Microbiome Tests: Why Experts Say Not Yet

Send one sample to six companies and you get six different answers. That's the finding, and it's the whole story.

What this is

Direct-to-consumer microbiome tests sequence what's in a stool sample and return a report: diversity scores, "good" and "bad" bacteria, food lists, and usually supplements to buy. The science underneath is genuine. The product is where it breaks down.

When researchers tested this directly — sending standardised material and split samples to commercial providers — they found major discrepancies both within and between companies, with variation between providers on the same scale as the biological variation between different people's guts. An international expert panel has been convened specifically because there is no agreed standard for clinical microbiome testing.

What the evidence says

  • The gut microbiome is genuinely important to health

    Gold· human data

    Not in question. The gap is between population-level science and individual-level interpretation.

  • Commercial tests disagree with each other on the same sample

    Gold· human data

    Analytical-performance studies found discrepancies within and across providers comparable to real between-person differences. If the measurement isn't reproducible, nothing built on it can be.

  • There is no consensus on clinical interpretation

    Gold· human data

    Expert reviews state plainly that evidence for clinical usefulness is scarce and that the tests are sold without agreed regulation or proven value in practice.

  • Their personalised food and supplement recommendations are validated

    Not supported

    The recommendations are the commercial layer. Experts reviewing real reports called the interpretations premature and unsupported — and they're frequently attached to the provider's own supplement range.

What this means in practice

  • Save the money. The general advice a report will give you — more fibre, more plant variety, fermented foods, less alcohol — is available without sequencing anything.
  • If you have symptoms, the useful tests are the ones a GP orders: coeliac serology, faecal calprotectin, FIT, iron studies.
  • Be sceptical of any test whose output is a list of the seller's products.
  • Revisit in a few years. This is a field that may well become clinically useful; it isn't yet.

When to see a doctor

  • Bleeding, unexplained weight loss, or persistent change in bowel habit — see a GP rather than testing at home.
  • Iron-deficiency anaemia, which needs a cause found.
  • A test report that recommends eliminating multiple food groups.

The honest summary

The science is real, the products aren't ready, and the measurement doesn't reproduce. Spend the money on food instead — the actionable advice was free all along.

See also

Frequently asked questions

Are gut microbiome tests accurate?
Analytical studies found major discrepancies between providers on identical samples, on the same scale as differences between different people. Reproducibility is the first requirement of a useful test, and it isn't met.
What about food intolerance tests?
IgG-based food intolerance tests are explicitly not recommended by allergy societies — food-specific IgG reflects exposure and tolerance, and positive results are expected in healthy people. We cover this on the at-home tests page.

Sources

Written and reviewed by Mathew Beale, MSc Biotechnology, University of Reading.

Last reviewed: