OUP user menu

Long-chain omega-3 fatty acids: time to establish a dietary reference intake

Michael R Flock, William S Harris, Penny M Kris-Etherton
DOI: http://dx.doi.org/10.1111/nure.12071 692-707 First published online: 1 October 2013


The beneficial effects of consuming omega-3 polyunsaturated fatty acids (n-3 PUFAs), specifically eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), on cardiovascular health have been studied extensively. To date, there is no dietary reference intake (DRI) for EPA and DHA, although many international authorities and expert groups have issued dietary recommendations for them. Given the substantial new evidence published since the last Institute of Medicine (IOM) report on energy and macronutrients, released in 2002, there is a pressing need to establish a DRI for EPA and DHA. In order to set a DRI, however, more information is needed to define the intakes of EPA and DHA required to reduce the burden of chronic disease. Information about potential gender- or race-based differences in requirements is also needed. Given the many health benefits of EPA and DHA that have been described since the 2002 IOM report, there is now a strong justification for establishing a DRI for these fatty acids.

  • cardiovascular disease
  • DHA
  • dietary reference intakes
  • EPA
  • omega-3 fatty acids


Considerable progress has been made over the past decade to better understand the biological effects of dietary fatty acids. Omega-3 polyunsaturated fatty acids (n-3 PUFAs), specifically eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), modulate both metabolic and immune processes and confer health benefits in areas of cardiovascular disease (CVD) and neurodevelopment. The shortest n-3 PUFA is alpha-linolenic acid (ALA), an 18-carbon fatty acid found in a variety of plant-based foods. In contrast, EPA and DHA (comprised of 20 and 22 carbons, respectively) are considered highly unsaturated fatty acids (HUFA; characterized as ≥20 carbons and ≥3 double bonds) and are found in marine sources (mainly fish and especially oily fish); DHA is also found in algae. Clinical trials, animal studies, and observational studies have demonstrated that fish and fish oil improve different inflammatory pathologies.1 Numerous mechanistic details as to how EPA and DHA modulate chronic disease have been reported.2

Acculmating evidence indicates that increasing EPA and DHA intake reduces risk of CVD35; however, the optimal dose of n-3 PUFAs remains to be resolved.

In 2002, the Institute of Medicine (IOM) report entitled Dietary Reference Intakes for Energy, Carbohydrates, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids6 concluded there was insufficient evidence to set a dietary reference intake (DRI) for EPA and DHA. Since the release of the report, the evidence base regarding EPA and DHA has grown dramatically, which has led a number of organizations and expert groups around the world to issue evidence-based recommendations for EPA and/or DHA, as described below. It is time for the IOM to address this question as well. A DRI for EPA and DHA will help inform nutrition policy decisions and will reduce consumer uncertainty about target n-3 PUFA intakes. In the present article, current recommendations for n-3 PUFA intakes set by various federal agencies and expert groups are reviewed, along with the evidence that was used to set them. This provides the rationale for convening an expert panel to set an official DRI for EPA and DHA. Key challenges and specific areas of research that need to be addressed when implementing such recommendations are also identified.

Dietary Intake of n-3 Polyunsaturated Fatty Acids

Dietary Reference Intakes

DRIs are a set of four nutrient-based reference values issued by the IOM of The National Academies. They are defined as follows:

Estimated average requirement (EAR)

“The average daily nutrient intake level estimated to meet the requirement of half of the healthy individuals in a particular life stage and gender group. The EAR reflects the estimated median requirement and is particularly appropriate for applications related to planning and assessing intakes for groups of persons.”6

Recommended daily allowance (RDA)

“The average daily dietary nutrient intake level sufficient to meet the nutrient requirement of nearly all (97–98%) healthy individuals in a particular life stage and gender group. The RDA is derived from the EAR and the corresponding population variability in requirements.”6

Adequate intake (AI)

“The recommended average daily intake level based on observed or experimentally determined approximations or estimates of nutrient intake by a group (or groups) of apparently healthy people that are assumed to be adequate – used when an EAR (and thus an RDA) cannot be determined.”6

Tolerable upper intake level (UL)

“The highest average daily nutrient intake level that is likely to pose no risk of adverse health effects to almost all individuals in the general population. As intake increases above the UL, the potential risk of adverse effects may increase.”6

There has been a transition in the DRI model from a “nutrient adequacy” approach to one that also focuses on “disease prevention.” A DRI review is no longer based only on whether a nutrient is essential, but also on whether a certain level of intake can reduce risk in their definition for a chronic disease (as opposed to a classic nutritional deficiency). Accordingly, the DRIs now include chronic disease endpoints and address concepts of probability and risk to define a DRI.7 This perspective is captured in the acceptable macronutrient distribution range (AMDR). This is an intake range for macronutrients, expressed as a percentage of total energy that is associated with reduced risk of chronic disease while providing adequate intake of essential nutrients. As stated in the IOM report,6 “A key feature of each AMDR is that it has a lower and upper boundary, some determined mainly by the lowest or highest value judged to have an expected impact on health. If an individual consumes below or above this range, there is a potential for increasing the risk of chronic diseases shown to affect long-term health, as well as increasing the risk of insufficient intakes of essential nutrients.” This new perspective on chronic disease risk was evident in the 2010 IOM report on DRIs for vitamin D and calcium, which used bone health as well as nonskeletal chronic disease outcomes as indicators.8

Although chronic disease is often addressed, DRIs are intended to meet the needs of healthy people, not individuals with disease. A challenge is defining the word “healthy.” Importantly, the intent is to offer dietary guidance for the promotion of health and the prevention of chronic diseases, particularly since the latter are rampant in the United States. Issues such as these make setting a DRI for the nonclassical nutrients (such as EPA and DHA) important.

There currently is no EAR, which is the basis for RDAs, for n-3 HUFAs. There is, however, an AI for ALA (1.6 g/day for adult men and 1.1 g/day for adult women). This is based on the observed median intake in the United States at which no nutrient deficiency is present. In addition, the AMDR for ALA is defined as 0.6–1.2% of energy,6 up to 10% of which can be consumed as EPA and/or DHA. It is important to note that this is not an AMDR (or a DRI) for EPA and/or DHA; it simply indicates that consumption of between 0.06% and 0.12% of energy as EPA and/or DHA will “count” toward meeting the AMDR for ALA.

Are EPA and/or DHA essential?

