{ "$schema_version": "5.4", "review_mode": "fresh", "claim_restatement": { "index_wording": "Elevated LDL particle concentration is a causal driver of atherosclerotic cardiovascular disease", "entry_form": "bundled", "gradeable_restatement": "In humans, a higher cumulative circulating concentration of LDL-class apoB-containing particles increases atherosclerotic cardiovascular disease risk by increasing the flux and retention of those particles in the arterial intima; interventions that reduce this exposure should reduce atherosclerotic events.", "restatement_required": true, "split_into": [ "SC1: arterial retention and lesion mechanism", "SC2: lifelong human genetic causality", "SC3: pharmacologic reversibility and clinical-event response", "SC4: particle-number specificity versus cholesterol cargo", "SC5: cross-context generalization and boundary mapping" ], "parent_claim": null, "scope_qualifiers": [ { "qualifier": "LDL-class apoB-containing particles", "determinable_before_outcome": true, "determined_from": "direct LDL-P measurement or a prespecified apoB/LDL-C intervention or genetic contrast, with non-LDL apoB particles identified where relevant", "action": "replaced with measurable antecedent" }, { "qualifier": "higher cumulative circulating concentration", "determinable_before_outcome": true, "determined_from": "concentration difference multiplied by exposure duration or a prespecified lifelong genetic proxy, rather than by who develops disease", "action": "replaced with measurable antecedent" }, { "qualifier": "atherosclerotic cardiovascular disease", "determinable_before_outcome": true, "determined_from": "prespecified coronary or major atherosclerotic endpoints; nonatherosclerotic cardiovascular death is separated", "action": "replaced with measurable antecedent" } ] }, "lay_summary": { "paragraphs": [ "Overall grade: Ζ (zeta), multiple independent causal lines converge; confidence is Moderate. The grade needs a careful reading, because two questions sit inside one sentence. The first question is whether LDL-type particles in the blood cause atherosclerotic heart and artery disease. The evidence for that is strong. People who inherit lifelong lower LDL have much less coronary disease, several unrelated drug classes that lower LDL reduce heart attacks and related events, and plaque imaging moves in the predicted direction. On their own, those parts of the claim grade at Η or Θ.", "The second question is what 'particle concentration' adds: whether the number of particles, rather than the cholesterol they carry, is the causal quantity. Here the evidence is real but unsettled. Genetic studies, large cohorts, and drug trials in which LDL cholesterol and particle count moved apart all agree that LDL cholesterol alone is not the whole story. But independent genetic analyses disagree about whether particle number beats total cholesterol content. They also show that particles of different types carry different risk per particle: remnant particles and lipoprotein(a) are more harmful per particle than LDL. Because the review's rules, fixed before scoring, require every part of the claim to hold, the unsettled particle-number part sets the overall grade.", "In plain terms, LDL-class particles are a primary cause of atherosclerotic disease. What is not settled is whether counting particles captures that cause better than measuring the cholesterol they carry. LDL is also not the only cause: blood pressure, smoking, diabetes, inflammation, plaque rupture and clotting, lipoprotein(a), and other illnesses all contribute." ], "grade_restated_faithfully": true, "claims_referenced": [ "C2", "C3", "C4", "C5", "C7", "C8", "C9", "C13", "C14", "C15", "C16", "C17", "C18", "C21", "C22", "C24", "C26", "C27", "C28" ], "jargon_glossed": [ "LDL-type particles", "particle concentration", "remnant particles", "lipoprotein(a)", "atherosclerotic" ], "no_practical_advice": true, "written_for": "an educated nonspecialist assessing causal evidence" }, "hypothesis": { "name": "LDL particle concentration as a causal driver of ASCVD", "central_claim": "Higher cumulative LDL particle exposure causally increases atherosclerotic cardiovascular disease.", "proposed_mechanism": "More circulating LDL-class apoB particles produce more arterial entry and proteoglycan-mediated retention, which initiates and propagates plaque; lowering exposure reduces lesion progression and atherosclerotic events.", "claimed_scope": "human ASCVD initiation and progression", "claimed_level": [ "initiation", "progression", "response" ], "is_conjunctive": true, "conjunctive_note": "The wording joins arterial mechanism, human causality, intervention reversibility, particle-number specificity, and broad ASCVD scope; each is scored separately." }, "rubric_routing": { "primary_rubric": "06-metabolism-aging.md", "secondary_rubrics": [ "05-immunology-inflammation.md (governs competing models M4 and the inflammatory component of M5 only)" ], "index_notes_applied": [ "No hypothesis-index entry exists for LDL/ASCVD; a stub is recommended (primary 06, secondary 05, SC1-SC5 decomposition, particle-specific conjunct preregistered explicitly).", "All seven rubrics (01-07) were checked in the red-team pass; none covers atherosclerosis or lipidology. Rubric 06 is the closest fit for the nominal lipid-exposure claim; rubric 05 caps bind inflammation claims and do not lower the nominal under the cross-field lowest-ceiling rule." ], "rubric_fit_note": "The supplied metabolism/aging rubric is an imperfect literal fit because its Δ language refers to lifespan and healthspan. Its upper-ladder section explicitly names Mendelian randomization, several pharmacological classes, dose-response, and lifelong natural experiments as field-valid equivalents. This review therefore applies those exact equivalents to prespecified hard ASCVD and plaque endpoints and states the adaptation rather than importing an unstated cardiovascular rubric." }, "subclaim_matrix": { "subclaims": [ { "id": "SC1", "statement": "LDL-class apoB particles enter and become retained in the arterial intima, initiating and propagating atherosclerotic lesions.", "subclaim_type": "mechanism", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "arterial initiation and progression", "scope": "Atherosclerotic lesion biology in arteries exposed to circulating LDL-class apoB particles; clinical plaque rupture and thrombosis are treated as later, partly separable processes.", "key_evidence_anchor": "Proteoglycan-binding perturbation, human lesion biology, and randomized plaque-trajectory imaging.", "hsef_level": "Η", "confidence": "Moderate", "governing_cap_or_threshold": "Robust cross-context intervention evidence and mechanistic specificity satisfy Η; species and surrogate limitations prevent Θ for this subclaim alone.", "rationale": "Arterial-retention perturbation reduces lesions in vivo [1], and intensive human lowering changes coronary atheroma trajectory [16]. Held at Η under the convergence rule because human genetics and trials establish that LDL-class exposure drives lesions; confidence Moderate because the retention step itself rests on one cited, single-species perturbation read at abstract depth." }, { "id": "SC2", "statement": "Inherited differences that create lifelong differences in LDL-related exposure change coronary and atherosclerotic risk in the predicted direction and approximately according to cumulative exposure.", "subclaim_type": "causality", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "human lifelong exposure", "scope": "Human variants in multiple lipid-handling genes with coronary or atherosclerotic outcomes; developmental pleiotropy and ancestry are retained as limitations.", "key_evidence_anchor": "PCSK9, NPC1L1, HMGCR, LDLR/APOB, multi-gene natural experiments, and FH-gene sequencing [2-6,38].", "hsef_level": "Θ", "confidence": "High", "governing_cap_or_threshold": "High reproducibility with few directional exceptions meets Θ for LDL-related cumulative exposure, without claiming that LDL-P is the only causal metric.", "rationale": "Multiple genes, ancestries, and effect durations produce consistent direction and a strong duration gradient; FH-mutation carriers have higher risk at any observed LDL-C with greater cumulative exposure [38]. Pleiotropy cannot plausibly explain the cross-target pattern, although the variants usually index LDL-C and apoB together. The clinically diagnosed FH odds ratio [6] is inflated by DLCN criteria that include premature CAD and is supportive only." }, { "id": "SC3", "statement": "Reducing LDL-related particle exposure through independent pharmacologic mechanisms reduces hard atherosclerotic cardiovascular events and can slow or reverse plaque progression.", "subclaim_type": "reversibility", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "human intervention response", "scope": "Primary and secondary prevention and non-dialysis CKD, with achieved exposure change, duration, baseline risk, event definition, and competing mortality specified before outcomes.", "key_evidence_anchor": "Statin, ezetimibe, PCSK9-antibody, bempedoic-acid, CKD, CETP-inhibitor and plaque-imaging randomized trials.", "hsef_level": "Η", "confidence": "High", "governing_cap_or_threshold": "Independent drug classes and endpoint concordance satisfy Η; the evacetrapib null [26], unresolved within SC3's scope (a particle-based explanation is plausible but not shown in the trial's own data), keeps the response claim below Θ. Advanced dialysis lies outside SC3's stated scope.", "rationale": "HMGCR, NPC1L1, PCSK9 and ATP-citrate-lyase interventions converge on fewer atherosclerotic events [7-14]. The CETP class is concordant when indexed by apoB/non-HDL-C (REVEAL rate ratio 0.91 with non-HDL-C -18%) but evacetrapib is a null under LDL-C indexing (HR 1.01 after a median 26 months) [25-27]. AURORA and 4D map a late-disease boundary [22,23]." }, { "id": "SC4", "statement": "When LDL particle number or apoB and LDL-C are discordant, causal and prospective risk tracks particle burden more closely than cholesterol mass per particle.", "subclaim_type": "measurement-specific mechanism", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "particle-number specificity", "scope": "Human LDL-P, apoB, LDL-C, and non-HDL-C discordance or multivariable genetic contrasts; apoB is explicitly recognized as counting LDL plus other atherogenic particles.", "key_evidence_anchor": "Multivariable and pathway-contrast MR, observational discordance cohorts, and pharmacological discordance (CETP inhibitors, pemafibrate) that separate particle burden from LDL-C, non-HDL-C, or remnant cholesterol.", "hsef_level": "Ζ", "confidence": "Moderate", "governing_cap_or_threshold": "Rubric 06 Ζ ('Reproducible intervention effects; Mechanistic narrowing; Evidence not reducible to general sickness, caloric reduction, or developmental artifacts') is met; Η ('Robust cross-context intervention evidence; Mechanistic specificity') and the framework's 'no major unresolved ... contradiction within the claimed scope' are not met, because no intervention separates particle number from total cholesterol cargo and independent groups report contradictions [20,35,36].", "rationale": "Against LDL-C as the unique causal quantity the evidence converges across MR [17-19,25], CETP trials [26,27] and discordance cohorts [21,30,31,33]. Against total cholesterol cargo (non-HDL-C) it is split: apoB-favoring MR [17-19] and cohort data [31,32] and PROMINENT [29] versus a cholesterol-favoring MR [20], higher per-particle atherogenicity of remnants and Lp(a) [35,36], and cohorts that cannot separate apoB from non-HDL-C [21,33,34]. Alternative models are narrowed but not excluded." }, { "id": "SC5", "statement": "The LDL-related causal effect generalizes across major human ASCVD contexts, while absolute benefit and event timing vary with cumulative exposure, baseline risk, competing pathways, and competing mortality.", "subclaim_type": "generalization", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "major human ASCVD contexts", "scope": "Coronary and major atherosclerotic outcomes across primary prevention, established ASCVD, and CKD; nonatherosclerotic events and universal individual prediction are outside the claim.", "key_evidence_anchor": "Primary- and secondary-prevention outcome trials, CKD trials, genetic ancestry contrasts, and late-dialysis boundary trials.", "hsef_level": "Η", "confidence": "High", "governing_cap_or_threshold": "Broad human generalization with an explicitly mapped dialysis boundary meets Η, while residual causes and late-stage nulls preclude Θ for unrestricted cardiovascular outcomes.", "rationale": "Directionally concordant benefit occurs across prevention settings and treatment mechanisms. The absolute effect depends on baseline risk, duration, competing events, and how much established disease remains modifiable." } ], "aggregation_rule": "weakest-link", "aggregation_preregistered": true, "aggregation_rationale": "The submitted wording requires arterial causality, human directionality, response to lowering, and the specific particle-concentration interpretation. SC1-SC4 therefore gate the overall claim; SC5 can lower it if generalization fails.", "primary_subclaim_id": "", "aggregation_derivation": "SC1=Η; SC2=Θ; SC3=Η; SC4=Ζ; SC5=Η. Preregistered weakest-link across SC1-SC4 returns Ζ; SC5 does not lower it.", "aggregation_alternatives": { "weakest_link_would_return": "Ζ", "primary_subclaim_would_return": "Ζ if SC4 (the content distinguishing this entry) is primary; Η if SC3 is primary", "scope_matched_would_return": "Η (SC5 matches the human-ASCVD scope; SC2's Θ is confined to genetically indexed LDL-class exposure)", "note": "A one-rung spread driven entirely by the particle-number conjunct. The preregistered weakest-link rule is applied unchanged; the scope-matched reading ('LDL-class particles cause ASCVD') is reported, not substituted." }, "index_rule_deviation": "" }, "competing_models": { "models_evaluated": [ { "rank_in_tournament": 1, "model_id": "M2", "is_nominal": false, "name": "Broader apoB-particle retention model", "core_claim": "Cumulative exposure to all arterial-retained apoB-containing particles is the causal lipid burden; LDL usually supplies most particles but is not the only atherogenic class.", "explains_what_nominal_does_not": "It accommodates LDL, remnants, and lipoprotein(a), and explains why apoB can outperform LDL-C when cholesterol content per particle differs.", "strongest_supporting_evidence": [ "retention perturbation [1]", "LPL/LDLR per-apoB genetic contrast [17]", "multivariable MR [18,19]", "CETP genetics [25]", "apoB-only association in UK Biobank and trial cohorts [32]", "cross-class outcome trials [7-15]" ], "strongest_contradictory_evidence": [ "Per-particle atherogenicity differs by class: remnant-rich SNP cluster OR 1.76 vs 1.33 per SD apoB [35]; Lp(a) ~6.6-fold LDL per particle [36].", "Helgadottir et al. found non-HDL-C improved an apoB model whereas apoB did not improve a non-HDL-C model [20]." ], "hsef_level_for_this_model": "Η", "rationale": "Mechanism, lifelong genetics and independent interventions converge; graded Η rather than Θ because unequal per-particle atherogenicity [35,36] and a cholesterol-favoring MR [20] are unresolved exceptions to apoB count as the complete causal burden." }, { "rank_in_tournament": 2, "model_id": "M1", "is_nominal": true, "name": "LDL-particle concentration model", "core_claim": "The cumulative number of circulating LDL particles drives atherogenesis through arterial entry and retention.", "explains_what_nominal_does_not": "It directly interprets LDL-P discordance and captures why equal LDL-C can represent different numbers of artery-entering particles.", "strongest_supporting_evidence": [ "LDL retention perturbation [1]", "LDL-P discordance in MESA and Framingham [30,31]", "apoB-prioritizing MR [17-19]" ], "strongest_contradictory_evidence": [ "apoB includes non-LDL particles that are causal [17,35,36]", "cholesterol-favoring MR [20]", "WHS non-HDL-C discordance at least as informative as LDL-P [21]", "most outcome trials move LDL-C, apoB and LDL-P together" ], "hsef_level_for_this_model": "Ζ", "rationale": "LDL-related causality is secure, but the distinctive particle-number-over-cargo content is contested; matches the overall Ζ." }, { "rank_in_tournament": 3, "model_id": "M3", "is_nominal": false, "name": "Cholesterol-cargo model", "core_claim": "The cholesterol carried into and retained within the arterial wall, better represented by non-HDL-C, is more causally important than particle number itself.", "explains_what_nominal_does_not": "It explains the Helgadottir multivariable genetic result and the strong dose relation between achieved cholesterol lowering and outcome benefit.", "strongest_supporting_evidence": [ "cholesterol-favoring MR [20]", "higher per-particle atherogenicity of cholesterol-rich remnants [35]", "LDL-C-scaled trial meta-analysis [15]" ], "strongest_contradictory_evidence": [ "CETP genetics and trials disfavor LDL-C-indexed cargo [25-27]", "apoB-prioritizing MR [17-19] and apoB-only association [32]", "PROMINENT: remnant cholesterol fell without apoB and without benefit [29]" ], "hsef_level_for_this_model": "Ζ", "rationale": "The LDL-C-only version is disfavored; the non-HDL-C (total cargo) version remains a live competitor supported by one direct MR discriminator." }, { "rank_in_tournament": 4, "model_id": "M5", "is_nominal": false, "name": "Plaque-instability and thrombosis model", "core_claim": "Inflammation, plaque rupture or erosion, and thrombosis dominate the conversion of existing plaque into an acute clinical event.", "explains_what_nominal_does_not": "It explains residual events despite low LDL and the limited response of mixed late-stage dialysis endpoints.", "strongest_supporting_evidence": [ "IL-1beta inhibition reduced recurrent events without lipid lowering (CANTOS) [39]", "residual event rates in outcome trials [9-13]" ], "strongest_contradictory_evidence": [ "it does not explain why lifelong LDL exposure and multiple lowering mechanisms consistently change plaque burden and atherosclerotic events" ], "hsef_level_for_this_model": "Ζ", "rationale": "Graded Ζ for the event transition as assessed on this ledger (antithrombotic trials not reviewed); complementary to, not a replacement for, LDL-driven plaque formation." }, { "rank_in_tournament": 5, "model_id": "M4", "is_nominal": false, "name": "Inflammation or endothelial-injury-first model", "core_claim": "Inflammation or endothelial dysfunction initiates atherosclerosis, while LDL is only a permissive substrate or downstream amplifier.", "explains_what_nominal_does_not": "It highlights immune amplification and nonlipid residual risk.", "strongest_supporting_evidence": [ "CANTOS: inflammation causally operative in recurrent events [39]", "persistent residual risk despite intensive lowering [9-13]" ], "strongest_contradictory_evidence": [ "LDL-related genetic exposure and several lowering mechanisms change disease without a shared anti-inflammatory intervention [2-16,38]" ], "hsef_level_for_this_model": "Ε", "rationale": "Inflammation is causally operative in a defined context (recurrent events) [39], but the claim that LDL is merely permissive fails against cross-target genetic and randomized lipid-lowering evidence; governed by rubric 05." } ], "tournament_summary": "The broader apoB-particle retention model ranks first (Η) because non-LDL apoB particles are causal [17,35,36]; it is qualified by unequal per-particle atherogenicity. The nominal LDL-particle model ranks second (Ζ) and the cholesterol-cargo model third (Ζ): LDL-C-only cargo is disfavored, non-HDL-C cargo is not excluded. Plaque instability (Ζ, event transition) and inflammation-first (Ε) complement rather than replace LDL-driven atherogenesis. The tournament informs role and wording only.", "study_separating_rank1_from_rank2": "Ference et al. 2019 [17]: LPL and LDLR scores gave nearly identical CHD risk per unit apoB, showing non-LDL apoB particles are causal; Björnson et al. [35,36] agree that non-LDL particles are causal while disputing equal per-particle effects.", "nominal_wins_only_by_funding_artifact": false, "level_unchanged_by_tournament": true }, "evidence_inventory": [ { "citation": "Skålén et al. 2002, Nature", "source_pointer": "PMID 12066187; DOI 10.1038/nature00804", "relates_to_subclaims": [ "SC1" ], "favors_models": [ "M1", "M2" ], "evidence_class": "Strong", "system": "animal model", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Changing apoB proteoglycan binding changed atherosclerosis in vivo.", "what_it_does_not_show": "It does not establish human outcome effect size or distinguish LDL from every other retained apoB class.", "confounders_present": [ "mouse model", "engineered apoB", "lesion rather than clinical event" ], "verification": "verified", "verification_note": "PubMed abstract and metadata read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Cohen et al. 2006, New England Journal of Medicine", "source_pointer": "PMID 16554528; DOI 10.1056/NEJMoa054013", "relates_to_subclaims": [ "SC2", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "human genetic natural experiment", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Lifelong PCSK9-associated LDL reduction strongly reduced CHD in two ancestry groups.", "what_it_does_not_show": "It indexes LDL-C and correlated particle traits together and cannot by itself select particle number over cholesterol cargo.", "confounders_present": [ "rare variants", "pleiotropy", "ancestry-specific variants" ], "verification": "verified", "verification_note": "PubMed abstract and metadata read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Ference et al. 2012, Journal of the American College of Cardiology", "source_pointer": "PMID 23083789; DOI 10.1016/j.jacc.2012.09.017", "relates_to_subclaims": [ "SC2", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "Mendelian randomization", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Multiple genes showed a duration-sensitive lifelong LDL effect on CHD.", "what_it_does_not_show": "The exposure scale is LDL-C and the source combines genetic studies rather than directly measuring LDL-P.", "confounders_present": [ "genetic pleiotropy", "variant weighting", "European-ancestry predominance" ], "verification": "verified", "verification_note": "PubMed abstract and metadata read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Stitziel et al. 2014, New England Journal of Medicine", "source_pointer": "PMID 25390462; PMCID PMC4335708", "relates_to_subclaims": [ "SC2", "SC3" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "human genetic natural experiment", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Loss of NPC1L1 function lowered LDL-C and CHD risk in the same direction as ezetimibe.", "what_it_does_not_show": "Small carrier counts and altered noncholesterol sterols limit exclusive mediation claims.", "confounders_present": [ "rare carriers", "winner's curse", "noncholesterol sterols" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Ference et al. 2016, New England Journal of Medicine", "source_pointer": "PMID 27959767; DOI 10.1056/NEJMoa1604304", "relates_to_subclaims": [ "SC2", "SC3" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "Mendelian randomization", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "PCSK9 and HMGCR variants had nearly identical and additive event effects per unit lower LDL-C.", "what_it_does_not_show": "It does not isolate particle count from cholesterol cargo.", "confounders_present": [ "gene-specific pleiotropy", "genetic proxy differs from short treatment" ], "verification": "verified", "verification_note": "PubMed abstract and metadata read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "WOSCOPS and 4S", "source_pointer": "PMIDs 7566020 and 7968073", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "human randomized trials", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Statin lowering reduced coronary events in primary and secondary prevention and improved survival in 4S.", "what_it_does_not_show": "A single drug class cannot exclude all HMGCR-related effects and does not discriminate LDL-P from LDL-C.", "confounders_present": [ "class-specific pleiotropy", "historical background therapy", "selected populations" ], "verification": "verified", "verification_note": "PubMed abstracts and metadata read; grouped as one drug-class line.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "IMPROVE-IT", "source_pointer": "PMID 26039521; DOI 10.1056/NEJMoa1410489", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "human randomized trial", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Adding NPC1L1 inhibition to statin therapy produced incremental LDL lowering and fewer cardiovascular events.", "what_it_does_not_show": "The effect was modest and all major LDL-related measures changed together.", "confounders_present": [ "post-ACS population", "background statin", "composite endpoint" ], "verification": "verified", "verification_note": "PubMed abstract and metadata read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "FOURIER, ODYSSEY OUTCOMES, and VESALIUS-CV (Bohula et al.)", "source_pointer": "PMIDs 28304224, 30403574, and 41211925", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "human randomized trials", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "PCSK9-antibody lowering reduced atherosclerotic events in established ASCVD, recent ACS, and patients with atherosclerosis or diabetes without prior MI or stroke.", "what_it_does_not_show": "They are one target/class line and lower LDL-C, apoB, and LDL-P together.", "confounders_present": [ "shorter follow-up in FOURIER", "composite endpoints", "one pharmacologic class" ], "verification": "verified", "verification_note": "PubMed abstracts and final metadata read; grouped as one target/class line.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "CLEAR Outcomes", "source_pointer": "PMID 36876740; DOI 10.1056/NEJMoa2215024", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "human randomized trial", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "ATP-citrate-lyase inhibition reduced a composite of major cardiovascular events in statin-intolerant patients.", "what_it_does_not_show": "Cardiovascular and all-cause mortality were not significantly reduced and particle-specific mediation was not isolated.", "confounders_present": [ "statin-intolerant population", "composite endpoint", "adverse-effect profile" ], "verification": "verified", "verification_note": "PubMed abstract and metadata read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "SHARP", "source_pointer": "PMID 21663949; PMCID PMC3145073", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "human randomized trial", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "LDL lowering reduced major atherosclerotic events in chronic kidney disease.", "what_it_does_not_show": "The simvastatin-ezetimibe combination does not isolate one target, and dialysis subgroup precision is limited.", "confounders_present": [ "CKD competing mortality", "combination therapy", "adherence and achieved LDL difference" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "CTT Collaboration 2010", "source_pointer": "PMID 21067804; PMCID PMC2988224", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Moderate", "system": "individual-participant meta-analysis", "decisive_perturbation": false, "endogenous": false, "direction": "supports", "what_it_shows": "Trial-level benefit scaled with absolute LDL-C reduction across broad subgroups.", "what_it_does_not_show": "It synthesizes mostly statin trials and is not another independent perturbation or a particle-number separator.", "confounders_present": [ "component-trial dependence", "LDL-C exposure scale", "historical trial mix" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "B" }, { "citation": "GLAGOV", "source_pointer": "PMID 27846344; DOI 10.1001/jama.2016.16951", "relates_to_subclaims": [ "SC1", "SC3" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "human randomized imaging trial", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "PCSK9 inhibition changed coronary atheroma trajectory and increased plaque regression.", "what_it_does_not_show": "Atheroma imaging is a surrogate and shares the PCSK9 mechanism with outcome trials.", "confounders_present": [ "surrogate endpoint", "statin background", "evaluable-imaging subset" ], "verification": "verified", "verification_note": "PubMed abstract and metadata read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Ference et al. 2019, JAMA", "source_pointer": "PMID 30694319; PMCID PMC6439767", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2", "M1" ], "evidence_class": "Strong", "system": "human Mendelian randomization", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "LPL and LDLR pathways produced similar CHD effects per unit lower apoB and apoB remained after multivariable adjustment.", "what_it_does_not_show": "ApoB counts all atherogenic particles and correlated traits challenge exclusive mediation.", "confounders_present": [ "overlapping cohorts", "multivariable collinearity", "genetic pleiotropy" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Richardson et al. 2020 and Zuber et al. 2021", "source_pointer": "PMIDs 32203549 and 33130851; PMC7089422 and PMC8271202", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2", "M1" ], "evidence_class": "Strong", "system": "human multivariable Mendelian randomization", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Two multivariable frameworks prioritized apoB over LDL-C and triglycerides.", "what_it_does_not_show": "Shared genetic datasets, correlated exposures, and apoB's non-LDL coverage prevent counting them as independent proof of LDL-P specificity.", "confounders_present": [ "trait collinearity", "dataset overlap", "model-selection dependence" ], "verification": "verified", "verification_note": "Complete PMC full texts read; grouped for independence.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Helgadottir et al. 2022, European Journal of Preventive Cardiology", "source_pointer": "PMID 36125206; DOI 10.1093/eurjpc/zwac219", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M3" ], "evidence_class": "Strong", "system": "human Mendelian randomization", "decisive_perturbation": true, "endogenous": true, "direction": "contradicts", "what_it_shows": "In this multivariable analysis, non-HDL-C improved an apoB model whereas apoB did not improve a non-HDL-C model.", "what_it_does_not_show": "One correlated-trait MR analysis cannot erase the opposing MR and discordance evidence.", "confounders_present": [ "correlated exposures", "instrument-model assumptions", "genome-wide instrument pleiotropy" ], "verification": "verified", "verification_note": "Authoring pass read the publisher PDF; red-team pass re-verified the abstract result (P = 3.9e-5 vs 0.69) on 2026-09-30. Affiliation is not used as a scoring input.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Mora et al. 2014, Circulation", "source_pointer": "PMID 24345402; PMCID PMC4501252", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Moderate", "system": "human prospective cohort", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "LDL-C misclassifies coronary risk when discordant with non-HDL-C, apoB or LDL-P; discordant non-HDL-C (HR 2.92) was at least as informative as apoB (2.48) or LDL-P (2.32).", "what_it_does_not_show": "It does not separate particle number from total cholesterol cargo; the authors state the study was not structured to compare the three measures and their confidence intervals overlap.", "confounders_present": [ "women only", "residual confounding", "median-based discordance" ], "verification": "verified", "verification_note": "Complete PMC full text read in both passes; inference corrected in the red-team pass (RT-03).", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "AURORA and 4D", "source_pointer": "PMIDs 19332456 and 16034009", "relates_to_subclaims": [ "SC3", "SC5" ], "favors_models": [ "M5" ], "evidence_class": "Strong", "system": "human randomized trials", "decisive_perturbation": true, "endogenous": true, "direction": "contradicts", "what_it_shows": "Large LDL-C reductions did not significantly reduce mixed primary cardiovascular endpoints in advanced hemodialysis; AURORA also showed no effect on individual components including nonfatal MI.", "what_it_does_not_show": "They do not show that LDL is noncausal for atherosclerotic lesion formation; competing event types, late disease, and modifiable fraction differ.", "confounders_present": [ "advanced dialysis", "mixed endpoints", "competing mortality", "PubMed-listed errata" ], "verification": "verified", "verification_note": "PubMed abstracts and status metadata read; interpreted with SHARP full text.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "EAS Consensus Panel 2017", "source_pointer": "PMID 28444290; PMCID PMC5837225", "relates_to_subclaims": [ "SC1", "SC2", "SC3", "SC5" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Moderate", "system": "consensus synthesis", "decisive_perturbation": false, "endogenous": false, "direction": "supports", "what_it_shows": "The field synthesis found dose-, duration-, genetic-, and intervention-concordant LDL causality.", "what_it_does_not_show": "Consensus does not add an independent causal line and cannot settle particle number versus cargo.", "confounders_present": [ "secondary synthesis", "author consensus", "relies on component studies" ], "verification": "verified", "verification_note": "Complete PMC full text read; used for framing only.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "B" }, { "citation": "Ference et al. 2017, JAMA (CETP and HMGCR genetic scores)", "source_pointer": "PMID 28846118; PMCID PMC5710502; DOI 10.1001/jama.2017.11467", "relates_to_subclaims": [ "SC4", "SC3" ], "favors_models": [ "M2", "M1" ], "evidence_class": "Strong", "system": "human Mendelian randomization", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Combined CETP+HMGCR scores lowered LDL-C without a proportional apoB fall and the risk reduction followed the attenuated apoB change; a discordant score had similar CHD risk per unit apoB but attenuated risk per unit LDL-C (OR 0.916 vs 0.831; P<.001).", "what_it_does_not_show": "It discriminates apoB from LDL-C, not particle number from total cholesterol cargo (non-HDL-C).", "confounders_present": [ "genetic proxy for drug action", "overlapping cohorts with other MR", "collinearity" ], "verification": "verified", "verification_note": "Abstract and key points read (PMC returned no body text) on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Lincoff et al. 2017, ACCELERATE (evacetrapib)", "source_pointer": "PMID 28514624; DOI 10.1056/NEJMoa1609581", "relates_to_subclaims": [ "SC3", "SC4" ], "favors_models": [ "M2", "M1" ], "evidence_class": "Strong", "system": "human randomized trial", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "LDL-C fell 31.1% at 3 months but the trial stopped for futility after a median 26 months (HR 1.01).", "what_it_does_not_show": "The trial's apoB change was not verified here; short duration is an alternative explanation. It is a null under LDL-C indexing, not a demonstrated test of particle number.", "confounders_present": [ "short exposure", "early termination", "CETP-specific effects" ], "verification": "verified", "verification_note": "PubMed abstract read via E-utilities on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "HPS3/TIMI55-REVEAL 2017 (anacetrapib)", "source_pointer": "PMID 28847206; DOI 10.1056/NEJMoa1706444", "relates_to_subclaims": [ "SC3", "SC4" ], "favors_models": [ "M2", "M1", "M3" ], "evidence_class": "Strong", "system": "human randomized trial", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Non-HDL-C was 17 mg/dL (-18%) lower and major coronary events were reduced (rate ratio 0.91) over a median 4.1 years.", "what_it_does_not_show": "Benefit is consistent with both the non-HDL-C and apoB changes; it does not separate particle number from total cargo.", "confounders_present": [ "background intensive statin", "CETP-specific effects", "modest effect size" ], "verification": "verified", "verification_note": "PubMed abstract read via E-utilities on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Schwartz et al. 2012, dal-OUTCOMES (dalcetrapib)", "source_pointer": "PMID 23126252; DOI 10.1056/NEJMoa1206797", "relates_to_subclaims": [ "SC3" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Moderate", "system": "human randomized trial", "decisive_perturbation": false, "endogenous": true, "direction": "null", "what_it_shows": "Dalcetrapib raised HDL-C with minimal effect on LDL-C and did not reduce events (HR 1.04).", "what_it_does_not_show": "It is a negative control for HDL-C raising, not a test of LDL-particle lowering.", "confounders_present": [ "post-ACS population", "CETP modulator pharmacology" ], "verification": "verified", "verification_note": "PubMed abstract read via E-utilities on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Das Pradhan et al. 2022, PROMINENT (pemafibrate)", "source_pointer": "PMID 36342113; DOI 10.1056/NEJMoa2210645", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2", "M1" ], "evidence_class": "Strong", "system": "human randomized trial", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Triglycerides fell 26.2% and remnant cholesterol 25.6% while apoB rose 4.8%, with no event reduction (HR 1.03): cholesterol cargo in remnants fell without particle number falling and without benefit.", "what_it_does_not_show": "Its non-HDL-C change was not verified here, so it does not by itself separate particle number from total cholesterol cargo; PPAR-alpha off-target effects possible.", "confounders_present": [ "type 2 diabetes population", "drug-specific effects", "LDL-C not a target" ], "verification": "verified", "verification_note": "PubMed abstract read via E-utilities on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Otvos et al. 2011, MESA LDL-P discordance", "source_pointer": "PMID 21392724; PMCID PMC3070150; DOI 10.1016/j.jacl.2011.02.001", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M1", "M2" ], "evidence_class": "Moderate", "system": "human prospective cohort", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "Among LDL-C/LDL-P discordant participants only LDL-P was associated with incident CVD (HR 1.45 vs 1.07 for LDL-C); IMT tracked LDL-P.", "what_it_does_not_show": "Observational; does not compare LDL-P with non-HDL-C; discordant subgroups differ in metabolic-syndrome traits.", "confounders_present": [ "metabolic confounding", "calculated LDL-C", "observational" ], "verification": "verified", "verification_note": "Complete PMC full text read on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Cromwell et al. 2007, Framingham Offspring LDL-P", "source_pointer": "PMID 19657464; DOI 10.1016/j.jacl.2007.10.001", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M1", "M2" ], "evidence_class": "Moderate", "system": "human prospective cohort", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "In multivariable models LDL-P was related more strongly to future CVD than LDL-C or non-HDL-C.", "what_it_does_not_show": "Observational; single cohort; NMR assay-specific.", "confounders_present": [ "residual confounding", "assay dependence" ], "verification": "verified", "verification_note": "PubMed abstract read (PMC returned no body text) on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Marston et al. 2022, UK Biobank plus FOURIER/IMPROVE-IT", "source_pointer": "PMID 34773460; PMCID PMC8590731; DOI 10.1001/jamacardio.2021.5083", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2" ], "evidence_class": "Moderate", "system": "human prospective cohort", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "When apoB, non-HDL-C and TG were assessed together only apoB was associated with MI (aHR 1.27 per SD) in primary- and secondary-prevention cohorts.", "what_it_does_not_show": "Observational mutual adjustment of collinear traits; not a causal discriminator.", "confounders_present": [ "collinearity", "residual confounding" ], "verification": "verified", "verification_note": "Abstract and key points read on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Johannesen et al. 2021, Copenhagen statin-treated", "source_pointer": "PMID 33736827; DOI 10.1016/j.jacc.2021.01.027", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Moderate", "system": "human prospective cohort", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "High apoB and non-HDL-C, but not LDL-C, were associated with MI; discordantly high apoB with low non-HDL-C was not associated with MI (HR 0.93).", "what_it_does_not_show": "Does not favor apoB over non-HDL-C for MI; observational.", "confounders_present": [ "statin-treated population", "residual confounding" ], "verification": "verified", "verification_note": "PubMed abstract read on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Welsh et al. 2019, UK Biobank prediction", "source_pointer": "PMID 31216866; PMCID PMC6693929; DOI 10.1161/CIRCULATIONAHA.119.041149", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Weak", "system": "human prospective cohort", "decisive_perturbation": false, "endogenous": true, "direction": "null", "what_it_shows": "apoB, direct LDL-C and non-HDL-C had similar associations (HR 1.23, 1.20, 1.21 per SD) and apoB added no discrimination beyond total and HDL cholesterol; apoB associated more strongly in the discordant subset.", "what_it_does_not_show": "Prediction is not causal primacy; traits correlated r>0.90.", "confounders_present": [ "collinearity", "prediction framework" ], "verification": "verified", "verification_note": "Complete PMC full text read on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Björnson et al. 2023, TRL remnants vs LDL (UK Biobank)", "source_pointer": "PMID 37358553; PMCID PMC10576615; DOI 10.1093/eurheartj/ehad337", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M3", "M2" ], "evidence_class": "Strong", "system": "human Mendelian randomization", "decisive_perturbation": true, "endogenous": true, "direction": "contradicts", "what_it_shows": "CHD OR per SD apoB was 1.76 in a lipolysis (remnant-rich) SNP cluster vs 1.33 in a receptor (LDL) cluster, replicated in CARDIoGRAMplusC4D; remnant cholesterol remained associated after apoB adjustment.", "what_it_does_not_show": "Remnant-C is indirectly derived; polygenic instruments carry pleiotropy risk; it contradicts equal per-particle atherogenicity rather than LDL causality.", "confounders_present": [ "indirect remnant measure", "polygenic pleiotropy", "European ancestry" ], "verification": "verified", "verification_note": "Complete PMC full text read on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Björnson et al. 2024, Lp(a) vs LDL per particle", "source_pointer": "PMID 38233012; PMCID PMC7616706; DOI 10.1016/j.jacc.2023.10.039", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2" ], "evidence_class": "Strong", "system": "human Mendelian randomization", "decisive_perturbation": true, "endogenous": true, "direction": "contradicts", "what_it_shows": "Lp(a) was about 6.6-fold (95% CI 5.1-8.8) more atherogenic than LDL per apoB particle.", "what_it_does_not_show": "Does not bear on LDL causality itself; apo(a) isoform effects not separated.", "confounders_present": [ "UK Biobank single cohort", "instrument clustering" ], "verification": "verified", "verification_note": "Complete PMC full text read on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Burgess et al. 2018, LPA Mendelian randomization", "source_pointer": "PMID 29926099; PMCID PMC6481553; DOI 10.1001/jamacardio.2018.1470", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2" ], "evidence_class": "Moderate", "system": "human Mendelian randomization", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "About 101.5 mg/dL change in Lp(a) was needed to match the CHD association of a 1 mmol/L LDL-C change; association proportional to absolute Lp(a) change.", "what_it_does_not_show": "Mass-concentration exposure; does not address LDL-P versus cargo.", "confounders_present": [ "Lp(a) mass assay", "European ancestry" ], "verification": "verified", "verification_note": "PubMed abstract read on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Khera et al. 2016, FH-gene sequencing", "source_pointer": "PMID 27050191; PMCID PMC5405769; DOI 10.1016/j.jacc.2016.03.520", "relates_to_subclaims": [ "SC2" ], "favors_models": [ "M1", "M2", "M3" ], "evidence_class": "Strong", "system": "human genetic natural experiment", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Within any stratum of observed LDL-C, FH-mutation carriers had higher CAD risk (OR 22.3 with LDL-C >=190 plus mutation vs 6.0 without mutation, against LDL-C <130 without mutation), with higher cumulative LDL-C exposure.", "what_it_does_not_show": "Indexes LDL-C; small carrier numbers; case-control components.", "confounders_present": [ "case-control design", "treated LDL-C imputation", "rare carriers" ], "verification": "verified", "verification_note": "Complete PMC full text read on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" }, { "citation": "Ridker et al. 2017, CANTOS (canakinumab)", "source_pointer": "PMID 28845751; DOI 10.1056/NEJMoa1707914", "relates_to_subclaims": [ "SC5" ], "favors_models": [ "M5", "M4" ], "evidence_class": "Strong", "system": "human randomized trial", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "IL-1beta inhibition (150 mg) reduced nonfatal MI, nonfatal stroke or CV death (HR 0.85) without lowering lipids; fatal infections increased and all-cause mortality was unchanged.", "what_it_does_not_show": "Supports inflammation in event transition; does not show LDL is merely permissive. Direction 'supports' refers to the competing models M5/M4, not the nominal.", "confounders_present": [ "post-MI with hsCRP >=2 mg/L", "dose-specific significance", "infection risk" ], "verification": "verified", "verification_note": "PubMed abstract read on 2026-09-30.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Citation-context (Scite) counts were not used in either pass; none are estimated." ] }, "source_tier": "A" } ], "verification_summary": { "load_bearing_claims_checked": 28, "verified": 28, "unverified": 0, "contradicted": 0, "note": "Red-team pass (2026-09-30) re-checked all 20 original claims against PubMed/E-utilities abstracts and PMC full text where available, corrected C9, C10 and the inference attached to C18, rewrote C20, and added C21-C28. 14 of 39 sources were read in complete full text across both passes; the rest are labeled abstract/metadata depth." }, "factual_issues_found": [ { "quoted_text": "LDL particle concentration", "problem": "LDL-P, apoB, and LDL-C are correlated but nonidentical exposures.", "source_says": "ApoB counts all circulating atherogenic particles, LDL-P directly estimates LDL number, and LDL-C measures cholesterol cargo; discordance and multivariable analyses do not agree perfectly on primacy.", "source_pointer": "Sources 17-21", "affects_score": true, "action": "Retained as a separate SC4 and used to stop the exact wording at Η rather than Θ." }, { "quoted_text": "causal driver", "problem": "Driver could be misread as sole, necessary, or sufficient cause of every clinical event.", "source_says": "LDL-related exposure drives plaque formation, while acute event transition and residual risk also depend on inflammation, plaque disruption, thrombosis, and other exposures.", "source_pointer": "Sources 9-16, 22-24", "affects_score": false, "action": "Scoped the role to atherogenesis and retained separate event-transition models." }, { "quoted_text": "lowering LDL prevents ASCVD in every setting", "problem": "The submitted claim does not say this, and advanced dialysis trials would contradict such an unrestricted response claim.", "source_says": "AURORA and 4D lowered LDL-C but did not significantly reduce mixed primary cardiovascular endpoints.", "source_pointer": "Sources 22 and 23", "affects_score": true, "action": "Mapped late hemodialysis as a context boundary and scored SC3/SC5 at Η rather than Θ." }, { "quoted_text": "multiple MR papers are independent replications", "problem": "Several particle-specific analyses share UK Biobank, CARDIoGRAM, or related lipid-genetic structures.", "source_says": "Different statistical frameworks are informative but do not fail independently in the way separate experimental modalities do.", "source_pointer": "Sources 17-20", "affects_score": false, "action": "Grouped them as one particle-discriminator line in the independence audit." }, { "quoted_text": "Sabatine et al. Evolocumab in Patients without a Previous Myocardial Infarction or Stroke", "problem": "Wrong first author.", "source_says": "PubMed PMID 41211925 lists Bohula EA as first author.", "source_pointer": "Source 13", "affects_score": false, "action": "Corrected (RT-09)." }, { "quoted_text": "bempedoic acid produced a 21.1% greater LDL-C reduction", "problem": "Percentage points reported as percent.", "source_says": "LDL-C reduction greater by 29.2 mg/dL; difference in percent reductions 21.1 percentage points.", "source_pointer": "Source 12", "affects_score": false, "action": "Corrected (RT-10)." }, { "quoted_text": "PCSK9 monoclonal-antibody trials reduced atherosclerotic events in secondary and primary-prevention contexts", "problem": "VESALIUS-CV enrolled patients with atherosclerosis or diabetes without prior MI or stroke, not a primary-prevention population in the usual sense.", "source_says": "'patients with atherosclerosis or diabetes and without a previous myocardial infarction or stroke'", "source_pointer": "Source 13", "affects_score": false, "action": "Reworded (RT-11)." }, { "quoted_text": "Direct LDL-P data point in the same direction", "problem": "Mora 2014 was cited as favoring particle number over cholesterol cargo; its non-HDL-C discordance result was at least as strong and the authors state the study cannot separate the measures.", "source_says": "Age-adjusted HR 2.92 (non-HDL-C), 2.48 (apoB), 2.32 (LDL-P); 'not structured to address specific differences'.", "source_pointer": "Source 21 (full text)", "affects_score": true, "action": "Re-described as evidence against LDL-C alone; neutral on particle vs cargo (RT-03)." }, { "quoted_text": "untreated definite or probable familial hypercholesterolemia ... adjusted odds ratio of 13.2", "problem": "DLCN diagnosis includes personal history of premature CAD, inflating the CAD odds ratio by design.", "source_says": "DLCN 'employs the personal and family history of premature coronary artery disease'.", "source_pointer": "Source 6; replacement anchor Source 38", "affects_score": false, "action": "Caveat added; Khera 2016 added as FH anchor (RT-12)." }, { "quoted_text": "SC4 therefore reaches Η with Moderate confidence", "problem": "SC4 fails the quoted Η requirements (intervention evidence isolating the quantity; mechanistic specificity; no major unresolved contradiction).", "source_says": "Contradictions from independent groups [20,35,36]; no intervention separates particle number from total cholesterol cargo.", "source_pointer": "Sources 17-21, 25-36", "affects_score": true, "action": "SC4 set to Ζ; overall Ζ under preregistered weakest-link (RT-01)." } ], "cap_rules_applied": [ { "subclaim_id": "SC1", "cap": "Γ", "cap_source": "rubric", "triggered": false, "trigger_text_quoted": "Single-species or single-strain evidence", "specific_trigger": "Binds the retention perturbation [1] in isolation; not re-applied to SC1 as a convergent set because human genetics, imaging and hard-event trials supply independent phenotype-relevant lines (rubric 06: 'Convergence is not re-capped').", "resulting_ceiling": "Γ" }, { "subclaim_id": "SC4", "cap": "Β", "cap_source": "universal", "triggered": false, "trigger_text_quoted": "Pure association evidence — however large, however replicated — cannot exceed Β.", "specific_trigger": "Binds the observational discordance cohorts [21,30-34] in isolation; not triggered for SC4 as a whole because multivariable MR [17-20,25,35,36] and pharmacological discordance [26,27,29] are also present.", "resulting_ceiling": "Β" }, { "subclaim_id": "SC3", "cap": "Γ", "cap_source": "rubric", "triggered": false, "trigger_text_quoted": "Biomarker-only improvement without organismal benefit", "specific_trigger": "Not triggered: multiple randomized trials measured myocardial infarction, stroke, revascularization, cardiovascular death, or plaque progression rather than lipid biomarkers alone.", "resulting_ceiling": "Γ" } ], "cap_precedence_resolutions": [ { "collision": "The closest rubric's literal Δ text addresses lifespan/healthspan, while its upper-ladder section explicitly recognizes human genetics, multiple pharmacological classes, dose-response, and lifelong exposure as field-valid equivalents.", "subclaim_scope_cap_binds_on": "Biomarker-only or isolated single-system lipid evidence would bind at Γ.", "subclaim_scope_threshold_binds_on": "Hard human ASCVD events, plaque trajectory, and convergent genetic/pharmacologic evidence bind on SC1-SC5.", "resolution": "Applied the rubric's explicit field-valid equivalents to ASCVD outcomes and did not re-cap the convergent set because one member is observational or animal.", "note": "This is a transparent rubric-fit adaptation, not a silent invention of a cardiovascular-specific threshold." } ], "boundary_check_BGD": [ { "subclaim_id": "SC1", "field_valid_perturbation_quote": "Field-valid equivalents in this domain: Mendelian randomization on the exposure itself, multiple pharmacological classes acting on the same node with concordant endpoint effects, dose-response across independent populations, and natural experiments of lifelong exposure.", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "Retention perturbation [1] plus randomized plaque imaging [16].", "gamma_hard_rules_triggered": [ "Biomarker-only improvement without organismal benefit — not triggered for the scored evidence set." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "Lesion initiation and plaque trajectory are phenotype-relevant; human hard outcomes provide convergent support.", "ceiling_from_boundary_check": "Δ or higher" }, { "subclaim_id": "SC2", "field_valid_perturbation_quote": "Field-valid equivalents in this domain: Mendelian randomization on the exposure itself, multiple pharmacological classes acting on the same node with concordant endpoint effects, dose-response across independent populations, and natural experiments of lifelong exposure.", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "PCSK9, NPC1L1, HMGCR, LDLR/APOB, multi-gene and FH-sequencing human natural experiments [2-6,38].", "gamma_hard_rules_triggered": [ "Biomarker-only improvement without organismal benefit — not triggered for the scored evidence set." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "Lifelong genetic exposure changes coronary disease, not merely a lipid biomarker.", "ceiling_from_boundary_check": "Δ or higher" }, { "subclaim_id": "SC3", "field_valid_perturbation_quote": "Field-valid equivalents in this domain: Mendelian randomization on the exposure itself, multiple pharmacological classes acting on the same node with concordant endpoint effects, dose-response across independent populations, and natural experiments of lifelong exposure.", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "Statin, ezetimibe, PCSK9-antibody, bempedoic-acid, and CKD randomized outcome trials [7-14].", "gamma_hard_rules_triggered": [ "Biomarker-only improvement without organismal benefit — not triggered for the scored evidence set." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "Several independent targets change hard atherosclerotic events; the hemodialysis boundary is separately retained.", "ceiling_from_boundary_check": "Δ or higher" }, { "subclaim_id": "SC4", "field_valid_perturbation_quote": "Field-valid equivalents in this domain: Mendelian randomization on the exposure itself, multiple pharmacological classes acting on the same node with concordant endpoint effects, dose-response across independent populations, and natural experiments of lifelong exposure.", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "Discordant-pathway genetics: LPL vs LDLR per apoB [17]; CETP+HMGCR scores [25]; multivariable MR [18,19].", "gamma_hard_rules_triggered": [ "Biomarker-only improvement without organismal benefit — not triggered for the scored evidence set." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "Positive for particle burden over LDL-C; contested for particle number over total cholesterol cargo [20,35,36]. Crosses Δ; the contest is adjudicated at Η.", "ceiling_from_boundary_check": "Δ or higher" }, { "subclaim_id": "SC5", "field_valid_perturbation_quote": "Field-valid equivalents in this domain: Mendelian randomization on the exposure itself, multiple pharmacological classes acting on the same node with concordant endpoint effects, dose-response across independent populations, and natural experiments of lifelong exposure.", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "Primary prevention, secondary prevention, ACS, CKD, and dialysis trials [7-14,22,23].", "gamma_hard_rules_triggered": [ "Biomarker-only improvement without organismal benefit — not triggered for the scored evidence set." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "Multiple contexts are directly tested, including negative boundary contexts.", "ceiling_from_boundary_check": "Δ or higher" } ], "thresholds_tested": [ { "subclaim_id": "SC1", "threshold": "Η", "requirement_quoted": "Requires: Robust cross-context intervention evidence; Mechanistic specificity; Strong alignment of lifespan/healthspan outcomes", "satisfied": "yes", "supporting_evidence": [ "Retention-specific animal perturbation [1]", "randomized human plaque regression [16]", "hard-event convergence [7-14]" ], "what_would_cross_it": "Already crossed using disease-trajectory and ASCVD-event analogues stated in the rubric-fit note." }, { "subclaim_id": "SC1", "threshold": "Θ", "requirement_quoted": "Requires high reproducibility with minimal contextual contradiction.", "satisfied": "no", "supporting_evidence": [ "The retention mechanism is strong, but human imaging is surrogate and apoB subclass contributions remain unresolved." ], "what_would_cross_it": "Replicated human lesion-level mediation that separates LDL particle entry, retention, and cholesterol delivery from other apoB particles." }, { "subclaim_id": "SC2", "threshold": "Θ", "requirement_quoted": "Requires high reproducibility with minimal contextual contradiction.", "satisfied": "yes", "supporting_evidence": [ "Multiple genes and ancestries [2-6]", "duration gradient [3]", "target concordance with trials [4,5,7-13]" ], "what_would_cross_it": "Already crossed for LDL-related cumulative exposure, not for LDL-P exclusivity." }, { "subclaim_id": "SC3", "threshold": "Η", "requirement_quoted": "Requires: Robust cross-context intervention evidence; Mechanistic specificity; Strong alignment of lifespan/healthspan outcomes", "satisfied": "yes", "supporting_evidence": [ "HMGCR [7,8]", "NPC1L1 [9]", "PCSK9 [10,11,13]", "ATP citrate lyase [12]", "CKD [14]", "CETP, apoB/non-HDL-C-indexed [25,27]" ], "what_would_cross_it": "Already crossed for prespecified ASCVD outcomes." }, { "subclaim_id": "SC3", "threshold": "Θ", "requirement_quoted": "Requires high reproducibility with minimal contextual contradiction.", "satisfied": "no", "supporting_evidence": [ "Evacetrapib null (HR 1.01) despite 31.1% LDL-C lowering [26]; the apoB-based explanation is not demonstrated in the trial's own data.", "Advanced-dialysis nulls [22,23] lie outside SC3's stated scope and are not used here." ], "what_would_cross_it": "Prospectively accurate prediction of benefit across late CKD contexts using achieved apoB/LDL-P change, atherosclerotic event fraction, duration, and competing risk." }, { "subclaim_id": "SC4", "threshold": "Ζ", "requirement_quoted": "Requires: Reproducible intervention effects; Mechanistic narrowing; Evidence not reducible to general sickness, caloric reduction, or developmental artifacts", "satisfied": "yes", "supporting_evidence": [ "Discordant-pathway genetics [17,25]", "multivariable MR [18,19]", "CETP trial pair [26,27]", "PROMINENT [29]", "discordance cohorts [30-33]" ], "what_would_cross_it": "Already crossed: LDL-C and remnant cholesterol as unique quantities are narrowed out; particle number versus total cholesterol cargo is not excluded." }, { "subclaim_id": "SC4", "threshold": "Η", "requirement_quoted": "Requires: Robust cross-context intervention evidence; Mechanistic specificity; Strong alignment of lifespan/healthspan outcomes", "satisfied": "no", "supporting_evidence": [ "No intervention separates particle number from total cholesterol cargo.", "Helgadottir et al. favored non-HDL-C over apoB [20].", "Per-particle atherogenicity differs by class [35,36].", "Framework Η: 'no major unresolved upstream/downstream contradiction within the claimed scope' is not met." ], "what_would_cross_it": "A replicated human discriminator (MR with opposing apoB/non-HDL-C instruments, or an intervention with measured discordant changes) showing that hard ASCVD outcomes track particle number after total cholesterol cargo and particle class are accounted for." }, { "subclaim_id": "SC4", "threshold": "Θ", "requirement_quoted": "Requires high reproducibility with minimal contextual contradiction.", "satisfied": "no", "supporting_evidence": [ "Η not met; Θ not reachable." ], "what_would_cross_it": "A replicated human discriminator that changes particle number and cholesterol cargo independently and predicts hard ASCVD outcomes without off-target differences." }, { "subclaim_id": "SC5", "threshold": "Η", "requirement_quoted": "Requires: Robust cross-context intervention evidence; Mechanistic specificity; Strong alignment of lifespan/healthspan outcomes", "satisfied": "yes", "supporting_evidence": [ "Primary and secondary prevention [7-13]", "CKD [14]", "genetic duration and ancestry [2-6]", "mapped dialysis boundary [22,23]" ], "what_would_cross_it": "Already crossed for atherosclerotic rather than all cardiovascular outcomes." }, { "subclaim_id": "SC5", "threshold": "Θ", "requirement_quoted": "Requires high reproducibility with minimal contextual contradiction.", "satisfied": "no", "supporting_evidence": [ "AURORA showed no effect on the composite or its individual components, including nonfatal MI, within SC5's CKD scope [22].", "4D composite null; fatal stroke RR 2.03 [23]." ], "what_would_cross_it": "A prespecified response model that accurately identifies contexts with and without event reduction and replicates prospectively." }, { "subclaim_id": "overall", "threshold": "Ω", "requirement_quoted": "Requires necessity and sufficiency of the proposed aging mechanism across the claimed scope. This will be rare.", "satisfied": "no", "supporting_evidence": [ "ASCVD can arise through interacting causes, and elevated LDL exposure is not sufficient for every individual event." ], "what_would_cross_it": "No ethically realistic design is expected to establish universal individual necessity and sufficiency; the submitted claim does not require Ω." } ], "generalization_check": { "generalization_tested": "yes", "context_boundary_mapped": "yes", "internal_negative_controls": [ "AURORA: 43% LDL-C reduction without reduction in the mixed primary endpoint or its components [22].", "4D: 42% LDL-C reduction, RR 0.92 for the composite; fatal stroke RR 2.03 [23].", "SHARP: proportional effect on major atherosclerotic events similar in dialysis and non-dialysis patients [14].", "ACCELERATE: 31.1% LDL-C reduction without event reduction over a median 26 months [26].", "dal-OUTCOMES: HDL-C raising without LDL lowering, no benefit [28].", "PROMINENT: remnant cholesterol lowered without apoB lowering, no benefit [29].", "CLEAR Outcomes: composite benefit without significant cardiovascular- or all-cause-death reduction [12]." ], "note": "The causal effect generalizes strongly across primary prevention, established ASCVD, ACS, statin intolerance, and CKD. Advanced dialysis maps a late-disease and endpoint-composition boundary; it narrows response prediction without reversing the atherogenesis conclusion." }, "falsification_patterns_met": [ { "subclaim_id": "SC3", "pattern": "Late-life intervention fails despite strong early-life effects", "triggered_within_scope": false, "evidence": "Advanced-dialysis nulls [22,23] lie outside SC3's stated scope (primary and secondary prevention, non-dialysis CKD); for SC5 they are a mapped boundary (pitfall 16), not a falsification.", "downgrade_applied": false }, { "subclaim_id": "SC5", "pattern": "Necessity fails when the proposed pathway is removed", "triggered_within_scope": false, "evidence": "Residual events after intensive lowering test necessity, which the claim does not assert; blocks Ω only.", "downgrade_applied": false }, { "subclaim_id": "SC3", "pattern": "Pathway manipulation changes biomarkers but not lifespan/healthspan", "triggered_within_scope": false, "evidence": "Not triggered for LDL-particle lowering (hard events fall across classes); echoed only under LDL-C indexing by evacetrapib [26], which is recorded as an exception in thresholds.", "downgrade_applied": false } ], "pitfalls_flagged": [ { "subclaim_id": "SC4", "pitfall": "Correlated biomarker treated as the unique causal quantity", "where": "LDL-C trial effects cannot by themselves prove that LDL particle number rather than cholesterol cargo is the operative unit." }, { "subclaim_id": "SC4", "pitfall": "ApoB silently equated with LDL-P", "where": "Each atherogenic particle carries one apoB, but apoB also counts remnants and lipoprotein(a)." }, { "subclaim_id": "SC3", "pitfall": "Reversibility mislabeled as mechanism", "where": "SC3 is explicitly typed as reversibility; outcome reduction does not alone identify the molecular mediator." }, { "subclaim_id": "SC3", "pitfall": "Drug classes and trials over-counted", "where": "WOSCOPS/4S are one HMGCR line and FOURIER/ODYSSEY/VESALIUS are one PCSK9-antibody line." }, { "subclaim_id": "SC4", "pitfall": "Overlapping genetic datasets counted as independent", "where": "The apoB-prioritizing MR papers are grouped as a discriminator family with shared failure modes." }, { "subclaim_id": "SC5", "pitfall": "Mixed cardiovascular endpoints treated as pure atherosclerotic endpoints", "where": "Dialysis trials include event types and competing risks less responsive to plaque modification." }, { "subclaim_id": "overall", "pitfall": "Primary driver read as sole, necessary, or sufficient cause", "where": "The review limits primary-driver status to atherogenesis and retains plaque instability, thrombosis, and other causal exposures." }, { "subclaim_id": "SC4", "pitfall": "Equal per-particle atherogenicity assumed across apoB classes", "where": "Remnant-rich and Lp(a) particles carry more risk per apoB than LDL [35,36]; apoB count is not a complete exposure metric." }, { "subclaim_id": "SC4", "pitfall": "Association dressed as mechanism (pitfall 2)", "where": "Observational discordance cohorts [21,30-34] show which measure predicts risk, not which quantity is causal." }, { "subclaim_id": "SC3", "pitfall": "Negative field-valid perturbation read as a crossed threshold (pitfall 17)", "where": "ACCELERATE and dal-OUTCOMES are nulls and are not counted toward any threshold; contradicting MR [20,35,36] is excluded from the supportive tally." } ], "causal_tally_audit": { "tally_claimed_by_document": 7, "tally_after_audit": 10, "items_rejected": [ "WOSCOPS and 4S are one HMGCR-class causal line.", "FOURIER, ODYSSEY OUTCOMES, VESALIUS-CV and GLAGOV share PCSK9 inhibition.", "NPC1L1 loss-of-function and IMPROVE-IT share the target.", "Richardson, Zuber, Ference 2017/2019, Helgadottir and Björnson use correlated lipid-genetic structures; their disagreement is more informative than a paper count.", "Helgadottir [20], Björnson [35,36], AURORA/4D [22,23] and ACCELERATE [26] are contradicting or null perturbations and are not counted as supportive (pitfall 17).", "CTT [15] and the EAS consensus [24] add no independent causal line." ], "note": "The 10 lines after audit are the 7 LDL-class lines plus 3 SC4-specific lines recorded in independence_basis; line (a) for SC4 is internally contradictory." }, "steelman": { "case_one_rung_higher": "A case for Η: particle-indexed measures consistently beat LDL-C across four MR analyses [17-19,25], three cohort studies [30-32], and a CETP genetics-plus-trial pair [25-27]; PROMINENT shows that lowering remnant cholesterol without lowering particle number does not reduce events [29]; and every LDL-class causal subclaim is at Η or Θ. On this view the remaining dispute is a refinement about how to weight apoB particles.", "case_one_rung_lower": "A case for Ε: the distinctive particle-number-over-total-cholesterol claim has no clean intervention, its MR support is contradicted by two independent groups [20,35,36], and the observational evidence cannot separate apoB from non-HDL-C [21,33,34].", "which_rejected": "Η is rejected: SC4 lacks the rubric's 'Mechanistic specificity' and intervention evidence isolating the quantity, and carries a 'major unresolved ... contradiction within the claimed scope'. Ε is rejected: the LDL-C-only alternative is excluded by independent genetic, pharmacological and observational lines [17-19,25-27,30-33], which is multi-line convergence beyond one context.", "blocking_cap_threshold_or_missing_experiment": "The Η threshold is blocked by the missing experiment: a replicated human discriminator separating particle number from total cholesterol cargo (and accounting for particle class) with hard ASCVD outcomes; contradictions [20,35,36] remain within SC4's scope." }, "final_score": { "hsef_level": "Ζ", "hsef_descriptor": "Multiple independent causal lines converge", "confidence": "Moderate", "confidence_rationale": "The binding subclaim (SC4) has dense but directionally inconsistent evidence from independent groups and heterogeneous assays (NMR LDL-P, immunoassay apoB, calculated non-HDL-C and remnant-C). Confidence in LDL-class causality (SC2, SC3, SC5) remains High and is reported separately.", "rests_on_non_peer_reviewed_work": false, "rationale": "Cumulative LDL-class apoB exposure is a primary causal driver of ASCVD (SC1, SC3, SC5 at Η; SC2 at Θ). The submitted wording adds that particle concentration (number) is the operative quantity; that is well supported against LDL-C but contested against total cholesterol cargo, and per-particle risk varies by particle class [20,35,36]. The preregistered weakest-link rule therefore returns Ζ; scope-matched aggregation would return Η." }, "role_of_mechanism": { "classification": "primary driver", "derivation_step_matched": "3", "derivation": "Step 1 (downstream consequence) does not match: disease does not arise or progress independently of LDL-class exposure. Step 2 (epiphenomenon) does not match: causal dependency is demonstrated. Step 3 matches on all three conditions: removal prevents or delays hard atherosclerotic outcomes [7-14]; lifelong elevation from FH mutations increases risk at any observed LDL-C with greater cumulative exposure [38], and lifelong lowering protects [2-5]; effects scale with absolute exposure and duration [3,15]. Scope: ASCVD through atherogenesis, at the level of LDL-class apoB particles; it does not certify particle number rather than cholesterol cargo as the rate-limiting property. Inflammation [39], plaque disruption and thrombosis co-determine event timing.", "tournament_used_as_tiebreaker": false, "note": "The broader apoB model ranks first; both it and the nominal place LDL-class particles upstream rather than downstream or merely permissive." }, "extended_scoring_MSD": { "momentum": { "value": "Mechanistic only", "artifact": "PREVAIL (NCT05202509; obicetrapib, CETP inhibitor) active, not recruiting, 9,541 enrolled, primary completion est. 2026-11; ORION-4 (NCT03705234; inclisiran) active, not recruiting, 16,124 enrolled, primary completion est. 2026-10 (ClinicalTrials.gov API, 2026-09-30).", "justification": "Neither pending outcome trial is designed to separate particle number from total cholesterol cargo; PREVAIL may help separate apoB from LDL-C if achieved apoB and non-HDL-C are reported." }, "stability": { "value": "Stable", "trigger": "PREVAIL readout (could shift SC4 confidence either way); full Lp(a)HORIZON report (NCT04023552, completed; sponsor topline 2026-09-04 states primary endpoint not met; Tier C, not scored), which bears on the broader apoB model rather than the nominal wording.", "justification": "No named pending result is expected to cross SC4's particle-versus-total-cargo boundary in either direction, so the Ζ assessment should persist." }, "divergence": { "implied_level": "Η", "implied_level_basis": "The sentence, as the field usually presents LDL causality, asserts an established causal driver without necessity or sufficiency, which implies strong convergent intervention-sensitive evidence.", "assessed_level": "Ζ", "gap_in_rungs": 1, "value": "Mild overstatement", "justification": "Implied Η, assessed Ζ, +1 rung. The overstatement lies entirely in the particle-number reading; read as 'LDL-class particles cause ASCVD' the wording is aligned." }, "calibration_note": "M/S/D do not modify HSEF level or confidence. Momentum and stability are tied to named registry artifacts; divergence to explicit rung arithmetic." }, "change_record": { "applies": true, "prior_level": "Η", "prior_confidence": "High", "prior_source": "2026-09-29 HSEF v5.4 review package (OpenAI Codex authoring run), red-teamed 2026-09-30.", "derived_before_consulting_prior": false, "new_level": "Ζ", "direction": "lowered", "what_moved_it": "SC4 (particle-number specificity, a preregistered weakest-link conjunct) was re-scored from Η to Ζ: the quoted Η requirements (intervention evidence isolating the quantity; mechanistic specificity; no major unresolved contradiction) are not met, and the contradiction is replicated across independent groups [20,35,36]. No new evidence lowered the LDL-class causal subclaims.", "divergence_from_other_reviewers": "The authoring run assigned Η/High while conceding in its own text that SC4 evidence conflicts and that no intervention isolates particle number; the red team applies the rubric text to that concession." }, "proposed_decisive_experiments": [ { "name": "Orthogonal particle-versus-cargo intervention", "experiment": "Use genetic instruments or a randomized therapy combination that produces prespecified opposite changes in apoB (or LDL-P) and non-HDL cholesterol, recording particle class (LDL, remnant, Lp(a)).", "readout": "Hard ASCVD events plus serial arterial plaque imaging and directly measured LDL-P, apoB, remnant cholesterol, and LDL-C.", "critical_control": "Equal achieved cholesterol cargo with different particle number, and the reciprocal contrast, with adherence and exposure duration verified.", "falsifying_result": "Risk and plaque change consistently follow cholesterol cargo after particle number is separated.", "supporting_result": "Risk and plaque change consistently follow particle number after cholesterol cargo is separated." }, { "name": "Human arterial flux and retention mediation", "experiment": "Use labeled autologous apoB-particle tracers and validated arterial-wall imaging before and after a selective lowering intervention, prospectively partitioning LDL, remnants, and lipoprotein(a).", "readout": "Particle-specific arterial entry/retention, plaque lipid change, inflammatory response, and subsequent plaque progression.", "critical_control": "Matched plasma cholesterol cargo and off-target inflammatory effects, with blinded image analysis.", "falsifying_result": "Plaque response is unrelated to change in arterial particle flux or retention once cargo is controlled.", "supporting_result": "Change in arterial particle flux and retention mediates plaque response independently of cargo." }, { "name": "Prespecified model comparison in pending CETP-inhibitor outcomes", "experiment": "Before unblinding, register the event reduction predicted by LDL-C, apoB and non-HDL-C models from achieved changes, then compare with the observed result; replicate in a second trial or cohort.", "readout": "Observed hazard ratio versus each model's prediction, with achieved LDL-C, apoB, non-HDL-C and LDL-P reported.", "critical_control": "Predictions locked before unblinding; adherence-adjusted exposure; duration matched to the genetic and CTT estimates.", "falsifying_result": "Benefit matches the LDL-C prediction and exceeds what the smaller apoB change predicts.", "supporting_result": "Benefit matches the apoB prediction and is smaller than the LDL-C prediction." } ], "knowledge_gaps": [ { "name": "Causal unit", "statement": "No hard-outcome intervention cleanly separates particle number from total cholesterol cargo.", "relevance": "Direct blocker to Η for SC4 and therefore for the overall letter." }, { "name": "Per-particle heterogeneity", "statement": "Remnant and Lp(a) particles carry more risk per apoB than LDL [35,36]; the mechanism and weighting are unresolved.", "relevance": "Particle count alone is an incomplete exposure metric; bears on M2 and on LDL-P versus apoB." }, { "name": "Instrument disagreement", "statement": "MR analyses disagree partly by instrument choice; LDLR variants also affect remnant clearance [35], and genome-wide sets carry pleiotropy [20].", "relevance": "Determines which discriminator results are credible." }, { "name": "CETP apoB versus LDL-C", "statement": "The apoB change behind the evacetrapib null was not established here; short duration remains an alternative explanation [26].", "relevance": "Bears on the 'any mechanism' argument and on SC4." }, { "name": "Arterial exposure metric", "statement": "Single-time plasma concentration is an imperfect proxy for cumulative arterial particle flux, residence, modification and retention.", "relevance": "A better exposure measure could sharpen mediation." }, { "name": "Late-disease modifiability", "statement": "The fraction of atherosclerotic risk still controlled by LDL lowering in advanced dialysis remains difficult to predict [14,22,23].", "relevance": "Defines SC3 and SC5 boundaries." }, { "name": "Residual event transition", "statement": "How LDL burden interacts with plaque inflammation [39], erosion or rupture, and thrombosis after intensive lowering remains incompletely modelled.", "relevance": "Separates plaque formation from event timing." } ], "review_metadata": { "review_date": "2026-09-29", "reviewer": "OpenAI Codex using HSEF v5.4 (authoring, 2026-09-29); red-team synthesis by Claude (Anthropic), 2026-09-30", "evidence_base_note": "Primary mechanistic, human genetic, randomized outcome, imaging, discordance, and counterevidence through 2026-09-29. Full text was read wherever accessible and abstract-only rows are labeled. The attached packet (SHA256 7E097D3D50D2068F68BDB1CD022DF4A045DAC46A9F25E51262EC7568E9DB11EB) was treated as methodology, not scientific evidence. Red-team pass added CETP-inhibitor, discordance-cohort, per-particle-heterogeneity, FH-sequencing and CANTOS evidence through 2026-09-30.", "literature_tools_available": true, "literature_tools_note": "Authoring pass: PubMed records, PMC full text, one publisher PDF. Red-team pass: PubMed MCP tools and NCBI E-utilities (full structured abstracts, errata/retraction links), PMC full text, ClinicalTrials.gov API. Scite counts not used.", "skill_version": "5.4", "validator_run": "pass (validator 2.2, red-team edition; see companion validation report)", "validator_warnings_addressed": "Validator 2.2 was run on the final pair after the red-team edits; see ldl-particle-concentration-ascvd-causality.validation.txt.", "redteam_synthesis_date": "2026-09-30", "consensus_consistency_audit": "2026-10-09; supplementary provenance/source reconciliation; original grading retained; prior grades visible." }, "is_calibration_entry": { "value": false, "entry_role": "not_calibration", "blind_graded": false, "completed_blinded_calibration": false, "expected_band_sealed_at": "", "target_sha256": "", "target_authorship": "No sealed calibration target was supplied.", "target_author_saw_review": null, "future_run_status": "not prepared", "unblinded_reason": "Standalone evaluation; no independently authored sealed expected band was supplied." }, "upper_ladder_check": { "final_level_below_zeta": false, "equivalents_considered": [ { "kind": "Mendelian randomization on the exposure itself", "available": true, "why_not_sufficient": "Carries SC2 to Θ for LDL-class exposure; for SC4, multivariable MR is internally contradictory [17-20,25,35,36]." }, { "kind": "multiple pharmacological classes acting on the same node with concordant endpoint effects", "available": true, "why_not_sufficient": "Carries SC3 to Η; no pharmacological class isolates particle number from total cholesterol cargo." }, { "kind": "dose-response across independent populations", "available": true, "why_not_sufficient": "Supports cumulative exposure [3,15] but does not select particle number over cargo." }, { "kind": "natural experiments of lifelong exposure", "available": true, "why_not_sufficient": "Supports Θ for SC2 [2-6,38] while changing LDL-C, apoB and LDL-P together." } ], "convergence_present": true, "cap_reapplied_to_convergent_set": false, "note": "Final level Ζ is at the Ζ boundary; no below-Ζ justification is required, but the equivalents are recorded because SC4 binds." }, "provenance_header": { "sources_total": 39, "claims_total": 28, "claims_verified": 28, "claims_unverified": 0, "claims_contradicted": 0, "tier_a_sources": 37, "generated_by": "HSEF v5.4 evidence-gated review workflow; red-team synthesis edition 2026-09-30", "gate_passed": true }, "claims": [ { "id": "C1", "text": "Mice expressing proteoglycan-binding-defective apoB-containing LDL developed significantly less atherosclerosis than mice expressing wild-type LDL, directly linking arterial-wall retention to lesion formation in that model.", "carries_number": false, "load_bearing_for": [ "SC1" ], "source_ns": [ 1 ], "verification": "verified", "verified_against": "PubMed abstract and bibliographic record", "verified_on": "2026-09-29", "note": "Direct mechanistic perturbation in one species, read at abstract depth ('significantly less'); not a human outcome experiment." }, { "id": "C2", "text": "In ARIC, PCSK9 loss-of-function variants associated with 28% lower LDL-C in Black participants were associated with an 88% lower CHD risk (HR 0.11), while a variant associated with 15% lower LDL-C in White participants was associated with a 47% lower risk (HR 0.50).", "carries_number": true, "load_bearing_for": [ "SC2", "SC5" ], "source_ns": [ 2 ], "verification": "verified", "verified_against": "PubMed abstract and bibliographic record; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Natural experiment supports lifelong exposure; the measured mediator was LDL-C rather than LDL-P." }, { "id": "C3", "text": "Across nine variants in six genes and 312,321 participants, lifelong exposure to 1 mmol/L lower LDL-C was associated with 54.5% lower CHD risk, about three times the reduction per unit seen when statin treatment begins later in life.", "carries_number": true, "load_bearing_for": [ "SC2", "SC5" ], "source_ns": [ 3 ], "verification": "verified", "verified_against": "PubMed abstract and bibliographic record; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Supports cumulative exposure and timing; variants are not direct measurements of LDL particle number." }, { "id": "C4", "text": "NPC1L1 protein-truncating variant carriers had about 12 mg/dL lower LDL-C and an odds ratio of 0.47 for CHD, providing a human target-based natural experiment concordant with ezetimibe therapy.", "carries_number": true, "load_bearing_for": [ "SC2", "SC3" ], "source_ns": [ 4 ], "verification": "verified", "verified_against": "complete PMC full text, results, and limitations; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "The carrier count was small and noncholesterol sterols could contribute; these limits are retained." }, { "id": "C5", "text": "PCSK9 and HMGCR genetic scores each predicted an odds ratio of 0.81 for cardiovascular events per 10 mg/dL lower LDL-C, and their effects were independent and additive.", "carries_number": true, "load_bearing_for": [ "SC2", "SC3" ], "source_ns": [ 5 ], "verification": "verified", "verified_against": "PubMed abstract and bibliographic record; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Target-concordant genetics reduces the plausibility of a statin-specific explanation." }, { "id": "C6", "text": "In the Copenhagen General Population Study, untreated definite or probable familial hypercholesterolemia was associated with coronary artery disease at an adjusted odds ratio of 13.2 compared with non-FH participants.", "carries_number": true, "load_bearing_for": [ "SC2", "SC5" ], "source_ns": [ 6 ], "verification": "verified", "verified_against": "PubMed abstract and status record", "verified_on": "2026-09-30", "note": "DLCN diagnosis includes personal history of premature CAD, inflating this odds ratio by design; erratum JCEM 2014;99:4758-9. Supportive only; Khera 2016 [38] is the FH anchor." }, { "id": "C7", "text": "In WOSCOPS, pravastatin lowered LDL-C by 26% and reduced definite coronary events by 31%; in 4S, simvastatin reduced all-cause mortality (RR 0.70), coronary death (RR 0.58), and major coronary events (RR 0.66).", "carries_number": true, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 7, 8 ], "verification": "verified", "verified_against": "PubMed abstracts and bibliographic records; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Two trials, but one HMGCR drug-class line in the independence tally." }, { "id": "C8", "text": "In IMPROVE-IT, adding ezetimibe after acute coronary syndrome lowered time-weighted LDL-C from 69.5 to 53.7 mg/dL and reduced the seven-year primary-event rate from 34.7% to 32.7% (HR 0.936).", "carries_number": true, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 9 ], "verification": "verified", "verified_against": "PubMed abstract and bibliographic record; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Independent intestinal-absorption mechanism; modest absolute and relative effect on background statin therapy." }, { "id": "C9", "text": "PCSK9 monoclonal-antibody trials reduced atherosclerotic events in established ASCVD (FOURIER HR 0.85), recent acute coronary syndrome (ODYSSEY OUTCOMES HR 0.85), and patients with atherosclerosis or diabetes without previous MI or stroke (VESALIUS-CV three-point MACE HR 0.75).", "carries_number": true, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 10, 11, 13 ], "verification": "verified", "verified_against": "PubMed abstracts and final bibliographic records", "verified_on": "2026-09-30", "note": "VESALIUS-CV first author is Bohula (corrected). Three trials, one pharmacologic class line." }, { "id": "C10", "text": "Among 13,970 statin-intolerant patients in CLEAR Outcomes, bempedoic acid lowered LDL-C 29.2 mg/dL more than placebo (a 21.1-percentage-point difference in percent reduction) and reduced the primary composite (HR 0.87), myocardial infarction (HR 0.77), and coronary revascularization (HR 0.81), without a significant reduction in cardiovascular or all-cause death.", "carries_number": true, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 12 ], "verification": "verified", "verified_against": "PubMed abstract and bibliographic record", "verified_on": "2026-09-30", "note": "Adds an ATP-citrate-lyase intervention line and a statin-intolerant population." }, { "id": "C11", "text": "In SHARP, simvastatin plus ezetimibe produced a 0.85 mmol/L LDL-C difference and reduced major atherosclerotic events from 13.4% to 11.3% (RR 0.83) in chronic kidney disease.", "carries_number": true, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 14 ], "verification": "verified", "verified_against": "complete PMC full text, methods, results, and discussion; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Supports CKD generalization while retaining the trial's distinction between atherosclerotic and mixed cardiovascular endpoints." }, { "id": "C12", "text": "The CTT individual-participant meta-analysis found a proportional reduction in major vascular events with greater achieved LDL-C lowering across the studied range and major subgroups.", "carries_number": false, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 15 ], "verification": "verified", "verified_against": "complete PMC full text, methods, results, and subgroup analyses; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Tier B synthesis; it summarizes component trials and is not an extra independent perturbation." }, { "id": "C13", "text": "In GLAGOV, adding evolocumab reduced percent atheroma volume by 0.95% versus a 0.05% increase with placebo, and 64.3% versus 47.3% of patients showed plaque regression by the primary imaging measure.", "carries_number": true, "load_bearing_for": [ "SC1", "SC3" ], "source_ns": [ 16 ], "verification": "verified", "verified_against": "PubMed abstract and bibliographic record; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Randomized lesion-trajectory evidence, but imaging is a surrogate and the intervention shares the PCSK9 class with outcome trials." }, { "id": "C14", "text": "In a 654,783-person genetic analysis, LPL and LDLR scores produced nearly identical lower CHD risk per 10 mg/dL lower apoB (OR 0.771 and 0.773); after mutual adjustment, apoB retained an association while LDL-C and triglyceride estimates were null.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 17 ], "verification": "verified", "verified_against": "complete PMC full text, methods, results, supplements discussion, and limitations; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Supports apoB particle burden, but not LDL particle number alone because apoB also counts remnants and other atherogenic particles." }, { "id": "C15", "text": "In multivariable Mendelian randomization, apoB retained a robust association with CHD (OR 1.92 per standard-deviation higher trait), whereas the LDL-C estimate was 0.85 and nonsignificant; the authors cautioned that strong trait correlation complicates exclusive causal interpretation.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 18 ], "verification": "verified", "verified_against": "complete PMC full text, methods, results, and limitations; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "Supports particle number while explicitly preserving collinearity limits." }, { "id": "C16", "text": "MR-BMA across 30 lipoprotein-related measures ranked an apoB-only model first and selected apoB in all sensitivity analyses, with no cholesterol or triglyceride measure consistently selected otherwise.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 19 ], "verification": "verified", "verified_against": "complete PMC full text, methods, results, and sensitivity analyses; re-verified against the PubMed abstract via E-utilities in the red-team pass", "verified_on": "2026-09-30", "note": "The analysis shares lipid-genetic data structures with other MR work and is not counted as a wholly independent causal line." }, { "id": "C17", "text": "A competing 235-variant Mendelian-randomization analysis found that adding non-HDL-C to an apoB model improved prediction of CAD effects, whereas adding apoB to a non-HDL-C model did not, favoring cholesterol burden over particle concentration in that analysis.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 20 ], "verification": "verified", "verified_against": "complete publisher PDF, abstract, methods, results, sensitivity analyses, and disclosures", "verified_on": "2026-09-30", "note": "Genuine directional contradiction to the particle-number formulation; replicated in direction (particle count is not the whole story) by an independent group's per-particle heterogeneity findings [35,36]." }, { "id": "C18", "text": "Among 27,533 women followed for a median 17.2 years, LDL-C misclassified coronary risk when discordant with non-HDL-C, apoB or LDL-P: discordantly high non-HDL-C (age-adjusted HR 2.92), apoB (2.48) and LDL-P (2.32) each identified higher risk; the authors state the study was not structured to separate the three alternative measures, whose confidence intervals overlap.", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 21 ], "verification": "verified", "verified_against": "complete PMC full text, methods, results, and limitations", "verified_on": "2026-09-30", "note": "Evidence against LDL-C as the unique quantity; neutral between particle number and total cholesterol cargo (red-team correction RT-03)." }, { "id": "C19", "text": "In advanced hemodialysis, AURORA and 4D lowered LDL-C by 43% and 42% but did not significantly reduce their mixed primary cardiovascular endpoints (HR 0.96 and RR 0.92); AURORA showed no effect on the individual components including nonfatal myocardial infarction, and 4D showed fatal stroke RR 2.03.", "carries_number": true, "load_bearing_for": [ "SC3", "SC5" ], "source_ns": [ 22, 23 ], "verification": "verified", "verified_against": "PubMed abstracts, status records, and SHARP full-text contextual analysis", "verified_on": "2026-09-30", "note": "Both PubMed records list errata. These are real boundary results, not evidence that LDL cannot cause atherosclerosis." }, { "id": "C20", "text": "Synthesis: cumulative exposure to arterial-retained LDL-class apoB particles is a primary causal driver of ASCVD, while the narrower claim that particle number rather than total cholesterol cargo is the operative quantity is supported against LDL-C but contested against non-HDL-C, and per-particle risk varies by particle class.", "carries_number": false, "load_bearing_for": [ "SC1", "SC2", "SC3", "SC4", "SC5" ], "source_ns": [ 1, 2, 3, 4, 5, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 35, 36, 38 ], "verification": "verified", "verified_against": "synthesis of the verified component claims C1-C19 and C21-C28 (not an independent factual check)", "verified_on": "2026-09-30", "note": "Review conclusion registered because the lay summary relies on it; rewritten in the red-team pass." }, { "id": "C21", "text": "Genetic scores mimicking CETP inhibition combined with statin action lowered LDL-C without a proportional apoB fall; the risk reduction followed the attenuated apoB change, and a discordant LDL-C/apoB score had similar CHD risk per unit apoB (OR 0.782 vs 0.793) but significantly attenuated risk per unit LDL-C (OR 0.916 vs 0.831; P < .001).", "carries_number": true, "load_bearing_for": [ "SC4", "SC3" ], "source_ns": [ 25 ], "verification": "verified", "verified_against": "PubMed abstract and PMC key points", "verified_on": "2026-09-30", "note": "Discriminates apoB from LDL-C, not particle number from non-HDL-C." }, { "id": "C22", "text": "Evacetrapib lowered LDL-C by 31.1% at 3 months without reducing events (HR 1.01; median 26 months; stopped for futility); anacetrapib lowered non-HDL-C by 17 mg/dL (-18%) and reduced major coronary events (rate ratio 0.91; median 4.1 years); dalcetrapib had minimal LDL-C effect and no benefit (HR 1.04).", "carries_number": true, "load_bearing_for": [ "SC3", "SC4" ], "source_ns": [ 26, 27, 28 ], "verification": "verified", "verified_against": "PubMed abstracts via E-utilities", "verified_on": "2026-09-30", "note": "Evacetrapib is a null under LDL-C indexing; its apoB change was not verified here." }, { "id": "C23", "text": "In PROMINENT, pemafibrate lowered triglycerides by 26.2% and remnant cholesterol by 25.6% while apoB rose 4.8%, with no reduction in cardiovascular events (HR 1.03).", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 29 ], "verification": "verified", "verified_against": "PubMed abstract via E-utilities", "verified_on": "2026-09-30", "note": "Consistent with particle models, not with a remnant-cholesterol-specific model; non-HDL-C change not verified." }, { "id": "C24", "text": "Discordance cohorts favor particle measures over LDL-C: in MESA only LDL-P was associated with CVD among discordant participants (HR 1.45 vs 1.07 for LDL-C); in Framingham Offspring LDL-P was related more strongly to CVD than LDL-C or non-HDL-C; in UK Biobank and two trial cohorts only apoB remained associated with MI when assessed with non-HDL-C and triglycerides (aHR 1.27 per SD).", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 30, 31, 32 ], "verification": "verified", "verified_against": "PMC full text (MESA); PubMed abstracts and key points (Framingham, Marston)", "verified_on": "2026-09-30", "note": "Observational; cap 2 binds this line in isolation." }, { "id": "C25", "text": "Other cohorts do not separate apoB from non-HDL-C: in statin-treated Copenhagen participants discordantly high apoB with low non-HDL-C was not associated with MI (HR 0.93), and in UK Biobank apoB, direct LDL-C and non-HDL-C had similar associations with CVD (HR 1.23, 1.20, 1.21 per SD) with no discrimination gain from apoB.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 33, 34 ], "verification": "verified", "verified_against": "PubMed abstract (Johannesen); PMC full text (Welsh)", "verified_on": "2026-09-30", "note": "Balance evidence; prediction is not causal primacy." }, { "id": "C26", "text": "Per-particle atherogenicity differs by class: CHD OR per SD apoB was 1.76 in a remnant-rich SNP cluster versus 1.33 in an LDL-rich cluster (replicated in CARDIoGRAMplusC4D), Lp(a) was about 6.6-fold more atherogenic than LDL per particle, and about 101.5 mg/dL of Lp(a) change matched the CHD association of 1 mmol/L LDL-C.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 35, 36, 37 ], "verification": "verified", "verified_against": "PMC full text (Björnson 2023, 2024); PubMed abstract (Burgess)", "verified_on": "2026-09-30", "note": "Second, independent contradiction to particle count as the complete causal quantity." }, { "id": "C27", "text": "In FH-gene sequencing of 26,025 participants, carriers had higher CAD risk than non-carriers within any stratum of observed LDL-C; LDL-C >=190 mg/dL with a mutation carried OR 22.3 and without a mutation OR 6.0 versus LDL-C <130 mg/dL without a mutation, with greater cumulative LDL-C exposure in carriers.", "carries_number": true, "load_bearing_for": [ "SC2" ], "source_ns": [ 38 ], "verification": "verified", "verified_against": "PMC full text", "verified_on": "2026-09-30", "note": "Replaces the DLCN-based estimate as the FH anchor." }, { "id": "C28", "text": "In CANTOS, canakinumab 150 mg every 3 months reduced nonfatal MI, nonfatal stroke or cardiovascular death (HR 0.85) without lowering lipids, with more fatal infections and no all-cause mortality difference (HR 0.94).", "carries_number": true, "load_bearing_for": [ "SC5" ], "source_ns": [ 39 ], "verification": "verified", "verified_against": "PubMed abstract via E-utilities", "verified_on": "2026-09-30", "note": "Supports inflammation in event transition (M5); does not make LDL merely permissive." } ], "citation_ledger": [ { "n": 1, "citation": "Skålén et al. Subendothelial retention of atherogenic lipoproteins in early atherosclerosis. Nature. 2002.", "tier": "A", "kind": "animal mechanism experiment", "pmid": "12066187", "pmcid": "", "doi": "10.1038/nature00804", "url": "https://pubmed.ncbi.nlm.nih.gov/12066187/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "SKALEN-2002-MOUSE-RETENTION", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C1", "C20" ] }, { "n": 2, "citation": "Cohen et al. Sequence variations in PCSK9, low LDL, and protection against coronary heart disease. The New England journal of medicine. 2006.", "tier": "A", "kind": "human genetic natural experiment", "pmid": "16554528", "pmcid": "", "doi": "10.1056/NEJMoa054013", "url": "https://pubmed.ncbi.nlm.nih.gov/16554528/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "ARIC-PCSK9-LOF", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C2", "C20" ] }, { "n": 3, "citation": "Ference et al. Effect of long-term exposure to lower low-density lipoprotein cholesterol beginning early in life on the risk of coronary heart disease: a Mendelian randomization analysis. Journal of the American College of Cardiology. 2012.", "tier": "A", "kind": "Mendelian-randomization meta-analysis", "pmid": "23083789", "pmcid": "", "doi": "10.1016/j.jacc.2012.09.017", "url": "https://pubmed.ncbi.nlm.nih.gov/23083789/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "FERENCE-2012-LIFELONG-LDL", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C3", "C20" ] }, { "n": 4, "citation": "Myocardial Infarction Genetics Consortium Investigators; Stitziel et al. Inactivating mutations in NPC1L1 and protection from coronary heart disease. The New England journal of medicine. 2014.", "tier": "A", "kind": "human genetic natural experiment", "pmid": "25390462", "pmcid": "PMC4335708", "doi": "10.1056/NEJMoa1405386", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC4335708/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NPC1L1-LOF-2014", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read, including methods, results, discussion, limitations, and available status information. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C4", "C20" ] }, { "n": 5, "citation": "Ference et al. Variation in PCSK9 and HMGCR and Risk of Cardiovascular Disease and Diabetes. The New England journal of medicine. 2016.", "tier": "A", "kind": "Mendelian randomization", "pmid": "27959767", "pmcid": "", "doi": "10.1056/NEJMoa1604304", "url": "https://pubmed.ncbi.nlm.nih.gov/27959767/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "PCSK9-HMGCR-2016", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C5", "C20" ] }, { "n": 6, "citation": "Benn et al. Familial hypercholesterolemia in the danish general population: prevalence, coronary artery disease, and cholesterol-lowering medication. The Journal of clinical endocrinology and metabolism. 2012.", "tier": "A", "kind": "prospective population study", "pmid": "22893714", "pmcid": "", "doi": "10.1210/jc.2012-1563", "url": "https://pubmed.ncbi.nlm.nih.gov/22893714/", "status_flags": [ "PubMed lists an erratum: J Clin Endocrinol Metab 2014;99(12):4758-9 (correction text not retrieved)", "DLCN diagnostic circularity noted in the red-team pass" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "COPENHAGEN-GENERAL-POPULATION", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C6" ] }, { "n": 7, "citation": "Shepherd et al., West of Scotland Coronary Prevention Study Group. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. The New England journal of medicine. 1995.", "tier": "A", "kind": "randomized controlled trial", "pmid": "7566020", "pmcid": "", "doi": "10.1056/NEJM199511163332001", "url": "https://pubmed.ncbi.nlm.nih.gov/7566020/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "WOSCOPS", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C7", "C20" ] }, { "n": 8, "citation": "Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet (London, England). 1994.", "tier": "A", "kind": "randomized controlled trial", "pmid": "7968073", "pmcid": "", "doi": "", "url": "https://pubmed.ncbi.nlm.nih.gov/7968073/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "4S", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C7", "C20" ] }, { "n": 9, "citation": "Cannon et al. Ezetimibe Added to Statin Therapy after Acute Coronary Syndromes. The New England journal of medicine. 2015.", "tier": "A", "kind": "randomized controlled trial", "pmid": "26039521", "pmcid": "", "doi": "10.1056/NEJMoa1410489", "url": "https://pubmed.ncbi.nlm.nih.gov/26039521/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT00202878", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C8", "C20" ] }, { "n": 10, "citation": "Sabatine et al. Evolocumab and Clinical Outcomes in Patients with Cardiovascular Disease. The New England journal of medicine. 2017.", "tier": "A", "kind": "randomized controlled trial", "pmid": "28304224", "pmcid": "", "doi": "10.1056/NEJMoa1615664", "url": "https://pubmed.ncbi.nlm.nih.gov/28304224/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT01764633", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C9", "C20" ] }, { "n": 11, "citation": "Schwartz et al. Alirocumab and Cardiovascular Outcomes after Acute Coronary Syndrome. The New England journal of medicine. 2018.", "tier": "A", "kind": "randomized controlled trial", "pmid": "30403574", "pmcid": "", "doi": "10.1056/NEJMoa1801174", "url": "https://pubmed.ncbi.nlm.nih.gov/30403574/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT01663402", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C9", "C20" ] }, { "n": 12, "citation": "Nissen et al. Bempedoic Acid and Cardiovascular Outcomes in Statin-Intolerant Patients. The New England journal of medicine. 2023.", "tier": "A", "kind": "randomized controlled trial", "pmid": "36876740", "pmcid": "", "doi": "10.1056/NEJMoa2215024", "url": "https://pubmed.ncbi.nlm.nih.gov/36876740/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT02993406", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C10", "C20" ] }, { "n": 13, "citation": "Bohula et al. Evolocumab in Patients without a Previous Myocardial Infarction or Stroke. The New England journal of medicine. 2026 (epub 2025).", "tier": "A", "kind": "randomized controlled trial", "pmid": "41211925", "pmcid": "", "doi": "10.1056/NEJMoa2514428", "url": "https://pubmed.ncbi.nlm.nih.gov/41211925/", "status_flags": [ "none_found", "first author corrected from 'Sabatine' to 'Bohula' in the red-team pass" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "VESALIUS-CV", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C9", "C20" ] }, { "n": 14, "citation": "Baigent et al. The effects of lowering LDL cholesterol with simvastatin plus ezetimibe in patients with chronic kidney disease (Study of Heart and Renal Protection): a randomised placebo-controlled trial. Lancet (London, England). 2011.", "tier": "A", "kind": "randomized controlled trial", "pmid": "21663949", "pmcid": "PMC3145073", "doi": "10.1016/S0140-6736(11)60739-3", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC3145073/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "SHARP", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read, including methods, results, discussion, limitations, and available status information. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C11", "C20" ] }, { "n": 15, "citation": "Cholesterol Treatment Trialists' Collaboration. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet (London, England). 2010.", "tier": "B", "kind": "individual-participant meta-analysis", "pmid": "21067804", "pmcid": "PMC2988224", "doi": "10.1016/S0140-6736(10)61350-5", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC2988224/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "CTT-2010-26-TRIAL-IPD", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read, including methods, results, discussion, limitations, and available status information. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C12", "C20" ] }, { "n": 16, "citation": "Nicholls et al. Effect of Evolocumab on Progression of Coronary Disease in Statin-Treated Patients: The GLAGOV Randomized Clinical Trial. JAMA. 2016.", "tier": "A", "kind": "randomized imaging trial", "pmid": "27846344", "pmcid": "", "doi": "10.1001/jama.2016.16951", "url": "https://pubmed.ncbi.nlm.nih.gov/27846344/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT01813422", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C13", "C20" ] }, { "n": 17, "citation": "Ference et al. Association of Triglyceride-Lowering LPL Variants and LDL-C-Lowering LDLR Variants With Risk of Coronary Heart Disease. JAMA. 2019.", "tier": "A", "kind": "Mendelian randomization", "pmid": "30694319", "pmcid": "PMC6439767", "doi": "10.1001/jama.2018.20045", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC6439767/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "FERENCE-2019-63-STUDIES", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read, including methods, results, discussion, limitations, and available status information. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C14", "C20" ] }, { "n": 18, "citation": "Richardson et al. Evaluating the relationship between circulating lipoprotein lipids and apolipoproteins with risk of coronary heart disease: A multivariable Mendelian randomisation analysis. PLoS medicine. 2020.", "tier": "A", "kind": "multivariable Mendelian randomization", "pmid": "32203549", "pmcid": "PMC7089422", "doi": "10.1371/journal.pmed.1003062", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC7089422/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "UKBB-CARDIOGRAM-MVMR-2020", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read, including methods, results, discussion, limitations, and available status information. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C15", "C20" ] }, { "n": 19, "citation": "Zuber et al. High-throughput multivariable Mendelian randomization analysis prioritizes apolipoprotein B as key lipid risk factor for coronary artery disease. International journal of epidemiology. 2021.", "tier": "A", "kind": "MR Bayesian model averaging", "pmid": "33130851", "pmcid": "PMC8271202", "doi": "10.1093/ije/dyaa216", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC8271202/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "MR-BMA-2021", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read, including methods, results, discussion, limitations, and available status information. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C16", "C20" ] }, { "n": 20, "citation": "Helgadottir et al. Cholesterol not particle concentration mediates the atherogenic risk conferred by apolipoprotein B particles: a Mendelian randomization analysis. European journal of preventive cardiology. 2022.", "tier": "A", "kind": "Mendelian randomization", "pmid": "36125206", "pmcid": "", "doi": "10.1093/eurjpc/zwac219", "url": "https://pubmed.ncbi.nlm.nih.gov/36125206/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "HELGADOTTIR-2022-UKBB-CAD", "access_depth": "full_text_complete", "access_note": "Complete publisher PDF and extracted text read, including methods, results, sensitivity analyses, discussion, limitations, and conflict-of-interest disclosure. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C17", "C20" ] }, { "n": 21, "citation": "Mora et al. Discordance of low-density lipoprotein (LDL) cholesterol with alternative LDL-related measures and future coronary events. Circulation. 2014.", "tier": "A", "kind": "prospective discordance cohort", "pmid": "24345402", "pmcid": "PMC4501252", "doi": "10.1161/CIRCULATIONAHA.113.005873", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC4501252/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "WOMENS-HEALTH-STUDY", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read in both passes (authoring 2026-09-29; red team 2026-09-30); inference corrected (RT-03).", "carries_claims": [ "C18", "C20" ] }, { "n": 22, "citation": "Fellström et al. Rosuvastatin and cardiovascular events in patients undergoing hemodialysis. The New England journal of medicine. 2009.", "tier": "A", "kind": "randomized controlled trial", "pmid": "19332456", "pmcid": "", "doi": "10.1056/NEJMoa0810177", "url": "https://pubmed.ncbi.nlm.nih.gov/19332456/", "status_flags": [ "PubMed lists an erratum: N Engl J Med 2010;362(15):1450; the abstract-level endpoint result was retained with this flag" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT00240331", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C19", "C20" ] }, { "n": 23, "citation": "Wanner et al. Atorvastatin in patients with type 2 diabetes mellitus undergoing hemodialysis. The New England journal of medicine. 2005.", "tier": "A", "kind": "randomized controlled trial", "pmid": "16034009", "pmcid": "", "doi": "10.1056/NEJMoa043545", "url": "https://pubmed.ncbi.nlm.nih.gov/16034009/", "status_flags": [ "PubMed lists an erratum: N Engl J Med 2005;353(15):1640; the abstract-level endpoint result was retained with this flag" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "4D", "access_depth": "abstract_metadata", "access_note": "PubMed abstract, bibliographic identity, and status metadata read; no abstract-external claim is attributed to this source. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C19", "C20" ] }, { "n": 24, "citation": "Ference et al., EAS Consensus Panel. Low-density lipoproteins cause atherosclerotic cardiovascular disease. 1. Evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European Atherosclerosis Society Consensus Panel. European heart journal. 2017.", "tier": "B", "kind": "consensus synthesis", "pmid": "28444290", "pmcid": "PMC5837225", "doi": "10.1093/eurheartj/ehx144", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC5837225/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "EAS-2017-CONSENSUS", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read, including methods, results, discussion, limitations, and available status information. Red-team pass 2026-09-30: identity (title, first author, year, journal) confirmed via PubMed; carried claims re-checked against the full structured abstract.", "carries_claims": [ "C20" ] }, { "n": 25, "citation": "Ference et al. Association of Genetic Variants Related to CETP Inhibitors and Statins With Lipoprotein Levels and Cardiovascular Risk. JAMA. 2017.", "tier": "A", "kind": "Mendelian randomization", "pmid": "28846118", "pmcid": "PMC5710502", "doi": "10.1001/jama.2017.11467", "url": "https://pubmed.ncbi.nlm.nih.gov/28846118/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "FERENCE-2017-CETP-HMGCR", "access_depth": "abstract_metadata", "access_note": "PMC record returned abstract and key points only (no body text); read 2026-09-30; no abstract-external claim is attributed to this source.", "carries_claims": [ "C20", "C21" ] }, { "n": 26, "citation": "Lincoff et al., ACCELERATE Investigators. Evacetrapib and Cardiovascular Outcomes in High-Risk Vascular Disease. The New England journal of medicine. 2017.", "tier": "A", "kind": "randomized controlled trial", "pmid": "28514624", "pmcid": "", "doi": "10.1056/NEJMoa1609581", "url": "https://pubmed.ncbi.nlm.nih.gov/28514624/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "NCT01687998", "access_depth": "abstract_metadata", "access_note": "PubMed full structured abstract (NCBI E-utilities) and bibliographic identity read 2026-09-30; no abstract-external claim is attributed to this source.", "carries_claims": [ "C22" ] }, { "n": 27, "citation": "HPS3/TIMI55-REVEAL Collaborative Group; Bowman et al. Effects of Anacetrapib in Patients with Atherosclerotic Vascular Disease. The New England journal of medicine. 2017.", "tier": "A", "kind": "randomized controlled trial", "pmid": "28847206", "pmcid": "", "doi": "10.1056/NEJMoa1706444", "url": "https://pubmed.ncbi.nlm.nih.gov/28847206/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "NCT01252953", "access_depth": "abstract_metadata", "access_note": "PubMed full structured abstract (NCBI E-utilities) and bibliographic identity read 2026-09-30; no abstract-external claim is attributed to this source.", "carries_claims": [ "C22" ] }, { "n": 28, "citation": "Schwartz et al., dal-OUTCOMES Investigators. Effects of dalcetrapib in patients with a recent acute coronary syndrome. The New England journal of medicine. 2012.", "tier": "A", "kind": "randomized controlled trial", "pmid": "23126252", "pmcid": "", "doi": "10.1056/NEJMoa1206797", "url": "https://pubmed.ncbi.nlm.nih.gov/23126252/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "NCT00658515", "access_depth": "abstract_metadata", "access_note": "PubMed full structured abstract (NCBI E-utilities) and bibliographic identity read 2026-09-30; no abstract-external claim is attributed to this source.", "carries_claims": [ "C22" ] }, { "n": 29, "citation": "Das Pradhan et al., PROMINENT Investigators. Triglyceride Lowering with Pemafibrate to Reduce Cardiovascular Risk. The New England journal of medicine. 2022.", "tier": "A", "kind": "randomized controlled trial", "pmid": "36342113", "pmcid": "", "doi": "10.1056/NEJMoa2210645", "url": "https://pubmed.ncbi.nlm.nih.gov/36342113/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "NCT03071692", "access_depth": "abstract_metadata", "access_note": "PubMed full structured abstract (NCBI E-utilities) and bibliographic identity read 2026-09-30; no abstract-external claim is attributed to this source.", "carries_claims": [ "C23" ] }, { "n": 30, "citation": "Otvos et al. Clinical implications of discordance between low-density lipoprotein cholesterol and particle number. Journal of clinical lipidology. 2011.", "tier": "A", "kind": "prospective discordance cohort", "pmid": "21392724", "pmcid": "PMC3070150", "doi": "10.1016/j.jacl.2011.02.001", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC3070150/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "MESA-LDLP-2011", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read 2026-09-30, including methods, results, discussion and limitations.", "carries_claims": [ "C24" ] }, { "n": 31, "citation": "Cromwell et al. LDL Particle Number and Risk of Future Cardiovascular Disease in the Framingham Offspring Study - Implications for LDL Management. Journal of clinical lipidology. 2007.", "tier": "A", "kind": "prospective cohort", "pmid": "19657464", "pmcid": "PMC2720529", "doi": "10.1016/j.jacl.2007.10.001", "url": "https://pubmed.ncbi.nlm.nih.gov/19657464/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "FRAMINGHAM-OFFSPRING-LDLP", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read 2026-09-30 (PMC returned no body text); no abstract-external claim is attributed to this source.", "carries_claims": [ "C24" ] }, { "n": 32, "citation": "Marston et al. Association of Apolipoprotein B-Containing Lipoproteins and Risk of Myocardial Infarction in Individuals With and Without Atherosclerosis: Distinguishing Between Particle Concentration, Type, and Content. JAMA cardiology. 2022.", "tier": "A", "kind": "prospective cohort (UK Biobank plus trial cohorts)", "pmid": "34773460", "pmcid": "PMC8590731", "doi": "10.1001/jamacardio.2021.5083", "url": "https://pubmed.ncbi.nlm.nih.gov/34773460/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "UKBB-FOURIER-IMPROVEIT-APOB", "access_depth": "abstract_metadata", "access_note": "PMC record returned abstract and key points only (no body text); read 2026-09-30; no abstract-external claim is attributed to this source.", "carries_claims": [ "C24" ] }, { "n": 33, "citation": "Johannesen et al. Apolipoprotein B and Non-HDL Cholesterol Better Reflect Residual Risk Than LDL Cholesterol in Statin-Treated Patients. Journal of the American College of Cardiology. 2021.", "tier": "A", "kind": "prospective discordance cohort", "pmid": "33736827", "pmcid": "", "doi": "10.1016/j.jacc.2021.01.027", "url": "https://pubmed.ncbi.nlm.nih.gov/33736827/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "COPENHAGEN-STATIN-DISCORDANCE", "access_depth": "abstract_metadata", "access_note": "PubMed full structured abstract (NCBI E-utilities) and bibliographic identity read 2026-09-30; no abstract-external claim is attributed to this source.", "carries_claims": [ "C25" ] }, { "n": 34, "citation": "Welsh et al. Comparison of Conventional Lipoprotein Tests and Apolipoproteins in the Prediction of Cardiovascular Disease. Circulation. 2019.", "tier": "A", "kind": "prospective cohort (prediction)", "pmid": "31216866", "pmcid": "PMC6693929", "doi": "10.1161/CIRCULATIONAHA.119.041149", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC6693929/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "UKBB-PREDICTION-2019", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read 2026-09-30, including methods, results, discussion and limitations.", "carries_claims": [ "C25" ] }, { "n": 35, "citation": "Björnson et al. Triglyceride-rich lipoprotein remnants, low-density lipoproteins, and risk of coronary heart disease: a UK Biobank study. European heart journal. 2023.", "tier": "A", "kind": "Mendelian randomization", "pmid": "37358553", "pmcid": "PMC10576615", "doi": "10.1093/eurheartj/ehad337", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC10576615/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "BJORNSON-2023-TRL-LDL", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read 2026-09-30, including methods, results, discussion and limitations.", "carries_claims": [ "C20", "C26" ] }, { "n": 36, "citation": "Björnson et al. Lipoprotein(a) Is Markedly More Atherogenic Than LDL: An Apolipoprotein B-Based Genetic Analysis. Journal of the American College of Cardiology. 2024.", "tier": "A", "kind": "Mendelian randomization", "pmid": "38233012", "pmcid": "PMC7616706", "doi": "10.1016/j.jacc.2023.10.039", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC7616706/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "BJORNSON-2024-LPA-LDL", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read 2026-09-30, including methods, results, discussion and limitations.", "carries_claims": [ "C20", "C26" ] }, { "n": 37, "citation": "Burgess et al. Association of LPA Variants With Risk of Coronary Disease and the Implications for Lipoprotein(a)-Lowering Therapies: A Mendelian Randomization Analysis. JAMA cardiology. 2018.", "tier": "A", "kind": "Mendelian randomization", "pmid": "29926099", "pmcid": "PMC6481553", "doi": "10.1001/jamacardio.2018.1470", "url": "https://pubmed.ncbi.nlm.nih.gov/29926099/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "BURGESS-2018-LPA", "access_depth": "abstract_metadata", "access_note": "PubMed full structured abstract (NCBI E-utilities) and bibliographic identity read 2026-09-30; no abstract-external claim is attributed to this source.", "carries_claims": [ "C26" ] }, { "n": 38, "citation": "Khera et al. Diagnostic Yield and Clinical Utility of Sequencing Familial Hypercholesterolemia Genes in Patients With Severe Hypercholesterolemia. Journal of the American College of Cardiology. 2016.", "tier": "A", "kind": "human genetic natural experiment (sequencing)", "pmid": "27050191", "pmcid": "PMC5405769", "doi": "10.1016/j.jacc.2016.03.520", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC5405769/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "KHERA-2016-FH-SEQ", "access_depth": "full_text_complete", "access_note": "Complete PMC full text read 2026-09-30, including methods, results, discussion and limitations.", "carries_claims": [ "C20", "C27" ] }, { "n": 39, "citation": "Ridker et al. Antiinflammatory Therapy with Canakinumab for Atherosclerotic Disease. The New England journal of medicine. 2017.", "tier": "A", "kind": "randomized controlled trial", "pmid": "28845751", "pmcid": "", "doi": "10.1056/NEJMoa1707914", "url": "https://pubmed.ncbi.nlm.nih.gov/28845751/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "NCT01327846", "access_depth": "abstract_metadata", "access_note": "PubMed full structured abstract (NCBI E-utilities) and bibliographic identity read 2026-09-30; no abstract-external claim is attributed to this source.", "carries_claims": [ "C28" ] } ], "independence_basis": { "required": true, "lines": [ { "label": "LDL-class: human lifelong genetic natural experiments", "sources": [ 2, 3, 4, 5, 6, 38 ], "excludes_error": "Reduces reverse causation, treatment-selection bias, and the possibility that benefit is limited to one drug; residual shared failures are pleiotropy, ancestry, and correlated lipid changes." }, { "label": "LDL-class: HMGCR inhibition randomized outcomes", "sources": [ 7, 8 ], "excludes_error": "Excludes inherited and behavioral confounding for statin treatment; both trials remain one class line and share possible HMGCR-specific effects." }, { "label": "LDL-class: NPC1L1 inhibition randomized outcome", "sources": [ 9 ], "excludes_error": "Excludes a statin-only explanation through a different lipid-lowering target; genetic NPC1L1 evidence [4] is target-concordant rather than wholly independent." }, { "label": "LDL-class: PCSK9-antibody randomized outcomes", "sources": [ 10, 11, 13 ], "excludes_error": "Excludes HMGCR- and NPC1L1-specific explanations across prevention contexts; the three trials are counted as one pharmacologic class line." }, { "label": "LDL-class: ATP-citrate-lyase inhibition randomized outcome", "sources": [ 12 ], "excludes_error": "Adds a separate target and a statin-intolerant population, reducing dependence on statin tolerance or one pathway." }, { "label": "LDL-class: direct lesion mechanism and trajectory", "sources": [ 1, 16 ], "excludes_error": "Links particle retention and lowering to lesion formation and plaque change rather than only event coding; residual failures are species translation and imaging-surrogate interpretation." }, { "label": "LDL-class: CETP inhibition (genetic and randomized)", "sources": [ 25, 26, 27, 28 ], "excludes_error": "A lipid-transfer mechanism distinct from synthesis, absorption, receptor-recycling and ACL targets; concordant when indexed by apoB/non-HDL-C, with the evacetrapib null under LDL-C indexing retained as an exception." }, { "label": "SC4 (a): multivariable and pathway-contrast Mendelian randomization", "sources": [ 17, 18, 19, 20, 25, 35, 36 ], "excludes_error": "Excludes confounding by lifestyle and treatment and reverse causation; shared failure modes are instrument choice, pleiotropy and collinearity. Internally contradictory [20,35,36]." }, { "label": "SC4 (b): observational discordance cohorts", "sources": [ 21, 30, 31, 32, 33, 34 ], "excludes_error": "Excludes MR instrument artefacts by measuring the traits directly in independent populations; shared failure mode is confounding by insulin resistance and adiposity (cap 2 binds this line alone)." }, { "label": "SC4 (c): pharmacological discordance", "sources": [ 26, 27, 29 ], "excludes_error": "Excludes genetic pleiotropy and lifelong-exposure differences by randomized drug-induced discordance between LDL-C or remnant cholesterol and apoB; shared failure mode is drug-specific off-target effects." } ], "shared_failure_mode_checked": true, "note": "Seven lines support LDL-class causality after grouping correlated publications (SC1-SC3, SC5). SC4 has three lines that converge against LDL-C as the unique quantity; only line (a) addresses particle number versus total cholesterol cargo directly, and it is internally contradictory, which is why SC4 is Ζ." }, "redteam_synthesis": { "date": "2026-09-30", "red_team_model": "Claude (Anthropic), cross-model red team; authoring model(s) differed", "inputs": [ "ldl-particle-concentration-ascvd-causality-hsef-v54-review.md", "ldl-particle-concentration-ascvd-causality-hsef-v54-scorecard.json (schema 5.4)", "ldl-particle-concentration-ascvd-causality-hsef-v54-validation.txt (validator 2.4)", "ldl-particle-concentration-ascvd-causality-hsef-v54-preregistration.md (kept byte-identical)", "README-ldl-particle-concentration-ascvd-causality-hsef-v54.txt" ], "mode": "single-version red team", "findings_total": 32, "findings_accepted": 25, "findings_rejected": 7, "findings_accepted_note": "25 accepted includes RT-20 (accepted as clarification) and RT-25 (preregistration issue logged, file not edited).", "findings_by_severity": { "Critical": 1, "Major": 6, "Minor": 17, "Editorial": 1, "Rejected_candidates": 7 }, "score_before": { "level": "Η", "confidence": "High" }, "score_after": { "level": "Ζ", "confidence": "Moderate" }, "grade_changed": true, "grade_change_rationale": "SC4 (particle-number specificity), a preregistered weakest-link conjunct, was graded Η although the framework's Η definition ('no major unresolved upstream/downstream contradiction within the claimed scope') and rubric 06's Η requirements ('Robust cross-context intervention evidence; Mechanistic specificity') are not met: no intervention separates particle number from total cholesterol cargo and contradictions are replicated across independent groups (Helgadottir 2022; Björnson 2023, 2024). SC4 meets Ζ; weakest-link returns Ζ. SC1-SC3 and SC5 are unchanged in level (Η/Θ); scope-matched aggregation would return Η. Confidence High -> Moderate because the binding subclaim is Moderate.", "citations_checked": 24, "citations_corrected": 1, "citations_removed": 0, "citations_added": 15, "citation_note": "All 24 original PMIDs resolve to the stated works (no PMID confabulation); one first-author misattribution corrected (ref 13, Bohula not Sabatine). 15 verified sources added. Registry artifacts NCT05202509, NCT03705234 and NCT04023552 (and the Lp(a)HORIZON sponsor topline, Tier C) are named for momentum/stability only and are not ledger rows.", "preregistration_sha256_verified_unchanged": "7f7b63a26177ab8f1e5243cc96cef7a8497ff89d97f46c510a5f8391556394a5", "audit_trail": "redteam/REDTEAM_REPORT.md; redteam/SYNTHESIS_CHANGE_LOG.md" }, "consensus_validation": { "slug": "ldl-ascvd", "completed": true, "provider_mode": "Consensus MCP keyword search and fetch; independent reconciliation against the frozen review ledger and its documented PubMed primary-record checks", "cutoff": "2026-09-29 (frozen review cutoff)", "queries": [ "LDL cholesterol causally causes atherosclerotic cardiovascular disease Mendelian randomization statin randomized trials PCSK9 genetic natural experiment", "primary randomized trials 4S Scandinavian Simvastatin Survival Study 1994 PCSK9 genetic variants lower LDL coronary heart disease Cohen 2006 IMPROVE-IT FOURIER primary studies" ], "raw_paths": [ "checks/ldl-ascvd/consensus-search-01.json", "checks/ldl-ascvd/consensus-search-02.json", "checks/ldl-ascvd/consensus-fetch-bfa63e6a82435b95a67f5b7b4638f7ef.json", "checks/ldl-ascvd/consensus-fetch-d976dec09c2754d2a7c605e3c5691019.json" ], "source_counts_preserved": { "evidence_inventory_rows": 33, "claims": 28 }, "primary_comparisons": [ { "pmid": "12066187", "doi": "10.1038/nature00804", "review_source": "Skålén et al. 2002", "source_type": "primary animal mechanism experiment", "access_depth": "PubMed abstract and metadata as documented in review", "result_comparison": "ApoB proteoglycan-binding change altered murine atherosclerosis; supports mechanism but not human outcomes or unique LDL particle-number causality.", "already_present": true }, { "pmid": "16554528", "doi": "10.1056/NEJMoa054013", "review_source": "Cohen et al. 2006", "source_type": "human genetic natural experiment", "access_depth": "PubMed abstract and metadata as documented in review", "result_comparison": "PCSK9 loss-of-function and LDL-C reduction associated with lower CHD risk; LDL-C and particle traits co-vary, leaving particle-versus-cargo unresolved.", "already_present": true }, { "pmid": "7968073", "doi": null, "review_source": "4S", "source_type": "primary randomized secondary-prevention trial", "access_depth": "PubMed abstract and metadata as documented in review; DOI absent from original ledger", "result_comparison": "Simvastatin reduced coronary events and mortality in established CHD; supports LDL lowering, not LDL-P versus LDL-C specificity.", "already_present": true }, { "pmid": "28304224", "doi": "10.1056/NEJMoa1615664", "review_source": "FOURIER", "source_type": "primary randomized outcome trial", "access_depth": "PubMed abstract and final metadata as documented in review", "result_comparison": "Evolocumab lowered LDL-C and cardiovascular events, while LDL-C, apoB, and LDL-P co-move; does not isolate particle count from cholesterol cargo.", "already_present": true } ], "score_before": { "hsef_level": "Ζ", "hsef_descriptor": "Multiple independent causal lines converge", "confidence": "Moderate", "confidence_rationale": "The binding subclaim (SC4) has dense but directionally inconsistent evidence from independent groups and heterogeneous assays (NMR LDL-P, immunoassay apoB, calculated non-HDL-C and remnant-C). Confidence in LDL-class causality (SC2, SC3, SC5) remains High and is reported separately.", "rests_on_non_peer_reviewed_work": false, "rationale": "Cumulative LDL-class apoB exposure is a primary causal driver of ASCVD (SC1, SC3, SC5 at Η; SC2 at Θ). The submitted wording adds that particle concentration (number) is the operative quantity; that is well supported against LDL-C but contested against total cholesterol cargo, and per-particle risk varies by particle class [20,35,36]. The preregistered weakest-link rule therefore returns Ζ; scope-matched aggregation would return Η." }, "score_after": { "hsef_level": "Ζ", "hsef_descriptor": "Multiple independent causal lines converge", "confidence": "Moderate", "confidence_rationale": "The binding subclaim (SC4) has dense but directionally inconsistent evidence from independent groups and heterogeneous assays (NMR LDL-P, immunoassay apoB, calculated non-HDL-C and remnant-C). Confidence in LDL-class causality (SC2, SC3, SC5) remains High and is reported separately.", "rests_on_non_peer_reviewed_work": false, "rationale": "Cumulative LDL-class apoB exposure is a primary causal driver of ASCVD (SC1, SC3, SC5 at Η; SC2 at Θ). The submitted wording adds that particle concentration (number) is the operative quantity; that is well supported against LDL-C but contested against total cholesterol cargo, and per-particle risk varies by particle class [20,35,36]. The preregistered weakest-link rule therefore returns Ζ; scope-matched aggregation would return Η." }, "grade_changed": false, "conclusion": "Consensus discovery does not resolve the frozen review's load-bearing LDL particle-number versus cholesterol-cargo discordance. Established LDL-class causality remains distinct from the stronger concentration-specific wording; no grade or inventory change.", "addendum_path": "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\consensus-addendum.md", "independent_blinded_audit": false, "additional_original_source_receipts": [ "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\primary-source-verification.json", "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\primary-source-verification.md" ], "primary_comparison_count": 4, "status": "completed_retrospective_original_source_reconciliation", "raw_receipts": [ { "path": "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\consensus-search-01.json", "sha256": "0a86333a1e88d8f05b3bc891f3a70e701a5785a6b5b1808253961723ee3db238" }, { "path": "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\consensus-search-02.json", "sha256": "9823fc01dd209e603f9d31c001fd0b67a135d6039f9453ad77edde41deb98a00" }, { "path": "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\consensus-fetch-bfa63e6a82435b95a67f5b7b4638f7ef.json", "sha256": "de62df25f4cce5c90196a67ffc7d45af91cccd3e26179fa336976f53fb5f0f08" }, { "path": "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\consensus-fetch-d976dec09c2754d2a7c605e3c5691019.json", "sha256": "9617618a4b7e23884b96254704e43860f42662d62e6dc83c01bcaff80e951dfa" }, { "path": "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\primary-source-verification.json", "sha256": "b328c9b02a5847585f92b4e6daef22c9d8a1252c642412ef7de69aabd8014a1b" }, { "path": "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\primary-source-verification.md", "sha256": "46b6f364ff4b3177a1daf651ec91e2a02d46b2e0c076ea1cd089b172ebb28e87" } ], "manifest_source": "deliverables\\consensus_consistency_audit_2026-10-09\\checks\\ldl-ascvd\\final-structured-consensus-validation.json", "literature_cutoff_preserved": true } }