Metabolic & Cardiometabolic
bempedoic acid decreases LDL cholesterol (~20-25%)
In plain terms: Does bempedoic acid lower LDL cholesterol, and by how much?
Strong support Metabolic & Cardiometabolic 💰 Industry COI noted
Part of: 💊 bempedoic acid
RefutedContestedStrong support
consensus score 1.00
Yes — consistently ~15-25% LDL-C reduction (≈17-18% on a statin background, ~21-28% as monotherapy/statin-intolerant; ~38% with ezetimibe), proven across the CLEAR phase 3 program and multiple meta-analyses.
📅 Last reviewed: 2026-07-15 ⓘ
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
14 support 0 contradict 0 tested null 0 mixed · 14 sources, 13 independent groups · 4 superseded — pooled inside a review, counted once
The evidence (19)
| Source | Grade | Stance | Quality | Finding |
|---|---|---|---|---|
| Giammanco A et al 2026 · Giornale italiano di cardiologia | observational | supports | moderate | MILOS (NCT04579367), 'a European prospective observational study' — Italian subset, 8-week interim: 'Of the 1310 patients enrolled, 855 (BA, n = 445; BA+EZE FDC, n = 410)' had paired LDL-C, and after a mean 59.3 days 'mean LDL-C reduction of 22.6 ± 32.3% was observed, with a median reduction of' '25.4% (interquartile range 8.1-43.2), from 2.9 ± 1.2 mmol/l (110.7 ± 44.8 mg/dl)' to 2.1 ± 0.8 mmol/l (79.2 ± 32.2 mg/dl). Supports the ~20-25% LDL-lowering claim, but held at moderate quality, not high: single-arm and uncontrolled, six authors are Daiichi Sankyo employees, the SD (32.3%) exceeds the mean effect, 410/855 (48%) took the bempedoic acid+ezetimibe fixed-dose combination so the pooled figure is not a monotherapy effect (no per-arm split in the abstract), and 'Before starting BA or BA+EZE FDC, 33.1% of patients were not receiving' lipid-lowering therapy at all. |
| Zhao X, Ma X, Luo X, Shi Z, Deng Z, Jin Y, Xiao Z, Tan L, Liu P, Jiang S, Shu Y, Tang B, Qiu C 2020 · BMC Pharmacol Toxicol 2020;21(1):86 | meta-analysis | supports | moderate | Zhao 2020 (BMC Pharmacol Toxicol 21:86), random-effects meta-analysis of 13 double-blind RCTs / 4858 hypercholesterolemic adults (mean age 61, baseline LDL-C ≥70 mg/dL, 72% high CVD risk). Primary endpoint is the net LSM percentage change in LDL-C, so the claim's outcome IS the paper's outcome. Monotherapy vs placebo: −25.01% (95% CI −30.66 to −19.35, I² = 86%) — the top of the claim's ~20–25% object. The increment on top of background therapy is smaller: bempedoic acid + statin vs statin −18.37% (−20.16 to −16.57, I² = 0); + ezetimibe vs ezetimibe −18.89% (−29.66 to −8.13, I² = 87%); + ezetimibe vs placebo −37.82% (−41.85 to −33.79, I² = 0). Two gradients temper the headline: effect decays with duration (−22% at wk 8 → −19.13% wk 12 → −18.65% wk 24 → −16.10% wk 52) and with background statin intensity (−28.49% none, −20.02% low/moderate, −17.64% high, p<0.001) — yet it stays large even on high-intensity statin. Head-to-head, 'difference in the magnitude of LDL-C reduction between bempedoic acid and ezetimibe was not obvious.' Secondary lipids move too: TC −11.40%, non-HDL-C −13.57%, ApoB −12.73%, hsCRP −20.72% vs statin alone. Non-industry synthesis of industry trials — Hunan provincial grants, 'The authors declare that they have no competing interests.' Quality held at MODERATE, not high: trial-level (not IPD) pooling, I² = 86% in the monotherapy pool, 'allocation concealment in 7 trials and completeness of outcome data in 3 trials were classified as 'unclear risk of bias'', publication bias judged by funnel-plot eyeball alone with no Egger test, and no PROSPERO protocol. Scope limit stated by the authors: 'this study failed to identify the cardiovascular risk reduction by bempedoic acid' — LDL only, CLEAR Outcomes was still ongoing. |
