点击显示 收起
1 From the Medizinische Klinik, Klinikum Innenstadt, University of Munich, Germany.
2 Address reprint requests to C von Schacky, Preventive Cardiology, Medizinische Klinik, Klinikum Innenstadt, University of Munich, Ziemssenstr 1, 80336 München, Germany. E-mail: vonschacky{at}medinn.med.uni-muenchen.de.
ABSTRACT
Epidemiologic studies have shown an inverse correlation between consumption of fish or other sources of dietary n-3 fatty acids and cardiovascular events. Numerous mechanisms of action for the favorable effect of dietary n-3 fatty acids on factors implicated in the pathogenesis of atherosclerosis have been described. Studies in dogs, swine, and nonhuman primates have consistently shown beneficial effects in various models of vasoocclusive diseases. Studies published currently do not indicate that dietary n-3 fatty acids prevent restenosis after percutaneous coronary angioplasty or induce regression of coronary atherosclerosis. However, in a recent study, occlusion of aortocoronary venous bypass grafts was reduced after 1 y by daily ingestion of 4 g fish-oil concentrate. In the Diet and Reinfarction Trial, 2-y overall mortality was reduced by 29% in survivors of a first myocardial infarction after consumption of n-3 fatty acid–rich fatty fish at least twice a week had been advised (Lancet 1989;2:757–61). When n-3 fatty acids were integrated into a diet resembling a traditional Mediterranean diet, 5-y cardiovascular mortality after a first myocardial infarction was reduced by 70% (Lancet 1994; 343:1454–9). Preliminary studies indicate that cardiac transplant patients could be an interesting focus of investigation. Currently, food sources rich in n-3 fatty acids are thought to be beneficial in secondary prophylaxis after a myocardial infarction. Large-scale clinical studies with endpoints such as morbidity and mortality are needed to more precisely define the role of n-3 fatty acids in primary and secondary prophylaxis of coronary atherosclerosis.
Key Words: n-3 Fatty acids • coronary atherosclerosis • primary prevention • secondary prevention • vascular disease • coronary artery disease • myocardial infarction • fish oil
EPIDEMIOLOGIC STUDIES
Epidemiologic studies relating intake of n-3 fatty acids with reduced all-cause mortality continue to be published. Recent evidence of such a relation is discussed elsewhere in this supplement (1). However, as early as 1953 there was curiosity about the diet and disease patterns of the Eskimos (reported in 2). More thorough investigations found surprisingly low cardiovascular mortality in Eskimos consuming a traditional diet rich in eicosapentaenoic (20:5n-3) and docosahexaenoic (22:5n-3) acids (2). The seminal findings in Eskimos were confirmed and extended in Western populations. In most studies, a dose-related, inverse correlation between the intake of fish or n-3 fatty acids and total mortality or cardiovascular mortality was shown (2–4).
MECHANISMS OF ACTION
A comprehensive review of the mechanisms of action described for n-3 fatty acids is beyond the scope of this discussion. Furthermore, data generated in vitro and in animal systems have at times conflicted with those in humans. However, it is worthwhile to note the mechanisms of action related to coronary atherosclerosis suggested by the following results of n-3 fatty acid supplementation studies: 1) platelet aggregation, as assessed ex vivo in response to various stimuli, is less vigorous after n-3 supplementation (2); 2) dietary n-3 fatty acids lower triacylglycerol concentrations (5); 3) large doses of n-3 fatty acids lower cholesterol concentrations (5); 4) LDL cholesterol appears to increase somewhat dose dependently with n-3 supplementation, the increase apparently being in the size, considered favorable, not the number of molecules (5–7); 5) increases in HDL-cholesterol concentrations in response to n-3 supplementation were observed in most, but not all, studies (5); 6) moderate reductions in blood pressure in response to moderate and large doses of n-3 fatty acids were observed consistently (2, 8–10) (a beneficial shift in the eicosanoid system toward vasodilation and antiaggregation is currently thought to contribute to the observed reductions in blood pressure and platelet aggregability; 2); 7) improvements in rheologic indexes after n-3 supplementation have been observed (2); 8) more recently, reductions in interleukin 1ß and tumor necrosis factor concentrations have been observed ex vivo (11, 12), bringing human cytokine synthesis into focus (11); 9) A more prominent role in the pathogenesis of atherosclerosis is ascribed to the platelet-derived growth factor (PDGF; 13), 10) dietary n-3 fatty acids down-regulate gene expression of both PDGF-A and PDGF-B in quiescent mononuclear cells of volunteers (14), indicating that a deliberate change in the diet can change human gene expression; 11) These findings were confirmed and extended to human mononuclear cells stimulated by adherence (15).
