A bacterium that normally lives in the human colon has emerged as an intriguing candidate in research on the gut-muscle axis. In a 2026 study, mice given Roseburia inulinivorans developed greater forelimb grip strength. Separately, older adults with detectable levels of the species had 29 per cent stronger handgrip than those in whom it was not detected.
Put those two findings together too quickly and the study can sound like a human probiotic trial. It was not. The mouse experiment supports a biological effect under carefully constructed laboratory conditions. The human figure came from a small, cross-sectional comparison, meaning bacteria and strength were measured at the same point in time.
That division is the key to reading the result fairly. The animal evidence is more than a correlation, but causality in mice does not automatically become causality in people. The human association is worth testing, but nobody in the study gained 29 per cent grip strength after taking the bacterium.
What the 29 per cent figure actually describes
The paper was published online in Gut on 10 March 2026. Researchers analysed data from two Spanish cohorts that had originally been recruited for other randomised trials. For this microbiome study, however, the relevant measurements were observational rather than a randomised bacterial intervention.
The older cohort contained just 33 adults aged 65 to 75. Their mean age was 68.4 years, their mean body mass index was 29.6, and 52 per cent were men. Stool sequencing was used to classify participants according to whether each Roseburia species was detectable.
Those with detectable R. inulinivorans, at an average relative abundance of 0.3 per cent, had 29 per cent greater handgrip strength than the group with no detectable signal. The difference had a reported p value below 0.01. Their peak oxygen uptake was not correspondingly higher, suggesting that the finding was not simply a mirror of cardiorespiratory fitness.
Two close relatives, R. faecis and R. intestinalis, were not significantly associated with grip strength in the older adults. That species specificity makes the result more interesting than a vague claim that “good gut bacteria” accompany fitness.
It also needs the right denominator. Twenty-nine per cent was the difference between two existing groups in a 33-person sample. It was not a 29 per cent improvement following treatment, and it does not tell us how much an individual’s strength would change if the bacterium appeared or disappeared.
The younger cohort broadened the association
The study also included 90 adults aged 18 to 25, with a mean age of 21.9. Researchers divided them into low, intermediate and high abundance groups for each bacterial species. The average relative abundance of R. inulinivorans ranged from 0.05 per cent in the low group to 2.7 per cent in the high group.
Higher abundance was associated with stronger handgrip and greater peak oxygen uptake. Across the wider analysis, the bacterium also tracked with leg-press and bench-press performance. Finding a related pattern in younger and older cohorts reduces the chance that the older result was a completely isolated accident.
It still does not establish direction. Stronger people may exercise more, consume different diets, carry different amounts of lean mass, take different medicines or have health conditions that shape both their muscles and microbiome. Resistance training itself has previously been reported to increase the abundance of the broader Roseburia genus.
This leaves at least three plausible stories. R. inulinivorans may influence muscle function. Strength-promoting behaviour may create a gut environment in which the bacterium does better. Or another factor may produce both. Cross-sectional human data cannot choose among them.
The mouse experiment supplied a causal test
To move beyond association, the team studied 32 six-week-old male C57BL/6J mice. All animals first received a cocktail of four antibiotics for two weeks to deplete their existing gut microbiota. They were then divided among four conditions: R. inulinivorans, R. faecis, R. intestinalis, or a phosphate-buffered control.
The human-derived bacterial strains were delivered orally three times a week for eight weeks. Forelimb strength was measured by allowing each mouse to grip a small bar while it was gently pulled backwards. The instrument recorded the peak force before the animal released its grip.
At weeks four, six and eight, mice receiving R. inulinivorans produced approximately 30 per cent greater grip strength than controls, with reported p values below 0.001. Neither of the related species generated a comparable strength effect. Correcting grip force for lean body mass did not erase the difference.
The change was specific to strength rather than general exercise capacity. None of the Roseburia strains improved running time to exhaustion. The treatment also did not increase whole muscle weight or change body weight and food intake in a way that readily explained the grip result.
