遇见数据集

Claim Verification: "Eating eggs significantly raises LDL cholesterol and heart-disease risk." — Disproved

收藏
Zenodo2026-04-09 更新2026-06-05 收录
官方服务:

资源简介:

Automated fact-verification of the claim: "Eating eggs significantly raises LDL cholesterol and heart-disease risk." Verdict: DISPROVED Key Findings SC2 (CVD risk) — directly contradicted: Two independent meta-analyses covering 1.7–2 million participants and 39–28 prospective cohort studies both find that moderate egg consumption (up to 1 egg/day) is not associated with increased cardiovascular disease risk (B1, B2 — both fully verified). SC1 (LDL) — threshold not met: The best RCT meta-analysis evidence (17 RCTs, healthy adults) shows an increase of 8.14 mg/dL in LDL-C with higher egg consumption (B3). This is statistically significant but falls below the 10 mg/dL threshold for clinical meaningfulness — and disappears in dietary contexts low in saturated fat. Compound claim fails: The claim requires BOTH sub-claims to hold. Neither does: SC1's "significantly" is not supported at the clinical level, and SC2 is directly contradicted by large-scale prospective evidence. Historical note: The 2015 U.S. Dietary Guidelines Advisory Committee removed the previous 300 mg/day dietary cholesterol limit, citing "no appreciable relationship between consumption of dietary cholesterol and serum cholesterol," signaling a major shift in the scientific consensus underlying this claim. Files proof.py — Re-runnable Python verification script proof.md — Structured proof report proof_audit.md — Full verification audit trail proof_narrative.md — Plain-language summary proof.json — Machine-readable structured data Generated by Proof Engine v1.0.0.

针对声明"食用鸡蛋会显著升高低密度脂蛋白胆固醇(LDL-C)与心脏病风险"的自动化事实核查。 结论:已证伪。 核心研究发现 SC2(心血管疾病(CVD)风险)——直接相悖:两项覆盖170万至200万参与者、纳入39至28项前瞻性队列研究(prospective cohort studies)的独立荟萃分析(meta-analyses)均发现,适量食用鸡蛋(每日最多1个)与心血管疾病(CVD)风险升高无关联(B1、B2,两项均已完全验证)。 SC1(低密度脂蛋白(LDL))——未达到临床显著性阈值:针对健康成年人的最佳随机对照试验(RCT)荟萃分析证据(共17项RCT)显示,更高的鸡蛋摄入量会使低密度脂蛋白胆固醇(LDL-C)升高8.14 mg/dL。该结果具有统计学显著性,但未达到具备临床意义的10 mg/dL阈值,且在饱和脂肪摄入较低的饮食场景中该差异会消失。 复合声明不成立:该声明需要两个子声明同时成立,但两者均不满足:SC1中的"显著"未达到临床层面的支持标准,而SC2则被大规模前瞻性研究证据直接驳斥。 历史背景:2015年美国膳食指南咨询委员会移除了此前每日300mg的膳食胆固醇摄入上限,理由是"膳食胆固醇摄入与血清胆固醇水平之间无明显关联",这标志着支撑该声明的科学共识发生了重大转变。 相关文件 proof.py —— 可重复运行的Python核查脚本 proof.md —— 结构化核查报告 proof_audit.md —— 完整核查溯源记录 proof_narrative.md —— 通俗语言总结 proof.json —— 机器可读的结构化数据 由Proof Engine v1.0.0生成。

提供机构:
Zenodo
创建时间:
2026-04-09
二维码
社区交流群
二维码
科研交流群
商业服务