Low-FODMAP diet reduces Bifidobacterium in IBS patients: direction-of-effect consistency analysis across 12 RCTs
The low-FODMAP diet consistently reduces Bifidobacterium abundance in IBS patients (9 of 11 studies, p=0.033), while its effect on F. prausnitzii remains inconsistent and statistically non-significant.
| Population | Adults diagnosed with irritable bowel syndrome (IBS) |
|---|---|
| Intervention | Low-FODMAP diet |
| Comparator | Control diet (habitual diet, sham diet, or other dietary interventions) |
| Outcome | Bifidobacterium abundance; Faecalibacterium prausnitzii abundance |
Summary of findings
| Outcome | Effect | 95% CI | Certainty | Clinical relevance | Notes |
|---|---|---|---|---|---|
| Bifidobacterium abundance | Sign test p=0.033 (primary, 11 studies, 9/11 decrease); sign test p=0.020 (sensitivity, 9 studies, 8/9 decrease); in the pooled estimate or 95% CI available | — | Low | — | 11 studies |
| Faecalibacterium prausnitzii abundance | Sign test not significant; 2/7 decrease, 4/7 in the change, 1/7 increase; in the pooled estimate or 95% CI available | — | Low | — | 7 studies |
Context
The low-FODMAP diet is recommended by clinical guidelines for symptom relief in IBS, but its impact on beneficial commensal bacteria raises relevant microbiological concerns. Both Bifidobacterium and F. prausnitzii are already reduced in IBS patients compared to healthy controls, making the question clinically important. High methodological heterogeneity across existing RCTs precluded conventional quantitative synthesis.
What the study showed
Of 11 RCTs measuring Bifidobacterium, 9 (81.8%) reported a decrease and 2 reported no change; the sign test confirmed a consistent reduction trend (p=0.033). Sensitivity analysis restricted to 9 studies with direct between-group comparisons showed 8 reporting a decrease (p=0.020). For F. prausnitzii (7 studies): 2 reported a decrease, 4 reported no change, and 1 reported an increase; the sign test was not statistically significant. Absolute effect sizes and 95% CIs were not calculable with the vote-counting methodology used.
How it was done
Systematic review of RCTs using direction-of-effect and consistency analysis (vote counting and sign test) instead of conventional meta-analysis, owing to high methodological heterogeneity. Twelve RCTs in adults with IBS comparing low-FODMAP diet to control diet were included. Systematic search up to September 2026; PROSPERO registration CRD420261351146.
Effect magnitude
Effect size with 95% CI cannot be expressed: vote counting and sign test methodology does not produce a pooled estimate. Sign test statistical significance for Bifidobacterium: p=0.033 (primary analysis) and p=0.020 (sensitivity analysis).
Risk of bias
The absence of quantitative meta-analysis prevents estimation of effect magnitude and 95% CI; the risk-of-bias tool is not explicitly stated in the available text (RoB 2 use cannot be confirmed). High methodological heterogeneity across RCTs (sequencing techniques, intervention duration, populations, and IBS definitions) limits comparability. Vote counting is a synthesis method considered hierarchically inferior to meta-analysis and susceptible to publication bias.
What this study does NOT prove
This study does not prove that Bifidobacterium reduction causes measurable clinical harm or that the low-FODMAP diet worsens microbiota-related health outcomes. Vote-counting methodology does not support quantitative causal inference or generalization to pediatric populations or those with specific comorbidities.
In clinical practice
Clinicians prescribing the low-FODMAP diet to IBS patients should inform them that the diet consistently reduces Bifidobacterium and that the effect on F. prausnitzii is uncertain. The gradual FODMAP reintroduction phase should be emphasized to mitigate microbiota impact. Supplementing with Bifidobacterium-containing probiotics during the elimination phase is clinically plausible but lacks direct RCT support designed for that purpose.
Limitations
The absence of quantitative meta-analysis prevents estimation of effect magnitude and 95% CI; the risk-of-bias tool is not explicitly stated in the available text (RoB 2 use cannot be confirmed). High methodological heterogeneity across RCTs (sequencing techniques, intervention duration, populations, and IBS definitions) limits comparability. Vote counting is a synthesis method considered hierarchically inferior to meta-analysis and susceptible to publication bias.
What is still missing
Studies monitoring Bifidobacterium and F. prausnitzii dynamics during the FODMAP reintroduction phase. RCTs stratifying patients by baseline microbiota subtype to identify those at highest risk of depletion.
Technical appendix
Version history
- 1.0 · 2026-10-23 — Auto-generated under Evidence Standard v1.0