A Low-Gluten Diet Reduces the Abundance of Potentially Beneficial Bacteria in the Gut Microbiota of Healthy Adults
Eight weeks of a low-gluten diet significantly reduced the abundance of Bifidobacterium spp. and other potentially beneficial bacteria in healthy adults without gluten-related disease.
| Outcome | Grade | Direction | Effect | Studies |
|---|---|---|---|---|
| Bifidobacterium spp. abundance | C | ▼ Unfavorable | log2FC <-1, FDR <0.1 (95% CI not reported) | 1 |
| Beta diversity (community structure) | C | ▼ Unfavorable | PERMANOVA Bray-Curtis p<0.05 (effect size R2 not reported) | 1 |
| Alpha diversity (Shannon, OTUs, Chao) | C | — Neutral | LME ANOVA p>0.05 at M2 and M4 (effect size not reported) | 1 |
| Abundance of carbohydrate-utilising bacteria | C | ▼ Unfavorable | log2FC <-1, FDR <0.1 (95% CI not reported) | 1 |
| Fecal fermentative metabolites (¹H NMR) | C | — Insufficient | quantitative values not extractable from available text | 1 |
| Bacteroidota/Bacillota ratio | C | — Insufficient | LME ANOVA applied; direction and magnitude not extractable from available text | 1 |
| Persistence of microbial changes at 16 weeks | C | ▼ Unfavorable | no reversal observed at M4 (n=20; statistical values not explicitly reported for M4 subset) | 1 |
Context
Adoption of gluten-reduced diets by healthy individuals has increased over the past decade without robust evidence. Changes in gut microbiota composition associated with this practice in populations without medical indication require better characterization. This study investigated the impact of 8 and 16 weeks of a low-gluten diet (LGD) on the microbiota and fecal metabolites of healthy French adults.
What the study showed
After 8 weeks of LGD, the abundance of Bifidobacterium spp. decreased significantly (qPCR; log2FC not numerically specified in available excerpt, statistically significant at FDR<0.1). Beta diversity (community structure) was altered between M0 and M2 (PERMANOVA, p<0.05 reported). No significant changes in alpha diversity (Shannon, observed OTUs, Chao) were detected at M2 or M4. Extension to 16 weeks (M4) did not reverse changes observed at M2.
How it was done
Single-centre, controlled, randomised trial with two consecutive 8-week dietary intervention periods. Forty healthy adults completed M0→M2; 20 continued to M4. Microbiota assessed by 16S rRNA sequencing (V3–V4 region, Illumina MiSeq), qPCR, cultural approach, and ¹H NMR for fecal metabolites. Sequencing performed on a subset (n=36 samples from 12+12+6 individuals).
Effect magnitude
Statistically significant reduction of Bifidobacterium spp. and carbohydrate-utilising taxa after LGD; effects reported as log2FC with FDR<0.1, without explicitly provided 95% CI values in the available text. Alpha diversity did not differ significantly at any timepoint.
Limitations
Small sample (n=40 at M2; n=20 at M4), with an even smaller sequencing subset (n=12 per timepoint). No parallel control group maintaining HGD throughout the entire period, limiting control for temporal variation. Risk of bias tool not explicitly applied (no RoB 2 report). Microbiota was a pre-specified secondary outcome. Fibre substitution (addition of psyllium to gluten-free bread) confounds attribution of effects solely to gluten. V3–V4 sequencing alone limits species-level taxonomic resolution.
In clinical practice
Healthcare professionals should not recommend low- or gluten-free diets to healthy adults without medical indication based on this study. Gluten reduction may adversely affect potentially beneficial gut bacteria such as Bifidobacterium spp. without demonstrated clinical benefit in this population. Patients self-adopting LGD should be counselled on potential microbiota impact and nutritional composition of the diet.
What is still missing
Randomised controlled trials with parallel control groups, larger samples, and duration exceeding 16 weeks are needed to establish causality and reversibility of microbial changes. The effect of gluten per se must be isolated from dietary fibre modification.