Semaglutide gut microbiome shifts in rats go beyond eating less

Semaglutide gut microbiome changes in obese rats went beyond what eating less would explain, a study in Comprehensive Physiology reports. Researchers at Penn State and the University of Calgary compared rats given the drug with rats fed exactly the same reduced diet. Several bacterial shifts, including a rise in Clostridium species, appeared only with the drug, and some patterns differed between males and females. The work used one stool sample per rat, did not test whether the changes help or harm, and has not been repeated in people.

How the pair fed design works

Semaglutide makes animals and people eat less, and eating less changes gut bacteria on its own. To pull those two effects apart, the team led by Karolina Skibicka used a pair feeding design. They studied 36 male and 36 female Sprague Dawley rats, housed alone and fed a high fat diet with 60% of calories from fat for six weeks before treatment and through four weeks of it. Each sex was split into three groups: untreated controls, rats given semaglutide injections with the amount raised in steps, and pair fed rats. Each pair fed rat received the same amount and mix of food that its matched semaglutide rat had eaten the day before, so any difference between those two groups points to the drug rather than the plate.

After four weeks, the researchers collected stool once and identified bacteria by sequencing the 16S rRNA gene, a standard marker that sorts microbes into families and genera. They also used a program called PICRUSt2 to predict what the bacteria might be doing, based on their genes rather than on direct measurement. As expected, semaglutide reduced food intake and weight gain in both sexes. In males, the drug cut weight gain more than pair feeding did, which suggests effects beyond appetite. In females, the drug outpaced pair feeding while the amount was being raised, but the difference during the steady phase was not statistically significant. Both the drug and pair fed rats ate less fibre than controls and took a larger share of calories from the fatty diet.

Semaglutide gut microbiome shifts by sex

Microbial richness, a count of how many kinds of bacteria can be detected, rose with semaglutide. In males it increased against controls while evenness, a measure of how equally those kinds are shared, fell. In females richness rose against both controls and pair fed rats, and evenness trended down without reaching significance. The Shannon index, which blends both, did not differ. The overall make up of the bacterial community differed among all three groups in both sexes, including between drug and pair fed rats. Bacteroidota, a major group of gut bacteria, fell by 76% in males on the drug and by 63% in pair fed males. In females the drops were 66% and 49%, so food restriction explained much, but not all, of that change.

The overall make up of the bacterial community differed among all three groups in both sexes, including between drug and pair fed rats.

Sex shaped several of the results. Verrucomicrobiota, the group that includes Akkermansia, fell in pair fed males but rose in pair fed and drug treated females, with the largest rise on the drug. Links between bacteria and body weight also split by sex. In males, more weight gain went with more Bacteroidota and less Firmicutes, while in females it went with more Bacteroidota and less Verrucomicrobiota. Among females, more Lachnospiraceae went with more weight gain and more Akkermansiaceae with less. Food intake and the share of fatty food added further differences, mostly in males. Earlier rodent work has shown that GLP 1 effects can depend on different brain wires, and the authors suggest a gut and brain loop here, but they did not test it.

Which bacteria rose on the drug alone

The clearest drug only signal came from the Firmicutes. The Clostridiaceae family rose about threefold in males and about fivefold in females given semaglutide. Several genera and species, including Clostridium sensu stricto 1, Clostridium perfringens and Terrisporobacter, were enriched with the drug in both sexes but not in pair fed rats. Females on the drug also showed more Peptostreptococcaceae and Akkermansiaceae. Other shifts, notably the loss of many Bacteroidota related taxa, tracked reduced food intake and appeared in both the drug and pair fed groups. The predicted functions also moved. In males, drug linked pathways included amino acid and lipid metabolism, secondary bile acid production and breakdown of foreign compounds, while females showed broader predicted changes.

Whether any of this is good or bad is open, and the authors say so. In the paper's own words the shifts are "not necessarily health positive". Clostridium perfringens can cause food poisoning and gut infections, while Akkermansia has been tied to better metabolic health in other studies. A higher relative share of a group does not mean the bacteria are causing disease, and 16S sequencing cannot tell harmless strains from toxin producing ones. The predicted pathways are also guesses from gene content, not measured activity. Read against our microbes section, where most stories involve peptides that kill bacteria, this one is a reminder that a peptide drug acting on hormones can also reshape the bacteria that live with us.

What a single stool sample cannot show

The design has clear limits. Each rat gave one stool sample at the end of treatment, so the study cannot show how the changes developed or whether they would settle with time. There was no sample taken before treatment and no group kept on standard chow alone, which makes it harder to know the starting point. The rats ate a very fatty diet for ten weeks, so the results may not carry over to animals or people on ordinary food. Two of the co authors work at Wright Labs, a private sequencing company. The funders, listed in the journal record, were the US National Institute of Diabetes and Digestive and Kidney Diseases and two Canadian research councils.

Still, the semaglutide gut microbiome question matters because millions of people now take GLP 1 drugs for months or years. Laboratory work on this site has shown that gut hormones of the same family protect the mouse intestine from immune attack, and human studies of these drugs in bowel disease are growing. The rat data suggest that bacterial shifts during treatment are not only a side effect of smaller meals. Whether similar shifts happen in people, and whether they matter for weight loss, nausea or infection risk, would need stool samples collected over time from patients. The paper in Comprehensive Physiology lists the full taxa and statistics.

Conclusion

This rat study separates the drug from the diet and finds that the semaglutide gut microbiome picture includes changes that eating less does not reproduce, with a sharp rise in Clostridiaceae and several sex specific patterns. It rests on one stool sample per animal, inferred rather than measured functions and rats on a fatty diet. Whether the shifts help, harm or matter in people is unknown.

Source Byun S, Sotzen MR, Bento MT and colleagues, Skibicka KP. Semaglutide Remodels the Gut Microbiome Beyond Its Appetite-Suppressing Action: Sex-Specific Gut-Brain Axis Implications. Comprehensive Physiology, October 2026. DOI 10.1002/cph4.70286

Amino Weekly reports research. It is not medical advice and never gives doses.