Icosapent ethyl (Vascepa)
Metabolic Health · Purified EPA ester
Tier A-
What this is
REDUCE-IT (2018) remains the pivotal Western trial. STRENGTH (mixed EPA+DHA carboxylic acid 2020) was negative — prompting debate whether (a) pure EPA is specifically beneficial vs. (b) REDUCE-IT's mineral oil placebo artifactually inflated effect by raising LDL/ApoB/CRP in the control arm. **RESPECT-EPA** (Japan 2024, n=2460, 1.8 g/day icosapent in CAD with low EPA:AA ratio) helps adjudicate: primary endpoint just missed (HR 0.79, p=0.055) but the secondary coronary endpoint reached significance (HR 0.73) — and crucially RESPECT-EPA used a non-mineral-oil control, weakening the placebo-confound objection. RESPECT-EPA also confirmed the AF dose signal (3.1% EPA vs 1.6% control, p=0.017) at 1.8 g — lower than the 4 g REDUCE-IT dose. 2024 Frontiers Nutrition review proposes analytical approaches to reconcile REDUCE-IT/STRENGTH discrepancies. Current guidelines support use in elevated-TG patients on statins; the RESPECT-EPA data support a lower-dose option in EPA:AA-deficient CAD patients but with explicit AF monitoring.
Mechanism
Highly purified EPA ethyl ester; lowers triglycerides; incorporates into membrane phospholipids; generates resolvins; plaque stabilization; restores EPA:AA ratio (a Japanese-specific clinical biomarker)
Dose & route
4 g/day PO (2 g twice daily with food); RESPECT-EPA-style protocols use 1.8 g/day in EPA:AA-deficient patients
Citations
- https://www.nejm.org/doi/full/10.1056/NEJMoa1812792
- https://www.ahajournals.org/doi/10.1161/CIRCULATIONAHA.123.065520
- https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2024.1490953/full
- https://pubmed.ncbi.nlm.nih.gov/33190147/
- https://pmc.ncbi.nlm.nih.gov/articles/PMC8292716/
- https://pubmed.ncbi.nlm.nih.gov/41483441/
- https://doi.org/10.1002/alz70856_101654
Links go to the source. If a link is dead or you want something re-checked, let me know.
This is an independent synthesis of published research by a non-clinician. Scores are opinions supported by citations, not prescriptions. See the full disclaimer and methodology for how this score was produced and what it does and doesn't mean.