SIB Prevalence
1.5 Self-Injurious Behaviour in Autism Spectrum Disorder: Person Characteristics and Operant Function and Operant Function
1.5.1 Association between self-injurious behaviour and person characteristics
Associations between a variety of behavioural and demographic variables and self-injury have been investigated in populations with intellectual disability of heterogeneous aetiology (see Section 1.3.2.1). However, little research has evaluated these associations in populations with ASD. The following sections describe and evaluate research investigating person characteristics and self-injury conducted in populations with ASD, contrasted with research conducted in individuals with intellectual disability of heterogeneous aetiology.
1.5.1.1 Age
Reported associations between age and self-injury in individuals with intellectual disability vary, with the majority of studies reporting that younger age is associated with the presence of self-injury (Emerson et al., 2001a; Collacott et al., 1998). Similarly, within a population with ASD, Esbensen et al., (2009) demonstrated that older age was correlated with significantly lower levels of self-injury. However, in a younger sample of children with ASD, age was not significantly associated with the presence or severity of self-injury (Bagdhadli et al., 2003).
The conflicting data may be reflective of a more complex curvilinear relationship between age and self-injury. Oliver et al., (1987) demonstrated this curvilinear relationship in a total population sample of individuals with intellectual disability of heterogeneous aetiology. The results demonstrated that the prevalence of self-injury peaks between the ages of 15 and 25.
An operant model successfully accounts for the increasing prevalence of self-injury as
behaviour becomes learnt and is shaped by the environment (see 1.3.3.1). However, the resulting decrease in prevalence as age increases is yet to be fully explained. Given the complex association between age and self-injury, it is necessary to evaluate this association in a sample of individuals with ASD which contains a sufficient number of participants across the lifespan. Delineation of the association between age and self-injury will inform theoretical understandings of the development of self-injury in ASD populations.
1.5.1.2 Gender
Whilst gender has been associated with aggressive behaviours (McClintock et al., 2003), the majority of studies have found no significant differences between males and females with regard to self-injury (Collacott et al., 1998; Griffin, Williams, Stark, Altmeyer, & Mason, 1986). A meta-analysis of 22 prevalence and cohort studies of challenging behaviour in individuals with intellectual disability found no evidence to suggest an association between self-injurious behaviour and gender (McClintock et al., 2003). In a sample of individuals with ASD, Baghdadli et al. (2003) replicated these results demonstrating no association between self-injury and gender. However, in a recent study by Cohen et al. (2010) in which self-injury was significantly associated with ASD, this association was stronger for females. Therefore, the association between self-injury and gender in ASD samples remains unclear. This association warrants further research, and is of particular importance for individuals with ASD, given the heightened prevalence of ASD diagnoses in males (Saracino et al., 2010).
1.5.5.3 Adaptive functioning
Low levels of adaptive functioning have been associated with self-injury in individuals with intellectual disability. A greater severity of intellectual disability is strongly associated with
the presence of self-injury (Collacott et al., 1998; McClintock et al., 2003; Schroeder, Schroeder, Smith & Dalldorf, 1978). Poor mobility has also been identified as a risk marker for self-injury (Murphy, Hall, Oliver & Kissi-Debra, 1999). In addition, impaired expressive language (Collacott et al., 1998; Emerson et al., 2001a; Murphy et al., 2005; Schneider, Bijam-Schulte, Janssen & Stolk, 1996) and social skills functioning (Murphy et al., 2005) have been identified as correlates of self-injury in individuals with intellectual disability.
The relationship between impaired social-communicative behaviours and self-injurious behaviour should be considered with respect to operant models which propose challenging behaviours to be analogous to communication. Given that impaired communication is a necessary characteristic for a diagnosis of autism, it is expected that this deficit will lead to increased susceptibility to develop self-injury in individuals with ASD. Unsurprisingly therefore, deficits in expressive language (Dominick et al., 2007) along with low levels of adaptive functioning (Baghdadli et al., 2003) and IQ (Dominick et al., 2007; McTiernan et al., 2011) are all associated with self-injury in individuals with ASD. Importantly, lower speech levels were found to predict the persistence of self-injurious behaviour in children with ASD over time (Baghdadli et al., 2008), indicating that lower speech levels may serve as a putative risk marker for self-injury in individuals with ASD. This hypothesis requires further investigation in both adult and child samples with ASD, given that adaptive functioning, expressive language and ASD symptoms are often confounded (McClintock et al., 2003).
