Is autism really just one condition?
No. A new study of nearly 2,000 brain scans found that autism splits into two biologically distinct subtypes, not a single condition. One subtype shows too little brain connectivity tied to faulty synapses, and the other shows too much connectivity tied to immune-related changes.
For years, doctors and scientists have noticed that no two autistic people look exactly alike. Symptoms, language skills, and brain patterns vary widely from person to person. Many assumed this surface-level variety reflected different underlying biology, but no one had solid proof. This study set out to find that proof by looking deep inside the brain’s wiring.
How the researchers cracked the problem
The team used functional MRI, a brain scan that measures how different brain regions talk to each other while a person rests. They focused on something called connectivity, which is just a measure of how strongly brain areas connect and communicate.
Here is the clever part. The researchers did not start with people. They started with mice. They scanned the brains of 20 different mouse models, each carrying a genetic change linked to autism. When they grouped the results, the mice fell into two clear camps. One group had unusually low brain connectivity, and the other had unusually high connectivity.
What the data show
The two mouse groups were not random. The low-connectivity group, called hypoconnectivity, was tied to problems with synapses, the tiny junctions where brain cells pass signals to each other. The high-connectivity group, called hyperconnectivity, was tied to gene activity and immune-related pathways instead.
Then the team checked whether the same split appears in people. They analyzed brain scans from 940 individuals with autism and 1,036 neurotypical individuals, drawn from a large multicenter dataset. The same two subtypes showed up in humans. The patterns were highly replicable, meaning they held steady across different groups and scanning sites. Each human subtype had its own brain network wiring and its own behavioral profile, with modest differences in autism severity between the two. Most striking, the human subtypes echoed the very same synaptic and immune pathways first spotted in the mice.
Dr. Kumar’s Take
This is the kind of study that quietly moves a field forward. For decades we have treated autism as one diagnosis, even though every clinician knows it presents in wildly different ways. What I find compelling here is the cross-species design. By anchoring the human brain patterns to mouse models with known genetic causes, the researchers did something most subtyping studies cannot. They grounded the categories in real biology, not just statistics.
I want to be honest about the limits, though. This does not change how autism is diagnosed today, and no one should walk into a clinic asking which subtype they are. We do not yet have a brain scan that sorts individuals reliably in everyday care. But the direction is exciting. If autism is truly two or more biological conditions, then treatments could one day be matched to the right person, instead of a one-size-fits-all approach.
Why this matters
The practical promise here is precision medicine. A drug that targets synaptic signaling might help the hypoconnectivity group but do nothing for the immune-linked hyperconnectivity group, and vice versa. Lumping everyone together in a single trial could hide a treatment that works beautifully for half of participants. Sorting people by their underlying biology first could rescue therapies that otherwise look like failures.
There is also a deeper point about how we understand autism. Only about 20 percent of autistic people carry a known high-impact genetic change, and no single mutation explains more than 1 percent of cases. That has made biological stratification nearly impossible from genes alone. Using brain connectivity as a readout offers a fresh path around that roadblock.
Practical Takeaways
- This research is early and lab-based, so do not expect a brain scan to classify your or your child’s autism subtype at a doctor’s visit yet.
- The findings support the idea that autism is biologically diverse, which is a useful reminder that what helps one person may not help another.
- If you join or consider an autism clinical trial in the future, ask whether the study accounts for biological subtypes, since this approach may shape better treatments.
- Keep following this space, because subtype-based research is likely to influence how new autism therapies are designed and tested.
Related Studies and Research
- Single-dose psilocybin shows rapid, sustained antidepressant effects
- A single gene-editing infusion cut bad cholesterol by 62%
- Being aerobically fit may protect you from 34 different diseases
- Fish oil supplements linked to faster brain decline
FAQs
What is the difference between hypoconnectivity and hyperconnectivity in autism?
Hypoconnectivity means brain regions communicate less than expected, while hyperconnectivity means they communicate more than expected. In this study, the low-connectivity pattern was linked to synaptic problems, the connection points between brain cells. The high-connectivity pattern was linked to gene activity and immune-related changes. These are not better or worse versions of autism, just biologically different routes that may need different treatments.
Does this mean my child’s autism can be diagnosed with a brain scan now?
No, not in routine care. This was a research study using specialized fMRI methods across many scans, not a clinical test you can request from your pediatrician. The patterns were reliable across large groups, but the science is not yet precise enough to sort one individual into a subtype with confidence. For now, diagnosis still relies on behavioral and developmental assessments by trained clinicians.
Why did the researchers study mice to understand human autism?
Mice with specific autism-linked genetic changes let scientists isolate biology without the many confounding factors present in humans. By scanning 20 different mouse models, the team could connect brain wiring patterns directly to known genetic and immune causes. They then checked whether the same patterns appeared in human scans, and they did. This cross-species approach gives the human subtypes a stronger biological foundation than statistics alone could provide.
Bottom Line
This study offers some of the clearest evidence yet that autism is not a single condition but at least two biologically distinct ones. By matching brain-connectivity patterns in nearly 2,000 people against 20 autism mouse models, researchers identified a synaptic subtype marked by low connectivity and an immune-related subtype marked by high connectivity. The work does not change diagnosis today, but it lays a foundation for precision medicine, where treatments could one day be matched to a person’s underlying biology rather than a one-size-fits-all label.
