# HSEF v5.3 review: Gut Microbiome as Master Regulator of Health

> Sources: 60 primary reports; claims: 65; verified: 65; unverified: 0. Verification is bounded attribution. Access: 4 partial-body and 56 complete abstract/status examinations. Cutoff: 2026-10-04. No complete-paper audit or blinded calibration claimed.
>
> **Adversarial review:** Generated by Sol 6.1; adversarially reviewed by Claude Opus 5.5 (Anthropic) on 2026-10-06. 0 corrections applied (see ADVERSARIAL REVIEW AND APPROVAL). Status: **Approved as revised.**

## AT A GLANCE

**Defined microbial functions causally alter phenotypes and recurrent-CDI restoration is strongly supported; autonomous upstream control of systemic human health remains Γ.**

| Level | Confidence | Causal role | Momentum | Stability | Divergence |
|---|---|---|---|---|---|
| Γ — Mechanistically plausible, partial system support | Moderate | reinforcing / stabilizing factor | Mechanistic only | Stable | Contested |

**Deciding experiment:** Randomize a defined microbial metabolic function against an otherwise matched inactive consortium in predeclared human metabolic and psychiatric or oncological settings, factorially controlling substrate and usual treatment, with equal contact, blinded outcomes, engraftment and longitudinal mediator measurements. Pair human contrasts with microbial-gene deletion and matched metabolite rescue in conventional animals.

**Belief versus evidence:** The preregistered scope-matched SC4 gate returns Γ. Positive recurrent-CDI restoration reaches Ζ and selected UC packages establish bounded human causal benefit; neither demonstrates autonomous multi-system primacy. Weakest-link would return Β because SC1 deliberately asks an observational existence question.

**Best-supported rival:** Niche-depletion-specific causality. Factorially manipulate a defined microbial function and substrate supply with matched host strata; a substrate-dependent effect favors conditional regulation over autonomy. Ranking of a narrow CDI model is not a claim that CDI explains systemic health.

## COMPETING MECHANISTIC MODELS

**Rank 1: Niche-depletion-specific causality (Ζ).** Microbial restoration repairs a depleted colonization-resistance niche; level refers to recurrent CDI. SER-109, VE303 and strain rescue support this defined scope [1,3,59]. UC benefits show that this model is not an exhaustive account of all microbial effects [26–31].

**Rank 2: Diet/substrate regulation (Ε).** Diet changes substrate availability and host/microbial outputs. Dietary interventions and diet-dependent twin-community rescue demonstrate conditions [11,13,49,60]. A direct diet effect does not prove that microbes mediate every dietary health benefit.

**Rank 3: Host-genome/physiology gating (Ε).** Host receptors, immune cells and physiological niches determine microbial effects in models. RGMb, Tfh, eosinophil and CNS1 contrasts demonstrate host gating [25,51,53,58]. No single host pathway explains all human systemic conditions.

**Rank 4: Disease/medication-to-microbiome reverse causation (Β).** Disease, diet selection and drugs can create diagnostic signatures. Medication stratification and ASD dietary associations support this direction [43,44,46]. Observational direction remains imperfectly identified; reverse causation cannot explain controlled microbial rescue.

**Rank 5: Autonomous microbial master regulator (Γ).** One autonomous microbial system sits upstream of human systemic health. Transferability, human treatment benefits and defined pathways make microbial causation plausible. Diet-dependent rescue, pasteurized products, host epistasis, null primary endpoints and treatment dependence do not support autonomy [9,25,49,55–58].

The tournament does not alter the nominal level. Compatible pathways and flexible alternatives are not automatically proved by explaining observations.

## FULL REVIEW

### Registered claim and scope

The gut microbiome operates as an autonomous upstream master regulator of human health across metabolic, oncological, psychiatric and immunological systems.

SC4 matches the broad central claim; recipient-animal and depleted-CDI-niche results remain separate. SC3 wording tests causal restoration rather than incorrectly presuming every intact microbiome is necessary.

Selection: public-interest rank 2 from the editorial ranking compiled 2026-09-23; categorical prior verdicts were visible. Private fresh review; earlier documents are preserved.

### Evidence and causal interpretation

Microbiome causation is supported, but the registered claim adds autonomy and upstream primacy across human systems. Compositional classifiers cannot establish that extra claim. Population studies show that stool consistency, medications and diet explain substantial signatures; metformin stratification changes T2D interpretation, while ASD dietary preferences can account for a diagnostic association [42–46]. These are practical confounding and reverse-direction controls rather than evidence that every microbial association is false.

The animal evidence contains more than crude donor transfer. Bifidobacterium, inosine-associated isolates, microbial genes and host receptor or Tfh contrasts move functional phenotypes [22,23,25,47,53]. Butyrate/propionate with CNS1 dependence provides a defined host-microbe route [58]. Such epistasis tests fail differently from a compositional signature. They justify Ζ for the specified SC2 settings, without converting the same organisms or pathways into a general human master regulator.

The twin-community study is especially discriminating: cultured bacterial collections reproduce adiposity, reducing the possibility that a nonbacterial stool component explains the transfer. Selected body sections show that cohousing rescue depends on dietary composition; the unfavorable diet prevents invasion and rescue [49]. This is causal microbiota-by-diet interaction. Calling it autonomous control would discard the experiment's demonstrated boundary.

Human recurrent-CDI evidence is stronger than a biomarker story. SER-109 reduced randomized clinical recurrence and VE303 tested a defined consortium [1,3]. RBX reports support restoration but share phase2 information through Bayesian borrowing [4,8]. Capsule-versus-colonoscopy equivalence establishes delivery comparison, not efficacy against no restoration [6]. The live-versus-filtered stool trial supplies a component discriminator, yet filtration changes more than live-cell presence, so it cannot prove universal microbial necessity [7].

Buffie's selected body and methods distinguish the strain rescue from the mediation assay. C. scindens and a consortium reduce infection, weight loss and mortality in antibiotic-exposed mice. Bile-salt sequestration removes inhibition in intestinal contents ex vivo [59]. The mechanism uses host-derived bile salts, and the human cohort is observational. This complements clinical restoration while preserving the untested precise human mediator and the difference between ex vivo and in vivo necessity.

Randomized UC trials invalidate a categorical claim that causal human microbiome intervention evidence is confined to CDI. Multidonor, oral and anaerobic donor FMT improve prespecified clinical/endoscopic outcomes [26–28,31]. Donor, antibiotics, dose, baseline disease and maintenance influence the effect; the FMT-plus-diet trial bundles components [30]. These are bounded treatment effects, and crude-community manipulation does not by itself identify a strain, explain disease initiation or establish systemic autonomy.

Metabolic translation is heterogeneous. Pasteurized Akkermansia improves selected indices in a pilot without significant weight/fat changes [9], while the later T2D trial has null overall weight/HbA1c contrasts and baseline-abundance subgroup signals [14]. A membrane-protein effect can be real without establishing a live-community controller [10]. Diet interventions and brief FMT studies support candidate contexts but do not isolate durable human healthspan mediation [11–15].

Earlier melanoma FMT studies combine stool transfer with active PD-1 rechallenge and have no randomized no-FMT counterfactual [20,21]. The 2026 TACITO article explicitly identifies 12-month PFS as primary: its P=.053 misses the prespecified superiority boundary. Favorable secondary median PFS is retained as a signal, not relabeled primary success [55]. LUMINate's single-arm response signal also cannot isolate incremental FMT efficacy. Its controlled mouse add-back does establish treatment-context recipient causation, using distinct lost-species cocktails rather than a universal favorable taxon [56].

Psychiatric evidence similarly needs endpoint discipline. Model transfer and defined-species/metabolite rescue are informative about capacity [33,35,37,39,40,50]. Open-label ASD packages and short probiotic symptom trials do not identify durable clinical mediation [34,36,38,41]. The 2026 adjunct depression RCT reports improved selected symptom scores with no week-8 remission difference; statistical bile-acid mediation is not a selective human intervention [57].

The resulting Γ is therefore a score for SC4 autonomy, not a denial of microbial biology. SC2 and SC3 reach Ζ on convergent, distinct causal controls. A conventional-host, defined-function intervention with diet and treatment factors separated would address the next broad boundary. No deliberate human mechanism RCT is demanded where genetics, selective rescue or another defensible equivalent supplies that discrimination.

## HYPOTHESIS STATE EVALUATION FRAMEWORK (HSEF v5.3)

<!-- consensus-consistency-2026-10-09 -->
### Consensus provenance reconciliation — 9 October 2026

Case-specific Consensus discovery records were located and linked to this final edition. This supplements the review's provenance; it leaves the original assessed claim, grades and source counts intact. Provider summaries and full-text badges do not establish examination of original primary passages by this reviewer. Source access and incorporation retain the limits stated in the original methods and any dated research update.

Recorded evidence artifacts:

- `deliverables/hsef_v53_popularity_batch_2026-10-04/microbiome-master-regulator/consensus/evidence-cards.json`
- `deliverables/hsef_v53_popularity_batch_2026-10-04/microbiome-master-regulator/consensus/evidence-cards.md`
- `deliverables/hsef_v53_popularity_batch_2026-10-04/microbiome-master-regulator/discovery/consensus-targeted.json`

The located artifacts do not state a reliable source-comparison count. No count is inferred.

