People with the same level of autistic traits show greater social attraction to each other, and their brains synchronize in unique ways during active conversations. A recent experiment published in the journal Biological Psychiatry suggests that the social difficulties associated with autism may be a matter of mismatched communication styles rather than an innate social deficit.
For decades, clinical psychology has treated autism primarily as a social disorder. This view assumes that people with autism lack a cognitive tool called theory of mind. This concept refers to the ability to intuitively understand what other people are thinking and feeling. Recently, alternative frameworks have emerged that challenge this deficit-based assumption.
One major alternative is the problem of dual empathy. This idea suggests that social friction is a two-way street. Neurotypicals and autistic people experience the world and process sensory information very differently. These differences can lead to mutual misunderstandings, and neurotypical people can have trouble reading the minds of autistic people.
Based on this concept, scientists developed the dialectical disharmony hypothesis. This hypothesis relies on predictive coding, a theory that explains how the brain constantly generates inferences about what will happen next. When events match your guesses, interactions feel smoother. When a person’s behavior violates those expectations, the brain experiences a prediction error, leading to social awkwardness.
Following this logic, people who share similar psychological characteristics should be able to more easily predict each other’s behavior. Autistic people who avoid eye contact can cause prediction errors for neurotypical people who expect steady gaze. However, other autistic people would not find this behavior unusual. This common wavelength should lead to more fluid interactions and a sense of mutual attraction.
First author Shuyuan Feng, in collaboration with Peng Zhang and Xuejun Bai from Tianjin Normal University in China, designed an experiment to test these ideas. Previous studies assessing how well people with different autism traits bond have yielded conflicting results. The researchers suspected that the initial experimental design may have caused this discrepancy.
Previous research has typically placed two people in a room to interact. This method makes it difficult to distinguish between a person’s general friendliness and their specific compatibility with their partner. By bringing together larger groups, researchers can use a mathematical approach called the social relations model. This model separates genuine interpersonal attraction from general social tendencies.
The research team measured autistic traits in hundreds of college students using a standardized questionnaire. The students had not been clinically diagnosed with autism. Instead, they conducted a study that scored common behavioral and cognitive characteristics associated with the autism spectrum. The research team selected the highest-scoring and lowest-10 percent students to represent the high-scoring and low-scoring autistic trait groups.
The researchers then assembled an isolated group of four people who had never met. Each group included two individuals with high autistic traits and two individuals with low autistic traits. In total, the study included 20 female groups and 10 male groups.
The researchers monitored the participants’ brain activity using functional near-infrared spectroscopy. This technology uses a small optical sensor placed firmly on the scalp. The sensor uses light to measure blood oxygen levels in specific areas of the brain, highlighting in real time which areas are working the hardest. Participants wore these sensor caps while completing a series of social tasks.
First, the group sat quietly and listened to the audio story. This passive task allowed the researchers to measure how similarly the brains responded to exactly the same information. Scientists use an analysis called between-subject correlation for this purpose. It measures the overlap in brain responses of different people experiencing the same audio stimulus.
Next, participants engaged in a structured group discussion. They discussed classic survival scenarios and decided which fictional character to rescue from a deserted island. The discussions followed strict turn-taking rules to prevent brain data from being clouded by people talking to each other. After the discussion, participants privately rated how much they wanted to continue talking to or be friends with each member of the group.
The responses revealed clear patterns of interpersonal relationships. Participants with similar levels of autistic traits reported a stronger desire to be social with each other. Those high on autistic traits were attracted to other high-scoring group members, and those low on autistic traits were attracted to each other.
This mutual affinity only emerges when there is consensus during a survival task. Personality similarities such as extraversion did not promote attraction. Instead, agreeing on the topic of survival allowed individuals with similar characteristics to recognize a deeper common perspective. This shared vision became the basis of their social appeal.
Brain scans have provided insight into how these connections are formed at a biological level. During a passive story listening task, pairs low in autistic traits showed similar neural responses to audio. Pairs with high levels of autistic traits had more diverse and unique brain responses to the same story.
When the group discussion was more communicative, brain activity was adjusted in different directions. The researchers measured brain-to-brain synchronization, a phenomenon in which two people’s brain waves coincide during a shared activity. A higher degree of synchronization indicates a smoother and more effective transfer of information between the two minds.
Pairs low in autistic traits showed higher brain synchronization at the right temporoparietal junction. This area of the brain is deeply involved in social cognition. People use this region to automatically process social cues and read the implicit intentions of their conversation partners.
Pairs with advanced autistic traits showed completely different neural patterns. Their brains were synchronized in the right dorsolateral prefrontal cortex. This area of the brain manages cognitive control, sustained attention, and deliberate problem solving.
This neural pattern suggests that people with highly autistic traits rely on different mental strategies to handle social interactions. Rather than relying on automatic social cognition, they may intentionally mobilize additional cognitive resources to build connections. This strategy allows you to better coordinate your brain activity with a like-minded partner.
These findings call into question models that depict autism solely as a disorder of social cognition. Instead of failing to communicate, people with advanced autistic traits appear to use different neural pathways that can fully support social bonds. Brain imaging studies support the idea that social conflict may arise from mismatched cognitive strategies rather than from an inherent inability to connect.
This study has several limitations that should be considered. Neuroimaging devices only detect blood flow near the surface of the brain. The deeper brain structures involved in social reward processing have not yet been observed. Additionally, the structured nature of time-limited experimental tasks may not capture the organic flow of everyday social life.
The participants in the study were college students with varying levels of autistic traits, rather than individuals formally diagnosed with autism spectrum disorder. The researchers noted that future studies could apply these methods to clinical populations. Using larger imaging machines could also help map the deeper neural networks associated with these unique communication styles.
The study, “Attraction through Similarity in Autistic Traits: A Group Communication Study Using Social Relationship Model and Functional Near-Infrared Spectroscopy Hyperscanning,” was authored by Shuyuan Feng, Mingliang Wang, Jianing Zhang, Lin Ding, Yuqing Yuan, Peng Zhang, and Xuejun Bai.
