Cardiovascular diseases and red meat

Abstract

Lean Australian red meat cuts are low in fat and have a ratio of cholesterol-raising saturated fatty acids (SFA) to cis-monounsaturated fatty acids (MUFA) to cis-polyunsaturated fatty acids (PUFA) of around 24:40:14. This is less cholesterol-raising than was earlier estimated, because cuts are now leaner and part of the SFA is stearic acid (that does not raise plasma cholesterol). and there are several other (cholesterol-lowering) PUFA as well as linoleic acid present. Low-fat, predominantly monounsaturated lean meat cuts have been shown to be acceptable in cholesterol-lowering diets. This does not mean that meat eaten with the fat on will not raise plasma cholesterol. Meat is low in sodium, high in potassium and has been shown in human dietary experiments not to raise the blood pressure. Meat is high in protein and contributes to weight reduction by increasing satiety and helping reduce intake in ad-lib weight-reducing diets. Overweight increases the risk of increased plasma cholesterol, increased blood pressure and diabetes. Meat is a good source of bioavailable iron. The hypothesis that people with high iron stores have increased risk of heart disease has not been confirmed in a number of epidemiological studies. Human studies suggest that dietary long-chain omega-3 PUFAs are protective against sudden cardiac death, consistent with lower risk of ventricular fibrillation. Most fat in the human diet is a mixture of triglycerides. The pattern of fatty acids attached to the glycerol has been known since 1956 to affect the plasma cholesterol concentration.1 The hardest evidence comes from strictly controlled metabolic ward experiments with metabolically normal human subjects. Saturated fatty acids (SFAs) raise the plasma cholesterol, polyunsaturated fatty acids (PUFAs) lower it, and monounsaturated fatty acids (MUFAs) have an intermediate, neutral effect.1, 2 The cholesterol-raising effect of SFAs is about twice as potent as the lowering effect of PUFAs. This is expressed in Ancel Keys classic equation:2Δcholesterol = 2.74 ΔSFAs − 1.3 ΔPUFAs (where Δ = change; plasma cholesterol is in mg/100 mL and fatty acids are estimated as percentage of total daily calories). Hundreds of human experiments have since been conducted, many papers published, and we now have meta-analyses which strongly confirm the original findings.3, 4 This is still the main message for public health education, but for health professionals, there are further modifications. Serum cholesterols nowadays are usually expressed in SI units, as mmol/L, so the numbers in the original Keys equation need to be divided by 38.6. Beyond serum total cholesterol, a knowledge of LDL-cholesterol (increases risk) and HDL-cholesterol (decreases risk) and triglycerides can better predict risk of coronary heart disease (CHD). Individual SFAs differ in their cholesterol-raising effect. PUFAs have different physiological effects if they belong to the omega-6 or omega-3 series. And different plant-derived oils, which contain predominantly MUFAs, do not all have the same effect on serum cholesterol, presumably because of the other lipids they contain (phytosterols, squalene).5 This paper outlines the effect of dietary components on risk of cardiovascular disease and considers the role of red meat in this context. Of the SFAs, only lauric (12:0), myristic (14:0) and palmitic (16:0) raise plasma cholesterol levels. Myristic is the most potent in this regard,6 and palmitic is the most abundant of all three in foods. SFAs with 10 or fewer carbon atoms (in medium chain triglycerides) do not appear to raise serum cholesterol.7 At the other end of the series, it has been repeatedly found that stearic acid (18:0) has little or no cholesterol-raising effect.8, 9 It is rapidly converted to oleic acid in vivo. To estimate the effect of fatty acid pattern on serum cholesterol, it is better to refer to fatty acids with carbon chains of 12:0 + 14:0 + 16:0, rather than total SFA. Among the PUFAs, by far the most abundant in foods and oils is linoleic acid, cis 18:2, n-6. Hence, any effect on cholesterol attributed to PUFAs is due very largely to linoleic acid. The intake of PUFAs from food reported in the Australian 1995 National Nutrition Survey of 12.5 g/day10 was around 10 times greater than that reported for the linolenic acid (18:3), and about 100 times greater than that reported for eicosapentaenoic acid (EPA 20:5) + docosapentaenoic acid (DPA 22:5) + docosahexaenoic acid (DHA 22:6)—of about 130 mg/day.11 Linoleic acid is the only fatty acid known to lower plasma and LDL-cholesterol levels even when it is added to the diet, increasing fat and energy intakes.1, 12 Other common PUFAs, linolenic (18:3, omega-3)7 and arachidonic13 (20:4, n-6), do lower plasma cholesterol, but