Dynamic Stimuli in a Social Incentive Delay Task: Examining the Need for More Ecologically Valid Stimulus Sets in ASD Reward Research

Background:

Although it has been suggested that the pervasive social deficits of autism spectrum disorders (ASD) are related to dysfunction of the brain’s reward system, very little is known about the neural mechanisms underlying reward processing in individuals with ASD. There are now multiple imaging studies on reward responsiveness in ASD, but the results of these studies are inconsistent, particularly regarding neural responsivity to social rewards. A limitation of this research has been the reliance on static face images to serve as social rewards.  By contrast, dynamic stimuli are perceived as more engaging than static pictures, and may therefore elicit more reliable activity in reward circuits, and reveal a clearer picture of putative ASD deficits.

Objectives:

The goal of this study was to evaluate the extent to which dynamic social reward stimuli would activate a key reward area, the nucleus accumbens (NAcc), in typically developing adults.

Methods:

We conducted a functional magnetic resonance imaging (fMRI) study using a social incentive delay task in order to examine participants’ striving to receive social approval or to avoid social disapproval (N=22; 11 females; age: 25.6 ± 3.5 years). The event-related fMRI task consisted of 48 incentive trials (per incentive condition) and 48 control trials.  Participants received social approval or avoided social disapproval by hitting a button-box during the presentation of a target symbol.  Condition cues signaled potential approval, potential avoidance of disapproval, or non-reward control outcomes.  Outcome stimuli consisted of a set of newly optimized video clips (validated for such factors as authenticity and likeability). The approval condition consisted of actors giving positive performance feedback by smiling, nodding and showing a ‘thumbs up’. The disapproval condition comprised actors giving negative feedback by frowning, head shaking, and showing a ‘thumbs down’. The non-feedback control condition included actors showing a neutral expression while snapping their fingers. Imaging data were collected on a Siemens 3T scanner and analyzed with FSL. Based on apriori hypotheses, we conducted region of interest (ROI) analyses for the NAcc, which was structurally defined from the Harvard-Oxford probabilistic atlas, applying a FWE corrected threshold of p ≤ 0.05.

Results:

On the behavioral level, participants showed faster response times under both incentive conditions relative to the control condition (p < 0.001). Consistent with our predictions, robust activation of the NAcc was observed during anticipation of social approval (vs. anticipation of control outcome). Additionally, to our knowledge, this is the first study to report NAcc activation while participants anticipated avoidance of social disapproval (which can be considered a ‘reward’).

Conclusions:

This data support that dynamic social reward stimuli elicit robust activity in the NAcc, a key reward area. Our lab is currently deploying variations on this paradigm to examine social reward responsivity to other types of incentives, such as food or money, in ASD. In the long term, these data may inform the choice of reinforcers for optimized behavioral treatment plans in ASD.