Red Wine and Diabetes Health: Getting Skin in the Game

  1. R. Paul Robertson
  1. Pacific Northwest Diabetes Research Institute and the Division of Metabolism, Endocrinology, and Nutrition, Departments of Medicine and Pharmacology, University of Washington, Seattle, WA
  1. Corresponding author: R. Paul Robertson, rpr{at}

It was an unexpected pleasure to be invited to write a Perspectives in Diabetes article considering the proposition that moderate consumption of red wine provides health benefits for people with diabetes. At the outset, I want to make clear that since Perspectives in Diabetes are not review articles in the conventional sense, this one does not exhaustively analyze the effects of beer, hard liquor, or alcohol per se on health. This one is all about wine, especially red wine. This is an important point because of the intrinsic psychic influences of wine. Beer is associated with boisterous behavior at sporting events. Hard liquor is associated with serious drinking and dark moods. Wine, on the other hand, is associated with relaxation, reflection, celebration, conviviality, toasting, and a certain amount of dry humor. So, in this spirit, I will lace this article with a modicum of these wine-related characteristics.

Archeologists tell us that humans made and drank wine in the Middle East (Fig. 1) beginning in ~7000 BCE—well before recorded time. During its 9,000-year history, wine has been used for many purposes, including religious (Fig. 2) and medical ones. The scientific literature over the past half century does not explicitly warn against drinking wine in moderate amounts, except during pregnancy. Yet, in the U.S. there continues to be a lurking hesitancy in some social circles about its use for relaxation or recreational purposes. Some religious faiths specifically proscribe wine, which also makes it interesting that some other religious faiths include wine in their services. Beyond use of moderate amounts, it is clear that drinking wine excessively leads to inebriation and likely endangerment of self and others, just as with any alcohol-containing beverage. In view of this spectrum of opinion and in keeping with the fashion of the day, it seems appropriate at the outset of this Perspectives in Diabetes and in the spirit of full disclosure to state that my mother, a full-blooded Italian from Vinchiaturo in Molise, always told her sons that a day without wine is a day without sunshine. Being a Seattle resident, I relish a sunny day.

Figure 1

A Persian woman pouring a Persian red wine (mey) in the ancient world (17th century, Persia). From a wall painting inside the Chehel Sotoun palace, Isfahan, Iran. Photo by Zereshk.

Figure 2

A monk cellarer keeping track of the progress of his sacramental wine (probably white) in the Middle Ages (late 13th century, France). Source: From Li Livres dou Santé by Aldobrandino of Siena.

In preparing to write, I submitted three topics to PubMed: wine and diabetes, wine and mental health, and wine and social health. Note that I purposely did not include beer, liquor, or alcohol in my search. These searches garnered 265 references from 1963–2013, 58 references from 1967–2013, and 200 references from 1970–2013, respectively. After reading titles and abstracts, I excluded articles that were not peer reviewed or original work, did not specifically address drinking wine in moderate quantities in humans, or did not consider the impacts on health in people with diabetes or cardiovascular disease. I read the full manuscripts for the remainder. This Perspectives in Diabetes is organized to consider 1) epidemiologic and metabolic studies of the consequences of red wine drinking for people with type 2 diabetes (T2D), 2) the question of whether the health effects of wine are intrinsic to grapes or a function of the alcohol in wine, and 3) potential mechanism(s) of action for the beneficial effects of red wine.

Epidemiological Studies

French Paradox, Cardiovascular Disease, and T2D.

The term French paradox appears to have been popularized in the early 1990s by 60 Minutes, the CBS news magazine (1), and an epidemiological article in The Lancet (2), which point out that high intake of saturated fat is positively related to high mortality from coronary heart disease (CHD) except in France, where there is high wine consumption. This observation was made as early as 1979 by St. Leger et al. (3). CHD has important linkage to the issue of the effects of wine use on T2D because of the marked increased risk for CHD in people with T2D. A steady accumulation of articles reporting an association of moderate alcohol use and protection against CHD independent of T2D has appeared over the past few decades. At the current time, a PubMed search for the terms “alcohol” and “coronary heart disease” yields 11,980 citations. Rather than trying to digest this unwieldy number, I will concentrate on a more refined and more manageable group of reports focused on the impact of red wine drinking on the health of people with T2D (423).

