Should Red Wine Be Considered a Functional Food?



Abstract:  Functional foods may be regarded as foods that have nutritional value, but in particular, they also have beneficial effects on one or more body functions. Thus, functional foods may improve health and/or reduce the risk of developing certain diseases when taken in amounts that can be consumed in a normal diet. Based on nearly 2 decades of research since the term “French paradox” was first coined in 1992, wine would appear to fit this definition. Yet there seems to be reluctance to consider wine as a functional food. In this review, we present an overview of the accumulated evidence for the health benefits of wine—and its key phenolic components such as resveratrol, quercetin, catechin—and show that these alone are not enough to firmly establish wine as a functional food. What is required is to create clearly defined products based on wine that are targeted to consumers’ needs and expectations when it comes to purchasing functional foods. Moreover, the crucial question of alcohol and health also needs to be addressed by the functional food industry. Suggestions are presented for working through this issue, but in many regards, wine is like any other food—it should be consumed sensibly and in amounts that are beneficial to health. Overindulgence of any kind does not promote good health.


Overview of functional foods

Functional foods have a long history (Weststrate and others 2002). Traditionally, many foods were regarded as a form of medicine in ancient China, Korea, and Japan. Chinese people developed a food culture which emphasized the preservation of health by diet and was based on the theory of Chinese medicine (Li and others 2004). Yet the “food as medicine” philosophy proposed by Hippocrates has been largely ignored by the general population in the West for many years (Smith 2004). Recently, however, interest in the functional food philosophy has been rekindled. Many definitions exist for the term functional foods (Subirade 2007); however, a simple definition is “those foods which are intended to be consumed as part of the normal diet and that contain biologically active components which offer the potential of enhanced health or reduced risk of disease” (Subirade 2007). Under this definition “unmodified whole foods such as fruits and vegetables represent the simplest form of a functional food” (Subirade 2007). An alternate definition emphasizes functionality beyond basic nutritional requirements: “a functional food is similar in appearance to, or may be, a conventional food that is consumed as part of a usual diet, and is demonstrated to have physiological benefits and/or reduce the risk of chronic disease beyond basic nutritional functions, i.e. they contain bioactive compound” (Agriculture and Agri-Food Canada 2009).

The current functional food concept was started in Japan in the late 1980s. The Japanese health authorities were interested in controlling health care costs and providing good quality of life for the elderly and recognized the benefits that a functional food philosophy could provide. In essence, the current functional food concept was developed to promote health and reduce the risk of disease (Subirade 2007). The European Union (EU) adopted the Japanese concept of a functional food and designated the European Food Safety Authority (EFSA) to evaluate and approve all proposed functional foods before they are launched as such in the EU market. Approval depends on the applicant's ability to demonstrate the safety and efficacy of each proposed product based on appropriate scientific research (Barreiro-Hurle and others 2008).

As studies have started to demonstrate that functional foods may be beneficial for certain age-related diseases, some countries have even begun to competitively support functional food developments without going through submission of applications and the review process (Hasler 1998). Moreover, some countries are proactively supporting the functional food industry because, as the global market for functional food is growing, it is creating opportunities for these countries to export niche products. In addition, as more countries support the functional food industry, the market has started to become globalized and is no longer controlled by countries with more established food industries such as in Western Europe and North America (Annadana 2008). In the long term this may have the effect of weakening standards, albeit with the benefit of stronger supply conditions.

The EU established the European Commission's Concerted Action on Functional Food Science in Europe (FUFOSE), to support functional food developments (Diplock and others 1999). Australia also set up The Natl. Center of Excellence in Functional Foods in 2003 to support the Australian food industry in developing a functional foods market by integrating knowledge from a range of sources. The reason behind this new development is the concept that “health is the future of food and all foods are fast becoming functional” (Droulez 2005). In addition, consumers are increasingly taking a more proactive role in optimizing their personal health and well-being by taking in a wholesome diet, exercising more regularly, and relying less on pharmaceutical drugs (Annadana 2008). If this trend continues, it will not only reduce health care costs but also improve economic opportunities for the functional food industries (Kotilainen and others 2006). Therefore, the functional food market has become one of the fastest-growing markets in the world (Annadana 2008).

The estimated size of the global functional food market in 2006 was between US$30 and $60 billion, depending on the definition used for functional food. Since then the market has been growing by approximately 10% annually (Kotilainen and others 2006). This growth rate is significantly higher in comparison with the overall food and beverage markets’ growth which has been about 2% per year (Weststrate and others 2002). The value of the Japanese functional food market and the U.S. functional food and nutraceutical market were US$27.1 and US$21.3 billion, respectively, in 2006. By comparison, the value of the EU's functional food and nutraceutical market was only US$8 billion (Basu and others 2007). The Australian and New Zealand nutrition industry represented approximately A$2.9 billion of the global nutrition market. Of that, functional food comprised only A$1 billion (35.1%) (Lethbridge and others 2006). The estimated forecast of the global functional food market in 2013 is at least US$90.5 billion (Annadana 2008).

Functional foods tend to sell at higher prices and with a greater profit margin than do conventional foods. The retail price of functional foods is 30% to 500% more than comparable conventional foods (Kotilainen and others 2006). The higher price of functional foods may be partially due to the high cost of scientific research and development, and the high cost of taking the necessary steps to gain approval from the relevant government authorities. Although in the past functional foods appealed only to niche market segments, this is changing rapidly as we see growth in demand. The economies of scale may one day see functional foods sold at similar or identical cost to conventional products, at least in some product categories.

Wine and health

Historically, there is evidence that wine has been used as a medicine from ancient times in countries like China and India. In Mesopotamian culture, use of wine as medicine was illustrated by Sumerian Pharmacopoeia inscribed on a clay tablet at Nippur in cuneiform script (Norrie 2005). During this time, medicinal properties of wine were used as an antiseptic for wounds, tranquillizer, hypnotic, anesthetic, anti-nauseant, appetite stimulant, diuretic, cooling agent, and in poultices (Norrie 2005).

The current wave of interest in the benefits of “moderate wine drinking” on health started after the observation of Renaud and de Lorgeril (1992) who reported that there was a low mortality rate from ischemic heart disease among French people despite their high consumption of saturated fats and the prevalence of other risk factors, such as smoking. This was thought to be related to the “Mediterranean diet,” which included a regular, but moderate, intake of red wine (Renaud and de Lorgeril 1992). Their paper received much attention by the press and the term “French paradox” became common. Some 5 y later, a Dutch epidemiological study showed that in elderly men coronary artery disease is inversely proportional to their flavonoid intake (Hertog and others 1997). Since then there have been several studies which have looked at the health benefits of wine. The purpose of this review is not to prove that moderate consumption of wine is healthy. Rather, given the well-documented health beneficial properties of wine (especially red wine) and its phenolic components, our purpose is to examine whether wine may be considered a good functional food.

Red wine polyphenols act as antioxidants

Wine contains a large and diverse class of phenolic compounds known variously as “phenolics,”“polyphenols,” or “biophenols.” Polyphenol classes, subclasses, and some prominent food flavonoids are listed in Figure 1. These compounds contribute to the characteristic colors and flavors of wine, and also act as natural wine preservatives that allows a long aging process (Waterhouse 2002). Polyphenols are extracted during crushing and fermentation when the juice is in contact with the grape skins and seeds. Other factors that influence the accumulation of phenolic compounds include variety, climate, viticultural practices (Stockley and Hoj 2005), soil type (De Andres-De Prado and others 2007), vineyard location, vine water status (Koundouras and others 2006), sulfur dioxide addition (Tao and others 2007), and oak aging (Gougeon and others 2009). The amount of phenolic compounds in red wine is about 6 times higher than that in white wine because red juice has longer contact time with the grape skins and seeds. The concentration of polyphenolic compounds in red wine is approximately 1800 to 3000 mg/L (Goldberg and others 1996b). These polyphenols in wine also have beneficial effects on health, purportedly because of their antioxidant properties (Valdez and others 2004; Logan and others 2008).

Figure 1–.

Classes and subclasses of phenolic compounds. Adapted from Erdman and others (2007).

Chemical compounds which have antioxidant properties are good for health because they protect cells from oxidative stress. For example, there is some evidence that certain age-related diseases occur because of oxidation of cell components by free radicals. Antioxidants protect the body by scavenging these free radicals (Zhang and others 2006). A free radical is a highly reactive and damaging species that contains an unpaired electron. The free radicals attack a neighboring molecule to gain an electron to become stable, however this process generates another free radical. This chain reaction is thought to contribute to lipid peroxidation, DNA damage, and protein degradation during oxidative-stress events (Clarkson and Thompson 2000). It is believed that this oxidation contributes to numerous diseases, such as Alzheimer's disease, rheumatoid arthritis, amyotrophic lateral sclerosis, cataractogenesis, various cardiovascular diseases, and some cancers (Stadtman 2004; Logan and others 2008). Potential diseases that can be caused by the damaging effects of free radicals are shown in Figure 2.

Figure 2–.

Overview of human diseases where excessive free radical production is thought to play a significant role. Adapted from Halliwell (1987).

Human cells are constantly attacked and damaged by free radicals and this accumulated damage over time results in early signs of aging or the onset of certain diseases. Therefore, various diseases can be delayed or even halted if cells can be protected from these free radicals. To defend itself against free radicals, the body produces endogenous antioxidants, but if radical damage is severe, there may not be sufficient quantities of antioxidants to neutralize free radicals completely. Therefore, use of additional exogenous antioxidants, which are present in certain diets, may help the body to fight against these free radicals. Some of these compounds are well known, for example, vitamins C and E and beta-carotene (precursor of vitamin A). Use of these compounds can help prevent cell damage and may also help repair, to some extent, damaged cells by scavenging free radicals. Polyphenols may act in a similar way in the body, but their role in human health has not been established to the same extent as for vitamins A, C, and E. Nevertheless, consumption of about 250 mL/d of red wine for 2 mo has been shown to significantly increase antioxidant status and decreases oxidative stress in circulation (Guarda and others 2005).

While historically the focus of research has been on the antioxidant activity of phenolic compounds, possible health benefits may arise from other bioactivities, for example, protein binding, DNA repair, enzyme modulation, and other “including the ability of biophenols to bind to cellular receptors and transporters thereby influencing gene expression (Moon and others 2003; Noe and others 2004; Zhang and others 2005), cell signaling (Rosenkranz and others 2002; Lee and Lee 2006; Aggarwal and Shishodia 2006), and adhesion” (Tucker and Robards 2008).

Cardioprotective Effects of Red Wine

In the United States, arteriosclerosis has been reported to account for almost 40% of all mortality. The process of atherogenesis (the accumulation of fatty substances in arteries known as plaques) has been thought to be accelerated by oxidation of low-density lipoprotein (LDL), popularly known as “bad cholesterol.” Oxidation of LDL polyunsaturated lipid components occurs with reactive free radicals and enzyme systems, such as 15-lipoxygenase, cytochrome p450, and myeloperoxidase; and red wine polyphenols can reduce LDL sensitivity to lipid peroxidation (Fuhrman and Aviram 2001). Red wine polyphenols are absorbed efficiently in humans and bind to LDL, thus protecting it from oxidation. Resveratrol also reduces the synthesis of lipids and eicosanoids (signaling molecules), which promote inflammation and atherosclerosis.

There is now almost universal acceptance among the scientific community that alcohol, when consumed in moderation, is associated with a lower incidence of cardiovascular disease and generally better health outcomes (Lang and Melzer 2009; Lee and others 2009b). Some controversy still exists over whether red wine has superior protective effects than other alcoholic beverages (Cordova and others 2005). Part of this may be explained by the different studies that have been conducted. That is, epidemiological studies reveal associations between various factors, but cannot be used to ascertain causality. They are useful to provide longer-term (years to decades) information on risk factors for disease and factors that may have a positive effect on health.

Human intervention studies can be used to establish causality between the type of beverage consumption and some measures of health and disease, but over a much shorter time frame—weeks to months (rarely years). However, over this shorter time frame the link between consumption and longer-term chronic disease cannot be established. In vitro and animal trials can be used to establish links between specific compounds and biomarkers of disease and are useful for understanding bioactivity at a molecular level. All these types of investigations have been used to learn about red wine. The results from many studies point to red wine having significant bioactivity, affecting short term biomarkers of health and disease, and being associated with better health outcomes over the longer term. Furthermore, from a chemical perspective, wine is unique among alcoholic beverages in that it does contain relatively high concentrations of phenolic compounds, which have been linked to cardioprotective effects (Lagrue-Lak-Hal and Andriantsitohaina 2006). Recent results from all these spheres will be presented here. The reader is directed to an extensive review by Cooper and others (2004) for results prior to 2004. The vast literature on the health benefits of wine is illustrated by a search of the ISI database using terms “wine and (health or disease),” which returned 2985 references on 12/3/10.

Epidemiological studies

Numerous epidemiological studies have demonstrated an association between moderate alcohol consumption and reduced risk of cardiovascular disease. For example, Thun and others (1997) examined the effect of alcohol consumption on mortality among U.S. adults. Of 490000 men and women who reported their alcohol and tobacco use, 46000 died during 9 y of follow-up. The death rates from cardiovascular disease were 30% to 40% lower among men and women reporting at least 1 drink (assumed average 12 g of alcohol) daily compared with nondrinkers. The overall death rates were lowest among men and women reporting approximately 1 drink daily (Thun and others 1997). Criqui and Ringel (1994) reached comparable conclusions, but further noted that wine was more strongly correlated with lower rates of coronary heart disease (CHD) than either beer or spirits.

More recently, Schroder and others (2007) analyzed the association of alcohol consumption with nonfatal myocardial infarction (MI). They performed a population-based case—control study consisting of 244 cardiac patients and 1270 control subjects. Male patients aged 25 to 74 y with past myocardial infarction who were matched with a group of healthy control subjects who were randomly taken from a representative sample of the population. After adjusting for lifestyle and cardiovascular risk factors, total alcohol consumption (up to 30 g/d) was inversely associated (odds ratio, 0.14; 95% confidence interval (CI), range 0.06–0.36), with the risk of nonfatal MI.

With many nations experiencing an aging population, attention has turned to disease prevention among the middle-aged to elderly. In this context, King and others (2008) investigated whether people who begin drinking alcohol in their middle years experience any health benefits from this change in lifestyle. The long-term, prospective study, Atherosclerosis Risk in Communities, involving individuals aged 45 to 64 y, has revealed that “Those who begin drinking moderately experience a relatively prompt benefit of lower rates of cardiovascular disease morbidity with no change in mortality rates after 4 years.”

Red wine can increase high-density lipoproteins (HDL), popularly known as “good cholesterol.” HDL is required for transport of cholesterol from the arteries and various other parts of the body back to the liver for metabolism and/or excretion. Hence, HDL protects against atherosclerosis. Several studies have shown that alcohol consumption increases HDL and, since wine contains 10% to15% alcohol, a part of its cardio-protective effects may be related to its alcohol content (Cordova and others 2005). In one study undertaken in France and Northern Ireland, it was shown that there was a significant increase in HDL cholesterol, from 0.47 to 0.59 g/L, in men who consumed 2.3 oz (65.2 g) per day of alcohol as compared to nondrinkers. These men consumed mostly red wine (Marques-Vidal and others 1995). Similar results have been reported from the Natl. Heart, Lung, and Blood Insti. Family Heart Study (Ellison and others 2004). The researchers concluded that among lifestyle behaviors, alcohol consumption is an important correlate of HDL-cholesterol.

Human intervention studies

Intervention with wine: With promising results from epidemiological studies, human intervention studies have further revealed the health benefits of wine in practical, applicable, and feasible doses. A review by Cooper and others (2004) lists 19 human intervention studies where wine has been shown to be beneficial in the reduction of risk factors associated with CHD. These include plasma lipid concentrations, lipid and serum oxidation, and plasma antioxidant capacity. More recently, Sacanella and others (2007) conducted a randomized cross-over study of 4 wk duration involving 35 women. They found that “moderate wine consumption is associated with beneficial effects on various inflammatory pathways related to endothelial activation” including decrease of interleukin-6, vascular CAM-1, and E-selectin; and enhanced adhesion of monocytes to stimulated endothelial cells was reduced by 51% after white wine consumption, and by 89% after red wine consumption.

The commonly used cardio-protective drug acetylsalicylic acid (aspirin) acts by attenuating platelet aggregation, thereby reducing blood clots. Polyphenols, especially resveratrol and quercetin, also inhibit platelet aggregation (Pace-Asciak and others 1995). A number of other studies have also shown reduced platelet aggregation upon consumption of red wine and these have been reviewed by Nardini and others (2007).

Some work is now beginning to emerge as to the possible interaction of wine with other dietary components that may be beneficial to human health. In a randomized trial involving supplementation with alpha-linoleic acid, de Lorgeril and others (2008) found that moderate alcohol consumption was associated with higher marine omega-3 lipids in the body. The authors concluded that while the “data have to be confirmed in large groups, this effect of wine comparable to that of fish may partly explain the protective effects of wine drinking against CHD.” Olive oil is another dietary component linked to the beneficial effects of the Mediterranean diet. Consumption of red wine and “green” olive oil was found to improve endothelial function in a synergistic manner (Karatzi and others 2008). Positive results were only obtained with one combination of wine and olive oil (red wine and green olive oil)—combinations involving white wine or refined oil (low in phenolic compounds) were not effective.

Alcohol in wine plays an important role in the reported health benefits (Kloner and Rezkalla 2007). Rimm and others (1999) conducted a meta-analysis of 42 human intervention studies involving alcohol consumption (up to 100 g/d) for a period of between 1 and 9 wk. Their analysis estimated that 30 g alcohol a day would reduce the risk of coronary heart disease by 24.7%. Since then, intervention studies consistently show that moderate alcohol consumption results in improvements in biomarkers for cardiovascular disease. For example, Brand-Miller and others (2007) investigated the effect of alcohol on postprandial glycemia and found that this marker of cardiac risk was lowered by 16% to 37% in a “realistic” meal setting. The researchers concluded that this might “provide a hitherto unrecognized benefit of moderate alcohol consumption for cardiovascular health.”

However, alcohol is not totally benign, especially at higher doses (Gronbaek 2009). It has been reported to elevate heart rate and blood pressure 8 to 10 h after drinking red wine (375 mL; 39 grams of alcohol) and beer (1125 mL; 41 grams alcohol) (Zilkens and others 2005; Mukamal and Rimm 2008). A consumption of high doses of alcohol has also been linked to the cause of some cancers (Hamid and others 2009; Mckillop and Schrum 2009; Petti 2009). Lachenmeier and others (2009) reported that acetaldehyde may be a contributing factor in the carcinogenicity of alcohol. However, as acetaldehyde is found in an enormous range of foods, including fresh fruits and vegetables and practically every food subjected to fermentation, this report awaits further scrutiny as to the true risks associated with acetaldehyde consumption. Despite the possible links between alcohol and cancer, the World Cancer Research Fund/American Insti. for Cancer Research (WCRF/AICR) recommends that if alcohol is to be consumed, the limit should be “no more than two drinks a day for men and one drink a day for women” (WCRF/AICR 2007). The controversial issue of ‘alcohol and risks and benefits’ will be discussed more in 2 sections below: Toxicology in the functional food market and Alcohol—the most controversial issue in wine and health research. The definition of what constitutes a “drink” of alcohol will also be explored.

Intervention with grape extracts and polyphenol compounds: There has been considerable interest in the polyphenol fraction of grapes and wine and ascertaining whether potential health benefits can be derived by these compounds without the consumption of alcohol. The results of human intervention studies have yielded mixed results (Perez-Jimenez and Saura-Calixto 2008). For example, an early report by Nigdikar and others (1998) showed an increased the lag time for oxidation of LDL isolated from healthy individuals fed an extract of red wine polyphenols. The increase in lag time was slightly more for consumption of red wine, yet there was no increase in lag time after consumption of white wine or alcohol. An increase in the lag time for LDL oxidation suggests the LDL is protected from attack by oxidants and this would be an indicator of a protective effect of the red wine polyphenols. Mixed results were obtained by Zern and others (2005) in a trial with 24 premenopausal and 20 postmenopausal women. They reported that grape polyphenols (5% to 8% in lyophilized grape powder) reduced plasma triglyceride concentration, plasma LDL cholesterol, and apolipoproteins B and E as well as cholesterol ester transfer protein activity; however, whole-body oxidative stress was increased. A recent study by Garcia-Alonso and others (2009) has looked at the anthocyanin components of red wine and found that monocyte chemoattractant protein 1 levels in plasma were reduced, while plasma antioxidant activity was enhanced in vivo. Both trends are consistent with reduced risk of cardiovascular disease.

In view of the uncertainty over the effects of alcohol alone versus the effects of alcohol in the presence of polyphenol compounds, several researchers have called for “randomized controlled trial(s) to assess the effects of alcohol consumption on clinical endpoints” (Lang and Melzer 2009). Until such trials are conducted, other researchers are pursuing animal models in an attempt to better understand the effects of wine and alcohol on specific biochemical and physiological processes.

Animal studies

Animal models allow more invasive experimentation than is possible with humans, for example, extracting target organs following a feeding trial or feeding higher doses than would be safe for human consumption. This is exemplified by studies on atheromatous plaque development, which are not possible in humans (Vinson and others 2001).

