Did icosapent ethyl reduce cardiovascular death specifically?

Yes. Cardiovascular death was reduced by 20% (HR 0.80, 95% CI 0.66-0.98, p = 0.03). Total mortality trended lower at 13% (HR 0.87) but did not reach statistical significance, as the trial was not powered for all-cause mortality.

Did the benefit depend on how high triglycerides were at baseline?

No. Patients with baseline triglycerides above and below 200 mg/dL had similar relative risk reductions (HR 0.73 and 0.76, respectively; interaction p = 0.54). This disconnect between triglyceride level and clinical benefit is part of the evidence for non-lipid mechanisms.

How soon did the Kaplan-Meier curves separate?

Curves began visibly diverging at about 12 months and continued to separate through the full 4.9-year follow-up. By year 2, the difference was statistically significant. This progressive pattern resembles statin trial kinetics more than fibrate trials.

What was the atrial fibrillation risk with icosapent ethyl?

Hospitalization for atrial fibrillation or flutter occurred in 3.1% of the icosapent ethyl group versus 2.1% on placebo (p = 0.003). This is listed as a warning in the prescribing information and should factor into shared decision-making, particularly in patients with existing atrial substrate.

Why is the mineral oil placebo controversial?

Mineral oil raised LDL-C by about 10 mg/dL and modestly increased hsCRP and apolipoprotein B in the placebo arm. Critics argue these changes inflated the apparent treatment benefit. FDA sensitivity analyses adjusting for this effect still showed a significant MACE reduction, and the agency approved the cardiovascular indication.

How does REDUCE-IT compare to the STRENGTH trial?

STRENGTH tested a mixed EPA/DHA omega-3 carboxylic acid formulation versus corn oil placebo and found no cardiovascular benefit. Key differences: STRENGTH used a different drug (lower EPA dose, added DHA), a different comparator, and enrolled a somewhat different population. Whether the divergence reflects EPA purity, placebo effects, or trial design differences remains debated.

Do over-the-counter fish oil supplements provide the same benefit?

No. Trials of low-dose omega-3 supplements (1 g/day EPA+DHA) such as VITAL and ASCEND showed no MACE reduction. The REDUCE-IT benefit is specific to prescription icosapent ethyl at 4 g/day, which delivers approximately 3.6 g of pure EPA daily, a dose 4-8 times higher than typical supplements.

Which guidelines recommend icosapent ethyl based on REDUCE-IT?

Both the 2019 AHA/ACC focused update and the 2019 ESC/EAS dyslipidemia guidelines issued Class IIa recommendations for icosapent ethyl in statin-treated patients with triglycerides 135-499 mg/dL and established ASCVD or high-risk diabetes.

REDUCE-IT Results in Detail: Numbers, Subgroups, and Time Course

Q: What was the primary endpoint of REDUCE-IT?

A 5-point composite of cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, coronary revascularization, or hospitalization for unstable angina. Icosapent ethyl reduced this composite by 25% (HR 0.75, 95% CI 0.68-0.83, p < 0.001) over a median 4.9 years of follow-up.

Q: What was the number needed to treat (NNT) in REDUCE-IT?

The NNT was 21 for the primary 5-point MACE composite and 28 for the harder 3-point MACE composite (cardiovascular death, nonfatal MI, nonfatal stroke) over 4.9 years. These are competitive with statin NNTs in secondary prevention populations.

By HealthRX.com Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

Clinical image for REDUCE-IT Results in Detail: Numbers, Subgroups, and Time Course Image: HealthRX.com clinical image

At a glance

| Parameter | Detail | |---|---| | Trial name | REDUCE-IT (Reduction of Cardiovascular Events with Icosapent Ethyl-Intervention Trial) | | N | 8,179 | | Intervention | Icosapent ethyl (Vascepa) 2 g twice daily | | Comparator | Mineral oil placebo | | Median follow-up | 4.9 years | | Primary endpoint | 5-point MACE: cardiovascular death, nonfatal MI, nonfatal stroke, coronary revascularization, unstable angina requiring hospitalization | | Key result | HR 0.75 (95% CI 0.68-0.83), p < 0.001; absolute risk reduction 4.8 percentage points |

Primary Endpoint: The Five-Point MACE Composite

The primary analysis used a time-to-first-event approach for the 5-point composite of cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, coronary revascularization, or hospitalization for unstable angina. A primary event occurred in 17.2% of patients assigned icosapent ethyl versus 22.0% on placebo (HR 0.75, 95% CI 0.68-0.83, p < 0.001).

That translates to an absolute risk reduction (ARR) of 4.8 percentage points and a number needed to treat (NNT) of 21 over 4.9 years. For context, the original statin mega-trials reported NNTs in the range of 20-30 for secondary prevention. A treatment layered on top of optimized statin therapy that matches those numbers is unusual.

