Benefits
Triglyceride reduction
The most consistent fish oil benefit. Pharmaceutical-dose EPA+DHA or EPA-only at 4 g/day reduces triglycerides by 30% or more — works as monotherapy or alongside statins. Lower over-the-counter doses (1-2 g/day) produce more modest 11-15% reductions. Important caveat: EPA+DHA combinations may slightly raise LDL when treating severe hypertriglyceridemia (≥500 mg/dL), while purified EPA-only does not. AHA endorses 4 g/day for severe hypertriglyceridemia.
Cardiovascular events — formulation matters
In statin-treated high-risk patients, purified EPA-only at 4 g/day (icosapent ethyl) reduces major cardiovascular events by 25%. EPA+DHA combinations have not replicated this benefit in similar populations. Whether the difference is genuinely about EPA vs. combination, or about contested placebo choices in the trials, is not fully resolved. Most relevant for high-risk cardiac patients under physician supervision; the cardiovascular case for routine OTC fish oil in healthy adults is much weaker.
Atrial fibrillation risk at high doses
High-dose omega-3 supplementation increases the risk of new-onset atrial fibrillation. The signal is consistent across multiple major trials — both EPA-only and EPA+DHA. Magnitude is modest in absolute terms (a few percentage points over years of use) but real. This is the most clinically important downside of pharmaceutical-dose fish oil. Anyone with arrhythmia history or AF risk factors should consult a cardiologist before starting high-dose omega-3.
Depression — EPA-rich formulas as antidepressant adjunct
Fish oil reduces depressive symptoms when added to standard antidepressant therapy. Effect size is modest but reproducible across many trials. Critically, formulation matters: EPA needs to be at least 60% of total omega-3 content for reliable benefit. Most retail 'fish oil' products aren't EPA-dominant enough — read the label. Best used as adjunct at 1-2 g EPA/day, not as monotherapy for clinical depression.
Pregnancy and infant brain development
DHA is the dominant fatty acid in fetal and infant brain tissue. Maternal DHA status during pregnancy and breastfeeding correlates with infant visual and cognitive outcomes. WHO and most national obstetric bodies recommend 200-300 mg DHA per day during pregnancy and lactation. Choose third-party-tested anchovy/sardine or algal oil to minimize mercury exposure. Avoid cod liver oil during pregnancy — its vitamin A content carries teratogenicity risk.
Rheumatoid arthritis adjunct
At higher doses (2-4 g/day combined EPA+DHA), fish oil produces small-to-moderate reductions in tender and swollen joint counts and morning stiffness in rheumatoid arthritis. Some trials show NSAID-sparing effects. Consistent across meta-analyses but the clinical magnitude is modest — useful adjunct, not disease-modifying. Reasonable component of comprehensive RA management under rheumatologist supervision.
Quality and oxidation matter — most cheap fish oil is rancid
Fish oil oxidizes readily, and rancid oil may be pro-inflammatory rather than anti-inflammatory. Independent testing (IFOS, USP, ConsumerLab) consistently finds wide variation in actual EPA/DHA content and oxidation levels across retail products. Practical guidance: choose products with antioxidants added, opaque packaging, and recent manufacturing dates. Refrigerate after opening. Discard any oil with strong fishy or rancid smell — that's the opposite of what you want.
Mechanism of action
Hepatic VLDL synthesis reduction (triglyceride mechanism)
Pharmaceutical doses (≥3 g/day) reduce hepatic synthesis and secretion of very-low-density lipoprotein (VLDL) by approximately 30%. Multiple kinetic studies (Shearer 2013, PMC3563284) confirm that reduced VLDL output, not increased clearance, accounts for most of the triglyceride lowering effect. This is dose-dependent — explains why low supplement doses don't significantly affect triglycerides while pharmaceutical doses do.
