Mayo Clinic Omega-3 Study

As Seen On...

The Mayo Clinic Omega-3 Study

Mayo Clinic conducted a meta-analysis of randomized controlled trials (RCTs) testing the effects of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) on coronary heart disease (CHD). The goal was to evaluate improvement in patients taking these 2 Omega-3 fatty acids. This sounds like it should be a good study. However, not everything is as it seems so lets dig in to this one and see if it pans out as they stated.

The title of the study was “A Meta-Analysis of Randomized Controlled Trials and Prospective Cohort Studies of Eicosapentaenoic and Docosahexaenoic Long-Chain Omega-3 Fatty Acids and Coronary Heart Disease Risk.”

The investigators searched medical literature databases including Ovid/Medline, PubMed, Embase, and the Cochrane Library for studies regarding EPA & DHA, either from food or from supplements associated with cardiovascular disease (myocardial infarction, sudden cardiac death, coronary death, and angina). They looked at studies from 1947 to 2015 and identified 18 randomized controlled trials (RCTs) & 16 prospective cohort studies.

They wanted to evaluate the association of Omega-3 intake with coronary heart disease (CHD). They also wanted to evaluate the effect of the dose of Omega-3s as well as the effect of EPA & DHA on specific outcomes such as heart attacks (myocardial infarction). Each identified study was required to report at least one of the following: myocardial infarction (heart attack – fatal or non-fatal), chest pain (angina), sudden cardiac death, & CHD incidence. The studies had to include outpatient adults at least 18 years of age with or without CHD. They grouped the studies based on Omega-3 intake levels of below 1 gram per day or above 1 gram per day.

There were 93,000 participants in the RCT data and 732,000 participants in the prospective cohort data. There were 2 trials for primary prevention (preventing a first CHD event) with 27,494 participants. There were 9 trials focusing on mixed prevention and 12 trials focusing on secondary prevention.

One of the problems with meta-analysis as a study type is that your conclusions are only as good as the studies in the meta-analysis. The problem here is that there is a very wide range of Omega-3 dosing (0.38 grams to 23.8 grams per day for the primary and mixed prevention groups). It is likely that beneficial effects of Omega-3s were not noticed in the trial where they only took 380mg of EPA+DHA (0.38 grams) daily. Conversely, 23.8 grams per day is quite a lot.

The authors of the trial did observe that higher doses of EPA+DHA had a “stronger impact” in those with elevated triglyceride levels compared to trials where the participants took less than 1 gram per day of EPA+DHA.

Prospective cohort trials did demonstrate statistically significant reductions in CHD. The number is the percentage reduction for that event in patients taking Omega-3s.

  • Any CHD event – 18%
  • Fatal CHD events – 23%
  • Coronary death – 18%
  • Sudden cardiac death – 47%

They noted that Omega-3s reduced triglycerides as well as reduced risk of CHD in patients with LDL cholesterol >130. They felt that the benefit in regards to LDL may have been due to the change in LDL to the larger LDL particle which is less atherogenic. They also stated that “Blood pressure is another well-documented CHD risk factor impacted favorably by n-3 LCPUFA administration.”


Their conclusion?

“Results indicate that EPA+DHA may be associated with reducing CHD risk, with a greater benefit observed among higher-risk populations in RCTs.” However, I believe that further review of their data demonstrates that their conclusion statement is understating the effect of Omega-3s.

For example, one of the studies was a food recall questionnaire asking participants how many servings of fish they ate on a weekly basis. The investigators then assigned an Omega-3 quantity based on what type of fish meat it was (dark meat fish, canned tuna, and other fish). One confounding factor noted was that there was an inverse relationship of fish intake and red/processed meats. In other words, those that ate more fish ate less red/processed meats. It should be noted that there is a huge potential for bias in a food/diet recall questionnaire. Regardless, the investigators stated “In this prospective cohort study of diabetic women, higher consumption of fish and ω-3 fatty acids was associated with a lower incidence of both CHD and total mortality, even after adjustment for established cardiovascular risk factors.”2

Here are some of the studies evaluated in the Mayo Clinic Omega-3 Study

The DART Trial3

Another trial was merely based on whether or not they received dietary recommendations. 2033 male participants who had recovered from a heart attack (myocardial infarction) were randomized to either receive or not receive dietary advice. If they did receive dietary advice it was in regards to one of 3 groups: reduced fat intake (and increase in ratio of polyunsaturated fat to saturated fat), increase the amount of dietary fatty fish, or increase the amount of cereal fiber. The investigators noted that “The subjects advised to eat fatty fish had a 29% reduction in 2 year all-cause mortality compared with those not so advised.”3 I have several issues with relying on this study. There may be a good correlation with dietary advice regarding fish intake and coronary heart disease but there is no way to confirm how much fish they ate, their Omega-3 Index levels, quality of fish, etc.

GISSI-Prevenzione Trial4

This trial consisted of 11,324 participants who had survived a recent heart attack. They were randomized to receive Vitamin E (alpha-tocopherol, 300mg daily), Omega-3 supplements (1 gram daily), both of these, or neither of these for 3-5 years. The investigators wanted to evaluate if these supplements made a difference in death, non-fatal heart attack (myocardial infarction), or stroke.
They discovered that vitamin E supplementation did not make a difference. However, Omega-3 supplementation did with a relative risk reduction of 15% by 4-way analysis. The results for the group taking Vitamin E and Omega-3 combined were similar to those taking Omega-3s alone.

The ORIGIN Trial5

This study is very interesting because they showed that “Daily supplementation with 1 g of n–3 fatty acids did not reduce the rate of cardiovascular events in patients at high risk for cardiovascular events.” 12,536 participants were randomized to receive 900mg of ethyl ester Omega-3 (EPA 465 mg and DHA 375 mg) or a placebo. The participants had impaired fasting glucose, impaired glucose tolerance, or diabetes. They also received ither insulin glargine or standard care. Their primary outcome measurement was death from cardiovascular disease.

Why did this study not show improvement with Omega-3s when so many other studies do show improvement? The answer is simple. Their placebo was 1 gram of olive oil which is well known to have cardiovascular disease protective benefits!

The JELIS Trial6

There were 18,645 Japanese participants with a total cholesterol level of 251 mg/dL or more who were randomized to receive a statin only or a combination of a statin plus EPA 1800mg. Primary endpoints were any major adverse coronary events (sudden cardiac death, heart attacks, unstable angina, angioplasty, stents, or coronary artery bypass grafting [CABG]).

There was a 19% reduction in these Major Adverse Coronary Events (MACE) in the EPA + statin group at the end of 4-6 years. This study shows that patients have lower risk of cardiovascular disease outcomes if they are on EPA in addition to a statin.

The OMEGA Trial7

The OMEGA trial was a multicenter trial in Germany and consisted of 3,851 participants. It was a randomized, placebo-controlled, double-blinded (RCDBT) trial. Participants were randomized between 3 & 14 days after a heart attack to receive either 1 g omega-3 ethyl esters (EPA 460 mg & DHA 380 mg) or a soft gelatin capsule containing 1 g olive oil. This is yet another trial using olive oil as a placebo inappropriately. Therefore, they didn’t notice any difference between the 2 groups.





  1. Mayo Clin Proc. 2017 Jan;92(1):15-29
  2. Circulation. 2003;107:1852–1857
  3. Lancet. 1989;334(8666):757-761
  4. Lancet 1999;354(9177):447-455
  5. N Engl J Med 2012; 367:309-318
  6.  Lancet. 2007;369(9567):1090-1098
  7. Circulation. 2010;122:2152–2159