Relationships between Self-Reported Social Reward and Neural Response to Peer Interaction in Children with Autism Spectrum Disorder and Typical Development

Background: The social motivation hypothesis proposes that individuals with autism spectrum disorder (ASD) have reduced motivation to engage in social interactions compared to children with typical development (TD). However, neuroimaging and behavioral support for this hypothesis is mixed. These inconsistent results could be a relic of neuroscience paradigms typically used to assess social reward, which rely on static images rather than the social reward experienced in a real-time interaction with a peer.

Objectives: To assess the relationship between self-reported enjoyment of social interaction and the neural response of a reward network brain region, the ventral striatum, to a real-time peer interaction in children with TD and ASD.

Methods: Children with ASD (n=26, 2 female, 7-15 years old) and TD (n=26, 2 female, 8-14 years old) engaged in a social chat paradigm while their brain activity was measured using functional magnetic resonance imaging. Groups did not differ on age or full scale IQ (ps>0.10). In the chat paradigm, participants sent messages about their interests to a perceived peer or computer. Participants then received messages back indicating engagement or disengagement from the peer (“Me too”/ “I’m away”) or computer (“Matched”/ “Disconnected”). Ventral striatum response to computer engagement was subtracted from ventral striatum response to peer engagement to create a metric of neural response to social engagement with a peer. This difference score was correlated with the Admiration subscale of the Social Reward Questionnaire (SRQ), measuring trait level enjoyment in receiving others’ approval, as well as children’s self-reported enjoyment of engaging with the peer minus enjoyment of engaging with the computer after the paradigm.

Results: Children with TD and ASD did not differ in their self-report of enjoyment of interacting with the peer compared to the computer (t(50)=-0.60, p>0.10) or in their self-report of enjoyment in receiving others’ approval (SRQ-Admiration; t(39)=-0.52, p>0.10). However, the relationship between these two measures of self-reported enjoyment of social reward and ventral striatum response to peer engagement differed between the two diagnostic groups. Increased enjoyment in the peer interaction compared to the computer interaction was associated with increased ventral striatum response to peer compared to computer in children with TD (r(24)=0.57, p<0.01), but not children with ASD (r(24)=-0.002, p>0.10). These correlations significantly differed from each other (p<0.05). In contrast, decreased enjoyment in others’ approval (SRQ-Admiration) was associated with increased ventral striatum response in children with ASD (r(21)=-0.35, p<0.10), while there was a non-significant association in the opposite direction in children with TD (r(16)=0.34, p>0.10). Again, these correlations significantly differed from each other (p<0.05).

Conclusions: The relationship between self-reported social reward and neural response to real-time peer interaction differs between children with TD and ASD. While increased self-reported social reward was associated with increased activation of a reward network brain region in children with TD, the opposite relationship was present in children with ASD. These findings highlight the importance of more thoroughly characterizing the neural correlates of social reward in children with ASD within ecologically valid contexts and determining their relevance to real-world social interactions.