Dietary and Pharmacological Fatty Acids and Cardiovascular Health

Abstract

Context

The effects of dietary intake of different fatty acids and pharmacological use of fatty acids, specifically long-chain n-3 polyunsaturated fatty acids (LC n-3 PUFAs), on cardiovascular health and atherosclerotic cardiovascular disease (ASCVD) prevention have been examined in a large number of observational studies and clinical trials. This review summarizes recent data and discusses potential mechanisms.

Evidence acquisition

The review is based on the authors’ knowledge of the field supplemented by a PubMed search using the terms seafood, fish oil, saturated fatty acids, omega-3 fatty acids, eicosapentaenoic acid, docosahexaenoic acid, polyunsaturated fatty acids, monounsaturated fatty acids, and ASCVD.

Evidence synthesis

We mainly discuss the recent clinical trials that examine the effects of different types of dietary fatty acids and pharmacological use of n-3 PUFA products on ASCVD prevention and the potential mechanisms.

Conclusions

While replacement of dietary saturated fat with unsaturated fat, polyunsaturated fat in particular, or intake of LC n-3 PUFA–rich seafood has generally shown benefit for ASCVD prevention and is recommended for cardiovascular benefits, data on effects of n-3 PUFA products on ASCVD health are inconsistent. However, recent clinical trials support benefits of prescription EPA in ASCVD prevention. n-3 PUFAs may contribute to ASCVD prevention through multiple mechanisms, including lowering plasma triglyceride levels, anti-inflammatory effects, antithrombotic effects, and effects on endothelial function.

Atherosclerotic cardiovascular disease (ASCVD) remains the leading cause of death and disability worldwide (1). Both diet and medication have been used for prevention and treatment of ASCVD. In this review, we discuss recent updates on dietary fat, including saturated fatty acids (SFAs), polyunsaturated fatty acids (PUFAs), and monounsaturated fatty acids (MUFAs) (Fig. 1), and pharmacological fatty acids, specifically n-3 PUFAs, in ASCVD prevention and treatment.

Figure 1.

Chemical Structure of Key Fatty Acids. From: (3). Abbreviations: DHA, docosahexaenoic acid; DPA, docosapentaenoic acid; EPA, eicosapentaenoic acid.

Dietary Fat and ASCVD Prevention

SFAs

In the United States, SFAs account for approximately 11% of daily energy intake on average (2). Studies show that replacing dietary carbohydrates with SFAs increases both low-density lipoprotein cholesterol (LDL-C) and high-density lipoprotein cholesterol (HDL-C) without significantly changing total cholesterol/HDL-C ratio (4). In contrast, replacing SFAs with PUFAs and MUFAs decreases total cholesterol and LDL-C but only slightly decreases HDL-C, resulting in decreased total cholesterol/HDL-C ratio (4). In prospective observational studies, replacing SFAs with carbohydrates, especially carbohydrates from refined grains and added sugars, did not significantly reduce coronary heart disease (CHD) risk (5–7), whereas replacing SFAs with whole grains or unsaturated FAs, PUFAs in particular, did reduce CHD risk (7). Replacing SFAs with PUFAs (mainly linoleic acid [LA]) or MUFAs also reduced CVD and all-cause mortality (8). However, SFA consumption was not associated with total stroke (9) or ischemic stroke (10) in meta-analyses of cohort studies.

PUFAs

n-3 PUFAs.

n-3 PUFAs include alpha-linolenic acid (18:3n-3), which is derived from plants, and those derived from animals, particularly seafood (fish oil), mainly eicosapentaenoic acid (EPA; 20:5n-3), docosahexaenoic acid (DHA; 22:6n-3), and docosapentaenoic acid (22:5n-3) (11–17). Although alpha-linolenic acid appeared to have cardiovascular benefits in observational studies, overall evidence is inconclusive (17–21). In contrast, seafood intake has been negatively correlated with CHD risk in primary prevention (22, 23). In a primary-prevention meta-analysis, risk for acute coronary syndrome was 22% lower in participants who ate seafood ≥4 times weekly than in participants who ate seafood <1 time monthly, with higher consumption associated with greater benefit (23).

In addition, the association of seafood n-3 PUFAs with recurrent events in individuals with CHD has been assessed in observational studies. In the Cardiovascular Health Study (24), plasma phospholipid n-3 PUFA level was inversely associated with mortality, and red blood cell content of EPA plus DHA was inversely correlated with mortality in the Framingham Heart Study (25), Women’s Health Initiative Memory Study (26, 27), and Heart and Soul Study (28), with similar results in several meta-analyses (29–31).

In the secondary-prevention Diet and Reinfarction Trial (12), 2033 men with recent myocardial infarction (MI) were randomized to receive or not to receive dietary advice on fish (2 servings of fatty fish per week; ~20% took EPA + DHA 855 mg/day), fat (total fat intake <30% of calories, PUFA/SFA ratio 1.0), and fiber (cereal fiber intake 18 g/day). At 2-year follow-up, no significant effects were seen with reduced total fat or increased cereal fiber, but patients randomized to increased fish had significantly lower total mortality (relative risk [RR] 0.71; 95% confidence interval [CI], 0.54–0.93) and CHD mortality (RR 0.84; 95% CI, 0.66–1.07) than those without fish advice.

A meta-analysis showed a negative association of high fish intake with total stroke risk (32). The inverse association was stronger for ischemic than for hemorrhagic stroke.

n-6 PUFAs.

The major n-6 PUFAs are LA and arachidonic acid. LA, found mainly in vegetable oils, nuts, and seeds, is the predominant dietary n-6 PUFA, whereas arachidonic acid, found mainly in red meat, eggs, algae, and fish oil, has low consumption (~100 mg/day compared with 15 g/day of LA). Controlled feeding trials, prospective cohort studies, and randomized clinical trials demonstrate that high intake of n-6 PUFAs, predominantly LA in substitution for SFAs, was associated with reduced CHD risk (33–38). Consistently, a recent meta-analysis of 30 prospective observational studies, including 68659 participants, indicated that higher circulating and tissue levels of n-6 PUFAs, LA in particular, were associated with reduced total CVD, cardiovascular mortality, and ischemic stroke (39). Another pooled analysis, which included 39740 participants, showed that higher percentages of LA biomarkers in total fatty acids in individual lipid compartments (phospholipids, plasma, cholesteryl esters, and adipose tissue) were correlated with reduced risk for type 2 diabetes; in contrast, arachidonic acid biomarkers were not significantly associated with type 2 diabetes risk (40).

MUFAs

The most common MUFA in human diets is oleic acid, accounting for approximately 90% of all dietary MUFAs (41). MUFAs comprise approximately 16% to 29% of total energy intake in a typical Mediterranean diet, mostly from olive oil, and epidemiological studies found low CHD prevalence in Mediterranean populations despite high total fat intake (42). MUFA intake improves a number of CVD risk factors including lipid profile, insulin sensitivity, inflammatory markers, and thrombogenic markers (4, 43–46). Whereas early prospective cohort studies on the association of MUFA intake with CHD risk were inconsistent (6–8, 31, 47–49), in part because of confounding variables such as concomitant SFA intake in common foods, a more recent study showed that substituting MUFAs for carbohydrates reduced risk for CVD and total mortality, and during the 32-year follow-up, the major food source of FAs in this study changed from red meat to olive oil and nuts (8). Olive oil intake was inversely associated with CVD risk in several studies. Increasing olive oil intake by 25 g/day reduced CVD risk by 18% in a meta-analysis of 9 studies including 1 clinical trial (50). MUFA intake was associated with reduced risk for hemorrhagic stroke but not total or ischemic stroke in a meta-analysis of 10 prospective cohort studies including more than 300000 individuals (51).

In the Prevención con Dieta Mediterránea study, Mediterranean-type diets supplemented with extra virgin olive oil or mixed nuts significantly lowered CVD events by 31% or 28%, respectively, compared with a low-fat diet, as primary prevention in individuals at high CVD risk at 4.8-year median follow-up (52). This study provides the strongest evidence to date for the benefit of dietary MUFAs for primary prevention of ASCVD in high-risk individuals.

Taken together, these studies indicate that reducing SFA intake and replacing SFAs with PUFAs and MUFAs are associated with reduced rates of CVD and total mortality. In contrast, replacement of SFAs with refined carbohydrates and sugars showed no benefit on lowering CVD risk.

Accordingly, the American Heart Association recommends lowering intake of SFAs and replacing SFAs with unsaturated FAs, especially PUFAs, to reduce risk of CVD and also to include 1 to 2 seafood meals per week to reduce risk of CHD, ischemic stroke, and sudden cardiac death (53). Consistently, the 2015–2020 Dietary Guidelines for Americans (2) and the Scientific Report of the 2015 US Dietary Guidelines Advisory Committee (54) recommend using nonhydrogenated vegetable oils high in unsaturated fats and relatively low in SFAs (soybean, corn, olive, canola) instead of animal fats or tropical oils (palm, palm kernel, coconut) (Fig. 2) (2). Instead of reducing total fat, the type of fat should be optimized. At least 2 servings (3.5 oz per serving) of seafood should be consumed weekly to provide approximately 250 mg/day EPA plus DHA (55) and replace other animal sources of protein (Table 1). While salmon, swordfish, and mussel contain high amounts of EPA and DHA, some other commonly consumed fish, such as tilapia, have very little EPA and DHA.

Figure 2.

Fatty Acid Profiles of Common Fats and Oils. * Coconut, palm kernel, and palm oil are plant derived but are solid or semisolid at room temperature because of high saturated-fat content and are considered solid fats for nutrition. ** Shortening may be made from partially hydrogenated vegetable oil, which contains trans fatty acids. From: (2).

Table 1.

