Omega-3: EPA vs DHA, Different Jobs, Different Doses

Fish oil is the second-most-purchased supplement in the United States, and its label is one of the most quietly misleading in the aisle. Most bottles report a single number — total omega-3, usually in the 1000 to 1500 mg range — as if that number captured what the product does. It does not.

EPA and DHA are two structurally distinct long-chain omega-3 fatty acids, and they do fundamentally different jobs in the body. EPA is the anti-inflammatory and cardiovascular workhorse. DHA is a structural lipid — a physical building block of brain and retinal membranes. Trials that lump them together produce muddy results. Trials that separate them produce cleaner and sometimes opposite signals.

This piece walks through what each does, the pair of large cardiovascular trials that got wildly different answers, how to translate that into your own dosing, and how to source a fish oil that has not gone rancid in the bottle.

What EPA and DHA are, structurally

Both are long-chain polyunsaturated fatty acids. EPA (eicosapentaenoic acid) is 20 carbons with 5 double bonds. DHA (docosahexaenoic acid) is 22 carbons with 6 double bonds. They come from the same biosynthetic pathway — plant ALA (alpha-linolenic acid, from flax and chia) is elongated and desaturated into EPA, which is further elongated and desaturated into DHA — but the conversion in humans is slow and the endpoint tissue distribution is very different.

EPA is the primary substrate for anti-inflammatory eicosanoids — specifically the series-3 prostaglandins and thromboxanes and the series-5 leukotrienes. These pathways compete directly with the arachidonic-acid-derived, pro-inflammatory series-2 and series-4 eicosanoids that dominate a typical Western diet. Raising the EPA-to-arachidonic-acid ratio in your membranes shifts the eicosanoid profile toward less inflammation.

DHA, by contrast, is a structural molecule. It makes up roughly 30 percent of the phospholipid fatty acids in the brain gray matter and up to 60 percent in the retinal photoreceptor outer segments. Its geometry gives neuronal and photoreceptor membranes the fluidity they need for receptor movement and signal transduction. It is not primarily an anti-inflammatory. It is a construction material.

Those different roles predict — and are borne out by — different effect patterns in trials.

The two big cardiovascular trials that disagreed

The clearest recent evidence for the EPA-versus-DHA distinction comes from two large randomized trials that used almost identical designs but different formulations.

REDUCE-IT (Bhatt 2019). 8,179 statin-treated patients with elevated triglycerides and CV disease or diabetes plus a risk factor. 4 g/day icosapent ethyl (essentially pure EPA) versus mineral oil placebo. Result: 25 percent relative reduction in the primary composite (CV death, nonfatal MI, nonfatal stroke, revascularization, unstable angina). One of the largest CV risk reductions ever recorded with a non-statin lipid-directed therapy.

STRENGTH (Nicholls 2020). 13,078 patients, essentially the same population. 4 g/day mixed EPA plus DHA carboxylic acid versus corn oil placebo. No reduction in major adverse CV events. Stopped early for futility.

Same daily dose. Same disease population. Opposite results.

The debate is not fully settled. Two possibilities dominate: the mineral oil placebo in REDUCE-IT may have raised LDL and inflammatory markers in the placebo arm, exaggerating the treatment-arm benefit; or DHA may blunt the mechanism by which EPA reduces cardiovascular events. The clinical read is that the EPA-to-DHA ratio matters, and EPA-dominated formulations appear to have a cleaner cardiovascular signal.

Depression: another EPA-favoring signal

Sublette 2011 published a meta-analysis of omega-3 trials in depression. The core finding: trials with EPA to DHA ratios greater than 2 to 1 produced meaningfully positive antidepressant signals; trials with ratios below that threshold, including DHA-dominant formulations, did not.

Later meta-analyses (Mocking 2016, Liao 2019) replicated the pattern. Effect size is modest — comparable to curcumin or exercise in mild depression — but the mechanism appears specifically EPA-driven. Not a replacement for standard depression care. A defensible adjunct in mild-to-moderate depression when the formulation is right.

Cognitive aging and neurodevelopment: where DHA shows up

The DHA story is stronger on the structural side — pregnancy, infancy, and cognitive aging.

