Associations between Clinical Features and Neural Markers of Sensory Processing in Children with Autism Spectrum Disorder

Background: Sensory and motor deficits are key features of the autism spectrum disorder (ASD) phenotype. Recently, we proposed that deficits in corollary discharge (CD), neural signals that allow actions to modulate perception, could underlie many core ASD symptoms. CD signals are copies of movement signals, which are sent to the sensory areas of the brain at the same time that the motor commands themselves are sent to lower motor regions. They allow organisms to suppress sensory consequences of self-generated action and attune more to external input. Decreased CD signaling in ASD could contribute to both hypo-responsiveness to external sensory stimuli and internal preoccupation seen in the ASD phenotype.

Objectives: To test whether (i) suppression of the neural response to self-generated tones (engaging CD) versus externally-generated tones (no CD) is reduced in children with ASD, and (ii) whether these CD alterations are associated with clinical symptoms in these children.

Methods: This study used an EEG paradigm to compare neural signatures of CD between typically developing (TD) children (n=10) and children with ASD (n=15), and to relate putative CD alterations with sensory hypo-responsiveness and overall symptom severity.CD was indexed by suppression of the N1 event-related potential (ERP) between two conditions: Self-generated tones, where participants initiated a tone by pressing a button on a response pad (CD condition), and externally-generated tones, where participants passively listened to tones played back by the computer (no CD condition). Mean N1 amplitudes were extracted from central electrodes, separately for each condition. Correlations between N1 suppression and clinical scores on the Social Experience Questionnaire (hypo-responsiveness and sensory-seeking subscales) and Autism Diagnostic Observation Schedule, 2^nd Edition (ADOS-2) total score were calculated.

Results: Typically developing participants exhibited significant suppression of N1 amplitude between externally- and self-generated conditions (t(10)= 4.87, p= 0.001). No significant N1 suppression was observed in ASD (t(15)= 1.70, p= 0.11). Participants with ASD showed attenuated suppression to self- versus externally-generated tones (1.17 μV) compared to the TD group (2.57 μV), suggesting a deficit in CD signaling (d = 0.63). The ASD group scored significantly higher in sensory hypo-responsiveness (t(23)= 3.96, p< 0.001) and sensory-seeking behaviors (t(23)= 2.19, p= 0.04) than the TD group. Lesser N1 suppression was associated with greater hypo-responsiveness on the SEQ across both groups (r = -0.41, p= 0.04). Though non-significant, the correlation between N1 suppression and sensory-seeking behaviors on the SEQ was of medium size across both groups (r= -0.34, p= 0.09). Lesser N1 suppression was also associated with higher ADOS-2 total scores (r= -0.55, p= 0.03).

Conclusions: These results suggest that CD signaling during self-generated action is attenuated in ASD and relates clinically to both sensory experiences and core symptom levels. If correlations between EEG measures and clinical symptoms are maintained in a larger sample, this study could lead to a breakthrough in our understanding of the neural basis of several core symptoms of ASD, in turn leading to novel, targeted treatment approaches that could be applied in the context of very early intervention.