- A new review proposes autism can emerge from a three hit combination of genetic sensitivity, early life triggers and prolonged cellular stress
- The model centres on a sustained cellular danger response, with extracellular ATP acting as an ongoing danger signal
- It is a unifying theory rather than proof of cause, and it needs careful testing before it can guide prevention or treatment
Researchers have long struggled to tie together the many genetic and environmental factors linked with autism into a single explanation.
A new review led by Dr Robert Naviaux at UC San Diego argues there may be a way to connect the dots, by treating autism as the downstream result of a prolonged metabolic stress response during early development.
The paper lays out a three hit model. First, some children inherit a sensitive biological setup, with mitochondria and signalling systems that react strongly to environmental change.
Second, triggers during a critical window, from pregnancy through early childhood, push those sensitive systems into a stress state.
The proposed triggers include maternal immune activation, pollution and metabolic stressors.
Third, the key step, that stress state does not switch off, persisting for months during late pregnancy or early childhood.
The mechanism used to link these hits is the cell danger response, a built in cellular programme that shifts the body towards defence.
- Soybean oil and obesity under the microscope
- How the gut microbiome links obesity to bowel cancer
- Ultra-processed foods are fuelling a rise in autism cases, study highlights
In Naviaux’s framing, extracellular ATP acts as a danger signal that keeps cells in that defensive mode.
If the signal stays high for too long, the body stays focused on protection rather than growth and development, which the model suggests could disrupt brain development and influence immune function and gut microbes.
This is also offered as a reason why some autistic children experience broader physical symptoms such as gastrointestinal problems or sleep disruption.
The review also points to phenylketonuria as a proof of concept that early metabolic intervention can prevent severe developmental impact in a condition with a clear biological trigger, arguing that autism could one day have its own prevention or early intervention playbook.
The paper goes further by suggesting that a substantial share of cases might be reduced through better support during pregnancy and early childhood, but that estimate is speculative and depends on the model being correct.
It is worth being blunt about what this is and is not.
This is a theory that tries to organise a messy body of evidence into one framework.
It does not prove that chronic activation of the cell danger response causes autism, and it does not yet give clinicians a validated test or a proven treatment strategy.
What it does offer is a structured way for researchers to ask sharper questions about metabolism, immune signalling and environmental exposures during early development.
Reference: Naviaux RK. A 3 hit metabolic signalling model for the core symptoms of autism spectrum disorder. Mitochondrion. 2026;87:102096. doi: 10.1016/j.mito.2025.102096
