At first, it was all very exciting. In 1971, a team of Danish researchers stationed on Greenland’s northwest coast found that a local Inuit community had remarkably low levels of diabetes and heart disease. The reason, the researchers surmised, was their high-marine-fat diet—in other words, fish oil. Incidence of heart disease, which once afflicted relatively few Americans, had shot up since the turn of the century, and here, seemingly, was a simple solution. “I remember how exciting those studies were when they first came out,” Marion Nestle, a professor emerita of nutrition and food studies at NYU, told me. “The idea that there were populations of people who were eating fish and were protected against heart disease looked fabulous.”

The hype didn’t stop with heart disease. Soon, fish oil was being hailed as a panacea. It could help prevent dementia! Depression! Obesity! Cancer! News stories and books parroted these claims. And supplement makers capitalized. By 2014, fish-oil supplements were a billion-dollar industry. Today, the market continues to grow at an astronomical rate. The growth of the science supporting fish oil’s curative properties, meanwhile, has been, shall we say, less astronomical. The early papers that sparked the initial enthusiasm were merely observational, meaning that they could establish only correlation, not causation. When the randomized control trials eventually began to trickle in, the results were mixed at best.

Tens of thousands of studies later, things haven’t gotten all that much clearer: We still don’t have anything close to a firm grasp of what fish oil can do and what it cannot. And lately, things have only gotten weirder.

Most experts acknowledge that fish oil does have some modest benefits in certain circumstances. Omega-3, its star nutrient, has been shown to lower levels of a fat associated with heart failure, help prevent premature births, and improve infant formulas. But these are a far cry from the game-changing promise of the early studies. That promise, over the years, has gotten lost in a tangle of theoretical possibilities, Nestle told me. Fish oil contains two distinct types of Omega-3, DHA and EPA; maybe only the former is providing the benefit. Or maybe only the latter. Maybe the benefit comes only from pairing the two. Maybe neither does anything unless it’s consumed with other parts of the actual fish.

And that’s just the beginning. Maybe the benefits have less to do with fish itself and more to do with the fact that if you’re eating fish, you’re probably not also eating a hamburger or a pork chop. Maybe they have to do with your overall diet. Maybe they don’t have to do with your diet at all. Maybe it’s just that fish eaters tend to be wealthier and, not unrelatedly, healthier in the first place. Maybe it’s something else entirely.

Read: The fishy science of omega-3s

Through much of the 2010s, one fish-oil study after another came up empty, Richard Bazinet, a nutrition researcher at the University of Toronto, told me—“null, null, null, null, null.” And then came REDUCE-IT, a trial funded by the pharmaceutical company Amarin to test its fish-oil-based heart drug, called Vascepa. The results, presented in 2018, found that, among high-risk adults already receiving another type of cholesterol-lowering treatment, the drug decreased the risk of heart failure and other serious cardiovascular events by an eye-popping 25 percent. Fish oil, it seemed, was back in business. When the study’s lead author, the Harvard cardiologist Deepak Bhatt, presented his findings at the American Heart Association’s annual meeting in Chicago, the crowd gave a standing ovation. The following year, the FDA approved the drug for the use studied in REDUCE-IT. (The agency had already approved the drug for a different use back in 2013.)

With triumph, though, came controversy. Even at the time of Bhatt’s presentation, some cardiologists noted that the study’s mineral-oil-based placebo—a pill selected because its color and consistency mimic those of fish oil, but whose use in fish-oil studies has been debated—seemed not to be entirely neutral. In fact, the placebo seemed to be harming people. Initially, nothing much came of these concerns. Then, last month, a new analysis published in the journal Circulation substantiated them and then some. It showed, based on elevated levels of several biomarkers in blood-test results, that the placebo may have increased volunteers’ risk of heart attack and stroke. Many researchers found these results to be compelling evidence that Vascepa’s eye-popping success could be due to a bad placebo, not a great drug.

“What’s somewhat shocking about that paper is that it looks like everything got worse in the placebo group and the treatment group stayed the same,” Bazinet told me. “You could have given the subjects a glass of water. Anything would have been better against that placebo.” Steven Nissen, a cardiologist at the Cleveland Clinic who was involved in a different Omega-3 trial, called the Circulation study’s findings “extraordinarily disturbing.” Two members of the expert panel that in 2019 recommended that the FDA green-light Vascepa even told Stat’s Matthew Herper that, if they’d had access to the new data at the time, they might not have voted to approve.

