Fish Oil for Heart Health: The Winning Trial Used a Drug

The statins article ended on the idea that a benefit is worth different amounts to different people. This one is about a different problem, and it is the last article in this section: what happens when the evidence everyone cites turns out to be about a product nobody is buying.

Fish oil is the most sold heart supplement there is, and the marketing is not inventing its science from nothing. There are large, serious, well-run randomised trials behind the claims. The trouble is which trials say what, and what exactly was in the capsule in each one.

We earn nothing from any of this. There is no affiliate link on this page and no brand being recommended, which is the whole reason a site like this can go through the omega-3 evidence without needing it to land somewhere particular.

What fish oil is, and what the label promises

Fish oil is oil pressed from oily fish, and the part everyone is interested in is two omega-3 fatty acids: EPA and DHA. MedlinePlus describes omega-3s as "a type of polyunsaturated fat" that the body uses "to build brain cells and for other important functions."

The heart claims on the shelf are downstream of a real, if mixed, literature. MedlinePlus lists several mechanisms attributed to omega-3s: they "reduce triglycerides," "reduce the risk of developing an irregular heartbeat," "slow the buildup of plaque" and "help to slightly lower your blood pressure."

None of that is in dispute here. The question this article is about is narrower and more practical: when a healthy-ish man over 40 takes the standard 1-gram capsule every morning, do the randomised trials show fewer heart attacks and strokes? That is a question with an answer.

What the bottle dose actually did in the trials

Two very large primary-prevention trials tested close to the dose sold over the counter, and both reported in the same journal within a year of each other.

ASCEND asked the same question in people at higher risk. It randomised 15,480 patients with diabetes but no established cardiovascular disease to "1-g capsules containing either n-3 fatty acids (fatty acid group) or matching placebo (olive oil) daily," and followed them for a mean of 7.4 years. A serious vascular event occurred in 8.9% of the fatty acid group and 9.2% of the placebo group — "rate ratio, 0.97; 95% confidence interval [CI], 0.87 to 1.08; P=0.55."

Forty thousand people between them, years of follow-up, and the primary endpoints landed on nothing. Neither trial is underpowered in the usual sense. But neither excludes a modest benefit either: VITAL's interval runs from a 20% reduction to a 6% increase, ASCEND's from a 13% reduction to an 8% increase. What they rule out is a large effect at this dose.

The one signal inside the null result

It would be easy, and wrong, to stop there. VITAL reported a list of secondary endpoints, and one of them moved: for total myocardial infarction the hazard ratio was 0.72 (95% CI 0.59 to 0.90), while total stroke was 1.04 and death from cardiovascular causes was 0.96.

That heart attack number is not nothing, and honest writing has to include it. It is also not a conclusion. A secondary endpoint in a trial whose primary endpoint was null is a lead for the next trial, not a finding the next trial can be skipped over. Trials report many secondary outcomes, and some move by chance; the primary endpoint is nominated in advance precisely to stop everyone from picking the winner afterwards.

So the fair statement is the awkward one: at 1 gram a day, the composite of heart attack, stroke and cardiovascular death did not fall, and inside that null result sits an unexplained signal on heart attacks alone that nobody has yet confirmed in a trial designed to test it.

The trial that did work, and what was in it

REDUCE-IT is the trial that fish oil marketing leans on, and its result is genuinely strong. It is also not a trial of fish oil.

Four differences from the capsule on the shelf, and they all matter. The dose was four times higher. The substance was purified EPA rather than the EPA-and-DHA mixture in fish oil. It is a prescription medicine, not a supplement. And the population was not healthy men hedging their bets — it was people already on a statin, with established cardiovascular disease or diabetes plus risk factors, and raised triglycerides.

The trial also reported costs. "A larger percentage of patients in the icosapent ethyl group than in the placebo group were hospitalized for atrial fibrillation or flutter (3.1% vs. 2.1%, P=0.004)," and serious bleeding events were 2.7% versus 2.1% (P=0.06). Anyone quoting the 0.75 should be quoting those too.

Why "just take four capsules" does not follow

The obvious inference from REDUCE-IT is that the supplement trials simply used too little. There is a trial that tests exactly that inference, and it does not support it.

STRENGTH randomised 13,078 statin-treated patients at high cardiovascular risk, with raised triglycerides and low HDL, to "4 g/d of omega-3 CA" — a carboxylic acid formulation of EPA and DHA — or corn oil, "which was intended to serve as an inert comparator." The trial was "prematurely halted based on an interim analysis that indicated a low probability of clinical benefit of omega-3 CA vs the corn oil comparator" — a stop that came with 1,384 of a planned 1,600 primary events recorded. The primary endpoint occurred in 12.0% versus 12.2%: "hazard ratio, 0.99 [95% CI, 0.90-1.09]; P = .84." Gastrointestinal adverse events were 24.7% versus 14.7%.

Same 4 grams a day, and one trial found a benefit while the other found none. Note that a null is not an opposite: STRENGTH did not find harm, it found "no significant difference," with a confidence interval of 0.90 to 1.09.

The two populations were also not the same, which is worth spelling out rather than waving at. Both enrolled statin-treated patients at high cardiovascular risk with raised triglycerides. But STRENGTH additionally required "low levels of high-density lipoprotein cholesterol," which REDUCE-IT did not; REDUCE-IT capped LDL cholesterol at "41 to 100 mg per deciliter," where STRENGTH states no LDL entry criterion; and REDUCE-IT reports that "70.7%" of its patients were enrolled "for secondary prevention of cardiovascular events," a figure STRENGTH does not give at all.

So the number of grams is not what separates these two results, but the trials differ in more than one way at once and neither paper identifies the cause. Purified EPA versus an EPA-and-DHA mixture is one candidate. The comparators are another, though the reports do not let you compare them: STRENGTH names its corn oil, while the REDUCE-IT abstract describes its own only as "placebo."

That is why "take four of the cheap capsules instead of one" is not a conclusion these trials license. Nobody has run that trial. The closest thing to it, STRENGTH, came back null.

What about just eating the fish

The supplement trials tested a capsule. They did not test a diet, and the sources are notably more confident about the food than about the pill. MedlinePlus puts it plainly: "most health experts agree that the best way to reap the benefits of omega-3 is from food."

The same passage then makes an exception, and leaving it out would misrepresent the source. It continues: "if you already have heart disease or high triglycerides, you may benefit from consuming higher amounts of omega-3 fatty acids. It may be hard to get enough omega-3s through food. Ask your health care provider if taking fish oil supplements might be a good idea." That is a conditional recommendation for a specific group, not a general one — and it points at a conversation rather than a shelf.

The amount is specified. MedlinePlus reports that "the American Heart Association (AHA) recommends eating at least 2 servings a week of fish rich in omega-3s," naming salmon, mackerel, albacore tuna, trout and sardines. The FDA's fish page lands in the same place, though it is quoting someone else and addressing someone else: it reports that "the Dietary Guidelines for Americans recommends: at least 8 ounces of seafood (less for children) per week based on a 2,000 calorie diet," with a serving measured against the palm of your hand. That page is written "for those who might become or are pregnant or breastfeeding and children ages 1 - 11 years," so it is a useful figure rather than advice aimed at this article's readers.

Both sources attach a caution, and both scope it to someone other than this article's reader. MedlinePlus says "some fish can be tainted with mercury and other chemicals," then that "eating tainted fish can pose health risks for young children and pregnant women," and then — in the same short section — "if you are middle-aged or older, the benefits of eating fish outweigh any risks." That last sentence is addressed squarely at the men this site is written for, and quoting the warning without it would invert what the source actually says.

The FDA's avoid-list, written for the same pregnancy-and-children audience as the rest of that page, names king mackerel, marlin, orange roughy, shark, swordfish, Gulf of Mexico tilefish and bigeye tuna. Its framing sentence carries its own scope: "while it is important to limit mercury in the diets of those who are pregnant or breastfeeding and children, many types of fish are both nutritious and lower in mercury."

Worth being honest about the limits of this too: a recommendation to eat fish twice a week is not the same class of evidence as a randomised trial, and eating fish is bound up with everything else about the people who eat it. It is guidance, not a trial result. But it is the form of the advice that both sources actually give.

The honest limits of this article

First, "no benefit found at 1 gram a day for major cardiovascular events" is not the same statement as "fish oil does nothing." Omega-3s at higher doses do lower triglycerides, which is why prescription versions exist at all, and the VITAL heart attack signal is unresolved rather than dismissed. The trials answer the question they were built to answer, in the populations they enrolled, at the doses they used. Two design details belong here too: VITAL was "a randomized, placebo-controlled trial, with a two-by-two factorial design, of vitamin D3 (at a dose of 2000 IU per day) and marine n-3 fatty acids," and ASCEND reports an "adherence rate, 76%."

Second, there is a real tension on this page that is worth naming rather than smoothing over. MedlinePlus says omega-3s "help keep your heart healthy and protected against stroke." VITAL and ASCEND, testing supplements, found no reduction in cardiovascular events, and VITAL found a hazard ratio of 1.04 for total stroke. Those statements are not talking about the same thing: one is about omega-3s as nutrients, largely from food, and the other is about capsules in randomised trials. Keeping them apart is most of the work in reading this literature.

Third, none of this transfers to anyone already prescribed icosapent ethyl. That is a prescription decision made with a doctor, for a specific situation — established cardiovascular disease or diabetes with risk factors, raised triglycerides, already on a statin — and the honest reading of REDUCE-IT is that it worked in exactly that group.

What to do with this

1. Check which trial a claim is citing. If the number is 0.75 or "25% lower," it is REDUCE-IT — a prescription drug at 4 grams a day, not the bottle.
2. Do not read "no benefit" as "harmful." VITAL observed "no excess risks of bleeding or other serious adverse events." The supplement trials found little effect in either direction.
3. Do not self-escalate the dose. STRENGTH tested 4 grams a day of a different preparation and was halted early after an interim analysis showed a low probability of clinical benefit, with far more gastrointestinal side effects.
4. Take the food advice seriously. Two servings a week of oily fish per the AHA via MedlinePlus, and at least 8 ounces of seafood a week per the Dietary Guidelines as quoted by the FDA. MedlinePlus adds that for the middle-aged and older, "the benefits of eating fish outweigh any risks."
5. If triglycerides are the issue, that is a prescriber conversation. The prescription preparation exists for a defined situation, and that is where REDUCE-IT applies.

This closes the heart section, and it is a fitting place to end it. The whole cluster has been about the distance between a number and what it means for you — a blood pressure threshold that moved, a cholesterol target with no floor, a benefit that is a share of your own risk. Fish oil is the same problem in its plainest form: a real trial result, attached in the marketing to a product that trial never tested.