Mendelian randomization studies estimate roughly 50-55% lower coronary heart disease risk per mmol/L lower LDL cholesterol
3 events · 1 assessment · 1 decision
Structured and assessed
First pass (structure_and_assess). Left the claim atomic: it is a single, directly verifiable factual report of a magnitude in the MR literature, uncontested, so no invented dependencies were added (avoided quota-filling per §6). Verified against primary sources: Ference et al. 2012 JACC (54.5%, 95% CI 48.8–59.5% per mmol/L) and the 2018 EAS consensus (OR 0.46 ≈ 54% per mmol/L); other genetic instruments cluster nearby, so "roughly 50–55%" is fair. Set importance to 0.25 (contestation 0.10): a supporting empirical premise in a specialized literature, low reach, essentially uncontested. Recorded verified, confidence 0.92, credence 0.95. Referenced the MR-validity assumption (05c19437) and the statin-trial comparator (d81f89e3) inline rather than as new edges since they already exist and the claim is descriptive. No named arguments to write/evaluate.
Assessed Verified
verdict confidence 0.92 · credence 0.95
Mendelian randomization studies of LDL cholesterol converge on a per-mmol/L reduction in coronary heart disease risk in the range this claim describes. The most cited analysis (Ference et al., 2012, JACC) pooled nine LDL-lowering polymorphisms across 312,321 participants and estimated a 54.5% (95% CI 48.8–59.5%) lower risk of coronary heart disease for each 1 mmol/L (38.7 mg/dL) genetically lower LDL-C, and the 2018 European Atherosclerosis Society consensus statement adopts an odds ratio of about 0.46 (roughly a 54% reduction) per mmol/L. Estimates from other genetic instruments cluster nearby, so "roughly 50–55%" is a fair summary of the central finding. The figure is a report of what the studies estimate and is not itself in dispute. Reading these estimates as the causal effect of lifelong lower LDL depends on standard Mendelian randomization assumptions, above all that the genetic instruments affect cardiovascular risk only through LDL, and the estimate is markedly larger than the roughly one-fifth-per-mmol/L effect seen in statin trials, a gap attributed to the much longer duration of exposure that genetic variants capture.
Claim entered the graph