Not so long ago, there was concern that drinking alcohol might raise blood glucose levels. Stampfer et al. (4) (Table 1) evaluated this hypothesis in 1988 by examining the status of 85,051 women participating in the Nurses’ Health Study who were 34–59 years of age in 1980 with no history of diabetes. A dietary questionnaire was used to gather information about the consumption of beer, wine, and liquor. Follow-up questionnaires were sent in 1982 and 1984. Ninety-eight percent responded to at least one follow-up. A total of 526 incident cases of T2D were confirmed. The risk of diabetes decreased with increasing moderate alcohol consumption. However, decreases in body weight were also associated with increasing alcohol use. The authors concluded that there was no support for the notion that moderate alcohol intake independent of weight loss increases or decreases the risk of T2D.

Table 1

Epidemiologic studies of the association between moderate alcohol consumption and the development of T2D

Four large epidemiological studies were reported from Australia (5), Sweden (6), a European consortium (7), and Denmark (8) in the past decade (Table 1). Hodge et al. (5) studied 36,527 adults aged 40–69 years at baseline, and incident cases of T2D were identified by questionnaire 4 years later. Eighty-seven and eighty-five percent of the women and men, respectively, responded to the questionnaires. Of women, 44% consumed no alcohol. Women with the highest consumption were younger, leaner, more active, and more educated, smoked more, and were less likely to have a family history of diabetes. Of men, 15% consumed no alcohol. Men with the highest consumption consumed less fat and were more likely to have gained weight. A total of 454 participants reported a diagnosis of diabetes after baseline. Of 397 who followed up with doctors, 76% were confirmed to have T2D. For women and men, after adjustment for BMI and waist-to-hip ratio, there were no associations between general alcohol consumption and a diagnosis of T2D. However, particularly pertinent for this Perspectives in Diabetes, wine was the most commonly consumed beverage (92% of total alcohol intake for women and 57% for men), and it was associated with lower risk of T2D for both women and men. Cullmann et al. (6) used oral glucose tolerance tests at baseline 8–10 years later to study the development of prediabetes or T2D in a group of 2,070 men and 3,058 women with normal glucose tolerance. The development of T2D was determined in this same group after enrichment with other subjects who had prediabetes, yielding groups of 2,217 men and 3,176 women with either normal glucose tolerance or prediabetes. Of the group with normal glucose tolerance, 105 men and 57 women developed T2D and 240 men and 174 women developed prediabetes. Data were examined before and after adjustment for confounding variables (age, BMI, tobacco use, physical activity, family history of diabetes, and education). Wine consumption by men was not associated with either prediabetes or T2D. In women with normal glucose tolerance at baseline, a reduced risk of prediabetes was seen in the high–wine intake group compared with occasional drinkers. Beulens et al. (7) investigated the association between alcohol consumption and T2D and determined whether it is modified by sex, BMI, and beverage type. They analyzed 455,680 subjects aged 35–70 years from the European Investigation into Cancer and Nutrition Consortium (26 centers in Denmark, France, Germany, Italy, the Netherlands, Spain, Sweden, and the U.K.). They were studied with an average follow-up of 9.9 years. The authors reported that moderate alcohol consumption was associated with a reduced risk of T2D; wine and fortified wine were most clearly associated with this reduced risk. Rasouli et al. (8) used data from the Nord-Trøndelag Health Survey to determine the incidence of T2D in 90,296 adults in relation to alcohol consumption, including quantity and frequency of drinking, type of drink, binge drinking, and alcohol use disorders. Consumption of wine rather than beer and spirits was associated with risk reduction of T2D.

Comparing the results of these epidemiology studies is difficult at best because they varied widely in design, size, definition of moderate alcohol drinking, and emphasis on weight change. Nonetheless, their message in common is uniform and very clear. Moderate drinking of alcohol was not associated with increased risk of T2D. Wine drinking specifically was associated with protection against the development of T2D. À votre santé!

Metabolic Studies

As we know, epidemiology is the study of associations and not ascertainment of cause-effect relationships. However, these associations have stimulated prospective metabolic research in humans. Dietary and metabolic approaches were used to ascertain whether consumption of alcohol and wine causes deterioration of carbohydrate tolerance (Table 2). Compared with epidemiologic studies, the metabolic studies I found were comprised of much smaller groups of subjects. Christiansen et al. (9) (Table 2) studied T2D subjects who consumed a light meal containing 300 mL tap water, dry white wine, sweet white wine with ethanol added, or dry white wine with glucose added. No deterioration of glycemic control was noted. Gin et al. (10) studied a group of T2D subjects, one-third of whom received red wine, tannic acid, or ethanol with their midday meal. Meals with wine or tannic acid (a component of wine) were associated with lower glucose levels after eating compared with meals with ethanol. The authors concluded that the beneficial effects of wine might not be mediated by alcohol. Bantle et al. (11) assessed the acute effects of wine in T2D subjects over 2 days during which they received 240 mL wine or grape juice with their evening meals. They also studied in random order the chronic effects of wine for 30 days and abstaining from alcohol for 30 days. Drinking wine had no effects on fasting glycemia. The authors concluded that people with T2D should not be discouraged from drinking wine in moderation. Napoli et al. (12) performed euglycemic-hyperinsulinemic clamps in T2D subjects before and after 2 weeks of red wine consumption (360 mL daily) and in T2D subjects who did not drink wine. Insulin-mediated whole-body glucose disposal improved by 43% after red wine consumption but did not change in the control group, suggesting that red wine attenuates insulin resistance. Ceriello et al. (13) explored the possibility that red wine consumption reduces oxidative stress produced by meals. T2D males in random order ingested a standard meal without wine, 300 mL red wine without a meal, or 300 mL red wine with the meal. Plasma total radical-trapping antioxidant parameter (TRAP) was used to evaluate plasma antioxidant capacity. TRAP decreased with the meal alone, increased with red wine alone, and did not change when red wine was combined with the meal. The authors concluded that the meal caused oxidative stress, that red wine reduced oxidative stress, and that wine with meals neutralized meal-induced oxidative stress. Marfella et al. (14) randomized subjects with diabetes who had sustained a first nonfatal myocardial infarction (MI) to receive or not receive a moderate daily amount of red wine for 1 year. At the end of a year, the group drinking red wine had lower levels of nitrosine, C-reactive protein (CRP), tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), and IL-8, and improved cardiac function. They concluded that moderate red wine accompanying meals may prevent complications after MI in T2D. Koivisto et al. (15) studied the effect of alcohol (1 g/kg as an aperitif before, wine during, and a drink after a meal) in T2D using mineral water as a comparator. They found that alcohol was associated with higher postprandial insulin levels with no effect on postprandial glucose levels but with slightly lower fasting glucose levels the next morning.

Table 2

Metabolic studies of the effects of moderate alcohol consumption on fasting glucose, glucose tolerance, and oxidative stress

Taken together, these metabolic studies strongly suggest that drinking moderate amounts of wine with meals does not cause adverse effects on glycemia and may improve several measures of carbohydrate and oxidative metabolism. Some of the information indicates that it is not the alcohol but something else in red wine that provides the beneficial effects. However, an important caveat is the possibility of hypoglycemia, especially in elderly people taking oral hypoglycemic drugs or insulin (16). It is well-known that use of alcohol can be hazardous, especially if used excessively, because of the adverse effects of alcohol on counterregulation (17) and symptom awareness of hypoglycemia, particularly at night while sleeping. This problem was recently revisited in an article by Turner et al. (18), who studied men with T1D who drank dry white wine (0.75 g/kg) or mineral water at 9:00 p.m. after eating a meal at 8:00 p.m. In the morning, fasting and postprandial glucose levels were significantly lower after consumption of wine the night before, and some subjects required treatment for hypoglycemia after breakfast the following morning. This finding was associated with reduced nocturnal growth hormone levels, which might have contributed to increased insulin sensitivity.

Benefits of Red Wine: Is It the Alcohol or the Grapes?

If, based on these epidemiologic and metabolic data, we accept the proposition that drinking red wine can be healthful, the next step is to ask what in red wine is responsible for its beneficial effects. Is it the alcohol, the grapes, or both? And if it is the grapes, what is in grapes that might be beneficial for people with or at risk for T2D?

To address these questions, Alberti-Fidanza et al. (19) examined the acute effects of lyophilized red wine after a meal on plasma oxygen radical absorbance capacity (ORAC), a measure of antioxidant capacity, in healthy subjects. The control was a meal without drinking wine afterward. The meal without wine decreased ORAC for 360 min, whereas the meal plus drinking red wine from which alcohol had been removed increased ORAC for 360 min. This suggests that meals decrease antioxidant capacity, whereas red wine ingredients in the absence of alcohol increase antioxidant capacity. Noguer et al. (20) studied whether red wine affects antioxidant enzyme activity in blood or blood products. Subjects consumed a low phenolic diet (to avoid interference from dietary polyphenols) while drinking or not drinking 300 mL red wine every day for a week. The low-polyphenol diet reduced enzyme activities for superoxide dismutase, catalase, and glutathione reductase; the same diet combined with red wine increased the activities of these enzymes. The authors posited that the increase in antioxidant enzyme activity was due to the polyphenolic composition of wine and not its alcohol content. Chiva-Blanch et al. (21) took this line of research important steps further by studying a much larger group of subjects and including the control of a nongrape alcoholic beverage. Men with high cardiovascular risk were randomized in a crossover trial during which all received red wine (272 mL, 30 g alcohol, 798 mg polyphenols/day), the equivalent amount of dealcoholized red wine (272 mL, 1.14 g alcohol, 733 mg polyphenols/day), and gin (100 mL, 30 g alcohol) for 4-week periods. Fasting glucose levels remained constant throughout the study. Compared with baseline values, fasting insulin levels decreased slightly but significantly during the periods of drinking red wine and dealcoholized red wine but not gin. The insulin data were used to calculate homeostasis model assessment of insulin resistance values, which led the authors to conclude that the nonalcoholic fraction of red wine contains polyphenols, which decrease insulin resistance. Nakamura et al. (24) studied subjects with T2D and nephropathy who were randomly assigned to drink 118 mL red wine or white wine daily for 6 months. A third group with T2D and nephropathy did not drink wine. Urinary 8-hydroxydeoxyguanosine was measured as a marker of oxidative stress, and urinary liver-type fatty acid–binding protein was measured as a marker for deteriorating renal function. The subjects who drank red wine had a reduction of these two markers at 3 and 6 months of study, whereas the group drinking white wine did not. The authors concluded that the renoprotective effect of red wine may be due in part to its ability to reduce oxidative stress.

These four studies point toward the beneficial effects of red wine being provided not by its alcoholic content but by something else. Is it the polyphenolic acid content? If so, which polyphenols, and are they modulators of oxidative stress? And where in the grape are polyphenols found? However, before we move on to more information about polyphenols, the definition of moderate drinking and how it relates to a glassful of wine needs to be addressed. Obviously, the size of the glass is a huge variable. Some red wine glasses can hold an entire bottleful of wine! The majority of studies define a moderate quantity as 300 cc or 2 glasses of red wine. This is a little less than one-half of a conventional 750-cc bottle of wine, which makes one wonder why bottle makers chose five as the number glasses a bottle can fill. However, this no doubt is the origin of the well-known gambit in which one diverts one’s dinner partner’s attention elsewhere in the restaurant while cadging the remaining 150 cc.

