Because the ketogenic diet has a long track record in epilepsy, and because epilepsy and autism share some biological threads, researchers have asked whether keto might help with some features of autism too. It is a reasonable question, and there is some early research on it. It is also an area where it is easy to find wildly overstated claims online, so this is a deliberately measured look at what the studies actually show.
What has been studied
The research base is small and mostly in children. It consists of pilot studies, a handful of small trials, and review papers that gather them up, including a 2021 scoping review in the journal Nutrients and a 2024 narrative review. The diets tested are usually modified or medium-chain-triglyceride versions of keto, chosen to be more tolerable, sometimes combined with other dietary changes. These are not large, definitive trials; they are the early, exploratory kind of research that tells you whether something is worth investigating properly.
What the early studies suggest
Within those limits, the reported signals have been broadly positive. Several studies found improvements on autism rating scales such as the CARS, with changes described across areas like social behaviour, communication, repetitive behaviours and sometimes hyperactivity and anxiety. A recurring observation is that the people who responded best tended to have milder or moderate presentations to begin with, while benefits were less apparent for those more profoundly affected. So the early picture is of a possible, modest support for some autistic traits in some people, not a transformation and not something that applies evenly.
The proposed mechanisms
Researchers offer several reasons why ketosis might matter here, none of them fully settled. Ketones may steady the brain’s energy supply, much as in epilepsy. The diet appears to influence inflammation and oxidative stress, both of which have been linked to autism in some studies. It also shifts the gut microbiome, and the gut-brain connection is an active area of autism research. These are plausible, mechanistically interesting threads, but they are explanations being explored, not proven pathways.
Why this is firmly preliminary
The honesty this topic demands is about the weakness of the evidence. The studies are few and small, often without strong control groups, the diets and measures vary from study to study, and keeping children on a restrictive diet is genuinely difficult, which affects both the results and their real-world usefulness. Most of the work is in children, so there is very little to say about autistic adults specifically. None of this means the idea is wrong; it means it is unproven, and that larger, better-designed trials are needed before anyone can speak with confidence.
What it means in practice
This is the part to take seriously. Autism is not an illness to be cured, and a diet should never be framed as fixing a person; at most the research is about easing certain symptoms or co-occurring difficulties for some individuals. Restricting the diet of a child, especially an autistic child who may already have a narrow range of accepted foods or particular nutritional needs, is not something to attempt off the back of a hopeful headline. If a family wants to explore it, it should be with a paediatrician and a dietitian involved from the start, monitoring growth and nutrition. For autistic adults curious about keto, the same sensible, supported approach as anyone else applies, with no special claims attached.
The bottom line
There is early, small-scale research suggesting a ketogenic diet may modestly help some features of autism in some people, particularly milder presentations, with several plausible biological mechanisms behind it. The evidence is preliminary and concentrated in children, so it sits firmly in the “worth studying, not yet proven” category, and any real-world trial of it, especially for a child, belongs with medical and dietetic support rather than alone.
This is general information, not medical advice. Autism is a neurodevelopmental difference, not a disease to be treated with diet. Do not put a child on a restrictive diet without a paediatrician and dietitian involved, and discuss any significant dietary change with a relevant professional.
Sources: Dietary therapies in autism spectrum disorder have been reviewed in: Ketogenic Dietary Therapies in Patients with Autism Spectrum Disorder: Facts or Fads? A Scoping Review. Nutrients. 2021;13(6):2057. Read it here. See also the 2024 narrative review in PubMed.