{ "$schema_version": "5.4", "review_mode": "revision", "review_metadata": { "review_date": "2026-09-27", "reviewer": "Codex (base grading, 2026-09-27); red-team synthesis: Claude (Anthropic), independent model, 2026-09-27", "evidence_base_note": "Base grading: fresh primary-literature assessment through 2026-09-27. Red-team pass: every ledger row re-resolved in PubMed, every registered claim re-checked against the record, and six sources added after an omission audit in both directions.", "literature_tools_available": true, "literature_tools_note": "Red-team retrieval used PubMed/PMC records, publisher pages and an authors' institutional copy. Full text newly obtained for Li 2026 (authors' institutional PDF), Green 2023 (publisher results tables) and Kelly 2016 (university repository PDF), each of which the base grading could not reach. Kazemi 2019, Zheng 2016, Valles-Colomer 2019, Chen 2022, Zhuang 2020 and the two 2025 syntheses were read at abstract depth; no open-access full text was located for them on 2026-09-27. No citation-context counts were available.", "skill_version": "5.4 packet; hsef-hypothesis-review skill 3.0", "validator_run": "pass", "validator_warnings_addressed": "Validator 2.4 re-run on the revised pair; full output in final/validation.txt.", "consensus_consistency_audit": "2026-10-09; supplementary provenance/source reconciliation; original grading retained; prior grades visible." }, "provenance_header": { "sources_total": 23, "claims_total": 23, "claims_verified": 23, "claims_unverified": 0, "claims_contradicted": 0, "tier_a_sources": 21, "generated_by": "Codex HSEF v5.4 primary-review workflow; red-team synthesis by Claude (Anthropic), independent model, 2026-09-27", "gate_passed": true }, "claims": [ { "id": "C1", "text": "In 1,054 Rotterdam and 1,539 HELIUS participants after exclusions, 12 genera associated with depressive symptoms in the same direction across cohorts; the design remained observational.", "carries_number": true, "load_bearing_for": [ "SC1", "SC5" ], "source_ns": [ 1 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: cohort restrictions, taxon replication and association results" }, { "id": "C2", "text": "A 1,269-person cross-sectional MDD cohort inferred possible Eggerthella and Hungatella causal directions with a graphical algorithm, while the authors stated that observational data cannot definitively establish causality.", "carries_number": true, "load_bearing_for": [ "SC1", "SC4" ], "source_ns": [ 2 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: 1,269-person design, causal-discovery methods and limitations" }, { "id": "C3", "text": "Stool pooled from three of the most severely depressed male patients versus three matched controls induced sucrose-preference and anxiety-like differences in 28 antibiotic-depleted rats (13 depressed-donor, 15 control-donor); forced-swim outcomes were null.", "carries_number": true, "load_bearing_for": [ "SC2", "SC4" ], "source_ns": [ 3 ], "verification": "verified", "verified_against": "primary full text, donor selection, recipient preparation and behavioural results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. University repository primary PDF: donor selection, rat experiment, behavioral results and limitations" }, { "id": "C4", "text": "Transfer of stress-altered mouse microbiota induced depressive-like behavior and hippocampal changes in recipient mice; subdiaphragmatic vagotomy abolished the transferred effects.", "carries_number": false, "load_bearing_for": [ "SC2", "SC4" ], "source_ns": [ 4 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML and primary publisher HTML: transfer, vagotomy, brain and behavior results" }, { "id": "C5", "text": "In 44 IBS patients, Bifidobacterium longum improved the two-point HADS-depression response at week 6 (14/22 versus 7/22; P=.04) and changed fMRI responses, but fecal community profiles were similar.", "carries_number": true, "load_bearing_for": [ "SC3", "SC4" ], "source_ns": [ 5 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA accepted primary manuscript PDF: prespecified endpoint, microbiota and fMRI results" }, { "id": "C6", "text": "Of 110 randomized MDD patients, 81 completed the three-arm trial; probiotic supplementation improved BDI scores versus placebo and prebiotic at eight weeks (P=.042), without gut composition measurement reported in the abstract.", "carries_number": true, "load_bearing_for": [ "SC3" ], "source_ns": [ 6 ], "verification": "verified", "verified_against": "primary abstract", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. PubMed primary abstract only: randomization, completers, BDI and blood outcomes; full text not retrieved" }, { "id": "C7", "text": "An eight-week double-blind trial in 79 unmedicated people selected for low mood found no probiotic–placebo difference on prespecified psychological outcomes or measured blood markers.", "carries_number": true, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 7 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: prespecified measures, ITT and null outcomes" }, { "id": "C8", "text": "Of 60 enrolled depressed inpatients, 47 completed the trial; the HAM-D time-by-treatment test was positive only in the compliant subset, while Lactobacillus abundance rose in the probiotic arm.", "carries_number": true, "load_bearing_for": [ "SC3", "SC4" ], "source_ns": [ 8 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: enrollment, primary ITT/mITT results, taxa and imaging" }, { "id": "C9", "text": "In a 49-person adjunctive probiotic pilot, self-reported depression and clinician-rated anxiety favored probiotics at week 8, while clinician-rated depression and GAD-7 did not show clear week-eight differences.", "carries_number": true, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 9 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: pilot design, 49-person ITT and symptom contrasts" }, { "id": "C10", "text": "A sham-controlled MDD FMT feasibility pilot randomized 15 adults 2:1 with successful masking; week-8 MADRS was 16.6 in both arms (Hedges g = 0.05; P = .67) and DASS did not differ between arms (P = .19).", "carries_number": true, "load_bearing_for": [ "SC3" ], "source_ns": [ 10 ], "verification": "verified", "verified_against": "primary full text, publisher Results tables", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. Publisher article read on 2026-09-27; the PMC deposit carries the abstract only, so the base grading's indexed-sections reading was re-done at the source." }, { "id": "C11", "text": "The 116-person PRO-DEMET randomized trial found no probiotic advantage on its primary MADRS change (P=.45), although a secondary minimal-important-change comparison favored probiotics.", "carries_number": true, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 11 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: prespecified MADRS outcome and primary versus secondary results" }, { "id": "C12", "text": "A 131-person bipolar/schizophrenia-spectrum probiotic RCT found no treatment effect on psychiatric symptom severity while serum and fecal gut-barrier markers changed; one cognitive domain, verbal memory, improved borderline (adjusted P = .059).", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 12 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: randomization, BPRS primary result and gut-marker results" }, { "id": "C13", "text": "A probiotic trial in depressed patients and controls found changes in some heart-rate-variability measures, but no probiotic effect on depression scores and no overall stool community shift over time.", "carries_number": false, "load_bearing_for": [ "SC3", "SC4" ], "source_ns": [ 13 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: HRV primary endpoint, mood scores and stool beta-diversity" }, { "id": "C14", "text": "In a 50-person MDD trial, probiotic plus fluoxetine lowered HDRS-24 more than placebo plus fluoxetine among 44 completers, while cortisol, ACTH and BDNF comparisons were null.", "carries_number": true, "load_bearing_for": [ "SC3", "SC4" ], "source_ns": [ 14 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: randomized design, completion, HDRS-24 and blood biomarkers" }, { "id": "C15", "text": "In a 60-person double-blind adjunctive MDD FMT trial, the prespecified primary endpoint was null (week-8 remission 43.3% versus 26.7%; P = 0.1760) while the secondary HAMD-17 contrast favoured FMT at week 8 (-2.9 points, 95% CI -5.5 to -0.3; P = 0.0303); masking was intact, newly acquired donor species fell from about 16% at week 2 to about 7% at week 8, and glycocholic acid statistically mediated 27% of the association between L. eligens and CASP-8 without any bile-acid manipulation.", "carries_number": true, "load_bearing_for": [ "SC3", "SC4" ], "source_ns": [ 15 ], "verification": "verified", "verified_against": "primary full text, endpoints, masking check, engraftment and mediation sections", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. Full text obtained on 2026-09-27 from the authors' institutional copy after the publisher site refused retrieval; this replaces the base grading's abstract-only reading, whose word 'durable' for engraftment the full text does not support." }, { "id": "C16", "text": "In 35 randomized adults with bipolar depression, allogenic FMT moved stool communities toward the donor and raised bacterial diversity at six months while autologous FMT did not, yet MADRS improvement did not differ (16.74 versus 15.4 points; t = -0.47; P = .64; 95% CI -7.3 to 4.6).", "carries_number": true, "load_bearing_for": [ "SC3", "SC4", "SC5" ], "source_ns": [ 16 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: 35-person design, MADRS contrast and donor similarity" }, { "id": "C17", "text": "A medication-plus-FMT versus medication-only trial reported greater two-week HAMD-24 reduction (11.50 versus 9.40 points; P = 0.040), but only 14 of 23 FMT-assigned versus 20 of 23 controls were analysed, there was no sham procedure, only ten selected bacterial targets were profiled, and post-treatment Enterococcus correlated positively with HAMD-24 scores (P = 0.030).", "carries_number": true, "load_bearing_for": [ "SC3", "SC4" ], "source_ns": [ 17 ], "verification": "verified", "verified_against": "primary full text, relevant methods/results", "verified_on": "2026-09-27", "note": "Verification is of the report and its scope, not proof of its causal interpretation. OA XML: intervention, attrition, two-week HAMD and limited qPCR flora" }, { "id": "C18", "text": "Germ-free mice colonized with fecal microbiota from patients with major depressive disorder developed depression-like behaviour relative to mice colonized with healthy-donor microbiota, with accompanying disturbances of carbohydrate and amino-acid metabolism.", "carries_number": false, "load_bearing_for": [ "SC2" ], "source_ns": [ 18 ], "verification": "verified", "verified_against": "primary abstract", "verified_on": "2026-09-27", "note": "Added by the red team as a second, independently conducted recipient system; germ-free rather than antibiotic-conditioned. No open-access full text located." }, { "id": "C19", "text": "In a 1,054-participant population cohort with validation in 1,070 further individuals, Coprococcus and Dialister were depleted in depression after correcting for antidepressant use, and butyrate producers tracked quality-of-life indicators.", "carries_number": true, "load_bearing_for": [ "SC1" ], "source_ns": [ 19 ], "verification": "verified", "verified_against": "primary abstract", "verified_on": "2026-09-27", "note": "Added by the red team; the landmark population-scale association the base grading omitted. It corroborates SC1 and cannot lift it past the Β cap." }, { "id": "C20", "text": "A two-sample bidirectional Mendelian randomization reported protective effects of Actinobacteria (OR 0.88, 95% CI 0.87-0.90) and Bifidobacterium (OR 0.89, 95% CI 0.88-0.91) on major depression, with the Ruminococcus1 and Streptococcaceae findings not supported by MR-Egger and no effect of depression on composition in the reverse direction.", "carries_number": true, "load_bearing_for": [ "SC1", "SC4" ], "source_ns": [ 20 ], "verification": "verified", "verified_against": "primary abstract", "verified_on": "2026-09-27", "note": "Added by the red team to populate the framework's Mendelian-randomization equivalent, which the base grading's upper-ladder check omitted." }, { "id": "C21", "text": "A bidirectional Mendelian randomization across Alzheimer's disease, major depression and schizophrenia reported a suggestive effect of genetically increased Bacilli on major depression (OR 1.07, 95% CI 1.02-1.12; P = 0.010), and found that the disorders in turn altered gut microbiota composition.", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 21 ], "verification": "verified", "verified_against": "primary abstract", "verified_on": "2026-09-27", "note": "Added by the red team. Its nominated taxa do not overlap with the other instrument study's, and its reverse-direction result cuts against the nominal model." }, { "id": "C22", "text": "In the same Rotterdam/HELIUS report, two-sample Mendelian randomization with major depression as the exposure gave a significant inverse-variance-weighted effect on Eggerthella (effect 0.237; P = 0.027), while the microbiome-as-exposure analysis returned an estimate inconsistent with the cohort association; the authors state the MR lacked power because instruments were taken below genome-wide significance.", "carries_number": true, "load_bearing_for": [ "SC1", "SC4" ], "source_ns": [ 1 ], "verification": "verified", "verified_against": "primary full text, Mendelian randomization results and limitations", "verified_on": "2026-09-27", "note": "Registered by the red team. The base grading read this source but did not report the analysis inside it that runs against the nominal model." }, { "id": "C23", "text": "Two 2025 systematic reviews of randomized trials in clinically diagnosed samples report a large favourable pooled effect of probiotics on depressive symptoms with between-study heterogeneity close to total, no prebiotic effect, and a recommendation that the gut microbiome be monitored and treated alongside pharmacological treatment; this is the field framing from which the divergence implied level is read.", "carries_number": false, "load_bearing_for": [ "SC3" ], "source_ns": [ 22, 23 ], "verification": "verified", "verified_against": "primary abstracts of both reviews", "verified_on": "2026-09-27", "note": "Tier B. Used for field framing and the divergence artifact only; no pooled figure is imported into this review and no level rests on it." } ], "citation_ledger": [ { "n": 1, "citation": "Radjabzadeh et al., Nature Communications 2022. Gut microbiome-wide association study of depressive symptoms.", "tier": "A", "kind": "primary", "pmid": "36473852", "doi": "10.1038/s41467-022-34502-3", "url": "https://doi.org/10.1038/s41467-022-34502-3", "status_flags": [ "none_found" ], "carries_claims": [ "C1", "C22" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "rotterdam-helius-2022", "access_depth": "full_text_partial", "access_note": "OA XML: cohort restrictions, taxon replication, association results, Mendelian randomization section and limitations" }, { "n": 2, "citation": "Fehse et al., Gut Microbes 2026. From gut to brain: data-driven evidence suggests causal contribution of gut-microbiota to major depressive disorder in humans.", "tier": "A", "kind": "primary", "pmid": "42627212", "doi": "10.1080/19490976.2026.2718589", "url": "https://doi.org/10.1080/19490976.2026.2718589", "status_flags": [ "none_found" ], "carries_claims": [ "C2" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "fehse-mdd-cohort-2026", "access_depth": "full_text_partial", "access_note": "OA XML: 1,269-person design, causal-discovery methods and limitations" }, { "n": 3, "citation": "Kelly et al., Journal of Psychiatric Research 2016. Transferring the blues: Depression-associated gut microbiota induces neurobehavioural changes in the rat.", "tier": "A", "kind": "primary", "pmid": "27491067", "doi": "10.1016/j.jpsychires.2016.07.019", "url": "https://doi.org/10.1016/j.jpsychires.2016.07.019", "status_flags": [ "none_found" ], "carries_claims": [ "C3" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "kelly-transfer-2016", "access_depth": "full_text_partial", "access_note": "University repository primary PDF: donor selection, rat experiment, behavioral results and limitations" }, { "n": 4, "citation": "Siopi et al., Molecular Psychiatry 2023. Gut microbiota changes require vagus nerve integrity to promote depressive-like behaviors in mice.", "tier": "A", "kind": "primary", "pmid": "37131071", "doi": "10.1038/s41380-023-02071-6", "url": "https://doi.org/10.1038/s41380-023-02071-6", "status_flags": [ "none_found" ], "carries_claims": [ "C4" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "siopi-vagus-transfer-2023", "access_depth": "full_text_partial", "access_note": "OA XML and primary publisher HTML: transfer, vagotomy, brain and behavior results" }, { "n": 5, "citation": "Pinto-Sanchez et al., Gastroenterology 2017. Probiotic Bifidobacterium longum NCC3001 Reduces Depression Scores and Alters Brain Activity: A Pilot Study in Patients With Irritable Bowel Syndrome.", "tier": "A", "kind": "rct", "pmid": "28483500", "doi": "10.1053/j.gastro.2017.05.003", "url": "https://doi.org/10.1053/j.gastro.2017.05.003", "status_flags": [ "none_found" ], "carries_claims": [ "C5" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "pinto-ibs-nct01276626", "access_depth": "full_text_partial", "access_note": "OA accepted primary manuscript PDF: prespecified endpoint, microbiota and fMRI results" }, { "n": 6, "citation": "Kazemi et al., Clinical Nutrition 2019. Effect of probiotic and prebiotic vs placebo on psychological outcomes in patients with major depressive disorder: A randomized clinical trial.", "tier": "A", "kind": "rct", "pmid": "29731182", "doi": "10.1016/j.clnu.2018.04.010", "url": "https://doi.org/10.1016/j.clnu.2018.04.010", "status_flags": [ "none_found" ], "carries_claims": [ "C6" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "kazemi-mdd-irct2015092924271n1", "access_depth": "abstract_metadata", "access_note": "PubMed primary abstract only: randomization, completers, BDI and blood outcomes; full text not retrieved" }, { "n": 7, "citation": "Romijn et al., Australian and New Zealand Journal of Psychiatry 2017. A double-blind, randomized, placebo-controlled trial of Lactobacillus helveticus and Bifidobacterium longum for the symptoms of depression.", "tier": "A", "kind": "rct", "pmid": "28068788", "doi": "10.1177/0004867416686694", "url": "https://doi.org/10.1177/0004867416686694", "status_flags": [ "none_found" ], "carries_claims": [ "C7" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "romijn-low-mood-2017", "access_depth": "full_text_partial", "access_note": "OA XML: prespecified measures, ITT and null outcomes" }, { "n": 8, "citation": "Schaub et al., Translational Psychiatry 2022. Clinical, gut microbial and neural effects of a probiotic add-on therapy in depressed patients: a randomized controlled trial.", "tier": "A", "kind": "rct", "pmid": "35654766", "doi": "10.1038/s41398-022-01977-z", "url": "https://doi.org/10.1038/s41398-022-01977-z", "status_flags": [ "none_found" ], "carries_claims": [ "C8" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "schaub-probiotic-mdd-2022", "access_depth": "full_text_partial", "access_note": "OA XML: enrollment, primary ITT/mITT results, taxa and imaging" }, { "n": 9, "citation": "Nikolova et al., JAMA Psychiatry 2023. Acceptability, Tolerability, and Estimates of Putative Treatment Effects of Probiotics as Adjunctive Treatment in Patients With Depression: A Randomized Clinical Trial.", "tier": "A", "kind": "rct", "pmid": "37314797", "doi": "10.1001/jamapsychiatry.2023.1817", "url": "https://doi.org/10.1001/jamapsychiatry.2023.1817", "status_flags": [ "none_found" ], "carries_claims": [ "C9" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "nikolova-probiotic-mdd-2023", "access_depth": "full_text_partial", "access_note": "OA XML: pilot design, 49-person ITT and symptom contrasts" }, { "n": 10, "citation": "Green et al., Canadian Journal of Psychiatry 2023. Feasibility, Acceptability, and Safety of Faecal Microbiota Transplantation in the Treatment of Major Depressive Disorder: A Pilot Randomized Controlled Trial.", "tier": "A", "kind": "rct", "pmid": "36637229", "doi": "10.1177/07067437221150508", "url": "https://doi.org/10.1177/07067437221150508", "status_flags": [ "none_found" ], "carries_claims": [ "C10" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "moving-moods-2023", "access_depth": "full_text_partial", "access_note": "Publisher article read on 2026-09-27: Results, mental-health outcome tables (MADRS, DASS) and blinding assessment; the PMC deposit carries the abstract only" }, { "n": 11, "citation": "Gawlik-Kotelnicka et al., Nutrients 2024. Metabolic Status Influences Probiotic Efficacy for Depression-PRO-DEMET Randomized Clinical Trial Results.", "tier": "A", "kind": "rct", "pmid": "38732635", "doi": "10.3390/nu16091389", "url": "https://doi.org/10.3390/nu16091389", "status_flags": [ "none_found" ], "carries_claims": [ "C11" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "pro-demet-2024", "access_depth": "full_text_partial", "access_note": "OA XML: prespecified MADRS outcome and primary versus secondary results" }, { "n": 12, "citation": "Borkent et al., Schizophrenia Bulletin 2025. Probiotic Formulation for Patients With Bipolar or Schizophrenia Spectrum Disorder: A Double-Blind, Randomized Placebo-Controlled Trial.", "tier": "A", "kind": "rct", "pmid": "39504580", "doi": "10.1093/schbul/sbae188", "url": "https://doi.org/10.1093/schbul/sbae188", "status_flags": [ "none_found" ], "carries_claims": [ "C12" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "borkent-bd-ssd-2025", "access_depth": "full_text_partial", "access_note": "OA XML: randomization, BPRS primary result and gut-marker results" }, { "n": 13, "citation": "Mörkl et al., Gut Microbes 2025. Multi-species probiotic supplement enhances vagal nerve function - results of a randomized controlled trial in patients with depression and healthy controls.", "tier": "A", "kind": "rct", "pmid": "40298641", "doi": "10.1080/19490976.2025.2492377", "url": "https://doi.org/10.1080/19490976.2025.2492377", "status_flags": [ "none_found" ], "carries_claims": [ "C13" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "morkl-hrv-2025", "access_depth": "full_text_partial", "access_note": "OA XML: HRV primary endpoint, mood scores and stool beta-diversity" }, { "n": 14, "citation": "Elahinejad et al., Food Science & Nutrition 2025. Effects of Probiotics as Adjunctive Therapy to Fluoxetine on Depression Severity and Serum Brain-Derived Neurotrophic Factor, Cortisol, and Adrenocorticotropic Hormone in Patients With Major Depressive Disorder: A Randomized, Double-Blind, Placebo-Controlled Trial.", "tier": "A", "kind": "rct", "pmid": "40177327", "doi": "10.1002/fsn3.4698", "url": "https://doi.org/10.1002/fsn3.4698", "status_flags": [ "none_found" ], "carries_claims": [ "C14" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "fluoxetine-probiotic-2025", "access_depth": "full_text_partial", "access_note": "OA XML: randomized design, completion, HDRS-24 and blood biomarkers" }, { "n": 15, "citation": "Li et al., Cell Host & Microbe 2026. Adjunctive fecal microbiota transplantation for major depressive disorder: A randomized, double-blind, placebo-controlled trial.", "tier": "A", "kind": "rct", "pmid": "42309058", "doi": "10.1016/j.chom.2026.05.017", "url": "https://doi.org/10.1016/j.chom.2026.05.017", "status_flags": [ "none_found" ], "carries_claims": [ "C15" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "li-fmt-chictr2300071421", "access_depth": "full_text_partial", "access_note": "Full text read on 2026-09-27 from the authors' institutional PDF: endpoint hierarchy and primary result, randomization and completion, masking check, engraftment analysis and mediation model; the publisher site returned 403/CAPTCHA to the base grading" }, { "n": 16, "citation": "Shekarriz et al., Canadian Journal of Psychiatry 2026. The Safety, Efficacy, and Feasibility of Fecal Microbiota Transplantation in a Population With Bipolar Disorder During Depressive Episodes: A Pilot Parallel Arm Randomized Controlled Trial: Sécurité, efficacité et faisabilité de la transplantation de microbiote fécal chez une population atteinte de troubles bipolaires, au cours d'épisodes dépressifs : essai pilote contrôlé à répartition aléatoire et à groupes parallèles.", "tier": "A", "kind": "rct", "pmid": "41785480", "doi": "10.1177/07067437261420877", "url": "https://doi.org/10.1177/07067437261420877", "status_flags": [ "none_found" ], "carries_claims": [ "C16" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "shekarriz-bipolar-fmt-nct03279224", "access_depth": "full_text_partial", "access_note": "OA XML: 35-person design, MADRS contrast and donor similarity" }, { "n": 17, "citation": "Wang et al., Scientific Reports 2026. A study on the efficacy and safety of fecal microbiota transplantation as an adjunctive therapy for treating depressive episodes.", "tier": "A", "kind": "rct", "pmid": "41826399", "doi": "10.1038/s41598-026-41801-y", "url": "https://doi.org/10.1038/s41598-026-41801-y", "status_flags": [ "none_found" ], "carries_claims": [ "C17" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "wang-fmt-depression-2026", "access_depth": "full_text_partial", "access_note": "OA XML: intervention, attrition, two-week HAMD and limited qPCR flora" }, { "n": 18, "citation": "Zheng et al., Molecular Psychiatry 2016. Gut microbiome remodeling induces depressive-like behaviors through a pathway mediated by the host's metabolism.", "tier": "A", "kind": "primary", "pmid": "27067014", "doi": "10.1038/mp.2016.44", "url": "https://doi.org/10.1038/mp.2016.44", "status_flags": [ "none_found" ], "carries_claims": [ "C18" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "zheng-gf-transfer-2016", "access_depth": "abstract_metadata", "access_note": "PubMed abstract; publisher full text paywalled and no repository copy located on 2026-09-27" }, { "n": 19, "citation": "Valles-Colomer et al., Nature Microbiology 2019. The neuroactive potential of the human gut microbiota in quality of life and depression.", "tier": "A", "kind": "primary", "pmid": "30718848", "doi": "10.1038/s41564-018-0337-x", "url": "https://doi.org/10.1038/s41564-018-0337-x", "status_flags": [ "none_found" ], "carries_claims": [ "C19" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "flemish-gut-flora-2019", "access_depth": "abstract_metadata", "access_note": "PubMed abstract; publisher full text paywalled on 2026-09-27" }, { "n": 20, "citation": "Chen et al., Journal of Affective Disorders 2022. Gut microbiota and major depressive disorder: A bidirectional Mendelian randomization.", "tier": "A", "kind": "primary", "pmid": "35961601", "doi": "10.1016/j.jad.2022.08.012", "url": "https://doi.org/10.1016/j.jad.2022.08.012", "status_flags": [ "none_found" ], "carries_claims": [ "C20" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "chen-mr-mdd-2022", "access_depth": "abstract_metadata", "access_note": "PubMed abstract; publisher full text paywalled on 2026-09-27" }, { "n": 21, "citation": "Zhuang et al., Journal of Neuroinflammation 2020. Associations between gut microbiota and Alzheimer's disease, major depressive disorder, and schizophrenia.", "tier": "A", "kind": "primary", "pmid": "33008395", "doi": "10.1186/s12974-020-01961-8", "url": "https://doi.org/10.1186/s12974-020-01961-8", "status_flags": [ "none_found" ], "carries_claims": [ "C21" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "zhuang-mr-neuropsych-2020", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read on 2026-09-27; the reported effect estimates are taken from it verbatim" }, { "n": 22, "citation": "Asad et al., Nutrition Reviews 2025. Effects of Prebiotics and Probiotics on Symptoms of Depression and Anxiety in Clinically Diagnosed Samples: Systematic Review and Meta-analysis of Randomized Controlled Trials.", "tier": "B", "kind": "meta_analysis", "pmid": "39731509", "doi": "10.1093/nutrit/nuae177", "url": "https://doi.org/10.1093/nutrit/nuae177", "status_flags": [ "none_found" ], "carries_claims": [ "C23" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "asad-synthesis-2025", "access_depth": "abstract_metadata", "access_note": "PubMed abstract; Tier B, used for field framing and the divergence artifact only, with no figure imported" }, { "n": 23, "citation": "Moshfeghinia et al., Journal of Psychiatric Research 2025. The impact of probiotics, prebiotics, and synbiotics on depression and anxiety symptoms of patients with depression: A systematic review and meta-analysis.", "tier": "B", "kind": "meta_analysis", "pmid": "40440772", "doi": "10.1016/j.jpsychires.2025.05.053", "url": "https://doi.org/10.1016/j.jpsychires.2025.05.053", "status_flags": [ "none_found" ], "carries_claims": [ "C23" ], "opened": true, "opened_on": "2026-09-27", "status_checked_on": "2026-09-27", "study_id": "moshfeghinia-synthesis-2025", "access_depth": "abstract_metadata", "access_note": "PubMed abstract; Tier B, used for field framing and the divergence artifact only, with no figure imported" } ], "claim_restatement": { "index_wording": "Gut-brain axis in behavior/psychiatry", "user_wording": "Gut microbial composition causally influences human psychiatric symptoms through gut–brain signalling.", "entry_form": "conjunction", "gradeable_restatement": "Changing human gut microbial composition changes psychiatric symptom severity through gut-to-brain signalling.", "restatement_required": true, "split_into": [ "SC1", "SC2", "SC3", "SC4", "SC5" ], "parent_claim": null, "scope_qualifiers": [ { "qualifier": "human", "determinable_before_outcome": true, "determined_from": "participant species at enrollment", "action": "retained" }, { "qualifier": "psychiatric symptom severity", "determinable_before_outcome": true, "determined_from": "prespecified validated clinical symptom scale", "action": "retained" }, { "qualifier": "gut community composition", "determinable_before_outcome": true, "determined_from": "prespecified fecal community contrast and strain-resolved target-engagement assay", "action": "retained" }, { "qualifier": "gut-to-brain signalling", "determinable_before_outcome": true, "determined_from": "pathway and mediator prespecified before outcomes; mediation must be tested, not selected post hoc", "action": "retained" } ] }, "hypothesis": { "name": "Gut microbial composition and human psychiatric symptoms through gut–brain signalling", "central_claim": "Changing human gut microbial composition changes psychiatric symptom severity through gut-to-brain signalling.", "proposed_mechanism": "A change in gut community membership or function changes neural, endocrine, immune or metabolite signalling to the brain and thereby changes clinical psychiatric symptom severity.", "claimed_scope": "disease-class", "claimed_level": [ "maintenance", "response" ], "is_conjunctive": true, "conjunctive_note": "A human symptom effect and its composition-dependent gut-to-brain mediation must both hold; no reviewed study isolates the complete chain." }, "lay_summary": { "paragraphs": [ "The claim says that changing the mix of gut microbes can change psychiatric symptoms in people by sending a signal from the gut to the brain.", "The full claim receives Γ: a plausible mechanism with partial support. Microbes can change mood-like behavior through a nerve pathway in mice, and some human depression trials report symptom improvement after probiotic or fecal-transfer treatment. Other well-controlled trials show no added symptom benefit, even when gut measures change. In the sharpest test so far, transplanted stool shifted the gut community toward the donor's and symptoms still improved no more than in the comparison group. No reviewed human study proves that a measured community change caused symptom relief through a particular gut-to-brain signal.", "A grade of Γ is a statement about how much causal work has been done, not an accusation against the researchers doing it, and not a verdict that gut microbes are irrelevant to mental health. It says the deciding human experiment — change the gut community, block the signal it is supposed to send, and see whether the symptom effect disappears — has not been run." ], "grade_restated_faithfully": true, "claims_referenced": [ "C3", "C4", "C5", "C8", "C11", "C12", "C15", "C16" ], "jargon_glossed": [ "Γ explained as plausible mechanism with partial support", "fecal transfer described as treatment", "gut community shifted toward the donor's described in ordinary words" ], "no_practical_advice": true, "written_for": "general reader, no field background" }, "is_calibration_entry": { "value": false, "entry_role": "not_calibration", "blind_graded": null, "completed_blinded_calibration": false, "expected_band_sealed_at": "", "target_sha256": "", "target_authorship": "no target band used", "target_author_saw_review": null, "future_run_status": "not prepared", "unblinded_reason": "No expected target band was consulted." }, "rubric_routing": { "primary_rubric": "04-microbiome-systemic.md", "secondary_rubrics": [ "07-psychiatry-neuromodulation.md" ], "index_notes_applied": [ "The indexed topic bundles existence, animal mechanism and human psychiatric generalization.", "The user asks about symptom influence through signalling, not primary initiation of major depression, bipolar disorder or schizophrenia.", "The index prescribes weakest-link. This review departs to scope-matched and records the index result (Β) as dissent rather than discarding it; see subclaim_matrix.index_rule_deviation.", "The index calls defined-consortium rescue Δ-qualifying for animal SC2; the governing rubric places first transfer at Δ and defined rescue with replication at Ζ.", "The index's secondary-rubric list (01 or 05) is not exhaustive. SC3 and SC4 are human psychiatric symptom-rating outcomes, so 07-psychiatry-neuromodulation governs them jointly with 04 under Rule 6 (lowest applicable ceiling)." ], "rubric_fit_note": "The microbiome rubric explicitly covers psychiatric phenotypes and gut-brain behavior and remains primary. Rubric 07 is added as secondary for the two human-trial subclaims because the outcome is a psychiatric symptom rating scale and the comparator problem it governs — expectancy, unblinded raters, open-label design — is live in this evidence base. Its functional-unblinding hard rule does not fire: capsule probiotics and encapsulated or enema-delivered FMT have no unmistakable acute subjective effect, and the two trials that measured masking found it intact. Its open-label rule fires on one source in isolation and is recorded there." }, "subclaim_matrix": { "subclaims": [ { "id": "SC1", "statement": "Human gut community composition correlates reproducibly with depressive symptom states.", "subclaim_type": "existence", "governing_rubric": "04-microbiome-systemic.md", "claimed_scope": "human observational symptom associations", "scope": "human observational symptom associations", "key_evidence_anchor": "independent cohorts with medication and host covariates", "hsef_level": "Β", "confidence": "Moderate", "governing_cap_or_threshold": "Association-only Β ceiling", "rationale": "The Rotterdam/HELIUS replication and Fehse cohort support associations. Cross-sectional stool and symptoms cannot set causal direction [1,2]." }, { "id": "SC2", "statement": "Microbial transfer can alter affect-like behavior through gut-to-brain pathways in rodents.", "subclaim_type": "mechanism", "governing_rubric": "04-microbiome-systemic.md", "claimed_scope": "germ-free and antibiotic-conditioned rodent recipient models", "scope": "germ-free and antibiotic-conditioned rodent recipient models", "key_evidence_anchor": "transfer with pathway interruption and behavioral endpoint", "hsef_level": "Ε", "confidence": "Moderate", "governing_cap_or_threshold": "Context-specific causal effect with limits", "rationale": "Depressed-donor transfer moved anhedonia and anxiety-like readouts in antibiotic-depleted rats [3] and, independently, in germ-free mice colonized with patient microbiota [18]; stress-donor transfer plus subdiaphragmatic vagotomy locates a vagal dependency [4]. Donor state, recipient conditioning and species still bound the inference, and the vagotomy result was replicated only within the originating laboratory." }, { "id": "SC3", "statement": "Microbe-directed treatment changes depressive symptom ratings in some adult clinical settings.", "subclaim_type": "sufficiency", "governing_rubric": "04-microbiome-systemic.md", "claimed_scope": "adult human depressive symptom response under intervention", "scope": "adult human depressive symptom response under intervention", "key_evidence_anchor": "blinded clinical randomization with symptom endpoint and target engagement", "hsef_level": "Δ", "confidence": "Moderate", "governing_cap_or_threshold": "First human interventional phenotype perturbation; no mapped generalization", "rationale": "Blinded randomized contrasts with prespecified symptom outcomes favour the microbial arm in some settings [5,6,14], and the one blinded human transfer trial with verified masking favours FMT on its continuous secondary outcome [15]; that crosses the first human interventional rung. It does not travel into the graded conjunction: SC3 is a treatment-response proposition that a product can satisfy without changing community composition at all [5,13,17], and the prespecified primary outcomes of the largest or best-controlled trials are null [7,10,11,15]." }, { "id": "SC4", "statement": "A changed human gut community alters psychiatric symptoms through a specified gut-to-brain signalling mediator.", "subclaim_type": "mechanism", "governing_rubric": "04-microbiome-systemic.md", "claimed_scope": "human composition-to-mediator-to-symptom causal chain", "scope": "human composition-to-mediator-to-symptom causal chain", "key_evidence_anchor": "randomized composition change plus independent microbial/mediator perturbation and symptom endpoint", "hsef_level": "Γ", "confidence": "Moderate", "governing_cap_or_threshold": "No selective positive human mediation perturbation crosses Δ", "rationale": "Li reports donor engraftment, bile-acid associations and symptom change [15], but the mediator was not independently manipulated and full methods were unavailable; fMRI [5], taxon [8] and HRV [13] cochanges are not causal mediation." }, { "id": "SC5", "statement": "The composition-dependent gut-to-brain symptom effect applies across human psychiatric conditions.", "subclaim_type": "generalization", "governing_rubric": "04-microbiome-systemic.md", "claimed_scope": "major depression, bipolar depression and schizophrenia-spectrum symptom domains", "scope": "major depression, bipolar depression and schizophrenia-spectrum symptom domains", "key_evidence_anchor": "independent target-engaged randomized symptom benefit across diagnostic classes", "hsef_level": "Γ", "confidence": "Moderate", "governing_cap_or_threshold": "Human cross-disorder phenotype generalization not established", "rationale": "Depression findings cannot be generalized to bipolar or schizophrenia symptom severity; randomized BD/SSD probiotic and bipolar FMT contrasts were null on psychiatric symptoms despite gut-marker and donor-similarity change [12,16]. Genetic-instrument work across diagnoses is suggestive at nominal significance only and disagrees with itself on direction [20,21]." } ], "aggregation_rule": "scope-matched", "aggregation_preregistered": true, "aggregation_rationale": "The graded sentence is decided by the two propositions it jointly requires — a human composition change moving symptoms, and a gut-to-brain signal carrying that effect — so the overall letter is taken at the scope of the human mediated chain; the association, rodent and treatment-response subclaims are reported at their own scopes.", "primary_subclaim_id": "SC4", "aggregation_derivation": "SC4 is the subclaim whose scope is the central claim's scope (human composition → mediator → symptom) and it sits at Γ; SC5 at the broader diagnostic scope is also Γ. SC3 reaches Δ at a narrower scope that drops the signalling clause, and SC2 reaches Ε in rodents. Scope-matched therefore returns Γ.", "aggregation_alternatives": { "weakest_link_would_return": "Β", "primary_subclaim_would_return": "Γ", "scope_matched_would_return": "Γ", "note": "REQUIRED disclosure of rule dependence. Weakest-link across all five subclaims returns Β, because SC1 is association-only and universal cap 2 binds it; the index prescribes that rule, so Β is a live reading and is recorded as dissent. Two things argue against adopting it. First, SC1 is not a conjunct the claim needs: a composition change could move symptoms with no detectable cross-sectional association, so letting an observational premise set the letter is pitfall 21, a universal cap re-applied through the weakest evidence class of a set assembled to cover each other's weaknesses. Second, weakest-link restricted to the conjuncts the sentence actually requires — SC3 and SC4 — returns min(Δ, Γ) = Γ, the same letter scope-matched returns. The two-rung swing therefore comes from which subclaims enter the minimum, not from the rule name. Primary-subclaim on SC4 also returns Γ." }, "index_rule_deviation": "The index prescribes weakest-link for 'Gut-brain axis in behavior/psychiatry', whose nominal model is the stronger proposition that dysbiosis is an upstream cause of major depression, anxiety, bipolar disorder and schizophrenia and that correcting it is disease-modifying. The graded sentence here is narrower: whether changing composition changes symptom severity through gut-to-brain signalling. We depart to scope-matched and state the cost openly: the index rule returns Β and this review returns Γ, a two-rung difference. The departure is defended on the necessity analysis in aggregation_alternatives, not on the letter it produces, and the Β reading is carried in the posted edition so a reader who prefers the index rule can find it." }, "competing_models": { "models_evaluated": [ { "rank_in_tournament": 1, "model_id": "M1", "is_nominal": false, "name": "Context-bounded microbial symptom modifier", "core_claim": "Some microbial interventions change depressive symptom ratings as adjuncts, without one established composition or mediator across diagnoses.", "explains_what_nominal_does_not": "Explains positive adjunctive depression signals alongside null primary outcomes and null bipolar/schizophrenia symptom trials.", "strongest_supporting_evidence": [ "5", "9", "14", "15" ], "strongest_contradictory_evidence": [ "7", "10", "11" ], "hsef_level_for_this_model": "Δ", "rationale": "Randomized symptom contrasts support a possible treatment contribution in limited depressive contexts. Several studies lack stool profiling and prespecified primary outcomes are mixed, so no context-defined generalization or pathway is established." }, { "rank_in_tournament": 2, "model_id": "M2", "is_nominal": false, "name": "Stress and illness reshape the gut niche", "core_claim": "Psychiatric state and upstream stress can alter diet, autonomic output and microbial ecology, generating some microbiome associations.", "explains_what_nominal_does_not": "Explains why stressed donors carry altered communities and why cross-sectional taxon patterns need not initiate symptoms.", "strongest_supporting_evidence": [ "1", "2", "4", "16" ], "strongest_contradictory_evidence": [ "5", "15" ], "hsef_level_for_this_model": "Γ", "rationale": "Human cohorts do not order cause and effect, and the Mendelian-randomization analysis inside the largest cohort report was significant with depression as the exposure rather than as the outcome [1]. Stress manipulation in mice can create transferable microbial changes [4]. This rival is credible but cannot dismiss blinded positive human intervention contrasts." }, { "rank_in_tournament": 3, "model_id": "M3", "is_nominal": true, "name": "Human composition-to-brain symptom mechanism", "core_claim": "Changing human gut community composition causes psychiatric symptom change through a specified gut-to-brain signalling mediator.", "explains_what_nominal_does_not": "Would unify microbial exposure, peripheral signal and brain/symptom response across human diagnoses if selectively demonstrated.", "strongest_supporting_evidence": [ "3", "4", "5", "15" ], "strongest_contradictory_evidence": [ "11", "12", "13", "16" ], "hsef_level_for_this_model": "Γ", "rationale": "Rodent causal pathways and human intervention signals are real but occur in different systems. No reviewed human trial independently manipulates a candidate mediator after verified composition change, and target-engaged clinical nulls constrain broad scope." }, { "rank_in_tournament": 4, "model_id": "M4", "is_nominal": false, "name": "Medication and diet ecological confounding", "core_claim": "Treatments and dietary behavior alter stool composition and can create illness-associated taxon differences without microbial symptom causation.", "explains_what_nominal_does_not": "Explains medicated-donor and nonrandomized cohort signatures, and why diet-linked taxa may track symptoms.", "strongest_supporting_evidence": [ "1", "3", "8" ], "strongest_contradictory_evidence": [ "5", "15" ], "hsef_level_for_this_model": "Β", "rationale": "Medication and diet are documented confounding paths in these designs, but this review does not establish that they account for every randomized microbial-treatment effect, and the two taxa a causal-discovery analysis nominated held their difference from controls in unmedicated patients [2]." }, { "rank_in_tournament": 5, "model_id": "M5", "is_nominal": false, "name": "Nonspecific procedure and usual-care improvement", "core_claim": "Expectancy, procedure effects and concurrent psychiatric treatment explain some symptom improvements in microbial-intervention arms.", "explains_what_nominal_does_not": "Explains similar improvement in both FMT arms and an unblinded medication-plus-FMT contrast with differential attrition.", "strongest_supporting_evidence": [ "10", "11", "16", "17" ], "strongest_contradictory_evidence": [ "5", "15" ], "hsef_level_for_this_model": "Γ", "rationale": "Blinded nulls, both-arm improvement in the bipolar FMT trial and an unblinded medication-plus-FMT contrast with differential attrition motivate this rival strongly, and the syntheses that pool these trials do so with near-total between-study heterogeneity [22,23]. It does not fully absorb the controlled positive HAMD signals reported in some blinded trials." } ], "tournament_summary": "The most supported current account is a context-bounded microbial modifier of depressive symptom response (M1, Δ), ahead of stress/illness-first ecology (M2, Γ). The nominal across-diagnosis human mediated mechanism ranks third at Γ. Rank affects role interpretation, not the score.", "study_separating_rank1_from_rank2": "Li et al. 2026: randomized donor FMT versus placebo during escitalopram treatment, with masking measured and intact. It separates a bounded microbial treatment component from a wholly host-first account, but weakly — its prespecified primary endpoint (week-8 remission) was null, the positive HAMD-17 contrast is secondary, and donor engraftment declined between weeks 2 and 8.", "nominal_wins_only_by_funding_artifact": false, "level_unchanged_by_tournament": true }, "evidence_inventory": [ { "citation": "Radjabzadeh et al., Nature Communications 2022", "source_pointer": "PMID 36473852; DOI 10.1038/s41467-022-34502-3; ref 1", "relates_to_subclaims": [ "SC1", "SC5" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Weak", "system": "human observational", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "In 1,054 Rotterdam and 1,539 HELIUS participants after exclusions, 12 genera associated with depressive symptoms in the same direction across cohorts; the design remained observational.", "what_it_does_not_show": "Cross-cohort replication does not establish the direction of microbiome–symptom causation.", "confounders_present": [ "residual diet and host-state confounding", "single-time-point stool" ], "verification": "verified", "verification_note": "OA XML: cohort restrictions, taxon replication and association results", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Fehse et al., Gut Microbes 2026", "source_pointer": "PMID 42627212; DOI 10.1080/19490976.2026.2718589; ref 2", "relates_to_subclaims": [ "SC1", "SC4" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Weak", "system": "human observational", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "A 1,269-person cross-sectional MDD cohort inferred possible Eggerthella and Hungatella causal directions with a graphical algorithm, while the authors stated that observational data cannot definitively establish causality.", "what_it_does_not_show": "Algorithmic causal discovery is not a microbial intervention or a measured signalling pathway.", "confounders_present": [ "cross-sectional direction ambiguity", "unmeasured behavioural and dietary pathways" ], "verification": "verified", "verification_note": "OA XML: 1,269-person design, causal-discovery methods and limitations", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Kelly et al., Journal of Psychiatric Research 2016", "source_pointer": "PMID 27491067; DOI 10.1016/j.jpsychires.2016.07.019; ref 3", "relates_to_subclaims": [ "SC2", "SC4" ], "favors_models": [ "M1", "M3" ], "evidence_class": "Moderate", "system": "animal model", "decisive_perturbation": true, "endogenous": false, "direction": "supports", "what_it_shows": "Stool pooled from three MDD donors versus three controls induced sucrose-preference and anxiety-like differences in antibiotic-depleted rats; forced-swim outcomes were null.", "what_it_does_not_show": "A pooled-donor rat result does not isolate a bacterial taxon, human symptoms, or an endogenous human pathway.", "confounders_present": [ "antibiotic recipient conditioning", "few pooled donors", "donor medication" ], "verification": "verified", "verification_note": "University repository primary PDF: donor selection, rat experiment, behavioral results and limitations", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Siopi et al., Molecular Psychiatry 2023", "source_pointer": "PMID 37131071; DOI 10.1038/s41380-023-02071-6; ref 4", "relates_to_subclaims": [ "SC2", "SC4" ], "favors_models": [ "M1" ], "evidence_class": "Strong", "system": "animal model", "decisive_perturbation": true, "endogenous": false, "direction": "supports", "what_it_shows": "Transfer of stress-altered mouse microbiota induced depressive-like behavior and hippocampal changes in recipient mice; subdiaphragmatic vagotomy abolished the transferred effects.", "what_it_does_not_show": "Vagotomy in antibiotic-conditioned male mice does not establish a vagal mechanism for human psychiatric symptoms or identify a causal strain.", "confounders_present": [ "stress-derived donor material", "antibiotic pretreatment", "vagotomy changes gut physiology" ], "verification": "verified", "verification_note": "OA XML and primary publisher HTML: transfer, vagotomy, brain and behavior results", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Pinto-Sanchez et al., Gastroenterology 2017", "source_pointer": "PMID 28483500; DOI 10.1053/j.gastro.2017.05.003; ref 5", "relates_to_subclaims": [ "SC3", "SC4" ], "favors_models": [ "M1" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "In 44 IBS patients, Bifidobacterium longum improved the two-point HADS-depression response at week 6 (14/22 versus 7/22; P=.04) and changed fMRI responses, but fecal community profiles were similar.", "what_it_does_not_show": "It does not show that a change in gut community composition mediated the mood or brain result; anxiety and IBS symptoms did not improve.", "confounders_present": [ "IBS-selected population", "pilot size", "brain signal not independently manipulated" ], "verification": "verified", "verification_note": "OA accepted primary manuscript PDF: prespecified endpoint, microbiota and fMRI results", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Kazemi et al., Clinical Nutrition 2019", "source_pointer": "PMID 29731182; DOI 10.1016/j.clnu.2018.04.010; ref 6", "relates_to_subclaims": [ "SC3" ], "favors_models": [ "M1" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "Of 110 randomized MDD patients, 81 completed the three-arm trial; probiotic supplementation improved BDI scores versus placebo and prebiotic at eight weeks (P=.042), without gut composition measurement reported in the abstract.", "what_it_does_not_show": "The abstract does not establish an intention-to-treat effect or a composition-mediated gut–brain route.", "confounders_present": [ "29 non-completers", "antidepressant co-treatment", "abstract-only methodological detail" ], "verification": "verified", "verification_note": "PubMed primary abstract only: randomization, completers, BDI and blood outcomes; full text not retrieved", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Romijn et al., Australian and New Zealand Journal of Psychiatry 2017", "source_pointer": "PMID 28068788; DOI 10.1177/0004867416686694; ref 7", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M2", "M5" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "An eight-week double-blind trial in 79 unmedicated people selected for low mood found no probiotic–placebo difference on prespecified psychological outcomes or measured blood markers.", "what_it_does_not_show": "A null for this two-strain product does not rule out every consortium or adjunctive treatment context.", "confounders_present": [ "low-mood sample rather than uniformly diagnosed MDD", "specific formulation" ], "verification": "verified", "verification_note": "OA XML: prespecified measures, ITT and null outcomes", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Schaub et al., Translational Psychiatry 2022", "source_pointer": "PMID 35654766; DOI 10.1038/s41398-022-01977-z; ref 8", "relates_to_subclaims": [ "SC3", "SC4" ], "favors_models": [ "M1" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Of 60 enrolled depressed inpatients, 47 completed the trial; the HAM-D time-by-treatment test was positive only in the compliant subset, while Lactobacillus abundance rose in the probiotic arm.", "what_it_does_not_show": "The reported main ITT interaction was not significant, and the taxon–symptom association does not prove microbial mediation.", "confounders_present": [ "dropout and compliant-subset selection", "baseline HAM-D imbalance", "usual inpatient treatment" ], "verification": "verified", "verification_note": "OA XML: enrollment, primary ITT/mITT results, taxa and imaging", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Nikolova et al., JAMA Psychiatry 2023", "source_pointer": "PMID 37314797; DOI 10.1001/jamapsychiatry.2023.1817; ref 9", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M1" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "In a 49-person adjunctive probiotic pilot, self-reported depression and clinician-rated anxiety favored probiotics at week 8, while clinician-rated depression and GAD-7 did not show clear week-eight differences.", "what_it_does_not_show": "A feasibility-sized symptom signal without a microbiome readout does not establish community composition or signalling mediation.", "confounders_present": [ "pilot efficacy estimates", "multiple symptom scales", "antidepressant co-treatment" ], "verification": "verified", "verification_note": "OA XML: pilot design, 49-person ITT and symptom contrasts", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Green et al., Canadian Journal of Psychiatry 2023", "source_pointer": "PMID 36637229; DOI 10.1177/07067437221150508; ref 10", "relates_to_subclaims": [ "SC3" ], "favors_models": [ "M2", "M5" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "A sham-controlled MDD FMT feasibility pilot randomized 15 adults (10 FMT, 5 placebo) with successful masking; week-8 MADRS was identical between arms (16.6 versus 16.6; Hedges g = 0.05; P = .67) and DASS did not differ (P = .19).", "what_it_does_not_show": "The 10-versus-5 feasibility design is far too small to exclude a moderate clinical effect; it was not powered for symptoms.", "confounders_present": [ "very small pilot", "short follow-up" ], "verification": "verified", "verification_note": "Publisher Results tables and mental-health-outcome text read on 2026-09-27; PMC deposit carries the abstract only.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "PRO-DEMET investigators, Nutrients 2024", "source_pointer": "PMID 38732635; DOI 10.3390/nu16091389; ref 11", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M2", "M5" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "The 116-person PRO-DEMET randomized trial found no probiotic advantage on its primary MADRS change (P=.45), although a secondary minimal-important-change comparison favored probiotics.", "what_it_does_not_show": "A secondary responder threshold cannot reverse the null primary symptom endpoint or identify a microbial mediator.", "confounders_present": [ "heterogeneous metabolic status", "secondary analyses" ], "verification": "verified", "verification_note": "OA XML: prespecified MADRS outcome and primary versus secondary results", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Borkent et al., Schizophrenia Bulletin 2025", "source_pointer": "PMID 39504580; DOI 10.1093/schbul/sbae188; ref 12", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M2", "M5" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "A 131-person bipolar/schizophrenia-spectrum probiotic RCT found no treatment effect on psychiatric symptom severity while serum and fecal gut-barrier markers changed.", "what_it_does_not_show": "Changing intestinal biomarkers without BPRS benefit does not demonstrate a psychiatric gut–brain effect.", "confounders_present": [ "mixed diagnostic groups", "no verified causal taxa", "biomarker validity" ], "verification": "verified", "verification_note": "OA XML: randomization, BPRS primary result and gut-marker results", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Mörkl et al., Gut Microbes 2025", "source_pointer": "PMID 40298641; DOI 10.1080/19490976.2025.2492377; ref 13", "relates_to_subclaims": [ "SC3", "SC4" ], "favors_models": [ "M2", "M5" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "A probiotic trial in depressed patients and controls found changes in some heart-rate-variability measures, but no probiotic effect on depression scores and no overall stool community shift over time.", "what_it_does_not_show": "Heart-rate variability is an autonomic proxy, not a selective test that gut microbes changed symptoms through the vagus.", "confounders_present": [ "antidepressant use", "sleep and smoking effects on HRV", "HRV is nonspecific" ], "verification": "verified", "verification_note": "OA XML: HRV primary endpoint, mood scores and stool beta-diversity", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Fluoxetine adjunct trial, Food Science & Nutrition 2025", "source_pointer": "PMID 40177327; DOI 10.1002/fsn3.4698; ref 14", "relates_to_subclaims": [ "SC3", "SC4" ], "favors_models": [ "M1" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "In a 50-person MDD trial, probiotic plus fluoxetine lowered HDRS-24 more than placebo plus fluoxetine among 44 completers, while cortisol, ACTH and BDNF comparisons were null.", "what_it_does_not_show": "The study did not measure community composition and its endocrine markers did not establish a gut–brain mediator.", "confounders_present": [ "small completer analysis", "co-administered fluoxetine", "no stool profiling" ], "verification": "verified", "verification_note": "OA XML: randomized design, completion, HDRS-24 and blood biomarkers", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Li et al., Cell Host & Microbe 2026", "source_pointer": "PMID 42309058; DOI 10.1016/j.chom.2026.05.017; ref 15", "relates_to_subclaims": [ "SC3", "SC4" ], "favors_models": [ "M1" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "In 60 randomized adults on escitalopram, adjunctive FMT capsules did not improve the prespecified primary endpoint (week-8 remission 43.3% versus 26.7%; P = 0.1760) but did improve the secondary HAMD-17 contrast at week 8 (between-group difference -2.9 points, 95% CI -5.5 to -0.3; P = 0.0303), with masking measured and intact.", "what_it_does_not_show": "It does not show composition-dependent mediation: donor engraftment fell from about 16% newly acquired species at week 2 to about 7% at week 8, and the bile-acid mediation is an observational model linking a taxon to an inflammatory marker, with no bile-acid manipulation.", "confounders_present": [ "escitalopram co-treatment", "post-randomization mediator correlation", "null prespecified primary endpoint", "single-centre" ], "verification": "verified", "verification_note": "Full text read on 2026-09-27 from the authors' institutional PDF: endpoint hierarchy, randomization, masking check, engraftment and mediation model.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Shekarriz et al., Canadian Journal of Psychiatry 2026", "source_pointer": "PMID 41785480; DOI 10.1177/07067437261420877; ref 16", "relates_to_subclaims": [ "SC3", "SC4", "SC5" ], "favors_models": [ "M2", "M5" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "In 35 randomized adults with bipolar depression, allogenic FMT moved stool communities toward donors but did not improve MADRS over autologous FMT (between-group P=.64).", "what_it_does_not_show": "Autologous FMT may itself be active and the pilot is small, so this null cannot exclude a more specific microbial intervention.", "confounders_present": [ "autologous comparator may be active", "attrition", "small pilot" ], "verification": "verified", "verification_note": "OA XML: 35-person design, MADRS contrast and donor similarity", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Wang et al., Scientific Reports 2026", "source_pointer": "PMID 41826399; DOI 10.1038/s41598-026-41801-y; ref 17", "relates_to_subclaims": [ "SC3", "SC4" ], "favors_models": [ "M1", "M5" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "A medication-plus-FMT versus medication-only trial reported greater two-week HAMD-24 reduction, but only 14 of 23 FMT-assigned versus 20 of 23 controls were analyzed and there was no sham procedure.", "what_it_does_not_show": "Differential attrition and an unblinded procedural contrast prevent clean attribution to composition; only ten prespecified bacterial targets were assayed.", "confounders_present": [ "9 versus 3 withdrawals", "procedure expectancy", "no sham FMT" ], "verification": "verified", "verification_note": "OA XML: intervention, attrition, two-week HAMD and limited qPCR flora", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Zheng et al., Molecular Psychiatry 2016", "source_pointer": "PMID 27067014; DOI 10.1038/mp.2016.44; ref 18", "relates_to_subclaims": [ "SC2" ], "favors_models": [ "M1", "M3" ], "evidence_class": "Moderate", "system": "animal model", "decisive_perturbation": true, "endogenous": false, "direction": "supports", "what_it_shows": "Germ-free mice colonized with fecal microbiota from patients with major depressive disorder developed depression-like behaviour relative to mice colonized with healthy-donor microbiota, with disturbed carbohydrate and amino-acid metabolism.", "what_it_does_not_show": "A germ-free recipient is not endogenous human biology, no taxon or consortium is isolated, and the colonization is forced rather than observed.", "confounders_present": [ "germ-free recipient physiology", "crude pooled donor material", "behavioural-despair readouts" ], "verification": "verified", "verification_note": "PubMed abstract read on 2026-09-27; no open-access full text located.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Valles-Colomer et al., Nature Microbiology 2019", "source_pointer": "PMID 30718848; DOI 10.1038/s41564-018-0337-x; ref 19", "relates_to_subclaims": [ "SC1" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Weak", "system": "human observational", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "In a 1,054-person population cohort with validation in 1,070 further individuals, Coprococcus and Dialister were depleted in depression after correction for antidepressant use, and butyrate producers tracked quality-of-life indicators.", "what_it_does_not_show": "Population-scale association with a confounder correction is still association; the neuroactive-potential catalogue is a genomic inference, not a measured signal.", "confounders_present": [ "cross-sectional design", "self-reported quality of life", "residual dietary confounding" ], "verification": "verified", "verification_note": "PubMed abstract read on 2026-09-27; publisher full text paywalled.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Chen et al., Journal of Affective Disorders 2022", "source_pointer": "PMID 35961601; DOI 10.1016/j.jad.2022.08.012; ref 20", "relates_to_subclaims": [ "SC1", "SC4" ], "favors_models": [ "M1", "M3" ], "evidence_class": "Weak", "system": "natural experiment", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "A two-sample bidirectional Mendelian randomization reported protective effects of Actinobacteria (OR 0.88, 95% CI 0.87-0.90) and Bifidobacterium (OR 0.89, 95% CI 0.88-0.91) on major depression, with no effect of depression on composition in the reverse direction.", "what_it_does_not_show": "The Ruminococcus1 and Streptococcaceae results were not supported by MR-Egger, instruments for taxa are weak, and the authors name sample overlap and low power as major limitations; no signalling pathway is tested.", "confounders_present": [ "sub-genome-wide instruments", "sample overlap", "horizontal pleiotropy", "taxon-level exposure definition" ], "verification": "verified", "verification_note": "PubMed abstract read on 2026-09-27; publisher full text paywalled.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Zhuang et al., Journal of Neuroinflammation 2020", "source_pointer": "PMID 33008395; DOI 10.1186/s12974-020-01961-8; ref 21", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Weak", "system": "natural experiment", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "A bidirectional Mendelian randomization across Alzheimer's disease, major depression and schizophrenia reported a suggestive effect of genetically increased Bacilli on major depression (OR 1.07, 95% CI 1.02-1.12; P = 0.010), and found that the disorders in turn altered composition.", "what_it_does_not_show": "Nominal significance on single taxa, no overlap with the taxa another instrument study nominates, and a reverse-direction effect in the same analysis; it does not identify or test a mediator.", "confounders_present": [ "weak instruments", "multiple testing across taxa", "reverse-direction effects in the same analysis" ], "verification": "verified", "verification_note": "PubMed abstract read on 2026-09-27.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "A" }, { "citation": "Asad et al., Nutrition Reviews 2025", "source_pointer": "PMID 39731509; DOI 10.1093/nutrit/nuae177; ref 22", "relates_to_subclaims": [ "SC3" ], "favors_models": [ "M1", "M5" ], "evidence_class": "Weak", "system": "human interventional", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "A systematic review of randomized trials in clinically diagnosed samples pools a large favourable probiotic effect on depressive symptoms and a moderate one on anxiety, reports high heterogeneity, finds no prebiotic effect, and recommends adjunctive microbiome-directed treatment in clinical practice.", "what_it_does_not_show": "A pooled estimate is field framing, not a causal experiment; it carries no number into this review and cannot show that the effect is composition-dependent.", "confounders_present": [ "between-study heterogeneity", "small-study effects", "formulation and population heterogeneity" ], "verification": "verified", "verification_note": "PubMed abstract read on 2026-09-27; used for divergence basis and field framing only.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "B" }, { "citation": "Moshfeghinia et al., Journal of Psychiatric Research 2025", "source_pointer": "PMID 40440772; DOI 10.1016/j.jpsychires.2025.05.053; ref 23", "relates_to_subclaims": [ "SC3" ], "favors_models": [ "M1", "M5" ], "evidence_class": "Weak", "system": "human interventional", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "A second systematic review of the same literature reports a significant pooled reduction in depression and anxiety scores with probiotic, prebiotic or synbiotic supplementation, alongside heterogeneity close to total.", "what_it_does_not_show": "Its pooled estimate is larger than any individual blinded trial's prespecified primary result in this ledger, which is a signal about the synthesis rather than about the effect.", "confounders_present": [ "near-total heterogeneity", "mixed intervention classes", "publication and small-study bias" ], "verification": "verified", "verification_note": "PubMed abstract read on 2026-09-27; used for divergence basis and field framing only.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [] }, "source_tier": "B" } ], "verification_summary": { "load_bearing_claims_checked": 23, "verified": 23, "unverified": 0, "contradicted": 0, "note": "All 17 inherited claims were re-checked against their records in the red-team pass and all 17 stand; three were re-verified at greater depth than the base grading reached (Li 2026 and Green 2023 from full text, Kelly 2016 from the repository PDF) and two were rewritten to match what the full text says rather than what the abstract said. Six added claims were verified on first entry. Verification of a reported result is not verification of its causal interpretation." }, "factual_issues_found": [ { "quoted_text": "Li's randomized FMT abstract reports donor engraftment, symptom-score change and bile-acid mediation analysis [15].", "problem": "The abstract's word 'durable' for engraftment was carried into the review; the full text reports engraftment declining between weeks 2 and 8, and the primary endpoint (week-8 remission) was null rather than merely 'not different'.", "source_says": "Week-8 remission 43.3% versus 26.7%, P = 0.1760 (prespecified primary); HAMD-17 week-8 difference -2.9 points (95% CI -5.5 to -0.3), P = 0.0303 (secondary); newly acquired donor species about 16% at week 2 falling to about 7% at week 8; glycocholic acid mediated 27% of the association between L. eligens and CASP-8, with no bile-acid manipulation.", "source_pointer": "PMID 42309058; DOI 10.1016/j.chom.2026.05.017", "affects_score": false, "action": "scored-on-corrected-understanding" }, { "quoted_text": "PRO-DEMET investigators, Nutrients 2024.", "problem": "A trial acronym is used in place of the authors, so the row cannot be matched to the record by author.", "source_says": "Gawlik-Kotelnicka O, Margulska A, Pleśka K, Skowrońska A, Strzelecki D. Nutrients 2024;16(9):1389.", "source_pointer": "PMID 38732635; DOI 10.3390/nu16091389", "affects_score": false, "action": "logged" }, { "quoted_text": "Fluoxetine adjunct trial, Food Science & Nutrition 2025.", "problem": "A description is used in place of the authors, so the row cannot be matched to the record by author.", "source_says": "Elahinejad V, Khorasanian AS, Tehrani-Doost M, et al. Food Sci Nutr 2025;13(4):e4698.", "source_pointer": "PMID 40177327; DOI 10.1002/fsn3.4698", "affects_score": false, "action": "logged" }, { "quoted_text": "Radjabzadeh et al. replicated 12 depression-associated genera across the Rotterdam and HELIUS cohorts ... the observational design cannot order the gut and mood.", "problem": "Accurate as far as it goes, but the same report contains a Mendelian-randomization analysis whose significant result runs with depression as the exposure, which cuts against the nominal model and was omitted.", "source_says": "With major depression as the exposure, the inverse-variance-weighted MR effect on Eggerthella was 0.237 (P = 0.027); with the microbiome as exposure the significant estimate was inconsistent with the cohort association, and the authors state the MR lacked power because instruments were taken below genome-wide significance.", "source_pointer": "PMID 36473852; DOI 10.1038/s41467-022-34502-3", "affects_score": false, "action": "scored-on-corrected-understanding" } ], "cap_rules_applied": [ { "subclaim_id": "SC1", "cap": "Β", "cap_source": "universal", "triggered": true, "trigger_text_quoted": "Pure association evidence — however large, however replicated — cannot exceed Β.", "specific_trigger": "SC1 uses cross-sectional human taxon–symptom associations [1,2], without intervention.", "resulting_ceiling": "Β" }, { "subclaim_id": "SC4", "cap": "Γ", "cap_source": "rubric", "triggered": true, "trigger_text_quoted": "Association plus mechanistic in vitro support, or host-pathway plausibility, caps at Γ.", "specific_trigger": "The complete human community-to-pathway-to-symptom chain has only co-moving stool, imaging, bile-acid, inflammation or HRV measures [5,8,13,15]; the pathway is not independently perturbed.", "resulting_ceiling": "Γ" }, { "subclaim_id": "SC5", "cap": "Γ", "cap_source": "universal", "triggered": true, "trigger_text_quoted": "Biomarker change without functional/phenotypic change caps at Γ.", "specific_trigger": "Gut biomarkers or donor similarity moved in BD/SSD and bipolar trials without psychiatric symptom advantage [12,16]; those studies cannot lift cross-diagnostic generalization.", "resulting_ceiling": "Γ" }, { "subclaim_id": "SC3", "cap": "Γ", "cap_source": "rubric", "triggered": false, "trigger_text_quoted": "Open-label, single-arm, retrospective, or naturalistic evidence alone. A positive open-label result never crosses Δ, however large.", "specific_trigger": "07-psychiatry-neuromodulation.md, evaluated under Rule 6. The rule fires on Wang [17] standing alone, which had no sham and an unblinded comparison. It does not bind SC3, because the subclaim also rests on blinded randomized contrasts with prespecified symptom outcomes [5,6,14] and on a blinded transfer trial whose masking was measured and intact [15]. The functional-unblinding rule is likewise not triggered: these interventions have no unmistakable acute subjective effect.", "resulting_ceiling": "Γ" }, { "subclaim_id": "SC4", "cap": "Γ", "cap_source": "rubric", "triggered": true, "trigger_text_quoted": "Mediation-by-correlation rule. Candidate mediator correlated with outcome, never manipulated.", "specific_trigger": "07-psychiatry-neuromodulation.md, evaluated under Rule 6 and concordant with the 04 ceiling already binding SC4. Every human mediator in this ledger — fMRI response [5], Lactobacillus abundance [8], heart-rate variability [13], serum bile acids and inflammatory markers [15] — is correlated with outcome and none is manipulated.", "resulting_ceiling": "Γ" } ], "cap_precedence_resolutions": [ { "collision": "Rubric in-vivo transfer threshold Δ versus universal forced-transfer Γ cap for endogenous propagation", "subclaim_scope_cap_binds_on": "SC4 and SC5, spontaneous human composition-mediated symptom change and across-diagnosis generalization", "subclaim_scope_threshold_binds_on": "SC2, behavior and vagal pathway in the actually manipulated rodent recipient", "resolution": "Kelly and Siopi cross the transfer threshold for recipient rodents. Their forced transfers do not by themselves establish endogenous human gut–brain symptom mediation.", "note": "The human microbial-intervention response SC3 is separately scored from a direct human randomized contrast; the rodent transfer is not being imported as that contrast." }, { "collision": "Two rubrics govern the same subclaims: 04-microbiome-systemic Δ (transfer paradigm in vivo) and 07-psychiatry-neuromodulation Δ (controlled human trial with a credible active comparator and a pre-registered primary outcome)", "subclaim_scope_cap_binds_on": "SC3 and SC4, human psychiatric symptom-rating outcomes, where 07's comparator and mediation rules bind", "subclaim_scope_threshold_binds_on": "SC2, rodent recipients, where only 04 applies", "resolution": "Rule 6: a cross-rubric hypothesis scores at the lowest applicable ceiling. Under 04, human FMT is a transfer paradigm and crosses Δ on a positive symptom result; under 07, the same trials must also carry a credible comparator and a pre-registered primary outcome. SC3 satisfies both on the blinded trials with positive prespecified symptom outcomes, so Δ stands. SC4 fails both rubrics' mediator requirements identically, so Γ stands with no conflict to resolve.", "note": "Recorded rather than resolved silently; the index's secondary-rubric list did not name 07 and was extended per Step 5." } ], "boundary_check_BGD": [ { "subclaim_id": "SC1", "field_valid_perturbation_quote": "Gnotobiotic or transfer experiment showing phenotype movement, at minimum", "qualifying_evidence_present": "no", "qualifying_evidence_citation": "1, 2", "gamma_hard_rules_triggered": [ "Association-only Β cap triggered for this observational existence proposition." ], "disease_phenotype_relevance": "no", "perturbation_result_positive": "no", "perturbation_result_note": "No microbial intervention in the human cohorts; the proposition is association-only.", "ceiling_from_boundary_check": "Γ" }, { "subclaim_id": "SC2", "field_valid_perturbation_quote": "Gnotobiotic or transfer experiment showing phenotype movement, at minimum", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "3, 4", "gamma_hard_rules_triggered": [ "'Any framework that has not crossed the transfer/rescue threshold in vivo' does not trigger within rodent recipient scope; transfers were positive." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "Anhedonia/anxiety-like assays changed after transfer [3,4], and vagotomy suppressed the stress-donor effect [4].", "ceiling_from_boundary_check": "Δ or higher" }, { "subclaim_id": "SC3", "field_valid_perturbation_quote": "Gnotobiotic or transfer experiment showing phenotype movement, at minimum", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "5, 6, 14, 15", "gamma_hard_rules_triggered": [ "No Γ hard rule blocks a human randomized treatment-response subclaim once a phenotype-positive intervention is credited; effects remain mixed." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "Clause 5 applied explicitly. The best-controlled human transfer trial returned a null prespecified primary endpoint [15], and three further blinded trials were null on their primary symptom outcomes [7,10,11]; those do not cross Δ and are not counted as if they did. Δ is carried by the blinded trials whose prespecified symptom outcomes were positive [5,6,14] together with the positive secondary contrast in the masked transfer trial [15].", "ceiling_from_boundary_check": "Δ or higher" }, { "subclaim_id": "SC4", "field_valid_perturbation_quote": "Gnotobiotic or transfer experiment showing phenotype movement, at minimum", "qualifying_evidence_present": "no", "qualifying_evidence_citation": "5, 8, 13, 15, 16", "gamma_hard_rules_triggered": [ "'Association plus mechanistic in vitro support, or host-pathway plausibility, caps at Γ' applies to the human pathway attribution; no independent mediator intervention is present." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "no", "perturbation_result_note": "The human treatment effect can be positive without proving that an observed taxon, fMRI, bile-acid or HRV signal carried it. The sharpest available test of the composition clause is negative: allogenic FMT moved the recipient community toward the donor's while autologous FMT did not, and the symptom trajectories were the same [16]. A field-valid design returning a null does not cross Δ (clause 5).", "ceiling_from_boundary_check": "Γ" }, { "subclaim_id": "SC5", "field_valid_perturbation_quote": "Gnotobiotic or transfer experiment showing phenotype movement, at minimum", "qualifying_evidence_present": "no", "qualifying_evidence_citation": "12, 16, 21", "gamma_hard_rules_triggered": [ "'Biomarker change without functional/phenotypic change caps at Γ' fires in the BD/SSD and bipolar donor-composition contrasts." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "no", "perturbation_result_note": "Relevant cross-diagnosis trials showed no added psychiatric symptom benefit, despite gut marker or donor-profile changes.", "ceiling_from_boundary_check": "Γ" } ], "thresholds_tested": [ { "subclaim_id": "SC1", "threshold": "Δ", "requirement_quoted": "Gnotobiotic or transfer experiment showing phenotype movement, at minimum", "satisfied": "no", "supporting_evidence": [], "what_would_cross_it": "Manipulate a prespecified associated community feature and move a symptom phenotype; the existence association itself remains Β." }, { "subclaim_id": "SC2", "threshold": "Δ", "requirement_quoted": "Gnotobiotic or transfer experiment showing phenotype movement, at minimum", "satisfied": "yes", "supporting_evidence": [ "3", "4" ], "what_would_cross_it": "Already crossed within recipient rodents." }, { "subclaim_id": "SC2", "threshold": "Ε", "requirement_quoted": "Requires context-specific causal effect with defined limits. A transferable phenotype exists, but microbial identity or generalizability remains incomplete.", "satisfied": "yes", "supporting_evidence": [ "3", "4", "18" ], "what_would_cross_it": "Already crossed: recipient conditioning, donor state and vagal integrity delimit the effect, and the phenotype transfers into both antibiotic-depleted rats and germ-free mice [3,4,18]." }, { "subclaim_id": "SC2", "threshold": "Ζ", "requirement_quoted": "Defined consortium or strain-level causal evidence; Rescue logic; Replication across at least two relevant models or cohorts", "satisfied": "no", "supporting_evidence": [ "3", "4", "18" ], "what_would_cross_it": "Subtract and restore a named microbial function, rescue the behavioural and brain endpoints, and replicate the same pathway outside the originating laboratory. The rubric's own falsification pattern — a defined consortium failing to recapitulate the crude-community effect — has not been tested for this phenotype, so rescue logic is absent rather than satisfied." }, { "subclaim_id": "SC3", "threshold": "Δ", "requirement_quoted": "At least one endogenous in vivo (or equivalent human interventional / natural-experiment) perturbation shows the mechanism contributes causally to the claimed phenotype.", "satisfied": "yes", "supporting_evidence": [ "5", "6", "14", "15" ], "what_would_cross_it": "Already crossed for bounded human depressive treatment response, with uncertainty around individual results." }, { "subclaim_id": "SC3", "threshold": "Ε", "requirement_quoted": "Context-specific causality established. The mechanism is causally operative in a defined context, subtype, stage, or model — but generalization to the full claimed scope fails.", "satisfied": "no", "supporting_evidence": [ "7", "8", "10", "11", "15", "16" ], "what_would_cross_it": "Replicate a prespecified formulation or donor-function response in an independently randomized depressed cohort with verified engagement, and map where it stops with the same intervention. Mapping requires the same product; the present nulls use different products in different populations, so the boundary is unmapped rather than mapped (pitfall 16)." }, { "subclaim_id": "SC4", "threshold": "Δ", "requirement_quoted": "Requires a transfer paradigm or causal perturbation in vivo.", "satisfied": "no", "supporting_evidence": [ "5", "8", "13", "15" ], "what_would_cross_it": "Randomize a composition-changing intervention and independently perturb the prespecified microbial signal or host receptor, with clinical symptom change tracking the functional pathway." }, { "subclaim_id": "SC5", "threshold": "Δ", "requirement_quoted": "Requires a transfer paradigm or causal perturbation in vivo.", "satisfied": "no", "supporting_evidence": [ "12", "16" ], "what_would_cross_it": "Show prespecified symptom improvement with verified microbial engagement in independent bipolar or schizophrenia-spectrum trials, followed by a common mediator test." }, { "subclaim_id": "SC1", "threshold": "Β", "requirement_quoted": "Pure association evidence — however large, however replicated — cannot exceed Β.", "satisfied": "yes", "supporting_evidence": [ "1", "2", "19" ], "what_would_cross_it": "Not a threshold to cross but a ceiling reached; three independent population-scale cohorts replicate the association and none of them can order it." } ], "generalization_check": { "generalization_tested": "yes", "context_boundary_mapped": "yes", "internal_negative_controls": [ "Kelly donor transfer changed sucrose preference and anxiety-like readouts but forced-swim outcomes were null [3].", "Siopi vagotomy removed stress-microbiota behavioural, neurogenic and neuroinflammatory effects in antibiotic-conditioned mice [4].", "In humans, the PRO-DEMET primary score, the BD/SSD psychiatric score and the bipolar allogenic-versus-autologous score were null [11,12,16].", "In the bipolar trial the community moved toward the donor in the allogenic arm and not in the autologous arm, and the symptom trajectories were the same — the closest thing in the ledger to an internal control on the composition clause itself [16]." ], "note": "Rodent causal scope is bounded and merits Ε, and it now spans two recipient systems in two laboratories [3,4,18]. Human depressive treatment response has mixed results from different products and settings, which does not yet map a stable product-specific Ε boundary. Cross-disorder generalization remains Γ." }, "falsification_patterns_met": [ { "subclaim_id": "SC4", "pattern": "Community moves as intended and the phenotype does not", "evidence": "Allogenic FMT shifted recipient communities toward the donor and raised diversity at six months while autologous FMT did not, and MADRS improvement was the same in both arms (16.74 versus 15.4 points; P = .64) [16]. This is the ledger's most direct test of the composition clause and it is negative.", "downgrade_applied": true }, { "subclaim_id": "SC4", "pattern": "Phenotype fails to move despite gut marker or donor-profile change", "evidence": "Borkent gut-barrier biomarkers changed with no BPRS treatment effect [12]; Shekarriz donor similarity changed with no MADRS difference [16].", "downgrade_applied": true }, { "subclaim_id": "SC3", "pattern": "Blinded replication of a putative benefit fails", "evidence": "Romijn and PRO-DEMET primary mood outcomes were null for tested formulations [7,11]; they do not falsify every bacterial product.", "downgrade_applied": true }, { "subclaim_id": "SC4", "pattern": "Proposed mediator moves while clinical phenotype does not", "evidence": "Mörkl HRV measures shifted, whereas depressive scores did not differ between treatment arms [13].", "downgrade_applied": true }, { "subclaim_id": "SC2", "pattern": "Defined consortium fails to recapitulate crude transfer", "evidence": "No direct defined-consortium recapitulation of Kelly or Siopi transfer was identified in this reviewed set.", "downgrade_applied": false }, { "subclaim_id": "SC3", "pattern": "Phenotype moves without the claimed community change", "evidence": "Depression scores improved in the probiotic arm with fecal profiles similar between arms [5]; symptom scores improved after FMT while only one genus differed by discriminant analysis and the authors report no change in overall community structure [17]; and a probiotic changed vagal parameters with no change in overall community composition [13]. A benefit that does not require a measured composition change does not test the composition-dependent claim (pitfall 13).", "downgrade_applied": false } ], "pitfalls_flagged": [ { "subclaim_id": "SC1", "pitfall": "Dysbiosis treated as mechanism", "where": "Large depression cohorts [1,2] show association, not an intervened cause." }, { "subclaim_id": "SC3", "pitfall": "Crude FMT treated as specific evidence", "where": "Li and Wang use mixed donor material; neither isolates a causal taxon or strain [15,17]." }, { "subclaim_id": "SC4", "pitfall": "Abundance treated as causality", "where": "Schaub Lactobacillus changes and Li donor engraftment are not independent mediation tests [8,15]." }, { "subclaim_id": "SC4", "pitfall": "Biomarker change treated as phenotype mechanism", "where": "Mörkl HRV and Borkent zonulin changes occurred without depressive or psychiatric symptom advantage [12,13]." } ], "causal_tally_audit": { "tally_claimed_by_document": 0, "tally_after_audit": 0, "items_rejected": [ "Human positive symptom trials do not jointly isolate composition and a signalling mediator [5,6,8,9,14,15].", "Rodent transfer and vagotomy cross a pathway threshold only within recipient rodents [3,4,18].", "Li reports statistical bile-acid mediation of a taxon-to-inflammatory-marker association, with no bile-acid manipulation and a null prespecified primary endpoint [15].", "Bipolar and schizophrenia-spectrum trials do not show the predicted psychiatric symptom movement despite gut measures changing [12,16].", "Mendelian-randomization results are nominal, taxon-level, mutually inconsistent and bidirectional; they do not count as a causal line for the mediated claim [1,20,21].", "Pooled estimates from systematic reviews are field framing, not causal items [22,23]." ], "note": "Recount of the human transfer paradigm specifically, since that is the design the rubric names: four randomized human FMT trials are in the ledger [10,15,16,17]. Two are null on their symptom comparisons [10,16], one is positive but unblinded, with nine of 23 assigned participants not analysed in the transplant arm [17], and one is blinded with a null prespecified primary and a positive secondary [15]. The audited number of studies showing the complete sequence — measured human community change, independently tested gut-to-brain mediator, psychiatric symptom effect — remains zero." }, "upper_ladder_check": { "final_level_below_zeta": true, "equivalents_considered": [ { "kind": "mendelian_randomization", "available": true, "why_not_sufficient": "Two-sample instrument studies exist and are nominally positive in places [20,21], and the largest cohort report ran its own MR [1]. They do not carry the claim: instruments for taxa are selected below genome-wide significance and explain little variance, the nominated taxa do not overlap between studies, one study's Egger analysis does not support its own hits, the reverse direction is significant in two of the three analyses, and none of them addresses the signalling clause at all." }, { "kind": "defined_strain_rescue", "available": true, "why_not_sufficient": "Vagotomy tests pathway necessity in mouse transfer [4], but a named microbial function was not subtracted and restored with replicated human symptom effects." }, { "kind": "therapeutic_perturbation_hard_endpoint", "available": true, "why_not_sufficient": "Clinical depression trials include randomized symptom endpoints [5–17], yet the exact human mediated chain is not isolated and several prespecified outcomes are null." }, { "kind": "natural_experiment", "available": false, "why_not_sufficient": "No selective human ecological exposure establishes composition-to-pathway-to-symptom ordering independent of stress, diet and medication." }, { "kind": "dose_response_across_populations", "available": false, "why_not_sufficient": "Replicated taxon–symptom associations [1] do not provide an intervention-anchored dose-response for a verified microbial function." }, { "kind": "withdrawal_rechallenge", "available": false, "why_not_sufficient": "No selective withdrawal and reinstatement of a specified bacterial function with reversible clinical symptoms and mediator readouts was identified." } ], "convergence_present": true, "cap_reapplied_to_convergent_set": false, "note": "Rodent transfer and pathway surgery, human randomized treatment results and human genetic instruments converge on a possible biological contribution, but not on one identified human composition-dependent signalling chain, and the genetic leg points both ways. No universal cap has been re-applied to the convergent set; the level rests on the absence of a positive mediated human test, not on capping the weakest leg." }, "steelman": { "case_one_rung_higher": "The Δ case, written first and at its strongest. A blinded randomized human transfer trial with masking measured and intact reports greater HAMD-17 reduction at weeks 2 and 8, donor strain acquisition, and a mediation model running from a donor-enriched taxon through bile acids to an inflammatory marker [15]; a second transfer trial reports a larger two-week HAMD-24 reduction [17]; blinded probiotic trials report positive prespecified symptom outcomes [5,6,14]; a vagotomy experiment shows the route is neural in mice, in two recipient systems and two laboratories [3,4,18]; and genetic instruments point the same way for some taxa [20,21]. On that reading the human phenotype perturbation has been made and the mediator has been identified.", "case_one_rung_lower": "The Β case. Most human taxon links are observational [1,2,19]; the prespecified primary outcomes of the largest and best-controlled trials are null [7,10,11,12,15,16]; several positive probiotic trials never measured composition [6,9,14]; the syntheses that pool them report heterogeneity close to total [22,23]; and the MR inside the largest cohort was significant with depression as the exposure [1]. SC5 could be argued at Β on the same basis without changing the aggregate, since scope-matched lands on SC4.", "which_rejected": "Δ is rejected for the exact human mediated claim. The blocker is not thinness of evidence but a specific negative result plus a specific missing experiment: the one trial that demonstrably changed the recipient community relative to its comparator produced no additional symptom improvement [16], and no human study has ever interrupted a candidate gut-to-brain mediator while leaving the microbial intervention in place. Β is rejected because blinded human symptom perturbations and a mapped rodent pathway are more than association, and letting an association-only premise set the letter is pitfall 21.", "blocking_cap_threshold_or_missing_experiment": "For SC4: the 04 Γ hard rule 'Association plus mechanistic in vitro support, or host-pathway plausibility, caps at Γ' together with the 07 mediation-by-correlation rule, and boundary-check clause 5 applied to the negative allogenic-versus-autologous contrast [16]. For SC5: replicated target-engaged symptom benefit outside depression, which does not exist." }, "final_score": { "hsef_level": "Γ", "hsef_descriptor": "Mechanistically plausible, partial system support", "confidence": "Moderate", "confidence_rationale": "The evidence base is sizeable and genuinely two-sided: four randomized human transfer trials, nine