Diets · Metabolic & Cardiometabolic
magnitude of LDL reduction correlates with cardiovascular disease
In plain terms: Does how much a statin lowers LDL predict how much it cuts heart-attack risk across trials?
Leans support Diets 🔬 Includes disconfirming
Part of: 💊 statins
RefutedContestedStrong support
consensus score 0.51
At the TRIAL level the size of LDL-lowering is a weak surrogate for benefit (R-squared near 0), yet per-mmol meta-analyses and Mendelian randomization still show LDL is causal — so this is a surrogacy gap, not proof LDL is harmless.
📅 Last reviewed: 2026-07-14 ⓘ
Evidence ladder
How far up the ladder this claim has climbed. A high consensus on a low rung means "consistent so far," not "proven in people."
Top evidence so far: All trials, pooled (Meta-analysis)
MechanismIn-vitroAnimalObservationalRCTMeta-analysis
How the studies fall
11 support 2 contradict 0 tested null 4 mixed · 17 sources, 11 independent groups
What the evidence shows
At the trial level, the MAGNITUDE of LDL lowering is a weak predictor of the size of a statin's cardiovascular benefit (Byrne 2022; Liaigre 2025, R^2~0). Norwitz uses this to attack LDL. The honest nuance: this is about dose-response SURROGACY across heterogeneous trials - it does NOT overturn LDL CAUSALITY, which Mendelian randomization and per-mmol CTT analyses support.
The evidence (18)
| Source | Grade | Stance | Quality | Finding |
|---|---|---|---|---|
| Wang 2020 · Lancet Diabetes Endocrinol | meta-analysis | supports | high | 327,037 participants: consistent ~21% RR reduction per mmol/L LDL-C lowering across baseline LDL-C and risk strata. |
| Ference 2017 · JAMA | observational | mixed | high | CETP/HMGCR variant study: CV benefit tracks LDL-lowering but is attenuated when LDL is lowered discordantly with ApoB. |
| Silverman 2016 · JAMA | meta-analysis | supports | high | 49 trials/312,175: across statin AND non-statin LDL-lowering, RR per 1 mmol/L LDL reduction ~0.77, consistent — trial-level dose-response for benefit. |
| Hao 2022 · BMJ | meta-analysis | supports | high | Risk-stratified guideline meta-analysis: adding ezetimibe/PCSK9i yields CV benefit proportional to additional LDL-C reduction. |
| Li 2021 · Eur J Clin Invest | meta-analysis | mixed | high | Umbrella review: many drug-outcome associations weak; anchors that magnitude-benefit links are often overstated |
| Byrne P, et al. 2022 · JAMA Intern Med | meta-analysis | contradicts | moderate | [FT-verified] Byrne2022: weak LDL-MI corr, R2=0 for mortality/stroke; benefit not strongly LDL-mediated at trial level |
| Preiss 2020 · JACC | meta-analysis | supports | high | >25 outcome trials: statins lower major atherosclerotic events ~22% per 1 mmol/L LDL-C, similar across subgroups. |
| Burger, Dorresteijn, Koudstaal, Holtrop, Kastelein, Jukema, Ridker, Mosterd, Visseren 2024 · Atherosclerosis | meta-analysis | supports | high | OFF-SCOPE for trial-level surrogacy. The paper's stated analysis is 'random-effects meta-regression analyses for the association between follow-up duration, age, and the HR for major vascular events per 1 mmol/L LDL-C reduction' — magnitude of LDL-C reduction is the denominator of the effect measure, not a tested predictor of effect size, so no surrogacy R^2 is produced. The prior extract also mis-stated the result: the effect did not grow over treatment time — 'Follow-up duration was not associated with a change in the HR for major vascular events (HR for change per year 0.994; 95 % CI 0.970-1.020; p = 0.66)', with the IPD sub-analysis agreeing (0.983; 0.943-1.025; p = 0.42). Headline number retained for context: 'The HR for major vascular events per 1 mmol/L LDL-C reduction was 0.78 (95 % confidence interval [CI] 0.75-0.81).' |
| Khan I 2020 · J Am Heart Assoc | meta-analysis | supports | moderate | A time-dependent treatment-benefit model derived from 22 RCTs (statins and non-statins) used 'magnitude of low-density lipoprotein cholesterol reduction' as a core parameter and outperformed CTT-based estimations in predicting benefit across 15 validation trials. |
| Ennezat PV 2023 · J Cardiovasc Pharmacol | meta-analysis | mixed | high | Systematic review/meta-analysis of 60 RCTs (323,950 participants) found that although lipid-lowering therapy reduced all-cause mortality overall (RR 0.92), 'intensive LDL-C percent lowering was not associated with further reductions in all-cause mortality' (RR 1.00) or |