A consequence—at times catastrophic—of coronary atherosclerosis is cardiac arrhythmias. Although each of the above described mechanisms for n-3 fatty acids by itself may not be sufficiently active to retard atherosclerosis or prevent its catastrophic sequelae, their joint activity may very well be.
STUDIES IN ANIMAL MODELS
In canines, swine, and nonhuman primates, studies with dietary n-3 fatty acids bore consistently positive results in all models of atherosclerosis and vasoocclusion investigated (16). However, results of studies in rats and rabbits have been inconsistent (16). Apparently, there are differences between species with respect to the effect of dietary n-3 fatty acids on models of atherosclerosis. Thus, it does not seem prudent to extrapolate findings in animal models to humans. Therefore, animal models will not be discussed further here.
HUMAN INTERVENTION STUDIES
Prevention of restenosis after percutaneous transluminal coronary angiography
The primary focus of many studies was the prophylaxis of restenosis after a percutaneous transluminal coronary angiography (PTCA; 16–18). On the whole, the first 7 studies conducted showed a favorable effect of dietary n-3 fatty acids (reviewed in 17). However, in subsequent and, more importantly, larger studies, no effect was detected (18, 19). Thus, fish oils are not currently considered to be effective in the prevention of restenosis after PTCA. Nevertheless, the question remains: Why are some studies positive and others negative in their results? Of note, 2 large studies with negative results used corn oil as a placebo, whereas others used olive oil or no placebo. The high rate of restenosis reported for corn oil, however, argues against an effect of corn oil obscuring the effect of n-3 fatty acids (18, 19). Pretreatment was thought to be important for all mechanisms of action to become effective. Although Leaf et al (18) included a minimum of 12 d of pretreatment, the results were negative. However, this question has been raised again by preliminary data of De Caterina et al (20) reported elsewhere in this supplement, pointing to an effect after a 1-mo pretreatment period. Dose-response, or better, incorporation-response analyses of n-3 fatty acids have not yet been conducted to my knowledge, and could yield information for the planning of future studies. The definition of restenosis differs from study to study, reflecting unresolved methodologic problems with quantitative coronary angiography (21, 22). However, a study relying on intravascular ultrasound to assess outcome has not been performed to my knowledge.
A long-held belief was that the prevailing mechanism of restenosis is a growth factor–dependent, rapid proliferation of subendothelial cells (18, 19). Coronary stents (intravascular prostheses made of a metal mesh) have been shown to reduce rates of restenosis from 40–50% to 20–25% (23). The lower rates of restenosis after placement of intracoronary stents argue against proliferation of subendothelial cells being the only mechanisms of restenosis after PTCA, because stents are not insurmountable obstacles to this proliferation. The success of stents, therefore, may be a result of the prevention of other mechanisms such as retractive processes and scar formation. Because of their success and relative ease of placement, stents will probably obtain a secure place in invasive cardiology. Thus, future strategies toward prevention of restenosis will probably aim toward the reduction of intimal growth in stents. Circumstantial evidence indicates that this process might be regulated by growth factors (13). n-3 Fatty acids had a mitigating effect on some of these growth factors (14, 15). This evidence taken together, the more select group of stented patients might be a better group in which to study prevention of restenosis after PTCA with n-3 fatty acids.
