The best-evidenced gut-brain treatment is not a supplement. It is hypnotherapy

"Gut-brain axis" in consumer health almost always means a supplement: a strain of bacteria sold on the premise that changing the gut changes the mind. The intervention with the strongest randomised human evidence for acting on that axis runs the other way — a psychological therapy directed at the gut. A systematic review and meta-analysis of gut-directed hypnotherapy in irritable bowel syndrome covering twelve studies in 1,158 patients found all twelve superior to their comparator, nine of them statistically significantly so [s1].

That is not a small literature by the standards of this field, and it points at a mechanism running from brain to gut rather than the reverse.

What the hypnotherapy meta-analysis found

Researchers searched PubMed, Embase and Web of Science, screened titles, abstracts and full texts, and assessed bias with the Cochrane Collaboration risk-of-bias tool [s1]. Twelve studies reported in eleven papers met inclusion criteria, involving 1,158 patients with IBS; eight provided continuous measures sufficient for meta-analysis [s1].

The headline pooled result is honest about its own weakness. Gut-directed hypnotherapy improved global IBS symptoms with a standardised mean difference of 0.73, but a confidence interval of −0.09 to 1.55 and an I² of 93% [s1] — an effect estimate that crosses zero, on trials that disagree with each other almost completely.

The subgroup results are where the signal firms up. High-volume delivery produced a statistically significant improvement in global symptoms (SMD 0.56, 95% CI 0.29 to 0.83) with an I² of 0% — complete consistency across those trials [s1]. Group delivery also produced significant improvement (SMD 0.41, 95% CI 0.05 to 0.77, I² 61%) [s1]. And on pain specifically, hypnotherapy significantly outperformed its comparators (SMD 0.25, 95% CI 0.01 to 0.49, I² 17%) [s1].

The authors' conclusion is that gut-directed hypnotherapy may improve global IBS symptoms, that it improved pain compared with other standard IBS interventions, and that group delivery was effective against standard interventions [s1].

Two things about that pattern are worth noting. A subgroup with an I² of 0% inside an overall analysis with an I² of 93% suggests the heterogeneity is coming from how the therapy was delivered rather than from whether it works. And group delivery being effective is a practical finding — group sessions are the version a health service can actually provide.

Where the guidelines put psychological therapy

The American College of Gastroenterology's IBS guideline suggests that gut-directed psychotherapy be used to treat global IBS symptoms [s2]. It sits in the same graded list as a limited trial of a low-FODMAP diet, chloride channel and guanylate cyclase activators for constipation-predominant IBS, and rifaximin for diarrhoea-predominant IBS [s2].

The guideline was developed using GRADE methodology, with 25 clinically important questions assessed after a comprehensive literature search — nine on diagnostic testing and sixteen on therapy — and consensus reached by a modified Delphi approach [s2]. Gut-directed psychotherapy earned its place on that list through the same process as everything else on it.

The other direction of travel, and why it is weaker

The popular version of the gut-brain axis is bacteria acting on mood. That claim's human evidence base is thinner than its coverage suggests, and the reason is structural.

Most causal claims about the gut microbiome rest on transferring human gut bacteria into germ-free rodents and observing whether a disease phenotype travels with it. A systematic review published in Cell found that 95% of published studies using this design — 36 out of 38 — reported that the pathological phenotype did transfer [s3]. The reviewers described that rate as implausible, argued it overstates the role of the gut microbiome in human disease, and called for a more rigorous approach to inferring causality specifically to avoid false concepts and unrealistic expectations that could undermine the field's credibility [s3].

That critique does not say the microbiome has no role in brain function. It says the standard experiment used to demonstrate one produces positive results at a rate that cannot be taken at face value, and that extrapolating from those animals to human disease is where the overclaiming happens [s3].

What an honest summary looks like

The gut-brain axis is real as anatomy and as physiology; nobody disputes that the gut and brain communicate. The question is what can be done about it, and the answer is asymmetric.

Acting on the brain to change gut symptoms has a randomised evidence base, a guideline recommendation, and a subgroup analysis showing perfect consistency where the therapy was delivered in sufficient volume [s1] [s2]. Acting on the gut to change the brain has a large, exciting, mechanistically appealing literature whose central experimental design has been publicly criticised by microbiome researchers for producing positive results 95% of the time [s3].

Both facts sit inside the same phrase. Only one of them is generally what a reader is being sold.

The limitations on the hypnotherapy side are real too: the overall pooled estimate is not statistically significant, heterogeneity across the full set of trials is extreme, and blinding a psychological therapy is close to impossible, so expectancy effects cannot be separated out [s1]. The subgroup findings are the strongest part of the analysis and subgroup findings are, by convention, hypothesis-generating.

This article is informational and is not medical advice. Decisions about therapy for gut symptoms, psychological or otherwise, belong with a reader and a clinician.

Sources

• Gut-Directed Hypnotherapy for Irritable Bowel Syndrome: A Systematic Review and Meta-Analysis — Neurogastroenterology & Motility, 2025-04-05
• ACG Clinical Guideline: Management of Irritable Bowel Syndrome — American Journal of Gastroenterology, 2020-12-14
• Establishing or Exaggerating Causality for the Gut Microbiome: Lessons from Human Microbiota-Associated Rodents — Cell, 2020-01-23