ResearchResearch summary
Comparing seven consumer gut microbiome testing services
Identical standardized stool material produced substantial within-service and between-service differences in a blinded commercial comparison.
In this pathway
- Current topic
- Comparing seven consumer gut microbiome testing servicesCurrent
Bottom line
Seven consumer services produced materially different profiles from standardized identical stool material, including conflicting classifications within one service.
Why this source matters
It tests analytical reproducibility directly instead of assuming that a commercial report accurately represents a person's microbiome or health.
Study and population
This laboratory comparison analyzed 20 reports returned from 21 kits ordered across seven anonymized services; it included no patients.
What was tested
Three kits per service received aliquots from the same homogenized pooled stool material, following each service's collection instructions.
Results
The workflows reported 1,208 unique taxa but shared only three genera across every analysis. One service labeled two identical-material replicates healthy and the third unhealthy.
Applicability
The findings concern analytical performance of seven services, not Candida-specific accuracy, diagnosis, or clinical outcomes.
Limitations
The material was not a ground truth, providers were anonymized, workflows differed, and one provider returned only two successful reports.
Contribution to the evidence
Supporting. The identical-sample design directly demonstrates reproducibility and comparability problems relevant to interpreting commercial reports.
Practical change
No practical change. The study supports caution but cannot decide whether one person has Candida-related disease.
Funding and conflicts
NIST researchers and reference material supported the work; no separate grant funder was reported. One author cofounded a microbiome-biotherapeutics company; all others declared no competing interests.
Source
Read the original publication (published 26 February 2026).
Study type, population, and sample scale
- Study type
- Blinded laboratory comparison of direct-to-consumer gut microbiome testing services using standardized fecal reference material
- Population
- Seven anonymized consumer testing services receiving homogenized aliquots from the same pooled human stool material; this was an analytical study, not a patient diagnostic study
- Sample scale
- 21 kits ordered (three per service); six services returned three reports each and one returned two reports after repeated quality-control failure, for 20 reports analyzed
Results
- Across seven services plus two NIST workflows, 1,208 unique taxa were reported but only three genera appeared in every analysis; within-service conservation ranged from 6% to 98% of identified genera.
- One service classified two replicate reports from the identical material as healthy and the third as unhealthy; seven of ten functional categories were below average in that third report versus one different below-average category in each of the other two.
Limitations
- Identical material produced substantial variation within and between services.
- The reference material was not an absolute ground truth, so the study could assess precision but not determine which reported composition was correct.
- Reports gave conflicting health classifications and taxonomic profiles.
- Services were anonymized, workflows differed, most profiling was bacterial, and no patient outcome or Candida diagnosis was tested.
Applies to
- Analytical reproducibility and comparability of the seven anonymized services and workflows tested with standardized material
Does not apply to
- Candida-specific diagnostic accuracy, clinical benefit, Polish-market surveillance, or proof that every consumer microbiome service performs identically
Contribution to the current conclusion
Supporting
This blinded comparison sent standardized identical fecal material to seven consumer services and directly exposes within-provider failures, between-provider variability, and conflicting health classifications; a summary can also make clear that it assessed analytical performance, mostly bacterial profiles, not a Candida diagnosis or patient benefit.
Related evidence mapPractical change or no change
No practical change. The comparison supports caution when interpreting consumer stool microbiome reports but cannot validate or invalidate a specific Candida diagnosis or determine patient benefit.
Funding and conflicts
- Funding
- The work included US National Institute of Standards and Technology researchers and NIST-developed reference material; the article reports no separate grant funder in its funding statement.
- Conflicts
- One author reports cofounding a biotechnology company focused on microbiome live biotherapeutics; the other authors declare no competing interests.
- Paper date
- 2026-02-26
Sources
View all 1 source
- Evaluating the analytical performance of direct-to-consumer gut microbiome testing services
- Source type
- Laboratory study
- Published
- 2026-02-26