Objectives: Since we currently have very little understanding of neural reactivity to motivational incentives in ASD, the present study aimed to investigate ERP and fMRI correlates in boys with and without autism while processing social compared to non-social rewards.
Methods:
Twenty ASD boys and 21 male TDCs, matched for age (mean age 14.4 years), full-scale IQ, and handedness, participated in the study. In both ERP and fMRI sessions, we applied an incentive go/nogo task with social (positive facial expressions), monetary, and non-reward contingencies for successful task performance. FMRI data were collected on a 3T scanner and analyzed with BrainVoyager. High-impedance EEG recordings were obtained from 64 electrodes and processed with BrainVision Analyzer.
Results: On the behavioral level, we found that both social and monetary incentives enhanced performance accuracy and response speed in all participants, with highest improvement under monetary reinforcement, confirming previous findings (Kohls et al., 2009). Contrary to our prediction, children with ASD showed comparable performance benefit and task motivation under social reward conditions as TDCs. By contrast, both imaging methods revealed aberrant brain responses in patients to social and to monetary reinforcers. Concerning ERPs, we found compromised P300 amplitudes to reward-predicting cues in children with ASD, which were most pronounced when social rewards were at stake, and particularly when a timely reaction was required to obtain a reward. Taken the recent locus coeruleus-norepinephrine (LC-NE) P300 theory into account (Nieuwenhuis et al., 2005), the ERP data imply an attenuated state of motivational attention allocation to incentives which trigger active approach behavior in individuals with ASD, - probably mediated by malfunctions in the reward circuitry which intervenes into the LC-NE system to boost the processing of motivational signals against other neutral stimuli. This interpretation is supported by our fMRI data which clearly demonstrate broad hypoactivations in the extended mesocorticolimbic pathway in patients including substantia nigra/VTA, dorsal striatum, and prefrontal cortex/ACC for both social and monetary reinforcement. Moreover, during facial reward processing we could replicate the typical lower brain reactivity in amygdala and fusiform gyrus in children with ASD relative to controls.
Conclusions: In sum, our findings are in line with recent social motivation deficit theories of autism which highlight a hyporesponsivity in the extended reward circuitry particularly to social incentives, what might cause the reduced socially motivated behavior in affected individuals. Noteworthy, our brain data strongly suggest that the processing of non-social incentives (such as money) is compromised, too, which deserves closer attention. Furthermore, the discrepant finding of undisturbed behavioral responses and the aberrant reward circuit responsivity in children with ASD indicate that imaging methods might be better suited to uncover deviant reward functioning in patients with neurodevelopmental disorders than behavioral measures.