Back

Mechanistic insights into microbiome-dependent and personalized responses to dietary fibre in a randomized controlled trial

Armet, A. M.; Li, F.; Deehan, E. C.; Nikolaeva, D. D.; Delannoy-Bruno, O.; Siegwald, L.; Berger, B.; Minehira Castelli, K.; Rodionov, D. A.; Arzamasov, A. A.; Liu, J.; Seethaler, B.; Cole, J. L.; Nguyen, K. N.; Jin, M.; Zhao, Y.-Y.; Sharma, A. M.; Curtis, J. M.; Proctor, S. D.; Bischoff, S. C.; Wismer, W. V.; Osterman, A.; Bakal, J. A.; Greiner, R.; Field, C. J.; Knights, D.; Prado, C. M.; Walter, J.

2025-11-21 nutrition
10.1101/2025.11.20.25340625 medRxiv
Show abstract

Dietary fiber supplementation can reduce cardiometabolic risk, but its effective use is limited by incomplete understanding of fibre-microbiome interactions and highly individualized responses. We tested acacia gum (AG; fermentable fibre), resistant starch type 4 (RS4; fermentable fibre), and microcrystalline cellulose (MCC; non-fermentable control fibre) in a six-week randomized trial in adults with excess body weight. Multi-omics profiling revealed distinct, structure-specific microbiota and short-chain fatty acid shifts with AG and RS4, which were not directly linked to physiological outcomes. Improvements in inflammation, gut barrier function, and satiety occurred across all arms, indicating fermentation-independent effects. AG reduced plasma ghrelin, linked to microbial carbohydrate-active enzyme genes targeting its structures. Machine-learning models predicted individualized, fiber-specific effects on blood pressure (AG) and C-reactive protein (RS4) from microbial pathways and fecal bile acids. These findings delineate fermentation-dependent and independent mechanisms of fibre action and provide a mechanistic basis for personalized fibre supplementation. Trial registration: ClinicalTrials.gov NCT02322112

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.