Saccharomyces boulardii is a yeast; the probiotics you usually see – Lactobacillus, Bifidobacterium and their relatives – are bacteria. For preventing antibiotic-associated diarrhoea, both the yeast and specific bacterial strains have evidence and guideline support, so the organism type alone does not decide the choice. The yeast is not killed by antibiotics, which makes it convenient to take during a course – a practical advantage, not a proven better outcome. The one indication where a guideline names only the yeast is preventing Clostridioides difficile–associated diarrhoea in children, and that recommendation rests on low-quality evidence. For acute infectious diarrhoea and for H. pylori treatment, the trials cited here did not show a clear benefit.
Are these even the same kind of thing?
Saccharomyces boulardii is a strain of yeast – a single-celled fungus, biologically closer to baker's yeast than to any gut bacterium. Bacterial probiotics are strains of live bacteria. That distinction is not pedantry; it drives the one difference that matters most in practice.
Not to be confused with: "probiotic" as a category name covers both, so a label reading "probiotic" tells you almost nothing about which organism is inside. And S. boulardii is sometimes listed as Saccharomyces cerevisiae var. boulardii – the same organism under a longer name, not a separate product.
Because it is a yeast, S. boulardii is intrinsically resistant to antibacterial antibiotics. Bacterial probiotics are not; an antibiotic taken at the same time can kill the bacterial strain you just swallowed. That single biological fact is the mechanism behind most of the situations where the yeast is preferred. It is a mechanism, not an outcome – it explains why the yeast tends to be the one studied alongside antibiotics, not that it produces a better result on its own.
What has actually been measured?
The two organisms do not share a head-to-head evidence base, so the claims below are kept separate by what was tested.
Acute infectious diarrhoea. A Cochrane systematic review, whose main analysis rested on large trials at low risk of bias, concluded that probiotics probably make little or no difference to the number of people whose diarrhoea lasts 48 hours or longer, and that it is uncertain whether they shorten diarrhoea at all (Cochrane, 2020). For a bout of ordinary infectious diarrhoea, this is not a reason to take a probiotic of either kind.
Antibiotic-associated diarrhoea in children. This is where the evidence is strongest. A Cochrane review of 33 trials (6,352 children) found a moderate protective effect: antibiotic-associated diarrhoea occurred in 8% of children taking probiotics versus 19% of controls (moderate-certainty evidence), with a clearer effect at doses of at least 5 billion CFU per day (Cochrane, 2019). The ESPGHAN working group recommends either Lactobacillus rhamnosus GG or S. boulardii for this purpose in children (moderate-quality evidence, strong recommendation) – one bacterium and one yeast, so the guideline does not favour either type (J Pediatr Gastroenterol Nutr, 2016). The PLoS One guide likewise lists S. boulardii CNCM I-745 among probiotics with strong evidence here, next to a three-strain Lactobacillus mixture and L. casei DN114001. Both reviews focus on children; adults were not their subject.
Helicobacter pylori eradication. This is where marketing most often puts the yeast, and where the trials here are least supportive. In a randomised trial of 404 healthy adults aged 40–64, adding S. boulardii CNCM I-745 (500 mg twice daily) to clarithromycin triple therapy produced no significant difference in eradication or side effects once the analysis was corrected for multiple comparisons; the authors conclude that on the 14-day regimen the yeast non-significantly lowered diarrhoea and did not increase eradication (Eur J Cancer Prev, 2024). A second trial in China compared a 10-day vonoprazan–amoxicillin regimen with S. boulardii against a 14-day bismuth quadruple regimen: eradication rates were similar (87.3% vs 88.9%), with fewer side effects and lower cost in the yeast arm (BMC Gastroenterol, 2024). That trial compared two different drug regimens, so it cannot tell you what the yeast itself added.
Where does the yeast genuinely have an edge – and where does the claim outrun the data?
Sorting the differences honestly:
- Specific to the yeast, but weakly supported: ESPGHAN suggests S. boulardii for preventing C. difficile–associated diarrhoea in children – the only indication where the guideline names the yeast alone, with low-quality evidence and a conditional recommendation. Not supported by the trials here: adding the yeast to standard H. pylori treatment did not significantly improve eradication.
- Demonstrated as biology, not yet as a head-to-head win: the yeast's antibiotic resistance is real and easy to explain. It makes the yeast a sensible candidate when an antibiotic is on board. It does not, by itself, show that the yeast produces better clinical outcomes than a bacterial strain given in the same setting – that comparison has not been made directly.
- Claimed but unresolved: broad marketing that positions the yeast as a universally "stronger" or "more resilient" probiotic runs ahead of the evidence. Resilience to stomach acid or antibiotics is a survival property, not a clinical benefit, and one does not automatically deliver the other.
An important practical point sits underneath all of this. A guide to choosing probiotics published in PLoS One made the case that most products on the market lack the trial evidence to support their specific use, and that effectiveness is strain- and indication-specific rather than a property of "probiotics" in general. In other words, the organism on the label and the exact indication both matter – a bacterial blend studied for one condition tells you little about a different blend for a different condition.
Who should choose the bacterial option instead?
There are concrete situations where the yeast is not the sensible default:
- No antibiotic involved. The yeast's headline advantage is antibiotic resistance. If you are not taking an antibiotic, that advantage does nothing, and the choice comes back to which organism has been studied for your specific indication.
- You need a strain studied for a bacterial-specific indication. Much of the probiotic evidence base – for the general infectious diarrhoea and paediatric settings above – was built on bacterial strains or on mixed groups, not on the yeast alone.
- Immunocompromise or a central venous line. Live yeast is a living organism, and the point at which a live-organism supplement becomes a clinical risk – rather than a shelf choice – is a decision for the treating clinician, not a label. This is the situation to raise with a doctor before starting either type.
How to weigh the criteria
Rank the criteria by how strongly the evidence supports them, not by how loudly a label states them:
- Match the organism to the indication (direct human outcome evidence). The strongest lever you have. The trials are indication-specific; "probiotic" alone is not a specification.
- Antibiotic resistance when an antibiotic is co-prescribed (mechanistically plausible; not shown to give better outcomes). A convenience reason to prefer the yeast during an antibiotic course, not a proven advantage.
- Survival and "resilience" marketing (untested as a clinical benefit). Treat as a manufacturing property, not a health outcome.
Verdict
There is no general winner. For preventing antibiotic-associated diarrhoea, S. boulardii and L. rhamnosus GG are recommended side by side, and the yeast’s resistance to antibiotics is a convenience rather than a demonstrated benefit. The yeast has a guideline-specific role only in preventing C. difficile–associated diarrhoea in children, on weak evidence. For acute infectious diarrhoea, probiotics of either kind probably make little or no difference, and adding the yeast to H. pylori treatment did not improve eradication in the trial that isolated it. No trial has pitted the yeast directly against a bacterial probiotic for the same indication. The useful question is which strain was tested, at what dose, for the exact thing you want to prevent – and on a label, look for the strain designation (CNCM I-745 in the studies here that name one) and the live-cell count guaranteed until the expiry date.
Sources
- Probiotics for treating acute infectious diarrhoea (Cochrane Database Syst Rev, 2020) – in large low-bias trials, probiotics probably make little or no difference to diarrhoea lasting 48 hours or longer; effect on duration uncertain.
- Probiotics for the prevention of pediatric antibiotic-associated diarrhea (Cochrane Database Syst Rev, 2019) – 33 trials, 6,352 children: AAD 8% vs 19%; clearer effect at ≥5 billion CFU/day.
- Probiotics for the Prevention of Antibiotic-Associated Diarrhea in Children (J Pediatr Gastroenterol Nutr, 2016) – ESPGHAN: L. rhamnosus GG or S. boulardii for AAD in children; S. boulardii suggested for C. difficile–associated diarrhoea (low-quality evidence).
- Choosing an appropriate probiotic product for your patient: An evidence-based practical guide (PLoS One, 2018) – basis for the point that effectiveness is strain- and indication-specific and many products lack supporting trials.
- Randomised clinical trial: efficacy and safety of H. pylori eradication treatment with and without Saccharomyces boulardii supplementation (Eur J Cancer Prev, 2024) – RCT, n=404: no significant improvement in eradication or diarrhoea after statistical correction.
- Effectiveness and safety of vonoprazan and amoxicillin dual regimen with Saccharomyces boulardii supplements on eradication of Helicobacter pylori (BMC Gastroenterol, 2024) – 10-day vonoprazan–amoxicillin + S. boulardii vs 14-day bismuth quadruple: similar eradication, fewer adverse events; does not isolate the effect of the yeast.
