Gut bacteria may engraft as a species after a fecal transplant, but the exact strain rarely does

If you or someone you love has gone through a fecal microbiota transplant (FMT) to treat a stubborn gut infection or inflammatory bowel disease, you have probably heard that success is a coin flip: it works beautifully for some patients and barely moves the needle for others, even when the same donor stool is used. A new study offers the clearest look yet at why, and the answer sits one level deeper than scientists have usually looked.

Key takeaways

• Researchers tracked 13 fecal transplant patients and their donors with genome-level (strain-resolved) sequencing, not just species-level testing.
• Donor bacterial species often showed up in the recipient's gut, but the exact donor strain (the specific genetic lineage within that species) rarely took hold. Only 22 of 465 apparent colonization events were confirmed as true donor-strain capture.
• Clinical remission tracked with donor-strain replacement specifically, not with how much the overall gut community resembled the donor's.
• This is an observational finding in FMT patients, not a randomized trial, so it points to a mechanism worth testing, not a guaranteed fix.

Strain-level filtering, not species overlap, governs gut colonization

Researchers at the University of Bern, led by Bahtiyar Yilmaz, found that a transplanted gut microbe can look like it has successfully colonized a new host while its actual genetic identity has already been swapped out. The team, publishing in Cell Reports, used longitudinal shotgun metagenomics (sequencing nearly all the DNA in a stool sample, repeated over time) to track 13 donor-recipient pairs before and after FMT, reconstructing 3,748 microbial genomes directly from the sequencing data. When they looked at species-level detection alone, donor microbes appeared to spread widely. But when they zoomed in to strain resolution, the picture changed: most of that apparent colonization was an illusion of the recipient's own resident bacteria rebounding, not the donor's bacteria taking over.

Why this matters for IBD, C. diff, and other gut-driven conditions

FMT is already the most effective treatment available for recurrent Clostridioides difficile infection, and it is being tested with much more mixed results in inflammatory bowel disease (IBD), metabolic conditions, and other disorders where gut bacteria are suspected to play a role. The patients in this study included cases of recurrent C. diff infection, IBD, a rare post-surgical condition called protein-losing enteropathy, and recurrent parasitic infection. Doctors and researchers have struggled to explain why the same transplant procedure produces remission in one patient and no benefit in another. This study suggests part of the answer is not which bacterial species arrive, but whether the recipient's gut lets the donor's specific bacterial strain permanently displace the strain already living there.

Tracking bacteria down to the strain, patient by patient, over months

The researchers sampled each donor-recipient pair at baseline and four weeks after transplant, with a subset also sampled at eight weeks; two patients underwent repeated FMT rounds, letting the team watch the same gut ecosystem get perturbed multiple times. Instead of relying on species-level classification (which groups together bacteria that can differ enormously at the genetic level), they used population-genetic tools to compare donor and recipient genomes at single-nucleotide resolution, the level of detail needed to tell whether a bacterium is genuinely the donor's strain or just a closely related one the patient already carried. Of 252 bacterial species found only in the donor before transplant, just 16.7% persisted through week 8, while 61.9% failed to establish at all. Among 465 cases that looked like new colonization at the species level, strain-resolved analysis confirmed only 22 (4.7%) as genuine donor-strain capture; another 116 turned out to be a different, recipient-native strain of the same species, and the rest could not be resolved with confidence.

What this could mean if you or a family member are considering FMT

For patients weighing FMT, the practical takeaway is not that the procedure is unreliable, but that "the donor's bacteria showed up" on a standard test may not mean what it sounds like. The study found one particularly striking pattern: Lachnospiraceae, a bacterial family that is normally abundant and beneficial in the healthy gut, was among the most frequent short-term colonizers (one member, Dorea formicigenerans, appeared in 9 of 13 patients) but had one of the worst strain-capture rates, with only about 3% of its colonization events confirmed as true donor-strain replacement, compared to roughly 15% to 16% for some less common families. In other words, a stool test that shows a favorable-looking bacterial species after your transplant is not proof that the transplant has durably rewired your gut ecosystem. If you are tracking your own or a family member's FMT progress, this is worth raising with your gastroenterologist: ask what kind of testing was used to confirm engraftment, and be cautious about treating early species-level improvement as a guarantee of lasting change.

What remains unproven and unresolved

This is an observational, correlational study, not a randomized controlled trial, so it cannot prove that strain replacement causes remission rather than simply accompanying it. The cohort was also small, just 13 recipients, all of whom already had an indication for FMT, so the findings describe strain dynamics in sick patients seeking treatment, not the general population. Strain-level resolution also has technical limits: most apparent colonization events (about 70%) had too little sequencing depth to confirm a strain match either way, so the true rate of donor-strain capture could be somewhat different from the confirmed figures. The researchers themselves frame this as identifying a candidate mechanism worth testing further, not a finished explanation for why FMT succeeds or fails.

What comes next for strain-tracking and transplant design

The team also found that repeating FMT multiple times in the same patient gradually increased how many donor species stuck around, building what they call "ecological memory," but this repetition still did not reliably overcome the strain-level bottleneck that determines whether the donor's exact bacterial lineage takes hold. That distinction points toward a next generation of FMT research: rather than optimizing for donor microbial diversity or species overlap, future trials may need to screen donors and recipients for strain-level compatibility before transplant, echoing work from other groups on the clinical and ecological determinants of strain engraftment, such as this multi-donor trial analysis in Cell Reports Medicine and a related randomized-trial engraftment analysis. A companion paper in Cell Host & Microbe similarly links strain-level genetic heterogeneity to how well microbiome therapies work, reinforcing that strain tracking, not just species counting, is likely to become standard in how FMT trials are designed and evaluated.

Common Questions

Does this mean fecal transplants don't work? No. FMT remains highly effective for recurrent C. diff infection specifically. This study explains variability in outcomes across different conditions by showing that species-level success and strain-level success are not the same thing.

What is the difference between a bacterial species and a bacterial strain? A species is a broad category (like "E. coli"), while a strain is a specific genetic lineage within that species. Two strains of the same species can behave very differently, so knowing a species "showed up" after a transplant does not tell you which strain it actually is.

Can a stool test tell my doctor if my transplant actually took? Standard species-level stool testing cannot reliably distinguish a donor's exact strain from the patient's own resident bacteria of the same species. This study used specialized genome-level sequencing not available in routine clinical testing, so ask your care team what level of detail their testing provides.

Why did Lachnospiraceae colonize so often but rarely persist as the donor's strain? The researchers describe this as a "colonization-persistence paradox": Lachnospiraceae bacteria are common and readily detected shortly after transplant, but the recipient's own strains of that same family tend to rebound and outcompete the donor's specific strain over time.

References

1. Isabel Baertschi, Sebastian B.U. Jordi, Lucie Jeanne Gardaz, et al. (2026). Strain-level ecological filtering governs microbial colonization of the human gut. Cell Reports. https://doi.org/10.1016/j.celrep.2026.117775
2. Anna H. Behling, Theo Portlock, Daniel Ho, et al. (2025). Cohort-specific determinants of donor strain engraftment following multi-donor faecal microbiota transplantation in two randomised clinical trials. Gut Microbes. https://doi.org/10.1080/19490976.2025.2597628