Omega-3 and cardiovascular outcomes: REDUCE-IT, STRENGTH, VITAL, and the EPA-versus-mixture question

Key takeaways

• REDUCE-IT (Bhatt 2018 NEJM) randomized 8,179 statin-treated patients with established CVD or diabetes plus risk factors and triglycerides 135 to 499 mg per dL to icosapent ethyl (purified EPA ethyl ester) 4 grams per day or mineral-oil placebo. The primary 5-point composite endpoint occurred in 17.2 percent versus 22.0 percent (HR 0.75) over 4.9 years .
• STRENGTH (Nicholls 2020 JAMA) randomized 13,078 statin-treated high-CV-risk patients to omega-3 carboxylic acid (EPA plus DHA, Epanova) 4 grams per day or corn-oil placebo. The trial was stopped early for futility; the primary endpoint hazard ratio was 0.99 .
• VITAL omega-3 arm (Manson 2019 NEJM) randomized 25,871 generally healthy adults to 1 gram per day marine n-3 fatty acids (840 mg of EPA plus DHA) or placebo for primary prevention. Major cardiovascular events were not reduced (HR 0.92, NS) .
• The class is not interchangeable. Purified high-dose EPA in a hypertriglyceridemic population reduced events; mixed EPA plus DHA carboxylic acid and low-dose mixed n-3 did not.
• FDA approved icosapent ethyl (Vascepa) for cardiovascular risk reduction as an adjunct to statins in 2019 based on REDUCE-IT .
• The mineral-oil placebo in REDUCE-IT raised LDL and hsCRP modestly, leading some commentators to argue that part of the apparent benefit reflects placebo harm. Independent analyses have judged this insufficient to explain the magnitude of effect, but the controversy is unresolved.

What the class actually is

Marine omega-3 fatty acids are a family of polyunsaturated fatty acids dominated by eicosapentaenoic acid (EPA, 20:5n-3) and docosahexaenoic acid (DHA, 22:6n-3), with smaller amounts of docosapentaenoic acid (DPA). Plant-derived alpha-linolenic acid (ALA) is also classified as omega-3 but converts to EPA only at very low efficiency in humans (typically less than 5 percent).

Mechanistically, omega-3 fatty acids modulate cell-membrane composition, eicosanoid signaling, platelet aggregation, and triglyceride metabolism. The triglyceride-lowering effect of high-dose EPA plus DHA (about 4 grams per day) is well-established, but triglyceride lowering does not automatically translate into hard cardiovascular endpoints; the trial record is the test of that claim.

REDUCE-IT: the positive trial that shifted the field

REDUCE-IT enrolled 8,179 statin-treated patients with established cardiovascular disease (70.7 percent) or diabetes plus at least one additional risk factor (29.3 percent), fasting triglycerides 135 to 499 mg per dL, and LDL 41 to 100 mg per dL (controlled on statin). Patients were randomized to icosapent ethyl 2 grams twice daily (4 grams per day total) or mineral-oil placebo. The primary 5-point composite was cardiovascular death, nonfatal MI, nonfatal stroke, coronary revascularization, or unstable angina .

Over a median 4.9 years, the primary endpoint occurred in 17.2 percent of icosapent ethyl patients versus 22.0 percent on placebo (HR 0.75; P less than 0.001). Cardiovascular death alone was reduced (HR 0.80). The number-needed-to-treat was 21 to prevent one primary endpoint event over 5 years, an unusually large effect for a supplement-class drug.

FDA approved icosapent ethyl for this CV risk reduction indication in December 2019, the first marine n-3 product with a cardiovascular outcomes label .

STRENGTH: same dose, similar population, neutral result

STRENGTH used a different formulation: omega-3 carboxylic acid (Epanova), an EPA plus DHA free-fatty-acid mixture, at 4 grams per day. The placebo was corn oil rather than mineral oil. The 13,078 enrolled patients overlapped substantially with the REDUCE-IT eligibility criteria (statin-treated, elevated triglycerides, high CV risk).

