Recent findings of long-chain n-3 polyunsaturated fatty acids (LCn-3 PUFAs) on atherosclerosis and coronary heart disease (CHD) contrasting studies in Western countries to Japan

Abstract

Recent long-term randomized clinical trials (RCTs) of long chain n-3 polyunsaturated fatty acids (LCn-3 PUFAs) on coronary heart disease (CHD) among high risk patients conducted in Western countries all failed to show their clinical benefits. In striking contrast, an RCT of LCn-3 PUFAs on CHD conducted in Japan, which is a combination of secondary and primary prevention, showed a significant 19% reduction. Potential reasons for this discrepancy are large differences in doses of LCn-3 PUFAs administered (300–900 mg/day in Western countries vs. 1,800 mg/day in Japan) and background dietary intake of LCn-3 PUFAs (<300 mg/day in Western countries vs. >1,000 mg/day in Japan). These observations suggest that higher doses of LCn-3 PUFAs than examined in RCTs in Western countries may be cardio-protective. Atherosclerosis is the major underlying cause of CHD. Recent observational studies and an RCT of LCn-3 PUFAs on atherosclerosis in Japan show that LCn-3 PUFAs are anti-atherogenic. In this brief review, we focus on recent epidemiological and clinical findings of LCn-3 PUFAs on atherosclerosis and CHD contrasting studies in Western countries to those in Japan. We also discuss mechanisms of high-dose LCn-3 PUFAs on atherosclerosis.

Introduction

Epidemiological studies in Eskimos in the mid-20^th century show very low prevalence of myocardial infarction (MI) and autoimmune diseases.¹ This population is characterized by substantially high dietary intake of long-chain n-3 polyunsaturated fatty acids (LCn-3 PUFAs), more than 40 times higher than the current median intake in the US general population ( <100 mg/day).² Since then, cardio-protective properties of LCn-3 PUFAs have been intensively investigated in basic, clinical and population studies. Several excellent reviews on this topic have recently been published, ^3–5 encompassing potential molecular mechanisms of LCn-3 PUFAs (anti-inflammatory, anti-atherogenic, and anti-arrhythmic and other properties); clinical trials of LCn-3 PUFAs on cardiovascular disease (CVD) and its risk factors; and epidemiological associations of LCn-3 PUFAs with CVD. In this brief review, we focus on recent epidemiological and clinical findings of LCn-3 PUFAs on atherosclerosis and coronary heart disease (CHD) contrasting studies in Western countries to those in Japan where dietary intake of LCn-3 PUFAs is >10 times higher than in the US. We also discuss mechanisms of high-dose LCn-3 PUFAs for anti-atherosclerosis and potential future direction.

Atherosclerosis

Atherosclerosis is the major underlying cause of CHD. Recent advancements in imaging technologies have made it possible to evaluate atherosclerosis in the general population non-invasively.⁶ These technologies include intima-media thickness of the carotid artery (cIMT) by ultrasound and coronary artery calcification (CAC) by computed tomography. Both cIMT and CAC are predictors of future cardiovascular events independent of other risk factors,⁶ though CAC predicts future events more powerfully than cIMT. ⁷

A large body of basic research documents that LCn-3 PUFAs have anti-atherogenic properties. ⁸ It is, however, speculated that for LCn-3 PUFAs to exert the anti-atherogenic properties in human, at least 1,000 mg/day of dietary intake of LCn-3 PUFAs is required.³ Current dietary intake of LCn-3 PUFAs in Western countries is far too low compared to this amount. In fact epidemiological studies in Western countries reported no significant association of biomarkers of LCn-3 PUFAs with atherosclerosis. ^{9, 10}

Recently, the ERA JUMP Study shows significant inverse associations of a biomarker of LCn-3 PUFAs with atherosclerosis in the general population in Japan but not in the US. ^11–13 The study is an international population-based prospective cohort study of atherosclerosis including cIMT and CAC in approximately 1,000 middle-aged men free from CVD at baseline (300 Japanese in Japan, 300 whites, 300 Japanese Americans and 100 blacks in the US.) The study showed that serum levels of LCn-3 PUFAs in men in Japan were three times higher compared to men in the US¹¹ and that serum levels of docosahexaenoic acid (DHA), one major component of LCn-3 PUFAs, but not EPA, had a significant inverse association with cIMT independent of other risk factors in Japan.¹² The significant inverse association was not observed in the US population. The study also demonstrated that men in Japan compared to the US had significantly lower prevalence as well as incidence of CAC, which remained after adjusting for traditional and other risk factors. However, after further adjusting for serum levels of LCn-3 PUFAs, the significant difference was attenuated and became non-significant.^{11, 13}

A recent cross-sectional study in Japan has documented that serum levels of LCn-3 PUFAs had a significant inverse association with the presence of non-calcified atherosclerotic plaque.¹⁴ Non-calcified atherosclerotic plaque is more prone to rupture and a strong predictor of future CHD events among symptomatic patients independent of other risk factors including CAC.¹⁵ The study examined 172 consecutive patients who underwent coronary computed-tomography angiography and found that multivariate-adjusted odds ratio of the presence of non-calcified atherosclerotic plaque in individuals whose serum levels of LCn-3 PUFAs were median or below was 3.51 (95% confidence interval (95% CI): 1.70–7.59), compared to individuals whose serum levels were above median. Results of these observational studies suggest that high levels of LCn-3 PUFAs have anti-atherogenic properties.

The notion that high level of LCn-3 PUFAs is anti-atherogenic is supported by a recent RCT of LCn-3 PUFAs on the progression of cIMT.¹⁶ Mita et al. conducted an open-labeled randomized trial of 1,800 mg/day of EPA on progression of cIMT among 81 patients with type 2 diabetes in Japan. During 2.1 years of intervention, treated groups had significantly less progression of cIMT compared to the control group (−0.029 ± 1.12 vs. 0.016 ± 0.109 mm/year, respectively, p=0.029). Decreased progression was not due to blood pressure, lipids (triglycerides, HDL-C), or glycemic control (HbA1c) because no differences were observed in these factors during the trial between the intervention and control groups.

Coronary heart disease

A recent meta-analysis of nine prospective cohort studies shows that dietary intake of LCn-3 PUFAs is associated with lower mortality from CVD, CHD, or sudden cardiac death. ¹⁷ Meta-regression analysis of these data shows a threshold effect: the association of LCn-3 PUFAs with mortality remains significant up to mean or median intake of 200 mg/day of LCn-3 PUFAs with no change in the risk at higher mean intakes. The dose-response odds ratio for CVD, CHD or sudden cardiac death is 0.64 to 0.84 per 200 mg/day of mean intake of LCn-3 PUFAs, using thresholds of 200 to 300 mg/day for the floor effect. The findings are consistent with those reported in a previous systemic review.¹⁸ It is postulated that the dose-response relationship for the anti-arrhythmic effect of LCn-3 PUFAs is initially steep until 250 mg/day of LCn-3 PUFAs with a subsequent plateau.¹⁸ Thus, the significant reduction in mortality from CVD, CHD, or sudden cardiac death in these prospective cohort studies is believed to be due to the anti-arrhythmic properties of LCn-3PUFAs.¹⁸

Two recent prospective cohort studies in Japan, the Japan Collaborative Cohort Study for Evaluation of Cancer Risk (JACC) Study ¹⁹ and the NIPPON DATA80, ²⁰ which were not included in the meta-analysis, have documented a significant inverse association of CVD mortality with dietary intake of LCn-3 PUFAs at much high levels of consumption (Figure 1). The JACC Study followed 57,972 Japanese men and women for 12.7 years and showed that the multivariable-adjusted hazard ratio (HR) for total CVD mortality in the highest quintile was 0.81 (95% CI): 0.67–0.98, p for trend =0.01) compared to the lowest quintile. The NIPPON DATA followed 9,290 Japanese men and women for 24 years and showed that the multivariable-adjusted HR for total CVD mortality in the highest quartile was 0.80 (95% CI: 0.66–0.96, p for trend =0.038) compared to the lowest quartile. In the JACC Study, the reduction in mortality from heart failure and fatal MI but not stroke contributes to the significant reduction in total CVD mortality whereas in NIPPON DATA80, the reduction in mortality from both CHD and stroke contributes to the significant reduction in total CVD mortality. Though dietary intake of LCn-3 PUFAs did not have a significant inverse association with CHD mortality in either study, this appears to be largely due to the lack of power in these studies, i.e., the small number of CHD death.

Fig 1. Multivariable-adjusted relative risk of non-fatal coronary event and total cardiovascular disease mortality from recent prospective cohort studies in Japan.

Relative risk of non-fatal coronary events from the Japan Public Health Center based cohort ²²

Relative risk of total cardiovascular disease mortality from the NIPPON DATA80 ²⁰

Relative risk of total cardiovascular disease mortality from the Japan Collaborative Cohort Study for Evaluation of Cancer Risk (JACC) Study ¹⁹

Prospective cohort studies in Western countries show that dietary intake of LCn-3 PUFAs has no significant association with non-fatal MI or if there is, the inverse association is much weaker compared to fatal MI. ^{18, 21} In sharp contrast, the Japan Public Health Center-Based (JPHC) Study has documented a significant and strong inverse association of dietary intake of LCn-3 PUFAs with non-fatal MI. JPHC followed 41,578 Japanese men and women for 10 years and observed 196 nonfatal and 62 fatal coronary events and showed that the multivariable-adjusted HR for non-fatal coronary events in the highest quintile was 0.33 (95% CI: 0.17–0.63; p for trend <0.01) compared to the lowest quintile (figure 1). ²² Though there was no significant association of fatal coronary events with dietary intake of LCn-3 PUFAs, this is due to the small number of fatal cases.

Mean dietary intake of LCn-3 PUFAs in these three prospective cohort studies in Japan ( JACC, NIPPON DATA80, and JPHC) are far above that of the US general population ( <100 mg/day) as well as the hypothesized threshold of the anti-arrhythmic effect of LCn3 PUFAs (250 mg/day) (Figure 1). These observations suggest that high dietary intake of LCn-3 PUFAs is protective against CHD through mechanisms other than the anti-arrhythmic effect.

Recent randomized controlled trials (RCTs) of LCn-3 PUFAs on CHD

Recent long-term RCTs of LCn-3 PUFAs on CHD (> two years of intervention) among high risk patients conducted in Western countries all failed to show their clinical benefits. In striking contrast, the largest RCT of eicosapentaenoic acid (EPA), one major component of LCn-3 PUFAs, conducted in Japan, the JELIS (Japan Eicosapentaenoic acid Lipid Intervention Study) documented a significant reduction in cardiovascular events in the intervention group (Table). The Alpha Omega Trial is an RCT of 376 mg/day of LCn-3 PUFAs among 4,837 patients with a history of MI who were treated with state-of-the-art medications. During 3.4 years of intervention, LCn-3 PUFAs did not reduce cardiovascular events (HR of 1.01 (95% CI: 0.87–1.17, p=0.93). ²³ The SU.FOL.OM3 is an RCT of 600 mg/day of LCn-3 PUFAs among 2,501 patients with a history of CHD or ischemic stroke. During 4.7 years of intervention, LCn-3 PUFAs did not reduce cardiovascular events (HR of 1.08 (95% CI: 0.79–1.47, p=0.64). ²⁴ The ORIGIN is an RCT of 840 mg/day of LCn-3 PUFAs among 12,536 patients who were at high risk for cardiovascular events and had dysglycemia. During 6.2 years of intervention, LCn-3 PUFAs did not reduce cardiovascular death (HR of 0.98 (95% CI: 0.87–1.10, p=0.72).²⁵ The Risk and Prevention Study is the most recent RCT of about 900 mg/day of LCn-3 PUFAs among 12,513 patients with multiple cardiovascular risk factors or atherosclerotic vascular disease but without MI.²⁶ During 5 years of intervention, LCn-3 PUFAs did not reduce time to cardiovascular death or admission to the hospital for cardiovascular causes (HR of 0.97, 95% CI: 0.88–1.08, p=0.58).

Table.

Recent randomized controlled trials (RCTs) of LCn-3 PUFAs on coronary heart disease in Western countries and Japan

Study name, Published year (Country) reference	Characteristics of participants	Age (years)	Dose of LCn-3 PUFAs (mg/day)	Placebo for controls	Duration (year)	No. of participants in treatment/control	Primary outcome	Funding source
Studies In Western countries
Alpha-Omega, 2010, (Netherlands) 23	Patients with AMI diagnosed up to 10 years before randomization	69 (60–80)	376 (226 EPA, 150 DHA)	Margarine without n-3 FAs	3.7	2,404/2,433	Major CVD events (fatal and non-fatal CVD, PCI, CABG)	Mixed
SU.FOL.OM3, 2010, (France) 24	Patients with acute coronary or cerebral event within 12 months	60.9 ± 8.8	600 (400 EPA, 200 DHA)	Placebo (not described fully)	4.7	1,253/1,248	Major CVD event (fatal and non-fatal MI, ischemic stroke or death from cardiovascular disease	Mixed
ORIGIN, 2011, (multiple) 25	Subjects with and without CVD	64	840 (465 EPA, 375 DHA)	1 g of olive oil	6.2	6,281/6,255	Death from CVD	Industry
The Risk and Prevention Study Collaborative Group, 2013, (Italy) 26	Patients with multiple CVD risk factors or atherosclerotic vascular disease (without MI)	64 ± 9	900 (EPA + DHA)	Olive oil	5	6,239/6,266	Time to death from cardiovascular cause or admission to the hospital for cardiovascular cause	Industry
A study in Japan
JELIS, 2007, (Japan) 27	Patients with hyper- cholesterolemia with and without CVD	61 (40–75)	1,800 EPA	None	4.6	9,326/9,319	Major coronary event (SCD, fatal and non-fatal MI, unstable angina, revascularization)	Industry

LCn-3 PUFAs: long-chain n-3 polyunsaturated fatty acids, SU.FOL.OM3: Supplementaion en Folates et Omage-3, ORIGIN: Outcome Reduction with Initial Glargine Intervention, JELIS: The Japan EPA Lipid Intervention Study, AMI: acute myocardial infarction, CVD: cardiovascular disease, EPA: eicosapentaenoic acid, DHA: docosahexaenoic acid, PCI: pericutaneous coronary intervention, CABG: coronary artery bypass graft, MI: myocardial infarction, SCD: sudden cardiac death

The JELIS is a randomized open-label trial of 1,800 mg/day of EPA among 18,645 patients with hypercholesterolemia without and with CHD (n=14,981 and 3,664, respectively). ²⁷ Statin was administered to both control and intervention groups. During 4.6 years of intervention, EPA significantly reduced cardiovascular events (HR of 0.81, 95% CI: 0.69–0.95, p=0.011). In JELIS, cardiovascular events were defined as the composite of sudden cardiac death, fatal and non-fatal MI, and non-fatal CHD including unstable angina, coronary artery bypass graft or percutaneous transluminal coronary angioplasty. Sub-group analyses show that the rates of sudden cardiac death and CHD death were not statistically significantly different between the intervention and control groups. The significant reduction was observed only in non-fatal coronary events.

What caused the discrepancy in the results of RCTs in Western countries and Japan? Negative results in RCTs in Western countries could be attributed to inappropriate sample size, low dose of LCn-3 PUFAs, state-of-the-art treatment in participants, and other reasons.^{28, 29} JELIS was an open label trial whereas RCTs in Western countries were placebo-controlled trials. Moreover, significant clinical benefits in JELIS were observed only in soft endpoints, i.e., non-fatal coronary events. Thus, the significant results might be attributed to bias. Nonetheless, the result was consistent with a prospective cohort study that reported dietary intake of LCn-3 PUFAs had a significant inverse association with non-fatal CHD²² described in the previous section.

Among other factors, at least two factors are remarkably different between the two groups: dose of LCn-3 PUFAs administered in RCTs and background dietary intake of LCn-3 PUFAs (Figure 2). The dose administered in JELIS was 1,800 mg/day whereas the doses administered in Western countries were much lower, ranging from 376 to 900 mg/day. Median dietary intake of LCn-3 PUFAs in Japan is >1000 mg/day, which is one of the highest in the world and much higher than that in Western countries.³⁰ These observations suggest that high dose or consumption of LCn-3 PUFAs exerts cardio-protective effects.

Fig. 2. Estimated ranges of dietary intake of long-chain n-3 polyunsaturated fatty acids (LCn-3 PUFAs) and dose of LCn-3 PUFAs administered in randomized clinical trials in Western countries and Japan.

denotes estimated ranges of background dietary intake of long-chain n-3 polyunsaturated fatty acids (LCn-3 PUFAs)

denotes the difference in LCn-3 PUFAs between control and intervention groups

Dose of LCn-3 PUFAs administered in each RCT is shown in the parenthesis.

RPSC: Risk and Prevention Study Collaborative group, JELIS: Japan Eicosapentaenoic acid Lipid Intervention Study,

Potential mechanisms of high-dose LCn-3 PUFAs on atherosclerosis

LCn-3 PUFAs possess various favorable effects on traditional cardiovascular risk factors (blood pressure, lipids, and diabetes), yet these effects are very mild in human except for the hypo-triglycerimic effect among subjects with hyper-triglyceridemia.^3–5 Thus, the anti-atherogenic properties of high-dose LCn-3 PUFAs are likely to be mediated through non-traditional risk factors. In fact, significant inverse associations of LCn-3 PUFAs with atherosclerosis observed in Japanese study are independent of traditional risk factors.^{11, 12, 16}

Atherosclerosis is a chronic inflammatory disease in the arterial wall driven by innate and adaptive immunity from its initiation, progression and through plaque rupture (Figure 3). ^{31, 32} LCn-3 PUFAs have anti-inflammatory and immunomodulatory properties which are applied to treating other chronic inflammatory diseases, e.g., rheumatoid arthritis, inflammatory bowel diseases.³³ Some clinical evidence supports the mechanistic link between atherosclerosis and LCn-3 PUFAs through their anti-inflammatory properties. For example, a recent meta-analysis of 18 RCTs of LCn-3 PUFA supplementation showed reduced plasma concentrations of soluble intercellular adhesion molecule-1 (sICAM-1) at a dose ranging from 270 to 6,600 mg/day over a median of 12 weeks.³⁴ Another example is a recent RCT of about 1,400 mg/day of LCn-3 PUFAs on 121 patients awaiting carotid endarterectomy which showed for a median of 21 days of intervention that LCn-3 PUFAs were associated with increase plaque stability.³⁵ However, clinical benefits for atherosclerosis of LCn-3 PUFAs through their anti-inflammatory and immunomodulatory properties have not been well established. One major difficulty in investigating the association of atherosclerosis with anti-inflammatory properties of LCn-3 PUFAs in human studies lies on the fact that there are generally very weak associations of circulating levels of inflammatory markers (e.g., C-reactive protein (CRP), cytokines), with measures of atherosclerosis (e.g., cIMT and CAC). ^{36, 37} This is partly because circulating levels of these inflammatory markers do not reflect local inflammation in the arterial wall.

Figure 3. Atherosclerosis is a chronic inflammatory disease driven by immunity.

(A) Initiation and progression of atherosclerosis

Atherosclerosis is a chronic inflammatory disease in the arterial wall driven by innate and adaptive immunity. ³² Plasma low-density-lipoprotein (LDL) particles that diffuse into the sub-endothelial space are prone to oxidative modification. These oxidized LDL (oxLDL) activates endothelial cells, induce expression of leukocyte adhesion molecules, i.e., vascular-cellular adhesion molecules (VCAM) and intra-cellular adhesion molecules (ICAM) and attract monocytes and T cells. Monocytes differentiate into macrophages, release pro-inflammatory mediators such as tissue-necrotizing factor-α and interluekin-1β. Macrophages evolve into foam cells after taking up large amounts of oxLDL. Dendritic cells serve as antigen-presenting cells (APCs) and activate T cells. The majority of T cells in atherosclerosis are type 1 helper T cells (Th1), which are pro-inflammatory and characterized by the production of high levels of interferon-γ (INF-γ). INF-γ promotes the recruitment of T cells and macrophages to the plaque, contributing to the expansion of atherosclerotic plaque burden. Type 17 helper T cells (Th17) are considered to be atherogenic whereas T-regulatory cells are protective against atherosclerosis.

(B) Atherosclerotic plaque

Th1 and macrophages contribute to plaque vulnerability. INF-γ secreted by Th1 inhibits the production of the interstitial collagen that is required to maintain the fibrous cap. Th1 also activate macrophages which produce interstitial collagenases, e.g., matrix metalloproteinases 1 and 8 that are associated with collagen breakdown in fibrous cap.³¹

LDL: Low density lipoprotein, oxLDL: oxidized LDL, APC, antigen presenting cell, Th1: type 1 helper T cell, INF-γ: Interferon γ, ICAM-1: intra-cellular adhesion modeluce-1, VCAM-1: Vascular-cellular adhesion molecule, MMP: Matrix metallopeptidase

Recent progress in vascular biology has revealed that immune cells play fundamental roles in atherogenesis from its inception through plaque rupture (Figure 3). ^{31, 32} The predominant T cells found in atherosclerotic plaque are type 1 helper T cells (Th1), which are pro-inflammatory and characterized by the production of high levels of interferon-γ (INF-γ). INF-γ promotes the recruitment of T cells and macrophages to the plaque, contributing to the expansion of atherosclerotic plaque burden.³² Th1 also contribute to plaque vulnerability through the secretion of INF-γ, which inhibits the production of the interstitial collagen required to maintain the fibrous cap. ³¹ Type 17 helper T cells (Th17) are also considered to be atherogenic whereas regulatory T cells are protective against atherosclerosis.³² Recently we have shown that circulating levels of Th1 had significant positive associations with both CAC and cIMT in 917 men and women from the US general population.³⁸ The significant associations were independent of traditional and novel risk factors including CRP and sICAM-1. Though the study was cross sectional, currently a large epidemiological study is ongoing to examine a longitudinal association of T cell phenotypes including Th1 with cardiovascular outcomes.

Since basic research shows an inverse association of LCn-3 PUFAs with Th1,³³ research in human to investigate the effect of high-dose LCn-3 PUFAs on T cell phenotype is of interest. However, feeding experiments of LCn-3 PUFAs on immune cells are limited to a small number of participants.³⁹ Moreover, large epidemiological studies on cellular phenotypes have always been complicated by the need for fresh whole blood. The use of cryopreserved cells is opening new avenues in cellular phenotyping that will allow further study of immune cells and atherosclerosis associated immunomodulatory properties of LCn-3 PUFAs.

Conclusions

Recent evidence from observational studies and RCTs in Japan where dietary intake of LCn-3 PUFAs is >10 times higher than in the US shows that LCn-3 PUFAs is anti-atherogenic and is inversely associated with CHD. These observations suggest that high-dose LCn-3 PUFAs may exert the anti-atherogenic, anti-inflammatory and immunomodulatory properties of LCn-3 PUFAs as opposed to the anti-arrhythmic properties at low dose with a ceiling effect. Though recent RCTs of LCn-3 PUFAs on CHD in Western countries have failed to show their clinical benefits, this could be due to low doses of LCn-3 PUFAs, sample size, state-of-the-art treatment in participants and other reasons. The results of VITamin D and OmegA-3 TriaL (VITAL) are awaited, which is an RCT of 1,000 mg/day of LCn-3 PUFAs and 2,000 IU of vitamin D supplements in the primary prevention of CVD and cancer among 20,000 US adults,⁴⁰ which is estimated to complete in 2017.