Pérez-Jiménez and Saura-Calixto (2008) reviewed 34 animal studies involving various wine and grape products and extracts. As with many studies in this field, the results are not unequivocal. For example, all animal studies reviewed showed that grape phenolic compounds were able to lower blood pressure. However, not all reviewed studies reported a cholesterol-lowering effect from grape-derived polyphenols. Other indicators of cardiovascular disease (CVD) reviewed included: endothelial function; platelet aggregation; atherosclerotic plaque development; markers of inflammation; oxidative stress; and glycemia. Generally wine and/or grape-derived phenols improve these markers in animal models.

Some insights into the mechanism of the action of red wine in improving risk factors for CVD have been provided by Lefevre and others (2007). Feeding apoE-deficient mice an antioxidant- rich Cabernet Sauvignon wine, there was improvement in a number of measures of recovery from ischemia including increased blood flow, decreased oxidative stress, and increased capillary density. Moreover, endothelial progenitor cells increased after red wine consumption as did the migratory capacity of these cells.

Notwithstanding the previous study, recent research in animal models has tended to focus more on polyphenolic grape or wine extracts rather than on whole wine in feeding trials. It is not clear why this trend has occurred, but it may be that researchers are trying to understand more fully the role of the phenolic compounds in isolation of alcohol, due to the potential for alcohol to confound results. Thus Noll and others (2009) investigated the effects of red wine polyphenols on hyperhomocysteinemia in mice. Supplementation with a phenolic extract in the drinking water of mice, over a 4-wk period, led to a reduction in plasma homocysteine levels and restored hepatic and plasma-decreased paraoxonase-1 activity. Together, these results are consistent with the extract having beneficial activity against endothelial dysfunction due to hyperhomocysteinemia.

Of all the phenolic compounds in wine, resveratrol has received most attention in animal supplementation studies (Das and Maulik 2006; Baur and Sinclair 2006; Guerrero and others 2009). A number of cardioprotective mechanisms have been elucidated for this compound. For example, a low dose of resveratrol induced similar anti-aging effects as a calorie-restricted diet including protection against age-related cardiac dysfunction.

In the past few years, several studies have shown that resveratrol given before ischemic arrest prevents the heart from ischemia and/or reperfusion injury. Such protection has been shown to be nitric oxide (NO)-dependent (Hattori and others 2002; Das and others 2005). In particular, resveratrol treatment results in a coordinated upregulation of iNOS-VEGF-KDR-eNOS.

A variety of roles for resveratrol as an anti-atherogenic agent were reported by Do and others (2008). These included modulating lipid profiles through lowering total cholesterol and LDL-cholesterol and raising levels of HDL-cholesterol; attenuating atherosclerotic lesions and periarterial fat deposition; and diminishing intracellular adhesion molecule-1 and vascular cell adhesion molecule-1 in atherosclerotic vessels. These studies were conducted in apoE-deficient mice fed on a normal diet with supplements of 0.02% and 0.06% (w/w) resveratrol. A study also showed that resveratrol reduces oxidative DNA damage and glycoxidative stress in vivo animal study (Mizutani and others 2003).

Apart from resveratrol, other phenolic compounds that have received attention for potential cardioprotective properties include catechin and quercetin. Auger and others (2005) found that catechin and quercetin (along with resveratrol) protected against aortic fatty streak accumulation in hypercholesterolemic golden Syrian hamsters. Mixtures of these phenolic compounds are also cardioprotective. Norata and others (2007) found that P183/1 (a mixture of catechin, resveratrol, and caffeic acid) affected inflammatory cell recruitment and expression of pro-inflammatory chemokines in vascular walls of apoE KO mice, thus significantly decreasing atherosclerosis.

Catechin (and derivatives) on its own has also been reported to demonstrate protective effects against CVD. These include a protective effect against tamoxifen-induced oxidative damage (Tabassum and others 2007); increased alpha-tocopherol concentration in blood plasma, liver, and lungs (Frank and others 2003); reduced plasma lipid peroxides resulting in protection against atherosclerosis (Miura and others 2001); prevention of endothelial dysfunction in Otsuka Long-Evans Tokushima fatty (OLETF) rats at the prediabetic stage (Ihm and others 2009); and modulation of cytokine expression and thus prevention of low-grade inflammation (Terra and others 2009).

In vitro studies

In vitro studies involving red wine, extracts, and specific phenolic compounds and their bioactivities encompass a very large number of studies ranging from chemically simple assays such as free-radical scavenging activity with, for example, 1,1-diphenyl-2-picrylhydrazyl (DPPH), to technically challenging cell-culture systems. In vitro studies allow more controlled conditions of specific biochemical events and are useful for detailed mechanistic studies. Alternatively, in vitro testing can be used to more rapidly identify bioactive compounds and extracts in screening studies. This has been recently illustrated by Appeldoorn and others (2009a) who showed a differential activity of phenolic compounds in their ability to increase nitric oxide (NO) levels in endothelial cells, thus improving function through relaxation of vessels. Two wine phenolic compounds—resveratrol and quercetin—were the most effective in enhancing cellular NO.

Apart from a possible role in NO production, wine polyphenols, and in particular resveratrol, have been associated with inhibition of the inflammation response. Thus red grape juice polyphenols have been shown (Davalos and others 2009) to reduce superoxide anion production in the NADPH oxidase system. Inhibition of inflammation was also demonstrated in arterial endothelial cells where sub-micromolar concentrations of resveratrol inhibited TNF-alpha-induced signal transduction through attenuation of a range of inflammation markers including IL-6, ICAM-1, and VCAM (Csiszar and others 2006). Similar anti-inflammatory properties for catechin (and catechin derivatives) and quercetin have been found in vitro at concentrations of between 10 and 100 μM (Choi and Hwang 2003; Huang and others 2006).

In vitro studies of wine phenolic compounds have revealed a possible role in inhibiting platelet aggregation. A recent review by Natella and others (2006) concluded that this activity may take place through a variety of different mechanisms and that “all polyphenols show some anti-aggregating activity.” The authors also note the wide range of inhibitory power among the various classes of wine phenols, while drawing attention to the difficulties in transferring results in vitro to in vivo, possibly due to synergistic interactions among compounds that promote biological activity (Nardini and others 2007).

Of all the in vitro tests available to investigate the bioactivity of wine polyphenols, by far the most attention has centred on antioxidant activity. Many such assays are available (Antolovich and others 2002) and it is not uncommon for different tests to assign different relative antioxidant ability to the same phenolic compound (McDonald and others 2001). Nevertheless, resveratrol consistently is rated as a relatively weak chemical antioxidant, despite the huge amount of attention given to resveratrol in health research. In wine, the polyphenols catechin, epicatechin, and quercetin are generally found to be the better antioxidants (Soleas and others 1997; Rifici and others 2002). The ISI database holds more than 1,700 records matching the search terms “wine” and “antiox*” (not counting in vivo studies) including 133 reviews. It is beyond the scope of this review to cover this vast body of literature. As such, recent and/or novel findings will be briefly discussed.

There is ongoing interest in testing the antioxidant activity of wines and their constituent phenolic compounds, from various countries and regions, using assays such as the DPPH assay. Thus de Quiros and others (2009) tested a number of Spanish white wines and correlated DPPH radical scavenging activity against phenolic composition for 11 polyphenols. Similar studies in 2009 have investigated wines from Serbia (Radovanovic and others 2009), Portugal (Feliciano and others 2009), Turkey (Anli and Vural 2009), Croatia (Maletie and others 2009), China (Li and others 2009), and Bulgaria and Greece (Gougoulias 2009). The effect of different oak types (Goncalves and Jordao 2009), variety of grapes (Hogan and others 2009), and organic production (Sacchi and others 2005) have also been explored in 2009 using chemical antioxidant assays.

Criticism of simple chemical antioxidant assays has grown in recent years, despite their continued widespread use (Frankel and Meyer 2000; McDonald and others 2001; Scalbert and others 2005). One of the criticisms is that biophenols are extensively metabolized in vivo and the metabolites may have different bioactivity. In this context, the report by Yu and others (2006) is relevant in that O-methylated flavanols were tested for their ability to alter the kinetics of oxidation of LDL in vitro. Differential effects were found depending on the number and position of the methylations and the study highlighted the importance of considering all metabolized forms of the parent polyphenols when ascertaining in vivo bioactivity.

Other Health Benefits

While the initial interest in red wine and health was directed towards cardiovascular disease, due to the French paradox, increasingly wine has been linked with other positive health outcomes. These include anti-mutagenic, anti-estrogenic, and anti-carcinogenic effects (Ferguson 2001). As with cardiovascular protection, evidence for these effects comes from a variety of sources. The possible exception is human intervention studies where it is difficult, over short time frames, to detect the action of wine, grape extracts, or polyphenols in diseases such as cancer. A number of bioactivities already discussed for cardiovascular protection are relevant for the other reported health benefits of wine. These include antioxidant activity, or more particularly scavenging of reactive oxygen species; and anti-inflammatory activity, both of which may help to protect against cancer development and progression.


Similar to the discussions previously mentioned, the effects of wine on health, as determined by epidemiological studies, are sometimes difficult to ascertain due to the potential for the study to look at alcohol and not distinguish between the various forms (Cummings and others 2009). Even where different types of alcoholic beverages are considered separately, the overall result may be presented as pertaining to “alcohol”. For example, Benedetti and others (2009) concluded that “moderate and high alcohol intake levels over the lifetime might increase cancer risk at several [cancer] sites.” This generalized statement does not encompass their finding that the main drivers of increased risk were beer and spirits.

On the other hand, a recent report has found a potential protective effect of alcohol on thyroid cancer (Meinhold and others 2009) and a specific protective effect for modest (< 50 to 90 g/wk) wine consumption was found for esophageal cancer (Pandeya and others 2009). Two recent studies have come to seemingly disparate conclusions on the effect of alcohol on pancreatic cancer with Rohrmann and others (2009) reporting “no association of alcohol consumption with the risk of pancreatic cancer” while Genkinger and others (2009) reported “a modest increase in risk of pancreatic cancer with consumption of 30 or more grams of alcohol per day.”

One study that has examined moderate intake of alcohol and cancer incidence in over one million middle-aged women in the U.K. has suggested that wine may not be different to other alcoholic beverages (Allen and others 2009). Increasing alcohol consumption was associated with increased incidence of cancers of the oral cavity and pharynx, esophagus, larynx, rectum, liver and breast; an overall increase of 15 cancers per 1000 women up to the age of 75 was reported. The low increase (0.015%) is an important discovery, though does need to be interpreted in context. The study made adjustments for only 5 of the known determinants of cancer; while some alternate causes are likely to me minor, the omission of family history and diet as contributing variables may account for the low total increase in cancer rates. The risk of thyroid cancer, non-Hodgkin lymphoma and renal cell carcinoma decreased with increasing alcohol consumption, further highlighting the need to account for family history and perhaps an in depth consideration of individual risk when using the results to provide public health advice.

