Exploring the Validity of the Social Anxiety Scale in Capturing Symptoms of Social Anxiety in Youths with Autism Spectrum Disorder

Background: Previous research indicates a high clinical prevalence of co-morbid anxiety in individuals with autism spectrum disorder (ASD). However, there has been less focus on the specific co-occurrence of social anxiety (SA) in ASD (Gillott et al., 2001). Limited research examines different aspects of SA or the relationship of SA to perceived social responsiveness in individuals with ASD (Bellini, 2004). Furthermore, while the Multidimensional Anxiety Scale for Children (MASC) has been found to be an appropriate measure of anxiety in ASD, there is a need for additional valid measures of SA specifically.

Objectives: The current study investigated SA in children with ASD and typically developing (TD) controls. We sought to evaluate the convergent validity of the Social Anxiety Scale for Adolescents/ Children, Revised (SAS-A/SASC-R) in capturing symptoms of SA in individuals with ASD through comparison to another established measure, the MASC. We also aimed to measure child-parent agreement in reports of the child’s SA, as well as the relationship between SA and social behavior in ASD.

Methods: This study’s sample was composed of 50 youth with ASD (37 males, 13 females) and 25 TD youth controls (18 males, 7 females), matched on gender, age and IQ (Table 1.). Data collection is ongoing. Social anxiety was measured using multiple standardized measures. Child self-report measures included the Multidimensional Anxiety Scale for Children (MASC-C) and Social Anxiety Scale for Adolescence/Children (SAS-A/SASC-R). Parent report of SA was also assessed by the MASC (MASC-P) and the child’s social behavior was reported on the Social Responsiveness Scale, Parent Report (SRS-P).

Results: Children with ASD reported higher symptoms of SA [M=44.62, SD=11.76] than those with TD [M=38.44, SD=12.36] on the SAS-A/SASC-R [F (1,73) = 4.45 p=.04]. A multitude of significant correlates were found (Table 2.). SAS-A/SASC-R Total scores were highly correlated with the MASC-C SA [r =.70, p<.001] and MASC-P SA [r=.43, p<.001]. The subdomains of the SAS-A/SASC-R, Fear of Negative Evaluations (FNE) and Social Avoidance and Distress in New Situations and with New Peers (SAD-N) were also significantly correlated with the subdomains of the MASC-C, Humiliation and Rejection [r=.71, p<.001] and MASC-C Performance Fears [r=.59, p<.001], respectively. Child and parent scores on the MASC SA domain were significantly correlated (r=.39, p<.0.01), as were the MASC-P SA score and the child’s SAS/SASC-R total [r=.44, p<.001]. The SAS-A/SASC-R, the MASC-C, and MASC-P SA domain did not correlate significantly with any domains of the SRS-P.

Conclusions: This study examined the validity of the SAS-A/SASC-R to measure symptoms of SA in individuals with ASD in comparison to the previously established use of the MASC. Results confirmed that the SAS-A/SASC-R can be used to measure symptoms of SA in ASD and can provide insight into the specific factors of SA that may be affecting the child. In this sample, social ability and symptoms of SA did not correlate. The SAS-A/SASC-R is likely capturing true symptoms of SA in children with ASD that are distinct from difficulties in social behavior as captured by the SRS-P.