To make matters more confusing, the Circulation study—as in, the very study that ignited this controversy—was also funded by Amarin, and one of the study’s 13 authors was Bhatt, the lead author on REDUCE-IT. In a statement, Amarin told me it “continues to stand by the results of REDUCE-IT” and is “very surprised” that the panel members would make such comments based on the Circulation paper. The company stressed that REDUCE-IT’s positive results “could not be explained” by the placebo, and that the effects found in the Circulation study were too minor to “correlate to any meaningful changes in outcomes.” Bhatt agreed, telling me he sees the new paper not as undermining REDUCE-IT but simply as clarifying Vascepa’s biological mechanisms. He defended the use of mineral oil as a placebo, arguing that it alone could not explain the significant risk reductions observed in the trial.

The lead author of the Circulation study, Paul Ridker, declined to comment on the controversial results. But other experts I spoke with were considerably less sanguine than Bhatt. Several would say only that, at this point, the REDUCE-IT results are basically uninterpretable. Nissen, who has in the past called REDUCE-IT “almost certainly a false-positive study,” went so far as to say that he thinks the benefits it found can be “entirely explained by the harms of the placebo” and that Amarin should have known not to use mineral oil. JoAnn Manson, the chief of preventive medicine at Brigham and Women’s Hospital in Boston and the leader of the largest-ever study of vitamin D and Omega-3 pills in healthy adults, was more sympathetic to the idea that the Circulation study’s findings likely don’t account for the full 25 percent risk reduction. But she also raised the possibility that the Vascepa, if ineffective, could be dangerous: Some studies have shown that a high daily dosage of fish oil can heighten one’s risk of developing a type of irregular heartbeat. (Amarin called the suggestion that Vascepa could be ineffective and dangerous “a gross distortion of fact,” saying that “the findings of independent, thorough, and impartial scientific and statistical reviews” had determined that the drug’s benefits to the at-risk patients for whom it is designed more than make up for its risks.)

The upshot of all this is that an already murky situation has become a good deal murkier, and there’s no end to the murk in sight. Which is a shame because, in one sense at least, the stakes are higher now than they’ve been in some time: REDUCE-IT suggested that Vascepa could legitimately save lives. If it can’t, that’s more than a scientific scandal; it’s a real, human loss. “I’ve never seen anything like this,” Bazinet told me. “In a way, it’s not surprising. The field’s been controversial all the time, and now we probably have the biggest controversy.”

The only way out of this mess, experts said, is to run a whole new trial comparing Vascepa (or its generic equivalent, icosapent ethyl) with something everyone agrees is a true placebo—one that we can be confident doesn’t harm people. Manson is leading a team applying for NIH funding to run such a study. (She said that Amarin told her it was not open to a replication trial and that the company declined to fund three related studies. When I asked Amarin about this, the company told me it would not replicate REDUCE-IT, because the outcomes “read out robustly,” and that it does not publicly discuss research proposals from third parties.) The study would also investigate a pair of promising leads turned up by her own major study, an ongoing project that has found that although Omega-3 did very little for the population as a whole, it might have considerable benefits for Black people and people who don’t eat much fish.

In the meantime, doctors are unlikely to ditch Vascepa, Clifford Rosen, a professor at Tufts University School of Medicine, told me. In the first quarter of 2022, Amarin sold nearly $100 million worth of the drug, which is its only product. “There’s such momentum to use this agent that until the next study comes around, I think there’s still going to be widespread use,” Rosen said. To his point: In 2019, the American Diabetes Association recommended icosapent ethyl for certain patients as part of its official standards of care, based explicitly on the REDUCE-IT results. Since the publication of the Circulation paper, the ADA has made no move to withdraw that recommendation. (When I asked whether the group is considering doing so, its chief scientific and medical officer said only that it had “followed the evidence based on what was available at the time.”)

Not that this state of affairs is particularly novel. We’ve known for years that fish-oil supplements have virtually no benefits for your average, healthy person, Pieter Cohen, a professor at Harvard Medical School, told me. That hasn’t stopped tens of millions of Americans from popping the pills every day. “People just love to take supplements,” Rosen said. “It’s religiosity … It’s magical thinking.” Vascepa is an FDA-approved drug, not merely a supplement, but in some ways the line isn’t all that clear. The dosage is certainly higher, the packaging is certainly better, and the regulations are certainly stricter. But if you don’t understand the biological mechanism behind either the drug or the supplement—and scientists do not—that makes it tough to assert with any confidence that they’re fundamentally distinct.

“If you don’t know how something works—like you have no idea how it works—it’s hard to say that they’re different!” Bazinet told me. “Because it could just be a little bit more of the same mechanism. It’s not clear.” When it comes to fish oil, very little is.