ALA is an essential fatty acid because humans lack the desaturase enzyme that inserts a double bond at the C-15 position of a fatty acid carbon chain to form n-3 PUFAs (Figure 1). In humans, endogenous synthesis of EPA and DHA from ALA is minimal, with between 0.01% and 8% of ALA being converted to EPA and less to DHA9,10; thus, plasma and tissue levels are determined largely by direct consumption. Even large amounts of dietary ALA have a negligible effect on plasma DHA.1113 This poses the question as to whether EPA and DHA also are essential fatty acids, due to their limited conversion from ALA, particularly with currently recommended linoleic acid intakes at 5–10% of energy. Since EPA and DHA have various metabolic functions not duplicated by other fatty acids, they could be viewed as conditionally essential fatty acids.

Figure 1

Metabolic pathway of n-3 PUFA synthesis via a series of desaturation and elongation reactions in the liver.

PUFAs are released from the sn-2 position of membrane phospholipids by the action of phospholipase A2. The released fatty acids are substrates for cytochrome P450 monoxygenases (CYP450), cyclooxygenases (COX), and lipoxygenases (LOX). Cyclooxygenase (COX) and lipoxygenase (LOX) enzymes metabolize EPA to form eicosanoids, including prostaglandins (PG), thromboxanes (TX), and leukotrienes (LT), as well as resolution-phase interaction products (resolvins).

Abbreviations: ALA, (alpha)-linolenic acid; DHA, docosahexaenoic acid; EDP, epoxydocosapentaenoic acid; EEQ, epoxyeicosatetraenoic acid; EPA, eicosapentaenoic acid; EpODEs, epoxyeicosatrienoic acid; HDoHE, hydroxydocosahexaenoic acid; HEPE, hydroxyeicosapentaenoic acid; HOTE, hydroxyoctadecatrienoic acid; PD, protectins; RvD, DHA-derived resolvins; RvE, EPA-derived resolvins.

Recommendations for n-3 polyunsaturated fatty acids

Many organizations and expert committees acknowledge the important role of EPA and DHA in human nutrition (Table 1).3,5,1417 The Dietary Guidelines for Americans, 20105 (DGA 2010) recommends consuming about 8 oz/week (i.e., two 4-oz servings) of a variety of seafood to reduce cardiac deaths among individuals with or without preexisting CVD. This would provide about 250 mg/day of EPA and DHA, which is sufficient to obtain cardioprotective effects for primary prevention of CVD. That this should be a minimum intake was recently argued by Musa-Veloso et al.,18 who reported that intakes in excess of 250 mg/day elicited a 35.1% greater reduction in risk of sudden cardiac death compared with intakes <250 mg/day in subjects free of known coronary heart disease (CHD). The American Heart Association (AHA) recommends that adults without CHD eat fish (particularly fatty fish) at least twice a week,3 which provides about 500 mg/day.19 Individuals with CHD are advised to consume 1 g/day of EPA and DHA for secondary prevention of CVD, preferably from seafood and in consultation with a physician.20 Those who need to lower triglyceride (TG) levels are recommended to take 2–4 g/day of EPA and DHA as capsules under a physician's care. In terms of their effects on blood levels of EPA and DHA, fish oil supplements provide benefits similar to those of seafood.21 Updated AHA recommendations for fish consumption22 are to consume two or more 3.5-oz servings/week (preferably oily fish). Although concern has been raised about high doses causing excessive bleeding in some individuals, there is little evidence of any increased risk for clinically significant bleeding.23,24 Nevertheless, individuals are advised to consult a physician before taking larger doses of n-3 PUFAs (=2 g/day). Unlike fish oil supplements, seafood contains many nutrients, such as high-quality protein and vitamin D, which may confer cardioprotective benefits beyond those of n-3 PUFAs.

View this table:
Table 1

Recommended intakes of EPA and DHA

OrganizationPopulationEPA and DHA recommendation
Academy of Nutrition and Dietetics14Adults≥500 mg/day
American Heart Association3Adults without CHDFatty fish ≥2 times/week (∼500 mg/day)a
Patients with CHD∼1 g/day
Patients with high TG2–4 g/day
US Department of Agriculture5Adults≥250 mg/day
International Society for the Study of Fatty Acids and Lipids15Adults≥500 mg/day
Pregnant/lactating women≥500 mg/day (≥300 mg/day of DHA)
European Food Safety Agency16Adults≥250 mg/day
Pregnant/lactating women≥250 mg/day (and100–200 mg/day DHA)
World Health Organization17Adults≥250 mg/day
  • a Calcuated to be approximately 500 mg/day.19

  • Abbreviations: CHD, coronary heart disease; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; TG, triglycerides.

Inadequate intake of DHA by pregnant women may impair neural development in the neonate.2529 Sufficient DHA intake during pregnancy and lactation is crucial for proper brain development of infants. Thus, recommendations have been made for pregnant and lactating women. The International Society for the Study of Fatty Acids and Lipids recommends that pregnant and/or lactating women consume ≥300 mg/day of DHA15; the European Food Safety Agency recommends women consume ≥250 mg/day of EPA and DHA plus an additional 100–200 mg/day of DHA.16 Children, pregnant women, and women who may become pregnant are advised to avoid eating fish with higher levels of mercury. The Dietary Guidelines for Americans, 20105 recommends eating a variety of seafood (two 4-oz servings per week) to reduce the amount of mercury consumed from any one type of seafood.

Fiber as a model nutrient for a DRI

The approach used in setting the DRI for fiber can be mimicked for setting a DRI for EPA and DHA. The 2002 IOM report was the first to set a DRI for total fiber intake, which included both dietary and functional fiber.6 Dietary fiber is the edible, nondigestible component of carbohydrates found in plants, whereas functional fibers are isolated or extracted from natural sources. This is simliar to EPA and DHA intake, where fish is the main dietary source and fish oils and functional foods containing EPA and DHA provide an auxiliary source. Thus, using the fiber DRI as a model, a DRI for EPA and DHA would be based on direct dietary sources of EPA and DHA (i.e., seafood) as well as foods fortified with EPA and DHA.

A convincing body of literature showed an increased CVD risk when diets were low in fiber. The IOM concluded that the best approach was to set an AI, since the available evidence was insufficient to determine an EAR and thus calculate an RDA.6 An AI was set for men and women at 14 g per 1,000 kcal, which is an intake observed to protect against CHD. This recommended intake for total fiber was considered sufficient to support intestinal function and reduce constipation in most heatlhy individuals, given adequate hydration.6 The same methodology used to establish a DRI for total fiber can be used for EPA and DHA. In fact, the evidence supporting a CVD benefit with EPA and DHA intake is substantaily greater than that for fiber. According to the National Health and Nutrition Examination Survey (NHANES) 2009–2010, adult men and women consume only approximately 120 mg and 90 mg of EPA and DHA per day from food, respectively.30 With additional evidence, the goal is to establish EARs and RDAs.

EPA and DHA and Cardiovascular Disease

The DRI process involves the identification of health outcomes that are consistently and causally linked to the nutrient of interest.8 Over the past decade, EPA and DHA have emerged as key modifiers of CVD risk.18,19,31 A significant body of epidemiological and clinical evidence has examined the cardioprotective effects of EPA and DHA consumption (Tables 2-3). Thus, just as CVD was used to help establish an AI for dietary fiber, it can serve as a suitable endpoint for developing an AI for EPA and DHA.

View this table:
Table 2

Meta-analyses of randomized controlled trials evaluating the effects of EPA and DHA on cardiac and/or all-cause mortality

ReferenceNo. of studiesPreventionSummary estimate (RR, 95% CI)
Cardiac deathAll-cause mortality
Bucher et al. (2002)3111Secondary0.70 (0.60–0.80)0.80 (0.70–0.90)
Hooper et al. (2004)4448Primary/secondary0.90 (0.61–1.33)0.98 (0.70–1.36)
Leon et al. (2008)4612Primary/secondary0.80 (0.69–0.92)0.92 (0.82–1.03)
Marik & Varon (2009)4511Primary/secondary0.87 (0.79–0.95)0.92 (0.85–0.99)
Filion et al. (2010)4329SecondaryN/A0.88 (0.64–1.03)
Trikalinos et al. (2012)417Primary/secondary0.89 (0.83–0.96)0.95 (0.89–1.01)
Kwak et al. (2012)4214Secondary0.91 (0.84–0.99)0.96 (0.90–1.02)
Rizos et al. (2012)4120Primary/secondary0.91 (0.85–0.98)0.96 (0.91–1.02)
  • Abbreviations: CI, confidence interval; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; N/A, not available; RR, relative risk.

View this table:
Table 3

Randomized controlled trials evaluating the effects of EPA and DHA on cardiac and/or all-cause mortality

ReferenceMean age in years (% males)No. treated/controlEPA and DHA (g/day)ControlIndicationDuration of study (years)Summary estimates RR (95% CI)
Cardiac mortalityAll-cause mortality
Burr et al. (1989)4757 (100)1,015/1,0180.24 or 0.86Non-fish-oil dietSecondary2N/A0.71 (0.54–0.93)
Sacks et al. (1995)4862 (93)31/286.0Placebo (olive oil)Secondary2.30.30 (0.01–7.13)0.30 (0.01–7.13
Singh et al. (1997)4949 (93)122/1181.8Placebo (non-oil)Secondary10.52 (0.29–0.95)N/A
Leng et al. (1998)5066 (68)60/600.27Placebo (sunflower seed oil)Secondary21.00 (0.15–6.87)1.00 (0.21–4.76)
von Schacky et al. (1999)5159 (76)111/1123.4 (3 mo), 1.7 (21 mo)Placebo (nonmarine fatty acids)Secondary20.33 (0.01–8.02)0.50 (0.05–5.48)
Johansen et al. (1999)5260 (78)250/2505.1Placebo (corn oil)Secondary0.50.33 (0.03–3.18)0.33 (0.03–3.18)
Durrington et al. (2001)5359 (73)30/293.2Placebo (corn oil)Secondary0.90.32 (0.01–7.61)0.32 (0.01–7.61)
Nilsen et al. (2001)5464 (80)150/1503.4Placebo (corn oil )Secondary1.51.00 (0.39–2.59)1.00 (0.45–2.24)
Marchioli et al. (2002)5560 (85)5,665/5,6580.85Vitamin E or no supplementSecondary3.50.81 (0.68–0.95)0.86 (0.77–0.97)
Burr et al. (2003)5661 (100)1,571/1,5430.34 or 0.86Non-fish-oil dietSecondary51.26 (1.00–1.58)1.15 (0.96–1.36)
Leaf et al. (2005)5766 (83)200/2022.6Placebo (olive oil)ICD11.01 (0.41–2.49)1.09 (0.51–2.34)
Raitt et al. (2005)5863 (86)100/1001.3Placebo (olive oil)ICD20.40 (0.08–2.01)0.40 (0.13–1.23)
Brouwer et al. (2006)5962 (84)273/2730.96Placebo (sunflower oil)ICD10.46 (0.18–1.20)0.57 (0.24–1.34)
Yokoyama et al. (2007)6061 (32)9,326/9,3191.8aStandard carePrimary/secondary4.60.93 (0.56–1.55)1.08 (0.91–1.27)
Tavazzi et al. (2008)6167 (78)3,494/3,4810.85Placebo (olive oil)Primary/secondary3.90.93 (0.85–1.02)0.94 (0.87–1.01)
Rauch et al. (2010)6264 (74)1,940/1,9111.0Placebo (olive oil)Secondary10.95 (0.57–1.59)1.24 (0.91–1.68)
Galan et al. (2010)6361 (80)1,253/1,2480.6Placebo (paraffin liquid)Secondary4.7N/A0.98 (0.69–1.39)
Kromhout et al. (2010)6469 (78)2,404/2,4330.4Placebo (margarine)Secondary3.40.99 (0.73–1.34)1.02 (0.84–1.24)
Einvik et al. (2010)6570 (100)282/2812.4Placebo (corn oil )Primary/secondary30.94 (0.88–1.01)0.58 (0.31–1.10)
Bosch et al. (2012)6664 (65)6,281/6,2551.0Placebo (olive oil)Primary/secondary6.20.98 (0.87–1.10)0.98 (0.89–1.07)
  • a EPA only.

  • Abbreviations: CI, confidence interval; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; ICD, implantable cardioverter defibrillators; mo, months; N/A, not available; RR, relative risk.