Eicosanoid pathway modulation
EPA and DHA partially displace arachidonic acid from cell membrane phospholipids. When inflammation occurs, this membrane composition shift produces less inflammatory series-2 prostaglandins/thromboxanes and series-4 leukotrienes, and more anti-inflammatory series-3 prostaglandins, series-5 leukotrienes, and specialized pro-resolving mediators (resolvins, protectins, maresins). The mechanism takes weeks to develop because membrane turnover is slow.
Membrane composition and cell function
DHA is preferentially incorporated into neuronal and retinal membrane phospholipids, where it influences membrane fluidity, receptor function, and synaptic activity. This is the basis for the developmental DHA recommendation in pregnancy/infancy and for proposed cognitive effects in aging. EPA is more evenly distributed across triglycerides, cholesterol esters, and phospholipids.
Why EPA and DHA are not interchangeable
Clinical effects differ. EPA-dominant formulations show stronger evidence for depression and cardiovascular outcomes (REDUCE-IT, JELIS). DHA is more important for fetal/infant neural development and may be more relevant to cognitive applications in adults. In humans, dietary DHA can be retroconverted to EPA, but EPA→DHA conversion is minimal — these are functionally distinct fatty acids despite both being marine omega-3s.
Resolution of inflammation (specialized pro-resolving mediators)
EPA and DHA are precursors to resolvins (E-series from EPA, D-series from DHA), protectins, and maresins — specialized pro-resolving mediators that actively turn off inflammation rather than simply suppress it. This is a different concept from anti-inflammatory drugs (which block inflammation initiation) and is the focus of much recent research (Serhan and colleagues, 2010s-2020s).
Clinical trials
8,179 statin-treated patients with established CVD or diabetes plus risk factors, elevated triglycerides (135-499 mg/dL), randomized to icosapent ethyl 4 g/day or mineral oil placebo over median 4.9 years.
Clinical population described in trial publication.
8,179 statin-treated patients with established CVD or diabetes plus risk factors, elevated triglycerides (135-499 mg/dL), randomized to icosapent ethyl 4 g/day or mineral oil placebo over median 4.9 years. Primary composite endpoint (CV death, MI, stroke, revascularization, unstable angina) occurred in 17.2% on EPA vs. 22.0% placebo — HR 0.75, 25% relative risk reduction. Cardiovascular death reduced 20%. Atrial fibrillation increased (5.3% vs. 3.9%, p=0.004). Led to FDA approval of Vascepa for cardiovascular risk reduction in 2019.
13,078 statin-treated patients with high CV risk and atherogenic dyslipidemia randomized to omega-3 carboxylic acid (EPA+DHA) 4 g/day or corn oil placebo.
Clinical population described in trial publication.
13,078 statin-treated patients with high CV risk and atherogenic dyslipidemia randomized to omega-3 carboxylic acid (EPA+DHA) 4 g/day or corn oil placebo. Trial stopped early for futility — no difference in primary cardiovascular endpoint. Atrial fibrillation increased ~70% on omega-3. Trial chair Steven Nissen argued the contrasting REDUCE-IT result reflects mineral oil placebo bias rather than true EPA benefit; debate ongoing.
Comprehensive review concluded prescription EPA+DHA or EPA-only at 4 g/day (>3 g/day total EPA+DHA) is effective and safe for reducing triglycerides as monotherapy or with other lipid agents.
Clinical population described in trial publication.
Comprehensive review concluded prescription EPA+DHA or EPA-only at 4 g/day (>3 g/day total EPA+DHA) is effective and safe for reducing triglycerides as monotherapy or with other lipid agents. Triglyceride reduction ≥30% at this dose; smaller effects at lower doses. EPA-only formulations don't raise LDL while EPA+DHA at high doses can cause modest LDL increase in very high triglyceride patients.
20 studies (15 clinical trials + 5 observational) with 2,300 participants.
2,300 participants
20 studies (15 clinical trials + 5 observational) with 2,300 participants. Pooled effect size Hedge's g = -0.45 for omega-3 vs. placebo on depressive symptoms in MDD. Effect strongest with EPA-predominant formulations and concurrent antidepressant use, consistent with meta-regression findings. Best positioned as adjunct rather than primary MDD treatment.