7). This explanation does not necessarily require there to be existing dysfunction in the production of endorphins.
What are endorphins?
Endorphins are neuromodulators found in the brain that have pain-relieving properties similar to morphine. Endorphins interact with opiate receptor neurons to reduce the intensity of pain. Among individuals afflicted with chronic pain disorders, endorphins are often found in high quantity. The effect of endorphins appears to be responsible for the so called “runner's high”, the temporary loss of pain when severe injury occurs, and the analgesic effects that acupuncture and chiropractic adjustments of the spine offer. Besides behaving as a pain regulator, endorphins are also thought to be connected to physiological processes including euphoric feelings, appetite modulation, and the release of sex hormones.

Medical conditions

There is some evidence in the literature that medical conditions such otitis media28 (middle ear infections) skin infections29 and irritation can lead to self-injurious behaviour in children and adults who have intellectual disability. (see Box 3.12) More recently it has been suggested that people who experience gastro-intestinal reflux also show self-injurious behaviour30. The relevance of this research to Cornelia de Lange Syndrome is clear given the health problems that we described in Chapter 2 and is discussed further in Chapters 5, 6 and 7.

Self-injurious behaviour associated with medical conditions
Self-injurious behaviours such as head banging are often reported in normally developing children (15-20%). De Lissovoy examined this behaviour in a series of papers in the early 1960’s and found that the children in his study that were head banging were significantly more likely to have had otitis media (middle ear infection) than children in a matched comparison group. A more recent example of this association comes from Colville and Mok (2003) who describe two children with no intellectual disabilities who began to show selfinjurious behaviour whilst in hospital. Both children were on ventilators in an intensive care ward and both began to show lip-biting behaviour. In both cases, the behaviour was extinguished through increased psychosocial input and anxiety reduction.

Psychological factors
Psychological theories of the cause of self-injurious behaviour in people with intellectual disability have dominated the research literature and are supported by a vast array of empirical research studies. The dominant perspective within the psychological theories is that of operant learning theory which proposes that self-injurious behaviour is a learned behaviour that occurs because it is rewarded (or reinforced) by either sensory or social consequences. The evidence for this perspective is very strong and has been built up over a period of forty years. However, the primary issue is the extent to which this perspective is applicable to all people who show self-injurious behaviour at all times. Certainly, there is very good evidence that when assessment procedures show that a behaviour is learned and is occurring because of reinforcement then interventions can be very effective 33. Given this very strong evidence base we will focus on this psychological approach in this book.

For the purposes of understanding self-injurious behaviour and the assessment process it is useful to break the operant learning theory into two components and consider the role of sensory reinforcement and social reinforcement separately.

Case studies of drug trials in self-injurious behaviour

Naltrexone and naloxone are opioid antagonists that work by blocking the uptake of endogenous opioids or endorphins. It has been suggested that rather than environmental reinforcers such as attention, self-injurious behaviour in some individuals is reinforced by the “high” produced when the endogenous opioids are released after self-injurious behaviour has occurred. Opioid antagonists such as naloxone and naltrexone block this release of endorphins and the individual therefore does not experience the “high”. Many studies have attempted to demonstrate the efficacy of these drugs in reducing self-injurious behaviour but, as highlighted by the studies below, the results have been inconsistent.

Study Drug tested Results
Sandman et al (1983) 48 Naloxone Substantial suppression of self-injurious behaviour but only whilst drug active in system.
Beckwith et al (1986)49 Naloxone- No effect on rate of self-injurious various doses behaviour, regardless of dose given.
Luiselli et al (1989)50 Naltrexone No effect on multiple forms of selfinjurious behaviour.
Symons and Thompson (1998)51Naltrexone both alone and with FCT* Naltrexone alone produced a 50% reduction in self-injurious behaviour.