randomized supplementation trials, three population-scale cohorts, two instrument studies and two rodent recipient systems. It is not dense where it matters — the trials are small, the products differ, the prespecified primary outcomes disagree with the pooled estimates, and no human mediator experiment exists. Moderate rather than High for that reason, and not Low because the key negative results are now replicated across products, diagnoses and delivery routes.", "rests_on_non_peer_reviewed_work": false, "rationale": "Rodent gut-brain causality reaches Ε across two recipient systems and some human depressive-treatment effects reach Δ. A human composition-dependent signalling mediator has not been established, the one human contrast that demonstrably changed the community relative to its comparator produced no extra symptom benefit, and bipolar/schizophrenia generalization is unsupported; the exact conjunction stays Γ." }, "role_of_mechanism": { "classification": "reinforcing / stabilizing factor", "derivation_step_matched": 5, "derivation": "Human mood symptoms are not shown to require one microbiome state, nor to be initiated or rate-limited by it, so steps 3 and 4 fail. Step 2 — epiphenomenon, no demonstrated causal dependency in either direction — was considered and rejected: randomized microbial interventions do move symptom scores in some settings [5,6,14,15], which is more than co-occurrence. Step 5 matches: some microbial interventions amplify or reduce symptom response alongside concurrent treatment. Tournament rank did not assign the role.", "tournament_used_as_tiebreaker": false }, "extended_scoring_MSD": { "momentum": { "value": "Mechanistic only", "artifact": "Li 2026 randomized adjunctive FMT, Shekarriz 2026 allogenic-versus-autologous FMT, the 2025 Mörkl vagal-function trial and the Siopi 2023 vagotomy transfer model", "justification": "Active trials and pathway work exist and are getting better controlled — masking is now measured, autologous comparators are being used — but no trial identified here independently manipulates a human signalling mediator while leaving a verified community change in place, which is the experiment the boundary check names." }, "stability": { "value": "Stable", "trigger": "No identified pending, independently mediator-perturbing human trial of a verified microbial community change and prespecified psychiatric symptom endpoint.", "justification": "Replication of Li with accessible prespecified analyses could move the depressive-treatment subclaim; it would not alone prove the full human pathway. A target-engaged mediator trial would move SC4 upward, while reproducible nulls would move the claim down." }, "divergence": { "implied_level": "Ε", "implied_level_basis": "Read from the field's own framing rather than from the claim's wording. Two 2025 systematic reviews of these trials report a large favourable pooled effect of probiotics on depressive symptoms and recommend that the gut microbiome be monitored and treated alongside pharmacological treatment [22,23]; primary reports in this ledger are titled and abstracted in causal and mechanistic language ('induces', 'require vagus nerve integrity', 'causal contribution'). A competent reader of that literature would infer context-specific human causality, Ε.", "assessed_level": "Γ", "gap_in_rungs": 2, "value": "Substantial overstatement", "justification": "On Α→Β→Γ→Δ→Ε, Ε is two rungs above the assessed Γ, so divergence is +2, substantial overstatement. A reader who takes the syntheses' clinical recommendation as implying converged causal lines would name Ζ and get +3; Ε is the conservative reading and is the one adopted. The gap concerns the complete human mediated claim, not the bounded animal result or all possible microbial treatments." }, "calibration_note": "M/S/D are reviewer judgments and do not modify HSEF level or confidence." }, "proposed_decisive_experiments": [ { "name": "Blinded defined-function depression trial with mediator contrast", "experiment": "Randomize adults with MDD on stable standard therapy to a sequence-defined, strain-resolved microbial intervention or matched control. Predefine the functional microbial output and a feasible second randomization that changes that output while keeping exposure and expectancy matched.", "readout": "Prespecified clinician-rated depressive symptom change, absolute strain engagement, gut and serum mediator time course, and brain/autonomic signal.", "critical_control": "Independent mediator perturbation or isogenic functional contrast, identical concomitant care, blinded allocation, adequate power, intention-to-treat analysis and microbiological target engagement.", "falsifying_result": "Community engraftment and mediator shift without differential symptoms, or symptom change with no mediator dependence.", "supporting_result": "Replicated symptom benefit that disappears when the prespecified microbial signalling function is selectively disabled and returns when restored." }, { "name": "Cross-diagnosis and reverse-direction factorial test", "experiment": "In separately powered bipolar-depression and schizophrenia-spectrum cohorts, randomize the same engaged microbial function versus matched control while prospectively measuring stress, diet, medication and gut transit; add a prespecified stress/diet intervention arm to test host-first causation.", "readout": "Diagnosis-specific psychiatric symptom scales, absolute microbial composition and function, candidate mediator and time-ordered host-state measures.", "critical_control": "Sham-equivalent procedures, stable psychiatric treatment, predeclared interaction test and correction for multiple diagnoses and symptom domains.", "falsifying_result": "Durable matched microbial engagement without symptom movement, or apparent effects fully accounted for by the host-state arm.", "supporting_result": "Consistent clinical benefit and mediator dependence in each diagnostic cohort despite reciprocal host-state controls." } ], "knowledge_gaps": [ { "name": "Functional exposure", "statement": "Which measured bacterial function, rather than a broad taxon count, reaches a human signalling compartment after intervention?", "relevance": "Composition alone does not identify a mediator." }, { "name": "Mediator necessity", "statement": "Does an independently changed bile-acid, immune or neural pathway abolish a clinical symptom effect?", "relevance": "This is the missing Δ test for the exact through-signalling clause." }, { "name": "Target-engaged clinical heterogeneity", "statement": "Which baseline communities, medications and diagnoses permit reproducible symptom benefit with the same product?", "relevance": "Mixed trial results cannot currently define an Ε context boundary." }, { "name": "Reverse direction", "statement": "How much of the microbial association is produced by psychiatric state, stress, diet and treatment?", "relevance": "This determines whether observational signatures can be read causally." } ], "change_record": { "applies": true, "prior_level": "Γ", "prior_confidence": "Moderate", "prior_source": "Grading A, OpenAI Codex, HSEF v5.4 schema, 2026-09-27 (archived unaltered in sources/A_v5.4_grading/)", "derived_before_consulting_prior": false, "new_level": "Γ", "direction": "held", "what_moved_it": "Held at Γ, and the reason it holds is now a named negative result rather than an absence. The red team re-derived the level from the caps after obtaining full text for the three sources the base grading could not reach and adding six sources in both directions. Nothing moved a subclaim: SC1 Β, SC2 Ε, SC3 Δ, SC4 Γ, SC5 Γ all stand. What changed is the support: SC2's scope widened to two recipient systems in two laboratories after Zheng 2016 was added; SC3's Δ is now explicitly carried by the blinded trials with positive prespecified outcomes rather than by the trial whose prespecified primary was null; SC4's Γ now rests on a positive falsification pattern — the allogenic-versus-autologous contrast that moved the community and not the symptoms — rather than only on the absence of a mediator experiment.", "divergence_from_other_reviewers": "Three disagreements with the base grading are recorded rather than absorbed. (1) Aggregation: the base grading departed from the index-prescribed weakest-link rule with the justification that the index rule was drafted for a stronger claim. The departure is upheld but re-argued on pitfall 21 and on a necessity analysis, and the index rule's answer, Β, is carried in the posted edition as dissent. (2) Divergence: the base grading derived its implied level from the claim's own wording and said so, which is not what the framework's divergence procedure asks for; the implied level is re-derived from the field's syntheses and lands on the same +2. (3) Rubric routing: the base grading declared no secondary rubric; 07-psychiatry-neuromodulation is added for the two human-trial subclaims under Rule 6, which changes no letter. Two factual corrections were also made against the base grading's abstract-only reading of Li 2026." }, "consensus_validation": { "completed": true, "status": "documented_historical_discovery_and_reconciliation", "provider": "Consensus", "provider_mode": "Historical browser/connector use as recorded in the source artifacts", "local_audit_date": "2026-10-09", "primary_comparison_count": null, "reported_usage": [], "grade_changed": false, "score_before": { "hsef_level": "Γ", "hsef_descriptor": "Mechanistically plausible, partial system support", "confidence": "Moderate", "confidence_rationale": "The evidence base is sizeable and genuinely two-sided: four randomized human transfer trials, nine randomized supplementation trials, three population-scale cohorts, two instrument studies and two rodent recipient systems. It is not dense where it matters — the trials are small, the products differ, the prespecified primary outcomes disagree with the pooled estimates, and no human mediator experiment exists. Moderate rather than High for that reason, and not Low because the key negative results are now replicated across products, diagnoses and delivery routes.", "rests_on_non_peer_reviewed_work": false, "rationale": "Rodent gut-brain causality reaches Ε across two recipient systems and some human depressive-treatment effects reach Δ. A human composition-dependent signalling mediator has not been established, the one human contrast that demonstrably changed the community relative to its comparator produced no extra symptom benefit, and bipolar/schizophrenia generalization is unsupported; the exact conjunction stays Γ." }, "score_after": { "hsef_level": "Γ", "hsef_descriptor": "Mechanistically plausible, partial system support", "confidence": "Moderate", "confidence_rationale": "The evidence base is sizeable and genuinely two-sided: four randomized human transfer trials, nine randomized supplementation trials, three population-scale cohorts, two instrument studies and two rodent recipient systems. It is not dense where it matters — the trials are small, the products differ, the prespecified primary outcomes disagree with the pooled estimates, and no human mediator experiment exists. Moderate rather than High for that reason, and not Low because the key negative results are now replicated across products, diagnoses and delivery routes.", "rests_on_non_peer_reviewed_work": false, "rationale": "Rodent gut-brain causality reaches Ε across two recipient systems and some human depressive-treatment effects reach Δ. A human composition-dependent signalling mediator has not been established, the one human contrast that demonstrably changed the community relative to its comparator produced no extra symptom benefit, and bipolar/schizophrenia generalization is unsupported; the exact conjunction stays Γ." }, "independent_blinded_audit": false, "raw_receipts": [ { "path": "deliverables\\prepublication-enrichment-2026-10-05\\website-series\\gut-brain-psychiatric\\consensus-browser-links.json", "sha256": "1f5b39a52e773276f8a60cffb7d46b2b4055a39e91892f1c38c47514fa610e0e" }, { "path": "deliverables\\prepublication-enrichment-2026-10-05\\website-series\\gut-brain-psychiatric\\consensus-browser.txt", "sha256": "4fb86556163c162aeb6ef125f23b84c973d6abfdef3916dc6b37bb731718a30a" }, { "path": "deliverables\\prepublication-enrichment-2026-10-05\\website-series\\gut-brain-psychiatric\\consensus-targeted-check.txt", "sha256": "84655d8e79743feb404e405d3306aaf62d7d818679a88de0a59a6b22e010cb08" } ], "limitations": "This verifies documented historical use and links the record to the final edition. It adds no primary-source, full-text, independence or verification credit. Original access remains limited to the source-specific record.", "existing_primary_reconciliation": { "slug": "gut-brain-psychiatric", "consensus_used": true, "rawproof": [ "deliverables/prepublication-enrichment-2026-10-05/website-series/gut-brain-psychiatric/consensus-browser.txt", "deliverables/prepublication-enrichment-2026-10-05/website-series/gut-brain-psychiatric/consensus-targeted-check.txt", "deliverables/prepublication-enrichment-2026-10-05/website-series/gut-brain-psychiatric/consensus-browser-links.json" ], "reconciliation_existing": "Yes; 8 timeline identifiers and selected mouse Results passage (vagotomy/sham prior to transfer) in dated update; limited model claim.", "primary_sources_matched": [ "27067014", "27491067", "28068788", "28483500", "36473852", "37131071", "39504580", "42309058" ], "limitations": "Keep separate gut-brain-psychiatric-disease run out. Animal perturbation does not establish human mediator." } } }