Groundbreaking study identifies four biologically distinct subtypes of autism, revolutionizing our understanding of diagnosis and treatment
For decades, autism has been described as a spectrum—a term that captures the tremendous variation in how it manifests but has long frustrated scientists trying to understand its biological underpinnings. Why do some autistic children speak in full sentences while others remain nonverbal? Why do some excel academically while others struggle with basic learning? Why do some receive diagnoses as toddlers while others aren't identified until adolescence?
A groundbreaking study published in July 2025 in Nature Genetics has identified biologically distinct subtypes of autism, each with different developmental trajectories, clinical presentations, and genetic signatures 1 .
This research doesn't just offer new clues to autism—it provides an entirely new framework for understanding what we now call the autism spectrum.
Autism spectrum disorder (ASD) affects approximately 1.5% of the global population 3 , characterized by alterations in communication, socialization, and repetitive behaviors or restricted interests. While the spectrum concept helpfully captures the diversity of autistic experiences, it has created significant challenges for researchers and clinicians alike.
The SFARI Gene Database lists 913 genes with varying levels of evidence linking them to autism 3 .
The transformative 2025 study took a radically different approach to unravel autism's complexity. Led by researchers from Princeton University and the Simons Foundation 1 6 , the team analyzed data from over 5,000 children enrolled in SPARK—the largest autism cohort study to date—using sophisticated computational methods that could detect patterns invisible to traditional analysis.
Instead of examining single traits in isolation, the team employed a "person-centered" approach that considered over 230 traits simultaneously for each individual 1 .
The researchers used general finite mixture modeling, uniquely suited to handle diverse data types—from simple yes/no questions to continuous measures like age at reaching developmental milestones 6 .
The analysis leveraged the SPARK consortium's massive dataset, which includes both extensive phenotypic information and genetic data 6 .
"As I spoke with autism researchers about leveraging our computational tools to analyze phenotypic and genotypic data from SPARK, I realized this cohort is the only one that has this combination of extensive phenotypic data as well as genetic data," noted Natalie Sauerwald, a study co-lead author 6 .
| Research Tool | Function in Research | Application in the 2025 Study |
|---|---|---|
| SPARK Cohort Data | Large-scale dataset with phenotypic and genetic information | Provided clinical and genetic data for over 5,000 autistic individuals |
| General Finite Mixture Models | Computational method for identifying subgroups in complex data | Identified the four autism subtypes based on trait combinations |
| Whole Genome Sequencing | Comprehensive analysis of genetic code | Identified both inherited and de novo genetic variations |
| SFARI Gene Database | Curated database of autism-associated genes | Provided reference for known autism genetic risk factors |
| Pathway Analysis Tools | Computational methods for identifying biological processes | Revealed distinct biological pathways for each autism subtype |
The analysis revealed that autistic individuals naturally clustered into four distinct groups, each with characteristic clinical profiles and outcomes. Critically, these categories emerged purely from the data without preconceived hypotheses about what autism "should" look like.
Prevalence: 37%
Core autism traits plus multiple co-occurring conditions. Typically reached developmental milestones on time.
Click for detailsPrevalence: 19%
Developmental delays but fewer psychiatric conditions. Reached milestones later than peers.
Click for detailsPrevalence: 34%
Milder core autism traits. Typically reached developmental milestones on time.
Click for detailsPrevalence: 10%
Widespread significant challenges across multiple domains. Reached milestones later than peers.
Click for detailsThe subtypes displayed markedly different developmental courses. Children in the Social and Behavioral Challenges group typically reached developmental milestones like walking and talking at ages similar to non-autistic children, whereas those in the Mixed ASD with Developmental Delay group and Broadly Affected group reached these milestones later 1 .
Typically reached milestones on time, but often diagnosed later due to typical early development.
Reached developmental milestones later than peers, with fewer psychiatric conditions.
Typically reached milestones on time with milder core autism traits.
Reached milestones later than peers with widespread significant challenges.
After establishing the four subtypes based on clinical characteristics, researchers made their most startling discovery: each subtype had distinct genetic signatures and biological pathways.
The research uncovered that genetic disruptions affected different developmental timepoints for each subtype.
Each subtype was linked to distinct molecular pathways previously implicated in autism but now specifically associated with different presentations 6 .
"What we're seeing is not just one biological story of autism, but multiple distinct narratives," explained Natalie Sauerwald. "This helps explain why past genetic studies often fell short—it was like trying to solve a jigsaw puzzle without realizing we were actually looking at multiple different puzzles mixed together." 1
These discoveries have profound implications for how we understand, diagnose, and support autistic individuals across the lifespan.
The ability to classify autism into biologically distinct subtypes could revolutionize clinical practice.
"If you know that a person's subtype often co-occurs with ADHD or anxiety, for example, then caregivers can get support resources in place and maybe gain additional understanding of their experience and needs," noted Natalie Sauerwald 6 .
The findings fundamentally shift how scientists can approach autism research.
"This opens the door to countless new scientific and clinical discoveries," said Chandra Theesfeld, senior academic research manager at the Lewis-Sigler Institute and Princeton Precision Health 1 .
| Subtype | Potential Educational Supports | Potential Medical Considerations | Long-Term Planning |
|---|---|---|---|
| Social & Behavioral Challenges | Social skills support, mental health resources | Monitoring for anxiety, depression, ADHD | Mental health support, workplace accommodations |
| Mixed ASD with Developmental Delay | Academic support, speech and language therapy | Physical therapy, monitoring for seizures | Supported living options, vocational training |
| Moderate Challenges | Mild accommodations, social communication support | Generally minimal | Independent living with some supports |
| Broadly Affected | Comprehensive special education supports | Multi-specialty care, medication management | 24-hour support, guardianship planning |
While these findings represent a monumental leap forward, researchers emphasize that this is just the beginning. The four subtypes likely don't represent the final word in autism classification, but rather a foundational framework for future discovery.
This research doesn't diminish the value of neurodiversity. Instead, it enhances our understanding of that diversity, acknowledging that different neurotypes may require different supports.
The discovery of biologically distinct autism subtypes marks a watershed moment in autism research—one that could ultimately transform how we diagnose, support, and understand autistic individuals.
By recognizing that autism comprises multiple conditions with different biological bases and developmental trajectories, we can move beyond one-size-fits-all approaches to personalized support strategies.
"Understanding the genetics of autism is essential for revealing the biological mechanisms that contribute to the condition, enabling earlier and more accurate diagnosis, and guiding personalized care" — Olga Troyanskaya, senior author 1
For the millions of autistic individuals and their families worldwide, these new clues to autism offer more than scientific insight—they offer the promise of a future where support and understanding are as unique as the individuals themselves.