Oxygen Levels in Your Gut May Decide Whether Yeast or Parasites Win

If you have spent months chasing gut symptoms, you have probably run into the word "candida" and found far more confident opinions than actual evidence. A new Dutch study of 2,848 stool samples finally puts numbers on who actually develops intestinal yeast overgrowth, and it lands on an unexpected suspect: oxygen. Two very ordinary exposures, antibiotics and long term acid reducers, sit at the centre of the pattern, which makes this immediately relevant if you take either one.

Key Takeaways

• Yeast in the gut is normal. It grew from 65.1% of 2,848 stool samples, but only 4.1% crossed the line into true overgrowth.
• Overgrowth tracked strongly with antibiotics taken in the previous three months (p<0.0001) and with proton pump inhibitor use (p=0.0001).
• As yeast took over, oxygen intolerant residents thinned out: Blastocystis fell from 26% to 14%, Dientamoeba fragilis from 14% to 5%, and overall bacterial diversity dropped.
• The oxygen explanation is the compelling part and also the weak link: it was shown in laboratory cultures, not measured inside anyone's intestine.

Yeast overgrowth appeared in 4.1% of 2,848 Dutch stool samples

Intestinal yeast overgrowth, defined as more than 100,000 yeast colony forming units per gram of stool (CFU/g, a count of the living yeast cells a sample will grow in culture), was found in 118 of 2,848 faecal samples, or 4.1%, in a Dutch retrospective single centre cohort study published in The Lancet Microbe. Yeast itself was thoroughly unremarkable: it grew from 1,855 of those samples, 65.1%. What separated people was quantity, not presence. Most samples sat in a normal band of 1 to 7,300 CFU/g (1,440 samples, 50.6%), a smaller group showed elevated growth between 7,301 and 100,000 CFU/g (297 samples, 10.4%), and only the top sliver tipped over into overgrowth. The thresholds were not invented for the paper; they were anchored against samples from 78 healthy donors.

On these numbers, finding yeast in a stool test is the default. Finding a lot of it is the rare event.

Antibiotics and proton pump inhibitors both tracked with yeast overgrowth

Two drug exposures were significantly associated with yeast overgrowth: antibiotic use in the three months before sampling (p<0.0001) and proton pump inhibitor use (p=0.0001), PPIs being the acid reducing drugs such as omeprazole that millions of people take for reflux, often for years. Neither association is shocking on its own. Antibiotics clear the bacterial competition, and PPIs are already known to reshape the gut community: a large Dutch cohort analysis in Gut found that PPI users carry a distinctly altered gut microbiome, with oral and upper gut bacteria surviving further down the tract than they otherwise would.

What is new is that both exposures converge on the same end state, and that this end state has a consistent, specific fingerprint rather than being generic "dysbiosis."

A matched 1:1:1 design separated overgrowth from recent antibiotic use

To test whether the fingerprint was just antibiotics doing the damage, the researchers built their comparison out of people who had taken none. Participants with overgrowth and no antibiotic exposure in the previous three months were matched by age and sex, in a 1:1:1 ratio, to participants who also had no recent antibiotics and who had either no yeast growth or normal to elevated growth. Yeast species were identified by CHROMagar culture plates and MALDI-TOF mass spectrometry (a technique that fingerprints an organism by the mass of its proteins). Parasites were diagnosed by microscopy plus PCR, and bacterial communities were profiled by 16S sequencing in a nested case control subset of 102 individuals.

Candida glabrata rose from 2% to 33% while parasites and diversity fell

Overgrowth had a distinct species profile. Candida albicans was present in 87 of 115 participants with overgrowth (76%), and Candida glabrata in 38 of them (33%). Compared with normal yeast growth, C glabrata prevalence rose sharply from 2% to 33% (p<0.001), while Candida parapsilosis moved the other way, falling from 10% to 2% (p<0.001). So overgrowth is not simply "more yeast." It is a different cast of yeast.

The parasite side of the ledger moved in the opposite direction. Overgrowth was associated with a reduced prevalence of Dientamoeba fragilis, from 14% to 5% (p=0.0055), and of Blastocystis species, from 26% to 14% (p=0.0062). Both are common single celled gut protozoa, frequently found in people with no symptoms at all. Alongside them, bacterial alpha diversity (a measure of how many different species share a sample) declined significantly (p=0.027), and the abundance of obligate anaerobes, bacteria that cannot tolerate oxygen, fell with it.

