Alpha-linolenic acid in the prevention and treatment of coronary heart disease

Transcription

1 European Heart Journal Supplements (2001) 3 (Supplement D), D26 D32 Alpha-linolenic acid in the prevention and treatment of coronary heart disease M. de Lorgeril 1, P. Salen 1, F. Laporte 2 and J. de Leiris 1 1 Laboratoire du Stress Cardiovasculaire et Pathologies Associées and 2 Laboratoire de Biologie du Stress Oxydant, UFR de Médecine et Pharmacie, Grenoble, France Alpha-linolenic acid is one of the two essential fatty acids in humans. Epidemiological studies and dietary trials strongly suggest that this fatty acid is important in relation with the pathogenesis (and prevention) of coronary heart disease. Like other n-3 fatty acids from marine origin, it may prevent cardiac arrhythmias and sudden cardiac death. The optimal dietary intake of alpha-linolenic acid seems to be about 2 g per day or 0 6 1% of total energy intake. Obtaining an optimal ratio of the two essential fatty acids, linoleic and alpha-linolenic acids i.e. a ratio of about 4 to 1 in the diet is a major issue. The main sources of alpha-linolenic acid for the European population should be canola oil (and canola-oil-based margarine if available), nuts (English walnut), ground linseeds and green leafy vegetables such as purslane. (Eur Heart J Supplements 2001; 3 (Suppl D): D26 D32) 2001 The European Society of Cardiology Key Words: Diet, essential fatty acid, walnut, canola oil, purslane, myocardial infarction. Introduction and basic biochemistry Alpha-linolenic acid (ALA or 18:3n-3) is one of the two essential fatty acids in humans. The other is linoleic acid (LA or 18:2n-6). The term essential means that these fatty acids must be supplied in the diet because the body needs them but cannot synthesize them. Humans lack the enzymes to introduce double bonds at carbon atoms beyond the C-9 in the fatty acid chain. LA and ALA obtained from foods furnished by the diet are the starting point for the synthesis of a variety of other unsaturated fatty acids. After ingestion, ALA is converted to very-long-chain polyunsaturated fatty acids (PUFAs), i.e. readily to eicosapentaenoic acid, 20:5n-3, and more slowly to docosahexaenoic acid, 22:6n-3. Using the same pathways (the same enzymes) and in competition with ALA, LA is converted into arachidonic acid, 20:4n-6, which is, in competition (again) with eicosapentaenoic acid, the starting point for the synthesis of eicosanoids and prostaglandins that are important mediators in many inflammatory diseases and in particular in cardiovascular diseases. Briefly, the prostaglandins, thromboxanes and leukotrienes derived from 20:5n-3 have different biological properties from those derived from 20:4n-6, i.e. their global effects result in less vasoconstriction, platelet aggregation and leukocyte toxicity. When the n-6/n-3 ratios decrease as a result of Correspondence: Dr M. de Lorgeril, Laboratoire du Stress Cardiovasculaire et des Pathologies Associées, UFR de Médecine et Pharmacie, Domaine de la Merci, Grenoble, France X/01/0D $35.00/0 dietary changes, 20:5n-3 competes with 20:4n-6 for eicosanoid metabolism at the cyclooxygenase and lipoxygenase levels in platelets and leukocytes. As a result, the balance between metabolites stimulating platelet and leukocyte (and also with vasoconstrictive properties) and metabolites with opposing properties is displaced toward those with antithrombotic, antiinflammatory and vasodilative properties. A major consequence of ALA deficiency is that its chief synthetic end-products, especially 22:6n-3, are not adequately produced. Because 22:6n-3 is a major component of the phospholipid membranes of the brain and retina, its deficiency in these organs leads to abnormal function [1,2]. The myocardium and the myocardial membrane phospholipids are also rich in 22:6n-3. This is very important in the case of ischaemia, in particular for the ability of the ischaemic or the post-ischaemic myocardium not to develop arrhythmias [3]. Deficiency in n-3 fatty acids is higher when the dietary supply of LA (found in corn and sunflower oils, for instance) is high, because of competition between ALA and LA for their entry in the elongation and desaturation pathways leading to the synthesis of long-chain PUFAs and then of eicosanoids. Thus, a high ratio of n-6 to n-3 PUFAs in the diet tends to accentuate a deficiency in ALA [4]. The paradox of the competition between LA and ALA is that there is a huge difference in their concentrations in the blood. For instance, in people following a Western type of diet, plasma concentration of ALA is about 0 30% of the total plasma fatty acids, whereas the concentration of LA is about 30%. Thus, the ratio of LA 2001 The European Society of Cardiology

