Abstract
Objective To explore the evidence for omega-3 fatty acid (O3FA) supplementation in primary and secondary prevention of cardiovascular disease (CVD).
Sources of information PubMed, Cochrane reviews, and Google Scholar were searched for meta-analyses and reviews related to O3FAs and CVD. Salient, recent randomized controlled trials referenced in these reviews were retrieved. Current lipid guidelines were reviewed.
Main message Most O3FAs are derived from marine or aquatic microalgae, which are consumed by fish. The essential fatty acids eicosapentaenoic acid and docosahexaenoic acid are mainly sourced from fish, with a small fraction coming from plants. Omega-3 fatty acids modestly lower triglyceride levels, but the major impact on CVD is through a variety of other mechanisms related to cell membrane function, antioxidant properties, and reduction of atherogenic small low-density lipoprotein cholesterol particles. Guidelines continue to recommend eating 2 servings of fish per week. There is little evidence of benefit of O3FAs in primary prevention of CVD. Given that 40% of Canadians have insufficient levels and that these low levels may be associated with other chronic diseases over time, supplementation with O3FAs could be considered, particularly in those with hypertriglyceridemia, in those who eat no fish, or for vegetarians or vegans. Doses up to 1 g daily are considered safe. For secondary prevention after statin optimization, if triglyceride levels are between 1.5 and 5.6 mmol/L, guidelines recommend with level 1A evidence taking 2 g of icosapent ethyl twice a day. This is also recommended in primary prevention for patients with diabetes and hypertriglyceridemia and additional CVD risk factors. As fish stocks dwindle over time, preserving fisheries for developing countries and obtaining O3FA from microalgal or genetically modified plant sources may become important.
Conclusion All guidelines recommend at least 2 servings of oily fish per week, although benefit from O3FAs is mostly seen in secondary prevention. Fish oil and combination preparations of eicosapentaenoic acid and docosahexaenoic acid have failed to show benefit at any dose at any level of prevention in patients who are appropriately prescribed statins. High-dose eicosapentaenoic acid shows substantial benefit in selected patients taking statins who have high triglyceride levels.
Our paleolithic ancestors first evolved as hunter-gatherers and opportunistic omnivores approximately 2.5 million years ago.1 In comparison with the modern Western diet they consumed more mono- and polyunsaturated fats, both omega-3 and omega-6, but less saturated fat and much less carbohydrates.1
Most Canadians do not consume fish at recommended levels. As a result, 40% of the population have omega-3 fatty acid (O3FA) blood levels lower than recommended, although it is important to note these data did not include people living in the 3 territories, on reserves, or in other Indigenous settlements.2 In concert with the rest of North America and most of Europe, our O3FA levels are among the lowest in the world, while areas around the Sea of Japan, Scandinavia, and regions with Indigenous populations who have not fully adopted a Westernized diet (including Indigenous populations in northern Canada) have high levels.3 Our generally low consumption of O3FAs combined with genetic factors could contribute to higher rates of morbidity. This article will explore the evidence for O3FA supplementation in primary and secondary prevention of cardiovascular disease (CVD).
Sources of information
PubMed, the Cochrane Library, and Google Scholar were searched for meta-analyses and reviews related to O3FAs and CVD. Salient randomized controlled trials (RCTs) published in 2019 or later and referenced in these reviews were retrieved. Current lipid guidelines were reviewed.
Levels of evidence were defined using GRADE (Grading of Recommendations Assessment, Development and Evaluation) taxonomy as outlined by the Canadian Task Force on Preventive Health Care4 using strength of recommendation and levels of evidence in evaluation. Research articles on fish intake were prospective cohort studies and could not be blinded, and they therefore provided moderate-quality evidence for conditional recommendations. Recent meta-analyses examining effects of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) combination on CVD provided high-quality evidence for strong recommendations. (Studies in Table 1 were heavily weighted by older trials that did not have broad statin use by contemporary standards,5-21 but there is good evidence from more recent trials in Table 2.6,12,15-18,22-24) Eicosapentaenoic acid–only studies and data on atrial fibrillation (AF) are from RCTs, which provided high-quality evidence for strong recommendations. A few primary prevention–only studies yielded high-quality evidence for strong recommendations.
Outcomes of recent meta-analyses evaluating potential cardiovascular benefits of marine O3FAs
Salient studies illustrating decision points for use of O3FAs in reducing residual risk in patients who are appropriately prescribed statins
Main message
Dietary sources. Omega-3 fatty acids cannot be synthesized and must come from exogenous sources. Photosynthesis creates these sources: microalgae, green leafy vegetables, nuts, seed oils, or grass (via grazing herbivores). Grain-fed meat sources contain very little O3FAs. Nonaquatic O3FA sources contain α-linolenic acid, a short-chain fatty acid, that undergoes slow and incomplete conversion25 to long-chain O3FAs, first EPA and then DHA. These 2 O3FAs comprise the primary active sources for the essential metabolic activities that involve O3FAs, and they are mainly obtained by eating fish that have fed on microalgae (Figure 1).26-29
Dietary sources of omega-3 and omega-6 fatty acids and simplified metabolic relationships between them: Omega-3 essential fatty acids are shown in boxes with dashed outlines.
Linoleic acid, the short-chain omega-6 fatty acid, is a much larger component of the Western diet and forms a longer chain to create arachidonic acid. It competes enzymatically with O3FAs, but the latter should have the advantage if ingested in sufficient quantities.27,28 Unfortunately, leafy vegetables and oily fish have not been a dietary priority for Canadians, 88% of whom fall below the dietary recommendation of 2 fish meals per week.2
Mechanisms of action. Of the 3 O3FAs shown in Figure 1,27-29 α-linolenic acid intake alone seems to have marginal or no effect on CVD and mortality in meta-analyses.30,31 Eicosapentaenoic acid and DHA have the greatest dietary impact. While both of these fatty acids are effective at lowering triglyceride levels, the benefit observed is higher than would be expected by this mechanism alone.32 Results from the PROMINENT (Pemafibrate to Reduce Cardiovascular Outcomes by Reducing Triglycerides in Patients with Diabetes) trial33 published in 2022 demonstrated that a 26.2% reduction in triglyceride levels by pemafibrate failed to reduce CVD events in primary- and secondary-risk populations in whom statins had been appropriately prescribed. Pleiotropic effects in addition to triglyceride level reduction may therefore be the main driver of CVD event reduction, and the properties of EPA and DHA may differ with respect to these mechanisms, which are outlined in Table 3.6,12,18,34-52 In particular, DHA is seen to raise low-density lipoprotein cholesterol (LDL-C) levels, while EPA lowers the atherogenic particle number with no effect on LDL-C levels.6 In addition, only EPA reduces oxidation of small LDL-C particles in the subendothelial space, reducing the potential for retention and phagocytosis and reducing foam-cell production. Atheroma production is therefore reduced.41 Eicosapentaenoic acid may therefore be superior to DHA in reducing atherogenesis.
Proposed mechanisms for benefit of O3FAs in patients with CVD
Evidence of benefit. Meta-analyses evaluating potential CVD effects of O3FA are listed in Table 1.5-21 All were published in 2019 or later and some contain the most recent definitive trials.12,15 Overall, they seem to suggest a reduction in myocardial infarction5,8,9,11,14,16,21 and CVD events,5,7-10,13,14,16,19-21 but no impact on mortality or stroke. However, a re-examination53 of the 2021 Khan et al meta-analysis16 suggests that inclusion of the older GISSI (Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto miocardico)-Prevenzione17 and GISSI–Heart Failure18 trials contributed to 41.5% of the relative weight of all trials evaluating EPA and DHA interventions. The participants in these trials did not broadly use statins and so the trials do not reflect contemporary care (Table 2).6,12,15-18,22-24 When the analysis was run again excluding these trials,54 no benefit was found for EPA and DHA at any dose, in agreement with findings of the Long-Term Outcomes Study to Assess Statin Residual Risk with Epanova in High Cardiovascular Risk Patients with Hypertriglyceridemia (STRENGTH) trial12 of high-dose EPA and DHA published in 2020. The authors of the redone analysis further stated that any CVD benefits were limited to EPA monotherapy when statins were appropriately prescribed.54
Eicosapentaenoic acid alone appears to be more effective, but studies were fewer and were done with patients at higher risk.6,24,55-58 Despite this, all outcome and mechanistic trials show benefit (Table 4).6,24,55-58 The JELIS (Japan EPA Lipid Intervention Study) trial24 had an open-label, blinded end point design with participants at high risk undergoing primary prevention. The Reduction of Cardiovascular Events with EPA–Intervention Trial (REDUCE-IT)6 showed a remarkable reduction in composite CVD end points using high-dose EPA for both primary and secondary prevention in patients at high risk. Neither REDUCE-IT6 nor the STRENGTH trial12 were adequately powered to distinguish relative effects in primary or secondary prevention.59
EPA-only effects on CVD: Mechanistic and outcome studies.
Taken together, there is no indication of benefit using EPA and DHA at any dose for prevention of CVD in primary or secondary prevention if patients are taking an appropriate statin at the appropriate dose. Eicosapentaenoic acid alone, when added to optimized statin therapy at a high dose, is effective in reducing residual risk in secondary prevention and for primary prevention in patients with diabetes that is accompanied by additional risk factors.
Trials of fish consumption are all long-term prospective cohort studies implying association rather than causation (conditional recommendation: moderate-quality evidence). Other components of fish, such as protein, trace minerals, and vitamins, may confer benefits in addition to O3FA.7,60 Two recent meta-analyses suggest an association between consuming 2 fish meals per week and reduced CVD incidence and mortality.7,10 Dose-response analysis suggested a 4% reduction in mortality for every 20 g per day of fish ingested, with a plateau reached at 175 g per week.10,61
A recent large pooled data set from studies of patients with or without CVD has suggested that the benefits of increased fish intake pertain only to secondary prevention, with no effect on patients without CVD.61
Concerns. No food or drug interactions have been reported with O3FA. Control of warfarin anticoagulation is not affected,62 but caution is warranted with use of any anticoagulant or antiplatelet agent. Although serious bleeding was not increased in the REDUCE-IT trial,6 a drug interaction database should be consulted. Gastrointestinal symptoms such as nausea, diarrhea, and dyspepsia are reported more frequently compared with control groups.12 Edema has been reported.63 Rates of serious bleeding, including hemorrhagic stroke, are not different from the placebo group.6,12
The major concern has been increased AF. There is an absolute increase of 1.4 percentage points in AF events using high-dose EPA6 and 0.9 percentage points using high-dose EPA and DHA carboxylic acids.12 Events are increased at O3FA doses of more than 1 g.64 No increase in AF is noted in 2 well-done primary prevention RCTs at doses of less than 1 g.23,65 Supplementation seems to be safe at low doses and in the absence of coronary artery disease. Patients with established heart disease and a high-dose indication need to be informed, although absolute numbers at risk may be quite small. Patients who have had ablation or paroxysmal AF might experience increased frequency of events, but no informative data currently exist.
A second concern is opportunity cost. Canadian lipid guidelines63 do not recommend O3FA to reduce risk of CVD in the absence of moderately raised triglyceride levels of 1.5 to 5.6 mmol/L in secondary prevention, or with diabetes plus 1 additional risk factor in primary prevention. While this may be appropriate based on evidence from existing RCTs in cardiology, we should consider the benefits we might lose when dietary choices are suboptimal and given that 40% of Canadians are found to have insufficient blood levels of O3FA.2 There are promising results from clinical trials and meta-analyses suggesting substantial benefit to patients with various chronic diseases including metabolic syndrome,66 diabetes,67 macular degeneration,68 inflammatory joint pain,69 Alzheimer disease prevention in apolipoprotein E4 carriers,70 nonalcoholic fatty liver disease,71 and depression.72
A third concern is incremental cost. It has been estimated that worldwide fish stocks peaked in 2000 and that the fishery could collapse by 2050 if trends continue.73 Similarly, fish farming is unsustainable with current practices, as two-thirds of fish products go to aquaculture, thus explaining the higher O3FA content of farmed fish.74 Currently consumed fish stocks are sourced equally from the wild fishery and aquaculture.26 Unless these 2 sources are uncoupled, both practices are unsustainable.
Most supplements and pharmaceuticals containing EPA and DHA are derived from fish oil35 (Table 5).6,12,26,75,76 With the increased demand for O3FA sources in the developed world there is also concern for food security in the developing world, as these countries rely on seafood for up to 20% of protein requirements.73 It requires 20 to 25 kg of fish to produce 1 kg of fish oil.74 Existing economic and social inequalities will be increased further. A possible solution may come from culture of microalgae, from which oils containing EPA and DHA can be extracted, or from genetically modified plant sources should they become widely accepted by the public.26,77
Omega-3 fatty acid supplements and pharmaceuticals
Mitigation
Diet: Derivation78 and validation30 studies have established an association between red blood cell O3FA levels (the Omega-3 Index) and CVD mortality. An index score of 4% or less suggests higher risk, with 8% or more considered ideal. It is usually unnecessary to determine blood levels. Persons with potential insufficiency most likely to benefit are those with established CVD who eat very little fish, who are vegetarians or vegans, or who have elevated triglyceride levels between 1.5 and 5.6 mmol/L.