| Yamashita 2025 · Circ J | RCT | supports | moderate | CLEAR-J (Japanese phase 3): bempedoic acid significantly lowered LDL-C vs placebo, replicating Western trials in a distinct population. |
| Ray 2019 · N Engl J Med | RCT | supports | high | CLEAR Harmony (n=2230, on max-tolerated statin): LDL-C -16.5% from baseline, placebo-corrected -18.1 percentage points at wk 12. ↩ SUPERSEDED — pooled in the review above, counted once |
| Laufs U, Banach M, Mancini GBJ, Gaudet D, Bloedon LT, Sterling LR, Kelly S, Stroes ESG 2019 · J Am Heart Assoc. 2019 Apr 2;8(7):e011662 (e-pub 29 Mar 2019) | RCT | supports | high | CLEAR Serenity (NCT02988115, n=345 statin-intolerant, 2:1, 24 wk): LDL-C placebo-corrected -21.4% at wk 12 (95% CI -25.1 to -17.7), attenuating to -18.9% at wk 24 (95% CI -23.0 to -14.9); on-treatment (adherent) estimate -23.9% (95% CI -27.5 to -20.2). apoB fell less than LDL-C (-15.0%). [FT-verified 2026-08-19 from PMC6509724] ↩ SUPERSEDED — pooled in the review above, counted once |
| Ballantyne CM, Banach M, Mancini GBJ, Lepor NE, Hanselman JC, Zhao X, Leiter LA 2018 · Atherosclerosis 2018 Oct;277:195-203 | RCT | supports | moderate | CLEAR Tranquility (NCT03001076), phase 3 double-blind placebo-controlled RCT, n=269 statin-intolerant patients (181 bempedoic acid, 88 placebo) already on ezetimibe after a 4-week run-in; primary endpoint percent change in LDL-C at 12 weeks. 'Bempedoic acid added to background lipid-modifying therapy that included ezetimibe reduced LDL-C by 28.5% more than placebo (p < 0.001; -23.5% bempedoic acid, +5.0% placebo).' Placebo-corrected effect is therefore 28.5%; -23.5% is the within-arm change from an already-ezetimibe-treated baseline (the prior extract had these two reversed, and its '~38% combo total' figure appears nowhere in the paper). Effect is incremental to ezetimibe, not monotherapy. Secondary endpoints moved with it: 'Significant reductions in secondary endpoints, including non-high-density lipoprotein cholesterol (-23.6%), total cholesterol (-18.0%), apolipoprotein B (-19.3%), and high-sensitivity C-reactive protein (-31.0%), were observed with bempedoic acid vs. placebo (p < 0.001).' Sponsor-run: Esperion Therapeutics funded the trial and two co-authors give their affiliation as 'Clinical Development, Esperion Therapeutics, Inc.' Abstract-only - full text not obtainable (not in PMC, subscription required), so baseline lipids, dropout handling and the disclosure statement are unverified. ↩ SUPERSEDED — pooled in the review above, counted once |
| Heikal M et al 2026 · Pharmaceutics | animal | supports | low | Pharmaceutical formulation study (MDPI Pharmaceutics 2026) whose in vivo arm gave 25 Sprague-Dawley rats (n = 5/group) a bempedoic acid + ezetimibe combination at 18/1 mg/kg for one week after acute hyperlipidemia induction with a single 100 mg/kg intraperitoneal Triton X-100 injection. Serum LDL fell from 170.39 ± 18.40 mg/dL in untreated hyperlipidemic rats to 30.25 ± 1.93 mg/dL on the test tablet and 27.89 ± 3.41 mg/dL on Nexlizet (p < 0.0001 for all treated groups vs untreated), and was statistically indistinguishable from the 14.92 ± 3.68 mg/dL uninduced control. The test formulation was NOT superior to the reference product — that comparison was explicitly non-significant (p = 0.9847); equivalence to Nexlizet was the study's objective. Directionally consistent with the claim but weak evidence for it: bempedoic acid was never given alone, so the LDL fall cannot be attributed to it rather than to ezetimibe; the Triton X-100 detergent model inhibits lipoprotein lipase rather than reproducing human dyslipidemia; and nothing here bears on the claim's ~20-25% human effect size. |