Progression of coronary atherosclerosis
The effect of dietary n-3 fatty acids on the progression or regression of coronary atherosclerosis has been investigated only twice. Sacks et al (24) reported a 2-y study in patients with angiographically documented coronary heart disease and normal plasma lipid concentrations (24). Forty-one patients (of whom 31 completed the trial) were randomly assigned to receive 6 g n-3 fatty acids from fish oil/d and 39 patients (of whom 28 completed the trial) received olive oil in identical capsules. The primary endpoint was the change in minimal luminal diameter, as assessed by quantification of coronary angiograms by a single operator blinded to treatment. In the fish-oil and olive oil groups, 179 and 126 coronary lesions were analyzed, respectively. The change in minimal luminal diameter was 0.104 mm in the fish-oil group and 0.138 mm in the olive oil group (NS). Forty percent of the coronary lesions investigated had a bypass, rendering the study population somewhat heterogenous.
Preliminary results of our own trial, the Study on Prevention of Coronary Atherosclerosis by Intervention with Marine Omega-3 Fatty Acids (SCIMO) have been published in abstract form (25). We tested the hypothesis that consumption of a fish-oil concentrate for 2 y would reduce a score to assess progression and regression of coronary atherosclerosis, as assessed by coronary angiography, by 50%. Identical opaque capsules contained either n-3 fatty acids (55% eicosapentaenoic and docosahexaenoic acids) or a placebo reflecting the average fatty acid composition of the European diet. The capsule contents weighed 1 g each, and the patients were instructed to ingest 6 capsules/d for the first 3 mo and 3 capsules/d for the subsequent 21 mo. The study was approved by the ethics committee of the medical faculty of the University of Munich and was conducted according to Good Clinical Practice, the pertinent guidelines of the European Union. From September 1992 to June 1994, 223 patients with mild coronary atherosclerosis gave informed consent and were recruited. Patient compliance was monitored by red blood cell phospholipid fatty acid analyses. Forty-nine patients did not complete the study, and follow-up coronary angiography was not performed on 12 patients who did complete the study. No patients were lost to follow-up. The code was broken after deposition of all relevant data at the trial monitor. One hundred eleven and 112 patients received n-3 fatty acids and placebo capsules, respectively. Randomization was successful with respect to age, sex, risk factors, lipid-lowering therapy, and other indexes. Analysis of all other results is being performed as of this writing in 1997.
Only one trial has addressed the question of whether n-3 fatty acids had an effect on coronary bypass patency 1 y after surgery (26). On the first postoperative day, 317 patients were randomly assigned to supplement their diet with 4 g 83% n-3 fatty acid concentrate/d; 293 patients received no supplement. Follow-up angiograms were performed on 302 patients in the n-3 fatty acid group and on 279 in the control group. The patency of internal mammaria grafts was not affected. However, of 635 and 595 distal anastomoses of venous bypass grafts, 174 and 196 were occluded in the n-3 fatty acid and control groups, respectively. This gave the patients receiving n-3 fatty acids an odds ratio of 0.77 (95% CI: 0.60, 0.99; P = 0.034) of distal anastomosis occlusion compared with the control group (26). Forty-three percent of the patients who received n-3 fatty acids had at least one vein graft occluded compared with 51% of the control patients (P = 0.05). n-3 Serum phospholipid concentration and response analysis showed a positive dose-response effect of n-3 fatty acids.
Few patients died in the studies mentioned previously, obviating a life table analysis for all studies combined. Unfortunately, definitions of other clinical endpoints, as well as the methods of reporting them, varied. Thus, a meaningful meta-analysis of the clinical course of the patients recruited for the studies cited is not possible from the published reports (17–19, 24–26).