A dated adversarial approval record exists for the separately identified microbiome-master-regulator.review.md (2026-10-06; see edition-approval-audit.json). Its scope is that recorded edition. This supplement has no separate approval and no coverage is inferred from topic identity. No score or confidence change is made by this reconciliation.
<!-- end-consensus-consistency -->


### Subclaim matrix and aggregation

| Subclaim | Scope | Level | Decision |
|---|---|---|---|
| SC1 Microbiome composition differs reproducibly between health and disease across systemic conditions. | human observational disease cohorts | Β | SC1 is reproducible association. Medication, stool consistency, diet and disease-state adjustment change the signature; observational replication cannot identify causal direction [42–46]. |
| SC2 Defined microbial manipulation causally moves phenotype in rodents. | specified gnotobiotic and conventional rodent settings | Ζ | SC2 reaches Ζ within specified rodent contexts. Defined bacteria/metabolites, transfer and rescue converge with host receptor/enhancer or immune-cell epistasis; diet and species-negative controls map limits [22,23,25,47,49,53,58]. This is causal modulation, not endogenous human primacy. |
| SC3 Microbiota restoration causally restores colonization resistance in recurrent C. difficile infection. | humans with recurrent CDI following antibiotics | Ζ | SC3 reaches Ζ for recurrent post-antibiotic CDI restoration. Randomized spore/defined-consortium clinical outcomes, live-versus-filtered component discrimination, and a distinct strain/bile-acid rescue route converge [1,3,7,59]. The human precise mediator and universally required strain remain unresolved. |
| SC4 The microbiome is an autonomous upstream driver of systemic human health. | human metabolic, oncological, psychiatric and immunological systems | Γ | SC4 remains Γ for autonomous upstream multi-system primacy. UC treatment causality and context-specific model pathways are real, but no selective intervention demonstrates independence from substrate, host physiology or concurrent drugs across the claimed systems [9–15,26–32,55–60]. |

Preregistered aggregation: scope-matched. The preregistered scope-matched SC4 gate returns Γ. Positive recurrent-CDI restoration reaches Ζ and selected UC packages establish bounded human causal benefit; neither demonstrates autonomous multi-system primacy. Weakest-link would return Β because SC1 deliberately asks an observational existence question. Sensitivities: weakest-link Β; primary-subclaim Γ; scope-matched Γ.

### Decisive-perturbation inventory and causal tally

[1] Post-antibiotic SER-109 reduced recurrent CDI at 8 weeks in the randomized trial. Restoration in a depleted niche does not establish autonomous regulation of systemic health.

[2] Primary-CDI FMT met the specified noninferiority comparison against vancomycin. The 25-point margin and rescue therapy limit claims of superiority; this is primary CDI rather than the exact recurrent-CDI subclaim.

[3] A defined eight-strain VE303 consortium reduced recurrence at the high dose in the phase 2 trial. Small groups and recurrence definitions require confirmation; the exact mediator was not randomized.

[4] RBX2660 had higher modeled treatment success than placebo following antibiotics. The Bayesian estimate borrows the phase 2 study and is not an independent replication of that report; linked correction remains attached.

[5] Early FMT improved resolution after antibiotics in first or second CDI. Early stopping and disease-stage differences limit magnitude and recurrent-CDI generalization.

[6] Capsule FMT was noninferior to colonoscopic FMT for preventing recurrence. Both groups received FMT; the route contrast does not isolate efficacy against no FMT.

[7] Sterile filtrate failed noninferiority to live donor stool, with lower recurrence-free proportions. Filtration changes multiple components; failure of noninferiority is not proof that every live organism is necessary.

[8] The two-dose RBX2660 primary contrast did not meet significance; a one-dose secondary contrast favored treatment. Secondary and open-label rescue outcomes cannot replace the null primary comparison; this cohort informs the phase 3 Bayesian estimate.

[9] Pasteurized Akkermansia improved insulin sensitivity and some metabolic measures in a small human pilot. Weight and fat changes were not significant; a pasteurized product effect does not establish live-community autonomy.

[10] Pasteurized Akkermansia and its membrane protein partially reproduced metabolic benefits in mice. Human observations in this report concern safety; a protein/TLR2 pathway depends on host machinery.

[11] Controlled dietary restriction and timing changed weight and microbial measures. Diet was manipulated directly; parallel changes do not isolate microbial mediation.

[12] Transfer, ethanolamine metabolism and probiotic rescue linked microbial function to barrier and metabolic phenotypes. Model rescue supports a defined pathway, not autonomous human multi-system primacy.

[13] An isocaloric Mediterranean dietary intervention changed cholesterol and microbial/metabolic profiles. Direct dietary effects and correlated microbiome changes do not identify the mediator.

[14] The Akkermansia T2D trial had overall null between-group weight/HbA1c contrasts, with benefits in a low-baseline-abundance subgroup and accompanying mouse work. Subgroup findings require prospective confirmation; baseline community dependence contradicts simple universal response.

[15] A small short-term FMT/metformin comparison reported improvements in metabolic indices. Limited size, duration and comparator structure do not establish durable broad human health benefit.

[16] Responder transfer and Akkermansia rescue enhanced PD-1 effects in mice with an IL-12-dependent route. Human antibiotic and taxon associations remain confounded; active anticancer treatment and host immunity gate the effect.

[17] Melanoma responder communities were associated with response and transferred favorable immune/tumor phenotypes into mice. The human observational comparison does not establish the microbial cause of patient benefit.

[18] Dietary fiber associated with human progression-free survival; low fiber or probiotics impaired ICI response in models. The human diet comparison is observational and the model effect depends on diet and immunotherapy.

[19] Exercise-associated microbial formate was tested as a required/sufficient contributor to antitumor Tc1 phenotypes, with Nrf2 mediation. Human formate is a biomarker candidate; exercise and host transcriptional gating remain upstream conditions.

[20] Responder-derived FMT plus anti-PD-1 produced clinical benefit in six of fifteen refractory melanoma patients. Single-arm drug rechallenge does not isolate the incremental FMT effect or human mediator necessity.

[21] FMT with PD-1 reinduction produced responses in three of ten refractory melanoma patients. The phase 1 single-arm design establishes feasibility and signals, not a controlled causal efficacy estimate.

[22] Cohousing, transfer and Bifidobacterium administration improved mouse tumor immunity and checkpoint efficacy. Mouse facility/community and immune context limit human generalization.

[23] Defined bacterial isolates and inosine enhanced checkpoint efficacy in several mouse cancer models through a costimulation-dependent A2A receptor route. The metabolite requires treatment/barrier and host immune context; it is not an autonomous human master switch.

[24] Baseline melanoma taxa associated with response; responder-derived microbiota improved tumor control in mice. Human association and recipient sufficiency do not identify endogenous human necessity.

[25] PD-L2/RGMb blockade or conditional T-cell RGMb deletion overcame microbial-context ICI resistance in mouse tumors. The host immune intervention can bypass microbial resistance; correction-body details were not examined.

[26] Intensive multidonor FMT improved the prespecified steroid-free clinical/endoscopic outcome in UC versus placebo. Crude-community benefit supports a UC treatment package, not strain-specific or autonomous systemic causation.

[27] Oral FMT after antibiotics improved the week-8 UC outcome; a very small maintenance randomization favored continuation. Early pandemic termination, pretreatment and tiny maintenance groups limit generalization and mechanism.

[28] Weekly donor FMT improved remission versus water enema in active UC. Donor dependence and early futility stopping complicate precision; the result does not establish universal microbiome primacy.

[29] Low FODMAP diet improved reported adequate symptom relief while reducing some putatively beneficial taxa. Inflammatory markers and diversity did not differ; health cannot be inferred from a universal favorable-diversity metric.

[30] FMT plus anti-inflammatory diet improved UC outcomes versus optimized standard therapy. The bundled unblinded intervention and continued diet do not isolate FMT or a universal microbial mechanism.

[31] Anaerobic pooled donor FMT improved week-8 steroid-free UC remission versus autologous FMT. Most initial responders did not retain remission at twelve months; the mediator is not selectively identified.

[32] Catered diets improved quality of life in a small UC crossover with microbial/metabolite changes. Some inflammatory contrasts were only significant versus baseline; diet-microbiome mediation was not isolated.

[33] ASD donor transfer induced mouse behavioral features; candidate metabolites improved model phenotypes. Recipient mouse behaviors are not human ASD causation or remission, and donor phenotypes include diet-related differences.

[34] A small probiotic RCT improved selected flexibility/stress measures in healthy older adults. Multiple outcomes and short follow-up limit general clinical interpretation; BDNF correlation is not manipulated mediation.

[35] Depression-associated donor stool induced anhedonia/anxiety and metabolic features in depleted rats. Recipient effects establish capacity in a model, not primary causation of endogenous human depression.

[36] A bundled antibiotic/cleanse/FMT open-label ASD study reported persistent GI and behavioral improvement. Eighteen participants without randomized masking cannot separate expectation, co-interventions or natural trajectory.

[37] Multi-cohort proline associations were combined with supplementation, transfer and transporter/RNAi or bacterial manipulations in models. Human dietary and depression relationships remain observational; cross-species behaviors do not establish clinical efficacy.

[38] A short probiotic/placebo comparison in test-anxious students reported improved symptom questionnaires. Short, selected student symptoms are not durable clinical psychiatric remission or established microbial mediation.

[39] Selected Bacteroides species reproduced behavioral/neurogenesis effects in antibiotic-treated mice whereas another species did not. Species and host-metabolite specificity constrain the result; the exact human primary pathway is not established.

[40] Human adolescent signatures and Roseburia/healthy-transfer rescue linked tryptophan pathways to stressed-mouse phenotypes. Sertraline alters human signatures; model rescue is not a randomized human mediator test.

[47] A targeted commensal bile-acid gene deletion in a complex community altered host colon inflammation. Gene-level specificity is stronger than dysbiosis association but does not establish a universal human regulator.

[49] Obese/lean twin communities and cultured collections transferred adiposity; cohousing rescue succeeded or failed depending on diet. Food intake contrasts do not explain adiposity here, but dietary dependence directly limits microbial autonomy; mild glucose changes are not established insulin resistance.

[50] B. fragilis improved barrier and selected behaviors in maternal-immune-activation offspring; a candidate metabolite induced behavioral abnormalities. Developmental immune injury precedes the microbial intervention; model modulation does not establish human ASD initiation.

[51] Eosinophil-deficient/colonization contrasts linked microbial IL-33 responses to gut physiological homeostasis. Host immune cells gate the microbial effect; the pathway is mutualistic rather than autonomous.

[52] C. scindens or bile-acid administration altered metabolic/obesity phenotypes in mice alongside human associations. Host bile-acid substrate and diet are required context; correction details remain unexamined.

[53] H. hepaticus-induced tumor control depended on Tfh cells, with host-cell rescue in a deficient mouse setting. Host immune epistasis supplies a specific causal route; it does not prove broad human cancer benefit.

[54] Antibiotic, mouse genetic and Lactobacillus/macrophage rescue contrasts linked microbes to postnatal niches and NEC severity. Developmental host-cell dependence and antibiotic pleiotropy limit autonomy; erratum-body details were not examined.

[55] TACITO primary 12-month PFS was 70% versus 41%, P=.053; secondary median PFS favored FMT. The primary superiority threshold was not met; secondary nominal signals, small groups and concurrent cancer drugs do not prove autonomous regulation.

[56] LUMINate reported single-arm clinical response signals; reintroducing lost baseline bacterial cocktails impaired ICI efficacy in mouse recipients. Clinical FMT benefit lacks a randomized counterfactual; the controlled mouse add-back establishes only recipient/treatment-context mediation.

[57] Adjunct FMT improved selected depression scores but not week-8 remission versus placebo with escitalopram. Statistical bile-acid mediation is not selective causal manipulation; durable remission and autonomous monotherapy are not shown.

[58] Butyrate/propionate promoted peripheral Treg generation with CNS1 dependence; acetate lacked the corresponding activity. Defined metabolite/host enhancer logic supports communication, not a universal human master regulator.

[59] C. scindens/consortium transfer reduced CDI illness in mice; bile-salt sequestration neutralized inhibition ex vivo. The human cohort is observational and the bile-acid blockade is ex vivo, not in vivo gene-specific necessity or human mediation.

[60] Short animal/plant diets rapidly altered human microbial structure and expression. This demonstrates diet-to-microbiome direction; downstream health mediation was not isolated.

Tally: 53 discriminating contrasts. Null and negative contrasts are discriminators, not positive threshold crossings. Reports and replications are not independent causal lines. A decisive flag denotes an intervention discriminator, including dietary and negative contrasts; it does not count positive nominal autonomy tests. CRC methylation report [48] contributes original observational measurements only; its meta-analysis is not an extra primary study.

### Cap audit, precedence and Β/Γ/Δ boundary

The microbiome rubric permits transfer to cross Δ in recipient animals and defined rescue/epistasis to support Ζ. Universal forced-transfer limits govern inference to endogenous human primacy, not the actual model result. Human randomized clinical recurrence and UC functional outcomes are assessed separately from biomarker associations. Convergent restoration evidence is not re-capped as merely crude FMT.

Gamma hard-rule audit: “Association plus cell-culture metabolite assays” and “Association plus in vitro immune readouts” constrain isolated associative inferences. “Any framework that has not crossed the transfer/rescue threshold in vivo” does not describe the complete corpus; SC2 crosses it. Forced transfer alone remains insufficient for endogenous human primacy.

SC1 cap: “Pure association evidence — however large, however replicated — cannot exceed Β.” SC1 is reproducible association. Medication, stool consistency, diet and disease-state adjustment change the signature; observational replication cannot identify causal direction [42–46].

SC4 cap: “Association plus mechanistic in vitro support, or host-pathway plausibility, caps at Γ.” SC4 remains Γ for autonomous upstream multi-system primacy. UC treatment causality and context-specific model pathways are real, but no selective intervention demonstrates independence from substrate, host physiology or concurrent drugs across the claimed systems [9–15,26–32,55–60].

SC1 field-valid definition: “
- Gnotobiotic or transfer experiment showing phenotype movement, at minimum
- Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per **Cap precedence** above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: no. SC1 is reproducible association. Medication, stool consistency, diet and disease-state adjustment change the signature; observational replication cannot identify causal direction [42–46].

SC2 field-valid definition: “
- Gnotobiotic or transfer experiment showing phenotype movement, at minimum
- Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per **Cap precedence** above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: yes. SC2 reaches Ζ within specified rodent contexts. Defined bacteria/metabolites, transfer and rescue converge with host receptor/enhancer or immune-cell epistasis; diet and species-negative controls map limits [22,23,25,47,49,53,58]. This is causal modulation, not endogenous human primacy.

SC3 field-valid definition: “
- Gnotobiotic or transfer experiment showing phenotype movement, at minimum
- Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per **Cap precedence** above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: yes. SC3 reaches Ζ for recurrent post-antibiotic CDI restoration. Randomized spore/defined-consortium clinical outcomes, live-versus-filtered component discrimination, and a distinct strain/bile-acid rescue route converge [1,3,7,59]. The human precise mediator and universally required strain remain unresolved.

SC4 field-valid definition: “
- Gnotobiotic or transfer experiment showing phenotype movement, at minimum
- Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per **Cap precedence** above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: no. SC4 remains Γ for autonomous upstream multi-system primacy. UC treatment causality and context-specific model pathways are real, but no selective intervention demonstrates independence from substrate, host physiology or concurrent drugs across the claimed systems [9–15,26–32,55–60].

### Upward threshold ladder

SC1 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: no. A prospective selectively microbial intervention linked to the same functional disease phenotype, beyond composition association.

SC2 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: yes. 

SC2 Ε: “Requires context-specific causal effect with defined limits. A transferable phenotype exists, but microbial identity or generalizability remains incomplete.” Satisfied: yes. 

SC2 Ζ: “Requires:
- Defined consortium or strain-level causal evidence
- Rescue logic
- Replication across at least two relevant models or cohorts” Satisfied: yes. 

SC2 Η: “Requires:
- Defined consortium or factor rescue in multiple contexts
- Narrowed mechanism
- Strong exclusion of diet, housing, batch, and handling confounds” Satisfied: no. Repeat narrowed causal routes across independent facilities with diet, handling, cage and host genetic controls sufficient for Η.

SC3 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: yes. 

SC3 Ε: “Requires context-specific causal effect with defined limits. A transferable phenotype exists, but microbial identity or generalizability remains incomplete.” Satisfied: yes. 

SC3 Ζ: “Requires:
- Defined consortium or strain-level causal evidence
- Rescue logic
- Replication across at least two relevant models or cohorts” Satisfied: yes. 

SC3 Η: “Requires:
- Defined consortium or factor rescue in multiple contexts
- Narrowed mechanism
- Strong exclusion of diet, housing, batch, and handling confounds” Satisfied: no. Selective human mediator manipulation or an independently discriminating defined-factor rescue with matched delivery and substrate controls for Η.

SC4 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: no. Demonstrate a positive selectively microbial human clinical perturbation identifying autonomous upstream action in the registered systemic scope.

Upper ladder: Randomized clinical recurrence, UC remission, live/filter component contrast, host epistasis and strain/metabolite rescue are eligible causal equivalents. Together they carry narrow SC2/SC3 to Ζ. Missing precise human mediation and matched cross-context autonomy keep the nominal claim below Ζ; no blanket RCT-only requirement or weakest-leg re-cap is imposed.

Independence: At Ζ, independence is assigned to distinct failure modes, not paper counts. Human product trials share modality; RBX2b is borrowed by phase3 and is one family. LUMINate reanalyses Davar/Baruch and counts no new independent patient cohorts. Clinical controls, defined microbial precision and host/biochemical epistasis cover different weaknesses; no one expectation, crude-matrix or sequencing error explains them all. Cross-system success remains narrower than autonomous human primacy.

SC3 defined microbial clinical product [1, 3]: Defined/spore randomized recurrence benefit reduces expectation/natural-resolution explanations and crude nonmicrobial donor-matrix confounding; the two products remain one clinical intervention family, not two orthogonal lines.

SC3 live-versus-filter component discriminator [7]: A randomized contrast changes the live component while retaining donor-derived material; it challenges a purely soluble-product explanation. Processing also changes components, so it is supporting discrimination rather than absolute necessity.

SC2/SC3 strain and biochemical rescue [59]: Controlled specific-strain transfer with illness and ex vivo bile-salt sequestration addresses taxonomic-signature correlation and supplies a pathway distinct from human expectation. Ex vivo mediation is not human in vivo necessity.

SC2 host genetic/immune epistasis [25, 53, 58]: Host receptor, Tfh and enhancer perturbations distinguish an actual host-microbe route from merely transferring stool contents or recipient housing artifacts.

SC2 cultured community and diet discriminator [49]: Cultured collections recapitulate adiposity and diet-switch rescue maps substrate dependence, addressing nonbacterial donor components and simple food-intake confounding.

### Falsification and pitfalls

Host phenotype persists after removing the claimed microbial driver. Host immune blockade can bypass microbiome-linked resistance [25]; dietary and drug context are material. This limits universal necessity without denying causal routes. Additional downgrade: False.

Rescue fails despite reinstating the proposed taxa. Community rescue depends on diet and colonization [49]; this is a mapped limit already reflected in the nominal scope score. Additional downgrade: False.

1. Correlation promoted to mechanism. Dysbiosis, engraftment and statistical mediation are distinguished from selective causal mediation.

13. Therapeutic effect misread as mechanistic proof. FMT packages and concurrent ICI/escitalopram are not proof of autonomous microbial primacy.

16. Untested generalization confused with mapped boundaries. Diet/species-negative controls define Ε or higher within actually tested rodent scopes.

22. Scope fitted to the result. Baseline abundance is a possible predictive context; post-treatment responders do not define the claim.

### Adversarial steelman: above, then below

The strongest Δ case is that several human randomized microbial interventions produce clinical benefit and model routes are unusually specific. This is persuasive for microbial contribution. It does not identify the extra autonomous upstream action across all four human systems, which is the registered SC4 target.

The strongest Β case is that most human multi-system signatures are observational and confounded. It understates demonstrated model pathways, clinical recurrence restoration and UC treatment contrasts. Those findings provide biological causal support beyond a signature-only hypothesis, while leaving the nominal autonomous scope unresolved.

Reject nominal Δ because the selectively identifying full-scope autonomy experiment is missing; reject Β because verified clinical/model contributions supply mechanistic support. Narrow SC2/SC3 remain Ζ rather than being dragged down by SC4.

### Final score, causal direction and M/S/D

The preregistered scope-matched SC4 gate returns Γ. Positive recurrent-CDI restoration reaches Ζ and selected UC packages establish bounded human causal benefit; neither demonstrates autonomous multi-system primacy. Weakest-link would return Β because SC1 deliberately asks an observational existence question.

