Autistic children show earlier, less specialised brain responses to faces

High-density EEG study of nearly 400 children finds that differences in face processing emerge earlier and extend more broadly than previously shown, but researchers caution against immediate clinical use.

A large new study suggests that the autistic brain processes faces differently from the neurotypical brain in ways that are broader and earlier than scientists had previously shown. Published in Nature Mental Health, the research used high-density electroencephalography, or EEG, to track electrical activity across the whole scalp in roughly 400 autistic children and a comparison group of neurotypical peers.

According to the Yale researchers behind the work, the brain patterns linked with seeing a face were less distinct in autistic children than in neurotypical children. In practical terms, the electrical activity recorded from the scalp gave a clearer signal for face viewing in the non-autistic group. The study also found a developmental split: in neurotypical children, the brain’s response to faces became more specialised with age, while that sharpening did not appear in the autistic group.

James McPartland, who leads the work at Yale, said the team used the full range of data from EEG rather than focusing on a small number of recording sites. That matters because earlier studies often concentrated on limited regions or on a single time point, such as the N170, a commonly studied brain response that appears about 170 milliseconds after a face is seen. By examining a wider time window, the team found evidence that differences in face processing may begin even earlier than that well-known signal.

The findings add to a long-running line of research on autism and face perception. Earlier studies have linked autism with reduced eye contact, less frequent looking at faces and difficulty reading facial expression. Research published in Nature Mental Health in 2024 also reported atypical face processing in the fusiform gyrus, a key brain area for recognising faces, while older work in journals such as the Journal of Child Psychology and Psychiatry and Nature Neuroscience pointed to differences in the timing of face-related brain responses and gaze fixation.

The new study could have implications for diagnosis and support, though researchers caution that much more work is needed before it can be used in clinics. Autism is still diagnosed mainly through behaviour rather than biological testing, and scientists are searching for biomarkers, measurable signs in the body or brain that might improve assessment. The Yale team says the face-related EEG patterns may eventually help identify children who could benefit from specific support, including social communication therapy. At the same time, the researchers emphasise that not every autistic person has difficulties that require intervention, and that forcing eye contact is not a useful remedy.