The study (Allen and others 2009) examined the role of wine compared with other beverages and reported no significant differences. This finding has received much media attention, although the finding has several important limitations for further research to address before a clear claim about cancer and wine can be made. Firstly, exclusive wine drinkers were compared to drinkers of beer, spirits or a mixture of all beverage types, including wine. If wine does have a protective effect, both exclusive wine drinkers and consumers of all beverage types experienced at least some of the protective effect; this could explain why no statistical difference was reported. Secondly, while no statistical difference was found between the groups, only a weak statistical similarity (P = 0.3) was found, such that a clear statement about wine compared with other beverages is premature. More compelling was the finding that a strong similarity exists between white and red exclusive wine drinkers for breast cancer, a finding that challenges the protective effect of red wine over white for breast cancer. This study is an important contribution and highlights a possible link between moderate wine consumption (including red wine) and the risk of some cancers in women, though further research is required to establish the unique contribution of wine, compared with other beverage types, specifically, where there may be protective compared with damaging effects.

While the jury may yet be out on the benefits or otherwise of wine on cancer, wine/alcohol consumption has been found to be beneficially linked with other diseases. These include rheumatoid arthritis (Kallberg and others 2009), upper gastrointestinal bleeding (Lanas and others 2000), and blood serum uric acid levels (Choi and Curhan 2004) (elevated levels associated with gout and a range of other diseases).

A range of epidemiological studies have found associations between the intake of specific polyphenols and reduced risk of various diseases. Most of these studies have focused on the flavonoids, and while these compounds are abundant in many foods, they are also present in wine. Table 1 lists some of these studies.

Table 1–.  Selected epidemiological studies on polyphenols and disease.
StudyNr of participantsCompoundDiseaseReference
Prospective study – Finnish Mobile Clinic Health Examination Survey10054quercetin kaempferol myricetinlung cancer asthma type 2 diabetes cerebrovascular disease type 2 diabetes prostate cancerKnekt and others (2002)
Case-control study in 6 Italian areas1225 colon cancer patients and 728 rectal cancer patientsisoflavones anthocyanidins flavones flavonolscolorectal cancerRossi and others (2006)
Prospective study – Multiethnic Cohort Study183518flavonols quercetin kaempferol myricetinpancreatic cancerNothlings and others (2007)
Case-control study in Los Angeles County558 lung cancer patientsepicatechin catechin quercetin kaempferollung cancerCui and others (2008)

A novel potential function of red wine has recently been reported (Gorelik and others 2008) that shows that red wine in the diet may inhibit the absorption of cytotoxic compounds that are associated with ingestion of lipid-containing foods. In a human intervention study, subjects were fed turkey cutlets, and then cytotoxic malondialdehyde (MDA) was measured in plasma and urine. When the cutlets were consumed with red wine there was zero MDA elevation in plasma. MDA is a lipid peroxidation product and the authors concluded that red wine may be effective in preventing such compounds being absorbed by the body thus protecting the body from possible harmful effects.

While not specifically about wine, Collie and Morley (2007) have reviewed the effects of polyphenols on cognitive function. They concluded that “there is evidence from human randomized, controlled trials that polyphenols and other antioxidant-rich substances can positively affect cognitive function,” while highlighting the need for further clinical trials. In the meantime, neuroprotective effects of polyphenols have been found in animal studies, as described subsequently.

Animal studies

Many animal studies support the findings of some epidemiological studies that red wine and/or its polyphenol content are beneficial to health beyond cardioprotective effects. Many of these studies have focused on anticancer properties and have been reviewed elsewhere (He and others 2008; Kraft and others 2009; Guerrero and others 2009). Briefly, animal models have shown that wine and/or polyphenol extracts may be beneficial in cancers such as breast (Hakimuddin and others 2008), colon (Dolara and others 2005), neurofibromatosis (Ebeler and others 2005), and liver (Kweon and others 2003). The antioxidant activity of wine extracts has been found to protect rat DNA from oxidative damage in vivo (Casalini and others 1999). This finding is consistent with anticancer effects of wine consumption. With regards to possible protection against breast cancer, Castillo-Pichardo and others (2009) found that a mixture of resveratrol, quercetin, and catechin was much more effective than the individual phenols in inhibiting mammary tumor growth and site-specific metastasis in a mouse model.

Animal models are helping to shed light on the link between wine consumption and reduced risk of Alzheimer's disease. Wang and others (2006) fed Cabernet Sauvignon to Tg2576 mice, which tend to develop a neuropathology of amyloid beta (A beta) protein deposits. The authors concluded that red wine exerts a “beneficial effect by promoting nonamyloidogenic processing of amyloid precursor protein, which ultimately prevents the generation of A beta peptides” supportive of epidemiological studies. Similar results were obtained in a more recent study by Wang and others (2009), but this time using grape seed extract and APP(Swe)/PS1dE9 transgenic mice. Reductions were found in brain and serum levels of A beta protein as well as in amyloid plaques and microgliosis.

Wine and polyphenol grape/wine extracts have also been shown, through animal models, to protect against oxidation in organs that are not part of the cardiovascular system. Such antioxidant activity is relevant to both anticancer and anti-Alzheimer's functions, but also plays a role in protection of organ function from oxidative stress. Thus red wine was able to protect liver function in Wistar rats over a 6-month study period (Assuncao and others 2009). The authors concluded that the protective effects of the polyphenolic fraction of red wine, may help to counter the pro-oxidative effects of the ethanol in the liver. Similar results have been found for the kidneys (Bertelli and others 2005). However in ochratoxin A-induced nephrotoxicity, both red wine and 13.5% ethanol were found to protect the kidney.

Animal models have been widely employed to investigate the potential health benefits (additional to cardioprotective effects) of specific wine phenolic compounds. Of the multitude of such compounds present in wine, by far the most attention has been given to resveratrol. Quercetin and catechin have received a reasonable amount of attention, and other compounds such as malvidin (Vanzo and others 2008) and myrecetin (Jung and others 2008) are starting to gain interest.

The prominence of resveratrol is highlighted in the titles of 2 reviews in 2009,“Wine, Resveratrol and Health: a Review” (Guerrero and others 2009) and “Fighting Cancer with Red Wine? Molecular Mechanisms of Resveratrol” (Kraft and others 2009). The list of diseases that have been investigated in animal models testing for resveratrol therapeutic or protective effects includes lung cancer (Lee and others 2006), colon cancer (Hope and others 2008; Sengottuvelan and others 2009), breast cancer (Whitsett and Lamortiniere 2006), gastric cancer (Woodall and others 2009), and prostate cancer (Harper and others 2007), and other cancers have also been tested with variable results (Guerrero and others 2009). Other activities, measured in vivo in animal models, have been reported: neuroprotection (Vingtdeux and others 2008; Bastianetto and others 2009), anti-aging activity (Baur and others 2006), anti-diabetic activity (Szkudelski 2008; Ramadori and others 2009), protection against pulmonary hypertension (Csiszar and others 2009), and protection against oxidative damage in liver (Kasdallah-Grissa and others 2007).

As mentioned, less attention has been given to quercetin or catechin, in the context of wine and health, than has been given to resveratrol. This is possibly due to the fact that these compounds occur in many other foods and beverages, whereas resveratrol has been almost uniquely associated with red wine consumption. Thus, quercetin is often associated with consumption of apples, onions, or tea, and catechin is mainly associated with consumption of tea (Scalbert and others 2005). Reports of animal models investigating potential health benefits of quercetin and catechin often lie outside of the wine literature and have been reviewed extensively (Middleton and others 2000; Scalbert and others 2005). In light of those reviews, selected examples of more recent in vivo animal testing of these compounds will be mentioned.

Quercetin and catechin were among 9 polyphenols added to different cereal fractions included in the diet of prematurely aging mice (Alvarez and others 2006). Mice fed the polyphenol-rich cereals showed better immune cell function and the authors concluded that “regular intake of these compounds could delay normal aging and improve quality of life.” Anticancer effects of quercetin continue to be demonstrated in animal models. Examples include the ability of quercetin (among several flavonoids) to inhibit melanoma lung metastasis by impairing endothelium interactions amongst tumor cells (Piantelli and others 2006); and protect against oxidative damage by UVA irradiation and thereby protect skin against premature aging or skin cancer (Kahraman and Inal 2002). Whilst neuro-protective effects of red wine extracts have been reported (Bastianetto and Quirion 2002), quercetin is not likely to play a significant role in this activity (Zbarsky and others 2005).

Similar activities have been reported for catechin (or catechin-like compounds such as Oligonol®, a mixture of catechin and oligomers of catechin units). Thus lung metastasis was inhibited by Oligonol fed to mice (Lee and others 2009a) and catechin on its own was effective in delaying tumor onset in a transgenic mouse model (Ebeler and others 2002). In contrast to quercetin, catechin has been found to be neuroprotective (Ruan and others 2009). It has been recently reported that the relatively subtle structural differences between these compounds (quercetin has a 4-oxo group and a double bond between C2 and C3) can result in very different in vivo bioactivities (Wiegand and others 2009).

In vitro

Consistent with the many reports from epidemiological studies linking wine with beneficial effects on various diseases and numerous in vivo animal studies, there is a vast body of literature detailing in vitro effects of wine, wine/grape extracts, and polyphenols. It is beyond the scope of this review to cover this research, especially when much of it has been reviewed recently. These include reviews on dietary polyphenols and cancer (Fresco and others 2006), resveratrol and multiple diseases (Guerrero and others 2009), resveratrol and cancer (Kraft and others 2009), quercetin and multiple diseases (Boots and others 2008), quercetin and cancer (Murakami and others 2008; Jagtap and others 2009), catechins and cancer (Jo and Lee 2007; Khan and Mukhtar 2008) (note that both reviews refer to catechin and related molecules such as proanthocyanidins), and catechin and neuroprotection (Mandel and Youdim 2004). Selected examples of work not covered by these reviews is reported subsequently.

One possible mode of anticancer action of resveratrol, revealed in an in vitro assay, was the conversion of resveratrol to piceatannol, which is a known antileukemic agent (Potter and others 2002). The conversion was achieved by the cytochrome P450 enzyme CYPIBI. The study is significant in that it shows that it is possible for compounds found in the diet to be converted to more bioactive forms by normal metabolic processes. Multiple anticancer actions of resveratrol itself were identified by Sun and others (2008) in malignant human pancreatic cells. These include increasing damage to mitochondrial function leading to increased reactive oxygen species, increased apoptosis, and possibly overcoming multi-drug resistance via inhibition of proteins associated with this condition.

In a recent study, red wine extracts and quercetin, but not resveratrol, were shown to have potent anticancer activity (Lee and others 2008). This study was undertaken using JB6 P+ cells derived from mouse epidermal cells. A key interaction identified was binding of quercetin to MEK1 and the authors concluded that “MEK1 is the most potent molecular target of [red wine extract] or quercetin for suppressing neoplastic transformation.” More recent research has also shed light on the possible neuroprotective effects of quercetin. Ossola and others (2009) concluded that while quercetin is unlikely to play an efficacious role in neurodegeneration, it may indeed be beneficial for neurodisorders involving cerebrovascular insult.

Compared to resveratrol and quercetin, catechin has received little attention recently in in vitro assays, apart from those related to cardiovascular disease above. For example, searching ISI across all years for the terms “catechin” and “wine” and “cancer” returned only 49 hits (100 with “quercetin,” 431 with “resveratrol”). Only 4 are from 2009 and none of those involved in vitro testing. A possible explanation for the lack of interest in catechin, with (+)-catechin the most abundant monomeric catechin in wine, is that recent attention on catechin-like molecules has focused on those found in tea and chocolate. These catechins tend to be oligomers with gallate groups, such as epigallocatechin gallate, and while there is much interest in these compounds and elucidating potential mechanisms of bioactivity, they are not relevant to wine.

Some in vitro research continues on the neuroprotective properties of catechin. Kraus and others (2007) showed that (+)-catechin and (−)-epicatechin increased the viability of microglial cells slightly when treated with amyloid-beta (25 to 35) and (1 to 40). In light of the varied bioactivities demonstrated for catechin in animal models, it is surprising that little in vitro research, particularly on (+)-catechin/(−)-epicatechin, has appeared in the literature in the last 4 to 5 y.

Earlier sections of this review have shown that there is a large body of literature demonstrating that wine and some of its polyphenolic components have the potential to protect against disease. The evidence is consistent across epidemiological, human intervention, in vivo animal, and in vitro studies. Nevertheless, to truly qualify as a functional food, a number of other factors must also be considered. First we explore the issue of bioavailability.


When applied to a pharmaceutical compound, bioavailability may be defined as the fraction of an administered dose of unchanged drug that reaches systemic circulation. When considering the bioactive constituents of wine, there is only one route of entry to circulation—oral—and as has been described above, the metabolites of wine polyphenols may themselves be bioactive. Therefore, for functional foods, the definition of bioavailability should include the original compound present in the food, as well as compounds derived from metabolic processes.

There are numerous studies demonstrating the bioavailability of the 3 main bioactive polyphenols in wine—resveratrol, quercetin, and catechin. While all 3 compounds are bioavailable, different studies have drawn attention to whether wine delivers sufficient amounts of these compounds to have health benefits (for example quercetin [de Vries and others 2001]). Nevertheless, one of the goals of functional food development is to increase levels of health-promoting compounds so that the functional food is capable of delivering an efficacious amount. This issue is discussed later on.

According to Baur and Sinclair (2006), “It is fair to say that the [pharmacokinetic] literature on resveratrol is, in many cases, contradictory and confusing.” While some researchers have concluded that resveratrol from wine, or indeed any dietary source, is unlikely to promote health (Vitaglione and others 2005), other authors, including Baur and Sinclair (2006), believe that the weight of evidence suggests that resveratrol is indeed likely to be beneficial. Resveratrol is rapidly and extensively metabolized (Walle and others 2004; Vitaglione and others 2005) to sulfate and glucuronide conjugates. Predictions of concentrations of resveratrol in systemic circulation vary (Kraft and others 2009), for example, “∼2.4 nM unmodified resveratrol and ∼180 nM total resveratrol from a dose equivalent to two glasses of red wine, and ∼9 μM authentic resveratrol and ∼680 μM total resveratrol from a high, but pharmacologically relevant, dose (based on rodent data) of resveratrol of 100 mg per kg body weight.” (Baur and Sinclair 2006). Using a dose of 400 mg resveratrol, Vaz-Da-Silva and others (2008) reported a maximum serum concentration of 200 nM. In the same study, the authors considered the effect of food on resveratrol absorption by co-administering the resveratrol with a standard high-fat meal. They found that while the rate of absorption decreased with food, the extent of absorption was unaffected. Despite the relatively low concentrations of resveratrol and its metabolites in systemic circulation, urinary resveratrol metabolites have been found to be reliable markers of wine consumption (Zamora-Ros and others 2009).

The bioavailability of quercetin has received more attention than that of resveratrol possibly as a consequence of the fact that quercetin has been identified as the most abundant polyphenol in the diet (Middleton and others 2000). The oral bioavailability of quercetin has recently been reviewed (Lesser and Wolffram 2006), however studies continue and Egert and others (2008) reported a median maximum plasma concentration of quercetin of 431 nmol/L from a dose of 150 mg quercetin. Quercetin is thus more bioavailable than resveratrol. In terms of wine consumption, de Vries and others (2001) questioned whether wine is a significant source of quercetin, although alcohol has been shown to increase quercetin absorption in a rate model (Dragoni and others 2006). The minor contribution of quercetin from wine or grape products is supported by Davalos and others (2006), who reported that quercetin from a glass of grape juice is bioavailable, but only leads to a small increase in plasma quercetin compared to ingestion of onions.

Catechin has similar bioavailability to quercetin (Manach and others 2005). Thus from a single dose of red wine containing 35 mg of catechin, 91 nM of catechin metabolites were found in human plasma 1 h after consumption (Donovan and others 1999). Similar to quercetin, very little catechin is present in a nonmetabolized form beginning from the first measurement of plasma following a dose. However, the study showed that alcohol does not affect absorption (using a de-alcoholized red wine), in possible contrast to quercetin. As mentioned above, recent studies on catechin and related compounds have focused on tea and chocolate. In this context, Roura and others (2007) have reported that milk does not affect the bioavailability of catechins. Further complications regarding the bioavailability of catechin arise because of stereomeric consideration ((+)-catechin and (−)-epicatechin are both found in wine) and the various polymeric forms available (tannins, proanthocyanidins, and so on). Studies in rats have shown that (−)-epicatechin is more bioavailable than (+)-catechin (Baba and others 2001), and others rat studies have been used to learn about the absorption of proanthocyanidin dimers (Appeldoorn and others 2009b). As far as we are aware these studies have not been undertaken in human trials, although there is interest in the different issues raised (Rasmussen and others 2005).

The 3 polyphenols that are the focus of this review have demonstrable bioavailability and bioactivity. Wine has been the focus for human consumption of resveratrol, yet amounts are low in wine. Likewise, quercetin and catechin are relatively low in wine. Thus, to further functionalize wine, increasing concentrations of these components would be advantageous. To what extent this can be achieved is dependent on at least 2 factors—one, the levels at which these compounds may become toxic to human consumption; and two, the ability to enhance wine with these compounds through either viticultural or enological practices. Both are addressed in turn.

Toxicology in the Functional Food Market

Determining the potential adverse effects of wine with enhanced levels of polyphenols is a challenge on a number of levels. First, it does not fall into the realm of what toxicology has traditionally studied, as argued by Rietjens and Alink (2006)“Toxicology historically has been directed at studying the mechanisms of adverse effects of isolated compounds on living organisms at high levels of exposure, forming the basis for risk and safety assessment.” The researchers also argued that rather than studying the adverse effects of high doses of single compounds, risk-benefit analyses on low-dose intakes should be aimed at understanding “the biological effects of chemical compounds on living organisms” at physiologically relevant levels. Another challenge is that any food or beverage is a complex chemical system and enhancing one or more components may result in unpredictable synergistic or antagonistic effects. Likewise, removing a compound from its complex naturally occurring matrix is generally assumed to result in higher toxicity (Milner 1999) and, although enhancement of polyphenols in wine through various practices may not constitute such a scenario, extreme manipulations should be treated cautiously.

Diet-derived active compounds such as flavonoids are generally regarded as safe based on their long history of consumption in the diet and use as natural remedies. However, the upper safe limits of these compounds need to be established to provide guidelines as to what levels are safe to produce a product that consumers can be confident is safe. At some concentration, all compounds become toxic and, therefore, safe upper limits are an essential part of describing a functional food.

Unlike wine health studies, wine toxicity studies are not well documented (there are no articles on the ISI Web of Science database with both wine and toxicology in the title; and only 3 where wine is in the title and toxicology is present as a topic word, none of which are relevant to this review). In fact the closest studies to wine toxicology that the authors have been able to locate are 3 studies from 2002 (Yamakoshi and others 2002; Wren and others 2002; Bentivegna and Whitney 2002) on the toxicology of grape seed/skin extracts. All studies found the no-observed-adverse-effect level (NOAEL) for rats to be around 2% (w/w) in the diet. No attempt was made to characterize these commercial extracts for individual phenolic compounds. However, by far the largest proportion of toxicological studies that have been conducted have been directed at single compounds of relevance to wine. Consistent with the earlier sections of this review, we consider more recent studies on resveratrol, quercetin, and catechin.


An early study (Juan and others 2002) investigating the safety of resveratrol utilized a daily dose of 20 mg/kg bw (bw = body weight) in rats over a 28-d period. This was considered to be 1000 times the dose expected in humans consuming 250 mL red wine per day. The researchers concluded that “The lack of harmful effects found in the hematology, clinical chemistry, and histopathology indicates that trans-resveratrol has a large safety margin.” This was backed up by Crowell and others (2004) who found the NOAEL in rats to be 300 mg/kg/d over a 28-d period. However, adverse effects were detected at 1000 mg/kg/d and became severe at 3000 mg/kg/d. The adverse effects included nephrotoxicity, increased kidney weight, changed gross renal pathology, and increased incidence and severity of histopathological changes in the kidney (Crowell and others 2004).

A wide battery of tests was recently conducted on Resvida (TM), a commercial, high-purity resveratrol supplement (Williams and others 2009). None of the in vitro tests, such as those to determine skin or eye irritation, returned adverse effects for Resvida (TM). Likewise, no adverse effects were observed in rats fed 700 mg/kg bw/d for 90 d. Other in vitro tests showed “that Resvida (TM) is readily absorbed, metabolized, and excreted.”

In the one human trial that the authors are aware of (Almeida and others 2009), the maximum dose of resveratrol was 150 mg (about 2 to 2.5 mg/kg bw) administered every 4 for 48 h (a total of 13 doses). The most commonly reported side effect was mild headache and all adverse effects were mild. No clinically significant abnormalities could be detected based on hematological parameters, vital signs, neurological examinations, or ECG recordings. The researchers concluded that “repeated administration [of resveratrol] was well-tolerated.”

