EPA vs DHA — which one should I take?

Depends on the goal. EPA dominates the anti-inflammatory, mood, triglyceride-lowering, and REDUCE-IT cardiovascular evidence. DHA dominates structural brain and retinal membrane composition, pregnancy and infant brain development, and dry-eye support. For depression specifically, formulations with ≥60% EPA outperform DHA-dominant blends in pooled meta-analyses. For pregnancy and infant brain, DHA is the priority. Most adults benefit from a formula containing both, with the ratio tilted toward EPA unless pregnancy/lactation is the indication.

REDUCE-IT showed 25% event reduction — but STRENGTH was null. What gives?

Two clean trials, two opposite results. REDUCE-IT used pure EPA (icosapent ethyl, 4 g/day) with a mineral-oil placebo; STRENGTH used combined EPA+DHA at the same dose with a corn-oil placebo. The two leading explanations: (1) pure EPA is uniquely cardioprotective and combined formulations don't deliver, or (2) REDUCE-IT's mineral-oil placebo modestly raised LDL/CRP in the placebo arm and inflated the relative effect. Both are plausible; neither is settled. The honest read: if you're going for the REDUCE-IT-style benefit, you want pure EPA, not generic fish oil — but treat the magnitude with appropriate humility.

How much EPA should I take for depression?

Pooled meta-analyses (Mocking 2016, Sublette 2011) support 1–2 g/day EPA from a formulation with ≥60% EPA — i.e., EPA:DHA ratio ≥2:1. The effect is most reliable as an adjunct to antidepressant therapy and in patients with elevated inflammatory markers (CRP, IL-6). Generic 1.5:1 EPA:DHA fish oil is below the empirical EPA-fraction threshold; if mood is the goal, read the EPA mg per serving on the label.

Is fish oil good for the heart at 1 g/day?

Largely no, based on the largest trials. VITAL (n=25,871, healthy adults), ASCEND (n=15,480, diabetes), and OMEMI (n=1,027, elderly post-MI) all came back null at 1–1.8 g/day combined EPA+DHA. The cardiovascular benefit was demonstrated at the pharmaceutical 4 g/day pure-EPA dose in already-statin-treated adults with elevated triglycerides (REDUCE-IT). One-a-day grocery-store fish oil for 'heart health' in healthy adults is not supported by randomized trial evidence.

I'm vegan — can I get EPA from plant sources?

Not efficiently. ALA from flax, chia, and walnut converts to EPA at ~5% efficiency in humans; conversion to DHA is <1%. Two tablespoons of flax delivers maybe 50–100 mg of converted EPA — well below trial-level doses. The practical vegan solution is algal oil. Most algal products are DHA-dominant, but EPA-containing algal formulations are increasingly available; read the EPA mg per serving on the label.

What's the Omega-3 Index target?

Above 8% RBC EPA + DHA is associated with the lowest cardiovascular risk in observational data; below 4% is high-risk. Most U.S. adults sit around 4–6%. Reaching the 8% target generally requires 1–2 g/day of combined EPA+DHA from a triglyceride-form supplement, or several servings of fatty fish per week. Commercial home tests (e.g., OmegaQuant) are inexpensive and the most direct way to confirm whether your supplementation is doing what you want.

Should I worry about atrial fibrillation with high-dose EPA?

It's a real, dose-dependent signal. REDUCE-IT showed 5.3% vs 3.9% afib on 4 g/day icosapent ethyl; STRENGTH and OMEMI replicated with combined EPA+DHA. Meta-analyses now consistently identify a dose-dependent afib risk at ≥1 g/day omega-3. The absolute risk increase is small but meaningful. If you have a personal or family history of atrial fibrillation or structural heart disease, discuss with your cardiologist before chronic high-dose EPA — the risk/benefit calculation shifts.

Best EPA Supplements

Q: What's the Omega-3 Index target?

Above 8% RBC EPA + DHA is associated with the lowest cardiovascular risk in observational data; below 4% is high-risk. Most U.S. adults sit around 4–6%. Reaching the 8% target generally requires 1–2 g/day of combined EPA+DHA from a triglyceride-form supplement, or several servings of fatty fish per week. Commercial home tests (e.g., OmegaQuant) are inexpensive and the most direct way to confirm whether your supplementation is doing what you want.

Q: Should I worry about atrial fibrillation with high-dose EPA?

It's a real, dose-dependent signal. REDUCE-IT showed 5.3% vs 3.9% afib on 4 g/day icosapent ethyl; STRENGTH and OMEMI replicated with combined EPA+DHA. Meta-analyses now consistently identify a dose-dependent afib risk at ≥1 g/day omega-3. The absolute risk increase is small but meaningful. If you have a personal or family history of atrial fibrillation or structural heart disease, discuss with your cardiologist before chronic high-dose EPA — the risk/benefit calculation shifts.

Research dossier

Clinical research on EPA

Last reviewed2026-05-22

10 trials reviewed across 6 indications.

Strongest evidence

Triglyceride lowering

Very Strong

Mechanism

EPA suppresses hepatic SREBP-1c and reduces VLDL-triglyceride synthesis and secretion. The effect is dose-dependent and reproducible — one of the most consistent biochemical findings in the entire omega-3 literature.

Each 1 g/day of combined EPA+DHA lowers fasting triglycerides by ~5–10%, with larger absolute drops at higher baseline values. At the 4 g/day pharmaceutical threshold, 20–30% reductions are routine; ≥30% in severe hypertriglyceridemia. Pure-EPA formulations achieve this without raising LDL — a key advantage over combined EPA+DHA at very high baseline triglycerides.

Clinically meaningful at ≥2 g/day, robust at 4 g/day. Standard 1 g/day capsules deliver only token triglyceride reduction. The 4 g/day threshold is a real clinical line, not a marketing one.

Trials cited

AHA Science Advisory — omega-3 for hypertriglyceridemia
positive · Systematic review
Skulas-Ray et al., 2019, Circulation
American Heart Association advisory: at the pharmaceutical 4 g/day dose, prescription omega-3 reliably drops triglycerides 20–30%, with ≥30% reductions in severe hypertriglyceridemia (≥500 mg/dL). Pure-EPA formulations lower triglycerides without raising LDL; combined EPA+DHA at very high triglycerides can modestly raise LDL.
The clinical effect requires the high (~4 g/day) dose. Standard 1 g/day capsules barely move triglycerides.
PubMed DOI
Harris — fish oil triglyceride dose-response
positive · Meta-analysis
Harris, 1997, American Journal of Clinical Nutrition (subsequently extended)
Classic dose-response analysis: each 1 g/day of combined EPA + DHA lowers fasting triglycerides by roughly 5–10%, with proportionally larger absolute drops in patients with higher baseline triglycerides. The triglyceride-lowering effect is one of the most reproducible findings in the entire omega-3 literature — and the basis for the 4 g/day pharmaceutical threshold.
Pre-statin-era trials over-represented. Effect size in modern statin-treated populations is similar in relative terms but smaller in absolute terms because baseline triglycerides are lower.
PubMed

Cardiovascular events at pharmaceutical dose

Mechanism

EPA enriches membrane phospholipids, displacing arachidonic acid and shifting eicosanoid production from inflammatory PGE2/LTB4 toward less-inflammatory PGE3 and pro-resolving E-series resolvins. EPA also lowers hepatic VLDL output, modestly inhibits platelet aggregation, and may stabilize atherosclerotic plaque membrane composition.

REDUCE-IT (4 g/day pure EPA, n=8,179) cut major cardiovascular events 25% in statin-treated high-risk adults with elevated triglycerides. But STRENGTH (same dose, combined EPA+DHA, n=13,078) was null and stopped early for futility. VITAL (1 g/day primary prevention) and OMEMI (1.8 g/day post-MI elderly) were also null. The defensible read: pharmaceutical-grade pure EPA at 4 g/day in high-CV-risk, statin-treated, elevated-triglyceride patients. Not a heart-health claim for healthy adults at 1 g/day.

Real benefit in high-CV-risk statin-treated adults with elevated triglycerides at the 4 g/day icosapent ethyl dose. Not a primary-prevention claim for healthy adults, and not supported for combined EPA+DHA formulations at any dose. The mineral-oil-placebo controversy keeps even the REDUCE-IT effect size genuinely debated.

