A major new study published in Nature Mental Health has found that autistic children's brains respond to faces in ways that are, prepare yourself, different from neurotypical children. Shocking, we know. The research, led by James McPartland, PhD, Harris Professor in the Child Study Center and director of the Center for Brain and Mind Health at Yale, is among the most extensive investigations yet into face perception in autism, involving approximately 400 autistic children.

The study used electroencephalography (EEG) to record electrical activity across the entire scalp - all 128 electrodes, to be precise - rather than the limited regions typically examined in previous research. McPartland compares traditional approaches to understanding face recognition with trying to grasp traffic patterns in a whole city by monitoring just one downtown street. "That's the way most research on face processing so far has worked," he says. The new method is more like using satellite imagery to see the entire city's traffic. Fancy.

Participants viewed pictures of faces and everyday objects while their brain activity was recorded. Machine learning techniques were then applied to determine whether the recordings could reveal what each child was viewing. The results showed that brain activity provided a clearer indication of face viewing in neurotypical children than in autistic children. In other words, the neural signals associated with seeing faces were less distinguishable in the autistic group. "The neural activity that we measure at the scalp is just not as distinct in autistic children," says first author Jason Griffin, PhD, assistant professor of psychology at University of Houston, who previously worked as a postdoctoral researcher in McPartland's lab.

The researchers also found that in neurotypical children, the ability to identify what a child was viewing from brain activity improved with age - a sign of increasing specialization. Autistic children did not show this age-related improvement. "Their face-specific processing is not following the same trajectory," Griffin says. So much for the old "they'll grow out of it" assumption.

Beyond confirming differences, the study offers new insight into when these differences emerge. Much prior research focused on the N170, a brain signal that typically occurs about 170 milliseconds after seeing a face and is known to be delayed in autistic individuals. But by analyzing activity over a wider time window, the researchers found evidence that differences begin before the N170 appears. "When you look at face processing through an approach that's more inclusive of the full temporal range of information, we're seeing that differences in autism occur even earlier," McPartland says. So the divergence starts sooner than we thought - because of course it does.

The findings could eventually influence autism diagnosis and support. Historically, some clinicians tried to force eye contact as a fix. Research has since shown that forcing eye contact is not effective. A better understanding of face processing could lead to more appropriate approaches. The study also suggests that differences are connected to underlying developmental characteristics of autism rather than simply reflecting life experiences. "When we're seeing this expected refinement and specialization in non-autistic people, we're seeing a reduced developmental change in autistic people," McPartland explains.

Because differences appear during development, support aimed at helping autistic children interpret facial emotions might be more beneficial if introduced earlier. However, the researchers emphasize that autism does not necessarily create difficulties requiring intervention. Many neurodivergent people experience little or no impairment and may neither need nor want support. For others, challenges with recognizing faces or expressions can complicate social interactions. "Imagine navigating life when it's hard for you to perceive the information in another person's face, from navigating a playground to navigating a job interview," McPartland says. "This is one of the critical areas that causes difficulty for many autistic people."

The study also contributes to the search for biomarkers - measurable biological characteristics that could improve diagnosis or help identify who might benefit from specific support, such as social skills therapy. Currently, autism is diagnosed primarily through behavioral evaluation because no reliable biological test exists for routine use. "This is a first step in introducing a new biomarker," McPartland says. More research is needed before clinical application, but the study provides a new way to investigate how the autistic brain processes faces and how those responses change throughout childhood. In other words, scientists have finally stopped staring at one downtown street and looked at the whole city - and what do you know, they found something new.