*FCT= Functional communication training (see Chapter 7)

One study of naloxone treatment (Richardson and Zaleski, 1983 52) reported an increase in SIB during drug administration which decreased again when medication was discontinued. They explained this increase in terms of an extinction burst. Extinction occurs when a normal reinforcer of a behaviour (i.e.attention) is removed (i.e. you ignore the behaviour instead). It is common for an extinction burst (a subsequent, often dramatic increase in the target behaviour) to occur as the individual strives to find reinforcement as usual (se Box 7.13). The subsequent decrease in self-injurious behaviour when drug administration was stopped might therefore be explained by the “re-activation” of the reinforcer.

Sensory reinforcement
The basis to the sensory reinforcement argument is that self-injurious behaviour occurs because the immediate sensory or perceptual consequences of the behaviour are experienced as pleasant either by the presentation of positive sensation or the removal of an unpleasant sensation. The evidence for this part of the operant theory comes from two sources. First, when alternative forms of stimulation are provided, and particularly when the form of stimulation is similar to that associated with the self-injurious behaviour, then the behaviour tends to decrease, at least temporarily. Second, when the stimulation from the self-injurious behaviour is blocked, this is called sensory extinction, and then the self-injury tends to cease. This theory has also been employed to understand stereotyped behaviours.

Social reinforcement

Since the late 1960s the evidence has grown that self-injurious behaviour can occur because it leads to social reinforcement (reward) from people who interact with the person showing the behaviour. In the last decade there has been a greater understanding of how this process unfolds and why it occurs 24. Broadly there are two ways in which the process operates at any one point in time. First, self-injurious behaviour can occur because it leads to the presentation of positive social contact or more tangible things such as food, drink or the presentation of activities. Second, self-injurious behaviour can occur because it leads to an unpleasant or aversive social contact being removed. The form of this unpleasant social contact is usually the presentation of tasks or demands that the individual simply does not want to do or finds too difficult or unrewarding. An understanding of this process of social reinforcement is critical to effective intervention consequently we have allocated much of the space on psychological assessment and intervention to this area.

It is important to note that the theories that have been developed that are described do not necessarily apply to all individuals at all times. Consequently, we emphasise the importance of assessment procedures to uncover which causes might be applicable to a given individual at any one point in time in order to guide the intervention process. However, it is equally important to recognise that for any individual more than one cause might be influential and it is almost certainly the case that causes may change over time.
This is particularly true for socially reinforced self-injurious behaviour. For this theory to be correct the behaviour has to be occurring in the first place in order for it to be reinforced. It is most likely therefore that self-injurious behaviour initially occurs in response to a minor illness or as a form of stereotyped behaviour that is reinforced by the sensory stimulation and then comes to be socially reinforced by those who interact with the person showing the behaviour. Our past research has shown us some evidence that this is the case in children with severe intellectual disability and there is no reason to think that this is not applicable to children with Cornelia de Lange Syndrome.


Self-injurious behaviour is associated with a greater degree of intellectual disability and some specific genetic syndromes. The prevalence of self-injury increases with age until the mid-20’s and the most common forms of self-injury are biting, picking and striking. Theories of the causes of self-injury in all people with intellectual disability are of two main types: biological and psychological. Biological theories tend to focus on possible neurotransmitter dysfunction and the role that might be played by minor illnesses, pain and discomfort. Psychological theories are concerned with the sensory stimulation that follows self-injury and the possible rewarding responses by carers. These theories seem to be applicable to children and adults with Cornelia de Lange Syndrome who show self-injurious behaviour and are explored in detail in subsequent chapters.


Chris Oliver

by Chris Oliver, Jo Moss, Jane Petty, Kate Arron, Jenny Sloneem, Scott Hall

Histórico da página
Modificado por Gerritjan Koekkoek em 2021/05/23 21:15
Criado por Gerritjan Koekkoek em 2021/05/23 20:22

Sobre o conteúdo do sítio web

Toda a informação contida neste WebSite é apenas para fins educativos. O local para obter aconselhamento médico específico, diagnósticos e tratamento é o seu médico. A utilização deste site é estritamente por sua conta e risco. Se encontrar algo que considere necessário corrigir ou esclarecer, por favor informe-nos em: 

Envie um e-mail: