First, the good news: I have 133 microbial species living in my gut. That’s a collection that ranks as ‘excellent’ according to NirvanaBiome, the company in Germantown, Maryland, that analysed my stool sample. The firm based its rating on the overall diversity of my gut’s residents plus my levels of specific beneficial and undesirable species.

Revealed: how this common gut microbe protects against heart disease
But then came a second, less promising assessment: Viome in Bellevue, Washington, identified only 61 kinds of active microorganism in my gut and suggested that some of them were not performing optimally. Viome recommended dietary changes and offered customized supplements to improve my health.
I was intrigued by the results, but also flummoxed by the different messages. That’s not an uncommon response, says microbiologist Rob Knight, director of the Center for Microbiome Innovation at the University of California, San Diego. When scientists tested a handful of microbiome testing kits alongside one another in formal studies, they found wide discrepancies in the results. These differences make it hard to understand or really trust the findings from any given provider, says Diane Hoffmann, director of the Law and Health Care programme at the University of Maryland in Baltimore, who co-authored one of those studies.
The companies stand by their products. “Of course there’s variability,” says Kelly Moffat, chief operations officer for Cmbio in Germantown, which does the laboratory work and genetic sequencing for NirvanaBiome. “There has to be because this is complex testing.”
Momo Vuyisich, chief science officer at Viome’s lab in Bothwell, Washington, also says he was unsurprised by the degree of variation found by formal studies. Labs use different methods to extract, sequence and analyse microbial genetic material. Viome, for example, sequences RNA rather than DNA, allowing it to uncover microbial gene activity, rather than only the presence of genes. Vuyisich argues that this is more relevant to human health. And he says that although every method will have biases, as long as a consumer sticks with the same provider, it should produce results that are comparable over time.
Interest in these kinds of test is booming. “There’s so much public enthusiasm and excitement about the microbiome,” says Knight. Some customers are simply curious about the microscopic beings that live inside them and potentially affect their health; others have health conditions such as irritable bowel syndrome and hope that their gut microbes will reveal the underlying cause or a path to better health.
Although people can order the tests and follow their recommendations without the help of a medical professional, some people are bringing the findings to their physicians. “I groan when I see someone come in with a microbiome test,” says Fergus Shanahan, a gastroenterologist and immunologist at University College Cork in Ireland, who has spent much of his career studying the microbiome. For someone struggling with gut health, a consumer microbiome profile isn’t usually specific enough to help. And if there’s a concern about an infection, such as Clostridioides difficile, it’s more straightforward to test for known pathogens than to assess the entire microbial community.

Microbiologists and gasteroenterologists who spoke to Nature generally agreed that microbiome testing has the potential to both reflect and promote health, but many said that the science isn’t yet mature enough to support microbiome testing for most consumers. In their view, the limited clinical usefulness of today’s consumer tests can make them poor value for money (costing around US$100–300). The specialists warn that acting on advice that doesn’t come from clinicians could lead customers to adopt unnecessarily restrictive diets or delay appropriate medical assessment. Critics say that consumer microbiome testing needs more regulatory scrutiny and oversight.
“This is an industry that’s running wild,” says genomicist Jacques Ravel, director of the Center for Advanced Microbiome Research and Innovation at the University of Maryland School of Medicine in Baltimore and a co-author on one of the studies that tested several providers. With regards to testing methods, he says, “there’s no standard”.
Moffat says, “I think the jury’s out on exactly what it is that we can actionably do today based on this data.” But she says that microbiome profiles, combined with other testing such as gut metabolomics, could yield interesting and useful health information.
An appealing proposition
The gut microbiome has emerged as an incredible correlate of overall health. From immune-system function to metabolism, skin disorders and mental health, study after study has found links between a person’s well-being and the trillions of microbes that live in their digestive tract. The premise of a head count of these species and assessment of their actions, to monitor and improve gut health in a personalized way, certainly sounds appealing. Viome’s website, for example, promises: “No more trial and error … Follow a plan built precisely for you.”
Companies can identify species and, on the basis of those lists, infer what the microbes are probably doing, such as manufacturing metabolites or inciting inflammation. But running even the simplest microbiome testing is expensive, says Alessandra Cervino, research director at PiLeJe, a science-based food-supplement company in Paris. She says her microbiome testing company, Luxia Scientific, closed in 2024 for financial reasons.

Many testing companies also offer lifestyle and nutritional advice or sell supplements. For example, Viome analysed diverse functions performed by the microbes in my gut, such as digestive efficiency, gas production and serotonin manufacture. I then received a detailed eating plan — which told me to eschew broccoli and load up on pecans, among other suggestions — along with an offer for precision supplements and probiotics with a price tag of more than $150 a month. Although it’s hard to change the composition of microbes in one’s gut, alterations in diet or even simply ageing can shift their activities, says Vuyisich. The company recommends re-testing every four to six months to check for shifts and update the plan.
He points to 19 peer-reviewed publications by Viome as evidence of the company’s rigour (see go.nature.com/4actdhq). For example, the firm’s analysis for consumers is supported, in part, by published data on how gut microbes influence individuals’ responses to glucose in various foods, based on 550 people in the United States who collectively ate more than 30,000 meals while wearing glucose monitors1. The company validated those findings in a further cohort in Japan, Vuyisich says, but those data are not published.
Microbiome tests are not created equal, says Tim Spector, a physician-scientist at King’s College London and co-founder of London-based ZOE, which offers microbiome testing and a supplement in addition to a gut-health app that I’ve been playing with. ZOE’s products, Spector says, are backed by research on thousands of people. “There definitely are some not-so-good players out there that give this field a bad name,” he says. “It’s not easy to do this well.”
Testing the tests
There are also several ways to do microbe testing; companies take varying approaches that yield disparate results, as at least two studies have shown. Scientists at the US National Institute of Standards and Technology (NIST) in Gaithersburg, Maryland, shipped homogenized stool samples from one person to seven companies2. NIST’s remit is researching and applying standards for measurement. This experiment was a step towards understanding the diversity of testing options out there, says one of the study authors, NIST microbiologist Stephanie Servetas, who is based in Sheridan, Wyoming. The researchers didn’t name the companies in their report because they wanted to focus attention on the comparison between tests rather than perceived accuracy of any individual provider.
The result: “Companies produce strikingly different results,” says co-author Scott Jackson, a molecular geneticist who used to direct the NIST microbiology program and is now a consultant in Saint Michaels, Maryland. “They couldn’t even agree on what’s healthy and unhealthy,” Jackson says. The microbe lists ranged from 34 to 906 identified genera; only 3 of which showed up in all reports. In fact, the results were as varied from company to company as they would be for different people, despite all tests being done on samples from the same individual.
A European and Canadian team obtained similar outcomes when they sent homogenized samples from one person’s stool to six companies, also anonymized. Three companies rated the microbial diversity as “excellent” or “good”; two called it “average”; one said it was “unfavourable”3. The results were so variable, that at least some of them must be lacking in validation or standardization, says co-author Céline Druart, executive director of the Pharmabiotic Research Institute (PRI) in Brussels, a non-profit organization specializing in regulation of microbiome innovations. “It’s very difficult to make a conclusion,” she says.

How gut bacteria could boost cancer treatments
This broad divergence in results is not surprising, says Jack Gilbert, director of the Microbiome and Metagenomics Center at the University of California, San Diego. The variation starts at sampling: scoop a bit of faecal matter into a tube, stab it with a coring device or swipe the sampler across some used toilet paper; ship it off as-is, or in some kind of preservative (as both Viome’s and NirvanaBiome’s kits asked me to). Each option could influence the balance of the microbes that are collected and what survives to be sequenced.
In the lab, Servetas says the method by which the researchers extract genetic material from a sample is a big source of divergence. Gentler protocols will bias a sample towards microbes that are easy to crack open, whereas harsher methods will capture a higher share of species with tougher exteriors.
After sequencing, another major source of difference, says Knight, is the bioinformatics pipeline that labs use. What length or number of genetic sequences is required to consider a microbe present? What list of sequences and microbes are they using as a reference? “All bets are off when you then computationally analyse it with your favourite method,” says Knight.
Would companies support having standards to validate their tests? “We are really big advocates on really having a standard here,” says Kimberley Sukhum, chief science officer at Tiny Health in Austin, Texas, which offers gut microbiome tests as well as vaginal microbiome tests.
But it’s hard to set those standards, in part because there is no known ‘ground truth’ for a natural microbial sample. Compare that with, say, a blood-glucose test; a lab can get a known value from a standard sample either right or wrong. But the microbiome, in its glorious complexity, does not allow for such simple validation. NIST has used mixed human donor samples to develop a standard4. And Tiny Health has experimented with generating a reference for benchmarking from synthetic communities. It was built from individual microbial species rather than sampled from faeces5. But neither of those standards is exactly like the microbiome of an individual person.
A healthy sample
Even if labs could assess a microbiome perfectly, they’d still have a problem: there’s no defined healthy microbiome like there is one range of healthy blood-glucose levels. Each person’s microbiome is unique, like a fingerprint. There are infinitely many ways to have a healthy microbiome, and infinitely many ways to have an unhealthy one, says Shanahan; it’s not entirely clear how to draw a line between the two.

How gut microbes could drive brain disorders
Companies must compare customer microbiomes with some standard of health, but these vary too. They might be based on past customers or large, public databases. But microbiomes differ considerably depending on factors such as ethnicity, sex, age, recent antibiotic use or even whether someone was born by caesarean section.
Enjoying our latest content?
Log in or create an account to continue
- Access the most recent journalism from Nature's award-winning team
- Explore the latest features & opinion covering groundbreaking research