Grapes Have Skin in the Game of Diabetes Health Care

What are polyphenolic acid and polyphenols? Polyphenols are found in plants, especially grapes and berries, and can be defined as macromolecules generally containing <12 phenolic hydroxyl groups with five to six aromatic rings per 1,000 daltons. Polyphenols sport exotic names, such as (in no particular order) phytoalexins, gallic acid, syningic acid, protocatechuric acid, caffeic acid, ferulic acid, stilbenes, flavonoids, quercetin, myricetin, kamempferol, rutin, catechin, delphinidin, malvidin, etc. Processing these names is challenging enough to tempt one to reach for a glass of sangiovese. One of these names that you will readily recognize is resveratrol of the stilbene group, commonly discussed in newspapers and on television shows. According to Wikipedia, this less-than-intuitive name comes from the fact that resveratrol is a resorcinol derivative coming from a Veratrum species (25). If you are in the mood to read an incredibly large amount of information about plants and polyphenols, I refer you to several reviews I thought were encyclopedic (26) or succinct (1,27). Many mechanisms of actions of polyphenols have been proposed, but the one that recurrently appears in the scientific literature is the role of polyphenols as antioxidants. Paganga et al. (22) reported that the antioxidant activities of 1 glass (150 mL) red wine is equivalent to 2 cups of tea, 3.5 glasses of black currant juice, 3.5 (500 ml) glasses of beer, 4 apples, 5 portions of onion, 5.5 portions of eggplant, 7 glasses of orange juice, 12 glasses of white wine, and 20 glasses of apple juice. Hence, the often-heard comment that red wine provides excellent protection against oxidative stress.

Tomé-Carneiro et al. (28) studied subjects undergoing primary prevention of cardiovascular disease in a triple-blinded, randomized, parallel, dose-response, placebo-controlled, 1-year follow-up trial of placebo versus grape supplement containing 8 mg resveratrol (a polyphenol) versus grape supplement lacking resveratrol. They observed at 1 year that the resveratrol-rich grape supplement decreased high-sensitivity CRP, TNF-α, plasminogen activator inhibitor type 1, and IL-6–to–IL-10 ratio. They suggested that grape-derived resveratrol as a dietary intervention could complement methods conventionally used in the primary prevention of cardiovascular disease. However, one should not assume that resveratrol is the only or even the principal antioxidant in red wine. German and Walzem (1) reported that the total phenolic acids and polyphenols in red wine is 900–2,500 mg/L. Of these, the concentration of flavonoids is 750–1,060 mg/L and consists of flavonols, flavanols, and anthrocyanins. The highest polyphenol concentrations in these three subgroups are, respectively, quercetin, procyanidins, and malvidin 3-monoglucaside. The nonflavonoids have a concentration of 240–500 mg/L and include hydroxybenzoic acids, hydroxycinnamic acids, and stilbene. The highest concentrations in these three subgroups, respectively, are protocatechuric acid, cis-/trans-caftaric acid, and trans-resveratrol. But resveratrol is at the bottom of the entire list in terms of concentration in red wine . . . a pitiful 1.0 mg/L (1,23), which makes one wonder why it gets all the attention. One review (27) suggests that it may be due to information published in the last decade suggesting that resveratrol is an inhibitor of cyclooxygenase and an activator of sirtuin enzymes. Regarding this point, Poulsen et al. (29) reported a study in which 24 obese men were randomized in a double-blinded, placebo-controlled fashion to receive either high-dose oral resveratrol or placebo for 4 weeks. They found no significant effect of resveratrol on a host of physical and metabolic parameters. However, Timmers et al. (30) and Brasnyó et al. (31) reported that resveratrol supplementation was associated with an improved homeostasis model assessment index, suggesting increased insulin sensitivity. Bhatt et al. (32) studied 62 subjects with T2D subjects using oral hypoglycemic agents who were randomized into an intervention group receiving resveratrol and a control group not receiving resveratrol. The former group experienced a slight but statistically significant decrease in mean HbA1c level. Rytter et al. (33) reported a study in which 47 subjects with T2D were randomized in a double-blind, placebo-controlled study of the effects of antioxidant supplements (unfortunately, resveratrol was not included) on blood levels of glucose and HbA1c and found no differences. Importantly, they also observed no effects of the supplements on markers of oxidative stress, which raises the question of whether the concentrations of supplements they used were sufficiently high to have effects on metabolic parameters. I did not find reports of comprehensive biochemical comparisons of the antioxidant potencies of the various polyphenolic acids found in red wine, which could be a more important factor than their total concentrations.

