Gluten and autism, ADHD or arthritis: what the research shows

Short answer

Reviews and trials find limited or mixed evidence that a gluten-free diet treats autism, ADHD or arthritis12, 22, 34. By condition:

  1. Autism: reviews of small trials disagree, with two reporting gains on some measures and two finding none or little, and England’s NICE tells clinicians not to use gluten- or casein-free diets for core features of autism in under-19s4, 7, 14, 12, 20.
  2. ADHD: a 2026 review found no eligible studies of a gluten-free diet in children and adolescents with ADHD and no diagnosed celiac disease23, 21.
  3. Rheumatoid arthritis: one 2001 trial of a vegan diet free of gluten concluded dietary modification may be of clinical benefit for certain patients, and a 2022 review of patients without celiac disease or non-celiac gluten sensitivity found no scientific evidence to recommend excluding gluten33, 34.
  4. Celiac disease is a separate claim: it is associated with higher rates of autism, ADHD and arthritis in people who have it37, 44.

Published by Gluten Verdict · updated · 21 sources, each linked below

Reviewed by Mark Peter, Registered dietitian (RD), on

Person at a kitchen table seen from behind with a laptop and a cereal box

Key points

  • A double-blind, placebo-controlled challenge in 14 children with autism found no statistically significant effect, and its authors say the study does not provide evidence to support general use of the diet16, 17.
  • Reviews of small autism trials reach different readings, and one of the more favourable ones says most included studies were single-blind5, 12.
  • For ADHD, the 2026 review found no eligible studies; a 2024 review of diet and supplement trials reported that gluten-free diets may alleviate symptoms but cautioned against broad generalizations21, 24, 25, 26.
  • The NHS says there is no strong evidence that specific dietary changes improve rheumatoid arthritis35.
  • A link between celiac disease and these conditions is about people with celiac disease; by itself it does not show that the diet treats the conditions in people without it, and the 2020 meta-analysis authors call for more research on any effect of the gluten-free diet37, 38.

Why these claims exist

The idea for autism is that a gluten-free and casein-free (GFCF) diet may ease symptoms and support neurodevelopment, and a 2022 meta-analysis said studies to date had been inconclusive1. A 2026 review adds that families increasingly use gluten-free diets for children’s psychiatric conditions despite limited evidence, and names proposed mechanisms including immune-mediated inflammation, altered gut-brain signaling and increased intestinal permeability2.

A 2022 review of gluten and rheumatoid arthritis describes a growing anti-gluten trend, attributed largely to information distributed through social networks, that, outside celiac disease and non-celiac gluten sensitivity, results in gluten being considered harmful and excluded from diets3.

The three claims have different evidence behind them, so each section below stands alone. Our verdict pages cover whether specific foods contain gluten; this page covers only the health claims.

Autism: reviews of small trials disagree

The table lists five reviews, and a double-blind trial follows it. Their study counts differ, so the totals are not additive.

Reviews of gluten-free and casein-free diets in autism, as their authors report them
ReviewWhat it includedWhat it reported
Quan and colleagues, Nutrition Reviews, 20228 studies, 297 participants45 studies reported significant reductions in stereotypical behaviors and 3 reported improvements in cognition; no significant change in other symptom categories. The authors call the benefits promising, note most studies were single-blind, and ask for larger studies4, 5.
Yu and colleagues, Frontiers in Neurology, 20227 randomized trials, 338 participants, covering gluten-free and/or casein-free diets and the ketogenic diet6, 8The authors conclude diet therapies can ameliorate core symptoms and gluten-free diets are conducive to improving social behaviors; the same abstract reports no statistically significant change in the GFCF and gluten-free subgroups, and asks for larger, higher-quality trials7, 9, 10.
Piwowarczyk and colleagues, European Journal of Nutrition, 20186 randomized trials, 214 participants11With few exceptions, no statistically significant differences in core symptoms on standardized scales; the authors conclude there is little evidence that the diet is beneficial11, 12.
Keller and colleagues, Nutrients, 20216 randomized trials, 143 participants14No effect on clinician-reported core symptoms, parent-reported functional level or behavioral difficulties; quality of evidence low to very low14, 15.
Stokan and colleagues, Cureus, 20263 autism studies, children and adolescents without diagnosed celiac disease18, 23The authors call the evidence across the three studies mixed: two randomized trials found no significant improvement in core symptoms, and one of them reported fewer gastrointestinal symptoms and modest behavioral gains19.