1.5.5.4 Aggressive behaviours
Almost half of all individuals with intellectual disability of heterogeneous aetiology who display self-injury also engage in aggressive behaviours (Emerson et al., 2001a; Griffin et al.,
1986). Studies have consistently found an association between self-injury and aggression (Schneider et al., 1996; Collacott et al., 1998), however, aggression does not appear to be associated with long term and persistent self-injury (Emerson et al., 2001b; Griffin et al., 1986). Additionally, temporal relationships have been identified between self-injurious and aggressive behaviours in individuals with intellectual disability of heterogeneous aetiology (Petty et al., 2009), indicating that mutual reinforcement contingencies may operate to maintain these different behaviours. Whilst the focus of this thesis is on self-injurious behaviour in ASD, investigation of the temporal associations between aggressive and self-injurious behaviours in this population is warranted, given the significant clinical implications of both behaviours.
1.5.5.5 Sensory sensitivity
Unusual sensory responses have frequently been documented within both preschool (Ornitz, Guthrie & Farley, 1977) and school aged children (Kientz & Dunn, 1997) with ASD when compared to typically developing controls. Using parental report questionnaires, children with ASD have been found to have significantly more symptoms of sensory sensitivity compared to comparison groups of children with intellectual disabilities and typically developing children (Rogers, Hepburn & Wehner, 2003). This profile of heightened sensory sensitivity is comparable to the levels of sensitivity seen in children with Fragile X syndrome and sensory modulation disorder (Miller, Reisman, McIntosh, & Simon, 2001). Theories regarding the development of self-injury have postulated that self-injurious behaviour may function to modulate levels of sensory arousal (Hutt & Hutt, 1965). However, despite the high levels of sensory sensitivity reported in individuals with ASD, there has been little consideration of the relationship between function of self-injury and sensory sensitivity in this population.
1.5.5.6 Repetitive behaviour
Self-injury is often viewed as a maladaptive repetitive or stereotyped behaviour.
Consequently, research has evaluated associations between these behaviours in an attempt to delineate the aetiology of self-injury. However, whilst there have been models proposed to delineate the development of self-injury from stereotypies (e.g., Guess & Carr, 1991), the database of evidence directly assessing these behaviours is lacking (Symons, Sperry, Dropik
& Bodfish, 2005). The available research appears to link stereotyped and self-injurious behaviours, showing that individuals with intellectual disability who engage in stereotypy are twice as likely to display self-injury (Collacott et al., 1998). Additionally, stereotyped behaviours are associated with self-injurious behaviour in individuals with intellectual disability (Emerson et al., 2001a).
Importantly the presence of repetitive and restricted behaviours are a diagnostic feature of ASD (DSM-IV; American Psychiatric Association, 1994; ICD-10; World Health Organization, 1992), and individuals with ASD are reported to display higher levels of repetitive behaviour than individuals with intellectual disability of heterogeneous aetiology (Estes et al., 2011; Turner, 1999) or typically developing individuals (Richler et al., 2007).
Associations have been identified between broad categories of repetitive behaviour and self-injury in individuals with ASD (Dominick et al., 2007). However, there has been little investigation of the associations between specific forms of repetitive behaviour and self-injury in ASD. Additionally, the predictive value of repetitive behaviours as a risk marker for self-injury in individuals with ASD has not been delineated.