While most studies generally find that resveratrol has no or minimal adverse effects, a cautionary note arises from the findings of Fujimoto and others (2009). In an in vitro study using rat thymocytes, the authors reported that at therapeutic concentrations of 10 μM resveratrol induced apoptosis in normal cells. Whether this in vitro study is pertinent to humans and whether resveratrol in its “natural” matrix is also cytotoxic requires further research.


There have been mixed reports on the safety of quercetin and the data to 2005 have been reviewed by Okamoto (2005). Early on, quercetin was classified as mutagenic, but in 1999 the Intl. Agency for Research on Cancer (IARC) released findings demonstrating that quercetin is not carcinogenic to humans. The safety of quercetin was later affirmed by Harwood and others (2007) who critically reviewed the conflicting in vitro and in vivo data.

More recently, Ruiz and others (2009) have reported that doses of quercetin up to 3000 mg/kg bw/day over 28 days do not result in any measurable adverse effects in Swiss mice. Parameters investigated included mortality, food and water consumption, weight gain, biochemical parameters and histopathology of clinical signs, and organ weights. A difference with the control group was noted for red blood cell count and hematocrit, both increasing after quercetin administration. These findings can be contrasted with some earlier findings where some adverse effects of quercetin were noted at lower doses. These include enhanced proteinuria at 1133 mg/kg bw/d in Wistar rats (Rangan and others 2002), and an increased severity of chronic nephropathy, hyperplasia, and neoplasia of the renal tubular epithelium (Dunnick and Hailey 1992). Nevertheless, as reviewed by Harwood and others (2007)“it may be concluded that quercetin, at estimated dietary intake levels, would not produce adverse health effects.”


No toxicological studies on catechin itself, comparable to those of resveratrol and quercetin were found. The nearest are the reports on grape seed extract as previously mentioned, where mixtures of monomeric and oligomeric catechins have been tested. Further evidence of the safety of such compounds comes from studies on Oligonol (Fujii and others 2008) an oligomerized polyphenol formulated from lychee and green tea extracts, consisting mainly of catechin monomers and oligomers. No adverse effects were recorded in a single dose of 2000 mg/kg bw to rats. Likewise, no adverse effects were observed in a 90-d feeding trial with doses up to 1000 mg/kg bw/d.

The toxicological data on resveratrol, quercetin, and catechin (or related compounds) overwhelmingly support their safety for human consumption. Nevertheless, the authors have not been able to locate data on the upper safe limits for these compounds. Such data are required to ensure that, if wines are produced with enhanced levels of these compounds, they do not exceed the safe upper limit. On the other hand, given that wine has relatively low levels of these compounds, with some varietal differences (Table 2), it is unlikely that manipulation of levels will reach anywhere near those used in toxicology studies. Apart from direct addition of pure polyphenols to a finished wine, wines may be enhanced in resveratrol, quercetin, and catechin through viticultural or vinification processes and these are reviewed in the next section.

Table 2–.  Concentration of phenolic compounds in Cabernet Sauvignon, Shiraz, and Pinot Noir wines from Australia, Hungary, and California.
Grape varietyCountry/regionResveratrol (mg/L)Quercetin (mg/L)Catechin (mg/L)
  1. aStockley and Hoj (2005).

  2. bNikfardjam and others (2006).

  3. cGoldberg and others (1996a).

  4. dGoldberg and others (1998a).

  5. eGoldberg and others (1998b).

Cabernet SauvignonAustraliaa1.9 ± 0.139.2 ± 1.032 ± 4.4
Hungaryb2.8 ± 2.45.6 ± 4.381.8 ± 47.3
California3.2 ± 0.8c7.0 ± 0.6d43 ± 2.6e
ShirazAustralia2.64 ± 0.4210.8 ± 1.222 ± 3.2
Hungary1.1 ± 0.213.4 ± 1.868.2 ± 5.4
Californianot available5.8 ± 0.9d26 ± 3.2e
Pinot NoirAustralia7.58 ± 1.661.8 ± 0.675 ± 10.2
Hungary3.2 ± 0.57.5 ± 2.0103 ± 46.4
California16.0 ± 1.2c5.0 ± 0.8d119 ± 7.6e

Methods to Increase Health Beneficial Compounds in Wine

Viticultural practices

The accumulation of phenolic compounds in berries can be varied depending on how the grapevine is treated (Jackson and Lombard 1993). Ultimately, a combination of environmental factors such as temperature, humidity, rainfall, sunshine, soil, wind, nutrients, and viticultural practices such as varietal selection, rootstock selection, crop load, soil/vineyard floor management, pests and disease control, canopy management, pruning, irrigation, and fertilization determine the grape composition and affect wine quality (Stockley and Hoj 2005). The relationship between any one of these parameters and concentration of phenolic compounds in berries is complex and it is not uncommon that contrasting effects are seen for individual phenols. For example, sun exposure on clusters increases the levels of quercetin-3-glucosides (Spayd and others 2002) and total anthocyanins, but decreases levels of caftaric acid, (+)-catechin, and (−)-epicatechin (Price and others 1995).

Berli and others (2008) have investigated the effects of UV-B exposure and altitude on levels of resveratrol, quercetin, and catechin in grape skins of the Malbec cultivar. The most pronounced effect was for grapes exposed to UV-B radiation at 1500 m above sea level where the level of resveratrol was significantly enhanced. Quercetin also increased with UV-B exposure at 500 m above sea level, whereas catechin concentrations did not appear to be significantly affected by light treatment at any altitude.

Cluster thinning is well known to impact on the composition of berries, with low crop levels associated with higher-quality wines (Prajitna and others 2007). In a study specifically looking at enhancing the antioxidant capacity of wine from the Chambourcin cultivar, Prajitna and others (2007) measured total phenols, antioxidant capacity, and resveratrol levels. They found that cluster thinning increased all of these parameters by up to 48%, 53%, and 200%, respectively, depending on the vintage. However, the authors commented that “the mechanism of how cluster thinning affected the polyphenolic content in berries and wines remains unclear and further studies are warranted.”

Cultivar is the major determinant of chemical composition in berries (Goldberg and others 2000). For example, “Pinot Noir wines generally have a high concentration of the phenolic acid, gallic acid, the flavanols, catechin, and epicatechin, and the stilbene, resveratrol” (see Table 2) (Stockley and Hoj 2005). However, once vines are planted, a wine-maker is limited to the cultivars at hand to produce certain styles of wine. On the other hand, there remains the possibility of developing high phenolic-content varieties, that may be used in ‘health enhanced’ blends. Blending would be necessary because single-variety wines with high levels of phenolic compounds are unlikely to be acceptable to consumers due to the bitterness and/or astringency inherent in polyphenols (see subsequently).

At the viticultural level, health beneficial compounds can be increased by controlling or adjusting the environmental factors and viticultural practices but, as there are many interacting combinations that can influence grape composition, it is difficult to control the quantities of such compounds. “There is no one simple rule for the synthesis and accumulation of the individual classes of phenolic compounds in berries” (Stockley and Hoj 2005). While ongoing research efforts will continue to provide more insight into the complex mechanisms involved, in the short term it is not possible to produce grapes with specified levels of polyphenolic compounds. A possible solution is to add specific compounds at well-defined levels during vinification; however, this may reduce the “authenticity” of wine and may be seen by consumers as “unnatural.” The practice is also banned in some countries (in Australia only a small number of compounds are allowed to be added to wine that are not grape-derived), presumably to maintain wine as a natural product. Vinification also provides another opportunity to enhance certain compounds to develop a “healthy” wine.

Vinification practices

Under commonly used vinification processes, approximately 50% of the phenolic compounds in berries are transferred into a finished wine (Stockley and Hoj 2005). While the factors that influence the phenolic composition are reasonably well understood (see subsequently), it would appear that significant increases above 50% may be difficult to obtain using traditional methods. During vinification, levels of phenolic compounds may be influenced by yeast selection (Caridi and others 2004; Clare and others 2005), longer skin maceration times (Gomez-Plaza and others 2001), longer skin fermentation time (Auw and others 1996), increasing fermentation temperature (Zimman and others 2002), selecting appropriate fining agents (Castellari and others 1998; Castillo-Sanchez and others 2008), oak aging (Sanchez-Iglesias and others 2009) and, logically, blending with wine known to contain high levels of target compounds. Few of these studies have considered the effect of vinification on specific compounds such as resveratrol, quercetin, and catechin, although resveratrol was the object of the studies of Castellari and others (1998) and Clare and others (2004, 2005).

Recently, alternative technologies have been looked at in an effort to increase phenolic content (especially resveratrol) in grapes, postharvest. Thus, Gonzalez-Barrio and others (2006) found that UV-C and ozone treatments increased stilbenoid compounds (resveratrol and related species) in a white table grape variety. While not a wine variety, the results demonstrated that resveratrol levels in grapes can be manipulated post-harvest. Jimenez and others (2007) have reported on the use of short anoxic treatments to increase resveratrol concentrations in grapes. In this technique, grape bunches are put into a dry nitrogen-filled chamber (up to 1.1 bar) for 6 to 48 h. Anoxic treatment for 15 h resulted in a 2.5-fold increase in resveratrol and no anoxic damage to the grapes. Longer treatments led to damaged grapes.

While there are many methods that can increase polyphenol levels in wine through viticultural and vinification practices, there are at least 2 issues that must be considered in the context of wine as a functional food. The first is standardization. As there are so many different factors that can influence polyphenol levels, it is not easy to control their concentrations. Thus it would currently be difficult to produce a wine with predetermined specified levels of resveratrol, quercetin, and catechin. While this would be desirable from a labeling and consumer point of view (consumers knowing exactly what “dose” of particular compounds they are receiving), it is simply not feasible given the many variables.

The other drawback in increasing polyphenol concentrations in wine is that they are responsible for the taste of bitterness and mouthfeel of astringency (Arnold and Noble 1978; Peleg and others 1999; Brossaud and others 2001; Lesschaeve and Noble 2005). Excessive bitterness and astringency are likely to be off-putting for many consumers, however, it remains unknown to what extent this may impact consumer decisions if there is an accompanying health benefit. However, the ultimate goal would be to produce wine with maximum levels of polyphenol compounds, as well as with well-regarded taste, because “consumers now want both health benefits and palatable food” (Lesschaeve and Noble 2005). The tastes of individual flavonoids are shown in Table 3.

