Embodied Emotions in Autism Spectrum Disorder: Somatosensory Evoked Potentials Reveal Atypical Patterns of Neural Activity during Perception of Emotional Expressions in ASD

Background: Previous research has shown the independent contribution of the right somatosensory cortex to emotion recognition in the neurotypical individuals (Adolphs & Damasio 2000, Pitcher et al. 2008, Sel et al. 2014), supporting the embodied emotion hypothesis (Niedenthal 2007). Atypical patterns of responses to emotions in Autism Spectrum Disorder (ASD) have been shown in visual cortical areas, but embodiment processes in the Somatosensory Cortex have not been studied in this group yet.

Objectives: The current study investigates embodied simulation in the somatosensory cortex during emotion perception in ASD and typically developed (TD) adults.

Methods: We tested two groups of 20 TD and 20 ASD participants, matched for age (mean TD= 40.3, mean ASD = 41.62), gender (1 female, 19 male) and IQ (mean IQV TD = 114, ASD = 108.08, mean IQP TD = 117, ASD = 112.62). We presented neutral, afraid and happy faces to participants while their brain activity was recorded with a 64 channels EEG. In separate blocks, participants were either discriminating the emotion expressions (emotion task) or the gender (gender task) of the faces shown answering either an emotion or gender question on 20% of trials per task. On 50% trials participant received mechanical tactile stimuli on their left index finger after 105’’ from visual onset. We measured the visual evoked potentials (VEPs) over the occipital lobe and the somatosensory evoked potentials (SEPs) over the Somatosensory Cortex. To isolate the pure somatosensory responses related to emotional processing from visual carryover effects, we subtracted the neural activity recorded in the visual-only condition (i.e., when no tactile stimuli were delivered) from activity in the tactile condition (following Sel et al., 2014).

Results: VEPs: we found significant main effects of emotion at P120, N170, and P300, and significant interactions between emotion, task and group at N170 and P300 (p<.05) with significant emotion effects only for the emotion task in both groups suggesting different visual processing of emotional expressions in the two groups during later processing stages. SEPs: Both group showed main effects of emotion in the right hemisphere in 60-80'', and in 80-100''. In 80-100'' time window, the effect was significant only in emotion task. Most importantly, we found significant interaction between emotion and group across all time windows between 100 and 220 ms, suggesting group differences at later processing stages. Follow up analyses showed that the main effect of emotion was persistent after 100’’ in the right somatosensory cortex of the TD but not in the ASD group, suggesting differential recruitment of somatosensory cortex for embodied processing between the two groups.

Conclusions: Our study provides novel evidence of the contribution of the somatosensory cortex in emotion processing in individuals with Autism Spectrum Disorder. In particular, our results show the presence of embodiment in TD and ASD in early Somatosensory Evoked Potentials, but this effect was prolonged only in the TD group.