Apply causal-direction steps in order: some signatures are downstream, but controlled microbial effects preclude classifying the whole mechanism as downstream (step 1) or epiphenomenal (step 2). Autonomous initiation and rate limitation are not demonstrated (step 3); a uniformly necessary prerequisite is not established across systems (step 4). Microbial functions amplify or stabilize host, diet and treatment-dependent phenotypes, matching step 5. This broad classification does not erase a narrower primary causal role in restoration of depleted CDI resistance.

Momentum: Mechanistic only. Named causal developments include microbial single-gene interrogation [47], exercise/formate/Nrf2 experiments [19] and LUMINate bacterial add-back [56]. Completed TACITO and depression trials are outcomes, not work still underway; none selectively tests the next autonomy boundary.

Stability: Stable. No specifically verified pending trial selectively identifying autonomous multi-system primacy was found in this bounded search. Completed 2026 trials are not a pending stability trigger.

Divergence: Contested; no single implied-level arithmetic. These dimensions do not change level or confidence.

## PROPOSED DECISIVE EXPERIMENTS

Randomize a defined microbial metabolic function against an otherwise matched inactive consortium in predeclared human metabolic and psychiatric or oncological settings, factorially controlling substrate and usual treatment, with equal contact, blinded outcomes, engraftment and longitudinal mediator measurements. Pair human contrasts with microbial-gene deletion and matched metabolite rescue in conventional animals.

Readout: Prespecified clinical phenotype or durable remission and adverse effects, alongside function-specific metabolite flux. Controls: Identical donor matrix/delivery and antibiotic exposure; baseline diet, medication and host strata; active versus function-inactive consortium. Supporting: Reproducible clinical and mediator effects attributable to the defined microbial function across the prespecified systems, including substrate-independent contrasts required for autonomy. Falsifying: Adequate function engagement without clinical change, benefits attributable to host/substrate treatment, or failure of the matched rescue.

## KNOWLEDGE GAPS

Precisely which live factors are necessary for human recurrent-CDI restoration?

Which UC treatment components cause sustained benefit and how durable is the effect?

Are the newer cancer and depression signals replicated on primary endpoints with clinically meaningful durability?

Do baseline abundance predictors prospectively identify treatment response rather than post hoc subgroups?

Can human selective mediation distinguish metabolite products, ecological remodeling and host immune effects?

## REVIEWER NOTES

Sixty primary reports were examined through complete abstract/status records; four selected primary-body examinations cover strain/bile-acid rescue, cultured twin communities, TACITO endpoint/statistical specification and LUMINate add-back. The older index claim that human causal support exists only in CDI is superseded by randomized UC outcomes. Correction notices remain attached for RBX2660, PD-L2/RGMb, IBD profiles, ASD diet, metformin, bile-acid obesity and postnatal niches; their detailed body impact was not fully examined, so corrected fine numerical claims are not used for level assignment. Publication versus online-year differences in Consensus are not new studies.

Search leads, reviews, protocols and correction notices do not pad the primary count. Downloaded XML does not earn examination credit. Linked notices remain attached; unexamined correction details limit inference. This is a coordinator source/claim consistency assessment, not independent blinded calibration.

## ADVERSARIAL REVIEW AND APPROVAL

**Authoring model:** Sol 6.1. **Adversarial reviewer:** Claude Opus 5.5 (Anthropic). **Date:** 2026-10-06.

**Scope of checks:** All 60 reference PMIDs re-fetched with the PubMed MCP tool get_article_metadata and cross-checked against NCBI E-utilities (title, first author, journal, year, DOI): 60/60 correct; no retraction or expression of concern; 7 linked errata, all disclosed on the cards. 19 load-bearing claims spot-checked against PubMed abstracts: refs 1, 3, 4, 6, 7, 9, 14, 20, 26, 28, 31, 44, 46, 49, 55, 56, 57, 58, 59. Scoring logic checked: scope-matched SC4 gate, reported alternatives (weakest-link Β from SC1), cap 2 on SC1, cap-precedence handling for the rodent-scope SC2, and the SC2/SC3 independence basis.

**Findings and corrections**

| ID | Severity | Location | Original → corrected | Evidence |
|---|---|---|---|---|
| — | — | — | No correction required | — |

**Rejected findings (checked, no change)**

- SC2 Ζ for rodent manipulation versus universal cap 5 (forced transfer): under the v4.0 cap-precedence rule a perturbation crosses a threshold within the system it was run in, and the cap governs the inference to human scope, which is where the review applies it (SC4). Correct as written.
- SC3 Ζ for recurrent-CDI restoration: the SER-109 phase 3, VE303 phase 2, the live-versus-filtrate non-inferiority failure and the C. scindens mouse rescue are distinct causal lines, so Ζ is defensible.
- Oncology FMT statements checked against the abstracts: the 2026 renal-cell trial's primary 12-month PFS did not meet its boundary while median PFS favoured FMT, and the lung/melanoma trial is single-arm with a controlled mouse add-back. The review states both accurately.
- The 2026 adjunct-FMT depression trial: no week-8 remission difference, symptom-score improvement, bile-acid mediation analysis. Matches the abstract.
- Akkermansia: pilot insulin-sensitivity benefit without significant weight/fat change, and the type 2 diabetes trial null overall with a low-baseline-abundance subgroup. Matches both abstracts.
- Year differences for refs 10, 15, 16, 17 (print vs e-pub, one year apart): acceptable under convention. PubMed title fields that drop italicised species names (refs 1 and 38) are rendering artifacts, not citation errors.

**Residual limitations**

- Verification was mostly at abstract depth; the 4 receipted partial-body passages were not re-read here.
- Scorecard provenance_header.generated_by reads 'Codex; HSEF v5.3' and was not altered.

**Approval:** Following adversarial review and the corrections listed above, the revised document is approved. — Claude Opus 5.5 (Anthropic), 2026-10-06.

## REFERENCES

1. Feuerstadt P, Louie TJ, Lashner B et al. SER-109, an Oral Microbiome Therapy for Recurrent Clostridioides difficile Infection. The New England journal of medicine. 2022. [PMID 35045228](https://pubmed.ncbi.nlm.nih.gov/35045228/). DOI: 10.1056/NEJMoa2106516. Access: abstract_metadata. Study: NCT03183128.

2. Juul FE, Bretthauer M, Johnsen PH et al. Fecal Microbiota Transplantation Versus Vancomycin for Primary Clostridioides difficile Infection : A Randomized Controlled Trial. Annals of internal medicine. 2025. [PMID 40523286](https://pubmed.ncbi.nlm.nih.gov/40523286/). DOI: 10.7326/ANNALS-24-03285. Access: abstract_metadata. Study: PMID40523286.

3. Louie T, Golan Y, Khanna S et al. VE303, a Defined Bacterial Consortium, for Prevention of Recurrent Clostridioides difficile Infection: A Randomized Clinical Trial. JAMA. 2023. [PMID 37060545](https://pubmed.ncbi.nlm.nih.gov/37060545/). DOI: 10.1001/jama.2023.4314. Access: abstract_metadata. Study: NCT03788434.

4. Khanna S, Assi M, Lee C et al. Efficacy and Safety of RBX2660 in PUNCH CD3, a Phase III, Randomized, Double-Blind, Placebo-Controlled Trial with a Bayesian Primary Analysis for the Prevention of Recurrent Clostridioides difficile Infection. Drugs. 2022. [PMID 36287379](https://pubmed.ncbi.nlm.nih.gov/36287379/). DOI: 10.1007/s40265-022-01797-x. Access: abstract_metadata. Study: RBX-phase2b+PUNCH-CD3-Bayesian.

5. Baunwall SMD, Andreasen SE, Hansen MM et al. Faecal microbiota transplantation for first or second Clostridioides difficile infection (EarlyFMT): a randomised, double-blind, placebo-controlled trial. The lancet. Gastroenterology & hepatology. 2022. [PMID 36152636](https://pubmed.ncbi.nlm.nih.gov/36152636/). DOI: 10.1016/S2468-1253(22)00276-X. Access: abstract_metadata. Study: PMID36152636.

6. Kao D, Roach B, Silva M et al. Effect of Oral Capsule- vs Colonoscopy-Delivered Fecal Microbiota Transplantation on Recurrent Clostridium difficile Infection: A Randomized Clinical Trial. JAMA. 2017. [PMID 29183074](https://pubmed.ncbi.nlm.nih.gov/29183074/). DOI: 10.1001/jama.2017.17077. Access: abstract_metadata. Study: PMID29183074.

7. Kao D, Wong K, Lee C et al. Effects of lyophilised faecal filtrate compared with lyophilised donor stool on Clostridioides difficile recurrence: a multicentre, randomised, double-blinded, non-inferiority trial. The lancet. Gastroenterology & hepatology. 2025. [PMID 40997843](https://pubmed.ncbi.nlm.nih.gov/40997843/). DOI: 10.1016/S2468-1253(25)00190-6. Access: abstract_metadata. Study: PMID40997843.

8. Dubberke ER, Lee CH, Orenstein R et al. Results From a Randomized, Placebo-Controlled Clinical Trial of a RBX2660-A Microbiota-Based Drug for the Prevention of Recurrent Clostridium difficile Infection. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2018. [PMID 29617739](https://pubmed.ncbi.nlm.nih.gov/29617739/). DOI: 10.1093/cid/ciy259. Access: abstract_metadata. Study: RBX-phase2b+PUNCH-CD3-Bayesian.

9. Depommier C, Everard A, Druart C et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study. Nature medicine. 2019. [PMID 31263284](https://pubmed.ncbi.nlm.nih.gov/31263284/). DOI: 10.1038/s41591-019-0495-2. Access: abstract_metadata. Study: PMID31263284.

10. Plovier H, Everard A, Druart C et al. A purified membrane protein from Akkermansia muciniphila or the pasteurized bacterium improves metabolism in obese and diabetic mice. Nature medicine. 2017. [PMID 27892954](https://pubmed.ncbi.nlm.nih.gov/27892954/). DOI: 10.1038/nm.4236. Access: abstract_metadata. Study: PMID27892954.

