Gluten-Free and Gluten-Restricted Diets in Multiple Sclerosis: A Systematic Review of Disease Activity and Patient-Reported Outcomes

Background: Dietary modification is widely used by people with multiple sclerosis (MS), and gluten avoidance is a common self-directed strategy. Biological plausibility has been proposed through gut-barrier and immune mechanisms, but whether gluten restriction improves clinical outcomes or objective disease activity remains uncertain. Objective: To systematically identify, appraise, and synthesise evidence on gluten-free and gluten-restricted diets in MS, with particular attention to whether observed effects can be attributed specifically to gluten. Methods: The protocol was registered with PROSPERO (CRD420261459657). MEDLINE (via PubMed), Embase, Scopus, the Web of Science Core Collection, and the Cochrane Library were searched from 1 July from the inception of each database to 22 July 2026, supplemented by backward and forward citation searching. Eligible studies enrolled adults with MS and evaluated a gluten-free diet, a clearly defined gluten-restriction component, or habitual gluten exposure. Screening, extraction, and appraisal were performed independently in duplicate. RoB 2, ROBINS-I, and the Newcastle-Ottawa Scale were applied according to design, and certainty was assessed with GRADE. Clinical and methodological heterogeneity precluded meta-analysis; results were synthesised narratively following SWiM guidance. Results: Of 694 records identified, 392 duplicates were removed and 302 records were screened by title and abstract, of which 292 were excluded; ten reports were sought for full-text retrieval and one could not be obtained; nine full texts were assessed, three were excluded (one non-English publication and two ineligible study designs), and six studies were included. Two parallel-group controlled trials were reported as randomised, one was a randomised crossover trial, two were non-randomised interventional studies, and one was observational. Only six studies were eligible, and they were small and clinically and methodologically heterogeneous; no randomised comparison isolated gluten restriction as the sole dietary change. Clinical outcomes (fatigue, quality of life, disability) are reported separately from biological or disease-activity outcomes (relapse, magnetic resonance imaging activity, and biomarkers). Fatigue and quality of life improved most consistently in multi-component dietary trials that did not isolate gluten. Evidence specifically isolating gluten restriction was much weaker, and these results should not be read as showing that a gluten-free diet itself confers clinical benefit. The two studies reporting improved disability used adherence-based, non-randomised comparisons. By contrast, the randomised crossover trial of marked wheat/gluten reduction found no disability benefit and had a negative primary immunological endpoint. This contrast reflects substantial uncertainty and does not support a causal benefit of gluten elimination on disability. Adherence was a recurrent limitation. GRADE certainty was very low for all seven outcome domains, including fatigue and quality of life, and most of the evidence was indirect with respect to gluten elimination. Conclusions: In six small and heterogeneous studies, fatigue and quality of life improved mainly with multi-component dietary interventions; current evidence does not establish that gluten elimination itself confers clinical benefit, is responsible for these improvements, or modifies MS disease activity. Adequately powered randomised trials that isolate gluten elimination, objectively verify adherence, and use clinically meaningful outcomes with longer follow-up are required.