Outside C. difficile, stool transplant's strongest evidence is 12 small colitis trials

Faecal microbiota transplantation has one established indication. Everywhere else it is a promising intervention with small trials, wide confidence intervals and unresolved questions about how to deliver it. The largest synthesis in ulcerative colitis — the leading candidate for a second indication — rests on twelve randomised trials, and its most striking finding is that the route of administration changes the result more than almost anything else does [s1].

Ulcerative colitis: route decides everything

Researchers performed a systematic review and network meta-analysis of randomised trials including at least one FMT arm in adults with active ulcerative colitis, searching MEDLINE, Embase and Cochrane CENTRAL [s1]. The primary endpoint was clinical remission — a total Mayo score of 2 or less with a Mayo endoscopic score of 1 or less — assessed between weeks 6 and 12 [s1]. Results were adjusted for bowel cleansing and pre-FMT antibiotics [s1].

Twelve studies were selected. In four of them participants were exclusively biologic-naive; in the others, between 9% and 32% had prior biologics exposure [s1]. Risk of bias was low across all domains in seven of the twelve [s1].

The delivery routes separated sharply. Upper gastrointestinal FMT was not distinguishable from placebo, with a relative risk of 1.1 (95% CI 0.2 to 7.7) [s1]. Oral capsules performed at a relative risk of 7.1 (1.8 to 33.3), lower gastrointestinal FMT at 4.5 (1.7 to 12.5), and the combination of both at 12.5 (2.1 to 100) [s1]. The combination was significantly better than upper gastrointestinal delivery, at a relative risk of 10.7 (1.1 to 104.2) [s1], and ranked highest overall with a SUCRA score of 0.93 [s1].

Look at the confidence intervals rather than the point estimates. An interval from 2.1 to 100 is not a precise result; it is a strong signal measured in very few patients. This is a body of evidence pointing consistently in one direction with almost no precision about magnitude.

Two negative findings are as informative as the positive ones. Multidonor FMT did not perform better than single-donor FMT [s1]. And autologous FMT — where patients receive their own stool back, the design intended as an inert placebo — ranked below placebo, at a SUCRA of 0.12 against 0.22, which the authors read as a potential detrimental effect and a reason to avoid it in trials [s1].

That is a methodological finding with consequences beyond this analysis: if the standard sham in FMT trials is not inert, the effect sizes in trials that used it are not what they appear.

Irritable bowel syndrome: a colder answer

A meta-analysis published in Gastroenterology in May 2026 updated an earlier synthesis, searching CENTRAL, MEDLINE and Embase to 20 February 2026 and including randomised trials reporting the proportion of patients with IBS symptom improvement between 4 and 24 weeks after FMT [s2]. Thirteen trials involving 693 patients were eligible [s2].

On intention-to-treat analysis — the conservative approach, counting everyone as randomised — FMT may reduce IBS symptoms at 12 weeks compared with placebo, but the authors state the evidence is very uncertain, with a relative risk of symptoms not improving of 0.72 (95% CI 0.50 to 1.03) [s2]. That interval crosses 1.

On per-protocol analysis, counting only those who completed treatment as planned, the result becomes statistically significant at 0.67 (95% CI 0.46 to 0.97) [s2]. Subgroups also reached significance: single-dose FMT at 0.62 (0.41 to 0.93), and trials using Rome IV criteria to diagnose IBS at 0.38 (0.17 to 0.86) [s2]. Adverse events did not differ between groups (RR 0.98, 95% CI 0.75 to 1.29) [s2].

The authors' own framing is the appropriate one: very low certainty that FMT improved IBS symptoms, effective in some subgroup analyses and in the overall per-protocol analysis, but needing cautious interpretation [s2]. A result that appears only when non-completers are excluded is a weaker result than one that survives intention-to-treat, because dropping out is not random.

Why the trials disagree with each other

An international expert panel convened for the first Rome consensus conference on gut microbiota and FMT in inflammatory bowel disease identified the reason directly. Several randomised trials of FMT for IBD, particularly ulcerative colitis, had recently been published with major variations in study design: differences in administered dose, route and frequency of delivery, type of placebo, and evaluated endpoints [s3].

The panel described outcomes as appearing promising overall, but highly dependent on both donor and recipient factors [s3]. Twenty-five experts in IBD, immunology and microbiology worked through donor selection and biobanking, FMT practices, and future study design, generating statements by an electronic Delphi process and a plenary consensus conference [s3].

Their stated goal was to move towards standardised practice and provide the general criteria required to promote FMT as a recognised strategy for treating IBD [s3] — which is a way of saying it is not yet one.

What this adds up to

For ulcerative colitis, the direction of effect is consistent across twelve trials and the combination of lower gastrointestinal delivery with oral capsules ranks best [s1]. For IBS, the evidence is very low certainty and depends on which analysis you privilege [s2]. In both, trials are small, protocols vary in nearly every parameter, and the field's own consensus body says standardisation has not yet happened [s3].

The thing to resist is the shortcut from "FMT works for recurrent C. difficile" to "FMT works on the microbiome, therefore it works on conditions the microbiome is implicated in". The route of delivery mattering by an order of magnitude in ulcerative colitis [s1], and the standard placebo appearing to be actively harmful [s1], both suggest the field does not yet understand its own intervention well enough for that inference.

This article is informational and is not medical advice. Faecal microbiota transplantation outside approved indications is an investigational procedure carried out in clinical trials, and self-administration carries infection risks that trial protocols exist to control.

Sources

• Efficacy of different modalities of faecal microbiota transplantation in ulcerative colitis: systematic review and network meta-analysis — Therapeutic Advances in Gastroenterology, 2025-08-25
• Fecal Microbiota Transplantation for Symptom Improvement in Patients With Irritable Bowel Syndrome: Systematic Review and Meta-Analysis of Randomized Controlled Trials — Gastroenterology, 2026-05-19
• The first international Rome consensus conference on gut microbiota and faecal microbiota transplantation in inflammatory bowel disease — Gut, 2023-06-20