| Ference 2017 · European Heart Journal | observational | supports | high | EAS consensus: genetic, epidemiologic and trial evidence show LDL causally lowers ASCVD proportional to absolute LDL exposure. |
| Marston 2019 · Circulation | meta-analysis | mixed | high | Per-mmol/L benefit differs by lipid lowered: LDL/ApoB-lowering predicts CV benefit whereas triglyceride-lowering does so weakly. |
| Liaigre L, et al. 2025 · Eur Heart J Cardiovasc Pharmacother | meta-analysis | contradicts | moderate | [FT-verified] Liaigre2025 umbrella 20RCTs/194k: trial-level R2 0-0.1 LDL (valid needs>=0.65); lit discrepant; individual surrogacy holds |
| Sabatine, Wiviott, Im, Murphy, Giugliano 2018 · JAMA Cardiol. 2018 Sep 1;3(9):823-828 | meta-analysis | supports | high | Per-mmol proportionality held at very low baseline LDL-C, across three concordant estimates: statin subgroup (CTTC data, 1,922 events) RR 0.78 (95% CI 0.65-0.94); 3 nonstatin add-on trials (50,627 patients, 9,570 events) RR 0.79 (0.70-0.88); combined RR 0.79 (0.71-0.87, P < .001) per 1-mmol/L LDL-C reduction. THREE CAVEATS: the previously quoted '22%' was the paper's BACKGROUND CTTC figure at a 3.4 mmol/L baseline, not its result; the claim is about STATIN benefit while most of this paper's evidence is nonstatin add-on (ezetimibe, a PCSK9 antibody, a CETP inhibitor) and the statin CI is wide, only just excluding 1; and the paper runs no trial-level regression of LDL-reduction magnitude against benefit size, so it speaks to per-mmol proportionality rather than to the trial-level surrogacy the claim's scope line names. |
| Gencer 2020 · Lancet | meta-analysis | supports | high | Supports, with the design caveat named. 'LDL cholesterol lowering significantly reduced the risk of major vascular events (n=3519) in older patients by 26% per 1 mmol/L reduction in LDL cholesterol (RR 0·74 [95% CI 0·61-0·89]; p=0·0019), with no statistically significant difference with the risk reduction in patients younger than 75 years (0·85 [0·78-0·92]; pinteraction=0·37).' This is a per-mmol dose-response estimate, so it is on-topic — but the proportionality is the meta-analysis's normalisation assumption applied to 'aggregated study-level data', not a trial-level surrogacy regression of benefit against achieved LDL reduction, which is the specific question this claim's nuance turns on. It also shares 24 of its 29 trials with the CTT collaboration. (Abstract-grade.) |
| Baigent, Blackwell, Emberson, Holland, Reith, Bhala, Peto, Barnes, Keech, Simes, Collins 2010 · Lancet 376(9753):1670-1681 | meta-analysis | supports | high | IPD meta-analysis, 26 RCTs / 170,000, addressing the trial-level question directly: across the 21 statin-vs-control trials 'larger absolute reductions in LDL cholesterol were associated with larger proportional reductions in risk (trend p<0.0001), but no significant residual variation remained after adjustment for LDL cholesterol differences (trend p=0.4)'; same pattern across the 5 more-vs-less-intensive trials (trend p=0.0004, residual p=0.05). Per 1.0 mmol/L, 22% fewer major vascular events (95% CI 20-24), with no dependence on baseline LDL (trend p=0.2). |
| Ray KK 2023 · Eur Heart J | RCT | supports | moderate | Pooled ORION-9/-10/-11 phase III trial analysis: inclisiran produced a 50.6% (1.37 mmol/L) LDL-C reduction and significantly reduced composite MACE (OR 0.74, 95% CI 0.58-0.94), consistent with the expected direction of the LDL-lowering-to-CV-benefit relationship, though |
| Khan 2022 · Eur J Prev Cardiol | meta-analysis | supports | moderate | Achieving lower (<70) vs higher LDL-C with intensive lipid therapy reduced major events, consistent with magnitude-benefit relation. |
Disagree, or know a study we missed?
We grade by evidence, not opinions. The way to weigh in is to point us to a study we haven't cited (check the evidence table above first), or to flag a problem with one we have. Every submission is reviewed; if it holds up, the grade updates and shows in Science Changes Its Mind.
Opens a short form. You'll sign in with Google so submissions are tied to a real account — we don't display your identity, and we only accept a link we can verify (PubMed, DOI, ClinicalTrials.gov).
Educational only, not medical advice. Grades and scores reflect published evidence weighted by study design and quality; see the methodology.