Brain Imaging Reveals Two Distinct Autism Subtypes With Different Biological Roots
Autism may include at least two distinct biological subtypes, defined by opposing patterns of brain connectivity, according to a new international study. Using advanced brain imaging and animal models, researchers identified one autism subtype with unusually high connectivity and another with reduced connectivity between brain regions.
The findings add to growing evidence that autism is not a single condition but a spectrum of related disorders with different underlying biological mechanisms. Scientists say the results could eventually support more personalized diagnostic and treatment approaches, although clinical applications remain years away. The study was published in Nature Neuroscience.
Two Brain Connectivity Patterns Identified
The research was led by teams at the Istituto Italiano di Tecnologia in Italy and the Child Mind Institute in New York, with support from the University of Trento and other international collaborators.
Researchers analyzed functional MRI scans from 940 children and young adults with autism and more than 1,000 neurotypical individuals. They focused on functional connectivity, a measure of how strongly different brain regions communicate with one another.
The analysis revealed two distinct connectivity patterns. One subgroup showed hypoconnectivity, meaning weaker communication across brain networks, while another displayed hyperconnectivity, or stronger-than-typical communication between regions. Together, these patterns were identified in approximately 25% of autistic participants.
The two subtypes also showed subtle differences in overall brain organization and standard clinical measures of autism. Individuals in the hyperconnected group tended to score slightly higher on autism severity assessments. However, both patterns appeared across a wide range of ages and cognitive abilities, highlighting the complexity and diversity of the autism spectrum.
Mouse Models Revealed Underlying Biology
To better understand the biological mechanisms behind these connectivity patterns, the researchers combined human imaging data with findings from 20 different mouse models carrying autism-related genetic or immune-system alterations.
In these animal models, scientists paired functional connectivity mapping with genetic and biochemical analyses at the cellular level. This approach allowed them to connect specific brain-network signatures with underlying molecular pathways.
The results showed that hypoconnectivity was strongly associated with alterations in synapses, the specialized junctions through which neurons communicate. Hyperconnectivity, by contrast, was linked to pathways involving immune function and neuroinflammation.
These biological signatures served as templates for interpreting the human brain scans. When similar connectivity patterns were observed in people, the findings suggested that comparable synaptic or immune-related processes might be influencing brain function, even though direct tissue analysis is not possible in living individuals.
Human Data Supported the Same Pattern
Most of the human brain scans came from the Autism Brain Imaging Data Exchange, a large international repository that combines MRI data from dozens of research centers worldwide. Additional scans were provided by the Child Mind Institute and collaborating institutions.
When researchers compared the connectivity patterns with molecular data, they found a remarkably consistent picture. Brain regions associated with the hypoconnected subtype were enriched for genes involved in synaptic function, while hyperconnected regions showed enrichment for immune-related genes.
Importantly, these findings were replicated across independent datasets and multiple imaging sites. According to the authors, this consistency reduces the likelihood that the results were caused by scanner differences, participant movement during scanning or other technical factors that commonly affect neuroimaging studies.
Potential Implications for Precision Medicine
The researchers caution that the findings do not yet translate into immediate clinical tools or treatments. Autism diagnosis continues to rely on behavioral assessments and developmental history, and the newly identified subtypes explain only part of autism's biological diversity. Many autistic individuals in the study did not fit clearly into either category.
Nevertheless, the work represents an important step toward precision medicine in neurodevelopmental disorders. Identifying subgroups linked to synaptic mechanisms versus immune-related pathways could help guide future drug development, improve patient selection in clinical trials and support more individualized treatment strategies.
The authors also note that larger and more diverse datasets, combined with higher-resolution genetic information, may reveal additional biological subtypes in the future. Further studies will investigate how these connectivity patterns change throughout development, respond to treatment and relate to co-occurring conditions such as anxiety, ADHD and epilepsy.
A Growing International Effort
The project was supported by the Simons Foundation Autism Research Initiative, the European Research Council, the Brain and Behavior Foundation, Fondazione Telethon and the U.S. National Institute of Mental Health.
Researchers say the study highlights the value of combining brain imaging, genetics and animal research to better understand the biological foundations of autism. As these approaches continue to evolve, they may help explain why autistic individuals can differ so substantially in their symptoms, strengths and support needs.
