Despite widespread claims linking gut health to autism spectrum disorder, a comprehensive analysis published in the journal Neuron has found no scientific evidence supporting the hypothesis that the gut microbiome causally contributes to autism.
The paper, authored by researchers from Trinity College Dublin, University College Cork, and the University of Oxford, systematically dismantles decades of research that has fuelled a lucrative wellness industry and sparked controversy over misleading public health messaging.
“Despite what you’ve heard, read, or watched on Netflix, there is no evidence that the microbiome causally contributes to autism,” says lead author Kevin Mitchell, a developmental neurobiologist at Trinity College Dublin in a press statement. “I don’t think it’s warranted to spend further time and funding on this topic. We know that autism is a strongly genetic condition, and there’s still loads to be worked out there”.
The hypothesis that autism stems from gut microbiome abnormalities has gained significant traction in recent years, with research funding reaching $20-25 billion annually since 2018 and over 100 published articles in 2024 alone. This interest largely derives from observations that many autistic individuals experience gastrointestinal symptoms and the mistaken belief that rising autism diagnoses must reflect environmental changes.
However, the authors argue that these foundational assumptions are themselves problematic.
According to their paper, strong evidence indicates that increased autism diagnoses reflect improved awareness and broadened diagnostic criteria rather than any biological epidemic. In simple terms, we have simply become much better at diagnosing autism, and it has nothing to do with gut health, or any external change or force.
The researchers meticulously examined the most influential human observational studies, those with 107 to 1,219 citations, that have been repeatedly cited as evidence for microbiome-autism associations. They found sample sizes ranging from just 7 to 43 individuals per group, far below the thousands required for statistically reliable microbiome research.
“Autism is not rare,” says study co-author Darren Dahly, a biostatistician at University College Cork. “There’s no reason to be having studies with only 20, 30, or 40 participants.”
Recent large-scale microbiome studies have demonstrated that detecting even substantial effects requires sample sizes of several hundred to thousands of individuals. The cited autism-microbiome studies fall dramatically short of these requirements, making their findings likely statistical noise rather than genuine signals.
More critically, when studies did report differences between autistic and non-autistic individuals, these findings were contradictory.
Some studies found lower microbial diversity in autistic individuals while others found the opposite. Better designed research studies comparing autistic children with neurotypical siblings generally found no significant differences.
Animal experiments using mice have been similarly unconvincing. The seminal studies most frequently cited as demonstrating causal links suffer from inadequate sample sizes, inappropriate statistical methods, and questionable relevance to human autism. The researchers highlight that there is no validated evidence that “autistic-like behaviors” in mice, such as marble burying or altered ultrasonic vocalizations, have any meaningful connection to autism in humans.
The study points out that researchers in this space take the behaviors of mice, and compare them to observed human autistic behaviors. However, there is one key problem: mice aren’t people.
The study argues that mice observation tests, where mice seem to have “autistic-like behaviors,” are purely based on “the superficial similarity between simple, isolated tasks in rodents and complex, context-dependent human behaviors.”
A 2018 workshop on autism animal models concluded that many behavioral assays “are rudimentary, do not engage similar neural circuitry as in humans, and lack translational face validity”.
The authors point to one highly cited 2019 study claiming that transplanting gut bacteria from autistic individuals into mice produced autistic behaviors. However, when the data was reanalyzed, it failed to hold up its previous conclusions.
Scientists found several critical errors. First, they treated each mouse as an independent sample even though multiple mice received bacteria from the same human donor, artificially inflating their sample size. Secondly, they used only eight autistic and five control donors, but analyzed data from just five and three, respectively, without scientific justification.
Third, they selected which animals and behavioral tests to focus on after seeing results rather than testing pre-specified hypotheses. Lastly, they used flawed statistical models that didn’t account for their experimental design. Though the flawed study is still highly cited by proponents of the “autism gut-health” theory, the study’s claims about gut bacteria causing autism-like behaviors were largely the result of improper analysis rather than real biological effects.
Human clinical trials testing fecal microbiota transplants (FMT) and probiotic therapies have similarly failed to demonstrate efficacy. A 2024 randomized controlled trial involving 103 children found no differences between those receiving FMT and placebo groups across three separate clinical scales. Meta-analyses of probiotic trials have reached contradictory conclusions, with the two largest reviews, including 12 randomized controlled trials with 630 participants, finding no evidence of benefit.
All meta-analyses noted methodological limitations, including small samples, high risk of bias, subjective parent-reported outcomes, and inconsistent measurements across 115 different outcome measures in just 12 trials.
So why is this theory still around?
Turns out, it’s big business. Popular media narratives, particularly Netflix’s 2024 documentary “Hack Your Health: The Secrets of Your Gut,” casually linked autism to gut problems. While the National Autistic Society condemned the documentary as “deeply irresponsible” and “offensive” for describing autism as a “disease” and suggesting it could be treated through gut interventions, it still sold countless people on the idea.
The microbiome-autism hypothesis has spawned a largely unregulated wellness industry offering specialized diets, probiotic supplements, direct-to-consumer microbiome profiling, and fecal transplant services. Many researchers in this field have competing commercial interests, including patents, spin-out companies, and industry funding.
“This field is characterized by misleading claims about the strength and reliability of these associations and the promise of various kinds of therapeutic approaches,” the study authors write. One researcher noted that “vigorous industry interest” likely creates publication bias, with negative results remaining unreported.
And in a world where health has become more about politics and money than actual science and treating people, the remaining unknown factors surrounding autism spectrum disorder make it ripe for snake oil salesmen and politicians.
Given the lack of convincing evidence and absence of progress despite substantial investment, the researchers argue that the microbiome-autism hypothesis has reached a dead end.
Dahly concludes that “the consensus across the studies that we surveyed is that when you do the trials properly, you don’t see anything.”.
For now, the evidence is clear: autism is a strongly genetic neurodevelopmental condition, not a gut disorder.
MJ Banias covers space, security, and technology with The Debrief. You can email him at mj@thedebrief.org or follow him on Twitter @mjbanias.