Epidemiological studies

Numerous epidemiological studies have reported an inverse relationship between EPA and DHA intake and CVD outcomes.3237 Of the intakes reported in these studies, 250 mg/day of EPA and DHA was the lowest level that significantly reduced the risk of cardiovascular events. Furthermore, Harris et al.19 pooled results from several observational studies, comparing the highest EPA and DHA intake groups with the lowest intake groups19; EPA and DHA intake and the relative risk of CHD mortality were significantly associated (P = 0.03). These results were derived from over 1.6 million person-years of follow-up, independent of other known risk factors, and were dose dependent.19 The average EPA and DHA intake associated with the greatest reduction in risk of CHD mortality (roughly 37%) was approximately 566 mg/day.19 A recent meta-analysis of prospective cohort studies found that higher dietary fish consumption was associated with about 15% lower risk of heart failure compared with lower fish consumption38; however, the association between EPA and DHA and atrial fibrillation has been less consistent.39,40 Understanding the specific cardiovascular outcomes associated with EPA and DHA intake will be imperative in determining optimal EPA and DHA intakes.

Intervention studies

Evidence supporting EPA and DHA recommendations in adults is based on primary and secondary prevention of CVD. The role of EPA and DHA in reducing sudden cardiac death and CVD risk factors has been the subject of numerous clinical studies. The DGA 20105 concluded that consuming two 4-oz servings of seafood per week, providing a total of 250 mg/day of EPA and DHA, reduces mortality from CHD or sudden death in persons with and without CVD.

Over the past decade, several meta-analyses have been published evaluating the potential benefits of EPA and DHA intake in primary and secondary prevention of cardiac and all-cause mortality (Table 2).4,31,4146 Virtually every meta-analysis has reported beneficial effects of EPA and DHA supplementation (or oily fish intake) on cardiac death, but total morality often is not statistically significant. Table 3 presents many of the randomized controlled trials (RCTs) included in these meta-analyses.4766 There is a need to determine why some trials show a benefit of EPA and DHA intake, whereas others do not. Discrepant findings may be partly explained by inherent differences between baseline populations in primary and secondary designs (i.e., generally healthy subjects compared with individuals with previous vascular disease). Differences in study design, subject population, duration, dosage, and the design of the meta-analyses exemplify inconsistencies among primary and secondary prevention trials. A recent meta-analysis of 12 RCTs reported that the pooled relative risk for cerebrovascular disease in EPA and DHA supplement groups was not significantly different from that in the control groups (relative risk 1.03; 95% confidence interval [95% CI] 0.94–1.12)67; the pooled relative risk for primary prevention (2 RCTs) was 0.98 (95% CI 0.89–1.08) and for secondary prevention (10 RCTs) 1.17 (95% CI 0.99–1.38).67 Competing risk events, such as CHD and cerebrovascular outcomes, may have limited power analysis by altering the probability of a specific outcome and thus impeding subsequent events.67 Therefore, more adequately powered intervention data, especially involving healthy populations, are needed.

Trikalinos et al.4 published a systematic review and meta-analysis of RCTs and prospective cohort studies evaluating the effects of EPA and DHA on cardiac, cardiovascular, or all-cause mortality. Of the 18 eligible RCTs, only 3 were conducted in the United States (13 were done in Europe, 1 in Japan, and 1 in India). Supplementation with EPA and DHA (approximately 0.27–6.0 g) reduced the relative risk of cardiac mortality (0.89; 95% CI 0.83–0.96), with no evidence of heterogeneity found. Evidence from prospective cohort studies (7 cohorts, 123,122 subjects) showed that EPA and DHA intake up to 200 mg/day was associated with reduced risk of cardiac, cardiovascular, or sudden cardiac mortality (0.64; 95% CI 0.45–0.89). These results suggest that the beneficial effects of EPA and DHA on mortality plateau once a mean intake threshold is reached.4 The authors acknowledged, however, that analyzing individual participant data with suitable methodologies and considering other outcomes would help refine this threshold. Regardless, convening an expert panel to set an official DRI for EPA and DHA will help clarify discrepancies in current recommendations and eliminate existing uncertainties. Effects of EPA and DHA in healthy adults (primary prevention) would be the most relevant for DRI development. Despite numerous prospective cohort studies reporting data on outcomes in healthy adults, few RCTs have been published.60,61,65,66 Nonetheless, more well-designed placebo-controlled clinical trials are needed to assess the role of EPA and DHA in the primary prevention of CVD.

Establishing a DRI for EPA and DHA

The available data for cardiac mortality provide a basis for establishing a DRI for EPA and DHA. Current intakes (approximately 100 mg/day) are not sufficient; setting a DRI for EPA and DHA is important for realizing health benefits, specifically CVD prevention. In the Harris et al.19 analysis, 566 mg/day of EPA and DHA was the average intake associated with the greatest reduction (37%) in risk of CHD mortality. Based on these data, an AI of 566 mg/day could be considered for EPA and DHA intake. However, if 566 mg/day is assumed as the median requirement, i.e., it reduced risk of CHD mortality in 50% of the population, then 566 mg/day would be considered as an EAR. The RDA could then be calculated using the standard deviation (224 mg/day). Two standard deviations above the EAR would be approximately 1 g/day, which would be the amount of EPA and DHA expected to meet the needs of nearly all healthy individuals (Figure 2). Of course, such requirements are speculative; the convening of a DRI committee is necessary to thoroughly examine the body of evidence and determine whether an RDA can even be established. Recommendations calling for =1 g/day of EPA and DHA are typically aimed at secondary prevention of CVD. To date, no UL for EPA and DHA has been set by any authoritative body. The US Food and Drug Administration (FDA) has stated that levels up to 3 g/day are generally recognized as safe,68 although other authorities have reported no adverse effects at intake levels up to 5–6 g/day.24,69

Figure 2

Potential dietary reference intakes for EPA and DHA based on cardiovascular benefit. Dashed lines highlight uncertainty regarding the indicated RDA, suggesting additional research is still needed.

a250 mg/day recommended by the US Department of Agriculture, 2010,5 the European Food Safety Agency,16 and the World Health Organization.17

b500 mg/day recommended by the Academy of Nutrition and Dietetics,14 the International Society for the Study of Fatty Acids and Lipids,15 and the American Heart Association3 (for those without CHD).

c1.0 g/day recommended by the American Heart Association3 for secondary prevention of CHD.

d2.0 g/day recommended by the American Heart Association3 for individuals with high triglyceride levels.

e≤3.0 g/day Generally Recognized As Safe (GRAS) by the US Food and Drug Administration.68

Abbreviations: AI, adequate intake; EAR, estimated average requirement; RDA, recommended daily allowance; UL, tolerable upper limit.

Potential mechanisms

The cardioprotective benefits of EPA and DHA are multifaceted.70 EPA and DHA have been shown to reduce susceptibility to cardiac arrhythmias,58,7176 stabilize atherosclerotic plaques,77 favorably affect serum TGs,71,7881 modestly reduce blood pressure,8286 produce less aggregatory eicosanoids compared with those from the omega-6 (n-6) fatty acid family,8791 and decrease markers of systemic inflammation and oxidative stress.9298 The mechanisms by which EPA and DHA exert their effects are also multifaceted. Due to their highly unsaturated nature (5–6 double bonds) they can increase membrane fluidity when incorporated into the phospholipid bilayer. Variation in the reported effects of EPA and DHA on cardiovascular outcomes complicates the interpretation; however, the physiological effects of EPA and DHA on CVD risk factors (i.e., hypertension, arrhythmias, TGs) clearly highlight the benefits of EPA and DHA consumption on cardiovascular health.

Mozaffarian and Rimm99 evaluated evidence in 2006 from RCTs and large prospective studies demonstrating the benefits of modest fish consumption (1–2 servings/week). In fact, 1–2 servings of fish per week, particularly fish high in EPA and DHA, reduced the risk of CHD by 36% (P < 0.001) and reduced total mortality by 17% (P = 0.046). The analysis also demonstrated a dose-response effect of fish intake, comparing 0 mg/day and 250 mg/day. Antiarrhythmic effects were reportedly strongest in reducing risk of CHD death and sudden death, having an effect within weeks of a modest intake (<750 mg/day EPA and DHA). At higher intakes (=750 mg/day EPA and DHA), however, maximum antiarrhythmic effects were achieved and lowering of TGs was greater. It is well established that EPA and DHA dose-dependently reduce fasting serum TG levels.100 This effect is, in part, attributed to decreased hepatic production as well as increased clearance of TG-rich particles.101103 Nonetheless, 1,000 mg/day of EPA and DHA was recommended for individuals with CHD to reduce CHD morbidity and mortality, which is greater than the 250–500 mg/day of EPA and DHA recommended for the general population.5,20 EPA and DHA are also ligands for certain G-protein-coupled receptors (GPRs) (e.g., GPR-120104) that inhibit proinflammatory, cell-signaling cascades. Finally, EPA- and DHA-oxygenated metabolites (eicosanoids and docosanoids, respectively) are known to have numerous physiological functions.

Increased intake of EPA and DHA results in the incorporation of these fatty acids into the membrane phospholipids of various cells, which can be enzymatically oxidized to generate eicosanoids and various lipid mediators that can modulate signaling events and alter a variety of metabolic activities (Figure 1). Arachidonic acid (AA) typically is the predominant substrate for eicosanoid synthesis, which leads to the production of prostaglandin (PG), PGE2, PGD2, leukotrienes, and thromboxanes. These eicosanoids are involved in the development of proinflammatory responses, pivotal to the progression of inflammation.105 The discovery of proresolution metabolites provides an alternative strategy to reduce inflammation, whereby EPA and DHA exert proresolving properties and reduce inflammatory complications. EPA can be enzymatically converted to eicosanoids that are very similar in structure to AA-derived eicosanoids, using the same enzymes (e.g., cyclooxygenase 2 [COX-2]), yet the EPA-derived eicosanoids exhibit very different properties. Three- and five-series eicosanoids produced from EPA tend to be more anti-inflammatory than AA-derived eicosanoids.2 Resolvins (resolution-phase interaction products), (neuro)protectins, and maresins are oxygenated metabolites derived from EPA and DHA that possess potent anti-inflammatory and proresolving actions. These specialized lipid mediators are not immunosuppressive but instead activate specific mechanisms to promote the resolution of inflammation.106 Mediators derived from EPA are designated E-series resolvins, and those biosynthesized from DHA are designated D series resolvins, whereas protectins and maresins are bioactive compounds derived only from DHA. Resolvins, protectins, and maresins can aid resolution by increasing the expression of chemokine receptor 5 on apoptotic neutrophils, thereby promoting chemokine clearance via macrophage engulfment and limiting additional neutrophil infiltration.107 This switch in macrophage phenotype exemplifies the potential role of EPA- and DHA-derived metabolites in promoting a proresolving environment.

The anti-inflammatory effects of EPA and DHA are promising, although results from observational and clinical trials have been somewhat mixed.108 There does appear to be evidence of the efficacy of EPA and DHA in patients with arthritis108111; however, clinical studies in other inflammatory conditions, such as asthma, inflammatory bowel disease, and psoriasis, have produced conflicting results.109,112118 Moreover, there is little evidence of an effect of EPA and DHA intake on the risk of cancer, which is thought to involve an inflammatory component.119124 Mechanistic studies have found that cytochrome P450-dependent metabolism of AA leads to formation of epoxy-signaling lipids that promote tumor growth125127; in contrast, cytochrome P450-dependent metabolism of EPA and DHA produces lipid mediators that inhibit tumor growth.121124 Thus, it is possible that a specific n-3 PUFA epoxide metabolite could reduce cancer risk by minimizing tumor growth.123 More research is clearly needed, particularly clinical studies, to assess the antitumor and proresolving properties of EPA and DHA.

EPA and DHA and Cognitive Function

In addition to enhancing cardiovascular health, EPA and DHA have been shown to beneficially affect cognitive function. Nevertheless, the role of EPA and DHA in cognitive function is not completely understood, despite increasing evidence linking deficiency in EPA and DHA to mental and neurological disorders.128 Both n-3 and n-6 PUFAs are present in large quantities in the brain, constituting about 30–35% of total brain fatty acids.129 Animal studies have shown consistently that diets deficient in n-3 HUFAs adversely affect learning.130136 Observational and clinical studies have focused largely on the role of n-3 HUFA levels, specifically DHA, during fetal development and infancy. Several comprehensive reviews of this literature have been published recently.129,137139 For purposes of the DRI discussion, the role of EPA and DHA in adult cognitive function will be the focus here.

Dementia, including Alzheimer's disease (AD), involves the loss of cognition to an extent that interferes with everyday function. Currently, there is no cure, treatment, or accepted prevention strategy for AD, which affects approximately 13% of the population aged 65 years and older.140 The number of Americans with AD and other dementias will continue to grow as the population over age 65 continues to increase.140 The current understanding of AD is that inflammatory and oxidative processes disrupt the functioning of neuronal cells over time, leading to DNA damage, accumulation of amyloid plaques, and eventual loss of neurons and synapses.141

Interest in the relationship between EPA and DHA and cognitive decline stems largely from the literature on DHA and neurodevelopment.142 It was discovered that AD patients have reduced DHA levels in their brain and peripheral tissues.143,144 This led many to question whether supplementation with EPA and DHA would be beneficial in preventing and/or treating cognitive decline. A recent meta-analysis reported that dementia was associated with lower blood levels of EPA (effect size = −0.47, P < 0.0001) and DHA (−0.33, P < 0.017), suggesting that EPA and DHA deficits may play an important role in the risk of dementia.145 The majority of observational research has suggested an association between EPA and DHA intake and lower risk of cognitive decline145149; however, causal inferences cannot be drawn from these studies. More clinical studies are needed, despite a large body of literature on the neurodevelopmental effects of EPA and DHA.150157

To date, only cases of mild cognitive impairment and early stages of AD have been shown to benefit from EPA and DHA supplementation.129,155,157 Freund-Levi et al.155 investigated the effects of EPA and DHA supplementation (0.6 mg/day and 1.7 g/day, respectively) or placebo for 6 months in 204 patients with AD. There was no overall effect on the Mini-Mental Status Examination score or the Alzheimer's Disease Assessment Scale (ADAS); however, in a subgroup with mild cognitive dysfunction, a significant reduction in the rate of cognitive decline was observed. Yurko-Mauro et al.152 examined the effect of DHA supplementation (900 mg/day) for 24 weeks in 485 adults over 55 years of age with mild-age related cognitive decline. Subjects doubled their plasma DHA values and exhibited improved performance in episodic memory tasks, yet showed no improvement in executive function or working memory tasks. Studies in healthy adults are far more limited. A few studies report improved reaction time on attention tests following EPA and/or DHA supplementation158,159; however, a recent meta-analysis of RCTs concluded there was no substantial cognitive benefit of EPA and DHA supplementation in healthy subjects.160

In summary, although tantalizing suggestions exist for a beneficial effect of n-3 PUFAs on risk of cognitive decline, intervention trials are necessary to better understand the role of n-3 PUFAs in ameliorating neurodegeneration. The current data are not sufficient to support an intake level different from that needed to achieve CVD risk reduction.

Dietary Strategies to Meet a DRI

Seafood is the major source of EPA and DHA in the diet. Thus, the first effective measure for increasing EPA and DHA intake is to increase seafood consumption. The DGA 2010 recommends consuming a variety of seafood in order to obtain approximately 250 mg/day of EPA and DHA.5 Species with higher EPA and DHA content should be given greater preference. Table 4 shows the variability in EPA and DHA levels, depending on the species and geographic region.161163 Eating a single serving of wild Atlantic salmon (3.5 oz or 100 g) three times a week would provide approximately 789 mg/day of EPA and DHA. However, one important question is whether fish oil producers can meet the growing demands without overfishing. If a DRI was established for EPA and DHA intake, it would be necessary to consider the ability of suppliers to meet the increased demand. Fish farms and aquacultures have helped meet the demand thus far, but it will be important to consider sustainability as the consumption of seafood increases globally. Recent development of innovative EPA and DHA products provides alternative sources of n-3 PUFAs. However, the changing composition of the food supply, including products enriched or fortified with n-3 PUFAs, adds additional complexity to quantifying n-3 PUFA consumption.

View this table:
Table 4

EPA and DHA content in selected seafood (100 g, cooked, dry heat)

SeafoodEnergy (kcal)EPA (mg)DHA (mg)EPA and DHA (mg)Mercury (ppm)
Salmon, Atlantic, farmed2066901,4572,1470.02
Salmon, Atlantic, wild1824111,4291,8400.02
Salmon, coho, wild1394016581,0590.02
Salmon, sockeye1691105246340.02
Mackerel, Atlantic2625046991,2030.05
Herring, Pacific2501,2428832,1250.08
Herring, Atlantic2039091,1052,0140.08
Catfish, wild1051001372370.03
Crab, blue83101671680.07
Bass, sea1242065567620.15
Tuna, albacore, canned1282306308600.35
Tuna, light, canned116502202700.13
Cod, Atlantic10541541580.11
Cod, Pacific85421181600.11
Mackerel, king1341742274010.73
Orange roughy105625310.57
  • Data from the USDA Agricultural Research Service (2012)162 and the Institute of Medicine (2006).161

  • Abbreviations: DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid.

Food fortification provides another means of obtaining EPA and DHA to meet the DRI. There are a variety of enriched food products, including eggs, yogurt, milk, and spreads, that contain fish-oil-derived EPA and DHA. Fish oil contains 20–80% EPA and DHA by weight, providing another option for consuming these fatty acids.99,164 Products are being fortified with EPA and DHA by feeding animals n-3 PUFAs and enriching their tissues with EPA and DHA, by adding EPA and DHA oils directly to foods, or by microencapsulating the oil prior to incorporation to maintain stability and avoid lipid oxidation.165 Biotechnology also has enabled the production of EPA and DHA oils from nonmarine sources, including specific strains of algae, genetically modified soybean oil, and bioengineered yeast.166 It has been reported recently that global consumer spending on EPA- and DHA-fortified products will increase from $25.4 billion in 2011 to $34.7 billion in 2016.167 North America accounts for 43% of these sales. Table 5 displays a variety of foods that contain EPA and/or DHA.162,165,168 Alternative dietary sources of EPA and DHA will be important to meet the growing demand for EPA and DHA among consumers; however, the rapid development of nontraditional foods fortified with EPA and DHA raises the question of food-nutrient interactions, particularly how EPA and DHA fortification influences the composition and stability of the food products.168 Foods fortified with EPA and DHA may not have the same overall health benefits associated with foods naturally high in EPA and DHA, and, depending on the food being fortified, could be more harmful than beneficial (e.g., EPA- and DHA-fortified cookies). Therefore, dietary recommendations should remain focused primarily on the consumption of quality foods naturally high in EPA and DHA.

