Reduced Heart Disease
Phenolic substances may be the compounds responsible for the reduced incidence of coronary heart disease seen in people who regularly consume wine. Wine has been espoused for centuries as a superior beverage. Louis Pasteur said, 'Wine is the most healthful and most hygienic of beverages,' and Plato remarked, 'No thing more excellent nor more valuable than wine was ever granted mankind by God.' Common to these endorsements is the view that wine is as good for the body as it is for the spirit. Modern scientific research supports this perception. David Goldberg at the University of Toronto has recently published an amusing and enlightening review of this subject.
Fifteen years ago, in an ecological epidemiology study, Selwyn St Leger and co-workers showed that there was a population-based association between a reduction in deaths from heart disease and increased wine consumption.(2) More recently. Serge Renaud and Michel de Lorgeril at INSERM in Lyon brought the issue to public attention with a similar study 'Wine, alcohol. platelets. and the French paradox for coronary heart disease.' They used World Health Organization data to show that dairy fat consumption is highly correlated with coronary heart disease (CHD) mortality. A few French cities, however, had very high fat consumption, yet low CHD mortality rates - thus the 'French paradox'. When they added wine consumption as a factor that affected CHD mortality, the researchers got a better correlation, with wine being a negative correlate - it appeared to reduce heart disease.(3) Michael Criqui and Brenda Ringel at the University of California, San Diego, subsequently investigated comparable data and came to a similar conclusion: wine was one of the few dietary factors that correlated with reduced CHD mortality.(4) Interestingly, they also showed that fruit consumption correlated with reduced CHD mortality.
There now appears to be no dispute that moderate wine consumption is associated with lower CHD mortality. A related question is whether or not total mortality rates decrease with increased wine consumption. and here there is still some argument. In an ecological study that compared entire populations, Criqui and Ringel showed that total mortality does not decrease as the population's wine alcohol consumption increases. The authors attribute this effect to a compensating increase in mortality from other causes, which offsets the (decreasing CHD mortality. However, in prospective studies which distinguish between subjects based on consumption rates, the lowest total mortality occurs with moderate alcohol consumption (1-3 drinks/day), whether from beer, wine or spirits.(5) For heavy drinkers, however, mortality is higher than non-drinkers, especially among women. Other studies agree that the lowest mortality occurs at moderate alcohol consumption levels. including the analysis of health professionals by Eric Rimm and co-workers, at Harvard School of Public Health.(6)
Despite the issues surrounding total mortality, it is clear that moderate alcohol consumption itself reduces CHD mortality. Several mechanisms for this are now recognised, of which the best known is alcohol's ability to alter blood lipid levels by lowering total cholesterol and raising high density lipoprotein (HDL) levels.(7)
But does wine confer any special benefit, and if so, is there an organic explanation for this effect that is unrelated to alcohol? Criqui and Ringel's ecological data clearly show that wine alcohol consumption (correlation coefficient, r = -0.66) is much more strongly correlated with reduced CHD mortality than total alcohol consumption (r = -0.39). Also, when Arthur Klatsky and Mary Ann Armstrong of the Kaiser Permanente Medical Center in Oakland, California, singled out wine drinkers in their prospective study, the drinkers exhibited a lower CHD mortality role than the other subjects.(8) Thus, in two types of studies, wine appears to have a special benefit. In their US-based study, Klatsky and Armstrong raised a concern that the correlation between wine consumption and reduced CHD mortality may not lie due to wine by itself, but perhaps to other lifestyle factors associated with wine consumption. For instance, in the US wine drinking correlates with increased income, which itself is related to reduced mortality rates. On the other hand, Criqui and Ringel's study is not subject to such a bias because the correlation between income and wine consumption does not hold true across the developed nations surveyed.
While future epidemiology studies can be designed to take into account such factors that still raise questions, epidemiological investigations are never capable of determining specific causes for observed effects. Therefore, it is important for chemists and biologists to establish whether or not there is a molecular mechanism by which wine nutrients could affect CHD.
One report has helped to define such an area for investigation. Michael Hertog and co-workers at the DLO State Institute for Quality Control of Agricultural Products, Wageningen, observed a direct correlation between reduced CHD mortality in elderly men and dietary levels of flavonols, one of the major classes of flavonoid phenolics.(9) The investigators divided the subjects into three groups based on the content of flavonols in their diet. Flavonol consumption was based on their reported diets and the flavonol content of those foods.(10) The best correlation between reduced CHD mortality and specific dietary components was with tea, onions and apples. The best correlation between CHD mortality and chemical constituents of the diet was with one flavonol: quercetin.(9) This was an interesting outcome because previous studies had shown negligible absorption and rapid clearance of pure quercetin in humans.(11) Surprisingly, Hertog's group discounted any other phenolic compounds, even other flavonoids, as having any potential effect on CHD.(9)
Wine is a particularly rich dietary source of flavonoid phenolics, so many studies to uncover a cause for wine's effects have focused on its phenolic constituents, particularly resveratrol and the flavonoids.
The phenolic compounds are well known to enologists for their sensory properties, and for this reason their chemistry has been investigated in wine for decades. These substances give wine its bitterness and astringency, and are the foundation of long ageing, since they are effective antioxidants. Wines low in these substances, such as white wines, rarely age gracefully.
Resveratrol has been studied intensely of late, partly because it was only in 1992 that E H Siemann and Leroy Creasy at Cornell University first described it in wine (see Figure 1 ).(12) The researchers focused attention on resveratrol by suggesting that it could reduce CHD mortality, a contention based on the compound's ability to inhibit platelet aggregation and reduce lipid levels in hyperlipidaemic rats. These effects were noted in studies to elucidate the activities of piceid, a glucoside of resveratrol, and the main component of 'Koji jon,' a traditional Chinese medicine prepared from the roots of Polygonum cuspidatum and used to treat atherosclerosis.(13) More. recently, resveratrol has also been shown to inhibit the oxidation of human low density lipoproteins (LDL).(14) In addition, another study has shown that resveratrol, as well as some flavonoids, inhibits eicosanoid synthesis, a key step in platelet aggregation.(15)
This flurry of research activity has been augmented by the more recent discovery of resveratrol's cis isomer and piceid in grapes and wine.(16) To date at least ten different chromatographic methods have been described to analyse wine and/or grapes for resveratrol. Perhaps the latest is a liquid chromatographic method that separates the cis and trans isomers of both resveratrol aud piceid. (16)
The low levels of resveratrol in wine, averaging l-2mg/l trans-resveratrol and perhaps 5mg/l for all derivatives, has not abated the interest in it. An estimate of the maximum resveratrol blood levels that could arise from wine consumption - making a number of generous assumptions, including the consumption of a substantial amount of wine (500ml) that contains high levels of resveratrol (5mg/l), and efficient absorption (50%) and slow metabolism, for a 75kg individual with 5 litres blood - is about luM. This is enough to attain the lowest resveratrol concentration known to elicit the in vitro effects mentioned above, typically 1-10uM. Certainly, the levels needed in vivo may be different from those seen to be effective in vitro, but based on these assumptions, the likelihood that resveratrol from wine would have a physiological effect on a wine drinking population appears to be low.
Of course there are many other phenolic phytochemicals in wine that occur at a much higher concentration than resveratrol. Some of the most abundant are found at more than 100mg/l. These are mainly the flavonoids (see Figure 2), which include four major classes: the flavonols, the anthocyanins, the catechins or flavanols, and oligomers (procyanidins) and polymers (tannins) of the catechins The total amount of these compounds present is 1-3g/l in red wines and 0.2g/l in whites. In addition, wine contains a significant amount of non-flavonoids, including the hydroxy-cinnamates, benzoic acids, stilbenes and others, typically 0.2-0.4g/l for all wines. The number of individual phenolic compounds in a particular wine varies, but a typical liquid chromatographic analysis of red wine will reveal 50 constituents - closer scrutiny will reveal many more.
In many situations, but not all, these pheno]ic compounds are antioxidants. To oversimplify the disease of atherosclerosis. or clogging of the arteries, the oxidation of LDL is an important step in the development of arterial plaque,(17) and so substances that can block this oxidation should slow the disease. In fact, some antioxidants have slowed atherosclerosis in animals. Thus the question arises, could wine phenolics affect atherosclerosis in humans? This issue has been addressed in a recent issue of Biochemist dedicated to the question of the nutritional value of antioxidants.(18)
A key study of wine by Edwin Frankel and colleagues at the University of California, Davis, (19) showed that wine contained antioxidants towards the oxidation of LDL in vitro. Thus it is possible that these substances inhibit LDL oxidation in vivo and, as a result, slow the development of arterial plaque. Based on this interpretation, these authors suggested that the pbenolic compounds in wine were responsible for the French paradox. A follow-up investigation on three wine phenolics showed that all were antioxidants for LDL. The most potent were epicatechin and quercetin, while resveratrol was less potent, and the control, alpha-tocopherol (vitamin E) was least so. (14)
These in vitro studies are interesting but do not establish whether or not the oxidation of LDL is inhibited by wine antioxidant constituents in vivo. Two studies have attempted to answer this question indirectly. One showed that wine consumption increased blood antioxidant activity. (20) Unfortunately, this method analysed serum which was prepared by first clotting the blood sample. Since wine phenolics inhibit clotting,(15,21) it is likely that the measured change in antioxidant capacity was at least in part a result of differential clotting. The other study showed that daily wine consumption reduced the tendency of isolated LDL to oxidise.(22) Unfortunately, in both these investigations, the chemical composition of the wine used was not characterised, and so it is impossible to ascribe the observed effects to any specific components.
Since the aggregation of platelets (thrombosis) is an important factor in precipitating a heart attack, compounds that reduce platelet activity could also reduce CHD mortality. Wine flavonoids have been shown to inhibit platelet aggregation.(21) They appear to do this by specifically inhibiting oxygenase enzymes.(23) It also seems that there are significant differences in the ability of the different flavonoids to affect platelets - quercetin is potent, while catechin is not. (15)
The significance of these in vitro studies has recently been complemented by in vivo studies using a well developed animal thrombosis model. Cyclic flow reductions are used in animals as a model for thrombosis - chemicals that reduce the cyclic flow reductions are thought to lower the chance of platelet aggregation and hence thrombosis. The consumption of wine or flavonoids greatly attenuated cyclic blood flow reductions.(24) Since grape juice also decreased the blood flow reductions (albeit with three times as much volume) it appears that non-alcoholic constituents are the active components. It is unknown whether the different potencies are the result of altered levels of phenolics or an effect of ethanol.
Phenolics must absorbed into the blood stream to have any direct effect on coronary diseases, but data in the field of flavonoid absorption by human are very sparse. In the available data, there is a striking difference between two of the compounds that have been well studied, catechin and quercetin. (25) About half the administred dose of catechin was absorbed, while there was no detectable absorption of quercetin. Clearly there are profound differences in tbe behaviour of these two compounds, and at present there is no explanation for this. In addition, if quercetin is not absorbed from wine or other foods by humans, then its nutritional significance would be very limited.
Wine is certainly not the only dietary source of phenolic compounds. Fresh fruits are a rich source and, notably, Criqui and Ringel found that fruit consumption correlated highly with reduced CHD mortality.(4) Wine phenolics all originate from the grape, although winemaking does alter the constitution. Tea is another potent phenolic source, but green and black tea have quite different compositions. Green tea mainly contains the monomeric catechins while black tea has the oligomeric and polymeric forms. To date, studies of tea's effects on CHD mortality have not distinguished between green and black tea,(26) but east Asian countries where green tea consumption is high are noted for their low CHD mortality rates.(3)
While some of the phenolic compounds are common to many plants, there are many differences as well. These differences have been exploited in the authentication of juices and wines.
For instance, Vitis vinifera, the wine grape, contains almost exclusively the monoglucosides of the anthocyanins while American grape species and crosses contain the diglucosides.
Compared with other dietary sources, wine contains relatively high levels of phenolics. It is also a diverse source, as it includes significant amounts of all the major classes of phenolics. While whole fruits are rich sources, the data on juices show low levels. Apparently, normal aerobic processing degrades these compounds.(27) Wine production, on the other hand, is largely anaerobic and thus the phenolic compounds are retained. The ethanol produced during fermentation is an effective solvent to extract the substantial amounts of flavonoid phenolics in both the skins and the seeds present in red grapes. In white wine production the skins and seeds are separated from the juice immediately after crushing the grapes, and thus flavonoid levels are much lower. However, the phenolics present solely in the white juice, mostly the non-flavonoids, are retained in anaerobic wine production.
The combination of epidemiological, clinical and mechanistic studies strongly suggests that wine phenolics are beneficial nutrients that can reduce CHD mortality. In order to make sound dietary recommendations for phenolic compounds and the foods that contain them, such as wine, much work remains to be done. The availability, absorption, metabolism and physiological effects of well-described chemical systems need to be thoroughly investigated. The striking differences between two flavonoids in absorption and effects on eicosanoid synthesis are a reminder that these compounds do not represent a simple class of substances with similar behaviours. A detailed understanding based on sound chemistry will not only clear up a current paradox, but may clarify the effects of other phenolic-rich foods, notably the anti-cancer activity of tea.
If future studies bear out these suspicions, chemists will be able to toast 'to your health' with enthusiasm and understanding!
1 Goldberg, D.M., Clin. Chem.,1995,41,14-6
2 St Leger, A.S.,Cochrane, A.L.,& Moore, F, Lancet, 1979, i, 1017-20
3 Renaud, S.,& deLorgeril, M., ibid, 1992,339,1523-6
4 Criqui, M.H., & Ringel, B.L., ibid. 1994,34, 1719-23
5 Marmot, M.G., Roso. G.,& Shipley, MJ., ibid, 1981,i,580
6 Rimm, E.,et al, ibid, 1991, 338,464-8
7 Hands. K., et al, Am. J. Card., 1990,65,287-9
8 Klatsky, A.L.,& Armstrong. MA.,ibid. 1993,71,467-9
9 Hertog, M.G., et al, Lancet, 1993,342,1007-11
10 Hertog, M .G., et al, J. Agric. Food Chem., 1992, 40, 2379-83
11 Gugler, R., et al, Eur. J. Clin. Pharmacol.,1975, 9, 229-34
12 Stemann, E.H., & Creasy, L-L., Am. J. Enol. Vitic., 1992, 43, 49-52
13 Arichi, H., et at, Chem. Pharm. Bull., 1982, 30, 1766-70
14 Frankel, E.N. el at, Lancet, 1993, 341, 1103-4
15 Pace-Asciak, C.R., et al, Clin. Chimica Acta, 1995, in press
16 Lamuela-Raventos, R.M.,etal, J. Agric. Food Chem., 1995,42,281-3
17 Esterbauer, H.,et al, Free Radical Commun.,1992,13,341-90
18 Leake, D.,Biochemist, 1995,17, No.l,12-15
19 Frankel, E.N. el at, Lancet, 1993, 341, 454-7
20 Maxwell, S., Cruickshank, A., & Thorpe, G., ibid, 1994, 344, 193-4
21 Seigneur, M.,et al, J. Appl. Card.,1990,5,215-22
22 Fuhrman, B., el al, Am. J. Clin. Nutr., 1995,61,549-54
23 Mower, R.,et al. Biochem. Pharmacol., 1994,36,317-22
24 Demrow, H.S.,Slane, P.R.,& Folts, J.D.,Circulation, 1995, in press
25 Hackett, A.M., in: 'Plant flavonoids in biology and medicine: biochemical pharmacological and structure activity relationships', (Eds. V. Cody, E.J. Middleton & J.B. Harborne), New York: Liss, 1986, 177-94
26 Klatsky, A.L.,et al, Ann. Epidemiol. 1993,3,375-81
27 Spanos.G.A.,& Wrolstad, R.E.,J. Agric. Food Chem, 1990,38, 1565-71
This article has been reproduced for the World Wide Web homepage of the Department of Viticulture & Enology with permission from Chemistry & Industry, 1 May 1995, pages 338-341.
For a copy of the article, and a free sample issue of Chemistry & Industry, please send your request and postal address to:
Editor, Chemistry & Industry
15 Belgrave Square
London SW1X 8PS