Major Fish and Other Seafood Sources of EPA and DHA

Fish/Seafood	Measure	EPA, g	DHA, g	EPA + DHA, g
Cod, Atlantic, canned, solids, and liquid	3 oz	0.003	0.13	0.13
Crab, queen, cooked, moist heat	3 oz	0.282	0.123	0.41
Flatfish (flounder and sole species), cooked, dry heat	1 fillet	0.213	0.168	0.38
Flatfish (flounder and sole species), raw, boneless	1 oz	0.039	0.031	0.07
Grouper, mixed species, cooked, dry heat	3 oz	0.03	0.181	0.21
Grouper, mixed species, raw	3 oz	0.023	0.187	0.21
Haddock, raw	3 oz	0.036	0.076	0.11
Herring, Atlantic, raw, boneless	1 oz	0.201	0.244	0.45
Herring, Pacific, raw	3 oz	0.824	0.586	1.41
Mackerel, jack, canned, drained solids, boneless	1 oz	0.123	0.226	0.35
Mackerel, Spanish, raw	3 oz	0.28	0.86	1.14
Mahi-mahi, cooked, dry heat	3 oz	0.022	0.096	0.12
Mussel, blue, raw	1 cup	0.282	0.38	0.66
Ocean perch, Atlantic, cooked, dry heat	1 fillet	0.037	0.093	0.13
Oyster, Eastern, cooked, breaded and fried	3 oz	0.172	0.185	0.36
Oyster, Eastern, farmed, raw	3 oz	0.16	0.173	0.33
Oyster, Eastern, wild, cooked, moist heat	3 oz	0.3	0.23	0.53
Oyster, Pacific, raw	1 medium	0.219	0.125	0.34
Pike, northern, cooked, dry heat	3 oz	0.036	0.081	0.12
Pollock, Alaska, cooked	3 oz	0.088	0.193	0.28
Roe, mixed species, cooked, dry heat	1 oz	0.357	0.495	0.85
Salmon, chum, cooked, dry heat	3 oz	0.254	0.429	0.68
Salmon, coho, wild, cooked, moist heat	3 oz	0.462	0.706	1.17
Salmon, pink, canned, drained solids	3 oz	0.284	0.632	0.92
Salmon, pink, canned, with bone and liquid	3 oz	0.718	0.685	1.4
Spiny lobster, mixed species, cooked, moist heat	3 oz	0.29	0.118	0.41
Swordfish, cooked, dry heat	3 oz	0.108	0.656	0.76
Tilapia, raw	1 fillet	0.006	0.1	0.11
Trout, mixed species, cooked, dry heat	1 fillet	0.161	0.42	0.58
Trout, rainbow, farmed, cooked, dry heat	1 fillet	0.184	0.437	0.62
Tuna, skipjack, fresh, cooked, dry heat	3 oz	0.077	0.201	0.28
Tuna, white, canned in oil, drained solids	3 oz	0.056	0.151	0.21

From: (55).

Effects and Mechanisms of n-3 PUFA on ASCVD Risks

Unsaturated FAs, LC n-3 PUFAs in particular, have been studied for their effects on CVD risk and tested in clinical trials for CVD prevention. Tissue culture, animal, and human studies indicate that n-3 PUFAs have multiple effects that may benefit CVD prevention (56).

Reductions of plasma triglycerides

n-3 PUFAs, particularly high-dose prescription formulations, lower plasma triglyceride (TG) levels (57–60). This effect may result from reduced hepatic very-low-density lipoprotein synthesis, which involves multiple mechanisms (56) of which decreased de novo lipogenesis appears to be predominant (61–64); other mechanisms include increased fatty acid β-oxidation, decreased hepatic nonesterified fatty acid delivery, reduced hepatic lipogenic enzyme activity, and increased hepatic synthesis of phospholipids in place of TGs (61, 65–68). The TG-lowering effect of n-3 PUFAs is generally linear and dose dependent but varies between individuals and is usually greater in individuals with higher TG levels at baseline (69, 70).

Anti-inflammatory effect

Inflammation contributes to ASCVD. The recent Canakinumab Anti-inflammatory Thrombosis Outcomes Study showed benefits of a therapy targeting interleukin-1β in secondary prevention of ASCVD (71). n-3 PUFAs lower several inflammatory markers associated with CVD risk. In clinical studies in patients with hypertriglyceridemia, prescription EPA as compared with placebo reduced plasma high-sensitivity C-reactive protein (hs-CRP), lipoprotein-associated phospholipase A2, and oxidized LDL levels and arachidonic acid/EPA ratio (72–77). TG reductions may contribute to decreasing inflammatory markers and other effects with n-3 PUFAs. However, n-PUFAs may also exert these effects independent of TG reductions. In our and others’ studies, short-term (2-week) treatment with EPA alone, with mild effects on plasma TG levels, improved monocyte phenotypes in individuals with hypertriglyceridemia (57, 78). Compared with DHA, EPA reduces the expression of genes linked to CVD, including those of the interferon pathway (79), 3',5'-cyclic adenosine monophosphate–responsive element protein 1, and hypoxia inducible factor 1 (79) and decreases production of inflammatory markers such as tumor necrosis factor–α and interleukin-1β in macrophages with lipopolysaccharide stimulation (80). All of these effects are involved in TG-independent benefits of EPA treatment on ASCVD events as revealed by the recent Reduction of Cardiovascular Events with EPA–Intervention Trial (REDUCE-IT) described below.

Antithrombotic effect

Hyper-reactivity of platelets leading to thrombus formation is the major cause for MI. Therefore, antiplatelet and antithrombotic therapies are effective in reducing ASCVD mortality. n-3 PUFAs reduce platelet aggregation and may therefore have antithrombotic effects, likely because of their incorporation into platelet membrane (see the following sections) and their role in reducing production and signaling of thromboxane A2, which plays a crucial role in platelet activation (81–83). Indeed, animal studies indicated that n-3 PUFA supplementation inhibited platelet activation and attenuated thrombus formation (84, 85). However, early human trials showed inconsistent effects of n-3 PUFA consumption on platelet aggregation and coagulation factors (86, 87). More recent studies indicated that adding n-3 PUFAs potentiated platelet response to clopidogrel in patients treated with combined aspirin and clopidogrel antiplatelet therapy after percutaneous coronary intervention and decreased lipoprotein-associated phospholipase A2, also known as platelet-activating factor acetylhydrolase, in patients with stable angina undergoing percutaneous coronary intervention (88–90). Overall, at the pharmacological dose of 4 g/day, antithrombotic effects do not appear to be the major mechanism by which n-PUFAs lower ASCVD risk, although subtle effects cannot be excluded.

Effect on endothelial function

Endothelial cell dysfunctions play a crucial role in development of ASCVD. Earlier tissue culture and animal studies indicate that n-3 PUFA improves endothelial function and reduces expression of adhesion molecules, thereby decreasing endothelial-leukocyte interactions and subsequent leukocyte infiltration (91), a key step in atherosclerosis. Several clinical studies showed that n-3 PUFAs improve flow-mediated vasodilation (92–95), a common noninvasive measure of endothelial function, and lower circulating levels of endothelial dysfunction markers such as E-selectin, vascular cell adhesion molecule–1, and intercellular adhesion molecule–1 (96–98). These endothelial function–improving effects may also contribute to n-3 PUFA–mediated protection against ASCVD.

Molecular Mechanisms

The molecular mechanisms whereby n-PUFAs exert the above effects are multifactorial and not fully understood. First, n-3 PUFAs are incorporated into the cell and organelle membrane and alter membrane lipid environments and fluidity, which are involved in various cellular functions such as signal transduction and protein trafficking (99–102) and may contribute to the observed beneficial effects of n-3 PUFAs on inflammation, platelet function, and monocyte phenotypes in humans with hypertriglyceridemia (78). Second, formation of specialized proresolving mediators derived from n-3 PUFAs such as EPA-derived E-series resolvins and DHA-derived D-series resolvins and protectins may play an important role in the resolution of inflammation (103–107). Third, n-3 PUFAs reduce the generation of intracellular secondary messengers such as diacylglycerol and ceramide, resulting in decreased inflammatory response (108). Further, n-3 PUFAs bind to and activate several nuclear receptors and transcription factors (109, 110), which are involved in regulating expression of multiple genes that participate in various functions related to ASCVD such as inflammation, glucose–insulin homeostasis, lipid metabolism, and production of adipocytokines (111–116). Activation of peroxisome proliferator-activated receptor–gamma by n-3 PUFAs reduces nuclear factor–κB translocation to the nucleus, thereby decreasing generation of multiple inflammatory cytokines (117).

Clinical Trials of Pharmacological LC n-3 PUFAs for ASCVD Prevention

Because of their effects as discussed above, n-3 PUFAs have been used in numerous clinical trials (12, 118–123) for ASCVD prevention (Table 2). However, considerable debate remains about the potential cardiovascular benefits of n-3 PUFAs.

Table 2.

Reviewed Outcomes Trials of n-3 PUFAs

Trial	Population	N	Follow-up, y	Interventions	Baseline Statin Use, %	Baseline LDL-C, mg/dL	Baseline TG, mg/dL	Primary Endpoint	Risk Ratio (95% CI)	Adverse Eventsa
Diet and/or supplement
DART (12)	Recent MI (men)	2033	2	Fatty fish 2 servings/week or supplement EPA + DHA 855 mg/d; total fat 30% of calories, PUFA/ SFA ratio 1.0; cereal fiber 18 g/d; no diet advice	NR	TC 250	NR	Total mortality	RR 0.71 (0.54–0.93)	NR
								CHD mortality	RR 0.84 (0.66–1.07)
EPA + DHA (low dose)
GISSI-P (118)	Recent MI	11,324	3.5	EPA + DHA 850 mg/d; vitamin E 300 mg/d; n-3 + vitamin E; no supplement	Cholesterol-lowering agents: 5 (end of study: 46)	137	162	Death or nonfatal MI or nonfatal stroke	RR 0.90 (0.82–0.99) [2-way analysis]; RR 0·85 (0.74–0.98) [4-way analysis]	GI disturbance, nausea
GISSI-HF (119)	Chronic HF	6975	3.9	EPA + DHA 850 mg; placebo (unspecified)	23	TC 188	126	Death	HR 0.91 (0.833–0.998)	GI disturbance
								Death or CV hosp	HR 0.92 (0.849–0.999)
OMEGA (124)	Recent MI	3851	1	DHA + EPA 840 mg; placebo	94–95	NR	NR	SCD	OR 0.95 (0.56–1.60)	Neoplasms (19 vs 8), cardiac device therapeutic procedures (16 vs 2)
Alpha-omega-3 (18)	Prior MI	4837	3.3	400 mg EPA + DHA ± 2 g ALA; 2 g ALA; placebo	Baseline lipid- lowering agents: 85–87	99–102	144–150	Fatal/nonfatal CV events or revasc	HR: 1.01 (0.87–1.17)	GI symptoms, prostate cancer incidence, death
SU.FOL.OM3 (125)	Recent acute coronary or cerebral ischemic event	2501	4.7	EPA + DHA 600 mg; placebo + vitamin B	Baseline lipid- lowering agents: 83–87	101–104	108 (no coronary event); 125 (coronary event)	MACE	HR 1.08 (0.79–1.47)	GI disturbance, nausea, cutaneous reactions
ORIGIN (126)	Dysglycemia or high-risk CVD	12,536	6.2	EPA + DHA 840 mg; placebo (olive oil)	53–54	112	140–142	CV death	HR 0.98 (0.87–1.10)	NR
ORIGINALE (127)	ORIGIN participants	4771	8.9	EPA + DHA 840 mg; placebo (olive oil)	57–59	108–109	142	CV death	HR 0.98 (0.88–1.09)	NR
Risk & Prevention (128)	High-risk CVD	12,513	5	EPA + DHA 850 mg; placebo	41	132	150	CV death or CV hosp	HR 0.97 (0.88–1.08)	GI symptoms
ASCEND (120)	Diabetes without CVD	15,480	7.4	DHA + EPA 840 mg; placebo (olive oil)	75	TC 161	NR	Nonfatal MI, nonfatal stroke, TIA, or vascular death	RR 0.97 (0.87–1.08)	Similar between groups
VITAL (121)	Healthy, no history of CVD	25,871	5.3	DHA + EPA 840 mg; placebo	Cholesterol-lowering agents: 37	NR	NR	MI, stroke, or CV death	HR 0.97 (0.85–1.12)	GI symptoms
EPA (high dose)
JELIS (122)	TC >250 mg/dL	18,645	4.6	EPA 1800 mg + statin; statin only	97	182	151	Sudden cardiac death, fatal or nonfatal MI, unstable angina, or revasc	HR 0.81 (0.69–0.95)	GI, skin, hemorrhage (cerebral, fundal, epistaxis, subcutaneous) (0.6% vs 1.1%, P = 0.0006)
REDUCE-IT (123)	CVD or diabetes + additional CVD risk	8179	4.9	EPA 4000 mg; placebo (mineral oil)	100	75	216	CV death, nonfatal MI, nonfatal stroke, revasc, or unstable angina	HR 0.75 (0.68–0.83)	Hosp for atrial fibrillation and atrial flutter (3.1% vs 2.1%, P = 0.004), serious bleeding events (2.7% vs 2.1%, P = 0.06)

