Skip to content
PaperFren

Concept · biology

Microbiome

Follow Microbiome — see important new research and changes in evidence.

Change log

What changed

Dated edits to this page's evidence: studies added or removed from a claim, claims added or withdrawn, and new explanations tagged here. Rewordings are not listed.

Show 5 earlier
  • Sep 14, 2026

    Mendelian randomisation proving a taxa–cancer association means the microbe’s mechanism in tumours is known.

    • Added a misconception
  • Sep 14, 2026

    B. dorei dominance before autoantibodies proves that B. dorei causes type 1 diabetes.

    • Added a misconception
  • Sep 14, 2026

    If two papers both study “the gut microbiome,” their disease conclusions can be pooled as one effect.

    • Added a misconception
  • Sep 14, 2026

    Microbiome findings from one model organism always transfer to humans.

    • Removed a misconception
  • Sep 3, 2026

    • Concept page published

The microbiome is the community of microorganisms living in or on a host (or in an environment), studied for composition, genes, and effects on host physiology. In this library it ranges from human gut taxa linked to metabolic traits and disease risk, through strain-level gene content and bile-salt enzymes, to experimental animal and bee systems that test how disruptions change the community.

Students meet microbiome claims that sound interchangeable—“gut bugs cause disease”—when the evidence is usually a specific host, assay, and outcome. The skill is saying what a study adds, where it applies, and what would change the position.

Evidence

What the evidence shows

Drawn from 11 studies in this library. Each finding starts with a plain-language takeaway, then the denser detail. Supports means evidence for a finding; Challenges means evidence against a stated position; Qualifies marks scope with a short note on each study’s contribution. Challenged positions are labeled — they are not findings.

  • Genetic instruments for gut taxa link a small set of microbiome–cancer associations under Mendelian randomisation.

    Using Mendelian randomisation on gut-taxa genetic instruments, eleven stringent microbiome-to-cancer associations were reported (for example higher breast-cancer odds with Actinobacteria/Bifidobacteriales signals; Tyzzerella3 linked with higher lung adenocarcinoma but lower colorectal cancer risk).

    1 supporting · 2 qualifying

    Qualifies

    1. 1B. dorei rises before T1D autoimmunitydifferent disease and design — observational T1D timing, not cancer MR
    2. 2Gut microbes, blood metabolites, and metabolic traitsdifferent outcome — metabolic traits/metabolites, not cancer

    Study comparison

    StudyRoleDesignNPopulationOutcome
    Mendelian randomisation associates specific gut taxa with several cancers2023SupportsMendelian randomisationTwo-sample MR of MiBioGen gut taxa instruments against eight cancer GWAS summary sets211 taxa instruments; cancer GWAS sample sizes vary by malignancy — no single primary NGut microbiome and cancer GWAS summary statistics (IEU OpenGWAS/consortia)Causal microbiome–cancer associations (stringent and sensitivity analyses)
    B. dorei rises before T1D autoimmunity2014Qualifiesdifferent disease and design — observational T1D timing, not cancer MRCohortLongitudinal DIPP stool sequencing in HLA-risk children before T1D autoimmunityN=76 · 947 stool samples; 29 seroconverters vs 47 autoantibody-negative controlsHLA-DQB1 moderate-to-high-risk children in DIPPGut microbiota (B. dorei dominance) prior to persistent T1D autoantibodies
    Gut microbes, blood metabolites, and metabolic traits2017Qualifiesdifferent outcome — metabolic traits/metabolites, not cancerCross-sectionalMETSIM subset microbiome–metabolite–metabolic trait association analysisN=531 · 531 middle-aged Finnish men from the METSIM cohort (parent cohort 10,197)Middle-aged Finnish men in METSIMAssociations of gut microbiota with plasma metabolites and metabolic traits
  • Before type 1 diabetes autoimmunity, Bacteroides dorei can dominate children’s gut communities.

    In children who later developed type 1 diabetes autoimmunity, Bacteroides dorei became dominant in the gut community before persistent autoantibodies (mean diagnosis age about 16.8 months).

    1 supporting · 2 qualifying

    Qualifies

    1. 1Mendelian randomisation associates specific gut taxa with several cancersdifferent disease and method — cancer MR in adults, not pediatric T1D timing
    2. 2Gut microbes, blood metabolites, and metabolic traitsdifferent population — middle-aged metabolic traits, not autoimmunity onset

