By now you have been offered a diet, a powder, a protocol, or a clinic that "treats the root cause". Some of it came from people who love your child. This page uses the same format the site uses for screens: what we can say, what we cannot, and what people claim anyway.
Why this page exists
Autism is a lifelong neurodevelopmental difference. Anything that promises to reverse it is promising something no trial has ever shown. That does not mean nutrition is irrelevant to your child; it means the honest questions are smaller and more useful than the ones being sold.
The honest-claims table
| Intervention | What we can say | What we can't say | What people claim anyway |
|---|---|---|---|
| Gluten-free, casein-free diet | Evidence for benefit in autism is poor. Safe only if planned so the child still gets calcium, protein and calories. | That it changes autism features. | "Leaky gut" causes autism; the diet reverses it. |
| Omega-3 supplements | No high-quality evidence of benefit for core or associated features. Generally safe. | That it improves attention, language or behaviour in autism. | "Brain food" that sharpens focus. |
| Vitamin D | Reviews of trials found no effect on most core features; one 2023 review reported a possible effect on repetitive behaviour only. Low vitamin D is common and worth treating on its own. | That it treats autism. | High-dose D as a protocol. |
| Vitamin B6 with magnesium | Studies too few and too small; no recommendation can be made. High-dose B6 has known nerve side effects. | That it helps. | A decades-old "megavitamin" claim that never died. |
| Probiotics | Trials are small, results inconsistent, evidence rated low. May help constipation or stomach discomfort in some children. | That they improve core autism features. | "Heal the gut, heal the brain." |
| Chelation | No evidence of benefit. Reports of serious harm including death. | Anything positive. | "Detoxes heavy metals that cause autism." |
| "Biomedical" protocols, MMS / chlorine dioxide | Dangerous. Health Canada: can cause poisoning and kidney failure; sale not permitted in Canada. | Anything positive. | "Kills the parasites that cause autism." |
The evidence tags below the table apply row by row.
Gluten-free, casein-free
The Cochrane review of gluten- and casein-free diets for autism concluded that "current evidence for efficacy of these diets is poor" and called for large, good-quality trials. strong The review is old and was later marked as withdrawn, which is Cochrane's way of saying it is out of date, not that the conclusion was reversed; no later high-quality trial has overturned it.
What is true: some autistic children have coeliac disease, milk protein allergy or lactose intolerance, like some children generally. Those are diagnosed by a doctor and treated by diet. And many autistic children are selective eaters; cutting two food groups from an already short list can make nutrition worse. If you try the diet anyway, do it with a dietitian, for a set period, with one clear thing you are watching.
Omega-3
The Cochrane review found no high-quality evidence that omega-3 supplementation improves core or associated features of autism. strong Fish is still good food. A capsule is not a treatment.
Vitamin D
A 2023 systematic review of randomised trials concluded that vitamin D supplementation "appears to improve stereotypical behaviors but does not improve other core symptoms and coexisting conditions". moderate Read that as: one small signal, no broad effect, and trials that differ a lot from each other. Separately, low vitamin D is common in Quebec winters in any child; if your doctor finds it, treat it because it is a deficiency, not because it is autism.
Vitamin B6 and magnesium
The Cochrane review found the studies so few, small and weak that "no recommendation can be advanced" for B6-magnesium as an autism treatment. strong High doses of B6 over time can damage nerves, which is a reason not to experiment.
Probiotics
A 2025 meta-analysis of eight placebo-controlled trials found a small improvement in overall behaviour ratings, mainly in preschoolers and longer courses, but rated the evidence low and noted the result disappeared if any one of three trials was removed. emerging Other reviews found no effect on core features. If your child is constipated or in stomach pain, that deserves a proper look from a doctor; gut discomfort makes any child harder to live with, and sorting it out is worthwhile in its own right. That is a different claim from "probiotics treat autism".
Chelation and "biomedical" protocols
Chelation removes metals from the body and is a legitimate treatment for real heavy-metal poisoning. Used for autism, the Cochrane review found "no clinical trial evidence" of benefit and, citing serious adverse events including low blood calcium, kidney damage and a reported death, concluded the risks "currently outweigh proven benefits". strong
"Miracle Mineral Solution", sold online as an autism treatment, is sodium chlorite that becomes a bleach-like chemical when mixed. Health Canada states that ingesting it "can cause poisoning, kidney failure, harm to red blood cells, abdominal pain, nausea, vomiting, and diarrhea", that it has approved no such product for human consumption, and that its sale is not permitted under the Food and Drugs Act. official People have been prosecuted in Canada for selling it. If anyone offers it, or an enema version of it, for your child, that is a report-to-Health-Canada situation, not a difference of opinion.
The wider "biomedical" world (hyperbaric chambers, antifungals, "detox" baths, stem-cell tourism) shares the same pattern: a root-cause story, a clinic outside the public system, a price, and no trial that survived scrutiny.
Where the disagreement is
Researchers do take the gut-brain connection seriously; there are real studies linking gut chemistry and brain activity in autistic children. The disagreement is about what follows. Scientists read those as early, correlational leads that may matter for a subgroup one day. Practitioners selling protocols read them as licence to treat now. The first reading is honest about what a small cross-sectional study can show. The second sends you a bill.
What is worth doing instead
- Ask your doctor for a check of iron, vitamin D and growth if your child eats fewer than about 20 foods. Treat what is found.
- Address constipation and reflux; they are common and they change behaviour.
- Treat selective eating as a sensory and anxiety issue, not a discipline one. An occupational therapist or dietitian can help; see Sensory systems and the regulation ladder.
- Protect sleep. It moves more than any supplement: see Sleep.
This week
- Write down every supplement or diet rule your child is currently on and what each is supposed to do. Bring the list to your next doctor visit.
- If one of them is chelation, MMS or a "detox" protocol, stop it and talk to your doctor or a pharmacist this week.
- Pick one real nutrition question (constipation, iron, food range) and book the person who can answer it.
Sources
- Gluten- and casein-free diets for autistic spectrum disorder (Cochrane review) — Millward et al., Cochrane Database of Systematic Reviews, 2008
- Omega-3 fatty acids supplementation for autism spectrum disorders (Cochrane review) — James et al., Cochrane Database of Systematic Reviews, 2011
- Effects of vitamin D supplementation on children with autism spectrum disorder: a systematic review and meta-analysis — Zhang et al., Clinical Psychopharmacology and Neuroscience, 2023
- Combined vitamin B6-magnesium treatment in autism spectrum disorder (Cochrane review) — Nye and Brice, Cochrane Database of Systematic Reviews, 2005
- A meta-analysis of randomized placebo-controlled trials on the effects of probiotics for autism spectrum disorders — Huang et al., Clinical Psychopharmacology and Neuroscience, 2025
- Chelation for autism spectrum disorder (Cochrane review) — James et al., Cochrane Database of Systematic Reviews, 2015
- Health Canada reminds Canadians of dangers of Miracle Mineral Solution after investigation leads to guilty plea and two-year sentence — Health Canada, 2018