Autonomic Responses to Social and Non-Social Reward Among Children with Autism

Background: Pervasive social impairments among those with autism spectrum disorders (ASDs) are often construed as stemming from reduced sensitivity to stimuli that are social in nature. Behaviorally, reduced attention to social stimuli (e.g., faces, voices) is among the earliest predictors of a later ASD diagnosis. A corresponding reduction in sensitivity to social stimuli is apparent at the neurobiological level. Individuals with ASDs fail to display selective sensitivity to social stimuli, although the familiarity of stimuli may moderate such effects. Dawson and colleagues (e.g., Dawson et al., 2005) argue that reduced sensitivity to social stimuli results from disrupted reward processing, reflected in behavioral and neurological impairments. However, the effects of reward on physiological functioning in ASD have not previously been tested.

 Objectives: We explored the effects of social and monetary reward on autonomic responding among 8- to 12-year-old boys with and without ASDs. Following from the neuroimaging literature, both unfamiliar and familiar social reward conditions were included to explore the potential moderating effect of familiarity. We also examined links between physiological response to reward and social functioning, both parent-reported and observed.

 Methods: Participants included 36 boys between the ages of 8 and 12 years. Eighteen children met ADOS, ADI, and clinical criteria for ASD and eighteen were typically-developing children with no history of developmental concerns. Physiological responding was assessed with (1) cardiac pre-ejection period (PEP), a measure of sympathetic cardiac influence that has been linked with reward sensitivity in typically-developing samples; and (2) cardiac respiratory sinus arrhythmia (RSA), a measure of parasympathetic cardiac influence that has been linked with social functioning. Autonomic responding was assessed while participants completed a simple matching task under three conditions of reward (monetary, unfamiliar social, and familiar social) and completed a social interaction task with unfamiliar and familiar partners.

 Results: Behaviorally, participants with and without ASDs had slower responses to the task under social versus monetary reward conditions. Participants with ASDs were less accurate than controls in their responses to the task. Autonomically, there were no effects of group or reward type on RSA or PEP reactivity, nor did participants with ASDs differ from controls on baseline PEP values. However, participants with ASDs displayed lower baseline RSA values relative to controls, consistent with hypotheses. Furthermore, baseline RSA was correlated with parent-reported social behavior but not with observational measures of social functioning.

 Conclusions: Our findings provide support for parasympathetic deficits among children with ASDs, but not for sympathetic deficits related to reward responding. According to polyvagal theory (e.g., Porges, 2004), parasympathetic deficits suggest reduced flexibility in adapting to changing social demands and argue for further exploration of relations between parasympathetic function and social behavior among individuals with ASDs. With regard to intervention, our findings suggest the need to integrate relationship-based and reinforcement-based strategies in intervention, and to explore the predictive or moderating effects of parasympathetic functioning on treatment outcome.