Eicosapentaenoic Acid and Docosahexaenoic Acid: Are They Different?
Trevor A Mori, Ph.D., Professor, School of Medicine and Pharmacology, Royal Perth Hospital Unit, University of Western Australia, Perth, Australia.
Many of the cardiovascular benefits ascribed to omega-3 (n-3) fatty acids were initially attributed to eicosapentaenoic acid (EPA) rather than docosahexaenoic acid (DHA). It was suggested that some of the effects of EPA were due to its being a competitive inhibitor of arachidonic acid for the cyclooxygenase and lipoxygenase enzymes, leading to prostaglandins and leukotrienes with attenuated bioactivity compared with the respective arachidonic acid analogues. However, we now know that DHA has many important, independent cardiovascular benefits. A limiting factor in determining the individual effects of EPA and DHA has been the lack of sufficient quantities of purified EPA and DHA for human studies. Although a number of EPA- and DHA-enriched oils are available, most are also contaminated with other fatty acids. Consequently, only those trials in which oils with >90% purity have been used can be considered useful when assessing the individual effects of EPA and DHA on cardiovascular risk factors in humans. In controlled studies comparing EPA and DHA as purified supplements provided to overweight hyperlipidemic men and in treated-hypertensive type 2 diabetic individuals, we have shown that both have important cardiovascular benefits.
EPA and DHA are differentially incorporated into membrane lipids. EPA supplementation increased EPA and docosapentaenoic acid (DPA), but decreased DHA in plasma and platelet phospholipids.1-5 In contrast, DHA supplementation increased DHA and EPA, the latter demonstrating retro-conversion of DHA to EPA, and decreased DPA in plasma and platelet phospholipids.1-4
EPA and DHA are equally effective in reducing plasma triglycerides.2,3,5,6 Although neither alters total cholesterol, they have differential effects on HDL-cholesterol. DHA increased HDL-cholesterol by 4-17%5 whereas EPA did not. Additionally, DHA increased HDL2-cholesterol in dyslipidemic3 and type 2 diabetic patients,2 whereas HDL3-cholesterol was reduced after EPA.2,3 Although LDL-C was unchanged, DHA and not EPA increased LDL particle size.3 These data demonstrate the lipid-regulating effects of DHA are at least as important as those of EPA.
In animal studies, DHA was more effective than EPA in retarding the development of hypertension in spontaneously hypertensive rats (SHR).7 DHA, but not EPA, also inhibited ischemia-induced cardiac arrhythmias and was more effective than EPA in inhibiting thromboxane-like vasoconstrictor responses in the aortas of SHR.7 In humans, EPA did not alter blood pressure in healthy subjects8 or those with dyslipidemia,6 angina9 or type 2 diabetes mellitus.2 However, DHA but not EPA, significantly lowered 24-hour and awake ambulatory systolic and diastolic blood pressure in overweight hypercholesterolemic men.4 These effects were accompanied by significant improvements in endothelial and smooth muscle function in the forearm microcirculation with DHA, but not EPA.10 The mechanisms likely include changes in the release of nitric-oxide, ADP, endothelium-derived hyperpolarising factor and prostanoids. In contrast, in patients with variant angina, EPA improved vasomotion at coronary sites exhibiting a slight vasoconstriction, but it did not prevent the persistence of vasospasms at severely constricting sites.9 EPA and DHA both improved arterial compliance in patients with dyslipidemia.6
A reduction in heart rate by long-chain omega-3 fatty acids (n-3 LC-PUFAs) suggests a significant cardiac component associated with the antihypertensive effects possibly mediated by effects on autonomic nerve function or b-adrenoreceptor activity. We showed DHA, but not EPA, reduced 24-hour awake and asleep heart rate in overweight hyperlipidemic men.4 These effects were substantiated by others,8 including Woodman et al.2 in treated-hypertensive type 2 diabetic patients.
