Moderate Diet

Omega3 Repletion

Name: Omega3 Repletion
Creator: Realised
Published: 2026-06-01

Summary

EPA-led omega-3 supplementation corrects a genuine modern-diet deficiency (seed-oil-dominant eating → omega-6:3 imbalance, low oily-fish intake) — moderate evidence for depression at an EPA-dominant dose and for raising a low Omega-3 Index; but it is NULL for cardiovascular prevention in already-replete people (VITAL / ASCEND / STRENGTH / OMEMI), so it is a repletion lever, never cardioprotection.

Why Moderate

Tier 2 (Moderate) for the deficiency-correction / mood / index-restoration use because: multiple replicated meta-analyses (Liao, Mocking) and an evidence-graded guideline (Guu/ISNPR) converge on a small-to-moderate, EPA-fraction-dependent, baseline-dependent effect, and the Omega-3 Index biomarker is independently validated with a clean dose-response (Walker). NOT Tier 1 because the mood literature is heterogeneous, EPA-fraction- and design-sensitive (adjunctive > monotherapy), and the supplement evidence base is funding-exposed and publication-bias-prone. NOT Tier 3 because the effect is replicated across independent meta-analyses with a coherent dose-response and a strong mechanism — it is more than emerging.

Separately NULL (not a tier — an absence) for cardiovascular prevention in replete people because four large, well-powered RCTs (one NIH-funded, one nonprofit, one industry-sponsored, all concordant) found no benefit, and the high-dose arms found net harm (AF). HIGH confidence in this null.

Practical takeaway

Who benefits:
• Definitely: seed-oil-heavy / low-oily-fish diets (the dominant modern gap); vegans and vegetarians (no direct EPA/DHA source; ALA conversion fails); those with a documented low Omega-3 Index (<4–6%); adjunctive use in depression (EPA-predominant), especially at low baseline status.
• Possibly: those with a high-omega-6 inflammatory dietary pattern.
• NOT (the excluded claims): general cardiovascular prevention in replete people (VITAL/ASCEND null); "optimisation" dosing of an already-replete person (no added benefit; AF risk at high dose).

How to show the gap (gate signals — the ideal → the practical):
• Ideal biomarker: Omega-3 Index (RBC EPA+DHA%); target 8–12%, deficient <4%. The direct gate — not in the v1 data diet, but the future ideal.
• Dietary-pattern proxy: low oily-fish intake (<1–2 servings/week) AND/OR seed-oil-dominant cooking; vegan/vegetarian pattern (no marine source). Chat/onboarding-reportable — the v1 practical gate.
• Mood picture: low mood in a user with a low-omega-3 dietary pattern (adjunctive framing only, never a monotherapy claim).

Protocol — standard repletion:
• Dose: 1–2 g/day combined EPA+DHA from a quality fish oil; EPA-predominant (EPA fraction ≥60%) for the mood angle.
• Net EPA for mood: 1–2 g/day NET EPA (ISNPR guideline range); above ~2 g/day net EPA the benefit is unestablished and bleeding/AF risk rises.
• Vegan route: algal-oil EPA/DHA (ALA from flax does NOT substitute — conversion fails).
• Form quality: triglyceride or re-esterified-TG forms absorb better than ethyl-ester; oxidation/rancidity is a real quality problem — store cool, check freshness.
• Target ("what working looks like"): raise the Omega-3 Index into 8–12% over ~3–4 months (where measurable); for mood, an adjunctive lift over weeks, not a monotherapy cure.
• Ceiling: do not chase high doses. >2 g/day net EPA adds no established benefit; high-dose (≥1.8–4 g/day) carries a replicated atrial-fibrillation signal.

Evidence detail

Why This Entry Exists

A Realised user who eats little oily fish, cooks in seed oils, or follows a vegan/vegetarian pattern is very likely sitting in the modern omega-3 gap — a low Omega-3 Index and a skewed omega-6:omega-3 ratio. They may also be carrying low mood. The bad advice this entry protects against runs in two opposite directions at once.

The first is the cardioprotection over-promise — the dominant fish-oil-marketing story ("take fish oil for your heart"). For an already-replete or unselected person, this is simply false: four large, well-powered RCTs have falsified it. A platform that frames omega-3 as cardioprotection is repeating an industry line the evidence has retired, and it risks sending high-dose "optimisation" doses into a population where the only thing that reliably rises is the atrial-fibrillation signal.

The second is the opposite dismissal — concluding from the dead CVD trials that "fish oil does nothing." That throws out the honest, surviving use. Where baseline status is genuinely low, EPA-predominant supplementation has a small-to-moderate antidepressant effect (adjunctive), and a measured low Omega-3 Index can be raised into the target range over a few months. The deficiency-correction use survives scrutiny precisely where the cardioprotection use dies.