A separate double-blind, placebo-controlled trial placed 14 children aged 3 to 5 on the diet, then challenged them with weekly snacks containing gluten, casein, both or placebo for 12 weeks16. Dietary challenges had no statistically significant effect on measures of physiologic functioning, behavior problems or autism symptoms; the authors say the findings must be interpreted with caution because of the small sample, and that the study does not provide evidence to support general use of the diet17.

Who reports an outcome matters: in the 2018 review, the remaining differences, if present, referred to parent-based assessment tools or other developmental or ASD-related features13.

Guidance, not a trial: England’s NICE guideline for autism in under-19s tells clinicians not to use exclusion diets, such as gluten- or casein-free diets, to manage the core features of autism. The page was last updated 14 June 202120.

ADHD: little trial evidence for gluten itself

A 2026 review set out to evaluate gluten-free diets in children and adolescents with psychiatric disorders and no diagnosed celiac disease23. It found no eligible studies for ADHD21; its search covered English-language studies published from January 2000 to August 2024 and required a gluten-free diet of at least six weeks53. It concluded that evidence for psychiatric disorders overall is sparse and that there is insufficient evidence to support routine use for core symptoms22.

A 2024 review took a wider question: 14 randomized controlled trials met its inclusion criteria, and it investigated dietary interventions and supplements alongside standard pharmacotherapy in ADHD24, 52. Its abstract says the findings suggest gluten-free diets may alleviate ADHD symptoms despite negative celiac serology, and also lists limitations of study diversity and short-term outcome assessment that caution against broad generalizations25, 26.

The two reviews asked different questions: the 2026 review looked only at gluten-free diets in people without diagnosed celiac disease23, while the 2024 review covered several diets and supplements used alongside medication24. England’s NICE ADHD guideline says there is no evidence about the long-term effectiveness or potential harms of a “few food” diet for children with ADHD and only limited evidence of short-term benefits27; its dietary-advice section does not mention gluten29.

Arthritis: one trial, one review and the NHS

One trial, published in 2001, randomized 66 patients with active rheumatoid arthritis to a vegan diet free of gluten or a well-balanced non-vegan diet for one year30. Among those who completed at least nine months, 40.5 percent (nine patients) on the vegan gluten-free diet met the ACR20 improvement criteria, against 4 percent (one patient) on the other diet; the corresponding intention-to-treat figures were 34.3 and 3.8 percent31.

Two points limit what that trial says about gluten. The diet changed more than gluten: it was vegan as well as free of gluten, and the comparison diet was non-vegan30. And no slowing of radiological damage was apparent in either group32. The authors concluded that dietary modification may be of clinical benefit for certain patients33.

A 2022 literature review set out to assess whether there is scientific evidence to justify eliminating gluten in patients with rheumatoid arthritis who have neither celiac disease nor non-celiac gluten sensitivity. It reports searching the Embase database and extracting data from 16 articles, and found no scientific evidence to recommend excluding gluten34.

The NHS says there is no strong evidence that specific dietary changes improve rheumatoid arthritis, although some people feel symptoms worsen after certain foods35.

Studies of people who have celiac disease do find links. A 2020 meta-analysis of 37 articles reported a significant increase in risk of autism (odds ratio 1.53) and ADHD (odds ratio 1.39) among people with celiac disease compared with controls; its authors say more research is needed on biological explanations and on any effect of the gluten-free diet37, 38.

For arthritis, a cohort in Sweden of 24,014 people with biopsy-proven celiac disease and 117,397 matched comparators found juvenile idiopathic arthritis in children occurred at 5.9 per 10,000 person-years against 2.2 (hazard ratio 2.68), and rheumatoid arthritis in adults at 8.4 against 5.1 (hazard ratio 1.70)42, 43, 44. The study notes that a definitive association had not been established before it41.

NIDDK lists immune-related disorders that some people have alongside celiac disease; its list includes rheumatic diseases such as Sjögren’s syndrome, and does not name autism, ADHD or rheumatoid arthritis45.

Results for people who already have celiac disease belong to that group. A 2011 report found 10 of 67 people with ADHD were positive for celiac disease, and that after starting a gluten-free diet the patients or their parents reported improved behavior compared with the period before diagnosis and treatment39. The Gluten Intolerance Group summarizes a study of children with celiac disease in which ADHD-like issues were associated with not following the diet40. These are findings about celiac disease. The trials and reviews above ask a different question: whether a gluten-free diet helps people with autism, ADHD or arthritis in general.

Where non-celiac gluten sensitivity fits

NIDDK says celiac disease is different from gluten sensitivity and that, unlike celiac disease, gluten sensitivity does not damage the small intestine46.

Coeliac UK describes non-coeliac gluten sensitivity as symptoms similar to coeliac disease with no antibodies produced and no apparent gut damage, and notes debate over whether gluten or other wheat components, such as FODMAPs, are the cause47, 48. A 2015 BMJ review says its pathophysiology is poorly understood but a cause of increasing interest in gluten-free diets in the general population49. The 2024 ADHD review above links its finding to this category25.

What the studies’ own authors flag

Several of the reviews and trials above point out weak spots in their own work. When you meet a new claim about gluten, these are the questions they raise:

  • How many took part? The double-blind trial had 14 children and its authors say the findings must be interpreted with caution because of the small sample16, 17.
  • Were the studies blinded? Most of the studies in the 2022 Nutrition Reviews review of GFCF diets were single-blind5.
  • Who reported the result? The 2018 review says its remaining differences, if present, referred to parent-based assessment tools13, and the celiac ADHD report relied on patients or parents reporting improvement39.
  • How good was the evidence overall? One review rated it low to very low, citing serious risk of bias, inconsistency and imprecision15.
  • How long did it run? The 2024 ADHD review lists short-term outcome assessment as a limitation26.
  • What else changed besides gluten? The arthritis trial compared a vegan diet free of gluten with a non-vegan diet30.

What agencies and charities say before a diet changes

NIDDK says doctors do not recommend starting a gluten-free diet before diagnostic testing for celiac disease, because the diet can affect test results51. Coeliac UK likewise does not recommend trying a gluten-free diet as a first option when there are symptoms, for the same reason50. NIAMS advises talking to a doctor before changing diet in rheumatoid arthritis36.

For ADHD, NICE’s guideline describes a route when a food seems to affect behaviour: where there is a clear link, keep a diary of food and ADHD behaviour, and if the diary supports a relationship, offer referral to a dietitian28.

Questions people ask

Does a gluten-free diet treat autism?

Reviews disagree. Two reported gains on some measures4, 7, while others found no effect on core symptoms or little evidence of benefit14, 12. NICE tells clinicians not to use gluten- or casein-free diets for the core features of autism in under-19s20.

Does a gluten-free diet help ADHD?

A 2026 review found no eligible studies for ADHD in children and adolescents without diagnosed celiac disease21, 23. A 2024 review of diet trials suggested gluten-free diets may alleviate symptoms, with limitations that caution against broad generalizations25, 26.

Does gluten cause arthritis or joint inflammation?

The sources here do not report that gluten causes rheumatoid arthritis. A 2022 review found no scientific evidence to recommend excluding gluten34, and the NHS says there is no strong evidence that specific dietary changes improve it35. People with celiac disease had higher rates of rheumatoid arthritis in one cohort in Sweden44.

Does celiac disease cause autism or ADHD?

A meta-analysis found celiac disease is associated with increased risk of both, and says more research is needed on biological explanations37, 38. An association is not a finding of cause.

More in Gluten basics and myths

Sources