New Insights into Autism Subtypes

Autism is a complex condition that manifests differently in each individual. Researchers have made significant strides in understanding autism over the years, but classifying distinct subtypes has proven challenging. Recently, though, a breakthrough may have been achieved.

An international team of researchers has identified two subtypes of autism, using evidence gathered from both human and mouse brains. By employing a cross-species analysis, they confirmed biological differences between these subtypes.

These findings could lead to more tailored therapies and support programs for individuals with autism, moving away from the traditional ‘one size fits all’ approach. Neuroscientist Alessandro Gozzi from the Italian Institute of Technology remarked on the variability in autism manifestations that they’ve observed for decades, noting that they lacked concrete evidence linking these differences to distinct biological factors.

Gozzi explained that their approach allowed them to pinpoint specific genetic and immune factors, which they then translated to human brain scans. This process revealed that different connectivity patterns correlate with various underlying mechanisms of autism.

The research team analyzed brain scans from mice with 20 different autism-like traits, along with scans from 940 children and young adults with autism, and 1,036 neurotypical individuals. From this, two notable groups began to emerge.

The first group, labeled as hypoconnectivity, displayed reduced brain connectivity, where specific genes related to synapse junctions were implicated. On the other hand, the hyperconnectivity group showed increased connectivity, linking them to genes associated with the immune system and indicating slightly more severe autism.

The consistency of these findings across both mice and humans adds credibility to the idea that these are legitimate subtypes of autism. However, it’s worth noting that only about one in four of the analyzed human brains with autism fell into these categorized groups.

Neuroscientist Adriana Di Martino from the Child Mind Institute likened the mouse studies to a biological ‘Rosetta Stone,’ allowing researchers to identify which biological pathways influence specific connectivity patterns, which could then be searched for in humans.

While this is a promising development, more work lies ahead. Confirming and diagnosing these hypoconnectivity and hyperconnectivity subtypes could pave the way for therapies targeting specific autism categories based on the biological characteristics identified in this study.

This isn’t the first attempt to categorize autism. A 2025 study identified four types in a sample of 5,000 children based on a wide range of behavioral traits rather than brain imaging. Other research suggests that autism’s expression can vary depending on when it develops, whether in early childhood, late childhood, adolescence, or young adulthood—all contributing to a richer understanding of autism.

While autism is often described as a spectrum to represent the variety of communication styles and behaviors, some experts argue that this characterization may not fully encapsulate what being neurodiverse entails. They advocate for new approaches that could lead to better identification and understanding.

Researchers believe that advancements in techniques and larger datasets will allow for the identification of even more subtypes in the future. In the meantime, they’ve made their data and analytical tools publicly available for other scientists to build on this work.

As stated in their published paper, the researchers hope their cross-species approach will create a more nuanced, biologically grounded framework for understanding autism. They have also made their database accessible to the research community, facilitating future studies into autism-related connectivity alterations.

The findings were published in Nature Neuroscience.