Omega-3s and Cardiovascular Health

Abstract

Background

Omega-3 (n-3) fatty acids have previously been shown to reduce the risk of cardiac events, cardiac death, and all-cause mortality in randomized controlled trials. However, recent data have challenged the benefits of n-3 fatty acids in the current era of optimal medical therapy.

Methods

We performed a literature review indicating important limitations that must be considered when interpreting the recent negative n-3 fatty acids trials.

Results

Our review found relative strengths and weaknesses of both the older and more recent studies, along with many possible explanations for the disparate results. The principal difference between the older and the more recent n-3 studies was a greater use of background optimal medical therapy that may have reduced the benefit from n-3s. Additionally, some of the more recent n-3 trials used relatively low doses or tested n-3 supplementation on top of a relatively high baseline intake of n-3s.

Conclusion

Despite the recent negative data about n-3 fatty acids, the overall evidence still supports the American Heart Association recommendation of 1 gram of eicosapentaenoic acid/docosahexaenoic acid per day for patients with coronary heart disease.

INTRODUCTION

Eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) (omega-3 polyunsaturated fatty acids [n-3-PUFAs]) are mainly present in marine fish oil. Ever since the low cardiovascular (CV) mortality rate in Greenland Eskimos, known for their high fish intake, came to light 28 years ago,¹ several studies have investigated the possible role of n-3-PUFAs in treating and preventing cardiovascular disease (CVD). Evidence from observational studies and randomized controlled trials (RCTs) regarding the role of fish oil in primary and secondary prevention of CVD has been promising. However, not all studies have shown consistent results. Many studies and metaanalyses report conflicting results that show positive and negative effects of n-3-PUFAs for prevention of CV events, CV death, and mortality.^2-4 Despite the conflicting data, current guidelines recommend 2 servings of fatty fish per week for the general population and 1 g/d of n-3-PUFAs for patients with coronary heart disease (CHD).³

Controversy regarding the efficacy of n-3-PUFAs in the primary and secondary prevention of CVD has led to recent debate. On a closer analysis of the current body of evidence, many of the discrepancies in the results of various trials can be explained on the basis of faulty study design (short follow-up, low baseline CV risk, lack of power, high baseline n-3-PUFA intake), inadequate dose of n-3-PUFAs used, and differences in study populations.⁴

Demographics and the baseline intake of fish in the study population are important points to consider. Many of the studies that have failed to demonstrate positive effects of n-3-PUFAs on CV outcomes have been conducted in populations with a high background level of fish intake.⁴ In such populations, maintaining a contrast between the n-3-PUFA intake levels of the intervention and the control groups is difficult. This hypothesis is supported by the low level of CV events in both arms of the studies showing a lack of benefit of n-3-PUFAs. One could argue that the maximum benefit of n-3-PUFAs was already present in both arms of the studies because of the high fish intake in the populations.

Several hypotheses have been presented regarding the role of n-3-PUFAs in preventing CVDs (Table 1).^5–88 The relative importance of these mechanisms for the reduced CV risk associated with replete n-3 status remains unclear.

Table 1.

Proposed Mechanisms of Action of Omega-3 Polyunsaturated Fatty Acids

In the following sections, we present observational studies, randomized trials, and recent metaanalyses.

OBSERVATIONAL STUDIES

Kromhout et al Study

Because Greenland Eskimos' consumption of a high amount of fish and marine mammals is generally regarded as being responsible for the low incidence of fatal CHD in this population, a cross-sectional study was conducted on 852 middle-aged men in the Netherlands.¹ The study recorded the fish consumption of these patients and followed them for 20 years. Mortality from CHD showed an inverse relation with the amount of fish consumption, with people consuming at least 30 g of fish every day showing a 50% lower mortality rate from CHD. Noteworthy is the fact that the amount of fish being consumed every day in this population is very large—possibly several times larger than the average fish intake in the US population.

Zutphen Study

The Zutphen Study was a cohort study that assessed the relationship between n-3-PUFA consumption and sudden cardiac death (SCD) in 1,373 males.⁸⁹ The risk of CHD (hazard ratio [HR] 0.32, 95% confidence interval [CI] 0.13-0.80 at age 50; HR 1.34, 95% CI 0.58-3.12 at age 80) and SCD (HR 0.46, 95% CI 0.27-0.78) was lower in men with increased fish consumption. The statistical significance of the decrease in the risk of CHD with fatty fish consumption decreased with increasing age. Additionally, no clear dose response of fish intake on the reduction of risk was apparent. Fish consumption, especially in people younger than 65 years, may significantly lower the risk for CHD and SCD. The lower incidence of SCD in these patients could be related to the antiarrhythmic effects of n-3-PUFAs.

Cardiovascular Health Study

A case-control study was conducted within the Cardiovascular Health Study, a cohort study, designed to assess the effects of n-3-PUFAs in older patients with CHD.⁹⁰ Patients with fatal myocardial infarction (MI) and other CHD and patients with nonfatal MI were matched to randomly selected controls. The plasma phospholipid n-3 concentration, a biomarker of n-3-PUFA intake, was measured in the blood samples obtained from the patients almost 2 years before the events. Patients with a higher plasma phospholipid concentration were at a lower risk of fatal CHD (odds ratio [OR] 0.32, 95% CI 0.13-0.78, P=0.01). Conversely, the plasma phospholipid concentration had no relation to nonfatal MI. The association of n-3-PUFAs with a lower risk of fatal CHD but not with nonfatal MI suggests a possible reduction in SCD with increased intake.

RANDOMIZED TRIALS

JELIS Trial

In the Japan EPA Lipid Intervention Study (JELIS), 18,645 patients with hypercholesterolemia were randomized to receive either 1,800 mg/d of EPA with statin or statin alone.⁹¹ The primary endpoint of the trial was any major CHD event, defined as SCD; fatal and nonfatal MI; and nonfatal CHD events including unstable angina pectoris, angioplasty, stenting, or coronary artery bypass graft (CABG). The primary endpoint occurred in 2.8% of the patients receiving EPA compared to 3.5% of the patients receiving the control (relative reduction [RR] 19%, P=0.011). The rates of SCD and CHD death were similar in both groups. Patients with a history of CHD in the EPA group had a significantly lower incidence of major CHD events, a 19% reduction compared to the control group (8.7% vs 10.7%, P=0.048). Conversely, patients without a prior history of CHD did not show a significant reduction in major CHD events with EPA treatment (18% reduction, 1.4% vs 1.7%, P=0.132). The results of this study show that in patients with hypercholesterolemia, EPA treatment reduces the incidence of major CHD events. This finding is especially prominent in hypercholesterolemic patients with a history of CHD. Additionally, the benefits of n-3-PUFAs were apparent in this population despite the statin therapy in both the intervention and control groups.

The absence of any benefit on the rates of SCD may be a consequence of the significantly higher fish intake in the Japanese population compared to the US population.⁹² N-3-PUFAs and fish oil have a nonlinear relation with the primary prevention of SCD and CHD death, presumed to reach maximum benefit around 250 mg/d.⁸⁴ One to 2 servings of fish per week is associated with a reduced risk of CV death in western countries,⁹³ but higher amounts do not seem to result in a greater reduction in SCD.^84,93 However, Japanese men consume more than 3 times this amount from childhood; thus the entire JELIS population might already have been ingesting levels of n-3-PUFAs sufficient to achieve optimal prevention of SCD.⁹⁴ This can be seen by the gross difference in the rates of SCD between the JELIS trial and other trials testing n-3-PUFAs.

A significant inverse association has been shown between levels of marine-derived n-3-PUFAs and carotid intima-media thickness in Japanese individuals, independent of traditional CV risk factors.⁹⁵ Japanese patients have been shown to have a 2-fold higher serum level of n-3-PUFAs than white patients and Japanese American patients in the United States.⁹⁵ Thus, the higher levels of n-3-PUFAs in Japanese patients may provide antiatherosclerotic benefits. Moreover, the incidence of CHD in Japan is less than half that of the United States,⁹⁶ and the percentage of surface involvement of raised lesions in coronary arteries in men aged 30-34 years is less than one-third for Japanese vs US white patients.⁹⁷

GISSI-P Trial

The GISSI (Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto miocardico)-Prevenzione trial was an RCT designed to assess the effect of n-3-PUFA supplementation on mortality in patients with a history of recent MI.^98,99 A total of 11,324 patients were randomized to receive n-3-PUFAs, vitamin E, both n-3-PUFAs and vitamin E, or to serve as controls in addition to receiving the standard medical management and lifestyle modifications. The study had a primary composite endpoint of death, nonfatal MI, and stroke. In patients treated solely with n-3-PUFAs, the risk of occurrence of the primary endpoint was decreased significantly (relative risk reduction [RRR] 10% by 2-way analysis; 15% by 4-way analysis). The risks of death (14% by 2-way analysis; 20% by 4-way analysis), CV death (17% by 2-way analysis; 30% by 4-way analysis), and SCD (26% by 2-way analysis; 45% by 4-way analysis) were also decreased significantly. However, no additional benefit of combined treatment with n-3-PUFAs and vitamin E was seen. In terms of absolute numbers, 164 patients would need to be treated for 1 year with n-3-PUFAs to prevent 1 death. Worthy of note is that n-3-PUFAs are inexpensive and free from any major adverse effects, making this reduction in all-cause mortality rather impressive and difficult to improve even when looking at trials testing statins. The reduction in SCD and total mortality points toward an antiarrhythmic effect of n-3-PUFAs in patients early post-MI. This beneficial effect of n-3-PUFAs is in addition to their well-known antiatherosclerotic and antithrombotic effects. In summary, GISSI-P showed that n-3 fatty acids cause a 20% reduction in death, a 30% reduction in CV death, and a 45% reduction in SCD in patients who have recently experienced an MI.

GISSI-HF Trial

The GISSI-HF trial (Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto miocardico – Heart Failure), an RCT, enrolled 6,975 patients with chronic heart failure of New York Heart Association functional class II–IV and randomized them to receive either 1 g/d of n-3-PUFA or placebo.¹⁰⁰ Deaths from any cause (27% vs 29%, adjusted HR 0.91, 95.5% CI 0.833-0.998, P=0.041) and hospitalizations for CV reasons (57% vs 59%, adjusted HR 0.92, 99% CI 0.849-0.999, P=0.009) were significantly lower in the group receiving n-3-PUFAs compared to those receiving placebo. Fifty-six patients would need to be treated with n-3-PUFAs to avoid 1 death, and 44 patients would need to be treated to avoid 1 event such as death or hospitalization because of CV reasons.

DART

The Diet and Reinfarction Trial (DART) was conducted to determine whether n-3-PUFAs have a role in the secondary prevention of MI in patients with a history of MI.¹⁰¹ In this trial, 2,033 patients with a history of MI were randomized to receive either no counseling or dietary counseling according to 1 of 3 dietary strategies: (1) reduced total fat intake with an increased intake of PUFAs, (2) increased intake of fatty fish, or (3) increased intake of cereal fiber. Patients in the group who received advice on reducing fat intake and increasing PUFAs did not show any difference in mortality compared to the other groups. Conversely, those who received advice on increasing fatty fish intake decreased all-cause mortality by 29% compared to those receiving no advice. The dietary counseling did not change the incidence of reinfarction. This study shows that an increase in the intake of fatty fish by patients who have a history of MI decreases the all-cause mortality rate.

Burr et al

In a similar study by the DART investigators, 3,114 men younger than 70 years with a history of angina received 1 of the following interventions: (1) dietary advice to increase oily fish intake or fish oil capsules; (2) dietary advice to increase fruits, vegetables, and oats intake; (3) dietary advice about both preceding points; or (4) no dietary advice.¹⁰² Although all-cause mortality remained the same across the 4 groups, cardiac death (adjusted HR 1.26, 95% CI 1.00-1.58, P=0.047) and SCD (adjusted HR 1.54, 95% CI 1.06-2.23, P=0.025) were more common in patients advised to increase their intake of oily fish. The increase in mortality was especially prominent in patients advised to increase their intake of fish oil capsules. Inferences from the results of this trial are limited because the researchers did not accurately measure the compliance of patients with the dietary advice.

SOFA Trial

The Study on Omega-3 Fatty Acids and Ventricular Arrhythmia (SOFA) trial was conducted to assess the antiarrhythmic effects of n-3-PUFAs.¹⁰³ A total of 546 patients with implantable cardioverter-defibrillators and a history of malignant ventricular tachycardia or ventricular fibrillation were randomized to receive either 2 g/d of fish oil or placebo. The primary endpoint for the trial was implantable cardioverter-defibrillator intervention for ventricular tachycardia, fibrillation, or all-cause death. No significant difference was seen in the incidence of the primary endpoint between the fish oil and the placebo groups (30% vs 33%, HR 0.86, 95% CI 0.64-1.16, P=0.33). The results of this study did not support the idea that fish oil has an antiarrhythmic effect. However, SOFA was designed to detect a 33% reduction in the primary endpoint with fish oil supplementation. Fish oil may have had a benefit smaller than 33%; therefore, the antiarrhythmic effect of n-3-PUFAs cannot be ruled out from this trial alone.

OMEGA Trial

In the OMEGA trial, 3,851 patients with a history of acute MI in the past 3 to 14 days were randomized to receive either 1 g/d of n-3-PUFAs in addition to the standard guideline treatment or the standard treatment only.¹⁰⁴ The incidence of SCD, total mortality rates, and nonfatal CV events was recorded. No statistically significant difference was seen in the incidence rates of SCD (1.5% vs 1.5%, P=0.84), total mortality (4.6% vs 3.7%, P=0.18), major adverse cerebrovascular and CVD events (10.4% vs 8.8%, P=0.1), or revascularization (27.6% vs 29.1%, P=0.34) between the intervention and the control groups, respectively. The results of this study implied that the addition of n-3-PUFAs to the best available treatment therapy for patients with a recent history of MI does not lead to any significant benefit. The rate of SCD was unexpectedly low in this study, most likely attributable to the aggressive treatment therapy used in these patients. In cases in which aggressive therapy has been used, the benefit of any new intervention becomes difficult to prove.

Alpha Omega Trial

The Alpha Omega trial was an RCT in which 4,837 patients between the ages of 60 and 80 years with a history of MI were randomized to receive margarine supplemented with one of the following: (1) EPA and DHA; (2) alpha linolenic acid (ALA); (3) EPA, DHA, and ALA; or (4) placebo.¹⁰⁵ All patients were already receiving optimal medical therapy, including antihypertensives, antithrombotics, and lipid-modifying therapies. Major CVD events, including fatal and nonfatal CVD events and CV interventions, were the primary outcome. No difference was seen in the incidence of the primary endpoint between the groups receiving EPA/DHA (HR 1.01, 95% CI 0.87-1.17, P=0.93) and solely ALA (HR 0.91, 95% CI 0.78-1.05, P=0.20). In this trial, supplementation with low doses of either EPA and DHA or ALA did not reduce the incidence of major CVD events in patients with a history of MI who were already receiving optimal pharmacologic therapy. However, patients with diabetes who received EPA and DHA supplementation showed a lower incidence of fatal CHD (HR 0.51, 95% CI 0.27-0.97) compared to the control group. This trial enrolled participants with a mean age of 69 years, which is significantly older than the mean ages in other trials. Additionally, patients not receiving statins (n=413) had a nominally significant reduction in CVD events (9% vs 18%, adjusted HR 0.46, 95% CI 0.21-2.02, P=0.051) when given n-3-PUFAs plus ALA; in contrast, no trend toward benefit was seen with this supplementation in statin users (HR 1.02, 95% CI 0.80-1.30). Alpha Omega did not show a benefit of n-3-PUFAs; however, the high background use of optimal pharmacologic therapy may have masked any potential positive effects.