REDUCE-IT — icosapent ethyl in high-risk adults
positive · RCT
Bhatt et al., 2019, New England Journal of Medicinen=8179Industry-funded
8,179 high-cardiovascular-risk statin-treated adults with elevated triglycerides randomized to 4 g/day pure EPA (icosapent ethyl) vs mineral-oil placebo. 25% relative reduction in major cardiovascular events; 20% reduction in cardiovascular death. The single strongest cardiovascular omega-3 result in the modern era — at a pharmaceutical 4 g/day dose, not a grocery-store capsule.
Industry-funded (Amarin). The mineral-oil placebo has been criticized for potentially raising LDL and CRP, which some argue inflates the relative effect. Atrial fibrillation rose to 5.3% vs 3.9% on placebo.
PubMed DOI
STRENGTH — combined EPA+DHA carboxylic acid (NEGATIVE)
negative · RCT
Nicholls et al., 2020, JAMAn=13078Industry-funded
13,078 patients in nearly the same population as REDUCE-IT, given combined EPA+DHA at the same 4 g/day dose. Stopped early for futility — no cardiovascular benefit vs corn-oil placebo. Atrial fibrillation rose ~70% on omega-3 (2.2% vs 1.3%). The most direct test of 'is the REDUCE-IT effect EPA-specific?' — and it argues yes.
Industry-funded (AstraZeneca). Corn-oil placebo is a more biologically inert comparator than mineral oil — making the divergent results from REDUCE-IT harder to dismiss as a placebo artifact.
PubMed DOI
VITAL — 1 g/day omega-3 for primary cardiovascular prevention
Null · RCT
Manson et al., 2019, New England Journal of Medicinen=25871
Largest primary-prevention omega-3 trial ever run. 1 g/day combined EPA+DHA did not reduce major cardiovascular events overall. A pre-specified MI signal appeared (28% reduction) and effects were larger in participants with low fish intake at baseline, but the headline primary composite was null. Closes the door on routine 1-g-a-day fish oil for primary CV prevention in already-well-fed adults.
Trial population had reasonable baseline fish intake (~1.5 servings/week). Magnitude of benefit in low-fish-intake populations remains an open question.
PubMed DOI
OMEMI — omega-3 in elderly post-MI patients
Null · RCT
Kalstad et al., 2021, Circulationn=1027
Elderly post-MI patients given 1.8 g/day EPA+DHA on top of guideline-directed therapy for 2 years. No reduction in the primary composite endpoint vs placebo. Atrial fibrillation, again, was numerically higher on omega-3 (7.2% vs 4.0%). Even in a high-risk secondary-prevention population, sub-pharmaceutical-dose combined omega-3 failed to deliver.
Smaller and older population than REDUCE-IT; combined EPA+DHA rather than pure EPA. Both factors plausibly contribute to the null result.
PubMed DOI
Curfman editorial — reconciling REDUCE-IT and STRENGTH
mixed · Systematic review
Curfman, 2020, JAMA
Editorial accompanying STRENGTH publication that crystallized the field's interpretive debate: was REDUCE-IT's 25% benefit a true EPA effect, or partly inflated by a pro-inflammatory mineral-oil placebo that raised LDL ~10% and CRP ~30% in the placebo arm? STRENGTH (corn-oil placebo, combined EPA+DHA) found no benefit. The honest read: either pure-EPA is uniquely cardioprotective, or REDUCE-IT's effect size was overstated by placebo confounding — and current data cannot fully resolve which.
Commentary, not a trial. The mineral-oil placebo controversy is unresolved; pre-specified subgroup analyses in REDUCE-IT showed benefit even after adjustment for LDL/CRP changes, but the question remains open.
PubMed DOI

Mood and major depression

Mechanism

EPA modulates neuroinflammation and prostaglandin signaling; elevated peripheral inflammatory markers (IL-6, CRP, TNF-α) predict EPA responsiveness in depression. The proposed pathway is anti-inflammatory rather than structural — which is why EPA outperforms DHA for mood despite DHA dominating neuronal membrane composition.

Pooled meta-analyses (Mocking 2016, Sublette 2011) show EPA-dominant formulas (≥60% EPA, EPA:DHA ≥2:1) at 1–2 g/day EPA produce small-to-moderate reductions in depressive symptoms (SMD ≈ -0.4 to -0.5). Pure DHA and DHA-dominant blends show no effect. The effect is most reliable as an adjunct to antidepressant therapy in patients with elevated inflammatory markers.

Real adjunctive effect in major depression with EPA-dominant formulas; do not expect monotherapy efficacy. Generic 1.5:1 EPA:DHA fish oil does not consistently hit the threshold — read the label.

Mocking meta-analysis — EPA-dominant omega-3 for major depression
positive · Meta-analysis
Mocking et al., 2016, Translational Psychiatryn=1233
Pooled across 13 RCTs in adults with MDD, omega-3 supplementation produced a small-to-moderate reduction in depression scores (SMD ≈ -0.4). The effect was driven by EPA-dominant formulas; DHA-pure and DHA-dominant blends showed no signal. Effect strongest as an add-on to antidepressant therapy rather than monotherapy.
Heterogeneous trials with varying baseline severity, antidepressant co-treatment, and EPA:DHA ratios. The pooled effect is robust to leave-one-out but individual trial quality varies.
PubMed DOI
Sublette meta-analysis — the EPA fraction threshold
positive · Meta-analysis
Sublette et al., 2011, Journal of Clinical Psychiatryn=916
Established the EPA-fraction threshold for depression: trials using formulations with ≥60% EPA showed clear benefit (SMD ≈ 0.53 favoring EPA); trials with <60% EPA showed no effect or favored placebo. The clearest single demonstration that 'omega-3 for mood' is actually 'EPA-dominant omega-3 for mood'.
Trials varied in patient severity and antidepressant background. The dichotomization at 60% EPA is data-driven but somewhat arbitrary; the underlying signal is continuous.
PubMed DOI

Anti-inflammatory eicosanoid shift

Mechanism

EPA competes with arachidonic acid for COX and LOX enzymes, producing series-3 prostaglandins (PGE3) and series-5 leukotrienes (LTB5) — substantially less inflammatory than the arachidonic-acid-derived PGE2 and LTB4. EPA is also the precursor to E-series resolvins, lipid mediators that actively resolve inflammation.

Mechanistically clean: EPA reshapes the eicosanoid landscape toward resolution rather than amplification of inflammation. The clinical translation is modest — in rheumatoid arthritis, ≥2.7 g/day for ≥3 months produces measurable but moderate symptom relief and NSAID-sparing effects. Not a substitute for disease-modifying therapy.

Adjunct in inflammatory conditions, not first-line. Effect size is real but smaller than mainstream anti-inflammatory pharmacotherapy.

Goldberg & Katz — omega-3 for inflammatory joint pain
positive · Meta-analysis
Goldberg & Katz, 2007, Painn=368
Pooled 17 RCTs of marine omega-3 in rheumatoid arthritis and related inflammatory joint conditions. Doses ≥2.7 g/day for ≥3 months produced modest but consistent improvements in patient-reported joint pain, morning stiffness duration, and NSAID requirement. Effect plausibly driven by EPA's anti-inflammatory eicosanoid shift.
Heterogeneous trials; most used combined EPA+DHA rather than pure EPA, so the relative contribution of each fraction is not cleanly separated. Modest effect — not a replacement for DMARD therapy.
PubMed DOI

Rheumatoid arthritis and joint inflammation

Mechanism

Same anti-inflammatory eicosanoid pathway: EPA-derived mediators (PGE3, resolvin E series) dampen the synovial inflammation that drives joint pain and morning stiffness in RA.

Pooled across 17 RCTs, ≥2.7 g/day combined EPA+DHA for ≥3 months reduced patient-reported joint pain, morning stiffness duration, and NSAID consumption in rheumatoid arthritis. Effect is modest but consistent — useful as an NSAID-sparing adjunct to standard RA care.

Reasonable adjunct in RA and other inflammatory arthritides. Do not expect transformative effects, and do not displace DMARD therapy.

Goldberg & Katz — omega-3 for inflammatory joint pain
positive · Meta-analysis
Goldberg & Katz, 2007, Painn=368
Pooled 17 RCTs of marine omega-3 in rheumatoid arthritis and related inflammatory joint conditions. Doses ≥2.7 g/day for ≥3 months produced modest but consistent improvements in patient-reported joint pain, morning stiffness duration, and NSAID requirement. Effect plausibly driven by EPA's anti-inflammatory eicosanoid shift.
Heterogeneous trials; most used combined EPA+DHA rather than pure EPA, so the relative contribution of each fraction is not cleanly separated. Modest effect — not a replacement for DMARD therapy.
PubMed DOI

Insulin sensitivity (mechanistic, weak clinical)

Mechanism

EPA reduces hepatic lipogenesis and may improve insulin signaling via PPAR-α activation and reduced ectopic fat deposition in liver and muscle. The mechanism is clean; the clinical-endpoint evidence in non-diabetic adults is weak.

Mechanistic plausibility for an insulin-sensitizing effect, particularly via reduced hepatic and intramyocellular lipid. Large RCTs (VITAL, ASCEND) did not show diabetes-prevention benefit at 1 g/day in non-diabetic adults. Treat metabolic benefit as a possible secondary effect at pharmaceutical doses in already-dysmetabolic patients, not a primary indication.

Mechanism is real but trial-level evidence for metabolic outcomes at supplement doses is weak. Do not supplement EPA primarily for blood sugar.

VITAL — 1 g/day omega-3 for primary cardiovascular prevention
Null · RCT
Manson et al., 2019, New England Journal of Medicinen=25871
Largest primary-prevention omega-3 trial ever run. 1 g/day combined EPA+DHA did not reduce major cardiovascular events overall. A pre-specified MI signal appeared (28% reduction) and effects were larger in participants with low fish intake at baseline, but the headline primary composite was null. Closes the door on routine 1-g-a-day fish oil for primary CV prevention in already-well-fed adults.
Trial population had reasonable baseline fish intake (~1.5 servings/week). Magnitude of benefit in low-fish-intake populations remains an open question.
PubMed DOI

5 forms of EPA compared

Triglyceride form (rTG / natural TG)
Reference standard — ~50–70% better absorbed than ethyl ester in head-to-head studies (Dyerberg 2010)
Best forEvery documented EPA benefit — cardiovascular, mood, triglyceride lowering, anti-inflammatory
The form omega-3 occurs in naturally in fish. Re-esterified triglyceride (rTG) is the concentrated supplemental form: ethyl-ester intermediates are converted back to triglycerides to combine high EPA concentration with reference-standard absorption. Best form per mg, most expensive.
Re-esterified triglyceride (rTG)
Matches natural triglyceride form — reference-standard absorption
Best forHigh-dose EPA delivery (≥2 g/day) without sacrificing absorption
Concentrate ethyl esters first to raise EPA content per gram, then convert back to triglycerides. The combination of high concentration and good absorption is why most premium EPA-focused products use this form.
Triglyceride (natural or re-esterified)
Reference standard for omega-3 absorption
Best forAll EPA indications
If the label says 'triglyceride form' or 'as TG' or 'rTG', the product is in the best-absorbed form. Always preferred over ethyl ester when budget allows.
rTG (re-esterified triglyceride)
Reference-standard absorption — the form used in most premium EPA-focused products
Best forHigh-dose EPA protocols
Shorthand seen on supplement labels. Confirms the product was concentrated as ethyl ester then converted back to triglyceride — the gold standard for high-EPA supplementation.
Ethyl ester (EE)
Roughly 70% of triglyceride-form absorption when taken without a fatty meal; gap narrows with dietary fat
Best forMost retail capsules use this form because it's cheaper to concentrate. The form used in icosapent ethyl (Vascepa) in REDUCE-IT.
Dyerberg 2010 showed ~50% lower bioavailability versus rTG. Always take with a fat-containing meal to narrow the gap. Note the irony: REDUCE-IT used ethyl ester at pharmaceutical 4 g/day — so the form penalty is partly compensated by sheer dose.

Are you deficient? Symptoms, risk groups, lab tests

Average U.S. EPA intake from non-fish-eating diets is ~50 mg/day — roughly 10× below the lower bound of clinically studied doses. The Western dietary pattern is heavily omega-6 dominant (ratio ~15–20:1 omega-6:omega-3 vs an ancestral ~1–4:1), which biases eicosanoid production toward inflammatory PGE2/LTB4 even in adults with 'adequate' total fat intake.