Stronger grip came with altered muscle fibres
Muscle tissue offered a possible physical explanation. In the soleus, R. inulinivorans treatment increased the proportion of type II fibres, often called fast-twitch fibres because they support relatively rapid, forceful contractions. The distribution also shifted towards larger fibre cross-sectional areas, including more fibres above 5,000 square micrometres.
That does not mean the bacterium simply built more muscle throughout the body. The reported total muscle weights did not rise, and some microscopic comparisons were stronger than others. The finding is better described as a change in fibre architecture that accompanied greater grip performance.
The experiment therefore provides evidence that the bacterium, or something set in motion by its repeated administration, caused stronger grip in this mouse model. It does not prove the same effect in an older person with an intact microbiome.
The model was deliberately artificial. Broad-spectrum antibiotics do much more than make space for an introduced strain. They alter microbial ecology, immune signalling and host metabolism. All the mice were young males, whereas the hoped-for application involves ageing adults and people with muscle-wasting disorders.
The bacteria did not permanently move in
There was another important limitation: none of the human-derived Roseburia species established lasting colonies in the mice. Samples collected more than 24 hours after administration contained little or no trace of the introduced organisms.
The effect may therefore have depended on transient signals produced while the bacteria passed through, or on interactions that temporarily changed the resident community. The R. inulinivorans group showed increases in the mouse gut genera Muribaculum, Blautia and Ruminococcus, leaving open the possibility that the administered species acted partly through other microbes.
This matters for translation. A product would not merely need to contain the right name on a label. Researchers would need to establish dose, viability, delivery, residence time and safety in the far more complex human gut. They would also need to learn whether transient exposure is sufficient or whether repeated dosing is required.
The suspected mechanism was not butyrate
Roseburia species are well known for producing butyrate, a short-chain fatty acid involved in intestinal and metabolic health. It was an obvious candidate mechanism, but caecal concentrations of butyrate and other short-chain fatty acids did not explain the mouse result.
Instead, R. inulinivorans reduced concentrations of several amino acids in the caecum and plasma, including methionine and branched-chain amino acids. Muscle profiling showed enrichment of purine metabolism and the pentose phosphate pathway, systems involved in nucleotide production, redox control and energy metabolism.
The researchers propose that these changes may support muscle protein synthesis and fibre remodelling. The sequence is not yet a complete mechanism. Inflammation and neuromuscular signalling were not directly assessed, and the measurements cannot identify one molecule that travelled from bacterium to muscle and produced the response.
That uncertainty is useful rather than disappointing. It prevents a familiar gut-microbiome shortcut in which every apparent benefit is attributed to butyrate. Here, the data point towards amino-acid handling and energy pathways, while leaving room for alternative or interacting explanations.
Promising candidate, not a probiotic recommendation
ScienceBlog’s earlier overview of the study described the animal result, the human cohorts and the possible metabolic route. The cautious update is that no randomised human trial has tested R. inulinivorans supplementation for strength, sarcopenia or athletic performance.
There is no validated clinical-grade supplement to recommend. The organism is strictly anaerobic, making exposure to oxygen a practical formulation problem. Eating inulin or another fermentable fibre also cannot be assumed to reproduce the mouse experiment. Diet changes many microbial species and metabolites at once, and a bacterium’s name does not turn a food into a targeted treatment.
The commercial context deserves disclosure. Several paper authors are named as inventors on an international patent family covering uses of R. inulinivorans for muscle mass and strength, assigned to Leiden University Medical Center and the University of Granada. A patent is not evidence against the science, but it gives some researchers and institutions a stake in successful development.
The strongest conclusion is therefore narrower than the most exciting one. Repeated administration of R. inulinivorans increased grip strength in antibiotic-treated young male mice. In 33 older adults, its detectable presence accompanied 29 per cent stronger handgrip, but did not prove that the bacterium caused the difference.
The next decisive evidence would be a controlled human trial measuring safety, microbial response and strength over time. Until then, R. inulinivorans is a serious experimental lead, not a substitute for resistance training, adequate nutrition or clinical care.