An international team of researchers has reported evidence that autism may not be a single condition but instead encompasses biologically distinct subtypes, each defined by a different pattern of brain connectivity. The findings, published in the journal Nature Neuroscience, add to a growing effort to map the biological diversity behind a diagnosis that varies widely from person to person.
The work was led by scientists at the Istituto Italiano di Tecnologia in Italy and the Child Mind Institute in New York, working with collaborators including the University of Trento. The team analyzed brain-imaging data from hundreds of autistic individuals drawn from a large shared research dataset, and combined those human scans with data from genetically engineered mouse models to probe the underlying biology.
Two reproducible subtypes emerged from the analysis. One was characterized by reduced communication between brain regions, a pattern the researchers describe as hypoconnectivity, which appeared linked to synaptic and cell-signaling pathways. The other showed increased communication between regions, or hyperconnectivity, and was associated with immune-related biological systems. Participants in the hyperconnectivity group tended to show moderately higher scores on measures of autism severity.
The researchers framed the results as an early, biology-based way of organizing autism’s complexity rather than a finished diagnostic system. By anchoring the subtypes to specific brain communication patterns and associated genetic pathways, the study points toward the possibility that different forms of autism could one day be approached differently in research and, eventually, in care.
Important caveats accompany the findings. The authors note that a large share of autistic participants did not fall cleanly into either subtype, suggesting their brain patterns may be more subtle, mixed or shaped by biological routes the study did not capture. The two categories, the researchers caution, are likely part of a much larger picture, and the work is not intended to create clinical labels or to oversimplify a condition that presents in many ways.
Autism has long challenged researchers precisely because people who share the diagnosis can differ enormously in their traits, support needs and co-occurring conditions. That heterogeneity has complicated the search for consistent biological markers and has made it difficult to develop tailored interventions. Studies that attempt to sort the condition into more defined biological groupings are seen by many in the field as a step toward so-called precision approaches.
Independent experts are likely to scrutinize whether the subtypes hold up in other populations and imaging datasets, a standard requirement before such findings can shape practice. Brain-connectivity research can be sensitive to how data are collected and analyzed, and replication across larger and more diverse groups will be key.
For families and clinicians, the immediate practical impact is limited, and the study does not change how autism is currently diagnosed or supported. Still, the researchers argue that establishing a biological framework, even a provisional one, is a necessary foundation for understanding the condition’s roots and for guiding future investigation.
