Omega-3 Fatty Acids (EPA/DHA)

At a glance

Also calledFish oil, omega-3s, EPA and DHA, krill oil, cod liver oil
Fish oil, marine omega-3s, n-3 PUFAs, eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), alpha-linolenic acid (ALA, plant-based omega-3), icosapent ethyl (Vascepa, prescription EPA), krill oil, cod liver oil
Typical doseUp to 1 g of EPA a day in EPA-rich oil, with antidepressants
General wellness / OTC supplementation: 250-500mg/day combined EPA+DHA (a typical single fish-oil capsule provides roughly 300mg combined EPA+DHA per 1000mg fish oil, though this varies widely by brand). Prescription cardiovascular risk reduction: 4g/day icosapent ethyl, purified EPA (studied in REDUCE-IT, Bhatt et al. 2019). Depression (adjunct to antidepressants): most consistent benefit seen with EPA-predominant (≥60% EPA) formulations at ≤1g/day EPA (Liao et al. 2019). Brain-health/cognitive claims: no dose in the current evidence base (up to 2g/day DHA tested in Yassine et al. 2026) has shown a clear cognitive benefit in a rigorous RCT.
Upper limitNone set — NIH guidance says no more than about 3 g a day
No Tolerable Upper Intake Level (UL) established by the Institute of Medicine for any omega-3 fatty acid. General safety guidance (NIH Office of Dietary Supplements): no more than ~3g/day combined EPA+DHA from all sources (food + supplements), including no more than 2g/day from supplements specifically, without medical supervision — higher doses (as used in some cardiovascular trials) increase bleeding and atrial fibrillation risk and should be physician-supervised.
OnsetBlood levels rise in hours to days; effects take weeks
Plasma EPA/DHA levels rise within hours to days of starting supplementation (Schmieta et al., 2025), but functional/clinical effects (e.g. triglyceride lowering, any cognitive or mood effect) generally require weeks of sustained daily use to build up in cell membranes/tissue — not established as a single precise figure across all outcomes.
LastsThought to last weeks after stopping, held in cell membranes
Not established as a single figure across outcomes. Incorporation into tissue and cell membranes is thought to persist for weeks after stopping regular supplementation, considerably longer than the single-dose plasma half-life.
Half-lifeAbout 45 hours for EPA, 66 for DHA, after one dose
Plasma: EPA ~45 hours, DHA ~66 hours after a single dose (Schmieta et al., 2025). This is single-dose blood clearance only — tissue/membrane incorporation with regular daily use follows a much slower, cumulative timeline over weeks, and is not the same measurement.
Found inOily fish such as mackerel, fish oil, prescription EPA
Fatty fish: mackerel (~2000mg EPA+DHA per 3oz cooked serving), Atlantic salmon (~1800mg per 3oz), sardines (~1400mg per 3oz). Standard over-the-counter fish oil supplement: often ~1000mg fish oil per capsule, but frequently only ~300mg of that is actual EPA+DHA. Prescription icosapent ethyl (Vascepa): 4g/day purified EPA, FDA-approved for cardiovascular risk reduction in specific high-risk patients. Plant sources (ALA only, not directly EPA/DHA): flaxseed, chia seeds, walnuts, soybean and canola oil.

Myth check

Fish oil / omega-3 supplements protect and boost brain health — sharpening memory and cognition, and protecting the brain against injury (including concussions) — essentially treated as a universal, low-risk 'brain supplement.'

The evidence is far more mixed and outcome-specific than the popular framing suggests. A 2026 USC-led RCT found that high-dose DHA reached the brain (confirmed via CSF biomarkers) but produced no measurable improvement in cognition or brain structure over 24 months in an at-risk older-adult population (Yassine et al., 2026). A separate 2026 mouse+human-tissue study found that EPA specifically can impair blood-vessel repair after repetitive mild brain injury, a pattern also seen in postmortem CTE tissue (Karakaya et al., 2026) — directly cutting against the 'protects against head injury' framing endorsed elsewhere (e.g. the ISSN's 2025 position stand). A 2026 observational analysis of the ADNI dataset found omega-3 supplement users declined faster cognitively than non-users (Liao et al., 2026), though as an observational study this cannot rule out reverse causation. Meanwhile, the pre-2026 literature on cognition is itself modest at best: a 2025 overview of systematic reviews found only a small pooled benefit (Barros et al., 2025). The strongest, most consistent evidence for omega-3 lies elsewhere: high-dose purified EPA measurably reduces cardiovascular events in high-risk patients (Bhatt et al., 2019; Khan et al., 2021), though an earlier meta-analysis of lower-dose supplements found no such benefit (Aung et al., 2018), and cardiovascular benefit comes with a real, dose-dependent increase in atrial fibrillation risk (Gencer et al., 2021).

The 'brain supplement' framing runs ahead of the 2026 evidence — the best-supported real benefit of omega-3 is cardiovascular (and dose/formulation-specific, mainly high-dose EPA), not cognitive, and the newest, most rigorous brain-focused research (a large RCT, a mechanistic injury study, and an observational cohort study) all point away from a general brain-protective effect rather than toward one.

The findingsStrongest first

1↓

Mood & wellbeingOnly the EPA kind of fish oil moved depression scores

Across 26 trials in 2,160 people taking antidepressants too, omega-3 modestly beat placebo, and nearly all of it came from capsules built on EPA rather than DHA.

★☆☆CautionSupport mood

Sources1 behind this page

Efficacy of omega-3 PUFAs in depression: A meta-analysisLiao et al., 2019DOI 10.1038/s41398-019-0515-5 (PMID 31383846)

Taken with something else

What the evidence says about combining Omega-3 Fatty Acids (EPA/DHA) with each of these. The marker is how well established the interaction is; a caution is about the hazard, which is a different question.

with Vitamin D3 (Cholecalciferol)No effectNothing changes.★★★Well replicated

Omega-3 Fatty Acids (EPA/DHA)

At a glance

Myth check

Taken with something else

What we don't know

Try one of these

Or start from a goal