Does fish oil lower cholesterol and triglycerides?

It lowers triglycerides, and the drop grows with the dose. A 2023 meta-analysis of 90 randomized trials in 72,598 people found 2 g a day of EPA plus DHA cut triglycerides by 42.61 mg/dL (95% CI 31.80 to 53.41) and 3 g a day by 68.90 mg/dL. The median trial lasted 13 weeks, and 52 of the 90 trials were in people with high blood lipids. LDL cholesterol, the number most people mean by cholesterol, did not fall in that analysis.

Established. Fish oil lowers triglycerides. A Cochrane review of 86 trials in 162,796 people, each lasting at least a year, rated a fall of about 15% as high-certainty evidence, and the fall grew with dose. Most of those trials used capsules, so this is evidence about supplements more than about eating fish.

Myth. Fish oil does not lower LDL cholesterol. In the dose-response analysis the LDL curve was J-shaped and peaked at 1.75 g a day with a rise of 2.91 mg/dL (95% CI 0.34 to 5.47). The Cochrane review found little or no effect on blood lipids other than triglycerides.

What the trials found

The higher your triglycerides, the more the dose matters

In people who started with high blood lipids, the fall kept growing roughly in line with the dose. In people without, it levelled off at roughly 40 mg/dL. Only 11 trials were limited to that second group, so the plateau rests on less data than the headline.

The biggest falls came from people who started highest. In 9 trials of people with HIV on antiretroviral therapy and high triglycerides, EPA plus DHA lowered triglycerides by 77.55 mg/dL (CI 33.25 to 121.85). In an 8-week trial of 116 people with coronary artery disease and triglycerides above 200 mg/dL, 1 g a day of DHA cut triglycerides by 21.8%, and a DHA plus EPA product by 18.3%. That trial had no placebo arm, so both numbers are falls from each group's own starting level.

Against corn oil, across 16 trials, fish oil supplements lowered triglycerides by 25.50 mg/dL (95% CI 8.57 to 42.44). Corn oil is not a fully inert comparison, and the authors ask for larger trials. A network meta-analysis of 64 trials put the effect at 8.971 mg/dL for each gram of omega-3 from fish oil and 9.838 mg/dL for each gram from krill oil. Gram for gram, krill oil did no better.

Cholesterol

The same 2023 analysis found HDL cholesterol rose by 3.48 mg/dL (95% CI 1.09 to 5.86) at the dose where LDL peaked. A rise in HDL from a supplement is a change in a blood test, and the review does not show it changes heart outcomes.

Unsettled. One small set of trials disagrees on LDL. In 10 trials of 633 people with ischemic heart disease, 1 to 4 g a day of EPA plus DHA lowered LDL by 9.43 mg/dL (95% CI 4.65 to 14.20) and triglycerides by 17.53 mg/dL. Those trials were small, and the larger syntheses above found no LDL fall, so this result does not overturn them.

Do lower triglycerides mean fewer heart attacks?

In the STRENGTH trial, 13,078 statin users at high cardiovascular risk with high triglycerides took 4 g a day of EPA plus DHA or corn oil. The main outcome, heart attacks, strokes and heart deaths together, occurred in 12.0% against 12.2% (hazard ratio 0.99, 95% CI 0.90 to 1.09). Triglycerides fell and events did not.

Unsettled. The American Heart Association cites a second large trial, REDUCE-IT, which used purified EPA alone and reported fewer events. We could not retrieve its abstract, so it has no card and no number on this page. The two trials used different products, and that is why the question of events stays open.

Who should be careful

Atrial fibrillation, an irregular heart rhythm, is the clearest harm. Across 8 large trials in 83,112 people, it occurred in 4.0% of those taking omega-3 against 3.3% on control, a relative rise of 24% (RR 1.24, 95% CI 1.11 to 1.38). The risk rose with dose, and triglyceride lowering uses the higher doses.

Bleeding overall did not rise. Across 11 trials in 120,643 people, the rate ratio was 1.09 (95% CI 0.91 to 1.31). The authors linked a small rise to high doses of purified EPA. In 21 trials of prescription omega-3 products in 24,460 people, a fishy taste and skin rashes such as itching or eczema were more common than on control.