Pregnancy and infancy. DHA accumulates rapidly in the fetal brain and retina during the third trimester and continues through the first two years of life. Multiple trials of prenatal DHA supplementation (usually 200 to 400 mg per day) have shown modest improvements in gestational duration, visual acuity development, and some cognitive markers in offspring. Effect sizes are modest but consistent.

Cognitive aging. Observational data associate higher blood DHA with lower dementia risk. Randomized data are more mixed — the MIDAS trial (Yurko-Mauro 2010) showed modest memory improvements at 900 mg per day of DHA in older adults with age-related cognitive decline. Effect sizes are small.

These are places DHA-dominant formulations are appropriate.

Hypertriglyceridemia: the FDA-recognized use case

The one area with clean FDA drug approval is severe hypertriglyceridemia. At 4 g/day of EPA plus DHA or 4 g/day of EPA-only formulations, triglyceride reductions of 20 to 50 percent are consistent across trials. This is drug-level dosing, requires prescription formulations in the US (Vascepa, Lovaza, Omtryg), and targets severe hypertriglyceridemia rather than general lipid optimization. The OTC 1 to 2 gram dose does not produce the same triglyceride effect — different endpoint.

Dose targets for regular supplementation

Working ranges based on the trial evidence for the common indications:

• General cardiovascular and metabolic support. Total EPA plus DHA of 1 to 2 grams per day. Prefer EPA-dominant formulations if cardiovascular endpoints are the goal. Serious cardiovascular indications should be discussed with a physician who can consider prescription EPA.
• Depression adjunct. 1 to 2 grams per day of EPA specifically, or a formulation with an EPA-to-DHA ratio above 2 to 1.
• Pregnancy and lactation. 200 to 400 mg per day of DHA. Prenatal-focused formulations tend to be DHA-heavy for this reason.
• Cognitive aging. 500 to 900 mg per day of DHA, or a balanced formulation. Modest expected effect.
• Severe hypertriglyceridemia. 4 grams per day, prescription route.

Note what these numbers refer to. The 1 to 2 gram number is EPA plus DHA — the sum of the two active fatty acids. Not the total fish oil weight, which is much larger. A 1000 mg fish oil softgel typically contains 300 to 400 mg of EPA plus DHA combined. To hit 1500 mg of EPA plus DHA, you need three to five typical softgels — or a single softgel of a concentrated product that delivers 500 to 1000 mg of active per capsule.

The label decoder is essential here. "Fish oil 1200 mg" tells you the total oil weight. Only the EPA and DHA numbers matter for dosing.

Vegetarian and vegan options

The vegan omega-3 story splits into two parts.

ALA from flax, chia, and walnuts. Alpha-linolenic acid is a short-chain omega-3. Humans convert ALA to EPA at roughly 5 percent efficiency and to DHA at less than 1 percent efficiency. Conversion is worse in men and in high linoleic acid diets. Practically, ALA supplementation does not produce EPA and DHA levels comparable to fish oil supplementation. Cardiac trial data with ALA supplementation are mixed and mostly weak.

Algae oil. The vegan solution. Marine algae are the original producers of both EPA and DHA — fish concentrate them by eating the algae. Algae-based supplements typically deliver DHA at higher concentrations than EPA, though EPA-forward algae products are increasingly available. Dose per dollar is higher than fish oil, and heavy-metal contamination is essentially zero because algae are cultivated in closed systems.

For strict vegetarians and vegans, algae oil is the honest recommendation. Do not rely on flax to deliver adequate EPA and DHA — the biochemistry does not cooperate.

The rancidity problem no one talks about enough

Fish oil is highly unsaturated. Multiple double bonds mean high susceptibility to lipid peroxidation — oxidation of the fatty acid molecules into aldehydes and other oxidation products. Rancid fish oil at a minimum tastes bad. It may also deliver oxidized lipids that undo some of the health benefit of the underlying EPA and DHA.

Independent testing has repeatedly found substantial fractions of the market exceeding recommended oxidation limits. Albert 2015 in New Zealand found more than 80 percent of tested products exceeded international voluntary standards on one or more oxidation markers.