Table 3–.  Flavonoid subclasses and their taste quality. Adapted from Drewnowski and Gomez-Carneros (2000).
Flavonoid subclassExamplesTaste quality
(−)-epicatechinBitter & astringent
(−)-epigallocatechinBitter & astringent
(−)-epicatechin 3-gallateBitter with sweet
(−)-epigallocatechin 3-gallateaftertaste
IsoflavonesDaidzeinBitter or astringent

Alcohol—The Most Controversial Issue in the Wine and Health Research

The most controversial issue in wine and health is the effect of alcohol on human health. The literature on this issue is vast and conflicting. In fact, if one were to ignore the potential negative health effects of alcohol, it would be fair to recommend that people ought to consume about as much wine as was appropriate given attention to caloric intake. Part of the controversy arises from the fact that the risk of certain diseases, most notably cardiovascular disease (as reviewed previously), is reduced with moderate alcohol consumption. On the other hand, there seems to be strong evidence that alcohol consumption is associated with an increased risk of other diseases, in particular certain forms of cancer. Compounding these conflicting effects is the fact that in trying to synthesize the huge body of literature, some authors do not distinguish between the various beverages in which alcohol is found. Given the large amount of evidence that is accumulating for the health benefits of the polyphenols in wine, including beneficial effects on cancer, it would seem appropriate to assess the risks of disease on a beverage-by-beverage basis, rather than grouping all alcoholic drinks together.

A further complication in this debate is the fact that, as far as the authors are aware, nobody has produced a full-health index of risk. For example, if moderate wine consumption increases the risk of larynx cancer by 10%, but decreases the risk of cardiovascular disease by 20%, is a person better off overall drinking wine or not? Of course, to be useful, such an index would also need to be linked with a particular individual's propensity for developing the different disease states and the likelihood that a particular diet or other interventions may be efficacious for them personally. Currently there is considerable effort at understanding such issues at the individual level through the burgeoning fields of metabolomics, metabonomics, and personalized medicine; however, a discussion of that literature is beyond the scope of this review. In the meantime, there are data on the association between alcohol consumption and total mortality; and, consistently, moderate alcohol consumption has been shown to be linked with lower total mortality (Fuchs and others 1995; Thun and others 1997; Di Castelnuovo and others 2006).

A recent review on the link between alcohol and cancer has been published by the Cancer Insti. of New South Wales (Lewis and others 2008). The authors examined 634 reviews of the literature in this area and found that only 31 of these studies matched their inclusion/exclusion criteria. Of these, 7 were examined in detail. Total of 19 different types of cancer (Table 4) were investigated for links between alcohol and increased risk of developing that cancer. We have re-examined some of the primary literature to determine if any studies have looked at different types of alcoholic beverages and their effects on cancer, and have included the results for wine in Table 4.

Table 4–.  Links between cancer and alcohol. Adapted from Lewis and others (2008).
Cancer siteRelationship between alcohol and cancerEvidence baseEffect of wine
  1. aModerate consumption is defined as up to 2 alcoholic drinks per day.

  2. bNo increased risk for <30 g alcohol per day (Bongaerts and others 2008).

  3. cLater findings report “In the absence of smoking, the relative and absolute risks [of laryngeal cancer] are small for moderate alcohol consumption” (La Vecchia and others 2008).

Cancer at any siteNo relationship with moderate consumptionaConvincing 
Increased risk with high consumption  
BreastIncreased risk, even with moderate consumptionConvincingHigh consumption of wine increased risks, presumably low or no effect from moderate consumption (Mattisson and others 2004)
ColonIncreased risk, even with moderate consumptionbConvincing“For several cancers (oesophagus, stomach, colon, liver, pancreas, lung, prostate) there was evidence of increased risk among alcohol consumers compared with abstainers and occasional drinkers. For most sites, it was beer and to a lesser extent spirits consumption that drove the excess risks.” (Benedetti and others 2009)
LiverIncreased risk, even with moderate consumptionConvincingSee Benedetti and others (2009).
RectumIncreased risk, even with moderate consumptionbConvincing 
StomachIncreased risk, even with moderate consumptionConvincingSee Benedetti and others (2009).
Upper aero-digestive tractIncreased risk, even with moderate consumptioncConvincingLower risk for moderate wine consumption (<3 drinks per day); higher risk for high wine consumption (>5 drinks per day)(Bosetti and others 2000)
CervixNo relationshipInsufficient 
GallbladderNo relationshipInsufficient 
KidneyNo relationshipInsufficient 
LungPossibly increased risk, heavily confounded by smokingInconsistentDecreased risk with red wine (Ruano-Ravina and others 2004)
MelanomaNo relationshipInsufficient 
Non-hodgkin's lymphomaPossibly decreased riskInsufficient 
OvaryConflicting-evidence of increased and decreased riskInconsistent 
ProstateNo relationship with low consumptionInconsistent 
Possibly increased risk with heavy consumption  
Small intestineNo relationshipInsufficient 
ThyroidPossibly decreased risk, confounded by smokingInadequate 
BladderNo relationshipConvincing 
EndometriumNo relationshipConvincing 
PancreasNo relationshipConvincingSee Benedetti and others (2009).

For some types of cancer (and in some studies), moderate wine consumption is not associated with increased risk of cancer, for example, upper aero-digestive tract cancers (Bosetti and others 2000) and others (Benedetti and others 2009). Further investigation of the primary literature reveals a number of studies reporting increased risk of cancer at >30 g/d (Smith-Warner and others 1998; Mattisson and others 2004). This quantity is higher than recommended daily intake. These studies remain silent on possible risk at lower levels of consumption. One study (Smith-Warner and others 1998), however, shows a linear increase in risk starting at zero alcohol consumption, the implication being that there is no safe dose of alcohol at any level. This ignores the basic principles of toxicology—before a toxic dose is reached there should be no effect or a beneficial effect. On the other hand, until a threshold level is determined, researchers are obliged to exercise caution. In terms of wine's status as a functional food, determining this threshold level for cancer is important. In some cases, these data exist but have not been interrogated fully. For example, Mattisson and others (2004) reported increased relative risk of breast cancer for 4 categories of wine consumers: abstainers, 1.7, 10.8, and >31.6 g/d. The increased relative risk (95% confidence intervals) calculated for these 4 groups are: 1.21 (0.86 to 17.2); 1 (no interval reported); 0.88 (0.69 to 1.13); and 2.11 (1.24 to 3.60), respectively. Indeed, Mattisson and others (2004) concluded “… our results support the hypothesis of a threshold effect of alcohol intake on breast cancer risk.” A threshold has also been proposed for alcohol and liver disease (Kamper-Jorgensen and others 2004).

The uncertainty over the place of alcohol in a normal diet is reflected in the advice given by various health groups. For example, the Cancer Council of Australia recommends that alcohol consumption should be limited or avoided to reduce the risk of cancers: “Drinking any type of alcohol (beer, wine, or spirits) increases the risk of developing cancer of the bowel, mouth, pharynx, larynx, oesophagus, liver, and breast. Cancer risk … increases from the first alcoholic drink you have” (Cancer Council Australia 2010). This view is supported by the WHO which concludes that the risks of alcohol consumption outweigh the benefits: “Consumption of alcohol beverage is not recommended; if consumed, do not exceed two units per day”(WHO 2003). On the other hand, The Natl. Heart Foundation of Australia provides information on healthy eating and drinking that includes: “If you drink alcohol, have no more than two standard drinks a day” (Natl. Heart Foundation 2009). The American Assoc. for the Study of Liver Diseases recommends that “Up to 2 drinks a day for men and 1 drink a day for women is considered safe with regard to the liver” (The American Assoc. for the Study of Liver Diseases 2005). In a strong statement in support of the health benefits of alcohol, the U.S. Dept. of Health and Human Services (U.S. Dept. of Health and Human Services 2005) reports that “in middle-aged and older adults, a daily intake of one to two alcoholic beverages is associated with the lowest all-cause mortality.” Clearly, given the divisions between these different government bodies, there is still some way to go before the issue of alcohol and health is settled.

Pathways to Wine Being Accepted as a Functional Food

With increasing interest in functional foods, many companies and governments are actively involved in functional food development. A multitude of possibilities exist: “a food that naturally contains sufficient amounts of a beneficial nutrient component; a food in which one of the components has been naturally enhanced through special growing conditions, new feed composition (animals), or genetic manipulation; a food with a modified recipe formulation that incorporates a functional ingredient; a food in which the nature of one or more components or their bioavailability in humans has been modified by means of specialized food processing technologies; a food from which a deleterious component has been removed, reduced, or replaced with another substance with beneficial effects” (Kotilainen and others 2006). Such development is necessary, but studies focused only on health itself are not enough to guarantee the success of a functional or health-enhancing food. Levitt (2004) suggested that identifying customers’ needs should be the priority in product development as customers are not interested in products that are not relevant to them. The opinion is that all industries should be focused on customer satisfaction, not on goods production; and it was stressed that “an industry begins with the customer and his or her needs, not with a patent, a raw material, or a selling skill” and goods should be created to suit the customers’ needs.

The need for functional food development to have a customer focus has been reported by Lethbridge and others (2006), who suggested that “providing a tasty, nutritional product at a competitive price point is very important for the market success of functional foods.” In their view “it is clear that the market is driven by taste, convenience, disease prevention, nutrition individualization, and snacking.” Another aspect of the marketing of functional foods is that of how to promote the health benefits. Weststrate and others (2002) have argued that to be truly effective, functional foods should primarily be aimed at function improvement or (longer-term) disease risk reduction for ‘healthy’ people, and not at disease treatment for ‘sick’ people.” As age-related diseases such as heart disease and cancers show their symptoms a substantial period of time after the onset of the disease, controlling these diseases at an early age is essential. Research has shown that symptoms of some diseases can be present early on in life. For example, in an alarming study, 20% of 15- to 19-y-old subjects already had symptoms of heart disease (raised fatty streaks in the abdominal blood vessels) and this percentage increased to “40% for 30- to 34-y-old subjects” (McGill and others 2000). In another study, American soldiers who were killed during the Vietnam War were autopsied and it was discovered that people as young as 18 had atherosclerotic plaque deposits. In fact, 45% of the soldiers autopsied had atherosclerosis (Joseph and others 1993); this population would partake in above-normal levels of exercise, suggesting that even higher levels may be reported in a general sample. Whilst early intervention to prevent atherosclerosis is essential, in terms of promotion of functional foods to younger consumers, this may be challenging.

The challenges of promoting health products to younger consumers were reported by Krystallis and others (2008) who studied the consumer motivations and cognitive structures for the purchasing of functional foods. Their aim was to identify the most purchased functional foods and define the functional food attributes that affect consumers’ purchasing decisions in 2 groups, young adult (25 to 34 y old) and early middle age (35 to 44 y old). In product preference, young adults preferred energy-boosting products, while the early middle age group preferred health functional products. Information on the product played a big role in the purchasing decision, especially consequence-related knowledge, such as “good for the bones,” which was more effective than attribute-related knowledge, for instance, “extra calcium added.” Positive product image such as naturalness and history of product as a functional food also influenced the purchasing decision. Taste was one of the most important factors that affected consumer choice. In the perceived quality attributes, “healthy product,”“quality product,”“safe food,” and “pure product” attributes were the priority order in the young adult group and “healthy product,”“safe food,” and “natural and pure products” were the priority order in the early middle age group. In the organoleptic attributes, “nice taste” was the number one consideration in the purchasing decision in both groups. In the functionality attributes, “contributes to good physical condition,”“enforces body defense and provides proved health claims,”“added vitamins and minerals;”“removed dangerous ingredients,” and “provides more energy” attributes were the priority order in the young adult group while “contributes to good physical condition,”“enforces body defense,”“provides proved health claims,”“contributes to digestion improvement, low cholesterol level, and removed dangerous ingredients” were the priority order in the early middle age group. With this result, it is clear that the most important attributes for consumer purchasing decisions were functionality. In the price attributes, “value for money” was the first priority in both groups. Cleary for wine to succeed as a functional food, its functionality must be demonstrated, and it must be available at a competitive price and with acceptable taste.

Wine functionality

These consumer preference studies give us a number of insights that are relevant to the wine industry. Functionally enhanced wine needs to be created for the middle to older age group as these age groups are more interested in health-beneficial products. While such products are being developed (see subsequently), it is still difficult to claim functionality, through labeling, for example, as wine contains alcohol, making government agencies reluctant to grant approval for health claims. This is despite the fact that other government agencies are prepared to recommend moderate alcohol consumption as part of a healthy diet (see above).

Are there any solutions that may enable wine to be marketed with health-benefit claims?Barreiro-Hurle and others (2008) suggested promoting “wine as a healthy product” and using the right channels to promote functional wine, such as: direct marketing to wine clubs; persuading government agencies with health claims that wine, which contains high amounts of polyphenols, is different from other alcoholic beverages which do not contain these health-beneficial compounds; and generic advertising of specific health-beneficial compounds such as resveratrol. Krystallis and others (2008) also suggested that a creative approach is required to position wine as a functional food or health-enhancing alcoholic beverage.

Already a small number of wines are being marketed with health claims in some countries. Not all of the examples given here are functionally enhanced wines, however, they do illustrate the trend towards wines labeled with health claims. In San Francisco, a wine company made the following controversial statement on a wine label that was approved by The U.S. Bureau of Alcohol, Tobacco, and Firearms: “We encourage you to consult your physician about the health effects of wine consumption” (Dornin 2000). Anti-alcohol groups interpreted this statement as the wine industry promoting wine as a healthy product. The wine industry claims otherwise. Subsequently in the U.S., winemakers pushed for a new rule that allowed wine labels to list health claims and the U.S Government left open the opportunity for wine makers to make specific health claims with balanced, sound scientific and nonmisleading evidence. The claim should be very specific such as “the biggest benefits would come only to those at greatest risk of heart disease and the risks for younger drinkers might outweigh the benefits” (Norris and Hesser 2003). In the EU, “beverages containing more than 1.2% by volume of alcohol shall not bear a health claim,” but “alcohol-free wine for which a market is growing may bear health claims” (Gorny 2007). In Germany, health claims for wine are not permitted, except “suitable for diabetics” and even that with certain limitations (Gorny 2007). In Australia, the labeling regulations for any food are very strict—“a food (including wine) may not have a ‘health claim’ on any label or in an advertisement” (Australian Wine and Brandy Corp. 2010). While not directly making a health claim, labels such as “REW” (Resveratrol Enhanced Wine) are allowed by FSANZ (Food Standards Australia New Zealand). Functionalization of a wine with resveratrol takes advantage of the huge publicity that resveratrol has received for nearly 2 decades. Such wines have been shown to attract higher prices because of the perceived health benefits of resveratrol.

With growing interest in the health functionality of wine, winemakers are likely to argue for the inclusion of more strongly worded health-related statements on the label. Nevertheless, until controversies surrounding the effects of moderate alcohol consumption are resolved (see above), it is unlikely that labeling laws will be changed in the near future.

Wine price

Pricing of a functional food is strongly related to consumer awareness of the functionality of the food. To what extent consumers are willing to pay extra for a functionalized wine was investigated by Barreiro-Hurle and others (2008). Their consumer acceptance study in Spain showed that 99.7% of consumers regarded wine as a healthy product, but only 6% of consumers knew the health-beneficial reasons, namely, the presence of antioxidants or polyphenols. Even so, generally consumers strongly believed that a functional food is good for their health. Resveratrol-enriched wine had the highest attribute coefficient in their study, followed by aging wine. In terms of wine price, consumers were willing to pay more for the resveratrol-enhanced wines. For example, the price that consumers were willing to pay for a regular young Andalusian wine (wine produced in a well-established red wine-producing region) was €3.11. On the other hand, the same wine with resveratrol enrichment commanded a much higher price of €9.01. Consumers were willing to pay 190% more for the resveratrol-enriched wine. However, the price differences were reduced as the wine price increased. For example, the wine price that consumers were willing to pay for the Rioja regular-aged wine was €9.84, but the price for resveratrol-enriched wine was €15.73, which was only 60% different.

Wine taste

As far as the authors are aware, no study has yet been reported in which functionalized wines have been taste-tested by consumers. Given that some phenolic compounds are bitter and astringent, enhancing levels of these compounds in wine may result in consumer rejection. Nevertheless, it is an interesting question that needs to be investigated—to what extent would consumers be willing to sacrifice taste to consume a healthier product?

In general, sensory perception of wine by consumers is still a vexed field even aside from the question of functionality. Wine-makers and wine judges have fixed ideas of how good-quality wine should taste. The wine-makers rely on marketing to promote the wine they think should be made. Taking the view of Levitt (2004) the industry could be more consumer-focused, where wine is made for targeted segments. Consumers, faced with a bewildering choice of varieties and labels, tend to take advice from wine judges. This too has some important limitations in both fidelity and reliability. For example, sometimes results are not consistent, depending on where or who evaluates the wine. Young (2003) revealed that “it has taken this writer 30 y to realize that most wine competitions are either a hoax, a myth, or entirely invalid; in some cases, absolutely fraudulent.” He concluded that “Any judging of wine can only be: by this group of people, from this bottle, at this place, at this time. No judgment can, possibly, be anything else.” The current authors do not ascribe to this view; however, it does highlight the level of frustration at the limited fidelity and reliability that can be associated with wine sensory evaluation. Since this is currently an issue with conventional wines, it will also need to be addressed for functionalized wine.

One method to assess the taste-impact of a single compound or group of compounds in wine has been proposed by Prescott and others (2005). The authors propose a consumer rejection threshold technique, which has been adapted by researchers to assess potential positive qualities as well as negative ones (Saliba and others 2009). Saliba and others (2009) propose that the addition of some compounds to wine may be significantly preferred by consumers, though at some point, possibly at high concentrations, they would be rejected. This quantitative methodology allows the determination of preferred levels that should be aimed for, based on taste preference alone, and levels that should be avoided, or further tested to determine whether a health claim changes taste acceptance.

Wine and moderate drinking

Due to the controversy still associated with alcohol consumption and health (see above), a much clearer picture of what constitutes drinking in moderation is required. Klatsky (2004) recognized that the “definitions of moderate and heavy drinking are arbitrary” and proposed an “operational boundary” defined as “the level of drinking above which net harm is evident in epidemiologic studies. Thus, 3 or more drinks per day is called ‘heavy’, and lesser amounts ‘light’, ‘lighter’, or ‘moderate’ drinking.” In light of more recent work by Di Castelnuovo and others (2006), that boundary may need to be revised down.

The “drinks” that Klatsky refers to are so-called “standard drinks.” This too is an area that the functional food market needs to address to facilitate a better understanding of the link between alcohol and health. Currently different regulations on what constitutes a standard drink apply in different countries, with the range being 8 g alcohol per standard drink in the UK to 19.75 g/standard drink in Japan (International Centre for Alcohol Policy 2010). Thus the point at which drinking becomes “heavy” would be 24 g/d in the UK and 59.25 g/d in Japan. However, as reviewed previously, significant health risks are associated with 30 g/d or over. Using grams of alcohol consumed per day to define drinking habits may assist the current situation where standard drinks vary from country to country. Thus “moderate” consumption may be about 20 g/d and light consumption 10 g/d or less. The authors believe that this approach provides the clearest information to consumers and is the only realistic method to communicate alcohol levels as they relate to healthy doses. Many consumers are familiar with culturally-derived measures, such as pints of beer, such that public health education may be required in some countries, while labelling laws may also need to be established.

A logical concern associated with the promotion of alcoholic beverages as “healthy” is that consumption may be encouraged beyond healthy limits. The popular notion of “if it is healthy, you can eat more” was investigated (Provencher and others 2009). Results showed that participants consumed 35% more oatmeal-raisin cookies when the snack was regarded as healthy. It is unlikely that the same effect would result from the perception that wine is healthy, since the over-consumption of alcohol is a focus of education and research programs (Croom and others 2009). Supporting this notion is research that has found that those who perceive wine as healthy tended to drink more frequently but not consume higher amounts across one week, suggesting healthy drinking patterns (Saliba and Moran 2010). Based on these results, there is no reason to suppose consumers’ well-being will be threatened by their behavioral response to positive information on health benefits of wine (Saliba and Moran 2010).

Even with the best of intentions, it may be difficult for consumers to determine dosage levels given current packaging. Innovative packaging could be developed for functionalized wine to assist the health-conscious consumer to monitor wine consumption. For example wine containers could be designed to hold multiples of standard drinks, for example, 2 or 4, or set amounts of alcohol 20 or 40 grams. A consumer who purchased a bottle of functionalized wine containing 20 g of alcohol would then know that this provides the total recommended intake of alcohol for one person for one day.


The functional food market is growing rapidly as people are more interested in their health and life expectancy, while health costs and disposable incomes have increased. The success factors for functional foods are conventional food form, good taste, convenience, disease prevention, and competitive price. Wine meets all of these criteria; however measurements of health-beneficial functionality have not been standardized yet. Nonetheless, in vitro, in vivo (animal), human intervention and epidemiological studies, all point to the strong possibility that wine, in moderate consumption, is beneficial to health.

To make wine successful in the functional food market further studies are required in consumer perception of healthiness of wine; the relationship between taste and health-enhancing properties (both perceived and actual); viticultural and vinification practices that influence polyphenol concentration and composition; the comparison between wine and other alcoholic beverages that contain no polyphenols; consumer preference on pricing; new wine evaluation methods that measure health benefits; resolving the controversy regarding alcohol function in wine; and the development of a wine and health education program.

We conclude with a quote from the Report on Functional Foods by the Food and Agriculture Organization. What is proposed for functional foods in general is especially true for wine in particular, should it be considered a functional food: “At the end, even though a functional food is considered to have specific health benefits, excessive consumption should not be encouraged as this may displace other foods in the diet. The emphasis should be on variety, moderation, balance, and a combination of foods to promote health and nutrition wellbeing” (Subirade 2007).