Why a Microbial Association Is Not a Treatment Target by Default
Why finding a microbe or microbial marker linked to a condition does not make it something you can treat, and the steps research needs before a link becomes a treatment target.
Content last updated . This page has not been medically reviewed.
Evidence labels on this page: association only, preclinical (cells/animals only), approved medicine, insufficient evidence and no good evidence of benefit. See How We Label Evidence in Specialty Topics.
Why an association is only a starting point
Human diseases are increasingly linked with an altered gut community, but whether such changes are causal, consequential or bystanders is, for the most part, unresolved. [1] A 2019 expert workshop agreed that it is not established whether dysbiosis is a cause, a consequence, or both. [2]Three common reasons a link can be real and still not be a target:
The condition changed the microbes. Illness changes diet, bowel habits, sleep and medicines, and these in turn shift the gut community. Stool consistency showed the largest effect size, and medication explained the largest share of variance, in a 2016 study of two cohorts of about 1,100 people each. [3] Antibiotics changed a small study's gut communities within 3 to 4 days. [4]
A third factor drives both. Alcohol intake and bowel movement quality differ between healthy and disease groups and are strong sources of variation. Matching cases and controls reduced spurious associations for several common diseases. [5]
The pattern does not hold up. A 2020 critique noted that many case-control studies are underpowered, and that meta-analyses are needed to sort consistent from spurious links. [6] A reanalysis of 680 stool samples from seven melanoma-immunotherapy studies began from the observation that available studies had not yet found reliable, consistent metagenomic markers of treatment response, then reported species that stood out when pooled. [7] Label: association only; pooling improves the pattern, but does not show that the species can be used to change outcomes.
The steps between a link and a target
| Step | Question | What it can support |
|---|---|---|
| Is the microbe or marker different in people with the condition? | A hypothesis |
| Does it appear in other groups, matched for diet, medicine and bowel habits? | A more reliable pattern |
| Does it come before the condition, in repeated samples over time? | Timing, not cause |
| Does a defined microbial product or pathway affect the body in animals or cells? | Preclinical plausibility only |
| In a randomized trial, does changing the microbe or its activity improve a patient-important outcome? | A treatment target |
| Has a regulator assessed an exact product for an exact use? | Approved medicine for that use only |
Animal evidence needs special caution: 95% of studies (36 of 38) that moved human gut microbes into rodents reported transfer of disease features, which the authors called implausible and said overstates the microbiome's role in human disease. [1] The idea of mechanisms vs benefit is covered in Mechanism vs Demonstrated Benefit.
Where a microbial approach has passed the later steps
Fecal-based products are the clearest example. The US FDA lists REBYOTA and VOWST as indicated to prevent recurrence of Clostridioides difficile infection in people 18 and older after antibiotic treatment for recurrent infection. [8][9] That is an approved medicine with a named indication, not a general "balance the microbiome" treatment. See Fecal Microbiota-Based Therapies.Where trials did not follow the association
- In a randomized, double-blind, placebo-controlled trial of healthy-donor fecal transplant in 27 people with sporadic ALS, the change in the main functional score did not differ from placebo, and the listed secondary efficacy outcomes (such as respiratory function and muscle strength) also showed no significant differences. The transplant group did have gastrointestinal side effects and improvements in constipation, depression and anxiety symptoms. [10] Label: no good evidence of benefit for the main outcome; small sample, larger trials needed.
- In a 2026 trial adding fecal transplant capsules to escitalopram in major depressive disorder, remission at week 8 did not differ significantly, though a symptom scale improved more. [11] Label: insufficient evidence. See Microbiome Research and Mental Health.
A symptom scale improving is a better outcome than a microbe count changing, but it is not the same as remission. See Surrogate vs Patient-Important Outcomes.
How to read a claim that says "target the microbiome for [condition]"
- Is the cited evidence a case-control comparison, a mouse study, or a human trial of the product?
- Did the study account for diet, medicines, alcohol and bowel habits? [5]
- Was the outcome how people felt or functioned, or only a change in microbes?
- Which exact product, strain or procedure, and for which exact use?
Related: Dysbiosis in Research and Marketing, Microbiome Studies: Association vs Causation, Observational Studies and Confounding.
Frequently asked questions
If a microbe is low in people with a condition, should I take it as a supplement?
That does not follow from the association alone. Whether such changes are causal, consequential or bystanders is mostly unresolved. [1]
What is the difference between an association and a cause?
An association is a pattern in data. A cause needs evidence that changing the factor changes the outcome, usually from a randomized trial.
Are microbiome markers used to guide treatment?
Pooled analyses suggest markers for some situations, such as response to cancer immunotherapy, but these remain research findings with inconsistent results across individual studies. [7]
Is fecal transplant a proven treatment?
For preventing recurrence of C. difficile infection, FDA-approved products exist. [8][9] Other uses are being studied, and two trials listed here did not show clear benefit on their main measures. [10][11]
More details: evidence, limits and cautions
The details behind this page: cautions, limits of the evidence and where sources disagree. Open any note to read it.
Safety note
References
- Walter J, Armet AM, Finlay BB, Shanahan F. Establishing or exaggerating causality for the gut microbiome: lessons from human microbiota-associated rodents. Cell 2020. PubMed 31978342. Read 2026-10-22.
- McBurney MI et al. Establishing what constitutes a healthy human gut microbiome. J Nutr 2019. PubMed 31373365. Read 2026-10-22.
- Falony G et al. Population-level analysis of gut microbiome variation. Science 2016. PubMed 27126039. Read 2026-10-22.
- Dethlefsen L, Relman DA. Incomplete recovery and individualized responses of the human distal gut microbiota to repeated antibiotic perturbation. PNAS 2011. PubMed 20847294. Read 2026-10-22.
- Vujkovic-Cvijin I et al. Host variables confound gut microbiota studies of human disease. Nature 2020. PubMed 33149306. Read 2026-10-22.
- Brüssow H. Problems with the concept of gut microbiota dysbiosis. Microb Biotechnol 2020. PubMed 31448542. Read 2026-10-22.
- Olekhnovich EI et al. Consistent stool metagenomic biomarkers associated with the response to melanoma immunotherapy. mSystems 2023. PubMed 36809182. Read 2026-10-22.
- US FDA, REBYOTA. Read 2026-10-22.
- US FDA, VOWST. Read 2026-10-22.
- Feng Y et al. Effect of fecal microbiota transplantation on patients with sporadic amyotrophic lateral sclerosis: a randomized, double-blind, placebo-controlled trial. BMC Med 2024. PubMed 39617896. Read 2026-10-22.
- Li et al. Adjunctive fecal microbiota transplantation for major depressive disorder: a randomized, double-blind, placebo-controlled trial. Cell Host Microbe 2026. PubMed 42309058. Read 2026-10-22.