A new study published in Nature Genetics (2025) has identified four distinct subtypes of autism, each differentiated by their own set of unique challenges and associated with specific genetic variations. According to co-lead author of the paper Natalie Sauerwald, “For families navigating autism, understanding their child’s specific subtype can provide greater clarity and open the door to more personalized care, support, and connection. 

The four subtypes that the researchers identified are based on the challenges individuals with autism face:

1. Social/Behavioral Challenges → difficulties with social communication and restrictive and repetitive behaviors and don’t experience developmental delays
2. Mixed ASD + Developmental Delay → showed developmental delays and had more varied social communication and repetitive behaviors
3. Moderate Challenges → consistently fewer difficulties with social communication and repetitive behaviors and do not experience developmental delays
4. Broadly Affected → severe and wide-range difficulties with social communication and repetitive behaviors, including developmental delays, compared to other autistic children

The research itself is no doubt impressive in scope and credibility. Published in one of the most prestigious scientific journals, Nature Genetics (impact factor: 29.0 in 2024), the findings were replicated in multiple cohorts, including the Simons Simplex Collection. The results reached statistical significance and offer a compelling perspective into the heterogeneity of autism. 

Still, while this is a fascinating study and could prove to be a critical bridge in our understanding, I have some reservations that are worth considering: 

Limited Representation of Participants

One major limitation is the cultural background of the sample. Most of the participants were of European ancestry, and the group skewed heavily toward white male boys. Obviously, this is the beginning stage for this research. If the researchers want to use this information in clinical settings, it would be more helpful to look at the data in a varying group of people. 

Complexity of Autism as a Spectrum

Stephen Shore, an autistic professor at Adelphi University once said, “If you’ve met one person with autism, you’ve met one person with autism.” This quote highlights that there are definitely core features in autism, but these features vary significantly in their severity and presentation among each individual. The experiences of those who have an autism diagnosis are incredibly diverse and vary broadly. In short, autism is very heterogenous and to classify autism into four broad categories seems a little dangerous. 

Another important nuance I wanted to bring up was that the four different classes were identified through extensive quantitative data, whereas most autism diagnoses require a lot of qualitative interpretation where clinicians observe and assess behavior in context. Bridging these two methods will definitely take time. 

What is particularly fascinating, however, is that there are noticeable genetic differences across the four subgroups. For example, the Broadly Affected group showed the strongest genetic signal and was made up of significantly more rare and potentially harmful mutations compared to individuals who don’t have autism and other autism subtypes. In contrast, the Social/Behavioral group had the lowest amount of these rare mutations. All four subtypes, however, did show a higher impact from these mutations compared to siblings who aren’t autistic. This kind of genetic lens provides underlying biological patterns that may drive the physical differences some individuals have. For example, certain genetic variants associated with increased sensory sensitivity may explain why differences occur in the first place.

Conclusion

This study is a meaningful step towards a more individualized understanding of autism, one that could eventually inform better diagnostic tools and individualized support. While I have some concerns about oversimplification, I appreciate that the researchers present these subtypes as a framework, rather than a definitive final answer. Overall, combining genetics and biology with our understanding of autism as a behavioral condition is incredibly important in our approaches to learning about autism.