this is of little practical importance because of their low levels in most foods. High-dosefish oils may raise plasma LDL-cholesterol a little. This is perhaps due to SFA in the fish oil, together with EPA and DHA, but they raise HDL-cholesterol too.7 Beneficial effects of fatty fish and fish oils on cardiovascular disease are not from plasma cholesterol lowering, but from effects that include reducing the risk of ventricular arrhythmias and the tendency to thrombosis. All that has been written above assumes that the fatty acids have double bonds in the usual, natural CIS configuration. But if the MUFA or PUFA is in the TRANS configuration, the effect on serum cholesterol is similar to that of SFA14 and, with high intakes, HDL-cholesterol may be lowered as well.15 Most human experiments have been conducted with trans 9, 18:1, called elaidic acid. Trans-unsaturated fatty acids occur naturally in small percentages in ruminant meat and milk fat. They are produced by microorganisms in the rumen. The main natural trans-fatty acid in red meat is trans 11, 18:1, called vaccenic acid. Trans-fatty acids are also produced during hydrogenation of vegetable and fish oils to make harder fats in food processing. Here the isomers are chiefly trans 9 and trans 10, 18:1. Around half of the trans-fatty acids in the British diet16 and Australian diet17 were estimated to come from animal foods, and half from margarines and other processed fats.16 With reduction of the industrially produced trans-fatty acids since then, the proportion from ruminant fat may be higher. Effects of different fatty acids on total-cholesterol levels are approximately mirrored by effects on LDL-cholesterol levels. As to HDL-cholesterol levels, SFAs 12:0, 14:0 and 16:0 raise it, and unsaturated fatty acids have little effect on it.18 Fasting plasma triglycerides are lowered by omega-3 PUFAs. Effects of fatty acids on different plasma LDL- and HDL-cholesterol are summarised in Table 1. There is now a vast body of literature on the role of fatty fish, fish oils and long-chain omega-3 PUFAs in the prevention and amelioration of heart disease, as well as in supporting the development of infants and having a positive impact on several chronic diseases and mental function. This summary concentrates on CHD and long-chain omega-3 PUFA provided in diet as fatty fish (and less on the more numerous human experiments with pharmacological doses of fish oil). Interest started with observations in Greenland Eskimos living traditional lifestyles, who had high intakes of animal fat but low incidence of CHD. Their animal fat was largely marine animals—fish, seal, etc. Dyerberg et al. in the 1970s found their plasma fatty acids contained unusually high amounts of eicosapentaenoic acid (EPA),19 and they suggested this reduced the tendency to thrombosis via an altered pattern of prostanoids.20 In the early 1980s, researchers reported that fatty fish, as well as fish oils (both rich in omega-3 PUFAs), produced a much greater reduction of plasma triglycerides and VLDL than oils or foods rich in n-6 linoleic acid.21 By the mid-1980s, there were two further major discoveries. Kromhout and colleagues (one of the Seven Country Study team) reported that in the Zutphen (Dutch) cohort, people who ate more fish (about 1/3 fatty) experienced significantly fewer CHD deaths.22 The mechanism was not via any of the known risk factors for CHD; possibly there was reduced platelet aggregation. They produced a recommendation for one or two fish dishes a week in dietary guidelines for the prevention of CHD.22 At about the same time, researchers in Australia were finding that rats fed dietary fat rich in PUFAs showed significantly fewer episodes of ventricular fibrillation when a coronary artery was occluded. The protective effect of tuna fish oil was stronger than sunflower seed oil, especially in the reperfusion phase.23 These animal experiments were confirmed by others in the United States, working with cultured neonatal rat cardiac myocytes, that can be seen to contract rhythmically under the microscope. It was then found that polyunsaturated fat in the medium counteracts the effects of pharmacological agents that usually initiate arrhythmias.24 Since then, at least 13 cohort studies have been completed, involving over 200 000 subjects. Most studies found protective effects of fish, and these were significant in countries with high rates of CHD and in studies of high scientific quality.25 Two meta-analyses (with a somewhat different selection of studies) both found significantly lower relative risk of CHD death in regular fish eaters,26, 27 and He et al.27 showed a striking dose–response relationship, where eating fish two to four times per week reduced the mortality risk to 0.77. From the animal experiments it would be expected