For people already on a statin, a 2012 review by the US Agency for Healthcare Research and Quality found low-grade evidence that adding omega-3 changed neither LDL, HDL, clotting tests nor bleeding time compared with the statin alone. Our search of a drug interaction database returned nothing for fish oil. We have no card on pregnancy, children or anticoagulants such as warfarin. That is a gap in what we found, and it does not mean these groups are safe.

Not for everyone. If you have atrial fibrillation, take a blood thinner or are about to have surgery, high-dose fish oil for triglycerides is a decision to make with whoever treats you.

What expert bodies say

The American Heart Association concluded in 2019 that prescription omega-3 at 4 g a day, more than 3 g of it EPA plus DHA, is an effective and safe option for lowering triglycerides, alone or alongside other lipid-lowering drugs. In very high triglycerides, 500 mg/dL or more, it says EPA plus DHA at that dose cuts triglycerides by at least 30%, with a rise in LDL cholesterol that EPA alone did not cause. These are prescription products, and supplement capsules vary in how much EPA and DHA they carry.

The European Food Safety Authority concluded that supplemental EPA plus DHA up to 5 g a day raises no safety concerns for adults, and it set no upper limit. The EU authorizes the label claim that DHA and EPA contribute to the maintenance of normal blood triglyceride levels. That claim is about keeping normal levels normal. It does not cover treating high triglycerides.

How we searched

What would change this answer

• A new dose-response meta-analysis. We found none after 2023 on triglycerides.
• Trials comparing purified EPA with mixed EPA plus DHA on heart events, such as RESPECT-EPA, and on LDL in people with very high triglycerides. These would shift the balance between benefit and the rhythm and bleeding risks.
• A check of trial registries for ongoing work. Our registry query failed on a syntax error and needs to be rerun.

Sources

1. ev-fotg-01 · Meta-analysis or systematic review · dose-response meta-analysis of randomized controlled trials (1-stage restricted cubic spline) · n = 90 RCTs, 72598 participants
   The mean change in triglyceride was −42.61 (95% CI, −53.41 to −31.80) mg/dL for 2 g/d and −68.90 (95% CI, −98.40 to −39.40) mg/dL for 3 g/d of DHA + EPA.
   
   Who: participants in randomized controlled trials of DHA, EPA or both, published 1990 to 2022, mean age 25.7 to 70.0 years
   Effect: TG -42.61 mg/dL (95% CI -53.41 to -31.80) at 2 g/d; -68.90 mg/dL (95% CI -98.40 to -39.40) at 3 g/d; near-linear dose-response
   Certainty: Best synthesis for the question; median trial length 13 weeks; most comparators olive or other vegetable oils; 52 of 90 trials in people with hyperlipidemia.
   J Am Heart Assoc, 2023 · checked 2026-10-07 · we read the fulltext
2. ev-fotg-02 · Meta-analysis or systematic review · subgroup dose-response meta-analysis of randomized controlled trials · n = 90 RCTs
   In subgroup studies stratified by prespecified hyperlipidemic status at entry, the approximately linear trend for triglyceride change was found only in the population with hyperlipidemia but not in the population without hyperlipidemia, where the effect stably plateaued at roughly 40 mg/dL.
   
   Who: trials stratified by hyperlipidemia at entry (52 trials with hyperlipidemia, 11 without)
   Effect: near-linear TG fall with dose in hyperlipidemia; plateau at roughly -40 mg/dL without hyperlipidemia
   Certainty: Subgroup analysis; only 11 trials restricted to people without hyperlipidemia.
   J Am Heart Assoc, 2023 · checked 2026-10-07 · we read the fulltext
3. ev-fotg-03 · Meta-analysis or systematic review · dose-response meta-analysis of randomized controlled trials · n = 80 RCTs for LDL-C
   The J‐shaped curve for LDL‐C change peaked at 1.75 g/d intake with a moderate LDL‐C increment of 2.91 (95% CI, 0.34−5.47) and HDL‐C increment of 3.48 (95% CI, 1.09−5.86) mg/dL, respectively.
   