11. Li L, Li R, Tian Q et al. Effects of healthy low-carbohydrate diet and time-restricted eating on weight and gut microbiome in adults with overweight or obesity: Feeding RCT. Cell reports. Medicine. 2024. [PMID 39454570](https://pubmed.ncbi.nlm.nih.gov/39454570/). DOI: 10.1016/j.xcrm.2024.101801. Access: abstract_metadata. Study: PMID39454570.

12. Mishra SP, Wang B, Jain S et al. A mechanism by which gut microbiota elevates permeability and inflammation in obese/diabetic mice and human gut. Gut. 2023. [PMID 36948576](https://pubmed.ncbi.nlm.nih.gov/36948576/). DOI: 10.1136/gutjnl-2022-327365. Access: abstract_metadata. Study: PMID36948576.

13. Meslier V, Laiola M, Roager HM et al. Mediterranean diet intervention in overweight and obese subjects lowers plasma cholesterol and causes changes in the gut microbiome and metabolome independently of energy intake. Gut. 2020. [PMID 32075887](https://pubmed.ncbi.nlm.nih.gov/32075887/). DOI: 10.1136/gutjnl-2019-320438. Access: abstract_metadata. Study: PMID32075887.

14. Zhang Y, Liu R, Chen Y et al. Akkermansia muciniphila supplementation in patients with overweight/obese type 2 diabetes: Efficacy depends on its baseline levels in the gut. Cell metabolism. 2025. [PMID 39879980](https://pubmed.ncbi.nlm.nih.gov/39879980/). DOI: 10.1016/j.cmet.2024.12.010. Access: abstract_metadata. Study: PMID39879980.

15. Wu Z, Zhang B, Chen F et al. Fecal microbiota transplantation reverses insulin resistance in type 2 diabetes: A randomized, controlled, prospective study. Frontiers in cellular and infection microbiology. 2022. [PMID 36704100](https://pubmed.ncbi.nlm.nih.gov/36704100/). DOI: 10.3389/fcimb.2022.1089991. Access: abstract_metadata. Study: PMID36704100.

16. Routy B, Le Chatelier E, Derosa L et al. Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors. Science (New York, N.Y.). 2018. [PMID 29097494](https://pubmed.ncbi.nlm.nih.gov/29097494/). DOI: 10.1126/science.aan3706. Access: abstract_metadata. Study: PMID29097494.

17. Gopalakrishnan V, Spencer CN, Nezi L et al. Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients. Science (New York, N.Y.). 2018. [PMID 29097493](https://pubmed.ncbi.nlm.nih.gov/29097493/). DOI: 10.1126/science.aan4236. Access: abstract_metadata. Study: PMID29097493.

18. Spencer CN, McQuade JL, Gopalakrishnan V et al. Dietary fiber and probiotics influence the gut microbiome and melanoma immunotherapy response. Science (New York, N.Y.). 2021. [PMID 34941392](https://pubmed.ncbi.nlm.nih.gov/34941392/). DOI: 10.1126/science.aaz7015. Access: abstract_metadata. Study: PMID34941392.

19. Phelps CM, Willis NB, Duan T et al. Exercise-induced microbiota metabolite enhances CD8 T cell antitumor immunity promoting immunotherapy efficacy. Cell. 2025. [PMID 40639377](https://pubmed.ncbi.nlm.nih.gov/40639377/). DOI: 10.1016/j.cell.2025.06.018. Access: abstract_metadata. Study: PMID40639377.

20. Davar D, Dzutsev AK, McCulloch JA et al. Fecal microbiota transplant overcomes resistance to anti-PD-1 therapy in melanoma patients. Science (New York, N.Y.). 2021. [PMID 33542131](https://pubmed.ncbi.nlm.nih.gov/33542131/). DOI: 10.1126/science.abf3363. Access: abstract_metadata. Study: Davar-FMT-antiPD1.

21. Baruch EN, Youngster I, Ben-Betzalel G et al. Fecal microbiota transplant promotes response in immunotherapy-refractory melanoma patients. Science (New York, N.Y.). 2021. [PMID 33303685](https://pubmed.ncbi.nlm.nih.gov/33303685/). DOI: 10.1126/science.abb5920. Access: abstract_metadata. Study: Baruch-FMT-antiPD1.

22. Sivan A, Corrales L, Hubert N et al. Commensal Bifidobacterium promotes antitumor immunity and facilitates anti-PD-L1 efficacy. Science (New York, N.Y.). 2015. [PMID 26541606](https://pubmed.ncbi.nlm.nih.gov/26541606/). DOI: 10.1126/science.aac4255. Access: abstract_metadata. Study: PMID26541606.

23. Mager LF, Burkhard R, Pett N et al. Microbiome-derived inosine modulates response to checkpoint inhibitor immunotherapy. Science (New York, N.Y.). 2020. [PMID 32792462](https://pubmed.ncbi.nlm.nih.gov/32792462/). DOI: 10.1126/science.abc3421. Access: abstract_metadata. Study: PMID32792462.

24. Matson V, Fessler J, Bao R et al. The commensal microbiome is associated with anti-PD-1 efficacy in metastatic melanoma patients. Science (New York, N.Y.). 2018. [PMID 29302014](https://pubmed.ncbi.nlm.nih.gov/29302014/). DOI: 10.1126/science.aao3290. Access: abstract_metadata. Study: PMID29302014.

25. Park JS, Gazzaniga FS, Wu M et al. Targeting PD-L2-RGMb overcomes microbiome-related immunotherapy resistance. Nature. 2023. [PMID 37138075](https://pubmed.ncbi.nlm.nih.gov/37138075/). DOI: 10.1038/s41586-023-06026-3. Access: abstract_metadata. Study: PMID37138075.

26. Paramsothy S, Kamm MA, Kaakoush NO et al. Multidonor intensive faecal microbiota transplantation for active ulcerative colitis: a randomised placebo-controlled trial. Lancet (London, England). 2017. [PMID 28214091](https://pubmed.ncbi.nlm.nih.gov/28214091/). DOI: 10.1016/S0140-6736(17)30182-4. Access: abstract_metadata. Study: PMID28214091.

27. Haifer C, Paramsothy S, Kaakoush NO et al. Lyophilised oral faecal microbiota transplantation for ulcerative colitis (LOTUS): a randomised, double-blind, placebo-controlled trial. The lancet. Gastroenterology & hepatology. 2022. [PMID 34863330](https://pubmed.ncbi.nlm.nih.gov/34863330/). DOI: 10.1016/S2468-1253(21)00400-3. Access: abstract_metadata. Study: PMID34863330.

28. Moayyedi P, Surette MG, Kim PT et al. Fecal Microbiota Transplantation Induces Remission in Patients With Active Ulcerative Colitis in a Randomized Controlled Trial. Gastroenterology. 2015. [PMID 25857665](https://pubmed.ncbi.nlm.nih.gov/25857665/). DOI: 10.1053/j.gastro.2015.04.001. Access: abstract_metadata. Study: PMID25857665.

29. Cox SR, Lindsay JO, Fromentin S et al. Effects of Low FODMAP Diet on Symptoms, Fecal Microbiome, and Markers of Inflammation in Patients With Quiescent Inflammatory Bowel Disease in a Randomized Trial. Gastroenterology. 2020. [PMID 31586453](https://pubmed.ncbi.nlm.nih.gov/31586453/). DOI: 10.1053/j.gastro.2019.09.024. Access: abstract_metadata. Study: PMID31586453.

30. Kedia S, Virmani S, K Vuyyuru S et al. Faecal microbiota transplantation with anti-inflammatory diet (FMT-AID) followed by anti-inflammatory diet alone is effective in inducing and maintaining remission over 1 year in mild to moderate ulcerative colitis: a randomised controlled trial. Gut. 2022. [PMID 35973787](https://pubmed.ncbi.nlm.nih.gov/35973787/). DOI: 10.1136/gutjnl-2022-327811. Access: abstract_metadata. Study: PMID35973787.

31. Costello SP, Hughes PA, Waters O et al. Effect of Fecal Microbiota Transplantation on 8-Week Remission in Patients With Ulcerative Colitis: A Randomized Clinical Trial. JAMA. 2019. [PMID 30644982](https://pubmed.ncbi.nlm.nih.gov/30644982/). DOI: 10.1001/jama.2018.20046. Access: abstract_metadata. Study: PMID30644982.

32. Fritsch J, Garces L, Quintero MA et al. Low-Fat, High-Fiber Diet Reduces Markers of Inflammation and Dysbiosis and Improves Quality of Life in Patients With Ulcerative Colitis. Clinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association. 2021. [PMID 32445952](https://pubmed.ncbi.nlm.nih.gov/32445952/). DOI: 10.1016/j.cgh.2020.05.026. Access: abstract_metadata. Study: PMID32445952.

33. Sharon G, Cruz NJ, Kang DW et al. Human Gut Microbiota from Autism Spectrum Disorder Promote Behavioral Symptoms in Mice. Cell. 2019. [PMID 31150625](https://pubmed.ncbi.nlm.nih.gov/31150625/). DOI: 10.1016/j.cell.2019.05.004. Access: abstract_metadata. Study: PMID31150625.

34. Kim CS, Cha L, Sim M et al. Probiotic Supplementation Improves Cognitive Function and Mood with Changes in Gut Microbiota in Community-Dwelling Older Adults: A Randomized, Double-Blind, Placebo-Controlled, Multicenter Trial. The journals of gerontology. Series A, Biological sciences and medical sciences. 2021. [PMID 32300799](https://pubmed.ncbi.nlm.nih.gov/32300799/). DOI: 10.1093/gerona/glaa090. Access: abstract_metadata. Study: PMID32300799.

35. Kelly JR, Borre Y, O' Brien C et al. Transferring the blues: Depression-associated gut microbiota induces neurobehavioural changes in the rat. Journal of psychiatric research. 2016. [PMID 27491067](https://pubmed.ncbi.nlm.nih.gov/27491067/). DOI: 10.1016/j.jpsychires.2016.07.019. Access: abstract_metadata. Study: PMID27491067.

