Abstract/Summary

Autistic individuals often struggle to recognise speech in noisy environments, but the neural mechanisms behind these challenges remain unclear. Effective speech-in-noise (SiN) processing relies on auditory processing, which tracks target sounds amidst noise, and semantic processing, which further integrates relevant acoustic information to derive meaning. This study examined these two processes in autism. Thirty-one autistic and 31 non-autistic adults completed a sentence judgement task under three conditions: quiet, babble noise, and competing speech. Auditory processing was measured using EEG-derived temporal response functions (TRFs), which tracked how the brain follows speech sounds, while semantic processing was assessed via behavioural accuracy and the N400 component, a neural marker of semantic processing. Autistic participants showed reduced TRF responses and delayed N400 onset, indicating less efficient auditory processing and slower semantic processing, despite similar N400 amplitude and behavioural performance. Moreover, non-autistic participants demonstrated a trade-off between auditory and semantic processing resources. In the competing speech condition, they showed enhanced semantic integration but reduced neural tracking of auditory information when managing linguistic competition introduced by intelligible speech noise. In contrast, the autistic group showed no modulation of neural responses, suggesting reduced flexibility in adjusting auditory and semantic demands. These findings highlight distinct neural processing patterns in autistic individuals during SiN tasks, providing new insights into how atypical auditory and semantic processing shape SiN perception in autism.