   Who: participants in randomized controlled trials of DHA plus EPA
   Effect: J-shaped LDL-C curve peaking at 1.75 g/d: LDL-C +2.91 mg/dL (95% CI 0.34 to 5.47); HDL-C +3.48 mg/dL (95% CI 1.09 to 5.86)
   Certainty: Answers the cholesterol half of the question: fish oil is not an LDL-lowering treatment.
   J Am Heart Assoc, 2023 · checked 2026-10-07 · we read the fulltext
4. ev-fotg-04 · Meta-analysis or systematic review · Cochrane systematic review and meta-analysis of randomized controlled trials · n = 86 RCTs, 162796 participants
   Increasing LCn3 and ALA had little or no effect on serious adverse events, adiposity, lipids and blood pressure, except increasing LCn3 reduced triglycerides by ˜15% in a dose-dependent way (high-certainty evidence).
   
   Who: adults at varying cardiovascular risk in RCTs lasting at least 12 months, mainly in high-income countries
   Effect: TG about -15%, dose-dependent (high-certainty evidence); little or no effect on other lipids, adiposity or blood pressure
   Certainty: Long trials only (12 to 88 months); most used capsules, so evidence is mainly from supplements, little from eating fish.
   Cochrane Database Syst Rev, 2020 · checked 2026-10-07 · we read the abstract
5. ev-fotg-05 · Meta-analysis or systematic review · systematic review and meta-analysis of randomized controlled trials · n = 16 RCTs
   The results revealed that the fish oil supplements significantly reduced TG (WMD: - 25.50 mg/dl, 95% CI: - 42.44, - 8.57, P = 0.000) levels compared to corn oil.
   
   Who: participants in randomized controlled trials comparing fish oil supplements with corn oil
   Effect: TG WMD -25.50 mg/dL (95% CI -42.44 to -8.57); HDL +2.54 mg/dL (95% CI 0.55 to 4.52); no significant change in TC or LDL
   Certainty: Corn oil is not fully inert; the authors call for larger trials.
   Syst Rev, 2024 · checked 2026-10-07 · we read the abstract
6. ev-fotg-06 · Meta-analysis or systematic review · network meta-analysis and meta-regression of randomized controlled trials · n = 64 RCTs
   One gram of n-3 fatty acids contained in fish oil and krill oil lowered median triglycerides by 8.971 mg/dL (95% credible interval [CrI], 2.27 to 14.04) and 9.838 mg/dL (95%CrI, 0.72 to 19.40), respectively.
   
   Who: participants in randomized controlled trials of krill oil or fish oil
   Effect: per 1 g n-3: TG -8.971 mg/dL (95% CrI 2.27 to 14.04) for fish oil, -9.838 mg/dL (95% CrI 0.72 to 19.40) for krill oil; krill vs fish oil TG WMD -4.07 (95% CI -15.22 to 7.08), not significant
   Certainty: The effect follows the omega-3 dose, not the oil type.
   Nutr Rev, 2020 · checked 2026-10-07 · we read the abstract
7. ev-fotg-07 · Meta-analysis or systematic review · systematic review and meta-analysis of randomized controlled trials · n = 633
   Omega-3 significantly reduced triglycerides (MD: -17.53 mg/dL, 95% CI: -30.64 to -4.41; p = 0.009) and LDL-C (MD: -9.43 mg/dL, 95% CI: -14.20 to -4.65; p = 0.0001).
   
   Who: adults with ischemic heart disease in RCTs of omega-3 (1-4 g/day EPA + DHA) versus placebo
   Effect: TG MD -17.53 mg/dL (95% CI -30.64 to -4.41); LDL-C MD -9.43 mg/dL (95% CI -14.20 to -4.65); moderate certainty (GRADE)
   Certainty: Small trials (633 people in all); LDL result differs from the larger general-population syntheses.
   Nutr Metab Cardiovasc Dis, 2026 · checked 2026-10-07 · we read the abstract
8. ev-fotg-08 · Meta-analysis or systematic review · systematic review and meta-analysis of randomized clinical trials · n = 9 RCTs
   The reduction of triglycerides level was -77.55 mg (IC of -121.85 to -33.25) in Omega 3 groups.
   