aNot significantly different between treatment groups except as noted.

Abbreviations: ALA, alpha-linolenic acid; ASCEND, A Study of Cardiovascular Events in Diabetes; CHD, coronary heart disease; CI, confidence interval; CV, cardiovascular; CVD, cardiovascular disease; DART, Diet and Reinfarction Trial; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; GI, gastrointestinal; GISSI-P, Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico Prevenzione; GISSI-HF, GISSI–Heart Failure; HF, heart failure; hosp, hospitalization; HR, hazard ratio; JELIS, Japan EPA Lipid Intervention Study; LDL-C, low-density lipoprotein cholesterol; MACE, major adverse cardiac events; MI, myocardial infarction; NR, not reported; OR, odds ratio; ORIGIN, Outcome Reduction with an Initial Glargine Intervention; ORIGINALE, Outcome Reduction with an Initial Glargine Intervention Legacy Effects; PUFA, polyunsaturated fatty acid; REDUCE-IT, Reduction of Cardiovascular Events with Icosapent Ethyl–Intervention Trial; revasc, revascularization; RR, relative risk; SCD, sudden coronary death; SFA, saturated fatty acid; SU.FOL.OM3, Supplementation with Folate, Vitamin B6 and B12 and/or Omega-3 Fatty Acids trial; TC: total cholesterol; TG, triglyceride; TIA, transient ischemic attack; VITAL, Vitamin D and Omega-3 Trial; y, years.

Outcomes trials of mixtures of EPA and DHA at low doses

The Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico Prevenzione (GISSI-P) study assessed effects of combined EPA and DHA ethyl esters 850 mg/day on recurrent ASCVD events in 11324 patients with recent MI (118). After a median follow-up of 3.5 years, the primary endpoint (death, nonfatal MI, and nonfatal stroke) was reduced by 15% (95% CI, 2%–6%; P = 0.02) with n-3 PUFAs compared with placebo. Of note, a low proportion (5%) of patients used concomitant statins at baseline, although 46% of patients used statins at the end of the study.

Similarly, a mixture of EPA and DHA ethyl esters 850 g/day was evaluated in 6975 patients with chronic heart failure in the GISSI–Heart Failure (GISSI-HF) study (119), in which approximately 22% of patients used statin therapy at baseline. Median TG level was 126 mg/dL at baseline and decreased only slightly during treatment. After a median follow-up of 3.9 years, n-3 PUFAs significantly reduced the coprimary endpoints of all-cause mortality by 9% (P = 0.041) and all-cause mortality or hospital admission for cardiovascular reasons by 8% (P = 0.009) compared with placebo.

However, several later studies showed no significant efficacy of n-3 PUFAs in lowering ASCVD risk in patients on statin treatment. A meta-analysis of 10 trials, which included 77917 individuals, showed no significant differences between n-3 PUFA and control groups for CHD, stroke, revascularization, and major vascular events in individuals with previous CHD (60).

In the Outcome Reduction with an Initial Glargine Intervention (ORIGIN) trial (126), 12536 patients at high risk for CVD and with impaired fasting glucose, impaired glucose tolerance, or diabetes were randomized to receive 1 g/day of n-3 PUFAs (containing 465 mg of EPA and 375 mg of DHA) or 1 g/day of placebo (olive oil). At a median follow-up of 6.2 years, TG levels were reduced by 14.5 mg/dL more with n-3 PUFAs than placebo (P < 0.001). However, the n-3 PUFA group did not show significant decreases in the incidence of the primary outcome of cardiovascular death (P = 0.72), major vascular events (P = 0.81), death from any cause (P = 0.63), or death from arrhythmia (P = 0.26) when compared with the placebo group.

The ORIGIN and Legacy Effects (ORIGINALE) study (127) extended follow-up for an additional 2.7 years to examine potential posttreatment impact of n-3 PUFAs. Of the 4771 individuals with extended follow-up, 2368 had been randomized to n-3 PUFAs and 2403 to placebo. At the end of ORIGIN, 89% of the initial n-3 PUFA group and 1% of the placebo group (P < 0.001) were taking n-3 PUFAs, whereas only 11% and 9% of the respective groups (P = 0.04) were taking n-3 PUFAs at the end of ORIGINALE. As in the main trial, at extended follow-up no difference was seen between the initial randomized treatment groups for the primary outcome of cardiovascular death (P = 0.68) or the other cardiovascular outcomes assessed.

More recently, 2 large clinical trials of n-3 PUFAs for primary prevention of ASCVD, A Study of Cardiovascular Events in Diabetes (ASCEND) (120) and Vitamin D and Omega-3 Trial (VITAL) (121), also failed to show benefit on the primary endpoint. In ASCEND, 15480 patients with diabetes were randomized to receive combined EPA and DHA 840 mg/day or olive oil placebo (120). At baseline, approximately 75% of patients were on statin therapy, and non-HDL-C level was 113 mg/dL. After a mean follow-up of 7.4 years, no significant difference was observed in the rates of primary outcome events (nonfatal MI, nonfatal stroke, transient ischemic attack, and vascular death; RR 0.97; 95% CI, 0.87–1.08; P = 0.55) and all-cause mortality between the n-3 PUFA and placebo groups. However, vascular death was modestly reduced with n-3 PUFAs compared with placebo (respective event rates 2.4% and 2.9%; RR 0.81; 95% CI, 0.67–0.99). VITAL used combined EPA and DHA 840 mg/day with or without vitamin D3 supplementation in 25871 healthy elderly individuals (≥50-year-old men and ≥55-year-old women) (121). After a median follow-up of 5.3 years, n-3 PUFAs did not reduce the primary endpoint of major CVD events (MI, stroke, and CVD death) compared with placebo. However, subgroup analyses demonstrated a significant interaction based on median weekly fish consumption, supporting benefits of pharmacological n-3 PUFAs versus placebo among individuals consuming <1.5 servings of fish per week at baseline. Secondary analyses from this trial also showed significant reductions of total MI, fatal MI, number of percutaneous coronary interventions, and total CHD (a composite of MI, coronary revascularization, and CHD death) in the n-3 PUFA group. Subgroup analysis of the total MI secondary endpoint suggested that n-3 PUFA supplementation may have greater coronary benefits in African Americans (~20% of the study population) than in other racial groups in the study. However, because the primary endpoint was not significantly achieved, the results generated from the secondary analyses should be interpreted cautiously.

Outcomes trials of EPA alone

EPA and DHA have biochemical differences including their membrane locations, lipid interactions, and effects on platelet function, membrane bilayer structure, and formation of cholesterol crystalline domains (129–132). Outcomes trials have examined effects of EPA alone on ASCVD prevention, but no trials of DHA alone with ASCVD endpoints have been reported or (according to ClinicalTrials.gov) are ongoing.

The open-label Japan EPA Lipid Intervention Study (JELIS) compared purified EPA at 1.8 g/day and low-intensity statin with low-intensity statin alone in 18645 Japanese patients with hypercholesterolemia (mean baseline total cholesterol ~275 mg/dL; median baseline TG levels ~154 mg/dL) (122). At a mean follow-up of 4.6 years, the primary endpoint of major coronary events (sudden cardiac death, fatal and nonfatal MI, unstable angina pectoris, angioplasty, stenting, and coronary artery bypass grafting) was significantly reduced by 19% in the EPA group compared with controls, and TG level was significantly decreased by 9% with EPA compared with 4% in controls; LDL-C levels were not significantly different between treatment groups (25% reduction from baseline in both groups). Of the component outcomes, unstable angina, but not sudden cardiac death or fatal MI, was significantly decreased with EPA. In subgroup analyses, reduction in risk for the primary endpoint was similar in primary and secondary prevention but was statistically significant only in secondary prevention (~20% of the study population). JELIS patients with high baseline TG level, low baseline HDL-C level, and impaired glucose metabolism had higher risk for CHD and greater benefit on CHD reduction with EPA (133, 134). Event reduction was associated with higher blood levels of EPA and higher ratios of EPA to arachidonic acid (135).

REDUCE-IT evaluated whether treatment with a highly purified EPA ethyl ester, icosapent ethyl, at 2 g twice daily (4 g/day), compared with placebo (mineral oil), reduces ASCVD events in statin-treated patients with relatively well-controlled LDL-C levels (>40 mg/dL and ≤100 mg/dL), high baseline TG levels (≥135 mg/dL and <500 mg/dL), and elevated risk for cardiovascular events (123). Of the 8179 participants randomized, 71% had ASCVD and 29% had diabetes with additional ASCVD risk factors but did not have established ASCVD; at baseline, 31% of patients took high-intensity statin, 62% took moderate-intensity statin, and 59% had diabetes. Baseline median (interquartile range) TG levels were 216 (176–272) mg/dL in the icosapent ethyl group and 216 (176–274) mg/dL in the placebo group; respective LDL-C levels were 74 mg/dL (62–88) mg/dL and 76 (63–89) mg/dL.

During a median follow-up of 4.9 years, primary endpoint events (first occurrence of cardiovascular death, nonfatal MI [including silent MI], nonfatal stroke, coronary revascularization, or unstable angina) occurred in 17% of icosapent ethyl patients and 22% of placebo patients (hazard ratio [HR] 0.75, 95% CI, 0.68–0.83; P < 0.001; absolute risk reduction [ARR] 4.8%) (123). Moreover, compared with placebo, the treatment group had significant reductions in secondary endpoint events including cardiovascular death (HR 0.80; 95% CI, 0.66–0.98; P = 0.03; ARR 0.9%), fatal or nonfatal MI (HR 0.69; 95% CI, 0.58–0.81; P < 0.001; ARR 2.6%), fatal or nonfatal stroke (HR 0.72; 95% CI, 0.55–0.93; P = 0.01; ARR 0.9%), unstable angina (HR 0.68; 95% CI, 0.53–0.87; P = 0.002; ARR 1.2%), and urgent or emergent revascularization (HR 0.65; 95% CI, 0.55–0.78; P < 0.001; ARR 2.5%), but all-cause mortality was not significantly different between groups (HR 0.87; 95% CI, 0.74–1.02).

In REDUCE-IT, icosapent ethyl treatment also improved lipids and other biomarkers. TG levels decreased by 18% (−39.0 mg/dL) at 1 year and by 22% (−45.0 mg/dL) at the final visit in the icosapent ethyl group, compared with a 2% (4.5 mg/dL) increase and 6% (−13.0 mg/dL) decrease in the placebo group at the respective timepoints. LDL-C decreased by 1% (−1.0 mg/dL) at both 1 year and the final visit with icosapent ethyl, compared with 11% (9.3 mg/dL) and 6% (5.7 mg/dL) increases with placebo at these timepoints. At the final visit, hs-CRP was reduced by 13% (−0.2 mg/L) with icosapent ethyl but increased by 30% (0.4 mg/L) with placebo.

In a subsequent analysis of REDUCE-IT, icosapent ethyl lowered the RR for total (first and subsequent) ischemic events included in the primary composite endpoint by 30%, with a 25% RR reduction for first events, 32% for second events, 31% for third events, and 48% for fourth and subsequent events (136). Total key secondary endpoint events (cardiovascular death, nonfatal MI, nonfatal stroke) were also reduced with icosapent ethyl compared with placebo (RR 0.72; 95% CI, 0.63–0.82; P < 0.0001).

Discussion of n-3 PUFA Outcomes Trials

Multiple reasons may have contributed to the discrepant results among the trials. The earlier GISSI trials were designed for secondary prevention and included low proportions of patients taking concomitant statins. ASCEND and VITAL were designed for primary prevention of ASCVD and included high proportions of patients on statins as did several other later trials that did not show significant efficacy of n-3 PUFAs (18,124,125,128,137,138).

In contrast, treatment with prescription EPA at high doses has shown great potential in ASCVD prevention, including in statin-treated patients with acceptable LDL-C levels. Despite different EPA doses in REDUCE-IT (4 g/day) and JELIS (1.8 g/day), circulating EPA levels were similar in these 2 trials (144 mg/L at 1 year in REDUCE-IT and 169 mg/L at 5 years in JELIS), which likely reflect greater fish consumption in the Japanese study. Treatment with EPA 4 g/day in the evaluation of the effect of two doses of AMR101 (Ethyl Icosapentate) on fasting serum triglyceride levels in patients with persistent high triglyceride levels (≥200 mg/dL and <500 mg/dL) despite statin therapy (ANCHOR) study resulted in a >6-fold increase in EPA and doubling of docosapentaenoic acid with no change in DHA levels (139). Prior studies involving EPA and DHA mixtures, including ASCEND and VITAL, used formulations containing much less EPA (460 mg/day in ASCEND and VITAL). The discontinued outcomes study to assess statin residual risk reduction with epanova in high CV risk patients with hypertriglyceridemia (STRENGTH) trial examining the efficacy and safety of high-dose (4 g/day) n-3 carboxylic acids containing EPA and DHA is expected to provide further insight into how n-3 PUFA formulation and dosage affect cardiovascular outcomes (140).

In addition, whereas ASCEND and VITAL focused on primary prevention, REDUCE-IT included a large proportion of secondary-prevention patients and higher TG levels at baseline, both higher-risk populations, which increases the power to detect risk reductions. Indeed, in REDUCE-IT, although no significant difference was observed between secondary and primary prevention, a trend toward greater risk reduction in secondary prevention suggests that higher-risk patients may benefit more from n-3 PUFA treatment.

TG levels at baseline were higher in REDUCE-IT (median 216 mg/dL) than in JELIS (median 153 mg/dL), and EPA led to greater TG reductions in REDUCE-IT than in JELIS. However, subgroup analysis from REDUCE-IT found consistent risk reduction with EPA in patients with baseline TG levels ≥150 mg/dL or <150 mg/dL as well as in patients with 1-year TG levels ≥150 mg/dL or <150 mg/dL, suggesting that EPA may have TG-independent benefits for ASCVD prevention. Nevertheless, REDUCE-IT showed significantly greater benefit in patients with baseline TG ≥200 mg/dL in conjunction with baseline HDL-C <35 mg/dL (HR 0.62; 95% CI, 0.51–0.77; compared with HR 0.79; 95% CI, 0.71–0.88 in patients without both lipid values; P = 0.04 for interaction). Similarly, in JELIS, patients with TG ≥150 mg/dL combined with HDL-C <40 mg/dL at baseline had a significantly greater risk for coronary events (HR 1.71; 95% CI, 1.11–2.64; P = 0.014) and a significantly greater risk reduction with EPA treatment (HR 0.47; 95% CI, 0.23–0.98; P = 0.043) compared with patients with neither of these dyslipidemias (134), consistently suggesting greater benefit of EPA treatment in higher-risk patients.

In addition to lowering TG, n-3 PUFAs have other effects that may benefit ASCVD prevention. The reduction of hs-CRP with EPA in REDUCE-IT substantiates previous reports on anti-inflammatory effects of EPA (141, 142). EPA also improves circulating monocyte phenotypes in individuals with hypertriglyceridemia (78). In addition, EPA may reduce atherothrombotic events by inhibiting platelet aggregation (81, 132), which may also explain the tendency for more bleeding events with EPA observed in REDUCE-IT. Additional analyses from this trial, including more information on the association of clinical event reduction with changes in EPA levels, lipids and lipoproteins, and inflammatory biomarkers, may provide more insight into the mechanism(s) of benefit with EPA treatment.