Finally, where in red grapes are polyphenols found? Here’s the skinny: Wine is composed of yeast fermentation products of the “must,” which is the juice pressed from the grape, the fruit of the genus Vitis. Fermentation produces a variety of chemical changes in the must, so wine is not a simple grape juice with ethanol added. Red wine is left in contact with the must for longer periods of time than white wine. The longer the must is exposed to the skins, seeds, and stems of the grapes, the more polyphenols are extracted into the juice and the more antioxidant activity is contained in the wine. This no doubt is why many wine drinkers believe choosing red is a must.

What Does Protection Against Oxidative Stress Have to Do With Protection Against Diabetes?

Oxidative stress is a topic that chemists, especially food chemists, love. They find moieties like O2, OH, H2O2, and ONOO fascinating, if not endearing. They teach us that reactive oxygen species (ROS), such as the ones just listed, are two-edged swords. In small concentrations, ROS are good things because they enhance physiological and cellular function. In excess concentrations, however, they cause havoc in cells. This is termed oxidative stress, which, as its name implies, is not a good thing. The complications of T2D have long been associated with chronic oxidative stress. Excessive circulating glucose concentrations spilling over into a variety of metabolic pathways form increasing amounts of ROS, which in excess destroy membranes and compartments in the tissues in which they are formed. This risky state of affairs is particularly true in the β-cell, which for unknown reasons does not contain two major antioxidants (catalase and glutathione peroxidase) (34,35). Thus, the β-cell is left without the major protection that other tissues have against accumulation of H2O2 and intracellular peroxides. β-Cells do contain superoxide dismutases (SODs), but this doesn’t help much because SODs, although classified as antioxidant enzymes, catabolize superoxide into the oxidant H2O2, thereby acting like a pro-oxidant enzyme.

In T2D, this all comes down to the issue of excess glycemia bathing tissues and creating ROS, which in excess can damage organs, including retina, kidney, nerves, vascular tissue, and β-cells (36,37). The pathogenesis of T2D by consensus originates as a polygenic disease amplified by epigenetic influences, such as environmental oxidants and obesity with associated insulin resistance, to damage β-cells, thereby causing impaired glucose tolerance and early diabetes. The glucose toxicity hypothesis holds that continual exposure to high postprandial glucose levels over time steadily generates excess ROS that constantly nibble away at cells and their normal function (38). Red wine comes into the picture of T2D because it contains antioxidants that can be used as an ancillary buffer against the oxidative stress caused by the ingestion of a meal or environmental toxins.


From my perspective, the articles that I have read strongly suggest that drinking red wine in moderation has no adverse effects on people with T2D. Moreover, red wine may have a preventive effect for people at risk for T2D. Overall, the metabolic studies of humans are consistent with the epidemiologic data, i.e., no harmful effects and possibly beneficial effects of red wine. Interestingly, the skin of the grape, rather than the alcohol or juice in red wine, is the source of polyphenols, and their antioxidant effects may play a role in the beneficial effects of red wine. Red wine has 12-fold higher levels of polyphenols than white wine and so has a clear edge on antioxidant potency. The actual mechanism of action of red wine has been popularly suggested to involve resveratrol, but the contention that this polyphenol is exclusively involved does not stand up to scientific evidence. This intriguing health-related area is ripe for clinical and basic research efforts designed to more precisely identify which of the many polyphenolic acids in red wine may be playing a role in the protective mechanism(s) involved and, beyond that, the full elucidation of these mechanism(s). You may want to get your own skin in this game. While thinking about it, sit back and relax with two glassfuls of your favorite red wine without fear. It may be that one significant mechanism of wine’s beneficial effects is just that…relaxing with good friends during a savory meal with a great cabernet.

Article Information

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

  • Received August 26, 2013.
  • Accepted October 15, 2013.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. See for details.


| Table of Contents