Microbiome learning path · Chapter 3
Your stool and microbiome test report, read properly
Standard stool tests answer specific questions well: infection, inflammation, hidden bleeding, pancreas function. A microbiome report is a different thing. It lists which bacteria were found in one stool sample and what share of the total each one made up. It gives shares, not numbers. The result changes with your stool consistency, your medicines, what you ate and which laboratory you used. There is no agreed healthy profile to compare it against. Read with those limits in mind, it can still add something useful. Read as a diagnosis, it misleads.
In short
- Standard stool tests answer specific questions well: infection, inflammation, hidden blood, pancreas enzymes.
- A microbiome report is not a diagnostic test. No microbiome test is approved to diagnose any condition.
- A report gives each bacterium’s share of the total, not how many there are. Two people with the same percentages can carry very different numbers.
- Healthy people differ enormously, so a “normal range” drawn from one group does not transfer to another.
- Stool consistency, medicines, diet and the laboratory’s own methods all change the result.
- A stool sample describes the end of the colon. It says little about the small intestine.
- A test is worth doing when its result would change the plan.
Which stool test answers which question?
| Test | The question it answers |
|---|---|
| Microscopy, culture and DNA-based infection panels | Is there an infection or a parasite? |
| Faecal calprotectin | Is the bowel inflamed? A normal result makes inflammatory bowel disease unlikely. |
| Test for hidden blood (FIT) | Is there bleeding too small to see? Used to screen for polyps and bowel cancer. |
| H. pylori stool antigen | Is this stomach infection present, and has treatment cleared it? |
| Faecal elastase | Is the pancreas making enough digestive enzymes? |
| Microbiome sequencing | Which bacteria are present, and in what proportions? |
The first five are well established and are used in gastroenterology clinics everywhere. Each has a clear question, a validated method and an agreed cut-off. The last one has none of those things, which is why the rest of this chapter is about it.
See amoeba, giardia or Blastocystis on a stool report and the faecal calprotectin test.
How does a microbiome test work?
A stool sample is collected at home and sent to a laboratory. There, the bacterial DNA is extracted and read. Two methods are in common use.
16S sequencing
- Reads one marker gene that all bacteria carry
- Names bacteria reliably down to the genus, less often the species
- Finds bacteria only
- Costs less, so most consumer tests use it
Shotgun sequencing
- Reads all the DNA in the sample
- Can reach species and strain, and shows which genes are present
- Also finds viruses, fungi and parasites
- Costs more, and needs more computing
A third kind of test measures chemicals rather than DNA, usually in urine. These are discussed further down.
Whichever method is used, the laboratory then compares your list against a reference database and a group of other people, and prints a report. Every one of those steps is a choice, and each choice changes the answer.
From sample to report. Five steps, and a choice at each one.
Why does my report show percentages?
Because sequencing counts the DNA it reads, not the bacteria in your gut. Each result is a share of the total reads: 40 per cent of this, 12 per cent of that.
This has a consequence that almost no report explains.
The same percentages, very different numbers. A report gives shares of the total. It does not say how many bacteria there are.
In 2017, researchers counted the bacteria in stool samples directly, using a cell-counting machine alongside the sequencing. Healthy people differed by up to ten times in the total number of bacteria they carried. Two people can therefore have identical percentages on a report and very different gut communities.
Worse, percentages move together. If one group rises, every other group falls on paper, even when its real numbers have not changed. In that study, a famous “trade-off” between two major bacterial groups turned out to be an artefact of this arithmetic. It was not a real seesaw in the gut.
So when a report says a bacterium is “low”, the honest reading is: a low share of what was sequenced in that sample, on that day, by that method.
Why do two laboratories give different answers?
Because the method decides a large part of the result.
The extraction step
In a study that put the same stool through 21 different DNA extraction protocols, the extraction method had the largest effect on the result, bigger than the effect of how the sample was stored.
The primers
16S tests copy one gene before reading it. A widely used primer fits some bacteria poorly, so those groups are under-counted. Bifidobacteria, a group many reports comment on, have been affected by this.
The reference database
The list of known bacteria differs between databases and versions. A bacterium can be named differently, or not at all.
The comparison group
“Low” or “high” is relative to whoever the laboratory compared you with. Those people are usually not from your population.
This is why repeating a test at another laboratory is not a second opinion in the way a repeated blood count would be.
What else changes the result, apart from my gut?
A surprising amount.
- Stool consistency. In a study of 53 healthy women, how loose or hard the stool was correlated with the variety of bacteria found, with the ratio of the two main groups and with the overall pattern. In a population study of over a thousand adults, stool consistency had the single largest effect of any factor measured. A loose sample and a hard sample from the same person are not the same reading.
- Medicines. In that same population study, medication explained more of the variation between people than anything else. Acid-blocking medicines, antibiotics and many others leave a mark.