that the effect of fatty fish consumption would be on prevention of sudden deaths due to ventricular fibrillation. This was indeed found in two intervention studies: with fish in Wales (the DART study),28 and with fish oil capsules in the Italian GISSI study.29 Later trials in patients with implanted cardioverter/defibrillators have not provided such clear answers,30, 31 but these are very complex cases to manage. In the largest randomised controlled trial of patients with implanted cardioverter/defibrillators (n = 402), the trend in favour of fish oil was impressive but not statistically significant, although technical difficulties with recordings and poor compliance in taking the large fish oil capsules were reported.30 Recently, five expert bodies have made recommendations regarding omega-3 PUFA intakes. In 2004, the US Food and Drug Administration concluded that supportive, but not conclusive, research shows that consumption of EPA and DHA omega-3 may reduce the risk of CHD. The Scientific Advisory Group of Food Standards Australia New Zealand likewise considered the evidence was probable that fatty fish containing omega-3 PUFAs reduces the risk of CHD. In 2003, the Joint WHO/FAO Expert Consultation32 concluded that the evidence was convincing that fish and fish oils (EPA and DHA) reduce the risk of cardiovascular disease. The (Australian) National Heart Foundation33 advises that, to lower the risk of CHD in the general population, adults should consume two to three serves of fish (preferably oily fish) per week, thus obtaining 500 mg/day of marine omega-3 PUFA. The National Health and Medical Research Council's Dietary Guidelines for Australian Adults (2003, pp. 117–122),10 as the third point in its 20 pieces of practical advice on fats, say ‘Try to include in your diet fish high in omega-3 polyunsaturated fats—for example, sardines, tuna, salmon and herring’. When considering any possible effects of long-chain omega-3 PUFAs, note that trials with fish oils use much higher intakes than are obtainable from ordinary diets. In trials with food relatively high in omega-3 PUFAs, the OPTILIP Study34, 35 found reduced fasting plasma triglycerides but no change in haemostatic factors or insulin. The Adelaide and Perth study (which did not get an acronym)36 found no differences in blood pressure, insulin or lipids, although red cell PUFA increased 50%. All this advice is about EPA (20:5) and DHA (22:6), the predominant omega-3 PUFAs in fish oil and oily fish. The third long-chain omega-3 PUFA is docosapentaenoic (DPA, or 20:5). Its concentration is about one-sixth of that of EPA and DHA in fish, and little seems to be known about its biological activity. The averages of these three fatty acids in herring, mackerel, pilchards, salmon, sardines and tuna are: 20:5 = 8.25%, 22:5 = 1.43% and 22:6 = 6.55% total fatty acids.37 When eaten, cholesterol, the sterol in the cell membrane of land animals, tends to raise plasma cholesterol, but less than might be expected. About half of the plasma cholesterol is synthesised endogenously from acetate in the liver, and there is a feedback mechanism so that this is downregulated if more is absorbed. Effects of dietary cholesterol on plasma cholesterol are inconsistent,38 evidently affected by the fatty acid composition of the diet and varying between individuals.39 In shellfish, the sterols are not of the cholesterol form;40 chromatography shows that they are mostly phytosterols from seaweed. Earlier bans on oysters, for example, in cholesterol-lowering diets were based on a chemical method that did not differentiate other sterols from cholesterol. Phytosterols (β-sitosterol, campesterol, stigmasterol, brassicasterol) are the corresponding sterols of plant cell membranes. They have a very similar chemical structure to cholesterol, only differing in one or two extra carbons at the side-chain end of the molecule. They reduce absorption of cholesterol by competitive inhibition. This affects both dietary cholesterol and cholesterol excreted in the bile, which is normally partly re-absorbed. In purified form (and higher intake), they can lower plasma cholesterol by a larger percentage than dietary cholesterol raises it.41 Dietary fibre,42, 43 even coffee,44 can affect plasma lipids, but the most important dietary factor, other than type of dietary fat, is excess energy, leading to overweight. Overweight people tend to have raised plasma triglycerides and higher total LDL-cholesterol and lower HDL-cholesterol.45 Weight reduction by diet and/or exercise will usually reduce their cholesterol and triglyceride levels. In 1981 Sullivan46 suggested that the lower incidence of CHD in premenopausal women, compared with men and postmenopausal women, could be due to higher iron stores in the latter two groups. A number of studies have