Of the recommended oily fish, sardines are probably the ideal seafood source, economically and environmentally. They are inexpensive, come from a wild source, and are rich in O3FAs. Salmon are at least twice the price and the source is often problematic.79 Two servings of oily (not fried) fish per week are recommended by American Heart Association guidelines (conditional recommendation; moderate-quality evidence).47 Fish oil supplements can no longer be recommended for CVD prevention in either primary or secondary prevention at any dose (strong recommendation; high-quality evidence). People may choose to use fish oil for other perceived benefits, but the dosage should remain below 1 g per day to minimize AF risk. Microalgal oil–derived supplements are available for vegetarians and vegans.
Pharmaceuticals: Recommendations in Figure 2 are based on the 2021 Canadian Cardiovascular Society dyslipidemia guideline.63 This is the most current and relevant of existing guidelines.
Flow chart for use of O3FA supplementation in mitigation of CVD
Primary prevention is best implemented by respecting patient preference for increasing fish or fish oil intake with the understanding that fish oils are not effective in CVD prevention. The Canadian Cardiovascular Society guideline recommends considering icosapent ethyl, an EPA ethyl ester, given at 2 g twice a day for patients who have diabetes and 1 or more additional CVD risk factor plus triglyceride levels between 1.5 and 5.6 mmol/L, and who are taking an appropriate statin at the appropriate dose (strong recommendation; high-quality evidence). This is based on the REDUCE-IT study,6 in which only 29.3% of participants were enrolled for primary prevention. Although the primary prevention cohort was believed to have derived no benefit in this study, other authors believe that the study had been insufficiently powered to find significant benefit.59
Patients who could benefit from O3FA for secondary prevention should first be taking an appropriate statin at the appropriate dose with or without adjuvant lipid-lowering agents. If triglyceride levels are between 1.5 and 5.6 mmol/L, non–high-density lipoprotein cholesterol or apolipoprotein B levels should be used as therapeutic thresholds for statin dosing and these patients should be offered 2 g of icosapent ethyl twice a day. A combination of EPA and DHA at similar doses should not be used because of absence of benefit in the STRENGTH trial, which was terminated early for futility12 (strong recommendation; high-quality evidence).
Conclusion
In considering CVD prevention using O3FAs, all guidelines continue to recommend at least 2 servings of oily fish per week, although benefit is probably seen only in secondary prevention. Fish oil preparations and combinations of EPA and DHA have failed to show benefit at any dose at any level of prevention in patients who are appropriately prescribed statins.
Eicosapentaenoic acid alone reduces residual risk in patients taking statins. High-dose EPA yields substantial benefit for patients with triglyceride levels between 1.5 and 5.6 mmol/L taking an appropriate statin at the appropriate dose who have CVD (secondary prevention), or who have diabetes and 1 or more additional CVD risk factors (primary prevention).
There is a substantial relative increase in AF incidence with use of high-dose O3FA, but the absolute increase is small. No indication of any harm is found at doses below 1 g daily.
Notes
Editor’s key points
▸ In considering cardiovascular disease (CVD) prevention by omega-3 fatty acids, all guidelines continue to recommend at least 2 servings of oily fish per week, although benefit is probably seen only in secondary prevention. Fish oil and combination eicosapentaenoic acid and docosahexaenoic acid preparations have failed to show benefit at any dose at any level of prevention in patients who are appropriately prescribed statins.
▸ Eicosapentaenoic acid alone reduces residual risk in patients taking statins. High-dose eicosapentaenoic acid yields substantial benefit for patients with triglyceride levels between 1.5 and 5.6 mmol/L taking an appropriate statin at the appropriate dose who have CVD (secondary prevention), or who have diabetes and 1 or more additional CVD risk factors (primary prevention).
▸ There is a substantial relative increase in atrial fibrillation incidence with use of high-dose omega-3 fatty acids, but the absolute increase is small. No indication of harm is found at doses below 1 g per day.
Footnotes
Competing interests
Dr N. John Bosomworth was on the secondary panel for review of the most recent 2021 Canadian dyslipidemia guidelines, which are referenced in the article.
This article is eligible for Mainpro+ certified Self-Learning credits. To earn credits, go to https://www.cfp.ca and click on the Mainpro+ link.
This article has been peer reviewed.
La traduction en français de cet article se trouve à https://www.cfp.ca dans la table des matières du numéro de juillet 2023 à la page e134.
- Copyright © 2023 the College of Family Physicians of Canada