2 Alpha-linolenic acid and coronary heart disease D27 18:1 (n-9) 18:2 (n-6) 20:4 (n-6) Table 1 Antioxidant, LA and ALA content of purslane and spinach leaves corresponding to 100 g (one serving) of fresh food Spinach Purslane (chamber-grown) Wild purslane 18:3 (n-3) 20:5 (n-3) Alpha-tocopherol (mg) Beta-carotene (mg) Ascorbic acid (mg) ALA (mg) LA (mg) ALA, alpha-linolenic acid; LA, linoleic acid. Modified from Reference min Figure 1 Each peak in the plasma chromatogram (obtained by capillary gas chromatography, with helium as the carrier gas) represents one specific fatty acid with a specific retention time on the column. The main unsaturated fatty acids are indicated. The peak corresponding to ALA, marked 18:3 (n-3), is well demarcated. Note the major difference with LA, marked 18:2 (n-6), in the size of the peak. All fatty acids are then expressed as proportion of total fatty acid peak area. to ALA in the blood is about 100 to 1. Since the enzymes involved in LA and ALA metabolism (elongases and desaturases) prefer ALA to LA [5], a small change in dietary intake results in highly significant changes in the n-6 to n-3 ratios in both the plasma and cell membranes [6]. Despite its low concentration, ALA is easily assayed with capillary gas chromatography, the corresponding peak being well identified on the chromatogram (Fig. 1). Nutrition background Coming back to the question of the potential role of ALA in pathology, symptoms of ALA deficiency have not yet been specifically outlined in the cardiovascular area. The potential effects of ALA on cardiovascular diseases were discovered when it was reported that populations whose plasma lipids are rich in ALA (the Greek and Japanese cohorts of the Seven Countries Study, for instance) are apparently protected from cardiovascular diseases [7], and that these low-risk populations consume foods rich in ALA [8,9]. Simopoulos and colleagues showed that purslane (largely consumed around the Mediterranean basin) is the richest source of ALA in any green leafy vegetables [8]. Purslane is also rich in antioxidants (Table 1). This is a major point, as ALA, because of its three double bonds, is highly sensitive to oxidation and high intake of ALA must be balanced with high intake of antioxidants to protect it from oxidation. The same group also reported that eggs 30 from range-fed Greek hens (which make a feast of purslane and other ALA-rich fresh green grass) are rich in ALA and other n-3 fatty acids compared with eggs bought in U.S. supermarkets which are, in turn, usually rich in LA [9]. The large difference in the fatty acid composition of the two types of eggs was also due to the fatty acid composition of the industrial foods given to U.S. hens. This suggested that egg yolk may have been a non-negligible source of ALA and other n-3 fatty acids for people living in the Mediterranean area, one of the regions of the world where the incidence of (and mortality from) coronary heart disease (CHD) is low. Taking the numbers given by Simopoulos, two egg yolks from range-fed Greek hens would provide about 0 5 gof ALA, about ten times the amounts provided by two supermarket eggs. Another important point in favour of encouraging ALA consumption and of reducing LA intake came from the observation that the lowest ex vivo platelet aggregation measured in humans in response to adenosine diphosphate, a possible indicator of the risk of acute CHD events, was recorded with a LA to ALA ratio in the diet of about 4 to 1 [10,11], i.e. much lower than that of the present Western diet and also lower than what many experts currently recommend. This pointed out the potential importance of ALA in the prevention of the thrombotic complications of CHD [12]. However, only trials could bring definite information regarding causal relationships. The next section will provide some data regarding recent dietary trials. Dietary trials The Lyon Diet Heart Study is a secondary prevention trial designed to test the hypothesis that a Mediterranean ALA-rich diet may improve the prognosis of patients having survived a first acute myocardial infarction [13]. The design, methods and results of the trial have been reported [13 15]. A striking protective effect of the Mediterranean diet was found, with a 50 70% reduction in the recurrence risk after 4 years of follow-up [16]. Briefly, in terms of lipids, the experimental Mediterranean diet tested in the trial supplied under

3 D28 M. de Lorgeril et al. Hydrogenated fat or coconut fat Butter Beef fat Lamb fat Pork fat Chicken fat Soybean oil Olive oil Sunflower oil Walnut oil Linseed oil Canola oil 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Saturated fat 18:1 (n-9) 18:2 (n-6) 18:3 (n-3) Figure 2 Fatty acid content (expressed as percentage of total fatty acids) of important fats and oils used in the Western diet. Hydrogenated fat, coconut fat, sometimes beef fat, which are resistant to oxidation and to repeated heating, are currently used to make French fries in fast food and non-fast food restaurants. This is obviously unhealthy, as high saturated fat intake has been associated with high risk of CHD in many epidemiological studies. The vegetable oils (and derived margarines) currently marketed in the Western countries are often too rich in polyunsaturated fatty acids (of the omega-6 family) which are very sensitive to oxidation. This is obviously unhealthy. The same goes for linseed oil, in which the amounts of both ALA and LA are too high. As shown in the table, the best combination between saturated, polyunsaturated and monounsaturated (essentially oleic acid) fatty acids is obtained in olive and canola oils. Canola oil and canola-oil-based margarine should be used as a major source of ALA in secondary prevention of CHD. 30% of energy from fats, and under 8% of energy from saturated fats. Regarding the essential fatty acids, the intake of LA was restricted to 4% of energy and the intake of ALA made up over 0 6% of energy. In practical terms, the dietary instructions were detailed and customized to each patient [13 15] and can be summarized as: more bread, more cereals, more legumes and beans, more fresh vegetables and fruits, more fish, less meat (beef, lamb, pork) and delicatessen foods, which were to be replaced by poultry; no butter and cream, to be replaced by an experimental, canola oil-based, margarine. This margarine was chemically comparable with olive oil but slightly enriched in LA and more so in ALA, the two essential fatty acids. Finally, the oils recommended for salad and food preparation were exclusively olive and canola (erucid acid-free rapeseed oil) oils. The scientific rationale for that dietary fat strategy is illustrated in Fig. 2. To meet the criteria of a Mediterranean diet, patients had to reduce drastically the consumption of foods rich in saturated (essentially animal) fat, shown in the upper part of the figure. Among the vegetable oils (the lower part of the figure), only olive oil (despite its lack of ALA) and canola oil (despite its quite high amounts of LA) have a fatty acid composition in line with our strategy. Thus, the patients were advised to use both oils. Because of their high content in LA, soybean, sunflower and walnut oils should not be used daily for food preparation and salad dressing. Peanut oil is too rich in saturated fatty acids and LA and linseed oil is too rich in PUFAs. In theory, the best option should be to vary the use of several oils. However, when considering the difficult conditions of daily life of many patients and of their families, we decided to try and simplify our advice and to recommend the exclusive use of olive and canola oils. The exclusive use of olive and canola oils (and of canola-oil-based margarine instead of butter to spread on bread) to prepare meals and salad was a major issue in that trial, as it resulted in significant differences in the fatty acid composition of both circulating plasma lipids (essentially, lipoproteins) and cell membrane phospholipids [6]. As shown in Table 2, the main differences between groups in platelet phospholipid fatty acids were not seen at the level of individual fatty acids (ALA is almost not detectable in cell membranes) but for the entire family of each group. Significant differences were also seen for the ratio between n-6 and n-3 fatty acids [6]. When comparing the dietary fatty acids in the two groups, control patients consumed about 0 7 g of ALA per day compared with about 1 8 g in the experimental group, with a ratio LA to ALA of about 10 to 1 in controls against about 4 to 1 in the experimental group.

4 Alpha-linolenic acid and coronary heart disease D29 Table 2 Platelet phospholipid fatty acids in the two groups of the Lyon trial Mediterranean (1) Control (1) P Saturated fatty acids 14: < : ns (2) 18: Sum ns Unsaturated fatty acids Omega-9 family 18:1n :1n :3n ns Sum Omega-6 family 18:2n ns 20:3n ns 20:4n ns 22:4n <0 05 Sum Omega-3 family 20:5n ns 22:5n ns 22:6n ns Sum (1) Results are mean SEM. (2) ns=non significant. Modified from Reference 6. Since the Mediterranean diet tested in that trial was different from the control diet in many other aspects than the LA to ALA ratio (less saturated fat, more antioxidants of various sources, more vitamins of the B group including folic acid, probably more vegetable proteins, etc.), the next question was to try and specify the precise role of ALA in the cardioprotection observed in the trial. Using multivariate analyses and adjustment for several confounding variables, we found that ALA plasma concentrations measured 2 months after randomization were significantly (and inversely) associated with the risk of recurrence, and in particular with fatal recurrences [16]. It could be said, however, that it is not ALA per se that was protective but the long-chain PUFAs derived from ALA, eicosapentaenoic and docosahexaenoic acids, which were also increased in the plasma of patients in the experimental group [11]. These long-chain n-3 have been shown to prevent ventricular fibrillation (VF) and sudden cardiac death (SCD) in animal experiments [17,20] and in human trials [21,22]. However, in the Lyon trial, these fatty acids were not significantly associated with a lower risk, suggesting that ALA was the main protective factor. Also, a specific antiarrhythmic effect of ALA itself was reported in animal studies [19,20]. This does not mean, however, that the benefits from ALA are not due, at least partly, to its conversion to long-chain PUFAs, and further studies are required to identify the individual effects of each n-3 fatty acid in the context of myocardial ischaemia and ventricular arrhythmias. Another randomized controlled trial in which the consumption of ALA was encouraged (patients in the experimental group were advised to eat more fruit, vegetables and nuts) reported a significant reduction of the risk of cardiac events in post-acute myocardial infarction (AMI) patients [23]. In that trial, the main source of ALA was nuts (Table 3). Earlier works on Seventh Day Adventists [24] and American nurses [25] suggested that eating nuts was associated with a diminished risk of CHD. Potentially protective constituents of nuts include ALA, folates, magnesium, potassium, fibre, vitamin E, arginine [26] and favourable lysine-to-arginine and methionine-to-arginine ratios [27,28]. One noteworthy point is that the fatty acid composition of the lipids in the walnuts (also called English walnut or noix de Grenoble or Californian nut) is apparently the same in Table 3 Essential fatty acid, oleic acid, folates, arginine and methionone content of various foods (per 100 g of edible portion) Oleate (g) LA (g) ALA (g) Arginine (mg) Methionine (mg) Folates (μg) Walnut Hazelnut Linseed Peanut Butternut Almond Pecan nut Pine nut Macadamia Mackerel Salmon Beef Pork Data from the USDA Nutrient Database (Web site) and Répertoire Général des Aliments (JC Favier et al., Lavoisier Editors).

