For years, mainstream neurodevelopmental research overwhelmingly focused on abnormalities within the cerebral cortex—the brain region responsible for thinking, learning, and processing information—as the primary driver of autism.
However, a May 2026 study suggests that another part of the brain, the cerebellum, may play a much larger role than ever thought.
Researchers discovered that perineuronal nets (PNNs), the tiny supportive structures surrounding cerebellar neurons, help maintain healthy communication across brain networks involved in social behavior.
When these structures were disrupted, social interactions changed, offering a fresh perspective on the biological mechanisms that may contribute to autism spectrum disorder (ASD).
Disrupted cerebellar support cells were found to affect social behavior linked to autism
Image credits: Unsplash (Not the actual photo)
The study was conducted by Japan’s Kanazawa University, and the results were published in the journal Translational Psychiatry on May 13.
The researchers examined two different mouse models of autism for the purpose.
Image credits: Pexels (Not the actual photo)
In one model, autism-like traits were induced by exposing mice to valproic acid before birth, while in the other, the mice carried a mutation in the autism-associated gene CHD8.
Despite their different origins, both models showed a significant loss of perineuronal nets (PNNs) in the cerebellum.
Image credits: Pexels (Not the actual photo)
To understand the impact of this change, scientists deliberately disrupted PNNs in healthy mice.
This yielded striking results, with mice showing damaged PNNs exhibiting reduced social interaction and less interest in unfamiliar mice — behaviors commonly associated with ASD.
Further experiments revealed why this happened.
Image credits: Pexels (Not the actual photo)
In healthy mice, social interactions activated neurons in the cerebellum, which then sent signals to other brain regions involved in processing social information. When PNNs were disrupted, this neural activity dropped significantly, weakening communication across multiple brain networks.
Image credits: Unsplash (Not the actual photo)
The team also identified a key molecular player called ARNT2. Levels of this protein increased when PNNs were lost, making neurons less responsive.
Reducing ARNT2 activity restored both normal brain function and social behavior in the mice.
The findings suggest that autism-related social difficulties may not stem solely from changes in the brain’s thinking centers. Instead, structural changes in the cerebellum and the networks it controls could also play an important role.
While more research is needed to determine whether the same mechanism exists in humans, the study offers a promising new direction for understanding the biology of autism.
The study came shortly before another research effort identified two subtypes of autism
Image credits: Unsplash (Not the actual photo)
This study, titled Autism Subtypes Identified Using Cross-Species Functional Connectivity Analyses, was led by researchers at the Instituto Italiano di Tecnologia in Rovereto, Italy, and the Child Mind Institute in New York, with additional contributions from the University of Trento.
Their findings were published in Nature Neuroscience on May 15.
Image credits: Pexels (Not the actual photo)
Researchers examined brain scans from 20 different mouse models that showed autism-like traits, as well as scans from 940 children and young adults with autism and 1,036 neurotypical individuals.
They were looking for differences in how various parts of the brain communicate with one another.
The analysis revealed two main patterns.
Image credits: Unsplash (Not the actual photo)
First was a hypoconnectivity group, where autism was associated with reduced brain connectivity. Here, brain activity was linked to genes involved in synaptic junctions that enable brain cells to communicate.
Then there was a hyperconnectivity group, associated with increased connectivity across the brain. The group’s brain patterns were linked to genes related to the immune system and showed measures of slightly more severe autism.
Because the same patterns were found in both mice and humans and replicated across multiple datasets, the researchers believe these patterns may represent real biological subtypes of autism.
If future research confirms these subtypes and reliable diagnostic methods are developed, it could lead to more personalized treatments tailored to each subtype.
This wasn’t the first time researchers tried to spot matching patterns and split autism into several types
Image credits: Unsplash (Not the actual photo)
A study published in July 2025 by researchers from Princeton University, the Simons Foundation, and the Flatiron Institute identified four types of autism in a group of 5,000 children.
Instead of using brain imaging, the researchers defined these types by analyzing more than 230 behavioral characteristics.
Other studies have suggested that autism can appear differently depending on when it develops, such as in early childhood, late childhood, adolescence, or young adulthood.
Together, these studies aim to improve how autism is identified and understood.
Image credits: Unsplash (Not the actual photo)
The researchers believe that larger datasets and more advanced analysis methods will help uncover additional autism subtypes in the future.
The team behind Autism Subtypes Identified Using Cross-Species Functional Connectivity Analyses has made its data and analysis tools publicly available, enabling other scientists to build on its findings.
“Our cross-species approach provides an advanced translational framework for a multidimensional, biologically grounded stratification of autism,” the researchers wrote in the published paper.
“Our database is openly available to the research community, supporting future investigations into autism-related connectivity alterations,” they added.
“I have three kids, all diagnosed with autism. I am realizing I am also autistic. Genetics!” a netizen wrote