Survival after myocardial infarction
The clinical course of survivors of a first myocardial infarction was investigated by the Diet and Reinfarction Trial (DART; 27). In a factorial randomized design, 2033 men either received or did not receive advice to alter their diet on average 41 d after the index event. Three recommendations were given. The first group was advised to reduce dietary fat and its composition according to the American Heart Association Step I diet. The second group was advised to eat fatty fish 2–3 times/wk or, for those unable to tolerate fish, to supplement their diet with fish-oil capsules, providing 0.5 g n-3 fatty acids/d. In this group, plasma eicosapentaenoic acid concentrations increased to 0.59% (P < 0.01 compared with the control group). The third group was advised to eat 18 g fiber/d. After 2 y, the patients given the fish advice had a 29% reduction in total mortality compared with the patients given no such advice (27). Interestingly, this reduction was not associated with a sustained reduction in serum cholesterol in the patients advised to consume fish or fish oil. Thus, the results of DART cannot be explained by extrapolation of similar results in trials in which cholesterol-lowering approaches were used (28). The authors speculated that the main reason for the 29% reduction in mortality was a reduced incidence of ventricular fibrillation during reinfarctions or some other acute myocardial ischemia. In the group advised to consume fish or fish oil, the incidence of nonfatal myocardial infarctions was higher, but not significantly so, than in the group given no advise. Thus, these data would also be consistent with smaller, and thus less lethal, infarctions occurring in the group advised to consume fish or fish oil. The results of DART have stimulated researchers of n-3 fatty acids. The fact that little money has been available for clinical studies with n-3 fatty acids in the cardiovascular area has, however, precluded confirmation of the results of DART.
Nevertheless, the results of the Lyon study (29) can, in a very broad sense, be regarded as supporting DART’s results. In a randomized study in patients similar to those studied in DART, subjects were given multifaceted dietary advice plus a margarine specifically designed to reflect a traditional Mediterranean diet. This diet incorporated an n-3 fatty acid [in this case -linolenic acid (18:3n-3)] and consisted predominantly of vegetables and fruit with some white meat, but no red meat, included. Plasma fatty acid concentrations were analyzed in 141 of the 302 experimental patients and in 139 of 303 control patients. In the experimental group, -linolenic acid increased to 0.62%, eicosapentaenoic acid increased to 1.03% (P < 0.001), and cardiovascular mortality was 70% lower at 5 y compared with the control group (29). Total cholesterol and LDL-cholesterol concentrations were identical in the control and experimental groups, pointing to effects independent of these 2 indexes.
Cardiac transplantation
In cardiac transplant patients, endothelium-dependent coronary vasodilation, as assessed after acetylcholine infusion, was normal after 3 wk of consuming 5 g dietary n-3 fatty acids/d, whereas it was pathologic in otherwise matched control subjects (30). This could contribute to the inhibition of transplant vasculopathy and longer cardiac graft survival seen in animal models of cardiac transplantation (31, 32). Whether n-3 fatty acids have an influence on cardiac transplant survival in humans is currently unknown.
Peripheral artery disease
In animal models of peripheral artery disease, beneficial effects of n-3 fatty acids have been shown in dogs, swine, and nonhuman primates (16). Studies in humans with clinical endpoints such as pain-free distance walking have been scarce and have shown no benefit of n-3 fatty acid consumption (16). Results derived from our ultrasound studies are pending. Arterial compliance seems to be improved after 6 wk of n-3 fatty acid supplementation in diabetic patients (33). Forearm vasoconstrictory response to pressors is reduced after ingestion of n-3 fatty acids, an effect thought to be mediated by the eicosanoid system (34). The effects of n-3 fatty acids on peripheral artery disease are much less well studied than are those on coronary artery disease. As yet, no clear-cut clinical benefit on peripheral arteries can be discerned.
CONCLUSION
In coronary artery disease, consumption of dietary n-3 fatty acids has been found to reduce mortality after a first myocardial infarction. Although confirming studies are needed, dietary n-3 fatty acids appear to be of value in secondary prevention of coronary artery disease. At present, we have no evidence that dietary n-3 fatty acids induce regression of coronary atherosclerosis. Ingestion of n-3 fatty acids appears to reduce occlusion of coronary venous bypass grafts. An interesting focus of research is on heart transplant patients. Large-scale clinical studies with endpoints such as morbidity and mortality are needed to more precisely define the role of n-3 fatty acids in primary and secondary prevention of coronary atherosclerosis.
ACKNOWLEDGMENTS
We deeply appreciate the contributions of P Angerer, W Kothny, K Baumann, S Störk, WE Kaminski, and E Jendraschak to the original work.
REFERENCES