Back

An EEG Investigation of Neural Dynamics of Empathy Influenced by Congruent and Incongruent Pain Expressions in Autistic and Neurotypical Adults

Wang, X.; Tong, S. X.

2025-10-15 neuroscience
10.1101/2025.10.14.681876 bioRxiv
Show abstract

Autistic individuals often show difficulties in empathy, but the underlying neural mechanisms of empathy in naturalistic contexts of pain have been less examined. This study employed a kinetic pain empathy paradigm, manipulating the congruence between pain expressions, i.e., body gestures and facial expressions based on a predictive coding framework. We collected EEG data from 51 autistic and 58 neurotypical adults during a pain observation task. Results indicated that autistic and neurotypical adults share a similar neural architecture for empathy processing and conflict resolution, involving an early stage of sensory arousal (i.e., N2 and theta) and a later stage of cognitive reappraisal (i.e., P3). However, the multivariate pattern analysis (MVPA) revealed nuanced but significant between-group differences in neural patterns. Compared to neurotypical peers, autistic adults demonstrated atypical processes in both empathy and conflict resolution. Specifically, they exhibited heightened early emotional arousal but expended greater cognitive effort to evaluate others pain. Autistic adults also showed increased alertness to unexpected sensory input and allocated more cognitive resources to resolve prediction errors from incongruent pairings. In contrast, neurotypical adults suppressed unnecessary cognitive efforts for meaningless errors. In summary, autistic adults may experience challenges in efficiently adjusting predictions to the external context, with their neural processing heavily depending on sensory input and less efficient in adapting cognitive resources to evaluate and respond to varied contextual demands.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.