After a median 41 months, the trial was stopped early for futility. The primary 5-point composite hazard ratio was 0.99 (95 percent CI 0.90 to 1.09; P 0.84). Atrial fibrillation was nominally more frequent on omega-3 (HR 1.69) .

Why the divergence? The leading hypotheses are (a) purified EPA versus EPA plus DHA acts differently on inflammation, plaque, and arrhythmia; (b) the mineral-oil placebo in REDUCE-IT modestly increased LDL and hsCRP versus baseline, slightly inflating the apparent treatment effect; (c) the corn-oil placebo in STRENGTH may have been a benign comparator. Independent reanalyses have argued that the placebo effect alone cannot explain the full magnitude of REDUCE-IT, but the controversy has not been definitively resolved.

VITAL: low-dose, primary prevention, neutral

VITAL was a 2-by-2 factorial trial of vitamin D3 (2000 IU per day) and marine omega-3 fatty acids (1 gram per day, providing 460 mg EPA and 380 mg DHA) for primary prevention of cardiovascular disease and cancer in 25,871 generally healthy adults (men 50 or older, women 55 or older).

Major cardiovascular events were not reduced by omega-3 supplementation (HR 0.92; P 0.24). Total cancer incidence was also not reduced. A pre-specified secondary analysis suggested possible benefit on myocardial infarction specifically, but the primary composite was negative .

The relevant difference from REDUCE-IT is the dose (1 gram per day in VITAL versus 4 grams per day in REDUCE-IT), the formulation (mixed EPA plus DHA in capsule form), and the population (generally healthy primary-prevention adults versus statin-treated high-risk secondary-prevention adults). Together with STRENGTH, VITAL bounds what omega-3 supplementation can claim outside the REDUCE-IT phenotype.

What current clinical guidance says

• Icosapent ethyl 4 grams per day is recommended in the 2018 AHA Triglyceride Reduction guidance and 2019 ESC/EAS lipid guidance as adjunctive therapy in statin-treated patients with persistent triglycerides 135 to 499 mg per dL who meet REDUCE-IT eligibility criteria .
• Routine over-the-counter fish-oil supplementation at 1 gram per day or less for primary cardiovascular prevention in healthy adults is not supported by RCT evidence (VITAL).
• The Cochrane 2020 omega-3 review concluded that increasing long-chain n-3 has little or no effect on all-cause mortality and a small effect on cardiovascular mortality (high-certainty evidence) .
• The atrial fibrillation signal seen in STRENGTH and in some REDUCE-IT subgroups is real and dose-dependent. Patients with paroxysmal atrial fibrillation should weigh the AF risk against the triglyceride or CV-event benefit.

What the evidence does not yet resolve

• Whether purified EPA at REDUCE-IT doses produces the same effect in patients without elevated triglycerides or without statin therapy.
• Whether the mineral-oil placebo in REDUCE-IT introduced enough effect to invalidate part of the apparent benefit. Independent reanalyses suggest no, but the question is not closed.
• Whether DHA-enriched formulations (or DHA alone) have a different cardiovascular signal than EPA alone.
• Whether the AF signal scales linearly with dose or is specific to certain formulations.
• Whether marine omega-3 supplementation has a meaningful primary-prevention role for adults without elevated triglycerides outside the REDUCE-IT phenotype.
• How n-3 supplementation interacts with established triglyceride-lowering therapies (statin, fibrate, niacin) and with newer agents like pemafibrate.

Editorial summary

Omega-3 is not one drug. The REDUCE-IT result is striking and prescription-grade icosapent ethyl now has an FDA cardiovascular-risk-reduction indication for a defined patient phenotype. STRENGTH and VITAL together limit what generic over-the-counter fish oil can claim. The class as a whole is illustrative of how dose, formulation, placebo choice, and patient selection can drive outcomes that look contradictory on the surface but are actually testing different propositions.