Common symptoms

Higher fasting triglycerides on routine lipid panels
Low Omega-3 Index (RBC EPA + DHA <4%)
Elevated systemic inflammatory markers (CRP, IL-6) in some pools
Higher baseline depressive symptom scores in epidemiologic cohorts with low fish intake
Dry skin and brittle nails (non-specific; often cited but weakly evidenced as EPA-specific)

Who is at risk

Adults eating little or no fatty fish
Direct dietary EPA comes almost entirely from fatty fish (salmon, sardines, mackerel, anchovy). Lean fish and shellfish contribute little. No fatty fish = near-zero EPA intake.
Vegetarians and vegans
Zero direct EPA intake. ALA from flax/chia/walnut converts to EPA at ~5% efficiency and to DHA at <1% — far too low to meet trial-level needs. Algal oil products are increasingly available with EPA content but most algal supplements remain DHA-dominant.
Adults with elevated triglycerides
High fasting triglycerides correlate with low long-chain omega-3 status. The therapeutic dose for hypertriglyceridemia is ~4 g/day — far above what diet alone supplies.
Adults with major depression and elevated inflammatory markers
Higher peripheral CRP, IL-6, and TNF-α predict greater EPA responsiveness in MDD trials. Population with the strongest mechanism-to-outcome link for EPA-specific supplementation.
Adults with rheumatoid arthritis or chronic inflammatory disease
EPA's eicosanoid-shifting and resolvin-precursor role is most clinically relevant in chronic inflammation. Pooled trials suggest modest symptomatic benefit at ≥2.7 g/day combined EPA+DHA.

Lab markers

Omega-3 Index (RBC EPA + DHA as % of total fatty acids)
The most validated marker for cardiovascular risk stratification. More informative than serum EPA/DHA, which fluctuates with the last meal. Available from commercial labs (OmegaQuant) at consumer-friendly cost.
High cardiovascular risk
<4%
Intermediate
4–8%
Cardioprotective target
>8%
Fasting triglycerides
Indirect marker — elevated triglycerides identify the population most likely to benefit from high-dose EPA. Not a measure of EPA status per se, but the most useful routine lab for identifying candidates for the REDUCE-IT-style intervention.
Optimal
<150 mg/dL
Borderline-high (REDUCE-IT entry range)
150–499 mg/dL
Severe (≥500 mg/dL) — pharmaceutical omega-3 indicated
≥500 mg/dL

Side effects and drug interactions

Side effects

Fishy burps and reflux
Common
The most common complaint with marine omega-3. Worse with low-quality, oxidized, or ethyl-ester capsules. Enteric-coating, refrigeration, and dosing with a meal reduce it substantially.
Worse with:ethyl ester
Gentler:Triglyceride form (rTG), Enteric-coated capsules
GI upset and loose stools
Common · Most pronounced above 3 g/day
Higher doses (≥3 g/day) commonly cause loose stools, stomach discomfort, or mild nausea. Splitting the daily dose across meals usually resolves it.
Atrial fibrillation
Uncommon · Signal detectable at ≥1 g/day; clearest at ≥4 g/day
REDUCE-IT showed 5.3% vs 3.9% atrial fibrillation on 4 g/day icosapent ethyl. STRENGTH replicated the signal with combined EPA+DHA (~70% relative increase). Multiple meta-analyses now describe a dose-dependent afib risk with omega-3 at ≥1 g/day. Real, dose-dependent, and worth flagging — especially in patients with prior afib or structural heart disease.
Increased bleeding tendency
Uncommon · Most relevant at ≥3 g/day
EPA mildly inhibits platelet aggregation. Clinically meaningful when stacked with anticoagulants, antiplatelets, or NSAIDs, or in the lead-up to elective surgery. Stop 1–2 weeks before scheduled surgical procedures.
Modest blood pressure reduction
Uncommon · Most relevant at ≥3 g/day
High-dose omega-3 produces a small (~2–4 mmHg) blood-pressure drop. Generally a feature, not a side effect — but watch for over-correction in well-controlled hypertensives or symptomatic hypotension when stacked with antihypertensive medication.
Rancidity and oxidation byproducts
Uncommon
EPA oxidizes readily. Spoiled capsules carry oxidation products (peroxides, aldehydes) that plausibly negate benefit. Smell-test the capsule contents; refrigerate after opening; choose products with third-party oxidation testing (TOTOX values).