- What you ate. The community shifts within days of a change in diet.
- Recent antibiotics. After a course, most bacteria return within one to two months, but in one study nine common species were still missing at six months.
- The time of collection, how long the sample took to reach the laboratory, and how it was stored.
Not only your gut. Five things that change a stool report.
A good report asks about these. Most do not.
What is a diversity score, and does mine matter?
Diversity has two everyday meanings, and reports mix them up.
- Richness is how many different kinds were found.
- Evenness is whether they are present in similar amounts or dominated by a few.
Richness and evenness. Two different things that one diversity score blends together.
Most scores combine the two. Higher is often described as better, and in broad terms more variety goes with a diet rich in plants. But there is no threshold that marks health, the score depends on sequencing depth and method, and it moves with stool consistency. A single diversity number is a weak thing to make a decision on.
Can a stool test tell me about my small intestine?
Not really, and this matters, because many symptoms come from there.
A stool sample describes the end of the colon. In 2023, researchers gave 15 healthy people a capsule that collects samples as it travels down the gut. The bacteria, the viruses and the bile acids found inside the intestine differed from those in the same person’s stool.
What stool describes. The end of the colon, not the small intestine.
So a stool report cannot diagnose bacterial overgrowth of the small intestine. That is what a breath test is for, with its own limits. See is SIBO over-diagnosed? and hydrogen breath tests.
Is there a normal range?
No agreed one.
The largest surveys of healthy adults found that the kinds and amounts of bacteria varied widely from one healthy person to the next, while the work the community did stayed much the same. Only about a third of microbial genes are shared by most healthy people.
India makes the point plainly. In a study of 110 healthy Indian adults, people from north-central India, eating a largely plant-based diet, carried mostly one dominant group; people from southern India, eating a mixed diet, carried a different set. Both were healthy. A second project sampled about 1,000 adults at 14 locations across the country to map that variation.
Most commercial reference ranges are built from Western populations. Compared against them, a healthy Indian gut can be made to look abnormal.
What about fungi and Candida on a report?
Finding Candida does not mean you have an infection. In stool from healthy adults, baker’s yeast was present in almost every sample and Candida albicans in about four out of five. These are ordinary residents. A yeast infection of the gut is a different clinical picture, seen mostly in people who are seriously unwell or immunosuppressed.
What about urine “organic acid” tests?
Some tests measure chemicals in urine and present them as a picture of gut bacteria. A few of those chemicals genuinely are made partly by gut bacteria and partly by the body.
Urine organic acid analysis is an established laboratory method, used in the diagnosis of inherited metabolic disorders against defined standards. Its use as a general gut or “yeast” test is a different claim, and that claim is not established.
So what is a microbiome report actually good for?
Not nothing. Used carefully, it can:
- show a strikingly low variety after repeated courses of antibiotics
- show a heavy overgrowth of one organism, which prompts a look for a cause
- give a before-and-after picture when the same laboratory and method are used twice
- in research, help to answer questions that no other method can
What it cannot do is diagnose a disease, choose a probiotic for you, or prescribe a diet by itself. An international consensus statement in 2025 reached the same conclusion: microbiome testing is not ready to guide routine clinical decisions.
From my practice
When someone brings me a microbiome report, I read it last. First the history, the examination, the bowel habit, the medicines and the standard tests. Then the report, as one more piece of information. In my experience the report is most useful when it confirms something the story already suggested, and least useful when it is the reason a person has come. I have never started a treatment on the strength of a bacterial percentage alone.
Microbiome testing is available through my clinic. It starts with a consultation, so that the right test is chosen and the report is read alongside your history. See gut microbiome testing.
| What this page says | How it is known |
|---|---|
| Healthy people differ by up to ten times in total gut bacteria | Bacteria counted directly alongside sequencing, in people |
| The extraction method changes the result most | The same stool put through 21 protocols |
| A common primer under-counts some groups | Laboratory comparison of primers |
| Stool consistency has the largest single effect of any measured factor | A population study of over a thousand adults |
| The intestine differs from stool | 15 healthy people sampled with a swallowed capsule |
| Healthy Indians carry different dominant bacteria by region | Two Indian studies, 110 and about 1,000 adults |
| Microbiome testing is not ready to guide routine care | An international consensus statement |
What should I ask before paying for a test?
- What question will this test answer?
- Would the result change my treatment?
- Is there a simpler, validated test for the same question?
- Who will interpret it, together with my history?
- Which method and which comparison group does the laboratory use?
When should I see a doctor instead?
Before buying any test, if you have bowel symptoms lasting more than three to four weeks, and promptly for blood in the stool, weight loss, fever, or symptoms that wake you at night. Those need a diagnosis, not a bacterial profile.
Questions patients ask me
Can a stool test diagnose IBS?