One limitation of REDUCE-IT was the use of mineral oil as a placebo (123). If mineral oil affected statin absorption in some placebo patients, this might have contributed to differences between treatment groups, but the small differences in LDL-C would not explain the observed 25% risk reduction, and a post hoc analysis suggested similar benefits regardless of whether there was an increase in LDL-C among placebo patients.

Safety of n-3 PUFAs

In clinical studies, gastrointestinal events were the main adverse effects with DHA-containing products (131, 143), whereas arthralgia was the main adverse event reported with EPA (130, 144). Some clinical studies showed that treatment with DHA-containing n-3 PUFA products was associated with increased LDL-C in individuals with severe hypertriglyceridemia (145, 146) while other studies involving subjects with moderate hypertriglyceridemia did not show this effect (147–155).

In REDUCE-IT, EPA treatment was associated with increased rates of atrial fibrillation (5.3% vs 3.9% with placebo) and peripheral edema (6.5% vs 5.0% with placebo). The mechanisms for these observations are unknown, but in vitro and animal studies demonstrated that n-PUFAs may affect ion channels of cardiac tissue (156–158). Previous observational studies and randomized trials that examined potential antiarrhythmic effects of n-3 PUFAs had inconsistent results (159). Some studies found trends toward increased atrial fibrillation with n-3 PUFAs (160, 161). In REDUCE-IT, the higher incidence of atrial fibrillation in the EPA-treated group was not associated with increased heart failure or stroke. Additional studies are needed to examine other potential consequences related to atrial fibrillation with EPA treatment.

Both JELIS and REDUCE-IT reported modest increases in bleeding events with EPA treatment compared with controls. Bleeding rates were lower in JELIS (1.1% with EPA; 0.6% in controls) than REDUCE-IT (2.7% with EPA; 2.1% with placebo); at baseline, antiplatelet therapy was used by 13% and 14% of the respective treatment groups in JELIS, compared with 79.7% and 79.1% of the respective groups in REDUCE-IT (136). Although no severe clinical bleeding events were observed in REDUCE-IT or other clinical trials of n-3 PUFAs, including in participants who used aspirin or warfarin and those who had surgery or percutaneous interventions (162–164), the potential interaction between EPA and antithrombotic medications warrants further investigation.

Conclusion

Accumulated evidence supports the recommendation of lowering intake of SFAs, replacing them with unsaturated FAs, especially PUFAs, and consuming seafood containing significant amounts of EPA and DHA at least 2 times per week (ideally ≥4 times per week (23)) to reduce the risk of ASCVD. Accordingly, a healthy diet for ASCVD prevention should include nonhydrogenated vegetable oils that are high in unsaturated FAs with a relatively low proportion of SFAs as well as seafood containing high amounts of LC n-3 PUFAs. With the increased popularity of low-carbohydrate, high-fat diets such as the ketogenic diet, it is important that health care providers educate patients on differential effects of FAs on lipids as MUFAs and PUFAs from plants (165) may avoid the frequent and highly variable LDL-C increases that may occur with the ketogenic diet (166). While statins alone or combined with ezetimibe or proprotein convertase subtilisin/kexin type 9 inhibitors reduce ASCVD risk by lowering LDL-C, other factors such as hypertriglyceridemia and inflammation still contribute residual ASCVD risk. Over the last few years, multiple trials of n-3 PUFAs have examined effects of various dosages and formulations in different patient populations, providing increasing evidence on the pharmacological and clinical effects of n-3 PUFAs in ASCVD prevention. In JELIS and REDUCE-IT, prescription EPA used in conjunction with statin therapy provided further benefit on reducing ASCVD events in high-risk patients compared with statin alone. On the other hand, n-3 PUFAs at low doses have not shown consistent efficacy in reducing ASCVD risk, particularly for primary prevention. Additional trials of n-3 PUFAs are ongoing and will provide further insights on the effectiveness of this class of medication in ASCVD prevention.

Glossary

Abbreviations

ARR

absolute risk reduction

ASCEND

A Study of Cardiovascular Events in Diabetes

ASCVD

atherosclerotic cardiovascular disease

CHD

coronary heart disease

CI

confidence interval

DHA

docosahexaenoic acid

EPA

eicosapentaenoic acid

FA

fatty acid

GISSI

Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico Prevenzione

HDL-C

high-density lipoprotein cholesterol

HR

hazard ratio

hs-CRP

high-sensitivity C-reactive protein

JELIS

Japan EPA Lipid Intervention Study

LA

linoleic acid

LC

long-chain

LDL

low-density lipoprotein cholesterol

MI

myocardial infarction

MUFA

monounsaturated fatty acid

ORIGIN

Outcome Reduction with an Initial Glargine Intervention

ORIGINALE

Outcome Reduction with an Initial Glargine Intervention Legacy Effects

PUFA

polyunsaturated fatty acid

REDUCE-IT

Reduction of Cardiovascular Events with EPA–Intervention Trial

RR

relative risk

SFA

saturated fatty acid

TG

triglyceride

VITAL

Vitamin D and Omega-3 Trial

Additional Information

Disclosure Summary: L.X. and H.W. have nothing to declare. C.M.B received grant/research support from Akcea, Amgen, Esperion, Novartis, Regeneron, and Sanofi-Synthelabo (all paid to Baylor College of Medicine) and consults for Akcea, Amarin, Amgen, Arrowhead, Astra Zeneca, Boehringer Ingelheim, Denka Seiken, Esperion, Intercept, Janssen, Matinas BioPharma Inc, Merck, Novartis, Novo Nordisk, Regeneron, and Sanofi-Synthelabo. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

Data Availability: Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.