since investigated this possibility. As red meat is a major source of bioavailable iron, this question is of potential concern for the safety of high intakes. The Institute of Medicine in its dietary reference intake (DRI) report on 14 nutrients,47 including iron, reviewed the literature on high iron status and CHD to determine the tolerable upper intake level. They found: Serum ferritin and CHD: three positive, five negative studies Transferrin saturation and CHD: all five studies negative Serum iron and CHD: all four studies negative Total iron binding capacity and CHD: all four studies negative Sempos explained why the US Food and Nutrition Board had concluded that there was not enough evidence to support the hypothesis.48 At least one of the positive studies was seriously flawed. Also, there is no consistent convincing evidence that having haemochromatosis, especially the heterozygous form, is associated with an increased risk of CHD. This was against the background that iron deficiency remains the most common nutritional deficiency in the USA.48 In the large West of Scotland pravastatin trial, genotypes for the haemochromatosis gene were assayed in 482 people who developed a CHD event and 1100 who did not.49 There were no significant differences. A recent trial of phlebotomy in 1277 US veterans with symptomatic peripheral arterial disease and 4.5-year follow up found no difference in outcome between phlebotomy and controls or with reduction of iron stores.50 Perhaps this topic can best be summed up by noting that the ANZ NRV report11 did not address the iron stores/CHD question. A recent paper by the Harvard epidemiologists Qi et al. suggests that dietary haem iron is a risk factor for CHD in women with type 2 diabetes.51 The three tables in the paper do not appear to show how much haem iron or meat (including pig meat) was recorded for the different quintiles. The authors themselves admit that residual confounding by saturated fat was unavoidable even after careful adjustment. The Australian Dietary Guidelines (ADG) Report10 presents the SFA and MUFA contents of lean red meat as quite low, especially in lean beef (total fat 1.8 g/100 g), and about equal, with smaller amounts of PUFA. Recent data on individual fatty acids in red meat52 expressed as percentages of total fatty acids suggest, for example that lean beef rump contains less cholesterol-raising SFAs and more PUFAs (Table 3). As per cent of total fatty acids, the values for: SFA (14:0 + 16:0) are 2.6 + 21.8 = 24.4% cis-MUFA (16:1 + 18:1) are 3.0 + 36.5 = 39.5% cis-PUFAs are 5.6 + 0.8 + 0.7 + 2.8 + 1.9 + 2.9 = 14.7% trans-fatty acids (18:1, 18:2 and 18:3) are 2.5 + 0.5 + 0.3 = 3.3% Most of the balance is stearic acid (13.2%) plus small amounts of 15:0, 17:0 and 14:1 fatty acids. The P/S ratio here is 0.60, whereas if the table in the Dietary Guidelines report is used, it would be only 0.22. Even if conjugated linoleic acid, 20:3, 20:5 and 22:6, are left out, the effective P/S for plasma cholesterol level is at least 10.3/24.4 = 0.42. The effect of red meat on plasma cholesterol in practice corresponds to what would be predicted from these recent Australian analyses of individual fatty acids in red meat cuts. Several controlled human studies find that lean red meat has a roughly neutral effect among foods on plasma cholesterol, suggesting it can be included in a cholesterol-lowering diet. A number of studies have shown the benefits of including lean red meat in prudent diets for cholesterol lowering. A group at St Thomas's Hospital London were able to reduce total- and LDL-cholesterol considerably in 15 free-living men with hyperlipidaemia on a diet, including increased P/S ratio and extra fruit and vegetables, that contained 180 g/day of lean red meat (8.5% fat). They concluded that providing care is taken to reduce total dietary fat, a moderate amount of meat and meat products may be included in a cholesterol-lowering diet.53 In Australia, Kestin et al. compared two fat-modified diets in 26 healthy men. One diet was vegetarian (LOV); the other contained 250 g/day lean meats (LOM). Total cholesterol fell 5% on diet LOM (10% on LOV). They concluded that a more widely accepted lean meat-containing prudent diet was almost as effective at lowering cholesterol levels as a plant-based prudent diet.54 Another Australian study involved 10 healthy subjects (men and women) given a very low-fat diet containing lean beef (500 g/day). Serum total cholesterol fell 20% and rose when beef dripping was added. The authors concluded that it was the beef fat, not lean beef itself, that was associated with elevations in cholesterol concentrations. The suggestion was that lean beef could be included in cholesterol-lowering diets (low saturated fat) provided that the meat was free of all visible fat.55 A further