View this table:
Table 5

EPA and DHA content in selected foods

FoodServing sizeEPA and DHA (mg)a
Nonmarine sources
 Eggs1 egg20–40
 Chicken, dark meat3.5 oz65–75
 Turkey, dark meat3.5 oz15–20
Fortified sources
 Milk8 oz25–50
 Soymilk8 oz32–50
 Orange juice8 oz50
 Yogurt4 oz16–32
 Eggs1 egg50–150
 Margarine/spread2 tbsp60–100
 Porkb3.5 oz50–800
 Peanut butter2 tbsp32
 Wheat bread1 slice30–50
  • a EPA and DHA content and source vary by brand; these ranges are based on nutrition labels and data from the USDA Agricultural Research Service (2012),162 Whelan et al. (2009),165 and Whelan & Rust (2006).168

  • b Hogs fed diets rich in algal sources of EPA and DHA.

  • Abbreviations: DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid.

Challenges in Setting a DRI

There are many challenges in establishing a DRI for n-3 HUFAs. Yet, with these challenges also come opportunities to explore areas of research needed. Some of these challenges are as follows: 1) EPA and DHA are distinct nutrients. Thus, although they are often consumed in combination, it remains unclear whether there is a greater need for one versus the other, and whether a DRI for each one should be set independent of the other. 2) Limitations in dietary intake assessment methods and nutrient databases for EPA and DHA present challenges in obtaining accurate, current intake data. 3) There is a lack of a consensus regarding the most appropriate and reliable biomarker for determining EPA and DHA status in humans. 4) The safety of markedly increased intakes of EPA and DHA needs to be defined.

EPA versus DHA

Numerous studies have demonstrated beneficial effects of EPA and DHA; however, relatively little is known about the unique effects of EPA compared with DHA. In addition, with increasing consumer awareness, many are beginning to question whether there is an appropriate form (TG or ethyl ester) or EPA-to-DHA ratio. As mentioned previously, additional research is needed to compare the specific effects of EPA versus DHA on health outcomes. It is not uncommon though to have DRIs for a mixture of nutrients, which is the case for macronutrient recommendations. Protein, fat, and carbohydrates, each containing mixtures of individual components with unique physiological effects, all have established AMDRs.6 In addition, both protein and carbohydrate have an RDA, while n-6 PUFA and ALA have an AI, as previously noted. Thus, concerns over an ideal EPA-to-DHA ratio should not deter the establishment of a DRI for these fatty acids.

Mozaffarian and Wu169 recently reviewed the evidence for shared or distinct cardiovascular effects of EPA and DHA. In animal and human studies, both EPA and DHA modulate inflammation, reduce platelet aggregation, lower TG levels, and may increase cardiac diastolic filling and arterial compliance. In contrast, observational and clinical studies suggest that DHA, in addition to increasing the proportions of large low-density lipoprotein and high-density lipoprotein particles, may have a greater TG-lowering effect. The clinical relevance of DHA effects on lipoprotein particle size is unclear. Mozaffarian and Wu169 concluded that EPA and DHA have collective as well as complementary benefits; available evidence is insufficient to specify recommendations for their individual intakes or the ratio of their intake. Both EPA and DHA provide cardiovascular benefits, and therefore increasing consumption of either EPA or DHA would be advantageous.169 More research is needed to understand the distinct effects of EPA and DHA on health outcomes; however, given their complementary effects and combined presence in fish and fish oil, it is appropriate to focus on their collective consumption in setting a DRI.

Dietary intake assessment

Reliable and valid dietary intake assessments are crucial in determining DRIs. This is a challenge for all nutrients, including EPA and DHA. Moreover, the quality of nutrient intake data can vary considerably between studies.6 The dosage, duration of intake, and source of EPA and DHA (i.e., supplements or fish), along with patient type, concomitant medications used, and other dietary and lifestyle factors, also contribute to the inconsistencies between studies and complicate the establishment of DRIs. Vegans and vegetarians have limited food choices for EPA and DHA and have low blood levels of EPA and DHA.170 Very few studies have examined whether vegetarians have greater ALA conversion efficiency compared with meat and fish eaters.171,172 Further research focusing on the endogenous production of EPA and DHA in vegans and vegetarians is needed. Reddy et al.173 found that infants born to vegetarians had lower DHA levels in plasma and cord artery phospholipids compared with infants born to nonvegetarians, while docosapentaenoic acid (DPA, 22:5, n-6) levels were greater (P < 0.001). Whether the partial replacement of DHA with DPA has any physiological consequence requires further investigation. Research on the endogenous production of EPA and DHA in vegans and vegetarians is needed in order to set a DRI for this population.

Interindividual variability in response to EPA and DHA also presents a significant challenge to defining DRIs, which is an issue IOM committees encounter when identifying standardized and consistent data on nutrient intakes. Differences in age, sex, weight, and perhaps race, as well as overall health status, complicate the discussion of optimal EPA and DHA intake. The impact of genetics on EPA and DHA levels (synthesis, absorption, metabolism) also is a likely determinant of response. A recent study in Framingham reported that, after controlling for dietary intake of EPA and DHA (and age, sex, smoking), heredity accounted for over one-third of the variability in the Omega-3 Index.174 There is increasing evidence to suggest that specific genotypes (e.g., apolipoprotein E, fatty acid desaturases, and peroxisome proliferator-activated receptors) modulate the response to EPA and DHA intake.175 Although the literature on genetic determinants of the response to increased EPA and DHA intake is very limited, with further research, genotyping may become a useful tool to determine optimal intakes for certain populations. A recent meta-analysis of genome-wide association studies identified genetic loci responsible for variations in phospholipid n-3 PUFA levels.176 Common genetic variations in the n-3 PUFA metabolic pathway may explain some of the individual variability in response to EPA and DHA intake; thus, additional research could help elucidate the role of genetics in EPA and DHA metabolism.

Maintaining and updating food composition databases (i.e., the USDA National Nutrient Database for Standard Reference162) is important in assessing n-3 PUFA intake. For example, NHANES, the major source of current dietary intake data for the US population, relies on this USDA database to accurately calculate nutritional values. The current version, Release 25, issued in September 2012, expanded the n-3 PUFA composition data for foods and dietary supplements. Although this improved the accuracy of n-3 PUFA intake estimates, challenges remain in accurately determining intake. A limitation of food records is that individuals frequently underestimate their intake. For this reason, a reliable and validated biomarker of n-3 PUFA status is necessary to validate dietary intake data.