Oxygen picked the winner in the lab: Candida grew, Blastocystis did not

The unifying explanation the authors propose is oxygen, and they tested it directly in culture. Candida albicans failed to grow at all under strictly anaerobic conditions but thrived in oxygen rich ones. Blastocystis did the reverse: it grew abundantly without oxygen and was inhibited when oxygen was present. That single experiment ties most of the pattern together. If colonic oxygen rises, the organisms that tolerate it expand and the ones that cannot are squeezed out, which is the direction nearly every measurement in this cohort moved, the fall in Candida parapsilosis being the loose end the oxygen model does not obviously explain.

That idea is not exotic. A healthy colon lining is deliberately oxygen poor, and colonocyte metabolism is now understood to shape which microbes can survive there: gut lining cells burn butyrate from fibre fermenting bacteria and consume oxygen in the process, keeping the lumen nearly anaerobic. When that metabolism falters, oxygen leaks into the gut lumen and oxygen tolerant organisms bloom, a pattern also documented in colitis, where increased epithelial oxygenation accompanies the expansion of oxygen tolerant bacteria. This study extends that logic from bacteria to fungi and protozoa.

What this means if you take antibiotics or long term acid reducers

For someone on either drug, the useful takeaway is a question, not an action. This study cannot tell you that stopping a PPI will fix yeast overgrowth, and it does not test any treatment, supplement, antifungal, or diet. What it does offer is a reason to have a specific conversation with your doctor: whether a long running acid reducer is still needed at its current dose, and whether repeated antibiotic courses are being prescribed for problems that need them. Both drugs are genuinely important, and stopping a PPI abruptly can cause rebound acid symptoms, so this is a conversation, not a decision to make alone.

It is also a reason to read stool test results more sceptically. A report noting yeast, or noting Blastocystis or Dientamoeba fragilis, is describing something common. This work suggests those parasites may be markers of an intact, oxygen poor gut rather than the problem to be eliminated, close to the opposite of how such results are often sold. Anyone managing IBS, IBD, or chronic diarrhoea should treat that reframing as worth raising with a clinician rather than acting on.

Limitations: a referral population, and gut oxygen was never measured

The two honest caveats are large. First, this is a retrospective, single centre cohort assembled from people referred for parasitology examination, not from the general public, so both parasite prevalence and yeast overgrowth rates here should not be read as national figures. Second, and more important, colonic oxygen was never measured in a single participant. The oxygen mechanism was demonstrated in vitro, in dishes, while everything human in this study is association. A retrospective design also cannot resolve direction: yeast overgrowth may follow a rise in gut oxygen, or contribute to it, or both.

What would settle it is measuring luminal oxygen directly in people with and without overgrowth, and following patients forward through an antibiotic or PPI course rather than looking backward at stored samples. Whether restoring an oxygen poor colon changes yeast overgrowth in humans is completely untested, and until it is, the mechanism remains a well supported hypothesis rather than a fact about your gut.

Common Questions

Does having Candida in a stool test mean I have candida overgrowth? No. Yeast grew from 65.1% of the 2,848 samples in this study, and only 4.1% exceeded the overgrowth threshold of 100,000 CFU per gram. Presence and quantity are different findings, and only the quantity was linked to the drug exposures and community shifts described here.

Do proton pump inhibitors cause yeast overgrowth? This study found a statistically significant association between PPI use and yeast overgrowth (p=0.0001), not proof of cause. It is observational, so it cannot rule out that the reasons people are prescribed PPIs, rather than the drugs, explain part of the link.

Is losing Blastocystis a good thing? Not necessarily. In this cohort Blastocystis prevalence fell from 26% to 14% where yeast had overgrown, alongside a drop in bacterial diversity and in oxygen intolerant bacteria. That pattern points to Blastocystis disappearing because the gut environment has changed, not because it has been successfully treated.

Which yeast species matter most in overgrowth? Candida albicans was the most common, in 76% of overgrowth participants, but the sharpest shift was Candida glabrata, which rose from 2% prevalence in normal growth to 33% in overgrowth.

References

Jason J Biemond, Robert F J Kullberg, Mèlanie V Bénard, et al. (2026). Assessment of relationships between faecal yeast overgrowth and yeast species, protozoan parasites, and bacterial population dynamics: a Dutch retrospective, single-centre, cohort study. The Lancet Microbe. https://doi.org/10.1016/j.lanmic.2026.101432