Each individual component of the primary composite moved in the same direction:

| Component | Icosapent ethyl (%) | Placebo (%) | HR (95% CI) | |---|---|---|---| | Cardiovascular death | 4.3 | 5.2 | 0.80 (0.66-0.98) | | Nonfatal MI | 6.1 | 8.7 | 0.69 (0.58-0.81) | | Nonfatal stroke | 1.9 | 2.6 | 0.72 (0.55-0.93) | | Coronary revascularization | 7.0 | 9.2 | 0.76 (0.65-0.88) | | Unstable angina hospitalization | 2.0 | 2.3 | 0.85 (0.66-1.10) |

Only unstable angina requiring hospitalization failed to reach individual statistical significance. Every other component did. Nonfatal MI showed the largest relative reduction at 31%.

Key Secondary Endpoint: The Three-Point MACE

The prespecified key secondary endpoint was a harder 3-point composite: cardiovascular death, nonfatal MI, or nonfatal stroke. This occurred in 11.2% of the icosapent ethyl group versus 14.8% on placebo (HR 0.74, 95% CI 0.65-0.83, p < 0.001). The ARR was 3.6 percentage points, yielding an NNT of 28.

HRX Hierarchical Endpoint Cascade

A useful way to read REDUCE-IT is through what we call an "endpoint cascade," stepping from the broadest composite down to individual hard endpoints. The pattern tells you whether the composite result is carried by a single soft component or distributed across hard outcomes.

| Endpoint level | HR | 95% CI | p-value | Signal | |---|---|---|---|---| | 5-point MACE (primary) | 0.75 | 0.68-0.83 | <0.001 | Broad benefit | | 3-point MACE (key secondary) | 0.74 | 0.65-0.83 | <0.001 | Confirmed on hard endpoints | | CV death alone | 0.80 | 0.66-0.98 | 0.03 | Significant | | Total mortality | 0.87 | 0.74-1.02 | 0.09 | Trend, not significant | | Fatal or nonfatal MI | 0.69 | 0.58-0.81 | <0.001 | Strongest single component | | Fatal or nonfatal stroke | 0.72 | 0.55-0.93 | 0.01 | Significant |

The cascade shows benefit was not driven solely by the softer components (revascularization, unstable angina). Hard atherosclerotic endpoints, particularly MI, carried substantial weight. Total mortality trended favorably but did not cross the significance threshold, which is common in trials not powered for all-cause death.

Time-Course Analysis: When Did the Curves Separate?

Kaplan-Meier curves for the primary endpoint began separating by approximately 12 months and continued to diverge through the end of follow-up. By year 2, the event-rate difference was already statistically significant. By year 5, cumulative event rates were 17.2% versus 22.0%.

This progressive divergence pattern differs from what you see with triglyceride-lowering fibrate trials, where separation was modest or absent. It more closely resembles statin trial kinetics, consistent with an antiatherosclerotic mechanism rather than a purely metabolic one.

The FDA's 2019 review of icosapent ethyl for the cardiovascular risk-reduction indication noted this time-course data as supportive of a durable treatment effect, not a transient metabolic signal.

Prespecified Subgroup Analyses

REDUCE-IT enrolled two strata: a secondary prevention cohort (70.7% of patients, with established atherosclerotic cardiovascular disease) and a primary prevention cohort (29.3%, with diabetes plus at least one additional risk factor). Both strata showed consistent benefit:

| Subgroup | n | HR (95% CI) | Interaction p | |---|---|---|---| | Secondary prevention | 5,785 | 0.73 (0.65-0.81) |, | | Primary prevention (diabetes + risk factors) | 2,394 | 0.78 (0.64-0.95) | 0.60 | | Baseline TG ≥ 200 mg/dL | 3,754 | 0.73 (0.64-0.84) |, | | Baseline TG < 200 mg/dL (150-199) | 4,425 | 0.76 (0.66-0.87) | 0.54 | | Diabetes at baseline | 4,787 | 0.77 (0.68-0.87) |, | | No diabetes at baseline | 3,392 | 0.73 (0.63-0.84) | 0.57 | | Baseline LDL ≤ 75 mg/dL | ~2,400 | 0.73 (0.62-0.87) |, | | Baseline LDL > 100 mg/dL | ~2,500 | 0.73 (0.62-0.85) | 0.91 | | eGFR < 60 mL/min | ~1,300 | 0.69 (0.55-0.86) |, | | Age ≥ 65 years | ~3,900 | 0.74 (0.65-0.84) | 0.78 |

No interaction p-value reached significance, meaning the treatment effect was consistent regardless of diabetes status, triglyceride level, LDL cholesterol, renal function, or age. This consistency across subgroups was a critical factor in the 2019 AHA/ACC guideline update that gave icosapent ethyl a Class IIa recommendation for patients with triglycerides 135-499 mg/dL on maximally tolerated statins.

Triglyceride Reduction Versus Clinical Benefit: A Disconnect Worth Noting

Median triglyceride levels dropped from 216 mg/dL at baseline to 175 mg/dL at 1 year in the placebo group (after a mineral oil-related increase from baseline). In the icosapent ethyl group, median levels fell from 216 mg/dL to 89 mg/dL, an 18.3% reduction versus placebo (p < 0.001).