SU.FOL.OM3 Trial

The SU.FOL.OM3 (Supplémentation en Folates et Omega-3) trial investigated the effects of n-3-PUFAs and vitamin B on the incidence of major CVD events in patients with a history of CHD or stroke.¹⁰⁶ A total of 2,501 patients with a history of MI, unstable angina, or ischemic stroke were randomized to receive 1 of 4 interventions: (1) a combination of 5-methyltetrahydrofolate, vitamin B6, and vitamin B12; (2) 380 mg of n-3-PUFAs; (3) both preceding therapies; or (4) placebo. The primary outcome measure was major CV events, including nonfatal MI, stroke, or CVD death. No significant effect on the primary outcome measure was seen with supplementation of either vitamin B (HR 0.90, 95% CI 0.66-1.23, P=0.50) or n-3-PUFAs (HR 1.08, 95% CI 0.79-1.47, P=0.64). Additionally, n-3-PUFA supplementation did not reduce the incidence of all-cause mortality, cancer morbidity, or any other secondary endpoint (HR 1.10, 95% CI 0.81-1.48, P=0.55). Several factors could have influenced the lack of apparent benefit of n-3-PUFA supplementation in the patients enrolled in this trial: small sample size; low dose of EPA/DHA (380 mg); greater use of angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers in the placebo group; more current smokers and patients with a history of MI in the omega-3 group; short trial duration; and fewer than expected major CVD events, leading to a 15% lower than expected power to test the benefits of n-3-PUFAs in this trial (ie, 20% power to detect a 25% benefit of n-3s). Additionally, the average start of n-3-PUFAs was 101 days post-CVD event in this trial, which is much longer than the 16 days post-MI in the GISSI-P trial.

Hemodialysis Patients

Svensson et al tested benefits of n-3-PUFAs for the secondary prevention of CV events in patients undergoing chronic hemodialysis in an RCT that assigned 206 patients to receive either n-3-PUFAs or control treatment.¹⁰⁷ A composite primary outcome of total CV events and death was seen in 59% of the patients. Although n-3-PUFAs did not reduce the incidence of the primary endpoint compared to the control (59 endpoints vs 62 endpoints, P=nonsignificant), the number of MIs in the n-3-PUFA group was significantly reduced compared to the control group (4 vs 13, P=0.036). Only half of the deaths in these patients were attributable to CV causes, and this study lacked a large sample size, especially considering the large number of withdrawals. Moreover, this trial was not adequately powered to detect a small benefit on the primary endpoint. The lower incidence of MI with n-3-PUFA treatment in secondary prevention patients undergoing chronic hemodialysis is clinically relevant as these patients are at a high risk of recurrent CV events.

Patients with CABG

Eritsland et al showed that n-3-PUFAs provide CV benefit in patients undergoing CABG; 610 patients undergoing CABG were randomized to receive either 4 g/d of fish oil or control therapy in addition to either aspirin or warfarin.¹⁰⁸ One year after the CABG, vein graft occlusion rates per distal anastomoses were significantly lower in the group receiving fish oil compared to the control group (27% vs 33%, OR 0.77, 95% CI 0.60-0.99, P=0.034). Similarly, 43% vs 51% of the patients in the fish oil group compared to the control group had ≥1 occluded vein graft (OR 0.72, 95% CI 0.51-1.01, P=0.05). Therefore, increased n-3-PUFA intake may be able to prevent occlusion of vein grafts and increase the rate and duration of patency of the graft in CABG patients.

RECENT METAANALYSES

Rizos et al

A metaanalysis conducted on 20 RCTs including 68,680 patients evaluated the effects of n-3-PUFAs on all-cause mortality, CV death, SCD, MI, and stroke.¹⁰⁹ The results did not show n-3-PUFA supplementation to be associated with a reduced risk of all-cause mortality (RR 0.96, 95% CI 0.91-1.02; absolute risk reduction [RD] −0.004, 95% CI −0.01-0.02), CV death (RR 0.91, 95% CI 0.85-0.98; RD −0.01, 95% CI −0.02-0.00), SCD (RR 0.87, 95% CI 0.75-1.01; RD −0.003, 95% CI −0.012-0.006), MI (RR 0.89, 95% CI 0.76-1.04; RD −0.002, 95% CI −0.007-0.002), or stroke (RR 1.05, 95% CI 0.93-1.18; RD 0.001, 95% CI −0.002-0.004). Lewis and colleagues explained the lack of apparent benefit of n-3-PUFAs in this metaanalysis by pointing out that the mean dose of n-3-PUFAs given to patients in studies included in this review was 1.51 g/d (perhaps not high enough), some of the studies had small sample sizes, and diverse sources of n-3-PUFAs may affect their efficacy.¹¹⁰ In addition, the authors chose <0.0063 as the P value for significance instead of the usual P value of 0.05. This measure was unnecessary and may have led to misinterpretation of the results.¹¹¹

Delgado-Lista et al

In a metaanalysis of 21 studies conducted on the CV effects of n-3-PUFAs, the risk of a CV event of any kind was significantly decreased by 10% (OR 0.90, 95% CI 0.85-0.96, P=0.001).² Similarly, CV death was decreased by 9% (OR 0.91, 95% CI 0.83-0.99, P=0.03), and fatal and nonfatal CHD events were decreased by 18% (OR 0.82, 95% CI 0.75-0.90, P<1×10⁻⁴). Total mortality was also reduced for the patients receiving n-3-PUFAs (5% reduction of risk, OR 0.95, 95% CI 0.89-1.02, P=0.15).

Kwak et al

These beneficial results were not shown for n-3-PUFAs in the metaanalysis by Kwak et al.¹¹² Inclusion of small trials, lack of inclusion of GISSI-P and DART, and inclusion of trials testing low doses of EPA/DHA (SU.FOL.OM3) could have contributed to the lack of demonstrated CV benefit of n-3-PUFAs.¹¹³ However, in a subanalysis limited to the 5 included studies in which lipid-lowering agents were not commonly used, a trend toward reduction in CVD events was noted in the omega-3 group (RR=0.74, 95% CI 0.54-1.03). This trend very likely would have been statistically significant if the large GISSI-P study had not been excluded for lack of double-blinding.

DISCUSSION

In light of the evidence that statin therapy exerts a prominent antiarrhythmic effect¹¹⁴ and the protection observed with n-3-PUFA supplementation among non–statin users (but not among statin users) in the Alpha Omega trial and the Kwak metaanalysis, concurrent statin therapy can be reasonably suspected to render superfluous the antiarrhythmic benefit of modest intakes of n-3-PUFAs, hence accounting for the null findings of several recent n-3-PUFA supplementation studies. Because SCD is the first symptom of CHD in a significant proportion of cases and most asymptomatic people developing CHD do not take statins, modest supplemental intakes of n-3-PUFAs may have important potential for prevention of SCD in non–statin users. Many patients cannot tolerate statins. Moreover, the JELIS study suggests that high intakes of n-3-PUFAs may be protective even in the context of statin usage, likely by evoking protective mechanisms independent of arrhythmia prevention. Hence, even among statin users, the possible benefits of high intakes of n-3-PUFAs merit further exploration in controlled trials.

Compared to populations without high intakes of n-3-PUFAs, those with very high lifelong intakes of n-3-PUFAs are notable for low CV risk and less arterial atherosclerosis. This trend may reflect the interaction of a large number of protective mechanisms, as demonstrated in rodent and clinical studies, as well as concurrent low consumption of red meat. Although high intakes of n-3-PUFAs may be required to evoke many of these mechanisms, epidemiology and some clinical trials suggest that intakes of n-3-PUFAs as low as 250 mg/d can provide protection from SCD and cardiac arrhythmias, especially in non–statin users. The failure of modest doses of n-3-PUFAs to confer protection in some recent trials may reflect the competing antiantiarrhythmic benefits of concurrent therapy with statins and possibly other CV drugs. Increased baseline intakes of n-3-PUFAs in recent study populations, owing to increased public awareness of the protection afforded by n-3-PUFAs, may also have obscured the impact of modest supplemental intakes of n-3-PUFAs in these studies (Table 2). Hence, supplemental n-3-PUFAs are most likely to be protective in non–statin users who rarely consume fatty fish. The quite low CV risk enjoyed by populations with very high dietary intakes of n-3-PUFAs and the favorable impact of 1.8 g of EPA daily among the statin users of the JELIS trial suggest that prolonged high intakes of n-3-PUFAs may confer benefit even in patients receiving modern statin-based pharmacotherapy; hence, high-dose n-3-PUFA supplementation merits further exploration in the secondary prevention of CV events. The antiatherogenic, antithrombotic, and antiinflammatory benefits of n-3-PUFAs appear to require considerably higher intakes than those adequate for prevention of SCD.

Table 2.

Summary of Important Positive and Negative Omega-3 Polyunsaturated Fatty Acids Trials

CONCLUSION

Despite the recent negative data about n-3 fatty acids, the overall evidence still supports the American Heart Association recommendation of n-3-PUFAs (1 g/d) for secondary prevention of CHD.

This article meets the Accreditation Council for Graduate Medical Education and the American Board of Medical Specialties Maintenance of Certification competencies for Patient Care and Medical Knowledge.