   Who: HIV/AIDS patients with hypertriglyceridemia in stable use of antiretroviral therapy
   Effect: TG -77.55 mg/dL (CI -121.85 to -33.25); trials with baseline TG >200 mg/dL: -114.15 mg/dL (CI -162.34 to -65.97)
   Certainty: Higher starting triglycerides, larger drop; small specific population.
   Cien Saude Colet, 2017 · checked 2026-10-07 · we read the abstract
9. ev-fotg-09 · Randomized controlled trial(s) · prospective randomized double-blind trial, 8 weeks · n = 116
   Triglycerides decreased by an average of 21.8% in the DHA group (p < 0.001) and 18.3% in the DHA + EPA group (p < 0.001).
   
   Who: patients with coronary artery disease and triglycerides greater than 200 mg/dL, mean age 69.4 years
   Effect: TG -21.8% with 1000 mg DHA and -18.3% with 1252 mg DHA + EPA (both p < 0.001); between-group difference not significant
   Certainty: No placebo arm; compares two omega-3 products.
   J Am Coll Nutr, 2006 · checked 2026-10-07 · we read the abstract
10. ev-fotg-10 · Randomized controlled trial(s) · double-blind randomized multicentre trial, 22 countries · n = 13078
    The primary end point occurred in 785 patients (12.0%) treated with omega-3 CA vs 795 (12.2%) treated with corn oil (hazard ratio, 0.99 [95% CI, 0.90-1.09]; P = .84).
    
    Who: statin-treated patients at high cardiovascular risk with hypertriglyceridemia and low HDL-C, median TG 240 mg/dL
    Effect: primary MACE 12.0% vs 12.2%, HR 0.99 (95% CI 0.90 to 1.09); trial stopped early for futility; GI adverse events 24.7% vs 14.7%
    Certainty: Lowering triglycerides with fish oil did not translate into fewer events here; REDUCE-IT with purified EPA did (via AHA advisory, card 12).
    JAMA, 2020 · checked 2026-10-07 · we read the abstract
11. ev-fotg-11 · Position of an expert body · science advisory reviewing trials of prescription omega-3 · n = —
    In treatment of very high triglycerides with 4 g/d, EPA+DHA agents reduce triglycerides by ≥30% with concurrent increases in low-density lipoprotein cholesterol, whereas EPA-only did not raise low-density lipoprotein cholesterol in very high triglycerides.
    
    Who: patients with very high triglycerides (500 mg/dL or more), US clinical guidance
    Effect: TG -30% or more with 4 g/d in very high TG; LDL-C rises with EPA+DHA, not with EPA-only
    Certainty: Prescription doses (>3 g/d EPA+DHA), not typical over-the-counter capsules; PMID outside the local corpus.
    Circulation, 2019, 140(12):e673-e691 (Skulas-Ray et al., AHA science advisory) · checked 2026-10-07 · we read the abstract
12. ev-fotg-12 · Position of an expert body · science advisory · n = —
    We conclude that prescription n-3 FAs (EPA+DHA or EPA-only) at a dose of 4 g/d (>3 g/d total EPA+DHA) are an effective and safe option for reducing triglycerides as monotherapy or as an adjunct to other lipid-lowering agents.
    
    Who: adults with hypertriglyceridemia, US clinical guidance
    Effect: prescription n-3 FA 4 g/d (>3 g/d EPA+DHA) recommended for TG lowering; cites 25% fewer major cardiovascular events with EPA-only in REDUCE-IT
    Certainty: Applies to prescription products; supplement capsules vary in EPA+DHA content.
    Circulation, 2019, 140(12):e673-e691 (Skulas-Ray et al., AHA science advisory) · checked 2026-10-07 · we read the abstract
13. ev-fotg-13 · Meta-analysis or systematic review · systematic review and meta-analysis of randomized controlled trials with meta-regression · n = 83112
    In meta-analysis, a significant association was noted between n-3 FA treatment and risk of AF (4.0% vs 3.3%; RR 1.24, 95% CI 1.11-1.38, p = 0.0002).
    