However, the magnitude of cardiovascular benefit exceeded what triglyceride lowering alone would predict based on prior trials like ACCORD-Lipid or FIELD. In a prespecified mediation analysis published in 2019, triglyceride reduction accounted for only a fraction of the observed MACE reduction. This finding supports pleiotropic mechanisms: anti-inflammatory effects (reflected in reduced hsCRP), membrane stabilization, antiplatelet activity, and improved endothelial function.

Baseline hsCRP was 2.2 mg/L (median) in both groups. By 2 years, hsCRP had fallen by approximately 13% more in the icosapent ethyl arm, suggesting a meaningful anti-inflammatory component, though smaller than the 37% CRP reduction seen with canakinumab in CANTOS.

The Mineral Oil Placebo Controversy

The most debated limitation of REDUCE-IT centers on the mineral oil placebo. Post-hoc analyses showed that mineral oil modestly raised LDL-C (by approximately 10.2 mg/dL), hsCRP, and apolipoprotein B in the placebo arm. Critics, including an FDA advisory committee minority, argued this could have inflated the between-group difference.

The STRENGTH trial (omega-3 carboxylic acids vs. corn oil placebo) found no cardiovascular benefit, and its proponents pointed to their biologically inert comparator as a reason for the divergent results. The counterargument: STRENGTH used a mixed EPA/DHA formulation at a lower EPA dose, different pharmacology, and a different patient population.

An FDA-commissioned sensitivity analysis adjusted the REDUCE-IT results for the placebo arm's LDL increase and still found a statistically significant MACE reduction (adjusted HR approximately 0.80-0.85 depending on the modeling assumptions). The agency ultimately approved the cardiovascular indication.

Total and Ischemic Events: The Burden Analysis

A prespecified total-events analysis (first plus recurrent ischemic events) showed icosapent ethyl reduced total primary endpoint events by 30% (rate ratio 0.70, 95% CI 0.62-0.78, p < 0.001). In absolute terms, there were 1,606 total primary endpoint events in the placebo group versus 1 to 117 in the icosapent ethyl group, a difference of 489 events among 8,179 patients.

This total-events framing matters clinically. Time-to-first-event analyses undercount the real-world burden of recurrent ischemic disease. A patient who has one MI is at high risk for a second. The 30% total-event reduction was numerically larger than the 25% first-event reduction, suggesting icosapent ethyl continued to protect after an initial event.

Safety Signal: Atrial Fibrillation

The primary publication reported a higher rate of atrial fibrillation or flutter requiring hospitalization in the icosapent ethyl group: 3.1% versus 2.1% (p = 0.003). Serious bleeding events were also numerically higher (2.7% vs. 2.1%, p = 0.06), though adjudicated fatal bleeding was rare in both arms.

The AF signal was confirmed in the Vascepa prescribing information and is now listed as a warning. For clinicians, this means monitoring for AF symptoms in patients starting icosapent ethyl, particularly those with pre-existing atrial substrate (left atrial enlargement, prior paroxysmal AF).

What the Guidelines Said After REDUCE-IT

The 2019 ESC/EAS dyslipidemia guidelines gave icosapent ethyl a Class IIa, Level B recommendation for high-risk patients with triglycerides between 135 and 499 mg/dL despite statin therapy. The 2019 AHA/ACC focused update on lipid management echoed this position, specifically calling out the REDUCE-IT evidence as the basis for recommending icosapent ethyl (not omega-3 supplements broadly) for persistent hypertriglyceridemia in ASCVD or high-risk diabetic patients.

This distinction is clinically important. Over-the-counter fish oil supplements, which contain variable EPA and DHA ratios at much lower doses, have not demonstrated MACE reduction. The VITAL and ASCEND trials of low-dose omega-3 (1 g/day) showed no cardiovascular benefit. The effect is specific to high-dose, purified EPA.

Frequently asked questions

›

References

Bhatt DL, Steg PG, Miller M, et al. Cardiovascular risk reduction with icosapent ethyl for hypertriglyceridemia. N Engl J Med. 2019;380(1):11-22. PubMed
Bhatt DL, Steg PG, Miller M, et al. Effects of icosapent ethyl on total ischemic events: from REDUCE-IT. J Am Coll Cardiol. 2019;73(22):2791-2802. PubMed
Arnett DK, Blumenthal RS, Baber B, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease. Circulation. 2019;140(11):e596-e646. PubMed
Vascepa (icosapent ethyl) prescribing information. Amarin Pharma. Revised December 2019. FDA Label
Bhatt DL, Miller M, Brinton EA, et al. REDUCE-IT USA: results from the 3,146 patients randomized in the United States. Circulation. 2020;141(5):367-375. PubMed
Nicholls SJ, Lincoff AM, Garcia M, et al. Effect of high-dose omega-3 fatty acids vs corn oil on major adverse cardiovascular events (STRENGTH). JAMA. 2020;324(22):2268-2280. PubMed