So this entry exists to hold the honest split: surface omega-3 as repletion only — correct a real gap, support mood at an EPA-dominant dose, raise a low index — and never as cardioprotection, and never at a high "more is better" dose.

Evidence

A. Deficiency / mood — the restoration use (the on-spine claim)

Depression — the EPA-predominant effect:
• Liao et al. (2019). Efficacy of omega-3 PUFAs in depression: a meta-analysis. Transl Psychiatry 9:190. Meta-analysis of 26 RCTs (n≈2,160). EPA-pure (100% EPA) and EPA-major (≥60% EPA) preparations were significantly more effective than placebo for depressive symptoms; DHA-pure and DHA-major preparations were not. Funding: academic (no commercial sponsor declared). Limitations: heterogeneity in dose, EPA:DHA ratio, and baseline severity; publication bias plausible in a supplement literature. (Note: Liao's benefit signal sat at an EPA dose ≤1 g/day; the higher 1–2 g/day net-EPA range below comes from the ISNPR guideline, not from Liao — see Guu.)
• Mocking et al. (2016). Meta-analysis and meta-regression of omega-3 PUFAs in major depressive disorder. Transl Psychiatry 6:e756. Meta-analysis + meta-regression of 13 RCTs (n≈1,233). Overall SMD favouring omega-3 (g≈0.40), effect driven by higher-EPA formulations and by adjunctive (add-on to antidepressant) use; monotherapy signal weaker. Funding: academic / non-commercial.
• Guu et al. (2019). International Society for Nutritional Psychiatry Research (ISNPR) practice guidelines for omega-3 in MDD. Psychother Psychosom 88:263–273. Expert, evidence-graded consensus guideline. Recommends EPA or EPA-predominant (EPA:DHA ≥2:1) at 1–2 g/day net EPA for major depression; explicitly notes pure DHA is ineffective; recommends omega-3 as an adjunct, not a substitute for standard treatment. Funding: professional society (ISNPR); declared expert panel. The EPA-dominance threshold is the load-bearing dosing fact.

Effect-size summary: small-to-moderate antidepressant effect (SMD ≈0.3–0.4), concentrated in EPA-predominant formulations and adjunctive use. Dose threshold: net EPA 1–2 g/day, EPA fraction ≥60% of total omega-3; above ~2 g/day net EPA the incremental benefit is not established and bleeding risk rises — more is NOT better. The effect is larger where baseline omega-3 status is low (the repletion logic — a replete person has little to gain). The mood literature is heterogeneous and the supplement evidence base is funding-exposed; this is Moderate, not Strong.

Omega-3 Index restoration — the biomarker gate:

The Omega-3 Index is the % of EPA+DHA in red-blood-cell membrane fatty acids — a validated, weeks-stable biomarker of long-term omega-3 status (the ideal gate for repletion, analogous to 25(OH)D for vitamin D).
• Harris & von Schacky (2004). The Omega-3 Index: a new risk factor for death from coronary heart disease? Prev Med 39:212–220. Proposed and validated RBC EPA+DHA% as a status biomarker; <4% = high-risk/deficient zone, 8–12% = target zone. Funding: academic (authors later commercially associated with index testing — note conflict; the biomarker itself is independently validated). Most Western populations sit in the 4–6% range (deficient-to-low) — the documented modern gap that repletion addresses.
• Walker et al. (2019). Predicting the effects of supplemental EPA and DHA on the omega-3 index. Am J Clin Nutr 110:1034–1040. Pooled dose-response model across multiple supplementation trials: ~1–2 g/day EPA+DHA raises the index into the target range over 3–4 months; baseline status and dose are the dominant variables. Funding: pooled academic trials.
B. Cardiovascular prevention — the NULL use (the EXCLUDED claim — Cui Bono)