    Who: participants in cardiovascular outcome RCTs with at least 1000 people and at least 1 year of follow-up
    Effect: AF 4.0% vs 3.3%, RR 1.24 (95% CI 1.11 to 1.38); <=1 g/d RR 1.12 (1.04 to 1.21); >1 g/d RR 1.51 (1.26 to 1.80); no rise in stroke (RR 1.04)
    Certainty: Dose-related risk; matters most at the higher doses used to lower triglycerides.
    Cardiovasc Drugs Ther, 2021 · checked 2026-10-07 · we read the abstract
14. ev-fotg-14 · Meta-analysis or systematic review · systematic review and meta-analysis of randomized clinical trials · n = 120643
    There was no difference in the pooled meta-analytic events of bleeding among patients receiving omega-3 PUFAs and those in the control group (rate ratio [RR], 1.09 [95% CI, 0.91-1.31]; P=0.34).
    
    Who: patients in randomized clinical trials receiving omega-3 PUFAs
    Effect: bleeding RR 1.09 (95% CI 0.91 to 1.31); high-dose purified EPA: 50% relative, 0.6% absolute increase; risk linked to EPA dose, not to antiplatelet use
    Certainty: Relevant for people on blood thinners or before surgery; clinical significance judged very modest.
    J Am Heart Assoc, 2024 · checked 2026-10-07 · we read the abstract
15. ev-fotg-15 · Meta-analysis or systematic review · systematic review and meta-analysis of randomized controlled trials · n = 24460
    Compared with the control group, RxOME3FAs were associated with more treatment-related dysgeusia (fishy taste; p = 0.011) and skin abnormalities (eruption, itching, exanthema, or eczema; p < 0.001).
    
    Who: participants in RCTs of marketed prescription omega-3 fatty acid products
    Effect: more dysgeusia (p = 0.011) and skin abnormalities (p < 0.001); EPA/DHA combination products: more belching and nausea, LDL-C +4.106 mg/dL (p = 0.009)
    Certainty: Prescription products at triglyceride-lowering doses.
    Prostaglandins Leukot Essent Fatty Acids, 2018 · checked 2026-10-07 · we read the abstract
16. ev-fotg-16 · Position of an expert body · agency scientific opinion (news summary) · n = —
    It concludes that supplemental intakes of EPA and DHA combined at doses up to 5g a day do not raise safety concerns for adults.
    
    Who: adults in the EU
    Effect: EPA+DHA up to 5 g/day supplemental: no safety concern; data insufficient to set a UL
    Certainty: Triglyceride-lowering doses (2-4 g/d) sit below this level; EU intakes are generally far below 5 g/d.
    EFSA, 2012, news: EFSA assesses safety of long-chain omega-3 fatty acids (Scientific Opinion EFSA Journal 2012;10(7):2815) · checked 2026-10-07 · we read the section
17. ev-fotg-17 · Position of an expert body · regulatory register entry · n = —
    POL-HC-8362 Docosahexaenoic acid and Eicosapentaenoic acid (DHA/EPA) DHA and EPA contribute to the maintenance of normal blood triglyceride levels Authorised
    
    Who: EU food labeling
    Effect: authorized health claim (conditions of use in Regulation (EU) No 432/2012, not quoted here)
    Certainty: The claim is about maintaining normal triglycerides, not treating high ones; dose condition sits in the regulation.
    EU Register on nutrition and health claims, claim POL-HC-8362, snapshot 2026-09-21 · checked 2026-10-07 · we read the section
18. ev-fotg-18 · Meta-analysis or systematic review · agency systematic review (AHRQ comparative effectiveness review) · n = 57 RCTs and 2 controlled trials (whole key question)
    Low grade evidence indicated no difference between omega-3 fatty acids plus statins compared with statins alone in the following measures: HDL-C, LDL-C, achieving LDL-C and HDL-C targets; diastolic blood pressure; C-reactive protein; blood coagulation parameters; and bleeding time.
    
    Who: adults taking cardiovascular drugs (statins) in randomized trials
    Effect: omega-3 plus statins vs statins alone: no difference in HDL-C, LDL-C, coagulation parameters or bleeding time (low strength of evidence)
    Certainty: Mostly small, short trials; triglyceride benefit on top of statins seen mainly with baseline TG of 200 mg/dL or more.
    Agency for Healthcare Research and Quality, 2012, Dietary Supplements in Adults Taking Cardiovascular Drugs (Comparative Effectiveness Review 51) · checked 2026-10-07 · we read the section