No. IBS is diagnosed from its pattern of symptoms. Stool tests are used to rule out conditions that look like it, such as infection and inflammatory bowel disease.
My report says one bacterium is low. Should I take a supplement for it?
Not on that basis. The figure is a share of one sample, there is no agreed normal range, and taking a named bacterium has not been shown to correct a reported “deficiency”.
Should I have a microbiome test to choose a probiotic?
It is not reliable for that purpose at present. Probiotics are chosen by the condition being treated and the evidence for a particular strain.
I had two tests and they disagree. Which one is right?
Possibly neither, as a measure of your gut. If the laboratories used different extraction methods, primers, databases or comparison groups, two different answers are expected. Comparing results is only fair when the same laboratory and method are used both times.
Does my diversity score need to go up?
There is no target to reach. Eating a wider range of plant foods tends to go with more variety, and is worth doing for other reasons as well. Chasing the number is not a plan.
If my stool looks normal, can there still be bleeding?
Yes. Small amounts of blood are invisible. That is what the test for hidden blood detects, and why it is used for screening.
How is a sample collected?
You are given a clean container and instructions. Avoid mixing the sample with urine or water, and deliver it within the time stated. For some tests, recent antibiotics or bowel-cleansing preparations affect the result, so mention them. Note the consistency of the stool as well, because it affects a microbiome result.
Sources
This page is written for patients and is based on the guidelines and studies below. It is for education and does not replace a consultation.
- International consensus statement on microbiome testing in clinical practice. Porcari S et al. Lancet Gastroenterol Hepatol. 2025. PubMed 39647502
- American College of Gastroenterology guideline on irritable bowel syndrome. Lacy BE et al. Am J Gastroenterol. 2021. doi:10.14309/ajg.0000000000001036
- American College of Gastroenterology guideline on colorectal cancer screening. Shaukat A et al. Am J Gastroenterol. 2021. PubMed 33657038
- Maastricht VI/Florence consensus on Helicobacter pylori infection. Malfertheiner P et al. Gut. 2022. PubMed 35944925
- Revised estimates of human and bacterial cell numbers in the body. Sender R et al. PLoS Biol. 2016. PubMed 27541692
- Study counting gut bacteria alongside sequencing, showing tenfold differences in total load. Vandeputte D et al. Nature. 2017. doi:10.1038/nature24460
- Comparison of 21 DNA extraction protocols on the same stool samples. Costea PI et al. Nat Biotechnol. 2017. doi:10.1038/nbt.3960
- Evaluation of two primers commonly used to amplify bacterial 16S rRNA genes. Frank JA et al. Appl Environ Microbiol. 2008. doi:10.1128/AEM.02272-07
- Study of stool consistency, gut bacterial richness and composition in 53 healthy women. Vandeputte D et al. Gut. 2016. doi:10.1136/gutjnl-2015-309618
- Population study of what shapes the gut microbiome in two cohorts of about 1,100 adults each. Falony G et al. Science. 2016. doi:10.1126/science.aad3503
- Study sampling the human intestine with a swallowed capsule, compared with stool. Shalon D et al. Nature. 2023. doi:10.1038/s41586-023-05989-7
- Human Microbiome Project: structure, function and diversity of the healthy human microbiome. Human Microbiome Project Consortium. Nature. 2012. doi:10.1038/nature11234
- Review of what a healthy human microbiome is. Lloyd-Price J et al. Genome Med. 2016. doi:10.1186/s13073-016-0307-y
- Multi-omics study of the Indian gut microbiome. Dhakan DB et al. Gigascience. 2019. PubMed 30698687
- Pan-India profile of the human gut microbiome. Dubey AK et al. Sci Data. 2018. PubMed 30375992
- Study of how quickly diet changes the gut microbiome. David LA et al. Nature. 2014. PubMed 24336217
- Study of gut bacteria recovery in healthy adults after antibiotics. Palleja A et al. Nat Microbiol. 2018. PubMed 30349083
- Study of acid-blocking medicines and the gut microbiome. Imhann F et al. Gut. 2016. PubMed 26657899
- Study of gut fungi in healthy adults of the Human Microbiome Project. Nash AK et al. Microbiome. 2017. doi:10.1186/s40168-017-0373-4
- Review of Candida albicans as a normal resident and occasional pathogen of the gut. Kumamoto CA et al. Curr Opin Microbiol. 2020. PubMed 32604030
- North American consensus on hydrogen and methane breath testing. Rezaie A et al. Am J Gastroenterol. 2017. PubMed 28323273
- Laboratory standard for urine organic acid analysis. Gallagher RC et al. Genet Med. 2018. PubMed 29543224
- American Gut citizen-science study of diet and the microbiome. McDonald D et al. mSystems. 2018. PubMed 29795809