study in Texas, USA, gave 38 free-living men with hypercholesterolaemia a reduced-saturated-fat diet that included about 170 g either lean beef (8% fat) or chicken (7% fat) for five weeks. LDL-cholesterol fell 9% and 11% respectively, but the difference was not statistically significant. The authors concluded that lean beef and chicken were interchangeable in the NCEP Step 1 diet.56 In a long-term trial conducted in Chicago, Minneapolis and Johns Hopkins Lipid Clinic, 191 men with moderate hypercholesterolaemia were given 170 g/day of either lean red meat or lean white meat in their NCEP Step 1 diets. At the end of 36 weeks, LDL-cholesterol declined 1.7% and HDL-cholesterol increased 2.3% on the red meat; changes in the white meat group were not significantly different. (Lean red meat here, as is usual in the USA, was beef, veal or pork; lean white meat was poultry or fish.)57 Two other points are important about meat and plasma cholesterol. First, there is as much cholesterol in lean meat as in meat fat, chicken contains rather more cholesterol than beef, and kidney and liver contain four or five times more than muscle meat (see Table 2). These organs are richer in several good nutrients as well.37 Second, the fat of red meat has a more unfavourable fatty acid pattern than lean muscle meat. It contains more 14:0 and 16:0 SFA and less PUFA (see Table 3). By the method suggested above (omitting stearic acid), the P/S ratio in this lean meat is 0.42, while in the fat, it is 0.07. Of course, it contains about 20 times more total fatty acids than the lean muscle meat. This difference was first noted by Crawford,58 who compared the fatty acid patterns between very lean wild herbivores like African antelopes and domestic cattle. The ADGs advise to choose foods low in salt.10 Red meat is a low-sodium food. Australian analyses average 54 mg (2.3 mmol) Na/100 g in lean beef,59 and British food tables give similar figures for lean beef cuts (raw).37 On the other hand, lean red meat is rich in potassium. Australian figures average 360 mg for beef and 343 mg/100 g for lamb, that is, about 9 mmol K/100 g. The Na : K molar ratio is therefore 0.26. Hodgson et al.60 measured blood pressures of two groups of 30 people with a moderate degree of hypertension. One group ate 250 g/day of red meat for eight weeks in place of carbohydrate-rich foods that the parallel control group was eating. Repeated systolic blood pressures were significantly lower (average 4 mmHg) in the meat group. The authors suggest that the extra taurine and arginine in the meat may have added to the effect of 25 mmol less sodium in reducing blood pressure. The ADGs advise against weight gain and the background papers discuss the role of diet in the prevention and management of overweight and obesity.10 The report focuses on the dangers of a high body mass index but does not provide a review of which diets can be recommended. In the few years since the ADGs report was prepared, there has been some research and much debate to find weight-reducing diets that are effective, acceptable and safe. The popular ‘CSIRO Total Wellbeing diet’61 may owe much of its success to the high protein content (33% of total energy intake). Noakes and Clifton based this diet on their experience with increased-protein, energy-restricted diets reported in at least seven papers published in the international literature during the period 2000–2004.61 In one paper,62 they report that over 12 weeks, the 27%-protein diet resulted in better changes in total- and LDL-cholesterol, in triglycerides and in blood glucose, compared with a 16%-protein diet, with no adverse effects on calcium metabolism or bone. Body weight and biomarker reductions in the men were not significantly different between the two diets, possibly because of the fairly short duration of the trial. Earlier, Skov et al.63 showed greater weight reduction with a protein intake of 25% energy, compared with 12% over a six-month ad libitum dietary trial. Other workers have had similar results with a six-month trial.64 Increased dietary protein has been shown in human experiments to increase satiety and diet-induced thermogenesis.65, 66 The mechanisms for these changes have not yet been elucidated.66 Low-fat, predominantly monounsaturated, lean meat cuts have been shown to be acceptable in cholesterol-lowering diets. Meat is a good source of bioavailable iron, which does not appear to be a risk factor for heart disease. Meat is low in sodium and high in potassium, and meat consumption does not appear to raise blood pressure. Overweight increases the risk of increased plasma cholesterol, increased blood pressure and diabetes. Meat is high in protein and contributes to weight reduction by increasing satiety and helping reduce energy intake in ad-lib weight-reducing diets.