In addition to the utility of repetitive behaviour as a risk marker for self-injury, repetitive behaviour has also been hypothesised as a stage in the development of self-injury (Guess &
Carr, 1991). Longitudinal study has associated stereotyped, proto-injurious and self-injurious behaviour (Richman & Lindaur, 2005), supporting a theory of progression from stereotypies to self-injury. This relationship has been further supported in populations with intellectual disability, where temporal relationships have been identified between repetitive behaviours and self-injury (Petty et al., 2009). Given the heightened prevalence of stereotyped and repetitive behaviours in individuals with ASD, it is essential that this relationship is further understood, in order to delineate the development of self-injury in individuals with ASD.
1.5.5.7 Overactive and impulsive behaviour
Despite the tentative causal models linking repetitive and self-injurious behaviours, evidence suggests that repetitive behaviours are displayed by typically developing children but rarely develop into self-injury. This may indicate that other factors are contributing to this relationship in individuals with intellectual disability and/or ASD. One hypothesis has been that an underlying executive dysfunction, evidenced by impulsivity, overactivity and compulsive behaviours, mediates the relationship between repetitive/stereotypic behaviours and self-injury (Bodfish et al., 1995).
This model is supported by evidence that self-injury is associated with overactivity in adults with intellectual disability (Collacott et al., 1998) and in individuals with Cornelia de Lange syndrome (Oliver, Sloneem, Hall & Arron, 2009). Hyperactivity has been found to be more common in individuals who display self-injury than a matched comparison group who do not display self-injury (Schneider et al., 1996). In addition to this, many individuals who engage
in self-injurious behaviour also actively seek out and engage in self-restraint behaviours (Forman, Hall & Oliver, 2002; Fovel, Lash, Barron, & Roberts, 1989; Oliver, Murphy, Hall, Arron & Leggett, 2003). This may suggest that self-injury is not fully under the individual’s control, perhaps due to a compromised ability to inhibit the behaviour. Compromised inhibition may influence self-injury in one or more ways (Barkley, 1997). It is plausible that impaired inhibition drives the individual to engage in self-injury as a prepotent response to triggering stimuli. Additionally, compromised inhibition may result in individuals being unable to terminate self-injury and thus self-restraint behaviours may reflect an attempt to suppress self-injury.
The relationship between self-injury and self-restraint was investigated in a group of individuals with CdLS (Hyman, Oliver & Hall, 2002). Hyman et al., (2002) found a significant association between self-injury and self-restraint. Importantly they also found that those individuals with CdLS who engaged in both self-injury and self-restraint displayed significantly more ‘compulsive’9 behaviours than those who did not display self-injurious or self-restraint behaviour. It is plausible that such an association exists within individuals with ASD. This hypothesis is supported by results which show that a group of individuals with ASD display significantly higher numbers of stereotypies and compulsions than a comparison group of individuals with intellectual disability. The group with ASD also had significantly greater severity ratings for ‘compulsions’, stereotypy and self-injurious behaviour (Bodfish et al., 2000).
9 ‘Compulsive’ behaviour was used as a term in this study to describe impulsive repetitive behaviours, as opposed to the anxiety relieving compulsive behaviours typically seen in conditions such as Obsessive
These preliminary findings may indicate a model of self-injury in ASD in which poor behavioural inhibition, as evidenced by overactivity/impulsivity and self-restraint behaviours, drives the development and maintenance of self-injury. However, thus far, no empirical study has quantified the associations between these behaviours and self-injury in individuals with ASD. This area warrants significant further attention, as the findings may lead to a more complete model of self-injury that could explain prevalence, maintenance and severity differences which are difficult to account for using purely operant and biological models.
1.5.5.8 Painful health conditions
Given the putative causal association between pain and self-injury (see Section 1.3.3.3), it is important to establish whether such an association may exist in individuals with ASD. Several studies have included individuals with ASD and provide preliminary evidence that pain and painful health conditions may have a causal role (Carr & Owen-DeSchryver, 2007;
Christensen et al., 2009). However, despite the emergent evidence of an association between painful health conditions and self-injurious behaviour, to date no empirical research has evaluated this association in large groups of individuals with ASD. Given the purported high levels of self-injury in individuals with ASD, it is critical that painful health conditions are quantified and evaluated in this population.