On 189–192 of this issue of Alcohol and Alcoholism, Bleich et al. (2001) report the results of an experimental study demonstrating that consumption of moderate amounts of alcohol (30 g per day) for 6 weeks by healthy social drinkers increases the circulating concentration of the cardiovascular risk factor homocysteine (HCy), irrespective of the type of the alcoholic beverage consumed (beer, red wine or spirit). Given the current emphasis on the cardiovascular health benefits of moderate alcohol consumption, this study merits the attention of the medical community, public health authorities and political decision-makers.
It must, however, be emphasized from the outset that this, the first of its kind, is a preliminary study in which Dr Stephan Bleich and his colleagues have assessed only 60 subjects, who were further divided into four groups of 15 each. Numbers are therefore very small indeed, and the need to replicate the findings cannot be over-emphasized. If confirmed (and there is no apparent reason to suspect their validity), these findings will have important implications for everyone who drinks alcohol and for those responsible for public health policies and related issues.
HCy is a cardiovascular risk factor of multifactorial implications. The first evidence linking HCy with vascular disease was provided by McCully (1969), who reported the presence of extensive arterial thrombosis and atherosclerosis in two children with elevated plasma HCy concentration and homocystinurea. Subsequent investigations confirmed McCully's findings and suggested that hyperhomocyst(e)inaemia is an independent risk factor for atherothrombosis and atherosclerosis in the coronary, cerebral and peripheral vascular system(s) (Ueland and Refsum, 1989; Kang et al., 1992; Stampfer et al., 1992; McCully, 1996), and a number of more recent studies have further confirmed the HCy–vascular relationship (see, e.g., Arnesen et al., 1995; Graham et al., 1997; Nygård et al., 1997; Wald et al., 1998). In their meta-analytical study, Boushey et al. (1995) calculated that 10% of the risk of coronary heart disease in the general population can be attributed to HCy, and it is noteworthy that an increase in HCy of about 20% is associated with a similar increase in the cardiovascular disease risk (Verhoef et al., 1998). Excellent reviews of the HCy–vascular relationship have recently been published (Weir and Scott, 1998; Welch and Loscalzo, 1998), some important points from which will be further emphasized below.
Homocysteine is formed by demethylation of the essential amino acid methionine, the average daily dietary intake of which is at least 1.8 g (Mayer et al., 1996). Total homocyst(e)ine (tHCy), which is measured in plasma, comprises homocysteine, homocystine, mixed disulphides of homocysteine, and also homocysteine thiolactone. The normal plasma tHCy concentration range is 5–15 μmol/l in the fasting state (Ueland et al., 1993; Jacobsen et al., 1994). However, as there is a gender difference in tHCy levels, two separate normal ranges are used in some laboratories. In that of Weir and Scott (1998), the normal ranges for women and men are 6–12 and 8–14 μmol/l, respectively.
Kang et al. (1992) classified abnormally high levels of tHCy under fasting conditions into: (1) moderate (15–30 μmol/l); (2) intermediate (30–100 μmol/l); (3) severe (>100 μmol/l). Apart from genetic abnormalities in homocysteine metabolism, which are generally rare, there are a number of conditions associated with elevated levels of plasma tHCy. These include chronic renal failure (Wilcken and Gupta, 1979; Chauveau et al., 1993), hypothyroidism (McCully, 1996), pernicious anaemia (Savage et al., 1994) and several types of carcinomas (Mayer et al., 1996). Plasma tHCy concentration can also be elevated by drugs and other xenobiotics, such as methotrexate, phenytoin (Ueland and Refsum, 1989; Ueland et al., 1992) and theophylline (Ubbink et al., 1996). These drugs act by depleting one or two of the cofactors of enzymes of HCy metabolism, namely folate and pyridoxal 5′-phosphate. It is noteworthy that cigarette smoking interferes with the synthesis of this latter cofactor (Nygård et al., 1995) and smokers have lower levels of pyridoxal 5′-phosphate (Vermaak et al., 1990), thus suggesting a potential mechanism for smoking-induced atherogenesis, if tHCy levels can be shown to be higher in smokers.
Vascular disease related to hyperhomocyst(e)inaemia can, in general, be attributed to three major pathogenetic mechanisms: vascular endothelial dysfunction, vascular smooth muscle proliferation, and coagulation abnormalities. Several specific mechanisms have been implicated. These include: (1) endothelial injury leading to platelet activation and thrombus formation (Harker et al., 1974, 1976; James, 1990), possibly initiated by (2) production of reactive oxygen species (such as superoxide anions, hydrogen peroxide and hydroxyl radicals) during auto-oxidation of HCy (Harker et al., 1974; Welch et al., 1997); (3) decreased availability of the vaso-protective nitric oxide by HCy, through decreased synthesis (Loscalzo, 1996; Welch et al., 1997), and decreased expression of nitric oxide synthase by lipid peroxides (Liao et al., 1995), whose production could be further enhanced by HCy's ability to also decrease the expression of glutathione peroxidase (Upchurch et al., 1997); (4) induction of proliferation of smooth muscle cells (Harker et al., 1983; Tsai et al., 1994, 1996). Other mechanisms are also possible, for which evidence is currently less robust.
Plasma tHCy levels can be increased if homocysteine synthesis is enhanced or if its degradation is impaired. Enhanced HCy synthesis occurs mainly when levels of its precursor methionine are raised after excessive dietary protein intake, and, in a recent study in rats, Stead et al. (2000) showed that plasma tHCy concentration is increased in animals fed a high protein diet; an effect that is most likely the result of not only enhanced synthesis in, but also increased transport from, the liver, the major organ of amino acid metabolism. HCy degradation occurs via: (1) remethylation to methionine by the action of methionine synthase, a cobalamin (vitamin B12)-dependent enzyme which also uses N5-methyl-tetrahydrofolate as a methyl donor; (2) trans-sulphuration with serine to form cystathionine by the action of cystathionine β-synthase, an enzyme requiring pyridoxal 5′-phosphate (the active form of vitamin B6) as a cofactor. Impaired degradation of HCy leading to its accumulation in plasma could therefore occur if availability of the above three cofactors is decreased secondarily to nutritional deficiencies of these three B vitamins. In fact, nutritional deficiency of vitamin B is by far the commonest cause of elevated plasma tHCy. Thus, elevated levels of tHCy occur in patients with nutritional deficiencies of vitamin B12 (Brattström et al., 1988b) and folic acid (Kang et al., 1987; Stabler et al., 1988), and negative correlations have been demonstrated in normal subjects between plasma tHCy concentration on the one hand and the serum levels of vitamins B12 and B6 and folate on the other (Selhub et al., 1993). These latter authors further suggested that low (or inadequate) levels of one or more of these B vitamins contribute to approximately two-thirds of all cases of hyperhomocyst(e)inaemia. It follows therefore that nutritional supplementation with the above B vitamins, particularly folate, could reduce the elevated tHCy to ‘normal’ levels and, in fact, a randomized double-blind placebo-controlled cross-over trial in 75 men and women with coronary artery disease of folic acid-fortified breakfast cereal has demonstrated such a reduction (Malinow et al., 1998). These latter authors also showed that, whereas daily intake of 127 mg of folic acid (the level recommended by the US Food and Drug Administration) produced only a 3.7% decrease in plasma tHCy, cereal supplementation with larger amounts resulting in daily folic acid intakes of 499 and 665 mg caused significant decreases in plasma tHCy of 11.0 and 14.0%, respectively.
Treatment of hyperhomocyst(e)inaemia of course depends on the underlying cause. That which cannot be attributed to the drugs and medical conditions listed above can be simply treated with dietary vitamin supplementation. In most cases, however, irrespective of aetiology, a 1–5 mg daily dose of folate will cause a rapid decrease in plasma tHCy concentration (Brattström et al., 1988a). As well as by folate alone, levels of tHCy have also been shown to be decreased by folate combined with vitamins B12 and B6, or by these two vitamins alone (Saltzman et al., 1994). It is noteworthy that the decrease in mortality from cardiovascular causes since 1960 has been correlated (McCully, 1996) with the increase in vitamin B6 supplementation of food.
Returning to alcohol, it is reasonable to suggest that, if moderate consumption of alcohol increases the circulating concentration of the cardiovascular risk factor homocysteine (Bleich et al., 2001), which is also elevated in chronic alcoholism (Carvo and Camilo, 2000; Bleich et al., 2001 and references cited therein), then two recommendations aimed at reducing the cardiovascular risk are worthy of consideration. First, the recommendation of moderate drinking as a cardiovascular health aid, particularly to those who have been shown not to benefit from it, namely males under 35 years of age, pre-menopausal women (Beaglehole and Jackson, 1992; Edwards, 1994; Anderson, 1995) and possibly also post-menopausal women treated with hormone-replacement therapy, as their tHCy levels are known to be low (van der Mooren et al., 1994) needs to be revised. There is a strong case for the argument that, if those who do not benefit from drinking alcohol do not drink, their HCy-related cardiovascular risk may even be decreased. These proposals should also be considered against the background of: (1) the relatively low strength of the cardiovascular protective effect (Friedman and Klatsky, 1993); (2) the modest contribution of this protective effect of only a 3-month increase in life expectancy in relation to mortality from cardiovascular causes (Dufour, 1994), which can be achieved by other less controversial health measures; (3) the increasing evidence against the beneficial effect of moderate alcohol consumption itself, due to the many confounders introduced in comparisons with different types of reference groups, including lifetime abstainers, each of which with different lifestyles, characteristics and vulnerability factors, which can inflate the level of benefit (Fillmore et al., 1998; Leino et al., 1998; Shaper and Wannamethee, 1999; Fillmore, 2000; Dawson, 2000). Secondly, since it seems that drinking moderate or larger amounts of alcohol is associated with an increase in this cardiovascular risk factor, a prudent policy would be to encourage those who drink alcohol at any level to have an adequate intake of folate and vitamins B6 and B12, to reduce their tHCy. Because tHCy presents this risk at any level, even the lowest within the normal range (Arnesen et al., 1995; Wald et al., 1998), it is important that levels should be reduced to the lowest possible extent with the help of vitamin intake. As the study by Malinow et al. (1998) showed, supplementation of breakfast cereal with folic acid at levels above those recommended by the Food and Drug Administration resulted in 11–14% decreases in plasma tHCy concentrations. Although these decreases seem modest, at the population level they could be very significant indeed. Another way to help keep levels of tHCy down is for the general public to avoid excessive intake of proteins and adhere to the daily recommended level, unless different amounts are clinically indicated.
It is clear also that more research is needed at various levels to assess and delineate the roles, relative contributions and interactions between alcohol consumption at its various levels, the nutritional vitamin B status, dietary protein intake, dietary lipid intake, tobacco smoking, and tHCy and other risk factors in cardiac and other vascular diseases, with adequate controls for variables and confounders. Bleich et al. (2001) suggested, as a consequence of their results, that the ‘French paradox’ will need to be re-explained. Could this paradox be explained by differences in tHCy levels? Do the French have a higher folate and other B vitamin intakes than the British and/or a lower protein intake, or just a greater variety of food selection? Or, is the ‘French paradox’ simply a myth created by an under-certification bias (Lang et al., 1999). Such studies as suggested above may yield new explanations in this important area of public health.