Investigating Social Reward Circuitry during Real-Time Peer Interaction in ASD

Background: Difficulties with social communication and social interaction, including reduced interest in both approaching and sharing information with peers, are core features of autism spectrum disorder (ASD). Some hypothesize that reduced social motivation in ASD is one cause of these social communication deficits. However, functional magnetic resonance imaging (fMRI) studies investigating the neural correlates of social motivation in ASD reveal inconsistent findings. Critically, these past paradigms do not involve social interactions, which is a core area of dysfunction in ASD. Rather, participants are asked to passively and independently view images (e.g., static photos of smiling faces), which may not approximate real-world social communicative contexts. This lack of reciprocal social interaction and real-world applicability is troubling since individuals with ASD perform within normal limits on non-interactive laboratory tasks related to social cognition, but may have the most difficulty in interactive social communicative contexts (Schilbach et al., 2013; Senju et al., 2009).

Objectives: The purpose of this study was to utilize a real-world interactive fMRI paradigm to investigate neural circuitry supporting social reward during reciprocal social interaction in ASD and neurotypical (NT) children.

Methods: 21 children with ASD (ages 7-14, three females) and 19 NT children (ages 8-14, three females) were informed they would be chatting online with a peer and a computer. The chat was simulated, but all children believed that the peer was real. Participants shared information about themselves with the peer and the computer, then during the “Reply” period, received either engaged (Peer: “Me too!”, Computer: “Matched!”) or non-engaged (Peer: “Away”, Computer: “Disconnected”) responses. For the current analyses, we focused on differences in neural activation between ASD and NT to the peer versus the computer during this “Reply” period.

Results: There were no differences between ASD and NT in a whole brain analysis when participants believed they were chatting with a peer versus a computer (“Me too!” vs. “Matched!”; p<0.005, k=20). Consistent with whole brain analyses, ROI analyses within classic reward regions (amygdala, ventral striatum, and ventral medial prefrontal cortex) revealed no group by condition interactions. However, posthoc analyses showed greater activation during the peer response in these reward regions in NT (ps<0.05), but not ASD participants. Whole brain and ROI analyses revealed no differences between ASD and NT when they believed they were chatting with a peer versus when they believed the peer was busy (“Me too!” vs. “Away”; p<0.005, k=20).

Conclusions: In summary, preliminary results suggest no differences between ASD and NT in classic reward brain regions when receiving a response from a peer. These findings suggest that the neural circuitry supporting social reward may be intact in ASD, and that perhaps other brain regions underlie the difficulties with social communication and social interaction that define the disorder. The investigation of neural and behavioral correlates of social reward within real-time social interactive contexts will help us better understand the core social deficits in ASD as well as typically developing children’s drive to orient to and interact with the social world.