Back

Rapid remodeling of the human gut microbiome in response to short-term animal product restriction

Emmanouil, C.; Anezaki, M.; Simistiras, A.; Glentis, S.; Scarmeas, N.; Hatzis, P.; Rouskas, K.; Dimas, A. S.

2025-08-02 systems biology
10.1101/2025.07.30.667695 bioRxiv
Show abstract

Diet strongly influences the gut microbiome, which in turn influences health, yet the effects of dietary patterns on microbiome composition and function remain underexplored in humans. We profiled a unique group of apparently healthy individuals from Greece, who alternate between omnivory and restriction of animal products for religious reasons (periodically restricted group, N=200). Using 16S rRNA sequencing, plasma metabolomics, and proteomics, we assessed the impact of three-to-four weeks of dietary restriction on gut microbiome composition and function and also explored links with host plasma biology. We compared findings to a continuously omnivorous group profiled in parallel (non-restricted group, N=211). Animal product restriction was found to reduce microbial diversity, primarily affecting rare taxa, and altered the abundance of nearly one-third of bacterial genera. Functional shifts included downregulation of cholesterol biosynthesis and purine degradation pathways, alongside upregulation of microbial biosynthesis of vitamin B2 and tryptophan, suggesting compensatory microbial responses to dietary nutrient depletion. Multi-omic integration revealed four microbial-metabolite-protein clusters, including a diet-responsive module linking Negativibacillus with potent metabolic regulator FGF21 and intermediate-density lipoproteins. Our findings demonstrate rapid adaptive plasticity of the human gut microbiome in response to short-term dietary restriction and highlight candidate microbial and molecular pathways that may mediate effects of animal product restriction on health.

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.