Drug interactions

Additive effect
warfarinapixabanrivaroxabandabigatranclopidogrelaspirin
EPA mildly inhibits platelet aggregation. Additive antiplatelet/anticoagulant effect at high doses.
Routine 1 g/day is generally tolerated on these medications, but discuss high-dose EPA (≥3 g/day) with your prescriber. Pause supplementation 1–2 weeks before elective surgery.
Additive effect
statinsfibratesezetimibe
Combined triglyceride-lowering effect. Clinically beneficial; REDUCE-IT was deliberately run on top of statin therapy.
Stacking is appropriate and intended in patients with elevated triglycerides on statin therapy. The combination is the population in which the REDUCE-IT benefit was demonstrated.
Other
antihypertensives (ACE inhibitors, ARBs, beta-blockers, calcium channel blockers)
High-dose EPA produces a small (~2–4 mmHg) blood-pressure reduction. Usually beneficial; rarely produces excessive drop.
Monitor BP when starting high-dose EPA on antihypertensive therapy. Dose adjustment is rarely needed but possible in well-controlled patients.
Combined-effect risk
other omega-3 sources (krill oil, algal oil, prescription omega-3)
Doses stack — running prescription icosapent ethyl alongside an additional fish oil capsule adds to total daily EPA exposure.
Account for all marine omega-3 sources when calculating daily EPA. The 4 g/day icosapent ethyl trial dose is already at the upper end of routine supplementation; layering more EPA on top is unstudied.

Other critical caveats

REDUCE-IT showed dose-dependent atrial fibrillation at 4 g/day icosapent ethyl (5.3% vs 3.9% on placebo). STRENGTH and OMEMI replicated similar signals with combined EPA+DHA. If you have a personal or family history of atrial fibrillation, discuss with your prescriber before chronic high-dose use.
STRENGTH (2020) was NEGATIVE for cardiovascular events with combined EPA+DHA at the same 4 g/day dose that worked in REDUCE-IT. Either pure-EPA is uniquely cardioprotective, or REDUCE-IT's effect was inflated by a pro-inflammatory mineral-oil placebo — current data cannot fully resolve which. Do not assume combined EPA+DHA fish oil will deliver REDUCE-IT-style benefits.
1 g/day omega-3 for primary cardiovascular prevention is not supported by large trials (VITAL n=25,871, ASCEND n=15,480, OMEMI n=1,027). The 'one-a-day fish oil for heart health' positioning is the dose-marketing gap of the supplement industry. Either go to ~4 g/day in an appropriate high-risk population or recognize the 1 g/day cap as a fish-intake supplement, not a cardioprotective intervention.
EPA, not DHA, drives the mood benefit. Pick formulations with ≥60% EPA and EPA:DHA ≥2:1 if depression is the target indication. Generic fish oil is typically ~1.5:1 EPA:DHA — below the empirical depression threshold.
Stop high-dose EPA (≥3 g/day) 1–2 weeks before elective surgery to avoid additive bleeding risk. Standard 1 g/day is generally fine to continue, but coordinate with your surgical team.
For pregnancy, infant brain development, retinal/vision support, and structural-membrane indications, DHA is the relevant omega-3, not EPA. See the DHA nutrient page for those use cases.

