Across a wide sweep of young minds, a research team has found that the brain's quiet electrical response to a human face is less clearly marked in autistic children than in their neurotypical peers, and that it does not sharpen with age the way it does for others. The finding asks us to see the familiar difficulty with faces, long noticed in autistic life, as something rooted in how the developing brain signals recognition itself. It also suggests that the early years may be when understanding this difference matters most, both for those who wish for help reading faces and for those who never
Full-Scalp EEG Study Finds Face Processing Signals Are Less Distinct in Autistic Children
Satellite imaging to track all the cars in every part of the city
Mark
What's actually new here? People have studied faces and autism for years.
Mimi
The scope. Earlier EEG work looked at particular brain areas. This one used all 128 scalp electrodes and about 400 autistic children, then asked a computer to guess whether a child was seeing a face or an object.
Mark
And the guess was worse for autistic kids?
Mimi
Yes. Face-specific signals were less distinct in them, and that's what Griffin describes.
Luke
Worth saying what that means. It's a statement about group-level signal separability, not a test that tells you whether one child is autistic. The source doesn't report accuracy figures for individuals.
Mark
What about the age finding?
Mimi
In neurotypical children, the signals got more refined with age. In autistic children, the researchers saw no change. Griffin calls it a different trajectory.
Luke
Careful, though. The source doesn't say how the age comparison was designed, whether it followed the same children over time or compared different ages. That matters for the word trajectory.
Mark
The researchers suggest early childhood may be the best time for intervention. Is that established?
Mimi
It's an inference. If differences arise early, support might work best then. They also stress that many autistic people don't want intervention.
Luke
Right, it's a hypothesis, not a tested result. Same for the biomarker. McPartland himself calls it a first step.
The Pulse
- Recordings from all 128 scalp electrodes reveal that face signals in autistic children are noticeably harder to tell apart from object signals than in neurotypical children.
- The divergence does not narrow with growing up: where neurotypical brains refine their face responses over time, autistic brains show no comparable change.
- Researchers are pressing to find where the difference begins, probing timing far earlier than the familiar 170-millisecond marker most prior work has examined.
- The team is redirecting interventions away from forcing eye contact, which they say does not work, toward helping children read facial emotions in early childhood.
- Face-processing signals are being tested as a possible objective biomarker, a step toward diagnosis that moves beyond behavior alone, though that work is only beginning.
Across a wide sweep of young minds, a research team has found that the brain's quiet electrical response to a human face is less clearly marked in autistic children than in their neurotypical peers, and that it does not sharpen with age the way it does for others. The finding asks us to see the familiar difficulty with faces, long noticed in autistic life, as something rooted in how the developing brain signals recognition itself. It also suggests that the early years may be when understanding this difference matters most, both for those who wish for help reading faces and for those who never needed it.
A research team at Yale has recorded electrical activity across the scalp of about 400 autistic children as they looked at pictures of faces and objects. Published in Nature Mental Health, the study offers a fuller picture of face perception than most earlier EEG work could provide.
The question is an old one. Autistic people often make less eye contact, and many find faces harder to remember and emotions harder to read. Earlier Yale research had already found less activity in face-processing regions of the brain. James McPartland, who has long used EEG in this field, likens older studies to watching a city's traffic through one camera; this work resembles satellite imaging that follows every car across the whole map.
Using data from the Autism Biomarkers Consortium for Clinical Trials, the team used machine learning to predict from brain activity alone what a child was looking at. The predictions were clearer for neurotypical children. Jason Griffin, now at the University of Houston, put it plainly: the signals are simply less distinct in autistic children.
Age revealed the second difference. Among neurotypical children, face-specific signals became more refined with age. Autistic children showed no such change, suggesting their development follows a different path. The team also found differences earlier than the N170 response that most research focuses on.
The findings carry practical hopes. Forcing eye contact, a common clinical approach, does not address the underlying difference. Early childhood may be the best time for help with reading emotions, though the researchers stress that many autistic people do not need intervention, and that for those who do, face-reading difficulties can matter at playgrounds and job interviews alike. The signals may also one day aid diagnosis, which still rests on behavior alone. McPartland calls this a first step, and the next task is testing whether these signals can serve as a reliable biomarker.
A research team at Yale has recorded electrical activity across the entire scalp of about 400 autistic children as they looked at pictures of faces and objects. The result, published in the journal Nature Mental Health, is a picture of face perception that is wider than most earlier EEG work could offer.
The backdrop is familiar. Autism is often marked by less eye contact and a tendency to avoid looking at other people's faces. Many autistic people also find it harder to remember faces and to recognize the emotions on them. Scientists have studied the brain basis of this for decades. In the early 2000s, Yale-led research found that brain regions involved in face processing were less active in autistic people than in neurotypical people.
James McPartland, Ph.D., the Harris Professor in the Child Study Center and director of the Center for Brain and Mind Health, has long used electroencephalography, which measures the brain's electrical activity, to study this question. Most EEG studies, he notes, have been confined to specific areas of the brain. He compares that to estimating a city's traffic from one camera on one downtown road. The new work, he says, is closer to satellite imaging that tracks every car in every part of the city.
The team drew on data from the Autism Biomarkers Consortium for Clinical Trials, which included autistic and neurotypical children across a range of ages. Researchers recorded from all 128 electrodes on the scalp while the children viewed faces and objects. They then used machine learning to predict, from brain activity alone, what a child had been looking at.
The predictions worked better for neurotypical children. It was easier to tell when they were viewing a face than when an autistic child was. First author Jason Griffin, Ph.D., now an assistant professor of psychology at the University of Houston and formerly a postdoctoral researcher in McPartland's lab, put it this way: the neural activity measured at the scalp is simply not as distinct in autistic children.
Age added a second finding. In neurotypical children, the ability to predict what they were viewing grew with age, which fits the idea that face-specific signals become more refined and specialized. In autistic children the researchers saw no such change. Griffin said their face-specific processing is not following the same trajectory.
The work also bears on timing. Most autism research on face processing focuses on the N170, a shift in brain activity about 170 milliseconds after a face appears, which is delayed in autistic individuals. Looking across the full time range of the recording, McPartland says, the team found differences occurring even earlier.
The researchers see practical implications. Clinicians have long tried to address face-processing differences by insisting that autistic patients make eye contact, an approach researchers now know does not work. A persistent difficulty in autism research is separating what the condition causes from what living with it causes, and the team says these findings point to the former. If the divergence appears early in development, interventions to help children understand facial emotions might work best in early childhood. The researchers note that many neurodivergent people are not hindered by autism and may not want or need intervention, but for others, trouble reading faces affects quality of life. McPartland points to settings from a playground to a job interview.
There is a diagnostic angle too. Autism is currently diagnosed from behavior, and no biologically based tools exist. Researchers are evaluating various biomarkers to improve diagnosis or define subgroups, and face-processing signals might help predict who would benefit from approaches such as social skills therapy. McPartland calls the study a first step in introducing a new biomarker, and what follows is testing whether the signals can work in that role.
Notable Quotes
The neural activity that we measure at the scalp is just not as distinct in autistic children.— Jason Griffin, Ph.D., University of Houston
This is a first step in introducing a new biomarker.— James McPartland, Ph.D., Yale Child Study Center
Want the full story?
Read the original at Medical Xpress ↗
Science & Space