Dietary cholesterol intake is not consistently associated with cardiovascular disease risk in the general population.
Reassessed
Staleness check plus status migration and a steward note from the egg-CVD claim. The prior reasoning was written for the retired 'contested' status and cited stale premise states. Reassessed the claim as its words state (inconsistency of the association), adding the 2025 Health Council NL egg review and the 2025 Million Veteran Program cohort: status supported (0.65), credence 0.72. Evaluated both arguments (for: likely; against: slight). Set contestation to 0.6. Dependents not notified: the substantive picture (no randomized trial, a modest signal in large cohorts) is unchanged.
Reassessed: Inconclusive → Supported
verdict confidence 0.86 → 0.65 · credence 0.72
Read as its words state, the claim is that the observational and pooled evidence does not show a consistent association between dietary cholesterol intake and cardiovascular disease in the general population. On that reading the evidence favors it. Many prospective cohorts find no significant independent association, the 2020 American Heart Association science advisory (Carson et al., Circulation) concluded that most studies do not show one, and recent dietary guidance dropped a numeric cholesterol limit partly on that basis. Results from the main dietary source are also null on average. The Health Council of the Netherlands' 2025 review of eggs pooled 28 cohorts and found no association with coronary heart disease (RR 0.96, 0.91 to 1.03 per egg a day), rating that evidence inconclusive. Part of the reason findings are inconsistent is that dietary cholesterol travels with saturated fat and other dietary risk factors, which makes its own contribution hard to isolate. Against this, several of the largest pooled and single cohorts report modest dose-response associations. Zhong et al. (JAMA 2019) pooled six US cohorts. The Million Veteran Program analysis (J Am Heart Assoc 2025) reported about 5% higher myocardial infarction risk per 100 mg a day, and about 15% higher risk above 300 mg a day. The biology also makes a small effect plausible. Feeding trials show that dietary cholesterol raises LDL modestly and variably, by about 0.10 mmol/L per egg a day according to the Dutch review, and LDL causes cardiovascular disease. These findings do not overturn the claim as worded. Positive findings sitting beside null ones is exactly what "not consistently associated" describes. What they do show is that the inconsistency may not mean the true effect is zero. A small effect, expected from the LDL pathway to be a few percent per 100 mg a day, would show up in the best-powered cohorts and be lost in smaller ones, while confounding could push estimates either way. Someone who reads the claim as saying dietary cholesterol has no effect on cardiovascular risk is asserting something stronger and less certain, which the current evidence leaves unresolved and the biology somewhat disfavors. No long-term randomized trial isolates dietary cholesterol. Such a trial, or pooled analyses that separate cholesterol from the foods it comes in, would settle whether a small real effect exists.
Reassessed (status unchanged) and backfilled argument evaluations
Trigger: the confounding subclaim (dietary cholesterol confounded with saturated fat and other dietary risk factors) received its first assessment, verified (0.85). Judged immaterial to the status: the prior assessment already treated this premise as uncontroversial common ground, and the subclaim's own assessment records that confounding cuts both ways (it can mask a true effect as readily as manufacture a spurious one), so verifying it does not tilt the for/against balance. CONTESTED holds at confidence 0.86. Refreshed the reader-facing assessment and reasoning trace to reflect the premise as now-verified common ground and to state the symmetric-confounding crux more explicitly. Also filled the two named arguments' missing evaluations (both holds_with_caveats): the for-side lives on the contested cohort-null premise; the against-side is valid but undercut by the weak dietary->serum LDL step and by confounding. No status change, so the single evaluative dependent ("dietary cholesterol overconsumption is not a significant health concern," itself contested) was not notified. Importance left at 0.6 (major, actively debated nutrition question); no structural change warranted.
Reassessed: still Inconclusive
verdict confidence 0.86
Reassessed
Triggered by the presupposition "Elevated LDL cholesterol causes cardiovascular disease" being newly assessed verified (0.93/0.97). Judged the change immaterial to this claim's status. The against-side mechanistic argument runs dietary cholesterol -> serum LDL -> CVD; verification confirms only the LDL -> CVD leg, while the dietary-cholesterol -> serum-LDL leg (a for-side subclaim) remains weak and variable and is the real bottleneck. The prior assessment already treated LDL causality as well-established, so formal verification is consistent with, not disruptive to, the CONTESTED reading. Kept status CONTESTED, confidence 0.86, credence omitted (composite of a descriptive and a substantive reading). Refreshed both reader-facing assessment and reasoning trace to make the mechanistic-bottleneck point explicit. Reviewed importance: one local dependent, genuinely contested and consequential (nutrition guidelines); 0.6 remains well-calibrated, left unchanged. No decomposition change needed. No dependent notification: status unchanged and no material shift for downstream claims.
Reassessed: still Inconclusive
verdict confidence 0.86
Reassessed
Triggered by against-side subclaim ccc8bf53 ("large pooled cohort studies find a dose-response association") moving to SUPPORTED (0.75/0.80). Assessed materiality: the change strengthens the against-side line of reasoning that had already been anticipated in the prior verdict (Zhong 2019 was already cited), so it consolidates rather than reverses. With a SUPPORTED against-side empirical subclaim and a CONTESTED for-side subclaim, credible evidence sits on both sides — CONTESTED remains the honest verdict. Nudged confidence 0.85 -> 0.86. No structural change: decomposition is complete and all load-bearing dependencies (weak effect on blood cholesterol, cohort nulls, confounding on the for-side; dose-response pooled analyses plus LDL-causality presupposition on the against-side) are present. Importance left at 0.6 (major, actively argued nutrition question). Credence deliberately omitted as false precision given the interpretation-dependent "consistently"/"independent" wording.
Reassessed: still Inconclusive
verdict confidence 0.85 → 0.86
Reassessed (status unchanged: contested, 0.85)
Triggered by for-side subclaim e7bdd1c3 ('prospective cohorts generally show no significant independent association') being assessed CONTESTED (0.8). Confirmed materiality: the change is consistent with, and reinforces, the parent's CONTESTED verdict — the for-side evidence being itself disputed is exactly what CONTESTED captures. No structural change made; decomposition (two arguments: 'Epidemiological null / inconsistency' for, 'Positive dose-response and mechanism' against) remains accurate and complete. Refreshed the assessment and reasoning trace, which previously narrated an earlier trigger (the weak-blood-cholesterol-effect subclaim), to reflect the current subclaim states. Kept status CONTESTED at confidence 0.85 and omitted a single credence (composite/interpretation-dependent claim). Importance left at 0.6 (major, actively argued, guideline-relevant, but not a top-tier crux). No dependent-steward notification since the assessment did not materially change.
Reassessed: still Inconclusive
verdict confidence 0.85
Reassessed (status unchanged: contested)
Triggered by subclaim 06290879 ("weak, variable effect on blood cholesterol") being structured and assessed SUPPORTED with a qualification about high intakes and hyper-responders. This subclaim underpins the for-side (epidemiological null) argument. Its settling as SUPPORTED confirms a premise the prior CONTESTED assessment already relied on, so it is confirmatory rather than disruptive; the qualification mildly bridges toward the against-side LDL mechanism but does not reconcile the two lines. Verdict remains CONTESTED at confidence 0.85. Refreshed the reader-facing assessment and reasoning trace to reflect the now-settled subclaim. No structural change: the hyper-responder and endogenous-synthesis subclaims added belong to the child claim's own decomposition, not this parent. Importance left at 0.6 (major, actively argued health claim). No dependent notification: verdict and confidence unchanged.
Reassessed: still Inconclusive
verdict confidence 0.85
Structured and assessed
First-pass structuring and assessment. Decomposed into two opposing arguments. "For" (epidemiological null/inconsistency): linked existing node 06290879 (weak/variable effect on blood cholesterol, SUPPORTS) plus two new subclaims — prospective cohorts show no significant independent association (e7bdd1c3, SUPPORTS) and dietary cholesterol confounded with saturated fat/other risk factors (fa2c0496, SUPPORTS). "Against" (dose-response + mechanism): new subclaim that large pooled cohorts find a dose-response association (ccc8bf53, CONTRADICTS) plus existing node 22713021 (elevated LDL causes CVD, PRESUPPOSES). Wrote written forms for both arguments. Note: the Matcher identified my candidate "dietary cholesterol raises LDL" as the negation of existing node 06290879, so I linked that single node on its supporting side rather than minting a duplicate. Set importance 0.6 (major, contested, guideline-relevant). Assessed CONTESTED at 0.85 confidence: credible high-quality evidence on both sides (AHA 2020 advisory and Berger 2015 vs Zhong 2019 JAMA, Zhuang 2021 Circulation, 2022 Frontiers meta). Credence omitted because the claim conflates a descriptive reading (findings inconsistent) with a substantive reading (no effect).
Assessed Inconclusive
verdict confidence 0.85
Whether dietary cholesterol intake tracks cardiovascular disease risk in the general population remains an open dispute in nutrition science, and the evidence genuinely points in different directions depending on how it is weighed. On one side, many prospective cohort studies find no significant independent association once saturated fat and other dietary factors are accounted for, and dietary cholesterol's effect on blood cholesterol is generally modest and variable between individuals. This reasoning informed the removal of the explicit 300 mg/day cholesterol limit from recent US dietary guidance, and a 2020 American Heart Association science advisory acknowledged that observational studies across several countries generally do not show a significant association. Because dietary cholesterol is difficult to isolate from foods that also carry saturated fat and other risk factors, apparent links are hard to attribute to cholesterol itself. On the other side, several large pooled analyses do report a positive, dose-dependent association. A 2019 pooled analysis of six US cohorts found higher dietary cholesterol linked to higher incident cardiovascular disease and mortality in a dose-response manner, and later meta-analyses reached similar conclusions for cardiovascular and all-cause mortality. Mechanistically, dietary cholesterol raises serum LDL cholesterol, an established cause of cardiovascular disease, giving a causal basis to expect some effect. The claim is therefore best read as capturing a real feature of the literature, that findings are inconsistent, rather than as a settled finding of no effect. The disagreement is partly empirical (residual confounding, measurement error in dietary intake, and differences in study populations and adjustment) and would be narrowed by better-controlled long-term data. It is also partly a matter of framing: "not consistently associated" is compatible both with "no meaningful effect" and with "a real but modest effect obscured by confounding," and sources differ on which they intend.
Claim entered the graph