36. Kang DW, Adams JB, Gregory AC et al. Microbiota Transfer Therapy alters gut ecosystem and improves gastrointestinal and autism symptoms: an open-label study. Microbiome. 2017. [PMID 28122648](https://pubmed.ncbi.nlm.nih.gov/28122648/). DOI: 10.1186/s40168-016-0225-7. Access: abstract_metadata. Study: PMID28122648.

37. Mayneris-Perxachs J, Castells-Nobau A, Arnoriaga-Rodríguez M et al. Microbiota alterations in proline metabolism impact depression. Cell metabolism. 2022. [PMID 35508109](https://pubmed.ncbi.nlm.nih.gov/35508109/). DOI: 10.1016/j.cmet.2022.04.001. Access: abstract_metadata. Study: PMID35508109.

38. Zhu R, Fang Y, Li H et al. Psychobiotic Lactobacillus plantarum JYLP-326 relieves anxiety, depression, and insomnia symptoms in test anxious college via modulating the gut microbiota and its metabolism. Frontiers in immunology. 2023. [PMID 37033942](https://pubmed.ncbi.nlm.nih.gov/37033942/). DOI: 10.3389/fimmu.2023.1158137. Access: abstract_metadata. Study: PMID37033942.

39. Zhang Y, Fan Q, Hou Y et al. Bacteroides species differentially modulate depression-like behavior via gut-brain metabolic signaling. Brain, behavior, and immunity. 2022. [PMID 35143877](https://pubmed.ncbi.nlm.nih.gov/35143877/). DOI: 10.1016/j.bbi.2022.02.007. Access: abstract_metadata. Study: PMID35143877.

40. Zhou M, Fan Y, Xu L et al. Microbiome and tryptophan metabolomics analysis in adolescent depression: roles of the gut microbiota in the regulation of tryptophan-derived neurotransmitters and behaviors in human and mice. Microbiome. 2023. [PMID 37386523](https://pubmed.ncbi.nlm.nih.gov/37386523/). DOI: 10.1186/s40168-023-01589-9. Access: abstract_metadata. Study: PMID37386523.

41. Chen Q, Wu C, Xu J et al. Donor-recipient intermicrobial interactions impact transfer of subspecies and fecal microbiota transplantation outcome. Cell host & microbe. 2024. [PMID 38367621](https://pubmed.ncbi.nlm.nih.gov/38367621/). DOI: 10.1016/j.chom.2024.01.013. Access: abstract_metadata. Study: ASD-2100043906+external-FMT-reanalysis.

42. Franzosa EA, Sirota-Madi A, Avila-Pacheco J et al. Gut microbiome structure and metabolic activity in inflammatory bowel disease. Nature microbiology. 2019. [PMID 30531976](https://pubmed.ncbi.nlm.nih.gov/30531976/). DOI: 10.1038/s41564-018-0306-4. Access: abstract_metadata. Study: PMID30531976.

43. Falony G, Joossens M, Vieira-Silva S et al. Population-level analysis of gut microbiome variation. Science (New York, N.Y.). 2016. [PMID 27126039](https://pubmed.ncbi.nlm.nih.gov/27126039/). DOI: 10.1126/science.aad3503. Access: abstract_metadata. Study: PMID27126039.

44. Yap CX, Henders AK, Alvares GA et al. Autism-related dietary preferences mediate autism-gut microbiome associations. Cell. 2021. [PMID 34767757](https://pubmed.ncbi.nlm.nih.gov/34767757/). DOI: 10.1016/j.cell.2021.10.015. Access: abstract_metadata. Study: PMID34767757.

45. Valles-Colomer M, Falony G, Darzi Y et al. The neuroactive potential of the human gut microbiota in quality of life and depression. Nature microbiology. 2019. [PMID 30718848](https://pubmed.ncbi.nlm.nih.gov/30718848/). DOI: 10.1038/s41564-018-0337-x. Access: abstract_metadata. Study: PMID30718848.

46. Forslund K, Hildebrand F, Nielsen T et al. Disentangling type 2 diabetes and metformin treatment signatures in the human gut microbiota. Nature. 2015. [PMID 26633628](https://pubmed.ncbi.nlm.nih.gov/26633628/). DOI: 10.1038/nature15766. Access: abstract_metadata. Study: PMID26633628.

47. Jin WB, Li TT, Huo D et al. Genetic manipulation of gut microbes enables single-gene interrogation in a complex microbiome. Cell. 2022. [PMID 35051369](https://pubmed.ncbi.nlm.nih.gov/35051369/). DOI: 10.1016/j.cell.2021.12.035. Access: abstract_metadata. Study: PMID35051369.

48. Liu Z, Zhang Q, Zhang H et al. Colorectal cancer microbiome programs DNA methylation of host cells by affecting methyl donor metabolism. Genome medicine. 2024. [PMID 38840170](https://pubmed.ncbi.nlm.nih.gov/38840170/). DOI: 10.1186/s13073-024-01344-1. Access: abstract_metadata. Study: PMID38840170.

49. Ridaura VK, Faith JJ, Rey FE et al. Gut microbiota from twins discordant for obesity modulate metabolism in mice. Science (New York, N.Y.). 2013. [PMID 24009397](https://pubmed.ncbi.nlm.nih.gov/24009397/). DOI: 10.1126/science.1241214. Access: full_text_partial. Study: PMID24009397.

50. Hsiao EY, McBride SW, Hsien S et al. Microbiota modulate behavioral and physiological abnormalities associated with neurodevelopmental disorders. Cell. 2013. [PMID 24315484](https://pubmed.ncbi.nlm.nih.gov/24315484/). DOI: 10.1016/j.cell.2013.11.024. Access: abstract_metadata. Study: PMID24315484.

51. Ignacio A, Shah K, Bernier-Latmani J et al. Small intestinal resident eosinophils maintain gut homeostasis following microbial colonization. Immunity. 2022. [PMID 35709757](https://pubmed.ncbi.nlm.nih.gov/35709757/). DOI: 10.1016/j.immuni.2022.05.014. Access: abstract_metadata. Study: PMID35709757.

52. Wei M, Huang F, Zhao L et al. A dysregulated bile acid-gut microbiota axis contributes to obesity susceptibility. EBioMedicine. 2020. [PMID 32408110](https://pubmed.ncbi.nlm.nih.gov/32408110/). DOI: 10.1016/j.ebiom.2020.102766. Access: abstract_metadata. Study: PMID32408110.

53. Overacre-Delgoffe AE, Bumgarner HJ, Cillo AR et al. Microbiota-specific T follicular helper cells drive tertiary lymphoid structures and anti-tumor immunity against colorectal cancer. Immunity. 2021. [PMID 34861182](https://pubmed.ncbi.nlm.nih.gov/34861182/). DOI: 10.1016/j.immuni.2021.11.003. Access: abstract_metadata. Study: PMID34861182.

54. Kim JE, Li B, Fei L et al. Gut microbiota promotes stem cell differentiation through macrophage and mesenchymal niches in early postnatal development. Immunity. 2022. [PMID 36473468](https://pubmed.ncbi.nlm.nih.gov/36473468/). DOI: 10.1016/j.immuni.2022.11.003. Access: abstract_metadata. Study: PMID36473468.

55. Porcari S, Ciccarese C, Heidrich V et al. Fecal microbiota transplantation plus pembrolizumab and axitinib in metastatic renal cell carcinoma: the randomized phase 2 TACITO trial. Nature medicine. 2026. [PMID 41606119](https://pubmed.ncbi.nlm.nih.gov/41606119/). DOI: 10.1038/s41591-025-04189-2. Access: full_text_partial. Study: NCT04758507.

56. Duttagupta S, Messaoudene M, Hunter S et al. Fecal microbiota transplantation plus immunotherapy in non-small cell lung cancer and melanoma: the phase 2 FMT-LUMINate trial. Nature medicine. 2026. [PMID 41606121](https://pubmed.ncbi.nlm.nih.gov/41606121/). DOI: 10.1038/s41591-025-04186-5. Access: full_text_partial. Study: LUMINate-new-cohorts+prior-FMT-reanalysis.

57. Li J, Wang Y, Li R et al. Adjunctive fecal microbiota transplantation for major depressive disorder: A randomized, double-blind, placebo-controlled trial. Cell host & microbe. 2026. [PMID 42309058](https://pubmed.ncbi.nlm.nih.gov/42309058/). DOI: 10.1016/j.chom.2026.05.017. Access: abstract_metadata. Study: PMID42309058.

58. Arpaia N, Campbell C, Fan X et al. Metabolites produced by commensal bacteria promote peripheral regulatory T-cell generation. Nature. 2013. [PMID 24226773](https://pubmed.ncbi.nlm.nih.gov/24226773/). DOI: 10.1038/nature12726. Access: abstract_metadata. Study: PMID24226773.

59. Buffie CG, Bucci V, Stein RR et al. Precision microbiome reconstitution restores bile acid mediated resistance to Clostridium difficile. Nature. 2015. [PMID 25337874](https://pubmed.ncbi.nlm.nih.gov/25337874/). DOI: 10.1038/nature13828. Access: full_text_partial. Study: PMID25337874.

60. David LA, Maurice CF, Carmody RN et al. Diet rapidly and reproducibly alters the human gut microbiome. Nature. 2014. [PMID 24336217](https://pubmed.ncbi.nlm.nih.gov/24336217/). DOI: 10.1038/nature12820. Access: abstract_metadata. Study: PMID24336217.

## Source-specific evidence cards

### [1] SER-109, an Oral Microbiome Therapy for Recurrent Clostridioides difficile Infection.

**Shows:** Post-antibiotic SER-109 reduced recurrent CDI at 8 weeks in the randomized trial.

**Does not show:** Restoration in a depleted niche does not establish autonomous regulation of systemic health.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC4.

### [2] Fecal Microbiota Transplantation Versus Vancomycin for Primary Clostridioides difficile Infection : A Randomized Controlled Trial.

**Shows:** Primary-CDI FMT met the specified noninferiority comparison against vancomycin.

**Does not show:** The 25-point margin and rescue therapy limit claims of superiority; this is primary CDI rather than the exact recurrent-CDI subclaim.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC4.

### [3] VE303, a Defined Bacterial Consortium, for Prevention of Recurrent Clostridioides difficile Infection: A Randomized Clinical Trial.

**Shows:** A defined eight-strain VE303 consortium reduced recurrence at the high dose in the phase 2 trial.

**Does not show:** Small groups and recurrence definitions require confirmation; the exact mediator was not randomized.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

### [4] Efficacy and Safety of RBX2660 in PUNCH CD3, a Phase III, Randomized, Double-Blind, Placebo-Controlled Trial with a Bayesian Primary Analysis for the Prevention of Recurrent Clostridioides difficile Infection.

**Shows:** RBX2660 had higher modeled treatment success than placebo following antibiotics.

**Does not show:** The Bayesian estimate borrows the phase 2 study and is not an independent replication of that report; linked correction remains attached.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID36342618. Subclaims: SC3.

### [5] Faecal microbiota transplantation for first or second Clostridioides difficile infection (EarlyFMT): a randomised, double-blind, placebo-controlled trial.

**Shows:** Early FMT improved resolution after antibiotics in first or second CDI.

**Does not show:** Early stopping and disease-stage differences limit magnitude and recurrent-CDI generalization.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

### [6] Effect of Oral Capsule- vs Colonoscopy-Delivered Fecal Microbiota Transplantation on Recurrent Clostridium difficile Infection: A Randomized Clinical Trial.

**Shows:** Capsule FMT was noninferior to colonoscopic FMT for preventing recurrence.

**Does not show:** Both groups received FMT; the route contrast does not isolate efficacy against no FMT.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

### [7] Effects of lyophilised faecal filtrate compared with lyophilised donor stool on Clostridioides difficile recurrence: a multicentre, randomised, double-blinded, non-inferiority trial.

**Shows:** Sterile filtrate failed noninferiority to live donor stool, with lower recurrence-free proportions.

**Does not show:** Filtration changes multiple components; failure of noninferiority is not proof that every live organism is necessary.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

### [8] Results From a Randomized, Placebo-Controlled Clinical Trial of a RBX2660-A Microbiota-Based Drug for the Prevention of Recurrent Clostridium difficile Infection.

**Shows:** The two-dose RBX2660 primary contrast did not meet significance; a one-dose secondary contrast favored treatment.

**Does not show:** Secondary and open-label rescue outcomes cannot replace the null primary comparison; this cohort informs the phase 3 Bayesian estimate.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

### [9] Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study.

**Shows:** Pasteurized Akkermansia improved insulin sensitivity and some metabolic measures in a small human pilot.

**Does not show:** Weight and fat changes were not significant; a pasteurized product effect does not establish live-community autonomy.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [10] A purified membrane protein from Akkermansia muciniphila or the pasteurized bacterium improves metabolism in obese and diabetic mice.

**Shows:** Pasteurized Akkermansia and its membrane protein partially reproduced metabolic benefits in mice.

**Does not show:** Human observations in this report concern safety; a protein/TLR2 pathway depends on host machinery.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

### [11] Effects of healthy low-carbohydrate diet and time-restricted eating on weight and gut microbiome in adults with overweight or obesity: Feeding RCT.

**Shows:** Controlled dietary restriction and timing changed weight and microbial measures.

**Does not show:** Diet was manipulated directly; parallel changes do not isolate microbial mediation.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [12] A mechanism by which gut microbiota elevates permeability and inflammation in obese/diabetic mice and human gut.

**Shows:** Transfer, ethanolamine metabolism and probiotic rescue linked microbial function to barrier and metabolic phenotypes.

**Does not show:** Model rescue supports a defined pathway, not autonomous human multi-system primacy.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

### [13] Mediterranean diet intervention in overweight and obese subjects lowers plasma cholesterol and causes changes in the gut microbiome and metabolome independently of energy intake.

**Shows:** An isocaloric Mediterranean dietary intervention changed cholesterol and microbial/metabolic profiles.

**Does not show:** Direct dietary effects and correlated microbiome changes do not identify the mediator.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [14] Akkermansia muciniphila supplementation in patients with overweight/obese type 2 diabetes: Efficacy depends on its baseline levels in the gut.

**Shows:** The Akkermansia T2D trial had overall null between-group weight/HbA1c contrasts, with benefits in a low-baseline-abundance subgroup and accompanying mouse work.

**Does not show:** Subgroup findings require prospective confirmation; baseline community dependence contradicts simple universal response.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

### [15] Fecal microbiota transplantation reverses insulin resistance in type 2 diabetes: A randomized, controlled, prospective study.

**Shows:** A small short-term FMT/metformin comparison reported improvements in metabolic indices.

**Does not show:** Limited size, duration and comparator structure do not establish durable broad human health benefit.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [16] Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors.

**Shows:** Responder transfer and Akkermansia rescue enhanced PD-1 effects in mice with an IL-12-dependent route.

**Does not show:** Human antibiotic and taxon associations remain confounded; active anticancer treatment and host immunity gate the effect.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC4.

### [17] Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients.

**Shows:** Melanoma responder communities were associated with response and transferred favorable immune/tumor phenotypes into mice.

**Does not show:** The human observational comparison does not establish the microbial cause of patient benefit.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC4.

### [18] Dietary fiber and probiotics influence the gut microbiome and melanoma immunotherapy response.

**Shows:** Dietary fiber associated with human progression-free survival; low fiber or probiotics impaired ICI response in models.

**Does not show:** The human diet comparison is observational and the model effect depends on diet and immunotherapy.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC4.

### [19] Exercise-induced microbiota metabolite enhances CD8 T cell antitumor immunity promoting immunotherapy efficacy.

**Shows:** Exercise-associated microbial formate was tested as a required/sufficient contributor to antitumor Tc1 phenotypes, with Nrf2 mediation.

**Does not show:** Human formate is a biomarker candidate; exercise and host transcriptional gating remain upstream conditions.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

### [20] Fecal microbiota transplant overcomes resistance to anti-PD-1 therapy in melanoma patients.

**Shows:** Responder-derived FMT plus anti-PD-1 produced clinical benefit in six of fifteen refractory melanoma patients.

**Does not show:** Single-arm drug rechallenge does not isolate the incremental FMT effect or human mediator necessity.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [21] Fecal microbiota transplant promotes response in immunotherapy-refractory melanoma patients.

**Shows:** FMT with PD-1 reinduction produced responses in three of ten refractory melanoma patients.

**Does not show:** The phase 1 single-arm design establishes feasibility and signals, not a controlled causal efficacy estimate.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [22] Commensal Bifidobacterium promotes antitumor immunity and facilitates anti-PD-L1 efficacy.

**Shows:** Cohousing, transfer and Bifidobacterium administration improved mouse tumor immunity and checkpoint efficacy.

**Does not show:** Mouse facility/community and immune context limit human generalization.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

### [23] Microbiome-derived inosine modulates response to checkpoint inhibitor immunotherapy.

**Shows:** Defined bacterial isolates and inosine enhanced checkpoint efficacy in several mouse cancer models through a costimulation-dependent A2A receptor route.

**Does not show:** The metabolite requires treatment/barrier and host immune context; it is not an autonomous human master switch.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

### [24] The commensal microbiome is associated with anti-PD-1 efficacy in metastatic melanoma patients.

**Shows:** Baseline melanoma taxa associated with response; responder-derived microbiota improved tumor control in mice.

**Does not show:** Human association and recipient sufficiency do not identify endogenous human necessity.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [25] Targeting PD-L2-RGMb overcomes microbiome-related immunotherapy resistance.

**Shows:** PD-L2/RGMb blockade or conditional T-cell RGMb deletion overcame microbial-context ICI resistance in mouse tumors.

**Does not show:** The host immune intervention can bypass microbial resistance; correction-body details were not examined.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID37264079. Subclaims: SC2.

### [26] Multidonor intensive faecal microbiota transplantation for active ulcerative colitis: a randomised placebo-controlled trial.

**Shows:** Intensive multidonor FMT improved the prespecified steroid-free clinical/endoscopic outcome in UC versus placebo.

**Does not show:** Crude-community benefit supports a UC treatment package, not strain-specific or autonomous systemic causation.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [27] Lyophilised oral faecal microbiota transplantation for ulcerative colitis (LOTUS): a randomised, double-blind, placebo-controlled trial.

**Shows:** Oral FMT after antibiotics improved the week-8 UC outcome; a very small maintenance randomization favored continuation.

**Does not show:** Early pandemic termination, pretreatment and tiny maintenance groups limit generalization and mechanism.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [28] Fecal Microbiota Transplantation Induces Remission in Patients With Active Ulcerative Colitis in a Randomized Controlled Trial.

**Shows:** Weekly donor FMT improved remission versus water enema in active UC.

**Does not show:** Donor dependence and early futility stopping complicate precision; the result does not establish universal microbiome primacy.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [29] Effects of Low FODMAP Diet on Symptoms, Fecal Microbiome, and Markers of Inflammation in Patients With Quiescent Inflammatory Bowel Disease in a Randomized Trial.

**Shows:** Low FODMAP diet improved reported adequate symptom relief while reducing some putatively beneficial taxa.

**Does not show:** Inflammatory markers and diversity did not differ; health cannot be inferred from a universal favorable-diversity metric.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [30] Faecal microbiota transplantation with anti-inflammatory diet (FMT-AID) followed by anti-inflammatory diet alone is effective in inducing and maintaining remission over 1 year in mild to moderate ulcerative colitis: a randomised controlled trial.

**Shows:** FMT plus anti-inflammatory diet improved UC outcomes versus optimized standard therapy.

**Does not show:** The bundled unblinded intervention and continued diet do not isolate FMT or a universal microbial mechanism.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [31] Effect of Fecal Microbiota Transplantation on 8-Week Remission in Patients With Ulcerative Colitis: A Randomized Clinical Trial.

**Shows:** Anaerobic pooled donor FMT improved week-8 steroid-free UC remission versus autologous FMT.

**Does not show:** Most initial responders did not retain remission at twelve months; the mediator is not selectively identified.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [32] Low-Fat, High-Fiber Diet Reduces Markers of Inflammation and Dysbiosis and Improves Quality of Life in Patients With Ulcerative Colitis.

**Shows:** Catered diets improved quality of life in a small UC crossover with microbial/metabolite changes.

**Does not show:** Some inflammatory contrasts were only significant versus baseline; diet-microbiome mediation was not isolated.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [33] Human Gut Microbiota from Autism Spectrum Disorder Promote Behavioral Symptoms in Mice.

**Shows:** ASD donor transfer induced mouse behavioral features; candidate metabolites improved model phenotypes.

**Does not show:** Recipient mouse behaviors are not human ASD causation or remission, and donor phenotypes include diet-related differences.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

### [34] Probiotic Supplementation Improves Cognitive Function and Mood with Changes in Gut Microbiota in Community-Dwelling Older Adults: A Randomized, Double-Blind, Placebo-Controlled, Multicenter Trial.

**Shows:** A small probiotic RCT improved selected flexibility/stress measures in healthy older adults.

**Does not show:** Multiple outcomes and short follow-up limit general clinical interpretation; BDNF correlation is not manipulated mediation.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [35] Transferring the blues: Depression-associated gut microbiota induces neurobehavioural changes in the rat.

**Shows:** Depression-associated donor stool induced anhedonia/anxiety and metabolic features in depleted rats.

**Does not show:** Recipient effects establish capacity in a model, not primary causation of endogenous human depression.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC4.

### [36] Microbiota Transfer Therapy alters gut ecosystem and improves gastrointestinal and autism symptoms: an open-label study.

**Shows:** A bundled antibiotic/cleanse/FMT open-label ASD study reported persistent GI and behavioral improvement.

**Does not show:** Eighteen participants without randomized masking cannot separate expectation, co-interventions or natural trajectory.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [37] Microbiota alterations in proline metabolism impact depression.

**Shows:** Multi-cohort proline associations were combined with supplementation, transfer and transporter/RNAi or bacterial manipulations in models.

**Does not show:** Human dietary and depression relationships remain observational; cross-species behaviors do not establish clinical efficacy.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [38] Psychobiotic Lactobacillus plantarum JYLP-326 relieves anxiety, depression, and insomnia symptoms in test anxious college via modulating the gut microbiota and its metabolism.

**Shows:** A short probiotic/placebo comparison in test-anxious students reported improved symptom questionnaires.

**Does not show:** Short, selected student symptoms are not durable clinical psychiatric remission or established microbial mediation.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [39] Bacteroides species differentially modulate depression-like behavior via gut-brain metabolic signaling.

**Shows:** Selected Bacteroides species reproduced behavioral/neurogenesis effects in antibiotic-treated mice whereas another species did not.

**Does not show:** Species and host-metabolite specificity constrain the result; the exact human primary pathway is not established.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

### [40] Microbiome and tryptophan metabolomics analysis in adolescent depression: roles of the gut microbiota in the regulation of tryptophan-derived neurotransmitters and behaviors in human and mice.

**Shows:** Human adolescent signatures and Roseburia/healthy-transfer rescue linked tryptophan pathways to stressed-mouse phenotypes.

**Does not show:** Sertraline alters human signatures; model rescue is not a randomized human mediator test.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [41] Donor-recipient intermicrobial interactions impact transfer of subspecies and fecal microbiota transplantation outcome.

**Shows:** Donor-recipient subspecies interactions associated with transfer and outcomes across FMT datasets.

**Does not show:** Reanalysis of shared datasets does not create independent clinical intervention lines or establish outcome mediation.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [42] Gut microbiome structure and metabolic activity in inflammatory bowel disease.

**Shows:** IBD microbiome/metabolome differences and classifiers replicated across discovery/validation cohorts.

**Does not show:** Inflammatory state and reverse causation remain possible; linked erratum-body impact is unresolved.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID30971771. Subclaims: SC1, SC4.

### [43] Population-level analysis of gut microbiome variation.

**Shows:** Population cohorts linked microbial composition to numerous covariates, especially stool consistency and medications.

**Does not show:** Associations identify confounders, not causal direction or autonomous regulation.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [44] Autism-related dietary preferences mediate autism-gut microbiome associations.

**Shows:** The ASD cohort showed negligible direct diagnostic associations and supported diet/stool-mediated signatures.

**Does not show:** Observational mediation does not prove all ASD microbial findings are reactive; correction impact is not fully examined.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID38242089. Subclaims: SC1, SC4.

### [45] The neuroactive potential of the human gut microbiota in quality of life and depression.

**Shows:** Neuroactive taxonomic/function potential associated with depression and quality of life in population cohorts.

**Does not show:** Covariate adjustment cannot exclude reverse causation or demonstrate pathway activity and necessity.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [46] Disentangling type 2 diabetes and metformin treatment signatures in the human gut microbiota.

**Shows:** Stratification distinguished metformin-associated microbial signatures from untreated T2D signatures.

**Does not show:** Medication confounding is material; associations do not prove the microbial fraction of drug efficacy; correction remains attached.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID28470190. Subclaims: SC1, SC4.

### [47] Genetic manipulation of gut microbes enables single-gene interrogation in a complex microbiome.

**Shows:** A targeted commensal bile-acid gene deletion in a complex community altered host colon inflammation.

**Does not show:** Gene-level specificity is stronger than dysbiosis association but does not establish a universal human regulator.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

### [48] Colorectal cancer microbiome programs DNA methylation of host cells by affecting methyl donor metabolism.

**Shows:** New CRC tissue/fecal profiling linked microbial methyl-donor features to host methylation patterns.

**Does not show:** Original observational data are admitted; the accompanying meta-analysis is not an additional primary source or causal perturbation.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [49] Gut microbiota from twins discordant for obesity modulate metabolism in mice.

**Shows:** Obese/lean twin communities and cultured collections transferred adiposity; cohousing rescue succeeded or failed depending on diet.

**Does not show:** Food intake contrasts do not explain adiposity here, but dietary dependence directly limits microbial autonomy; mild glucose changes are not established insulin resistance.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: S2, S4, S8. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

### [50] Microbiota modulate behavioral and physiological abnormalities associated with neurodevelopmental disorders.

**Shows:** B. fragilis improved barrier and selected behaviors in maternal-immune-activation offspring; a candidate metabolite induced behavioral abnormalities.

**Does not show:** Developmental immune injury precedes the microbial intervention; model modulation does not establish human ASD initiation.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

### [51] Small intestinal resident eosinophils maintain gut homeostasis following microbial colonization.

**Shows:** Eosinophil-deficient/colonization contrasts linked microbial IL-33 responses to gut physiological homeostasis.

**Does not show:** Host immune cells gate the microbial effect; the pathway is mutualistic rather than autonomous.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

### [52] A dysregulated bile acid-gut microbiota axis contributes to obesity susceptibility.

**Shows:** C. scindens or bile-acid administration altered metabolic/obesity phenotypes in mice alongside human associations.

**Does not show:** Host bile-acid substrate and diet are required context; correction details remain unexamined.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID33845237. Subclaims: SC1, SC2.

### [53] Microbiota-specific T follicular helper cells drive tertiary lymphoid structures and anti-tumor immunity against colorectal cancer.

**Shows:** H. hepaticus-induced tumor control depended on Tfh cells, with host-cell rescue in a deficient mouse setting.

**Does not show:** Host immune epistasis supplies a specific causal route; it does not prove broad human cancer benefit.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

### [54] Gut microbiota promotes stem cell differentiation through macrophage and mesenchymal niches in early postnatal development.

**Shows:** Antibiotic, mouse genetic and Lactobacillus/macrophage rescue contrasts linked microbes to postnatal niches and NEC severity.

**Does not show:** Developmental host-cell dependence and antibiotic pleiotropy limit autonomy; erratum-body details were not examined.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID37703832. Subclaims: SC2.

### [55] Fecal microbiota transplantation plus pembrolizumab and axitinib in metastatic renal cell carcinoma: the randomized phase 2 TACITO trial.

**Shows:** TACITO primary 12-month PFS was 70% versus 41%, P=.053; secondary median PFS favored FMT.

**Does not show:** The primary superiority threshold was not met; secondary nominal signals, small groups and concurrent cancer drugs do not prove autonomous regulation.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: Sec5, Sec17, Sec29, Sec30. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

### [56] Fecal microbiota transplantation plus immunotherapy in non-small cell lung cancer and melanoma: the phase 2 FMT-LUMINate trial.

**Shows:** LUMINate reported single-arm clinical response signals; reintroducing lost baseline bacterial cocktails impaired ICI efficacy in mouse recipients.

**Does not show:** Clinical FMT benefit lacks a randomized counterfactual; the controlled mouse add-back establishes only recipient/treatment-context mediation.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: Sec7, Sec41. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC4.

### [57] Adjunctive fecal microbiota transplantation for major depressive disorder: A randomized, double-blind, placebo-controlled trial.

**Shows:** Adjunct FMT improved selected depression scores but not week-8 remission versus placebo with escitalopram.

**Does not show:** Statistical bile-acid mediation is not selective causal manipulation; durable remission and autonomous monotherapy are not shown.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

### [58] Metabolites produced by commensal bacteria promote peripheral regulatory T-cell generation.

**Shows:** Butyrate/propionate promoted peripheral Treg generation with CNS1 dependence; acetate lacked the corresponding activity.

**Does not show:** Defined metabolite/host enhancer logic supports communication, not a universal human master regulator.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

### [59] Precision microbiome reconstitution restores bile acid mediated resistance to Clostridium difficile.

**Shows:** C. scindens/consortium transfer reduced CDI illness in mice; bile-salt sequestration neutralized inhibition ex vivo.

**Does not show:** The human cohort is observational and the bile-acid blockade is ex vivo, not in vivo gene-specific necessity or human mediation.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: BODY, S4, S5. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC3.

### [60] Diet rapidly and reproducibly alters the human gut microbiome.

**Shows:** Short animal/plant diets rapidly altered human microbial structure and expression.

**Does not show:** This demonstrates diet-to-microbiome direction; downstream health mediation was not isolated.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.