5 D30 M. de Lorgeril et al. the U.S.A. and in Europe, as the concentrations of ALA, LA and oleic acid measured in our laboratory are respectively 13 4, 60 5 and 15 6% in the noix de Grenoble and 12 9, 62 5 and 13% in the Californian nut. In Table 3, the contents of two major amino acids, arginine (the precursor of nitric oxide) and methionine (the precursor of homocysteine), in various nuts are detailed because of their importance in cardiovascular physiology and pathophysiology. The contents are compared with those of meat (beef and pork) and fatty fish (salmon and mackerel), which are the main sources of proteins and indispensable amino acids in the Western diet. Folic acid is also indicated because of its importance in these pathways: for retro-conversion of homocysteine to methionine on one hand, and for recycling tetrahydrobiopterin (which is folic acid-dependent), which may account for a dysfunction of nitric oxide synthase when its availability is reduced, on the other hand. It is clear from Table 3 that, beside their fatty acid composition, certain nuts may be important because they provide large amounts of arginine (even more than meat) with the major advantage that they are quite poor in methionine (the precursor of the vasculo-toxic homocysteine) compared with meat and fish. This is especially true for walnut, almond and hazelnut which are commonly eaten in great amounts by the Mediterranean populations. The last, but certainly not least, point concerns the high folate content of most nuts. Low serum folate levels [29] and high homocysteine levels [30] have been clearly associated with an increased risk of CHD. Results of a recent trial are consistent with a decreased CHD risk following homocysteine-lowering treatment with folic acid (plus vitamin B6) and suggest a causal relationship between low folic acid with high homocysteine levels and CHD [31]. Thus, although there is no opportunity here to fully discuss each of these points, nuts (which are both rich in ALA and important in the Mediterranean diet) can obviously be included as part of a healthy diet. In contrast, walnut oil, too rich in LA, should not be advised. Finally, with its very high content in ALA, linseed oil should also not be advised owing to its excessive global content in PUFAs (LA plus ALA). As outlined [11], despite their cardioprotective properties when given in small amounts, n-3 PUFAs, as well as PUFAs in general, should not be given in large amounts: small is beautiful. Thus, when considering a given food and its potentially protective constituents, we must also look at the other, potentially harmful (depending on the dose) components. Good examples are fish and fish oil. When taken in moderation, for example in DART and GISSI [21,22], they are cardioprotective. In contrast, when taken in large doses in low-risk population, they do not appear to be protective [32]. Epidemiological studies Although the association of ALA and cardiovascular end-points has received much less attention than that with long-chain n-3 PUFAs (usually from fish oil), several studies have been published reporting significant inverse associations between ALA (measured in the diet, plasma or adipose tissue) and the risk of CHD in quite large populations [33 38]. In two of these studies, however, this association was no longer significant after adjusting for other variables [34,36]. In the Health Professional Follow-up Study in the U.S.A. ( health professionals aged years followed-up for 6 years), Ascherio and colleagues reported that ALA intake was inversely associated with the risk of AMI and of cardiac death; the association was strengthened after adjustments (in particular for traditional risk factors and total fat intake) giving a relative risk of 0 41 for each 1% increase in energy [37]. It is interesting that in the same cohort the intake of (marine) long-chain n-3 PUFAs was not associated with any risk increase or decrease [39], suggesting again that the cardioprotective effects of ALA itself may be different from those of these