| Nissen 2023 · N Engl J Med | RCT | supports | high | CLEAR Outcomes (n=13,970 statin-intolerant): bempedoic acid cut LDL-C ~21% vs placebo at 6 mo (baseline 139 mg/dL), durable over trial. |
| Ballantyne 2022 · Am J Cardiol | observational # was RCT — single-arm open-label extension, no concurrent control (reread #141, precedent k) | supports | low | CLEAR Harmony 78-week OPEN-LABEL EXTENSION, single-arm (no concurrent control; efficacy = % change from parent-study baseline). Of 1,462 OLE enrollees, 970 prior bempedoic acid: 'reductions in LDL-C, other lipid parameters, and hsCRP observed in the parent study remained stable through 130 weeks of treatment'; the 492 former-placebo patients 'experienced reductions in LDL-C, other lipid parameters, and hsCRP, mirroring reductions observed in patients who received bempedoic acid in the parent study'. DIRECTION ONLY - no LDL-C effect size is reported in the available text, so this cannot corroborate the claim's ~20-25% object. Two authors are Esperion Therapeutics employees; abstract-grade provenance. |
| Hsieh IC et al 2026 · J Am Heart Assoc | observational | supports | moderate | CLEAR Taiwan (NCT06925100): prospective, pragmatic, phase 4, multicenter SINGLE-ARM open-label study — no randomization, no comparator arm — n=180 enrolled, 160 (88.9%) completers analyzed (complete-case, no imputation). Bempedoic acid 180 mg/d added to background lipid-lowering therapy for 12 weeks: median LDL-C change -19% (IQR -36.4% to -3.6%), 117.5 to 92 mg/dL, P<0.01; 31.3% reached LDL-C targets. Because there is no control arm this is a within-person pre/post change, not a placebo-corrected effect, and adherence explained part of the spread (poor adherence accounted for 25% of patients whose LDL-C rose >10%). Funded by Daiichi Sankyo Taiwan; three sponsor employees are authors and one wrote the first draft. |
| Ridker, Lei, Ray, Ballantyne, Bradwin, Rifai 2023 · J Clin Lipidol | RCT | supports | moderate | Post-hoc biomarker substudy of CLEAR Harmony (NCT02666664), restricted to the 817 statin-treated patients with residual inflammatory risk (hsCRP >=2 mg/L), randomized 2:1 to bempedoic acid 180 mg vs placebo for 12 weeks: placebo-corrected median LDL-C change -21.1% (-23.7 to -18.5), ApoB -13.1% (-15.5 to -10.6). LDL result is randomized and CI-clear, but this is a subgroup of a trial already represented in the corpus by s30865796, an Esperion Therapeutics employee is a co-author, and the abstract's 'pattern ... almost identical to ... statin therapy' framing rests on one inflammatory hit (hsCRP -26.5%) against two nulls (IL-6 -3.7% [-11.5, 4.3]; fibrinogen 2.1% [-2.0 to 6.4]). |
| Khattak S, Ochoa-Ferraro A, Khan N, George S, Khan SQ, Townend JN, Dawson C, Thomas MR 2025 · J Clin Med | observational | supports | low | Single-centre retrospective observational cohort (University Hospitals Birmingham lipid clinic, 2017-2024), n=256 across four agents, of whom '30 patients received bempedoic acid'. Direction is clear: 'In patients treated with bempedoic acid, LDL-C decreased by 36% compared to baseline' ... 'to a mean of 2.68 mmol/L after 12-24 months, with 12% of patients reaching the ESC target and 29% reaching the NICE target' (baseline LDL-C 4.1 mmol/L). Magnitude is NOT usable and runs well above the randomised evidence, as the authors themselves note: 'Our study showed that bempedoic acid reduced plasma LDL-C levels by approximately 35%, a greater reduction than the 18% reported in clinical trials' - their explanation (80% female subgroup) is post-hoc, and there is no control arm, no adjustment, and no treatment of regression to the mean, with patients compared only to their own pre-initiation value ('Data were extracted retrospectively from electronic medical records.'). The subgroup is also unusually lightly pre-treated for an add-on agent (22.7% on a high-intensity statin, 72.7% on ezetimibe, Table 1), which would further inflate the observed drop. Reporting is unreliable at the margins: the abstract's interval is incoherent ('bempedoic acid with a reduction of 36% (95% CI, 22 to 69'), and the Discussion restates the target attainment as 'only 30% and 10% of patients, respectively, achieved the ESC- and NICE-recommended LDL-C targets on bempedoic acid', contradicting its own Results. Authors concede 'The sample size was relatively small for patients treated with bempedoic acid but was sufficiently powered to detect a 30% reduction in LDL-C'. J Clin Med (MDPI) 2025;14(22):7946; no manufacturer funding for bempedoic acid disclosed. Counts as real-world directional corroboration only; the ~20-25% magnitude in the claim object should continue to come from the randomised/meta-analytic evidence, not from here. |
| Nelson J, Bloedon L, Lewandowski D, Vaduganathan M, Ajose M, Bonafede M, Sarnes E 2026 · Atherosclerosis Plus | observational | supports | moderate | Retrospective, UNCONTROLLED pre-post cohort in the Veradigm Network EHR linked to Komodo open claims, US 2020-2024; n=900 bempedoic acid alone, n=615 BA+ezetimibe fixed-dose combination. Median baseline LDL-C 137 (100-167) and 127 (93-163) mg/dL. By 3 months median LDL-C fell to 108 (80-138) mg/dL on BA alone, a 21.2% reduction 'maintained at 6 and 12 months'; the combination fell to 85 (61-114) mg/dL (33.1%) at 3 months but WANED to 96 (67-128) mg/dL (22.4%) by 12 months, so 'sustained' holds for BA monotherapy and not for the combination. Attainment of LDL-C <70 mg/dL rose from 6.3%/8.5% at baseline to 16.8%/33.3% at 3 months and 18.0%/27.5% over 12 months. Reductions were larger without prior statin use (-24.2% / -36.7% vs -16.7% / -29.8%) and with continuous use (-26.1% / -40.2%). A closed-claims sensitivity analysis (n=178/163) reproduced the direction (-24.5% / -28.4% at 3 months). WHY THIS IS ONLY MODERATE, NOT HIGH: (1) no comparator whatsoever - 'The major limitation of the analysis is the inability to make causal inference based on these observational results due to the lack of a control cohort of alternative LLT initiators' - so with an elevated-selecting baseline draw, regression to the mean is uncontrolled and unmentioned; (2) requiring >=2 follow-up LDL-C labs including one at 12 months meant 'this resulted in a loss of ∼75% of the sample population, potentially biasing towards patients who receive more comprehensive care'; (3) an inference mismatch the authors state outright - 'For descriptive reporting in this paper, median laboratory levels are described though mean lipid values were used for inferential testing' - combined with unpaired t-tests across timepoints where 'patients were not required to have lab tests available at every timepoint', so the p<0.0001 values compare partly different people rather than within-patient change; (4) the primary analysis runs on open claims that 'are not all fully adjudicated'; (5) funding and authorship: 'This study was funded by Esperion Therapeutics, Inc.', the manufacturer, with all seven authors reporting Esperion financial support. The direction is nonetheless corroborated externally - 21.2% here matches the 21% mean LDL-C reduction in CLEAR Outcomes - which is why this stays supports rather than being discounted further. |
| Masuda D et al 2026 · J Atheroscler Thromb | observational | supports | moderate | Phase 3 long-term single-arm study (CLEAR-J LONG), n=130 Japanese patients, bempedoic acid 180mg/day for 52 weeks; LDL-C decreased 21.6% overall (25.3% in newly enrolled subgroup), well tolerated. |
| Venkatraman, Das, Eerike, Cherian, Bagepally 2023 · Eur J Clin Pharmacol 2023;79(11):1453-1463 | meta-analysis | supports | moderate | SR + random-effects MA of 17 RCTs (n = 21,131) to June 2023: 'Treatment with bempedoic acid led to a significant reduction in the mean serum total cholesterol [- 34.41 mg/dl (95% CI: - 42.43 to - 26.39), p < 0.001], low-density lipoprotein cholesterol (LDL-C) [- 33.91 mg/dl (95% CI: - 39.66 to - 28.17), p < 0.001]'. 'The GRADE of evidence was moderate for all outcomes.' Supports the direction; the abstract gives no baseline LDL-C, so the claim's ~20-25% magnitude cannot be checked from available text. Author-independent (ICMR/JIPMER/AIIMS) but pools the CLEAR programme, so not data-independent of CLEAR-based appraisals. |
| Uddin 2023 · Curr Probl Cardiol | meta-analysis | supports | high | 7 RCTs (n=17,816): LDL-C mean difference -22.4% (95% CI -25.9 to -18.8) vs placebo; total cholesterol -13.9%. |
| De Filippo 2023 · Cardiovasc Diabetol | meta-analysis | supports | high | Systematic review + meta-analysis of 11 RCTs / 18,315 patients (9,854 bempedoic acid 180 mg/day vs 8,461 placebo or no treatment; PRISMA, PROSPERO 399,867; 'Funding None' and 'Competing interests None declared'). LDL-C at 12 weeks pooled across 8 trials / 18,130 participants: mean difference -22.42% vs control (95% CI -24.0% to -20.8%) - inside the claim's stated ~20-25% band. Companion atherogenic-lipid endpoints move the same way: total cholesterol -16.5% (-19.2 to -13.8), non-HDL-C -20.3% (-22.6 to -18.0), ApoB -19.6% (-22.7 to -16.4), hs-CRP -28.1% (-31.7 to -24.4), and the effect held at latest follow-up as well as at 12 weeks. The size of the LDL-C effect depends on background therapy: -24.1% with no statin, -19.6% on low/moderate-intensity statin, -15.7% on high-intensity statin (the drug shares the statin pathway upstream), while ezetimibe background made no difference (-19.0% with vs -18.5% without). Two caveats the authors state themselves: 'Significant heterogeneity was observed for all the laboratory efficacy outcomes' (low heterogeneity applies to the clinical-event endpoints, not the lipid ones), and over 70% of the pooled sample is the single CLEAR OUTCOMES trial (Nissen 2023), so this is not an independent vote alongside that trial. |
| Goldberg 2019 · JAMA | RCT | supports | high | CLEAR Wisdom (n=779, on max-tolerated statin): LDL-C placebo-corrected -17.4% at wk 12; also lowered non-HDL-C, ApoB, hsCRP. ↩ SUPERSEDED — pooled in the review above, counted once |
| Feingold KR. 2026 · MDText.com, Inc., South Dartmouth (MA) | observational # grade enum is closed; review-ness lives in evidence_role (zero weight), fixed after vocab ERROR 2026-08-19 | supports | low | Endotext textbook chapter (Feingold KR, chapter revised 2026 Apr 23), not a study: 'Bempedoic acid lowers LDL-C by 15-25% by inhibiting hepatic ATP citrate lyase activity resulting in a decrease in cholesterol synthesis in the liver, a decrease in hepatic cholesterol content, and an up-regulation of LDL receptors.' No new data, no trial named, no n; mono vs add-on still not specified. Regraded observational to review to match the evidence_role already set; zero weight either way. |
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.