Interactive Social Neuroscience to Assess Reward Processing in the Broad Autism Phenotype: An Event-Related Potential Study

Background: Autism Spectrum Disorder (ASD) is hallmarked by interpersonal difficulties, yet scant published neuroscience research investigates actual social interactions. Atypical reward processing is implicated in ASD, but ecologically valid studies dissociating social versus non-social reward in ASD are lacking. This study utilized interactive social neuroscience methods to examine reward processing in a social context and its association with autistic traits in typically developing (TD) adults. Using simultaneous recording of electroencephalogram (EEG) and electrocardiogram (ECG) in pairs of people competing in a computer game, we examined reward processing (indexed by event-related potentials [ERP]), heart rate variability, and their association with autistic symptomatology.

Objectives: Characterize neural and cardiac markers of reward and feedback processing and their association with autistic traits during face-to-face social interactions.

Methods: 16 TD adults, grouped in same-sex dyads, (a) sat quietly apart versus together while making eye contact, and (b) played a competitive treasure hunt game against a computer and against each other. EEG and ECG were recorded from each member of the dyad using wireless recording devices. ERP components marking early visual processing (P1, N2) and reward processing (FRN, P3) were contrasted based on competitor (computer/human), outcome (win/loss), and player (self/opponent). To quantify social function and dysfunction, the Autism Quotient (AQ) and Broad Autism Phenotype Questionnaire (BAPQ) were administered. 

Results: Preliminary analyses of heart rate variability and resting data revealed differences associated with the presence of another person. Autistic traits were associated with attenuated FRN and P3 indices of reward processing. Sensitivity to feedback, indexed by the FRN, during live, not computer, competitors was correlated with autistic traits. There was a significant correlation between the differences in FRN amplitude between self-win and opponent-win when playing against another person and AQ (r=-0.505, p=0.046) and BAPQ (r=-0.558, p=0.025) scores. Additionally, BAPQ score was correlated with differences in FRN amplitude between self-win and opponent-miss when playing against another human (r=-0.544, p=0.030), as well as with the difference in FRN amplitude for opponent-win and self-miss (r=0.526, p=0.036). Sensitivity to reward during live interaction, indexed by the P3, was correlated with autistic traits. For the P3, BAPQ scores were correlated with the difference in amplitude between self-win and opponent-win conditions when playing against another person (r=-0.516, p=0.041), as well as between opponent-win and self-miss conditions when playing against another person (r=0.524, p=0.037). There were no significant effects of outcome, player, competitor, or interactions for low-level sensory components (P1 and N2; all ps>0.10).

Conclusions: This study applied interactive social neuroscience to investigate reward processing during live interaction in TD adults. This was the first study to examine electrophysiological indices of reward-feedback monitoring during live dyadic interaction. Results reveal task-specific modulation of brain activity during live interaction that is absent during computer interaction. Individuals with lower levels of autistic traits exhibited greater sensitivity to outcome during live interaction versus playing alone. Our findings provide new insight into differences in reward processing mechanisms associated with autistic traits and further emphasize the import of utilizing more ecologically valid approaches in neuroscientific studies of social brain function.