Reward Responsiveness across Autism and Autistic Traits. Evidence from Neuronal, Autonomic, and Behavioural Levels

Atypicalities in processing of social rewards have been suggested to lie at the root of social difficulties in autism spectrum conditions (ASC). While evidence for atypical reward function in ASC is mounting, it remains unclear whether it manifests specifically in hypo- or hyper-responsiveness, and whether it appears only in the social domain or more generally. Moreover, stimuli used as social rewards in studies often lack familiarity and relevance, which are known to enhance reward-related responses. In this study, we investigated behavioural (reaction times and ratings), neuronal (event-related potentials), and autonomic (pupil sizes) responses to three conditions -- relevant social rewards, money, and neutral informative outcomes -- in 26 ASC and 53 non-autistic subjects varying in levels of autistic traits, as measured with the Autism Spectrum Quotient (AQ). We used both a population-based approach (low AQ vs. high AQ) and a psychopathological approach (low AQ vs. ASC) to investigate the effects of both sub- clinical and clinical autistic traits on reward responsiveness. As hypothesised and preregistered, autism and autistic traits did not differently influence responses to social, monetary, and neutral outcomes on behavioural, neuronal or autonomic level. Although the ASC group rated the motivational and rewarding values of stimuli lower than the other groups, the task performance was similar for all participants. Moreover, the ASC group in contrast to low AQ group showed enhanced brain responses (the CNV) in early anticipation and larger pupil constrictions in reward reception. Both effects were also predicted by autistic traits (AQ). Together, our results do not offer evidence for specifically social reward deficits in ASC. Instead, the data suggest enhanced neuronal and autonomic reward responsiveness linked to autism with simultaneously typical performance and reduced self-reported motivational and rewarding values of stimuli. Together, these results emphasise the need to investigate multiple processing levels for a broader picture of reward responsiveness in ASC.