In a short-term uncontrolled study in healthy individuals, von Schacky C et al.11 first showed differential effects of EPA and DHA on platelet responsiveness. Both reduced ex-vivo platelet aggregation to collagen but only DHA attenuated ADP-stimulated platelet aggregation.11 In the only controlled study in humans, we showed EPA did not alter collagen- or PAF-induced ex-vivo platelet aggregation in type 2 diabetic patients.1 However, DHA reduced collagen-induced aggregation and platelet thromboxane B2 (TXB2) release, but had no effect on PAF-induced aggregation.1 The reduction in platelet TXB2 following DHA may be due to competitive inhibition of cyclooxygenase, inhibition of TXA2 synthetase or inhibition of TXA2 receptor function. In contrast, Park et al.12showed that mean platelet volume, a marker of platelet activation, was decreased by EPA, but not DHA. Platelet count was also increased by EPA and not DHA.12
Disparate effects on glycemic control in type 2 diabetic patients likely relate to the dose of n-3 LC-PUFAs, oral diabetic medication, presence of obesity and/or insulin resistance, presence of other conditions such as hypertension, not controlling subjects’ diets and the duration of intervention. In dyslipidemic men, we showed a borderline significant increase in fasting glucose after 4g/day EPA but no change with DHA.3 Fasting serum insulin significantly increased after DHA, but not EPA, relative to placebo.3 In patients with type 2 diabetes, fasting glucose was increased following 4g/day EPA or DHA relative to control, but insulin, C-peptide and HbA1c, and insulin secretion and insulin sensitivity were unchanged.2
In vitro studies showed DHA, but not EPA, decreased pro-inflammatory cytokine expression, cell-adhesion molecules and monocyte adhesion to endothelial cells.13 The resolvins from EPA and DHA and protectins from DHA are potent agonists promoting active resolution of inflammation. They provide another example of different mechanisms for the anti-inflammatory effects of EPA and DHA.14
Concern that n-3 PUFAs increase lipid peroxidation and oxidative stress is unfounded. We showed EPA and DHA were equally effective in reducing plasma and urinary F2-isoprostanes.15,16 F2-isoprostanes derive from non-enzymatic free radical oxidation of arachidonic acid in membrane lipids and are the most reliable biomarkers of in vivo lipid peroxidative damage. Reduced F2-isoprostanes following EPA or DHA likely relate to decreased leukocyte activation and the immunomodulatory actions of n-3 PUFAs.
EPA and DHA have many different yet complementary hemodynamic and anti-atherogenic properties. Human data suggest DHA may be more favorable in lowering blood pressure and improving vascular function, raising HDL-cholesterol and attenuating platelet function. However, EPA has important bioactive properties relevant to cardiovascular risk reduction. Further studies are needed to carefully assess the independent effects of EPA and DHA on other clinical and biochemical measures, and in other populations, before recommendations can be made with respect to the ratio of EPA to DHA in dietary supplements and food fortification. The greatest benefit on cardiovascular risk reduction is likely to be gained from a combination of both EPA and DHA. With the limited available data it is difficult to speculate what that proportion should be, but an approximate equal proportion of each would seem judicious. Exceptions may be situations such as in pregnancy-fetal development where DHA may be more important than EPA.
1 Woodman RJ, Mori TA, Burke V, et al. Atherosclerosis 2003;166:85-93.
2 Woodman RJ, Mori TA, Burke V, et al. Am J Clin Nutr 2002;76:1007-1015.
3 Mori TA, Burke V, Puddey IB, et al. Am J Clin Nutr 2000;71:1085-1094.
4 Mori TA, Bao DQ, Burke V, et al. Hypertension 1999;34:253-260.
5 Grimsgaard S, Bonaa KH, Hansen JB, et al. Am J Clin Nutr 1997;66:649-659.
6 Nestel P, Shige H, Pomeroy S, et al. Am J Clin Nutr 2002;76:326-330.
7 Mclennan P, Howe P, Abeywardena M, et al. Eur J Pharmacol 1996;300:83-89.
8 Grimsgaard S, Bonaa KH, Hansen JB, et al. Am J Clin Nutr 1998;68:52-59.
9 Yamamoto H, Yoshimura H, Noma M, et al. Jap Circ J 1995;59:608-616.
10 Mori TA, Watts GF, Burke V, et al. Circulation 2000;102:1264-1269.
11 von Schacky C, Weber PC. J Clin Invest 1985;76:2446-2450.
12 Park Y, Harris W. Lipids 2002;37:941-9466.
13 De Caterina R, Liao JK, Libby P. Am J Clin Nutr 2000;71:213S-223S.
14 Serhan CN, Savill J. Nature Immunology 2005;6:1191-7.
15 Mori TA, Woodman RJ, Burke V, et al. Free Rad Biol Med 2003;35:772-781.
16 Mas E, Woodman RJ, Burke V, et al. Free Rad Res 2010 44:983-990.