{
  "slug": "sauna-cardiovascular-prevention",
  "file": "sauna-cardiovascular-prevention.review.md (shipped review)",
  "authoring_model": "Sol 6.1",
  "adversarial_reviewer": "Claude Opus 5.5 (Anthropic)",
  "date": "2026-11-02",
  "level_before": "Γ",
  "level_after": "Γ",
  "confidence_before": "Moderate",
  "confidence_after": "Moderate",
  "pmids_checked": 57,
  "pmids_failed": 0,
  "claims_spot_checked": 21,
  "findings": [
    {
      "id": "S1",
      "severity": "Major",
      "location": "SC4/SC5 subclaims; Competing models Rank 2; rubric07 routing; AT A GLANCE best-supported rival; References 31-56; scorecard final_score.confidence_rationale",
      "original": "Randomized exercise-plus-sauna35785965 demonstrates that heat can add measured effects beyond exercise allocation, whereas sham-controlled TREAT-AF32192678 demonstrates a different exact electrical intervention on objective AF burden. These distinguish real scoped effects from a universal interchangeable wellness narrative; neither identifies sauna mortality prevention. | Multiple independent studies across 57 primary reports ...",
      "corrected": "Randomized exercise-plus-sauna35785965 shows that heat can add measured blood-pressure and estimated-fitness effects beyond exercise allocation, and the CKD crossover42113470 shows acute vascular and walking gains. Neither tests hard cardiovascular events, so they identify scoped physiological benefit, not sauna event prevention. | SC4, SC5 and Rank 2 moved to a separately registered claim or prefixed 'Out-of-scope companion; not evidence for or against the registered sauna claim.' | Confidence rationale counts the 31 heat/sauna reports (refs 1-30, 57), noting the hard-event evidence is one Finnish cohort family plus one matched Waon cohort.",
      "evidence": "Registered claim (tools/hypothesis_register.json; scorecard claim_restatement, split_into empty) is sauna-only. 16 inventory records relate only to SC5 and 10 only to SC4. HSEF competing-model rule requires alternatives to explain the same core phenomenon; the cold/vagal model explains no sauna-event observation. Lay box cites AF device trial PMID 32192678. Headline level unaffected."
    },
    {
      "id": "S2",
      "severity": "Major",
      "location": "AT A GLANCE Causal role; Final score paragraph; scorecard role_of_mechanism; published reviews/sauna-cardiovascular-prevention.html",
      "original": "epiphenomenon (lay gloss: 'This travels alongside the process without causing it or being caused by it.')",
      "corrected": "Not determined for hard events: sauna causally changes blood pressure and some heart-failure function in randomized trials, but whether it reduces heart attacks, strokes or cardiovascular deaths has not been tested causally. (Exact category from the five-value vocabulary to be lead-adjudicated; epiphenomenon should not stand.)",
      "evidence": "Own derivation: 'no causal dependency ... established in either direction' = undetermined, not 'no causal role'. Own Β-rejection cites controlled perturbation beyond co-occurrence; framework assigns epiphenomenon when support is only associational. RCTs: PMID 35785965 SBP -8.0 mmHg (95% CI -14.6 to -1.4) vs exercise alone; PMID 15564698 PVC reduction. ROLE_PLAIN gloss in tools/build_v5_reviews.py appears verbatim on the published page."
    },
    {
      "id": "S3",
      "severity": "Minor",
      "location": "Subclaim matrix SC1; Competing models Rank 1",
      "original": "SC1 Sauna heat exposure changes cardiovascular physiology. (Γ) while rival M3 'Controlled physiological and restricted functional benefit' is Ε",
      "corrected": "SC1 Sauna-induced physiological changes mediate a reduction in cardiovascular events. (Γ retained) — or keep the statement and grade Ε",
      "evidence": "SC1 rationale grades the prevention-mediation inference, not the stated physiology; the literal statement is met by randomized human interventions (PMIDs 35785965, 42113470, 39004886), which the review grades Ε for M3. Headline unaffected."
    },
    {
      "id": "S4",
      "severity": "Minor",
      "location": "Cap audit SC2 cap; Adversarial steelman Β-rejection sentence",
      "original": "SC2 cap: 'Biomarker change without functional/phenotypic change caps at Γ.' | 'Β is rejected for the assembled mechanistic proposition because controlled human thermal perturbation provides partial system support beyond co-occurrence.'",
      "corrected": "SC2 ceiling from boundary check (no field-valid hard-event perturbation: Γ) plus universal cap 2 binding the KIHD event evidence at Β | 'Β is rejected for SC2 because its own scope contains randomized human thermal interventions with disease-relevant functional and rhythm endpoints in heart failure (WAON-CHF secondary NYHA and 6MWD; Holter PVC reduction) and a matched, non-randomized Waon cohort with fewer hard events; these exceed co-occurrence, but none is a field-valid hard-event perturbation, so Γ holds.'",
      "evidence": "SC2 contains functional change (PMIDs 27001189, 15564698) so the quoted biomarker cap is not literally triggered; steelman concedes the clinical endpoint alone might be Β although it is the preregistered primary subclaim. Re-derived on SC2's own evidence, Γ stands."
    },
    {
      "id": "S5",
      "severity": "Minor",
      "location": "Evidence and causal interpretation, KIHD paragraph; Knowledge gaps item 1",
      "original": "(no mention of incident non-fatal events or of any non-Finnish cohort)",
      "corrected": "Add: KIHD incident stroke HR 0.38 (0.18-0.81) for 4-7 vs 1 session/week (PMID 29720543), an added endpoint, not an independent cohort; independent Japanese JPHC cohort of 30,076 adults, near-daily tub bathing HR 0.72 (0.62-0.84) for incident CVD, no association with sudden cardiac death (PMID 32209614), different modality and observational. Knowledge gap amended accordingly.",
      "evidence": "Verified in PubMed: Kunutsor 2018 Neurology DOI 10.1212/WNL.0000000000005606 (PubMed article type 'Editorial' despite cohort abstract); Ukai 2020 Heart DOI 10.1136/heartjnl-2019-315752. Registered claim covers clinical events; all SC2 cohort sources are fatal-event analyses. No level change."
    },
    {
      "id": "S6",
      "severity": "Minor",
      "location": "Cap audit, SC1-SC4 'Single-species or single-strain evidence' lines (5 in review, 6 in scorecard)",
      "original": "Any hamster mechanistic component remains model-scoped; human clinical trials and cohorts are evaluated separately.",
      "corrected": "Not triggered: no animal-model source is in the inventory; all 57 records are human studies.",
      "evidence": "Template bleed; scorecard evidence_inventory system = human interventional (52) or human observational (5); no hamster source."
    },
    {
      "id": "S7",
      "severity": "Minor",
      "location": "reader-summary.json limitations",
      "original": "57 primary reports with 3 selected partial-body examinations; density, heterogeneous controls and incomplete access set the confidence, not correctness.",
      "corrected": "Confidence reflects evidence density and reproducibility: multiple heat/sauna studies with partial replication but heterogeneous comparators, modalities and endpoints; reviewer access depth does not set the band.",
      "evidence": "Prior correction A2 fixed this sentence in the scorecard on the policy that access depth is not a confidence criterion; the lay-facing summary was not updated."
    },
    {
      "id": "S8",
      "severity": "Editorial",
      "location": "scorecard review_metadata.reviewer",
      "original": "Codex",
      "corrected": "Sol 6.1 (or a reconciling note)",
      "evidence": "Review banner and approval section state the authoring model is Sol 6.1."
    },
    {
      "id": "S9",
      "severity": "Editorial",
      "location": "CONSENSUS_CROSSCHECK.md and CONSENSUS_CROSSCHECK.html (4 instances each)",
      "original": "the2315-men; absenceIRR0.71; underlyingTable2; OriginalTREAT-AF; secondaryNCT04143269",
      "corrected": "the 2315-men; absence IRR 0.71; underlying Table 2; Original TREAT-AF; secondary NCT04143269",
      "evidence": "Known spacing defect; fixed earlier in review.md and video-insert.md (A1, A3) but not in these public files. No value changed."
    }
  ],
  "rejected_findings": [
    "Tally 27 discriminating contrast units vs 28 flagged records: [53] and [54] share NCT04143269, so 28 records = 27 study units. Correct.",
    "Decisive perturbations: 16 printed equal 16 flagged in the scorecard.",
    "Primary-subclaim aggregation returns SC2 Γ; weakest-link Β (SC3) and scope-matched Γ correctly reported; Δ correctly withheld (no field-valid hard-event perturbation).",
    "Cold-shower retrospective registration and unverifiable compliance confirmed in PMC full text (PMC5025014; NTR5183 registered 25 June 2015 prior to analysis, after recruitment). Resolves the prior pass's residual.",
    "Waon prognosis numbers (64 vs 65; 31.3% vs 68.7%; 8 vs 12 deaths; matched, randomisation not stated) and the percentage-point vs relative-reduction statement are correct.",
    "WAON-CHF BNP primary described as null: abstract says the BNP change did not reach statistical significance. Correct.",
    "Momentum 'Mechanistic only' and Stability 'Stable' correct; 41831305 says a controlled trial is being planned.",
    "Rubric06 lifespan wording (SC1-SC4) and rubric07 psychiatry rules (SC5) are a declared adaptation, not undisclosed bleed; the SC4/SC5 problem is scope (S1).",
    "Reference years within ±1 (13 instances) and the ref 12 subscript: not errors.",
    "No collective-author or empty-DOI defect."
  ],
  "lead_policy_applied": {
    "confidence_bands": "Access depth not a determinant; Moderate stands, but rationale must count only in-scope heat/sauna reports (S1, S7).",
    "momentum": "Value 'Mechanistic only'; planned HFpEF trial not mislabelled as underway.",
    "collective_authors": "Checked; none found.",
    "role_of_mechanism": "'epiphenomenon' contradicts the review's own derivation and Β-rejection; replacement category requires lead adjudication (S2)."
  },
  "validator": "Not re-run in this pass; recorded package validator PASS - 0 fail, 0 warn, 30 pass",
  "scorecard_changed": false,
  "related_files": {
    "published_page": "reviews/sauna-cardiovascular-prevention.html renders the S2 epiphenomenon gloss",
    "note": "The review embeds an earlier adversarial approval dated 2026-10-14 (A1-A3); this pass is independent and did not edit it."
  },
  "residual_limitations": [
    "Spot checks mostly at abstract depth; one paper read in PMC full text.",
    "Omitted-literature search targeted, not systematic.",
    "S2 replacement category depends on a lead decision about the role vocabulary."
  ]
}