Frequently asked

EPA vs DHA — which one should I take?
Depends on the goal. EPA dominates the anti-inflammatory, mood, triglyceride-lowering, and REDUCE-IT cardiovascular evidence. DHA dominates structural brain and retinal membrane composition, pregnancy and infant brain development, and dry-eye support. For depression specifically, formulations with ≥60% EPA outperform DHA-dominant blends in pooled meta-analyses. For pregnancy and infant brain, DHA is the priority. Most adults benefit from a formula containing both, with the ratio tilted toward EPA unless pregnancy/lactation is the indication.
REDUCE-IT showed 25% event reduction — but STRENGTH was null. What gives?
Two clean trials, two opposite results. REDUCE-IT used pure EPA (icosapent ethyl, 4 g/day) with a mineral-oil placebo; STRENGTH used combined EPA+DHA at the same dose with a corn-oil placebo. The two leading explanations: (1) pure EPA is uniquely cardioprotective and combined formulations don't deliver, or (2) REDUCE-IT's mineral-oil placebo modestly raised LDL/CRP in the placebo arm and inflated the relative effect. Both are plausible; neither is settled. The honest read: if you're going for the REDUCE-IT-style benefit, you want pure EPA, not generic fish oil — but treat the magnitude with appropriate humility.
How much EPA should I take for depression?
Pooled meta-analyses (Mocking 2016, Sublette 2011) support 1–2 g/day EPA from a formulation with ≥60% EPA — i.e., EPA:DHA ratio ≥2:1. The effect is most reliable as an adjunct to antidepressant therapy and in patients with elevated inflammatory markers (CRP, IL-6). Generic 1.5:1 EPA:DHA fish oil is below the empirical EPA-fraction threshold; if mood is the goal, read the EPA mg per serving on the label.
Is fish oil good for the heart at 1 g/day?
Largely no, based on the largest trials. VITAL (n=25,871, healthy adults), ASCEND (n=15,480, diabetes), and OMEMI (n=1,027, elderly post-MI) all came back null at 1–1.8 g/day combined EPA+DHA. The cardiovascular benefit was demonstrated at the pharmaceutical 4 g/day pure-EPA dose in already-statin-treated adults with elevated triglycerides (REDUCE-IT). One-a-day grocery-store fish oil for 'heart health' in healthy adults is not supported by randomized trial evidence.
I'm vegan — can I get EPA from plant sources?
Not efficiently. ALA from flax, chia, and walnut converts to EPA at ~5% efficiency in humans; conversion to DHA is <1%. Two tablespoons of flax delivers maybe 50–100 mg of converted EPA — well below trial-level doses. The practical vegan solution is algal oil. Most algal products are DHA-dominant, but EPA-containing algal formulations are increasingly available; read the EPA mg per serving on the label.
What's the Omega-3 Index target?
Above 8% RBC EPA + DHA is associated with the lowest cardiovascular risk in observational data; below 4% is high-risk. Most U.S. adults sit around 4–6%. Reaching the 8% target generally requires 1–2 g/day of combined EPA+DHA from a triglyceride-form supplement, or several servings of fatty fish per week. Commercial home tests (e.g., OmegaQuant) are inexpensive and the most direct way to confirm whether your supplementation is doing what you want.
Should I worry about atrial fibrillation with high-dose EPA?
It's a real, dose-dependent signal. REDUCE-IT showed 5.3% vs 3.9% afib on 4 g/day icosapent ethyl; STRENGTH and OMEMI replicated with combined EPA+DHA. Meta-analyses now consistently identify a dose-dependent afib risk at ≥1 g/day omega-3. The absolute risk increase is small but meaningful. If you have a personal or family history of atrial fibrillation or structural heart disease, discuss with your cardiologist before chronic high-dose EPA — the risk/benefit calculation shifts.

References

Last reviewed2026-05-22

EPA

What the evidence says

Top EPA supplements

Clinical research on EPA

Triglyceride lowering

AHA Science Advisory — omega-3 for hypertriglyceridemia

Harris — fish oil triglyceride dose-response

REDUCE-IT — icosapent ethyl in high-risk adults

STRENGTH — combined EPA+DHA carboxylic acid (NEGATIVE)

VITAL — 1 g/day omega-3 for primary cardiovascular prevention

OMEMI — omega-3 in elderly post-MI patients

Curfman editorial — reconciling REDUCE-IT and STRENGTH

Mocking meta-analysis — EPA-dominant omega-3 for major depression

Sublette meta-analysis — the EPA fraction threshold

Goldberg & Katz — omega-3 for inflammatory joint pain

Goldberg & Katz — omega-3 for inflammatory joint pain

VITAL — 1 g/day omega-3 for primary cardiovascular prevention

Triglyceride form (rTG / natural TG)

Re-esterified triglyceride (rTG)

Triglyceride (natural or re-esterified)

rTG (re-esterified triglyceride)

Ethyl ester (EE)