    Study comparison

    StudyRoleDesignNPopulationOutcome
    B. dorei rises before T1D autoimmunity2014SupportsCohortLongitudinal DIPP stool sequencing in HLA-risk children before T1D autoimmunityN=76 · 947 stool samples; 29 seroconverters vs 47 autoantibody-negative controlsHLA-DQB1 moderate-to-high-risk children in DIPPGut microbiota (B. dorei dominance) prior to persistent T1D autoantibodies
    Mendelian randomisation associates specific gut taxa with several cancers2023Qualifiesdifferent disease and method — cancer MR in adults, not pediatric T1D timingMendelian randomisationTwo-sample MR of MiBioGen gut taxa instruments against eight cancer GWAS summary sets211 taxa instruments; cancer GWAS sample sizes vary by malignancy — no single primary NGut microbiome and cancer GWAS summary statistics (IEU OpenGWAS/consortia)Causal microbiome–cancer associations (stringent and sensitivity analyses)
    Gut microbes, blood metabolites, and metabolic traits2017Qualifiesdifferent population — middle-aged metabolic traits, not autoimmunity onsetCross-sectionalMETSIM subset microbiome–metabolite–metabolic trait association analysisN=531 · 531 middle-aged Finnish men from the METSIM cohort (parent cohort 10,197)Middle-aged Finnish men in METSIMAssociations of gut microbiota with plasma metabolites and metabolic traits
  • In middle-aged men, gut composition associates with metabolites and metabolic-syndrome traits.

    In middle-aged men, gut microbiota composition was associated with plasma metabolites and metabolic-syndrome–related traits, placing the microbiome alongside genetics and lifestyle as a correlated host factor.

    1 supporting · 2 qualifying

    Qualifies

    1. 1Mendelian randomisation associates specific gut taxa with several cancersdifferent outcome — cancer odds via MR, not metabolic metabolites
    2. 2Selective bacterial BSH shifts host metabolismmechanistic enzyme study — not a population metabolic association

    Study comparison

    StudyRoleDesignNPopulationOutcome
    Gut microbes, blood metabolites, and metabolic traits2017SupportsCross-sectionalMETSIM subset microbiome–metabolite–metabolic trait association analysisN=531 · 531 middle-aged Finnish men from the METSIM cohort (parent cohort 10,197)Middle-aged Finnish men in METSIMAssociations of gut microbiota with plasma metabolites and metabolic traits
    Mendelian randomisation associates specific gut taxa with several cancers2023Qualifiesdifferent outcome — cancer odds via MR, not metabolic metabolitesMendelian randomisationTwo-sample MR of MiBioGen gut taxa instruments against eight cancer GWAS summary sets211 taxa instruments; cancer GWAS sample sizes vary by malignancy — no single primary NGut microbiome and cancer GWAS summary statistics (IEU OpenGWAS/consortia)Causal microbiome–cancer associations (stringent and sensitivity analyses)
    Selective bacterial BSH shifts host metabolism2018Qualifiesmechanistic enzyme study — not a population metabolic associationAnimal / in-vitroBacteroidetes BSH selectivity screen plus B. thetaiotaomicron BT2086 deletion in gnotobiotic mice~20 Bacteroidetes strains screened; monocolonization used 12 mice per Bt WT/KO group (8 GF controls) — no single primary NHuman-gut Bacteroidetes strains and germ-free C57BL/6 miceHost metabolic phenotypes after selective bile-salt hydrolase loss
  • Bacterial bile-salt enzymes can change host bile acids and weight-related phenotypes in models.

    Selective bacterial bile-salt hydrolases can change host bile-acid chemistry and weight-related phenotypes in colonization models—evidence that specific microbial enzymes, not only community membership lists, can matter for host physiology.

    1 supporting · 2 qualifying

    Qualifies

    1. 1How much do gut bacterial gene contents differ?different question — strain gene-content variation across people
    2. 2Antibiotics hurt bee gut and survivaldifferent host — honey bees, not mammalian colonization models

    Study comparison

    StudyRoleDesignNPopulationOutcome
    Selective bacterial BSH shifts host metabolism2018SupportsAnimal / in-vitroBacteroidetes BSH selectivity screen plus B. thetaiotaomicron BT2086 deletion in gnotobiotic mice~20 Bacteroidetes strains screened; monocolonization used 12 mice per Bt WT/KO group (8 GF controls) — no single primary NHuman-gut Bacteroidetes strains and germ-free C57BL/6 miceHost metabolic phenotypes after selective bile-salt hydrolase loss
    How much do gut bacterial gene contents differ?2015Qualifiesdifferent question — strain gene-content variation across peopleCross-sectionalMetagenomic gene-deletion detection quantifying within-species gene content across gut communitiesN=207 · 252 fecal metagenomes from 207 individuals (HMP and European MetaHIT)Human gut metagenomes from public cohortsInter-individual within-species gene-content variation
    Antibiotics hurt bee gut and survival2017Qualifiesdifferent host — honey bees, not mammalian colonization modelsAnimal / in-vitroWorker honeybees fed tetracycline; 16S community and survival vs germ-free controlsCup cages of 30 bees × 15 replicates per condition; 16S profiles n≈14–15 per arm — no single primary NWorker honeybees with conventional or germ-free gutsCore gut microbiota disruption and mortality after tetracycline
  • Strain-level gene content varies a lot within the same gut species — species labels are coarse.

    Even within the same named gut species, strain-level gene content varies substantially among healthy people—so species labels alone understate functional diversity.

    1 supporting · 2 qualifying

    Qualifies

    1. 1Is the gut microbiome the same along the intestine?different scale — mucosal vs luminal niches along the gut
    2. 2Human body microbiome biogeographydifferent scale — body-site biogeography, not within-species gene content

Open questions

Tensions and limits

Some items are genuine disagreements on the same question. Others mark different assays, populations, or outcomes — limits on how far one study travels — not a forced fight between papers.

