Can Diet Treat IBD? What the Best Trials Actually Found

Restrictive diets for inflammatory bowel disease often arrive with a promise: remove the wrong foods, calm the inflammation, avoid stronger treatment. The controlled evidence is less obliging. The best individual-level trial found no clinically meaningful difference between two popular elimination diets in children with IBD, while the dietary approach with the strongest pediatric Crohn's data gets roughly half its calories from formula.

Diet can matter to people with IBD. It can affect symptoms, food tolerance, nutrition, and daily life. But treating symptoms is not the same as healing intestinal inflammation, and neither is the same as replacing proven medical care.

The most personalized trial did not find a personalized winner

Kaplan et al.'s 2022 PRESTO trial put the personalization idea to a demanding test. Across 19 sites, 54 patients aged 7 to 18 with Crohn's disease or ulcerative colitis completed personalized N-of-1 crossover trials of the Specific Carbohydrate Diet (SCD), a modified SCD (MSCD), and usual diet. Each participant acted as their own control.

The result did not show a reliable advantage for SCD over MSCD. For fecal calprotectin, a stool marker associated with intestinal inflammation, the geometric mean ratio for SCD versus MSCD was 0.77, with a 95% credible interval of 0.51 to 1.10. That interval includes no difference. The average symptom difference between diets was −0.3 points (95% CI, −1.2 to 0.75), and the investigators calculated less than a 1% posterior probability of a clinically meaningful difference between the diets.

Neither diet consistently outperformed usual diet. That finding matters because an N-of-1 design is supposed to reveal meaningful individual differences that disappear in group averages. Here, it did not produce a dependable way to identify whether SCD or MSCD would work better for a given child.

The trial also exposed the practical cost of restrictive eating. Twenty-four of 54 participants, or 44%, withdrew before completing the protocol. Reported reasons included lack of response, adverse events, and inability to continue the diet. A diet that works only for the people able to sustain it will look better in completer analyses than it does in a clinic.

Adult Crohn's disease produced the same basic answer

The pediatric study compared two restrictive diets. An adult randomized trial asked whether SCD could beat a less restrictive alternative. In Lewis et al.'s 2022 trial, 194 adults with mild-to-moderate Crohn's disease were randomized to SCD or a Mediterranean diet.

At weeks 6 and 26, the diets did not differ statistically in symptomatic remission or endoscopic response. Neither showed a consistent mucosal-healing advantage over the other. Endoscopic response carries more weight than symptom relief alone because inflammation can continue even when a person feels better.

This does not prove that food choices are irrelevant. It means SCD did not show consistent superiority to a Mediterranean diet in this population, over these follow-up periods, on these outcomes. That is a much narrower and more useful conclusion than declaring one diet a treatment for Crohn's disease.

The broader literature has more symptoms than scopes

A 2025 systematic review by Gleave et al. examined 11 controlled studies covering 10 dietary interventions for IBD. Its conclusion was limited evidence for specific diets, especially in adults with mildly to moderately active disease. The review also called for trials that use endoscopic outcomes rather than relying mainly on symptoms.

No diet in that review achieved majority endoscopic remission in a controlled trial. This is the central limitation of the field. Symptoms are important patient outcomes, but they do not establish mucosal healing. A diet that reduces abdominal pain or changes bowel habits may still leave active inflammation untreated.

A 2024 umbrella review by Christensen et al. reached a compatible population-level view: dietary and nutritional exposures relate to IBD risk and progression, but newer evidence has not established consistent objective remission with a dietary intervention for active IBD.

CDED plus formula has the strongest dietary evidence—and important strings attached

The Crohn's Disease Exclusion Diet (CDED) combined with partial enteral nutrition (PEN) is different from SCD and MSCD. CDED limits foods thought to aggravate gut-barrier disruption, including processed foods, emulsifiers, animal fats, and gluten-containing grains. PEN supplies about 50% of calories as formula, which helps maintain nutritional intake.

That combination has the strongest objective evidence among whole-food-oriented dietary approaches, but mainly in pediatric Crohn's disease. A 2024 systematic review by Correia et al. reported about 75.6% clinical remission at week 6 and 75% sustained remission at week 12 with CDED plus PEN in pediatric Crohn's disease, alongside biomarker improvements. A second 2024 systematic review similarly identified CDED plus PEN as the solid-food diet with the strongest evidence for induction and maintenance of remission.

Those results deserve more attention than SCD marketing receives. They also need accurate framing. CDED plus PEN is not a food-only protocol; formula provides roughly half of caloric intake. Its controlled evidence base is limited, concentrated in children, and has not been replicated in adult randomized trials. It also could not fit into PRESTO's crossover comparison of whole-food diets in isolation, so PRESTO cannot answer whether CDED plus PEN works.

For pediatric Crohn's disease, the reviewed ECCO/ESPGHAN guidance recommends exclusive enteral nutrition (EEN), complete diet replacement with formula, as first-line dietary induction therapy. It does not recommend SCD, autoimmune protocol diets, or comparable elimination diets as standard care.

Biomarkers can track response, but cannot pick responders

A 2023 study by Verburgt et al. examined 26 pediatric Crohn's disease patients who responded to dietary therapy. Responders showed fecal calprotectin normalization, enrichment of Faecalibacterium prausnitzii, and inflammatory metabolite patterns that shifted toward those of healthy controls.

These findings are biologically plausible response markers, not a pre-treatment test. They appeared among responders to both EEN and CDED-based treatment, so they cannot identify which diet will work before treatment starts or prove that a particular food restriction caused the microbiome change. No validated biomarker, microbiome signature, or clinical feature currently selects dietary responders in advance.

What the evidence supports

For adults with IBD, no dietary protocol has shown consistent objective remission in controlled trials. SCD has not beaten either MSCD in pediatric N-of-1 testing or a Mediterranean diet in an adult Crohn's trial. Restrictive diets also impose adherence, nutritional, and social burdens that trials often measure poorly; PRESTO did not provide detailed categories for adverse-event withdrawals, growth outcomes, or nutritional deficiencies.

For children with Crohn's disease, CDED plus PEN has the best evidence among the dietary approaches discussed here, but it should be treated as medical nutrition therapy rather than a DIY elimination diet. It requires formula, evidence-based clinical supervision, and realistic attention to whether a child can sustain it.

Diet may help manage IBD, but current trials do not support presenting SCD, MSCD, or other whole-food elimination diets as reliable substitutes for standard treatment.

Finally, no trial in this evidence base directly compares dietary therapy with biologics or immunomodulators. We therefore do not know the absolute effect of diet against the treatments patients would otherwise receive. Anyone considering a restrictive diet for IBD should involve their gastroenterology and nutrition team, monitor disease activity with appropriate objective measures, and avoid mistaking a quieter gut for healed inflammation.