Biomarkers of EPA and DHA intake

In the previous DRI report, the IOM acknowledged the biological potency of EPA and DHA, yet also indicated there was a lack of accepted biomarkers of intake for EPA and DHA.6 Significant progress has been made since the DRI report, and several markers of EPA and DHA intake are now available, including levels in plasma, erythrocytes, and adipose tissue. The use of blood markers of fatty acid intake has made it possible to evaluate outcome measures related to disease. Plasma levels of fatty acids reflect intake over the past few days, whereas adipose tissue levels of fatty acids are more reflective of long-term fatty acid intake.177,178 The erythrocyte content of EPA and DHA, i.e., the Omega-3 Index, is a useful biomarker of n-3 PUFA status. It reflects tissue levels and correlates with intake. The use of the Omega-3 Index has made it possible to evaluate EPA and DHA status relative to diseases. Harris and von Schacky179 proposed that a low score on the Omega-3 Index be considered a risk factor for CHD death. A low Omega-3 Index is associated with an increased risk of nonfatal acute coronary syndromes.180 The Omega-3 Index is highly correlated with cardiac membrane EPA and DHA levels (r = 0.81, P < 0.0001).181 Measuring EPA and DHA content of erythrocyte membranes (as opposed to the content in whole plasma or plasma phospholipids) provides insight into longer-term intake versus short-term intake.182 Utilizing a standardized assessment method, such as the Omega-3 Index, is important for assessing n-3 HUFA status and the biological effects of EPA and DHA intake.

Safety of increased EPA and DHA intake

There are some concerns over excessive bleeding at higher doses of n-3 HUFAs (≥3 g/day), even though several studies have examined this question and have uniformly reported no increased risk of clinically significant bleeding.23 In order to obtain ≥3 g/day of n-3 HUFAs, one would have to consume multiple servings of seafood each day, an achievable yet uncommon practice in the United States. However, as food fortification continues to increase, concerns over excessively high n-3 HUFA intakes could become a more relevant topic for discussion. In 2011, the Norwegian Scientific Committee for Food Safety conducted a safety review of EPA and DHA and found no adverse effect on bleeding time with levels as high as 6.9 g/day.69 More recently, the European Food Safety Authority concluded that intakes up to approximately 5 g/day of n-3 PUFAs do not appear to increase the risk of bleeding complications or affect glucose homeostasis, immune function, or lipid peroxidation, provided that the oxidative stability of the EPA and DHA is guaranteed.24 While additional research will continue to clarify the effect of EPA and DHA supplementation on bleeding time, current evidence indicates that n-3 HUFAs do not increase the risk of adverse bleeding episodes.

Consumption of fish raises the issue of human exposure to methylmercury, a toxic form of mercury found in long-lived fish and top-level predators, such as king mackerel, swordfish, shark, tilefish, and albacore tuna. In 2004, the FDA and the Environmental Protection Agency issued a joint advisory for women who may become pregnant and for pregnant women, nursing mothers, and young children to avoid certain types of fish high in mercury.183 This message should not discourage individuals from eating fish; rather, it should encourage the replacement of high-mercury fish with low-mercury fish. Moreover, fish oil supplements contain little to no mercury.184 A recent analysis of prospective cohort studies examined the combined effects of methylmercury and n-3 PUFA on the risk of myocardial infarction in middle-aged men, reporting that eating fish high in n-3 PUFAs and low in methylmercury was associated with a reduced risk of MI.185 Risk-benefit analyses indicate that lowering fish consumption would have serious public health consequences.99,185187

Research Needs

A large body of convincing evidence supports a CVD benefit related to increased intakes of EPA and DHA, and there is strong justification for setting a DRI for these fatty acids. However, much remains to be done to establish a DRI for EPA and DHA. The following specific areas of research would help further characterize and define optimal EPA and DHA intake: 1) improve understanding of the requirements for the individual long-chain n-3 fatty acids and the factors that affect their conversion and biological function; 2) identify the appropriate endpoints to use for establishing a DRI for n-3 fatty acids; 3) conduct intervention trials to determine the role EPA and DHA play in the primary prevention of CVD in the general population; 4) elucidate the dose-response relationship between dietary intakes and various biomarkers; 5) improve understanding of differences between the individual n-3 PUFAs (ALA, stearidonic acid, EPA, DPA, DHA) with respect to health outcomes, conversion efficiency, and metabolism; 6) expand the evidence base about the impact of EPA and DHA intake on other health outcomes, such as cancer and immune function; 7) improve understanding of the effects of genetic variation and epigenetic regulation of EPA and DHA on health outcomes; and 8) characterize the safety issues related to different EPA and DHA intakes in total and in at-risk populations.


Numerous recommendations for EPA and DHA intakes have been made by several national organizations and expert committees, largely based on data supporting health outcomes for CVD and cognitive development. It is clear, based on the extent and nature of the current body of evidence, that there is a need to reevaluate the establishment of a DRI for EPA and DHA. A minimum daily intake of 250–500 mg of EPA and DHA for healthy adults appears to be the general consensus among the scientific community; however, until a DRI is established, consumers and health professionals will continue to be uncertain as to how to interpret various recommendations for appropriate EPA and DHA intake. Clearly, establishing a DRI for EPA and DHA is a large undertaking that will require input from many experts across multiple fields. Nonetheless, there is overwhelming evidence that justifies the public health importance of establishing a DRI for EPA and DHA.


Funding. MRF is supported by an AOCS Thomas H. Smouse Memorial Fellowship Award.

Declaration of interest. WSH is a member of the scientific advisory boards of Omthera and Aker Biomarine and has been a consultant to Monsanto, Acasti, GlaxoSmithKline (GSK), and Amarin. He was on the speakers' bureau for Reliant and GSK. During his career, he has been the principal investigator on five National Institutes of Health grants dealing with n-3 fatty acids. He is currently the president of OmegaQuant Analytics, LLC, and is a senior scientist at Health Diagnostic Laboratory, Inc., two companies that offer blood omega-3 fatty acid testing. MRF and PKE have no relevant interest to disclose.


View Abstract