What raises oxidation risk:

• Time on shelf. Oxidation accumulates. Fresh oil is less oxidized than year-old oil.
• Heat exposure. A product that spent months in a warm warehouse or shipping container is more oxidized than one kept cold.
• Light exposure. Clear bottles let light drive oxidation. Amber or opaque bottles protect the oil.
• Oxygen exposure. Bottle openings and softgel manufacturing quality matter.
• Lack of antioxidants. Vitamin E (tocopherols) or rosemary extract in the formulation slows oxidation.

What to look for on a label: recent bottling date, third-party oxidation testing (IFOS gives an oxidation score), added tocopherol antioxidants, opaque packaging, refrigeration recommendation for larger bottles.

What to do if you open a bottle: smell it. Fresh fish oil has a mild ocean smell. Strongly fishy or rancid smell means the oil is already oxidized. Return it.

Sourcing: small fish, third-party purity

Heavy metal and PCB contamination is the other purity issue with fish oil. Large predatory fish (tuna, swordfish, mackerel, shark) concentrate methylmercury and other contaminants through biomagnification. Small oily fish (sardines, anchovies, herring, menhaden) sit lower on the food chain and are much less contaminated.

Almost all serious fish oil brands source from small fish for this reason. The distillation and refining process also removes most contaminants when done properly. What matters at the buyer end is third-party testing that confirms heavy metals, PCBs, and dioxins are below detection or below regulatory limits.

Two third-party programs worth recognizing:

• IFOS (International Fish Oil Standards). Star ratings for CoQ purity, oxidation, and heavy metals. Publicly searchable by lot number. A five-star IFOS score is a legitimate quality signal.
• GOED (Global Organization for EPA and DHA Omega-3). Industry group with voluntary standards. Membership is a lower bar than IFOS certification.

Krill oil versus fish oil

Krill oil delivers EPA and DHA in the phospholipid form rather than the triglyceride or ethyl ester form of most fish oils. Head-to-head trial data are mixed. The absorption advantage per gram of active may exist but is smaller than marketing suggests, and krill oil costs several times more per gram than comparable fish oil. Not obviously superior for cardiovascular or depression endpoints.

An honest fish oil checklist

1. Read the EPA and DHA numbers, not the fish oil weight. Add them. That is your active dose.
2. Match the ratio to the goal. Cardiovascular or depression: prefer EPA-dominant. Cognitive aging or pregnancy: DHA-dominant is fine.
3. Verify third-party testing. IFOS certification or equivalent. Heavy metals and oxidation should both be reported.
4. Check the bottling date. Fresher is better. Avoid deep-discount bulk buys.
5. Take with a fat-containing meal. Omega-3s are fat-soluble.
6. Refrigerate after opening for larger bottles. Slows oxidation once the seal is broken.
7. Consider algae oil if vegetarian. Do not rely on flax.

Why supplement scores in this category cluster Provisional

Fish oil products score Provisional under the current Vyvata rubric for the same industry-wide reasons that apply across the supplement category. Even brands that submit to IFOS testing do not always publish the results in a form that is easy for a buyer to verify. Brands that publish detailed COAs with heavy metal, oxidation, and specific EPA and DHA content per capsule (Nordic Naturals, Carlson, Wiley's Finest, Thorne) would score higher on a fish-oil-specific transparency upgrade. Until that upgrade lands, most fish oil products cluster in the same Provisional band as the rest of the supplement catalog.

The honest summary

EPA and DHA are not interchangeable. EPA carries the anti-inflammatory and cardiovascular load and appears to be the specific driver of the depression signal. DHA is a structural lipid for brain and retina, and it is the version that matters most in pregnancy, infancy, and cognitive aging. The right formulation depends on what you are trying to do.

For most healthy adults aiming at general cardiovascular and inflammatory endpoints, an EPA-dominant fish oil at 1 to 2 grams total EPA plus DHA per day, from a brand with third-party purity and oxidation testing, taken with a fatty meal, is a defensible baseline. For pregnancy, prefer DHA. For serious hypertriglyceridemia, prefer prescription EPA under a physician's care.

Read the EPA and DHA numbers on the label. Smell the oil when you open it. Do not accept fish oil that has already gone rancid on the shelf. That is more of the game than any of the marketing suggests.