A major new study has uncovered important differences in how the brains of autistic children respond to human faces. Published in Nature Mental Health, the research is among the most extensive investigations yet into how face perception works in autism.
Some autistic people tend to look at faces less frequently or make less eye contact. They may also find it more difficult to remember faces or interpret facial expressions compared with neurotypical individuals. Scientists have spent decades investigating the brain activity behind these differences. In the early 2000s, for example, researchers at Yale reported that brain regions responsible for processing faces showed lower activity in autistic individuals than in people without autism.
James McPartland, PhD, Harris Professor in the Child Study Center and director of the Center for Brain and Mind Health, studies how autistic people process faces using electroencephalography (EEG), a technique that records the brain’s electrical activity. Previous EEG studies, however, have typically examined signals from limited regions rather than analyzing activity across the entire scalp.
To gain a more complete picture, McPartland and his colleagues conducted a study involving approximately 400 autistic children, examining electrical signals recorded across the scalp. They discovered that brain activity associated with face recognition was less clearly distinguishable in autistic participants. They also identified a striking difference in development: In neurotypical children, the brain’s response to faces became more specialized with age, while autistic children did not show the same pattern.
“This is the first time we’ve leveraged the full volume of very rich information provided by an EEG recording to understand face perception and autism,” McPartland says.
How Scientists Discovered Differences in Autistic Brain Activity
McPartland compares traditional approaches to studying face recognition with attempting to understand traffic throughout an entire city by monitoring just one downtown street.
“That’s the way most research on face processing so far has worked,” McPartland says.
The new approach provides a much broader perspective, similar to observing traffic across an entire city using satellite imagery. To investigate these patterns, the researchers analyzed information collected through the Autism Biomarkers Consortium for Clinical Trials (ABC-CT). The dataset included autistic and neurotypical children across different age groups.
During the experiments, participants viewed pictures of faces and everyday objects while EEG equipment recorded their brain activity. Instead of concentrating on a small number of recording locations, the scientists examined electrical signals from all 128 electrodes positioned around each participant’s scalp.
They then applied machine learning techniques to determine whether patterns in 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 electrical 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. Griffin previously worked as a postdoctoral researcher in McPartland’s laboratory.
The researchers also found that the ability to identify what a child was viewing from brain activity improved with age among neurotypical participants. This age related improvement was not observed in autistic children, suggesting that their neural responses to faces develop along a different pattern.
“Their face-specific processing is not following the same trajectory,” Griffin says.
Face Recognition Differences May Begin Earlier Than Previously Thought
Beyond identifying differences in brain activity, the study offers new insight into how early those differences emerge during face perception.
Recognizing someone involves multiple stages of neural processing. The sequence begins when visual information enters the eyes and continues as the brain interprets that information until a face is recognized.
Much of the earlier research on autism and face perception has concentrated on a brain signal called the N170. This measurable change in electrical activity typically occurs about 170 milliseconds after a person sees a face. Previous studies have shown that the N170 response tends to occur later in autistic individuals.
By analyzing brain activity over a wider period, the researchers found evidence that differences in face processing begin before this familiar signal appears. This suggests that autism related differences may arise at earlier stages of visual processing than researchers had previously established.
“But 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.
What the Discovery Could Mean for Autism Diagnosis and Early Support
The findings could eventually influence how researchers approach autism diagnosis and the development of supportive interventions.
Historically, some clinicians attempted to address difficulties with face perception by requiring autistic individuals to maintain eye contact. Research has since shown that forcing eye contact is not an effective solution. A better understanding of how face processing develops in the brain could help identify approaches that are more useful and appropriate.
An ongoing challenge in autism research is separating neurological differences associated with autism itself from changes that may develop through the experience of living with the condition. According to the researchers, the patterns identified in this study suggest that differences in face processing are connected to underlying developmental characteristics of autism rather than simply reflecting accumulated 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 the differences appear during development, the findings also raise the possibility that support aimed at helping autistic children interpret facial emotions could be particularly beneficial when introduced earlier in childhood.
However, the researchers emphasize that autism does not necessarily create difficulties that require intervention. Many neurodivergent people experience little or no impairment related to their autism and may neither need nor want additional support. For others, challenges with recognizing faces or understanding facial expressions can complicate everyday 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.”
New Brain Signals Could Help Improve Autism Diagnosis
Another potential application involves the search for more objective ways to assess autism. Currently, clinicians diagnose autism primarily by evaluating behavior because reliable biological tests are not yet available for routine diagnosis.
Scientists are investigating several possible biomarkers, which are measurable biological characteristics that could improve diagnostic assessments or help distinguish different groups within the autism spectrum.
The newly identified patterns of face related brain activity could contribute to that effort. By examining how these neural signals differ among individuals, researchers may eventually develop additional diagnostic tools or better identify autistic people who are likely to benefit from particular forms of support, including social skills therapy.
Although more research is necessary before these findings can be used in clinical practice, the study provides a new way to investigate how the autistic brain processes faces and how those responses change throughout childhood.
“This is a first step in introducing a new biomarker,” McPartland says.