> ⚑ The honest reversal. The omega-3 → cardioprotection claim — the dominant fish-oil-industry marketing story — is dead in replete people. Four large, well-powered RCTs are concordant. This entry exists partly to record that null so the spine never frames omega-3 as cardioprotection.
• Manson et al. (2019). Marine n-3 Fatty Acids and Prevention of Cardiovascular Disease and Cancer (VITAL). N Engl J Med 380:23–32. 1 g/day fish oil vs placebo — NO significant reduction in the primary composite of major cardiovascular events (HR 0.92, 95% CI 0.80–1.06) and NO effect on cancer. Landmark RCT (2×2 factorial), n=25,871 generally-replete US adults, 5.3 years. Funding: NIH (independent / government).
• ASCEND Study Collaborative Group (2018). Effects of n-3 Fatty Acid Supplements in Diabetes Mellitus. N Engl J Med 379:1540–1550. 1 g/day omega-3 vs placebo in diabetics — NO reduction in serious vascular events (RR 0.97, 95% CI 0.87–1.08). Large RCT, n=15,480 adults with diabetes, 7.4 years. Funding: British Heart Foundation + others (nonprofit / government-adjacent); supplement supplied by industry but trial independently run.
• Nicholls et al. (2020). Effect of high-dose omega-3 (STRENGTH trial). JAMA 324:2268–2280. 4 g/day carboxylic-acid EPA/DHA vs corn-oil comparator — TERMINATED for futility; NO cardiovascular benefit; increased atrial-fibrillation signal in the omega-3 arm (new-onset AF 2.2% vs 1.3%, p<0.001). Large RCT, n=13,078 high-CV-risk adults. Funding: industry-sponsored (AstraZeneca) — and it still found null, which strengthens the null. The high-dose, mixed-EPA/DHA preparation is directly relevant to the "more is better" framing.
• Kalstad et al. (2021). Effects of n-3 Fatty Acid Supplements in Elderly Patients After Myocardial Infarction (OMEMI). Circulation 143:528–539. 1.8 g/day EPA+DHA vs corn-oil placebo post-MI — NO reduction in clinical events (HR 1.08, 95% CI 0.82–1.41); a (non-significant directional) atrial-fibrillation signal again. RCT, n=1,027 elderly post-MI patients. Funding: investigator-initiated / academic, with supplement support.

The icosapent-ethyl caveat (do not use this to resurrect the claim):
• Bhatt et al. (2019). Cardiovascular Risk Reduction with Icosapent Ethyl (REDUCE-IT). N Engl J Med 380:11–22. High-dose PURIFIED EPA (4 g/day icosapent ethyl) in statin-treated patients with high triglycerides reduced events (HR 0.75) — BUT the mineral-oil placebo and the specific high-TG, high-EPA, prescription-drug context make this NOT generalisable to OTC fish-oil repletion, and the effect may be partly EPA-specific / partly a placebo-arm artefact (the mineral-oil arm showed rises in hsCRP, LDL, and other inflammatory biomarkers). RCT (heavily debated). Funding: industry-sponsored (Amarin). Do NOT use REDUCE-IT to resurrect a general cardioprotection claim. It is a high-TG pharmacotherapy result, not a repletion result. The four trials above govern the replete-population null.

Cardiovascular summary: in already-replete / unselected populations, omega-3 supplementation does NOT prevent cardiovascular disease. The platform must NEVER frame omega-3 as cardioprotection. High-dose omega-3 (STRENGTH 4 g, OMEMI 1.8 g) carries a real, replicated atrial-fibrillation signal — a harm that rises with dose, reinforcing that more is not better.

Mechanism

Plausibility is high for the membrane / anti-inflammatory / neuronal routes; the outcome is where the evidence is mixed, not the mechanism.

1. Membrane incorporation. EPA and DHA are incorporated into cell-membrane phospholipids, altering fluidity and the substrate pool for signalling. The brain is ~60% fat and DHA is a major structural component; EPA is the more bioactive anti-inflammatory/mood species.

2. Eicosanoid / resolvin shift. Omega-3-derived eicosanoids and specialised pro-resolving mediators (resolvins, protectins) are less inflammatory than the omega-6 (arachidonic-acid)–derived series. The modern ~15–20:1 omega-6:omega-3 ratio (vs an ancestral ~1–4:1) pushes the balance toward a pro-inflammatory set point — the structural insult repletion offsets.

3. Neurotransmission / neuroinflammation. EPA's antidepressant plausibility runs through reduced neuroinflammation and membrane-mediated effects on monoamine signalling; this is why EPA-predominant (not DHA-pure) formulations carry the mood effect.

Why flax is not a substitute (conversion failure): ALA (plant omega-3, from flax/chia/walnut) converts to EPA at only ~5–10% and to DHA at ~0–5% in most adults — so flax is NOT a substitute for fish/algal EPA+DHA. Vegans and the fish-averse cannot close the gap with ALA alone; an algal-oil EPA/DHA supplement is the vegan-compatible repletion route.

Risks And Contraindications

Harm tier: low at the repletion dose (a modest, well-tolerated supplement at 1–2 g/day) — but NOT negligible. The anticoagulant interaction and the dose-dependent AF signal are real; the repletion dose sits below the AF-signal range, so the harm is a property of high-dose "optimisation," which this entry excludes.
• Anticoagulant / antiplatelet interaction. Omega-3 has a mild antiplatelet effect. With warfarin, DOACs, or antiplatelet agents (aspirin/clopidogrel), high-dose fish oil can additively raise bleeding risk — defer to the prescriber; dose is informational.
• Atrial fibrillation. High-dose omega-3 (≥1.8 g/day in OMEMI, 4 g/day in STRENGTH) is associated with increased incident AF. A reason the repletion dose stays modest (1–2 g/day) and high-dose "optimisation" is excluded.
• Perisurgical. Consider pausing high-dose fish oil before surgery (bleeding) per clinician guidance.

Controversy

Position A (industry / popular framing): "Fish oil is cardioprotective — take it for your heart." This narrative built a multi-billion-dollar supplement category.

Position B (the large-RCT evidence): In replete / unselected populations, omega-3 supplementation does not prevent cardiovascular disease. VITAL, ASCEND, STRENGTH, and OMEMI are concordant nulls; high-dose arms add an atrial-fibrillation harm. The only surviving outcome uses are deficiency correction, EPA-led adjunctive mood support, and raising a low Omega-3 Index.

Funding-bias dimension: the cardioprotection claim was the commercial engine of the category, yet the trials that falsified it include NIH-funded (VITAL) and nonprofit/government-adjacent (ASCEND) work — and even the industry-sponsored high-dose trial (STRENGTH) returned null. When industry-funded trials fail to find the effect their sponsor would profit from, the null is especially trustworthy. REDUCE-IT (industry-sponsored, positive) is the apparent exception, but its mineral-oil placebo and high-TG prescription-drug context make it a pharmacotherapy result, not a repletion result.

Realised Position: author and surface omega-3 as repletion only — correct a real modern gap, support mood at an EPA-dominant dose, raise a low index — never as cardioprotection, and never at a high "more is better" dose. The deficiency-correction use survives the reversal test; the cardioprotection use does not.

Cross-Pillar Connections

• diet_practical_implementation — the dietary-pattern grounding: SMASH-style oily fish 2–3×/week, the 15–20:1 vs 1–4:1 omega-6:omega-3 ratio, reduce seed oils. Food-first sits upstream of supplementation.
• vegan_vegetarian_diet_health_optimisation — the ALA-conversion-failure / algal-oil EPA-DHA grounding for plant-based users.
• magnesium_supplementation_sleep — sibling repletion entry (the gated-repletion shape: correct a real deficiency, don't over-dose).
• vitamin_d3_high_dose_supplementation — sibling repletion entry (the deficiency-stronger / null-in-replete parallel).

What would change our mind

Falsifiability: explicit upgrade/downgrade criteria from source

Upgrade the mood use toward Tier 1 if: a large, independent, pre-registered RCT of EPA-predominant supplementation (≥60% EPA, 1–2 g/day net EPA) in low-baseline-status depressed adults replicates a clinically meaningful effect with low heterogeneity, ideally stratified by baseline Omega-3 Index.

Re-open the cardiovascular question if: a well-designed RCT with an inert placebo (resolving the REDUCE-IT mineral-oil confound) demonstrates event reduction in a defined population — and even then, the claim would be specific to that population (e.g., high-TG, purified high-dose EPA), never a general OTC repletion claim.

Downgrade the deficiency use if: meta-analytic re-evaluation shows the mood effect is an artefact of publication bias or driven entirely by methodologically weak trials, or if Omega-3 Index restoration is shown not to track any meaningful outcome.

Industry bias note

Structural incentives the evidence base may reflect

⚑ The fish-oil industry has systematically over-claimed cardioprotection. Omega-3 is a multi-billion-dollar supplement category, and the cardioprotection narrative drove much of that market — a narrative the four large independent/government-adjacent RCTs (VITAL [NIH-funded], ASCEND [BHF/others], STRENGTH [industry, still null], OMEMI) have now falsified for replete people. The reversal test matters here: the deficiency-correction / mood / index-restoration use survives honest scrutiny (NIH-independent meta-analyses, baseline-dependent effect, mechanistically grounded); the cardioprotection use does not. Author and surface omega-3 as repletion only.

A second, subtler bias runs the other way and must be named: the Omega-3 Index biomarker, while independently validated, is commercially promoted by testing companies associated with its originators — so "get your index tested" carries a mild sell-side incentive. Realised uses the index as the ideal future gate and the dietary-pattern proxy as the v1 practical gate, not as a paid-test upsell.

The REDUCE-IT positive result is a high-triglyceride prescription-drug context with a contested mineral-oil placebo — it must not be used to resurrect a general OTC cardioprotection claim.

Sources (10)

Open in the Library: search, filter, every entry →