I have a confession to make. For at least the last ten years, I’ve taken an omega-3 supplement every day. Religiously. Why? I literally have no idea. I don’t remember exactly when or why I started taking it. Presumably I read somewhere that it was a good idea. This was before I started studying medicine, and the decision certainly wasn’t based on any thorough evaluation of the scientific evidence. So I figure it’s probably about time I actually take a look at the evidence, before I decide whether to continue spending hundreds of dollars a year on omega-3 supplements.
The most commonly claimed benefit when it comes to omega-3 is that it prevents heart disease. This dates back to the 1970’s, when it was noted that inuit following a traditional diet rich in meat from whales and seals suffered from remarkably little cardiovascular disease. It was suggested that this might be due to the high levels of omega-3 fatty acids in meat from marine mammals. Of course, correlation is not causation, and I can think of several other major differences between the traditional inuit diet and the standard western diet that could explain the lack of heart disease, such as the complete absence of sugar.
Since then, there have been many randomized trials looking specifically at omega-3 for the prevention of heart disease. In 2018 the Cochrane collaboration carried out a systematic review and meta-analysis, with the goal of answering the question definitively. The review included both randomized controlled trials of omega-3 supplementation and of advice to eat more fatty fish. The requirements for inclusion in the systematic review were that the trials follow participants for at least one year, and provide data on mortality or outcomes related to heart disease.
79 randomized trials were identified that fulfilled these criteria. The total number of participants in the 79 trials was 112,059, which is a nice big number that should allow some pretty firm conclusions to be drawn. 33 of the trials were looking at omega-3 for secondary prevention (i.e. to prevent people who already have heart disease from progressing to more severe disease or death), while the remaining 46 were looking at omega-3 for primary prevention (i.e. to prevent healthy people from developing heart disease in the first place).
71 of the studies were investigating long-chain omega-3 (the kind found in fish), while eight were looking at short-chain omega-3 (the kind found in walnuts and flax seeds).
63 of the studies were of omega-3 supplements, eight were of omega-3 enriched foods (such as for example breakfast spreads), and four were of advice to eat more foods rich in omega-3. In other words, the overwhelming majority of the studies were of fish oil supplements, and that is where the strongest conclusions can be drawn from the results. In the supplement trials, the control group usually received a “placebo” supplement containing some other type of oil (for example olive oil, sunflower oil, or corn oil).
Ok, let’s get to the results.
We’ll start with cardiovascular events, and then move on to overall mortality (the researchers also looked at cardiovascular mortality, but that is a nonsense metric – people care about whether they’re alive or dead, not what cause of death is listed on their death certificate – so I’m not going to bother wasting my time or yours discussing it).
38 of the trials reported on cardiovascular events (for example heart attacks, strokes, and revascularisations). In total, 14,737 participants experienced a cardiovascular event in these trials. That is plenty, so if there is an effect of omega-3, it should be possible to see. Unfortunately, increased intake of omega-3 was not associated in any meaningful reduction in cardiovascular events. There was a tiny 1% reduction in the relative risk of experiencing a cardiovascular event that was nowhere close to being statistically significant.
The lack of effect remained when only studies at low risk of bias were included. Furthermore, there was no sign of a dose-response effect. In other words, no benefit was seen when the higher dose studies were analyzed separately. In fact, the studies giving the highest omega-3 doses had the worst results of all, showing a 9% increased risk of cardiac events (although as with the lower doses the effect wasn’t statistically significant).
Nor was any benefit seen when looking only at studies of secondary prevention, i.e. of people who had already experienced a cardiovascular event, and who should be most likely to benefit from a treatment that decreases the risk of new events. Nor was there any benefit seen when longer term studies (i.e. with more than four years of follow-up) were analyzed separately. Basically, there was no signal that omega-3 had any beneficial effect on heart disease risk whatsoever.
When looking specifically at heart attacks (as opposed to cardiovascular events more generally), the researchers initially found a small 5% reduction in relative risk. However, when they only included studies at low risk of bias, this instead transformed in to a 3% increase in relative risk. I should point out here that we’re talking relative risks, not absolute risks. If we were to talk absolute risks instead, then the 5% decrease in relative risk of a heart attack, even if real, would actually only be a 0.065% decrease in absolute risk, taking you from a 3.80% risk of having a heart attack down to a 3.73% risk – In other words so marginal as to not be remotely worth bothering about even if the results had been statistically significant, which they weren’t. If you’re confused about the difference between relative risk and absolute risk, or the difference between statistical significance and clinical significance, then I recommend you read this.
Ok, so omega-3 doesn’t appear to be useful for preventing heart disease. But it might have other effects that result in an improvement in overall longevity. There are, for instance, those who claim that omega-3 supplements can prevent cancer. If that is the case, then there should be some improvement in overall mortality. Let’s check.
8,189 people died during the follow-up period in the studies that provided mortality data. As with the cardiovascular events, that is plenty of data. Unfortunately, as with the cardiovascular events, increased intake of omega-3 was not associated with any meaningful reduction in overall mortality over the course of follow-up. To be more precise, omega-3 was associated with an extremely marginal 2% reduction in the relative risk of death. As before, the reduction wasn’t anywhere near being statistically significant.
Apart from that, the funnel plot (a type of graph used to look for publication bias – i.e. the fact that researchers are more likely to publish studies that show benefit) suggested that there were some negative studies “missing”. If those were included, the marginal and statistically insignificant reduction in deaths would likely disappear completely.
It isn’t suprising that some studies are missing. Many of the studies of fish oil supplements and omega-3 enriched foods have been produced by companies that sell fish oil supplements and omega-3 enriched foods. It’s likely that they’ve done their best to hide away studies that failed to show benefit.
The lack of effect on mortality remained when the reviewers looked separately at the studies of omega-3 supplements, the studies of fortified foods, and the studies of dietary advice. It also remained when studies were separated based on whether they were looking at primary prevention or secondary prevention. And it remained when studies with more than four years of follow-up were analyzed separately. In other words, there was no signal that an effect started to appear with time.
The lack of effect even remained when the reviewers separated out the studies by dosage, and only looked at the higher dose studies. There was no suggestion of a dose related effect whatsoever, with the studies giving a relatively high 2.4-4.4 grams of omega-3 per day (equivalent to three to five concentrated fish oil capsules) even showing a marginally increased risk of death.
What can we conclude?
Increasing intake of omega-3 does not protect against heart disease. More importantly, it doesn’t appear to result in any improvement in longevity whatsoever. With that being the case, I’m going save myself some money and stop taking omega-3 supplements, at least until I see some real evidence of benefit.