longer chain n-3 fatty acids. In the Nurses Health Study ( women followed-up for 10 years), a higher intake of ALA was associated with a lower relative risk of fatal CHD [38]. The adjusted relative risk from the lowest to the highest quintile was 0 55 [38]. Further adjustments for intake of fish (n-3) fatty acids, oleic acid, trans fatty acids, cholesterol, folate or fibres did not materially alter the association. For non-fatal AMI, the association was not significant. When the relationship between ALA and fatal CHD was examined using vitamin E supplements, the investigators observed a stronger association among supplement users, the relative risk between the extreme quintiles becoming as low as Finally, in support of this association, Hu and colleagues found that specific foods which are primary dietary sources of ALA in the U.S. diet (oil and vinegar salad dressings) were also associated with a risk reduction [38]. Thus, these epidemiological data suggest that moderate amounts of ALA in the diet may be cardioprotective and, importantly, no harmful side-effects have been detected. From theory to practice There are different ways to obtain g of ALA per day in the diet without using ALA-containing capsules or fortified foods. The simplest (and easiest) way is to use canola oil for food preparation and salad dressing. As 100 g of canola oil provide about 8 g of ALA, two small (U.S.) tablespoons may provide about 2 g of ALA. The canola-oil-based margarine, which contains about 5 g of ALA per 100 g of margarine, may be useful. To obtain 2 g of ALA, 35 g of margarine, or about six teaspoonsful are needed. Alternatively, one can eat one tablespoonful of canola oil (with salad, for instance) and two teaspoonsful of canola margarine on a piece of bread. Both canola oil and margarine can be used in association with olive oil (which does not contain ALA) for food preparation because of the large amounts in oleic acid and flavonoids and also because of its particular taste. Regarding nuts, to obtain 2 g of ALA, four to

6 Alpha-linolenic acid and coronary heart disease D31 five English walnuts (or noix de Grenoble) are needed. If English walnuts are liked enough to eat them every day (as Mediterranean people often do), it is possible to use exclusively olive oil for food preparation. A good (and succulent) way is to add the nuts to the salad. One can also use ground linseeds (not linseed oil) with the salad or other green leafy vegetables knowing that 100 g of ground linseeds provide about 23 g of ALA. Thus, only one tablespoonful of ground linseeds provides about 2 g of ALA. Eating purslane (or other equivalent ALA-rich leafy vegetables) to obtain large amounts of ALA is less easy, in particular because in many areas purslane is not available. In addition, one large portion (100 g of purslane) provides less than 0 4 g of ALA. Thus, for most adults, purslane should be associated with either walnuts or a canola-oil-containing salad dressing. One advantage of eating green leafy vegetables (and not only purslane) is to increase the diversity of the diet which is probably a major component of any healthy diet [40]. For those who think that it is easier to take supplements (capsules, for instance) of ALA rather than ALA-rich foods, it is important to remember that nutrients are not drugs. It means that when we eat foods, we eat a combination of nutrients in proportions which are not the result of chance but the product of a very long natural evolution. For instance, the proportions of certain PUFAs and many different antioxidants in a given food are probably the best combination to protect PUFAs against peroxidation. Conclusion Epidemiological studies as well as dietary trials including moderate amounts of ALA in the experimental diet suggest that this essential fatty acid, despite its low concentrations in blood and tissues, may be important in relation with the pathogenesis (and prevention) of CHD. It is still not known whether ALA is cardioprotective of itself or also through its conversion in long-chain (n-3) PUFAs and then in the corresponding eicosanoids and prostaglandins. The definite demonstration of a specific cardioprotective effect of ALA would require a trial specifically designed for that purpose. Nevertheless, one potential advantage of ALA, compared with the longchain n-3 PUFAs found in marine products, is that ALA competes with LA at the first step of the elongationdesaturation pathway, whereas the long-chain n-3 PUFAs compete with the n-6 PUFAs only at the late steps of eicosanoid metabolism. Thus, ALA (but not long-chain n-3 PUFAs) is capable of preventing the accumulation of long-chain n-6 PUFAs. According to our current knowledge, dietary ALA should represent about 0 6 1% of total daily energy or about 2 g per day. With the exception of algae (or seaweeds) and other rarely consumed marine products, the main sources of ALA in our diet could be canola oil and canola-oil-based margarines in moderation. Nuts are also a good source of ALA (and also of folates and arginine) but they are rich in LA and therefore they should be eaten in moderate amounts, about three to five walnuts per day. The consumption of green leafy vegetables rich in ALA, for instance purslane, should be encouraged not only because of their content in ALA but also because of the antioxidant nutrients they contain. References [1] Connor WE, Neuringer M, Reisbick S. Essential fatty acids: the importance of n-3 fatty acids in the retina and brain. Nutr Rev 1992; 50: [2] Neuringer M, Connor WE, Lin DS, Barstad L, Luck S. Biochemical and functional effects of prenatal and postnatal omega-3 fatty acid deficiency on retina and brain in rhesus monkeys. Proc Natl Acad Sci USA 1986; 83: [3] Nair SD, Leitch J, Falconer J, Garg ML. Cardiac (n-3) non-esterified fatty acids are selectively increased in fish oilfed pigs following myocardial ischemia. J Nutr 1999; 129: [4] Connor WE. Alpha-linolenic acid in health and disease. Am J Clin Nutr 1999; 69: [5] Emken EA, Adolf RO, Rakoff H, Rohwedder WK. Metabolism of deuterium-labeled linolenic, linoleic, oleic, stearic and palmitic acid in human subjects. In: Baillie TA, Jones JR, eds. Synthesis and Applications of Isotopically Labeled Compounds. Amsterdam: Elsevier Science Publishers, 1989: [6] de Lorgeril M, Salen P. Modified Mediterranean diet in the prevention of coronary heart disease and cancer. World Rev Nutr Diet 2000; 87: [7] Sandker GN, Kromhout D, Aravanis C et al. Serum cholesteryl ester fatty acids and their relation with serum lipids in elderly men in Crete and Netherlands. Eur J Clin Nutr 1993; 47: [8] Simopoulos AP, Norman HA, Gillapsy JE, Duke JA. Common purslane: a source of omega-3 fatty acids and antioxidants. J Am Coll Nutr 1992; 11: [9] Simopoulos AP, Salem N. n-3 Fatty acids in eggs from range-fed Greek chickens. N Engl J Med 1989; 321: [10] Renaud S. Linoleic acid, platelet aggregation and myocardial infarction. Atherosclerosis 1990; 80: [11] Renaud S, Nordoy A. Small is beautiful: alpha-linolenic and eicosapentanoic acids in man. Lancet 1983; i: [12] Renaud S, de Lorgeril M. Dietary lipids and their relation to ischemic heart disease: from epidemiology to prevention. J Intern Med 1989; 225: [13] de Lorgeril M, Renaud S, Mamelle N et al. Mediterranean alpha-linolenic acid-rich diet in secondary prevention of coronary heart disease. Lancet 1994; 343: [14] de Lorgeril M, Salen P, Martin JL et al. Effect of a mediterranean type of diet on the rate of cardiovascular complications in coronary patients. Insights into the cardioprotective effect of certain nutriments. J Am Coll Cardiol 1996; 28: [15] de Lorgeril M, Salen P, Caillat-Vallet E, Monjaud I, Hanauer MT, Mamelle N. Control of bias in dietary trial to prevent coronary recurrences. The Lyon Diet Heart Study. Eur J Clin Nutr 1997; 51: [16] de Lorgeril M, Salen P, Martin JL et al. Mediterranean diet, traditional risk factors and the rate of cardiovascular complications after myocardial infarction. Final report of the Lyon Diet Heart Study. Circulation 1999; 99: [17] McLennan PL, Abeywardena MY, Charnock JS. Reversal of arrhythmogenic effects of long term saturated fatty acid intake by dietary n-3 and n-6 polyunsaturated fatty acids. Am J Clin Nutr 1990; 51: [18] McLennan PL, Abeywardena MY, Charnock JS. Dietary fish oil prevents ventricular fibrillation following coronary occlusion and reperfusion. Am Heart J 1988; 16:

7 D32 M. de Lorgeril et al. [19] Isensee H, Jacob R. Differential effects of various oil diets on the risk of cardiac arrhythmias in rats. J Cardiovasc Risk 1994; 1: [20] Billman GE, Kang JX, Leaf A. Prevention of sudden cardiac death by dietary pure omega-3 polyunsaturated fatty acids in dogs. Circulation 1999; 99: [21] Burr ML, Fehily AM, Gilbert JF et al. Effects of changes in fat, fish, and fibre intakes on death and myocardial reinfarction: Diet And Reinfarction Trial (DART). Lancet 1989; 2: [22] GISSI Prevenzione Investigators. Dietary supplementation with n-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: results of the GISSI Prevenzione trial. Lancet 1999; 354: [23] Singh RB, Rastogi SS, Verma R et al. Randomised controlled trial of cardioprotective diet in patients with recent acute myocardial infarction: results of one year follow-up. Br Med J 1992; 304: [24] Fraser GE, Sabaté J, Beeson WL, Strahan TM. A possible protective effect of nut consumption on risk of coronary heart disease. Arch Intern Med 1992; 152: [25] Hu FB, Stampfer MJ, Manson JE et al. Frequent nut consumption and risk of coronary heart disease in women: prospective cohort study. Br Med J 1998; 317: [26] Cooke JP, Tsao P, Singer A et al. Anti-atherogenic effect of nuts: is the answer NO? Arch Intern Med 1993; 153: [27] de Lorgeril M. Dietary arginine in prevention of cardiovascular diseases. Cardiovasc Res 1998; 37: [28] Kritchevski D, Tepper SA, Czarnecki SK, Klurfeld DM. Atherogenicity of animal and vegetable protein: influence of the lysine to arginine ratio. Atherosclerosis 1982; 41: [29] Robinson K, Arheart K, Refsum H et al. Low circulating folate and vitamin B6 concentrations. Risk factors for stroke, peripheral vascular disease and coronary heart disease. Circulation 1998; 97: [30] Bostom AG, Selhub J. Homocysteine and arteriosclerosis. Subclinical and clinical disease associations. Circulation 1999; 99: [31] Vermeulen EG, Stehouwer CD, Twisk JW et al. Effect of homocysteine-lowering treatment with folic acid plus vitamin B6 on progression of subclinical atherosclerosis: a randomised, placebo-controlled trial. Lancet 2000; 355: [32] Marckmann P, Gronbaek M. Fish consumption and coronary heart disease mortality. A systematic review of prospective cohort studies. Eur J Clin Nutr 1999; 53: [33] Miettinen TA, Naukkarinen V, Huttunen JK, Mattila S, Kumlin T. Fatty acid composition of serum lipids predicts myocardial infarction. Br Med J 1982; 285: [34] Wood DA, Riemersma RA, Butler S et al. Linoleic and eicosapentanoic acids in adipose tissue and platelets and risk of coronary heart disease. Lancet 1987; i: [35] Dolecek TA, Grandits G. Dietary polyunsaturated fatty acids and mortality in the Multiple Risk Factor Intervention Trial (MRFIT). In: Simopoulos AP, Kifer RR, Martin RE, eds. Health Effects of Polyunsaturated Fatty Acids in Seafoods. Basel: Karger, 1991: [36] Guallar E, Aro A, Javier Jimenez F et al. Omega-3 fatty acids in adipose tissue and risk of myocardial infarction. The EURAMIC Study. Arterioscler Thromb Vasc Biol 1999; 19: [37] Ascherio A, Rimm EB, Giovanucci EL et al. Dietary fat and risk of coronary heart disease in men: cohort follow-up study in the United States. Br Med J 1996; 313: [38] Ascherio A, Rimm EB, Giovanucci EL et al. Dietary intake of marine n-3 fatty acids, fish intake and the risk of coronary disease among men. N Engl J Med 1995; 332: [39] Hu FB, Stampfer MJ, Manson JE et al. Dietary intake of alpha-linolenic acid and risk of fatal ischemic heart disease among women. Am J Clin Nutr 1999; 69: [40] Kant AK, Schatzkin A, Ziegler RG. Dietary diversity and subsequent cause-specific mortality in the NHANES I Epidemiologic Follow-up Study. J Am Coll Nutr 1995; 14: