Exploring factors related to the anger superiority effect in children with Autism Spectrum Disorder
Introduction
Identifying emotional faces is an important social skill that allows people to identify the mental states of others and respond appropriately during social interactions. Individuals with Autism Spectrum Disorder (ASD), a neurodevelopmental condition affecting 1% of the population, have a core deficit in their ability to function socially (American Psychiatric Association, 2013). They experience deficits in the way they process faces and impairment in understanding the mental states and intentions of others, often referred to as Theory of Mind. Atypical face processing in ASD is marked by paying more attention to the mouth rather than eyes when processing faces (Baron-Cohen et al., 1997, Dalton et al., 2005). Although primary facial expressions such as happy and sad often appear intact in ASD, the recognition of subtle emotions, such as guilt, fear and anger, may be impaired (Harms et al., 2010, Kuusikko et al., 2009). It has been proposed that because individuals with ASD do not recognize emotions in a typical way, they compensate use more time consuming processes.
In typically developing individuals, an angry face in a crowd is detected faster than a happy face in a crowd (Hansen & Hansen, 1988), a phenomenon known as the anger superiority effect (ASE). Angry or threatening faces provide early warning signs of possible danger, hence, the ability to rapidly identify an angry or threatening face at the pre-attentive level may be underpinned by a neural network evolved for this purpose (Krysko and Rutherford, 2009, Vuilleumier and Schwartz, 2001). Given the face processing deficits in ASD, atypical performance on this type of task, which requires differentiation of facial expressions could be indicated. However, it has also been suggested that the ASE being at the pre-attentive or implicit processing level, may involve more basic brain areas than those required for identifying faces or Theory of Mind, with these basic areas intact in ASD (Krysko & Rutherford, 2009). The amygdala is thought to be involved in the ASE given its well established role in processing faces and threatening stimuli. The amygdala may underpin the neural circuitry of an “emotional attention” network, with this type of attention potentially resulting in the “pop out” effect found in the ASE. This emotional attention may be different to the other types of executive functioning attention components such as sustained and selective attention. In ASD, the executive functionings are often impaired. However, the amygdala is also implicated.
To date, five past studies, three focusing on children and adolescents and two on adults have examined the ASE in ASD, Table 1. Ashwin, Wheelwright, and Baron-Cohen (2006) found a threat detection over friendly face advantage in adult males with ASD similar to a comparison group. Krysko and Rutherford (2009) examined a group of adult males with ASD and similarly found them to be as fast as comparison adults in identifying a threatening face in a group of faces; although adults with ASD made more errors, indicating some atypical processing. Children with ASD (N = 30) aged 7–17 years were found by Rosset et al. (2011) to experience a similar ASE to typically developing children. This was also the case in Isomura and colleagues (Isomura, Ogawa, Yamada, Shibasaki, & Masataka, 2014) in one study of children aged 7–17 years, but not in their second study of younger children aged 7–10 years (Isomura, Ito, Ogawa, & Masataka, 2014). Isomura (Isomura, Ito, et al., 2014) suggest this contradictory finding might indicate that there is a developmental effect of the ASE. However potentially the small sample of these studies may have produced the inconsistency. Past studies of normally developing individuals have found that the ASE is present in infants and 5 year old children, as well as adults, with faster detection in adults compared to children (LoBue, 2009, LoBue and DeLoache, 2010).
Overall, these studies suggest the ASE is present in ASD, thus supporting the notion of an intact implicit threat detection system. However, there were a number of limitations in these studies. Firstly, all of the studies have employed schematic faces. Schematic faces lack ecological validity and may have provided a way to for individuals with ASD to compensate for emotion recognition difficulties and potentially even provide them with an advantage given their local feature biases (Isomura, Ogawa, et al., 2014). In fact, children with ASD have been found to process cartoon faces but not photographic faces similar to typically developing children (Rosset et al., 2008). The need for replication of the findings using ecologically valid photographic faces may therefore be important.
Secondly, the adult studies have only examining males with ASD, and the child studies have had too few females to examine gender differences. A gender difference in attention bias has been found, such that anxious females show threat bias, whereas anxious males do not (Tran, Lamplmayr, Pintzinger, & Pfabigan, 2013). Meta-analytic reviews suggest that females have a face emotion processing advantage over males which may relate to socialisation or neurological maturation differences between the genders (McClure, 2000). Males with ASD outnumber females by on average 4:1, with the cause of this gender discrepancy much debated and still unclear (American Psychiatric Association, 2013). It has been proposed that females may have better superficial social skills which results in subthreshold symptoms of ASD (American Psychiatric Association, 2013). Whether gender plays a role in the anger superiority effect and how this might interact with ASD is unknown.
Thirdly, studies of the anger superiority effect show that anxious individuals exhibit an enhanced effect over non anxious controls (Ashwin et al., 2012, Byrne and Eysenck, 1995, Fox et al., 2008). This enhanced threat bias is thought to maintain and potentially cause anxiety (Mathews, 1990, Mathews and Mackintosh, 1998). Given extremely high levels of anxiety in ASD (White, Oswald, Ollendick, & Scahill, 2009), anxiety symptoms may potentially be a confounding factor and may resulted in an enhanced effect in ASD resulting in apparent similar performance to controls.
This study aimed to extend past research in threat detection in ASD to address these three potentially confounding factors. Firstly, the studied employed a similar number of males and females to enable the exploration of gender differences. Secondly, the study used colour photographs to examine the ASE using ecologically valid stimuli rather than schematic faces which could confer an advantage to individuals with ASD. Thirdly, anxiety symptoms were measured to determine if these were associated with the ASE in ASD. Finally, given the proposal by Isomura, Ito, et al. (2014) that there may be an age effect in ASD such that attainment of the ASE is delayed put may ‘catch up’ by adolescence, age was also investigated.
Predictions were as follows. Firstly, if photographic faces resulted in more complex processing for children with ASD, they should exhibit slower and less accurate performance than typically developing children on the anger superiority task. Secondly, if females are more likely to show threat bias than males, females, with and without ASD, would be expected to show an enhanced ASE compared to males. Thirdly if anxiety symptoms are driving the ‘intact’ anger superiority performance in ASD, there should be a positive association between anxiety symptoms and the anger superiority in individuals with ASD. Finally, if the effect is related to development in ASD, there should be significant associated between age and the ASE in the ASD group.
Section snippets
Participants
In total, eighty-one children participated in this study. Participants with ASD were 42 children, 18 male and 24 female, with Autistic Disorder or Asperger’s Disorder aged between 7 and 12 years of age. This sample was taken from our larger study of gender difference in children with ASD, hence females were oversampled compared to the normal male prevalence in ASD (May, Cornish, & Rinehart, 2012). All clinical children had been diagnosed with ASD by registered psychologists and paediatricians.
Correlations with demographic variables
Pearson correlations between the log 10 RTs of the array conditions and demographic variables were conducted, Table 3. In the TYP group there was a significant negative correlation between age and Threat target in neutral array RT, indicating faster speeded detection of threat in older TYP children. In the ASD group, age correlated with Happy target in Neutral array RT, indicating speeded detection of happy faces in neutral arrays in older children with ASD. In the ASD group verbal IQ negatively
Discussion
This is the largest study of the ASE in ASD to date and also the first to use unaltered ecological photographic stimuli and a gender balanced group. Findings confirmed that 7–12 year old children with ASD exhibited the ASE in detecting threatening faces faster than happy faces in a neural crowd, which was similar to typically developing children. This is consistent with the past majority of studies of the ASE in ASD which have employed smaller groups and schematic rather than more complex
Conclusions
Together our findings suggest that high-functioning children with ASD, regardless of gender, age and level of anxiety, exhibit a similar ASE to typically developing children when using photographic faces. This provides further evidence that more basic evolutionary mechanisms such as implicit threat detection are intact in ASD. However, the different associations with age suggest that although there is similar task performance, different underlying processing or compensatory mechanisms may be at
References (37)
Fear, faces, and the human amygdala
Current Opinion in Neurobiology
(2008)- et al.
Enhanced anger superiority effect in generalized anxiety disorder and panic disorder
Journal of Anxiety Disorders
(2012) - et al.
Finding a face in the crowd: Testing the anger superiority effect in Asperger syndrome
Brain and Cognition
(2006) - et al.
The amygdala theory of autism
Neuroscience & Biobehavioral Reviews
(2000) - et al.
The neurodevelopmental frontostriatal disorders: Evolutionary adaptiveness and anomalous lateralization
Brain and Language
(2000) - et al.
A threat-detection advantage in those with autism spectrum disorders
Brain and Cognition
(2009) Why worry? The cognitive function of anxiety
Behaviour Research and Therapy
(1990)- et al.
A parent-report measure of children’s anxiety: Psychometric properties and comparison with child-report in a clinic and normal sample
Behaviour Research and Therapy
(2004) - et al.
More than just another face in the crowd: Evidence for an angry superiority effect in children with and without autism
Research in Autism Spectrum Disorders
(2011) A measure of anxiety symptoms among children
Behaviour Research and Therapy
(1998)