  • Scope / different questions

    Cancer MR, pediatric T1D timing, and adult metabolic-trait associations all involve gut microbiota, but they answer different disease and design questions. Treating them as one interchangeable “gut microbiome disease effect” collapses distinct estimands.

    3 studies
    1. 1Mendelian randomisation associates specific gut taxa with several cancers
    2. 2B. dorei rises before T1D autoimmunity
    3. 3Gut microbes, blood metabolites, and metabolic traits

    Study comparison

    StudyRoleDesignNPopulationOutcome
    Mendelian randomisation associates specific gut taxa with several cancers2023SupportsMendelian randomisationTwo-sample MR of MiBioGen gut taxa instruments against eight cancer GWAS summary sets211 taxa instruments; cancer GWAS sample sizes vary by malignancy — no single primary NGut microbiome and cancer GWAS summary statistics (IEU OpenGWAS/consortia)Causal microbiome–cancer associations (stringent and sensitivity analyses)
    B. dorei rises before T1D autoimmunity2014SupportsCohortLongitudinal DIPP stool sequencing in HLA-risk children before T1D autoimmunityN=76 · 947 stool samples; 29 seroconverters vs 47 autoantibody-negative controlsHLA-DQB1 moderate-to-high-risk children in DIPPGut microbiota (B. dorei dominance) prior to persistent T1D autoantibodies
    Gut microbes, blood metabolites, and metabolic traits2017SupportsCross-sectionalMETSIM subset microbiome–metabolite–metabolic trait association analysisN=531 · 531 middle-aged Finnish men from the METSIM cohort (parent cohort 10,197)Middle-aged Finnish men in METSIMAssociations of gut microbiota with plasma metabolites and metabolic traits
  • Scope / different questions

    Human observational and MR studies do not automatically generalise to bee antibiotic/pesticide disruptions or to enzyme-focused colonization experiments. Animal and insect systems test different host–microbe questions; they limit how far a human association travels rather than falsifying it.

    4 studies
    1. 1Antibiotics hurt bee gut and survival
    2. 2Pesticides reshape honey bee gut microbes
    3. 3Selective bacterial BSH shifts host metabolism
    4. 4Mendelian randomisation associates specific gut taxa with several cancers

    Study comparison

    StudyRoleDesignNPopulationOutcome
    Antibiotics hurt bee gut and survival2017SupportsAnimal / in-vitroWorker honeybees fed tetracycline; 16S community and survival vs germ-free controlsCup cages of 30 bees × 15 replicates per condition; 16S profiles n≈14–15 per arm — no single primary NWorker honeybees with conventional or germ-free gutsCore gut microbiota disruption and mortality after tetracycline
    Pesticides reshape honey bee gut microbes2016SupportsAnimal / in-vitroIn-hive pesticide exposures across sites; 16S and ITS pyrosequencing of foragers and broodN=24 · 24 sequenced samples (4 treatments × 3 sites × 2 bee types), each pooled from five beesHoney bees (Apis mellifera) exposed to in-hive pesticidesGut bacterial and fungal community composition under pesticide treatments
    Selective bacterial BSH shifts host metabolism2018SupportsAnimal / in-vitroBacteroidetes BSH selectivity screen plus B. thetaiotaomicron BT2086 deletion in gnotobiotic mice~20 Bacteroidetes strains screened; monocolonization used 12 mice per Bt WT/KO group (8 GF controls) — no single primary NHuman-gut Bacteroidetes strains and germ-free C57BL/6 miceHost metabolic phenotypes after selective bile-salt hydrolase loss
    Mendelian randomisation associates specific gut taxa with several cancers2023SupportsMendelian randomisationTwo-sample MR of MiBioGen gut taxa instruments against eight cancer GWAS summary sets211 taxa instruments; cancer GWAS sample sizes vary by malignancy — no single primary NGut microbiome and cancer GWAS summary statistics (IEU OpenGWAS/consortia)Causal microbiome–cancer associations (stringent and sensitivity analyses)

Common misconceptions

Exam-style questions

Short-answer questions that ask you to explain or compare, not recall.

A classmate says the MR cancer paper and the B. dorei T1D paper disagree about whether the microbiome is harmful. How do you respond?

They are not answering the same question. One reports genetic-instrument associations with specific cancers; the other reports a temporal dominance pattern before T1D autoimmunity. Different diseases and designs are a scope limit, not a forced disagreement about a single “harm” effect.

What would most strengthen a causal position that a named gut taxon increases a cancer’s risk?

Convergent evidence beyond MR assumptions—such as strain-resolved exposures, experimental models that recover the pathway, and replication outside the original ancestry GWAS—while keeping the cancer endpoint specific.

Why does strain-level gene-content variation qualify species-level microbiome claims?

People can share a species name while differing in gene content, so functions (for example enzyme capacity) may not travel with the species label alone.

The studies

11 studies in this library bear on Microbiome, ordered by citations. The first 8 are shown.

Show 3 more studies

Learn alongside