Decreased Ventral Striatum Activation during Receipt of Social Response in Children with ASD and Co-Occurring Social Anxiety Symptoms

Background: Children and adolescents with Autism Spectrum Disorder (ASD) are frequently diagnosed with co-occurring anxiety disorders, with meta-analytic results indicating social anxiety disorder (SAD) co-occurs in approximately 16.6% of children and adolescents with ASD (van Steensel, Bogels, & Perrin, 2011). Research focused on SAD in adults has pointed to reduced neural activity in the ventral striatum (VS) during social reward processing when being observed by others (Becker, Simon, Miltner, & Straube, 2017). In one study comparing activation during receipt of social versus monetary reward in adults with ASD or SAD, adults with SAD exhibited greater amygdala activation relative to ASD in response to social reward (Richey et al., 2014). Little work has explored the co-occurrence of social anxiety (SA) symptoms and ASD and its impact on neural activation, especially in children during a real-time social interaction.

Objectives: The current study investigated whether co-occurring SA symptoms in children with ASD were associated with amygdala and VS activation in response to social versus nonsocial engagement during a real-time social interaction.

Methods: Participants were 26 children with a confirmed diagnosis of ASD aged 7 to 15 years (M = 12.3, SD = 1.9). Symptoms related to SA were assessed using the social phobia scale of the Screen for Child Anxiety Related Emotional Disorders, parent version (SCARED; Birmaher et al., 1997). SA scores ranged from 1 to 18 (M = 7.0, SD = 4.3). Neural activity within the VS and amygdala in response to social reward was assessed using a socially interactive fMRI paradigm (Warnell, Sadikova, & Redcay, 2018). Participants chatted with either a peer or a computer (social interaction context) that contingently responded to their interest (engaged) or was away (non-engaged), creating a 2 (peer vs. computer) x 2 (engaged vs. non-engaged) design. We were interested in the comparison between social contingency (i.e., peer-engaged) and nonsocial contingency (i.e., computer-engaged). We hypothesized that there would be a positive association between SA symptoms and amygdala activation and a negative association between SA symptoms and VS activation during the peer-engaged relative to the computer-engaged context.

Results: Results supported a negative association between levels of SA symptoms and activation in the inferior (r = -.54, p = .01) and superior VS (r = -.50, p = .01) when receiving a socially contingent (i.e., engaged) response from a peer versus the computer. No significant association was observed between SA symptoms and amygdala activation during peer versus computer response (r = -.13, p > .05).

Conclusions: Results extend previous work in social anxiety by demonstrating a negative relation between social anxiety and social reward activation in the striatum, but not the amygdala, in children with ASD. Further, unlike previous studies, these social reward effects were found within a naturalistic peer interaction which may have greater relevance to real-world social interaction.