5 - (SYM 95) Anger Experience and Expression in Children with Autism Spectrum Disorder

Background: Anger and associated behaviors such as temper tantrums and noncompliance are common in autistic children and often represent the primary clinical concerns. While there is extensive research on parent-ratings of these behaviors, little is known about subjective experience of anger in autism. This is partly due to a historical the lack of confidence in the validity of self-report in autism.  Here we examine reliability and validity of the self-reported State-Trait Anger Experience Inventory (STAXI, Spielberger, 1988).

Methods: Participants included 90 autistic children, aged 8 to 15 years and 30 matched neurotypical controls. All children were characterized with regard to ASD diagnosis as well as cognitive and adaptive functioning.  Children completed the STAXI the Multidimensional Anxiety Scale for Children–2^nd edition (MASC-2). Parents completed the Social Responsiveness Scale-2 (SRS-2) and the Affective Reactivity Index (ARI).  The study was approved by the Yale University IRB and children and their parents signed informed consent for participation.

Results: Cronbach’s alpha coefficients for the state and trait anger experience subscales were 0.92 and 0.85, respectively. The three anger expression subscales evidenced lower internal consistency with anger-in alpha=0.62, anger-out alpha=0.67 and anger-control alpha=0.87. Autistic children reported higher levels of anger experience and expression than matched controls on all five subscales. There were significant correlations between self-report of anger experience and self-report of anxiety (r₈₆=0.46, p< 0.001) and parent-ratings of irritability and self-report of outward (r₈₆=0.28, p< 0.01) and inward (r₈₆=0.46, p< 0.01) anger expression. There were no significant correlations of the STAXI subscales with full scale IQ and with the SRS ratings of social communication and restricted repetitive behavior.    

 
Conclusions: This is the first study to examine anger experience in autistic children using the STAXI. Cronbach’s alphas for the state-anger, trait-anger, and anger control subscales were in the adequate range of 0.85 to 0.92 but anger-in and anger-our subscales showed low reliability in the 0.62 to 0.67 range. Self-reported anger was associated with self-reported anxiety and parent-rated irritability, demonstrating convergent validity of the STAXI.  Anger was not correlated with full scale IQ or severity of ASD symptoms indicating that autistic children across different levels of functioning are able to describe their subjective anger experiences.