Critique 174: Two-year clinical trial of wine administration in diabetics who were previous abstainers or very-light drinkers — 19 October 2015

Gepner Y, Golan R, Harman-Boehm I,  . . .  Stampfer MJ, Shai I, et al.  Effects of Initiating Moderate Alcohol Intake on Cardiometabolic Risk in Adults With Type 2 Diabetes.  A 2-Year Randomized, Controlled Trial.  Ann Intern Med 2015; pre-publication.  doi:10.7326/M14-1650 

Authors’ Abstract

Background: Recommendations for moderate alcohol consumption remain controversial, particularly in type 2 diabetes mellitus (T2DM). Long-term randomized, controlled trials (RCTs) are lacking.

Objective: To assess cardiometabolic effects of initiating moderate alcohol intake in persons with T2DM and whether the type of wine matters.

Design: 2-year RCT (CASCADE [CArdiovaSCulAr Diabetes & Ethanol] trial. Clinical NCT00784433)

Setting: Ben-Gurion University of the Negev–Soroka Medical Center and Nuclear Research Center Negev, Israel.

Patients: Alcohol-abstaining adults with well-controlled T2DM.

Intervention: Patients were randomly assigned to 150 mL of mineral water, white wine, or red wine with dinner for 2 years.  Wines and mineral water were provided.  All groups followed a Mediterranean diet without caloric restriction.

Measurements: Primary outcomes were lipid and glycemic control profiles. Genetic measurements were done, and patients were followed for blood pressure, liver biomarkers, medication use, symptoms, and quality of life.

Results: Of the 224 patients who were randomly assigned, 94% had follow-up data at 1 year and 87% at 2 years.  In addition to the changes in the water group (Mediterranean diet only), red wine significantly increased high-density lipoprotein cholesterol (HDL-C) level by 0.05 mmol/L (2.0 mg/dL) (95% CI, 0.04 to 0.06 mmol/L [1.6 to 2.2 mg/dL]; P < 0.001) and apolipoprotein(a)1 level by 0.03 g/L (CI, 0.01 to 0.06 g/L; P = 0.05) and decreased the total cholesterol–HDL-C ratio by 0.27 (CI, 0.52 to 0.01; P = 0.039). Only slow ethanol metabolizers (alcohol dehydrogenase alleles [ADH1B*1] carriers) significantly benefited from the effect of both wines on glycemic control (fasting plasma glucose, homeostatic model assessment of insulin resistance, and hemoglobin A1c) compared with fast ethanol metabolizers (persons homozygous for ADH1B*2).  Across the 3 groups, no material differences were identified in blood pressure, adiposity, liver function, drug therapy, symptoms, or quality of life, except that sleep quality improved in both wine groups compared with the water group (P = 0.040). Overall, compared with the changes in the water group, red wine further reduced the number of components of the metabolic syndrome by 0.34 (CI, – 0.68 to – 0.001; P = 0.049).

Limitation: Participants were not blinded to treatment allocation.

Conclusion: This long-term RCT suggests that initiating moderate wine intake, especially red wine, among well-controlled diabetics as part of a healthy diet is apparently safe and modestly decreases cardiometabolic risk.  The genetic interactions suggest that ethanol plays an important role in glucose metabolism, and red wine’s effects also involve nonalcoholic constituents.

Forum Comments

Background:  Observational epidemiologic studies have consistently found that moderate drinkers are at lower risk of cardiovascular disease (CVD), and that such alcohol consumption also lowers the risk of diabetes as well as CVD among diabetics.  For example, a meta-analysis by Koppes and associates indicates that the risk of coronary heart disease is 34% to 55% lower among diabetics who are moderate drinkers than in those who consume no alcohol.  The Kaiser Permanente group in California reported that among more than 38,000 diabetics, those who consumed alcohol had evidence of considerably better control of their diabetes than did nondrinkers (Ahmed et al); these authors concluded that this finding “supports current clinical guidelines for moderate levels of alcohol consumption among diabetes patients.”

Alcohol has been shown to affect both the macrovascular sequelae of diabetes and the microvascular complications of diabetes.  In a report from the EURODIAB Prospective Complications Study, involving the follow-up of 3250 type 1 diabetic patients from 16 different European countries, Beulens and colleagues found a lower occurrence of retinopathy, neuropathy, and nephropathy among those individuals who consumed alcohol moderately in comparison with nondrinkers.  The association was strongest among wine drinkers, and to some extent among beer drinkers, but was not seen among those consuming spirits

Previous clinical trials:  There have been a few randomized clinical trials (RCT) on this topic.  In 2007, Shai et al reported on a RCT in which 150 ml/day of wine (versus non-alcoholic beer) was given for 3 months to 109 previously abstaining patients with diabetes.  Patients in the alcohol group showed a significant lowering of fasting blood glucose; those with higher baseline A1C levels had the largest reductions.  No significant changes were observed in the levels of bilirubin, alkaline phosphatase, alanine aminotransferase, or aspartate aminotransferase, and no notable adverse effects were reported.  However, participants in the alcohol group reported an improvement in the ability to fall asleep (P < 0.001).

A cross-over trial by Joosten et al in 2008 showed that 6 weeks of the administration of 2 drinks/day of white wine (versus grape juice) improved insulin sensitivity and lipid profile among 36 healthy, post-menopausal women.  Both fasting triacylglycerol (8.2%; p = 0.04) and LDL-cholesterol levels (7.8%; p < 0.0001) decreased, whereas HDL-cholesterol increased (7.0%; p < 0.0001) after prolonged moderate alcohol intake.  No notable adverse effects were reported.

Specific comments on the present study:  Forum members were very glad to see the present paper describing the results of a relatively large, well-planned, and well-executed RCT among diabetics of the effects of alcohol consumption on cardiometabolic factors.  As stated by Forum member Finkel, “It is so invigorating to behold a randomized, prospective, therapeutic trial, so rare in our little besieged world of alcohol effects.”

Reviewer Skovenborg noted a number of important aspects of the study, stating that “There were many innovative and clever design-features of the study:

(1) The participants were abstainers/very light drinkers with the logical conclusion that any effects from the two years intervention had to be “de novo” effects and not the results of previous moderate wine consumption.

(2) The exclusion criteria were both exhaustive and sensible.

(3) The length of the intervention, two years, is unique concerning a lifestyle factor like wine and will propose a new standard for what it is possible to achieve in RTCs of lifestyle factors.

(4) The beverage adherence was very good and better than seen in many RCTs of new medications.

(5) The genotyping of participants made a closer look at alcohol effects possible by comparison of the results between slow and fast ethanol metabolizers.

(6) The outcomes were modest but bear the RCT quality stamp.

(7) The intervention was safe: no increase of waist circumference, nor blood pressure or liver enzymes – and the benefit of the modest wine intake on sleep quality is noteworthy.”

Forum member Thelle agreed that “This is a good study, well planned and executed.  It is larger than that of Hansen et al in 2005 that only assessed red wine (with slightly larger effects on HDL-C).  I wonder whether diabetics are the ideal subjects for such a study.  One may consider diabetes as a metabolic breakdown situation from which some may recover, whereas others remain there.  This may explain why the results are relatively modest.  However, I cannot see any alternative research avenues to the parallel arm study design.”  Reviewer Ellison stated: “In my opinion, diabetics, because of their greatly increased risk of cardiovascular diseases, are indeed a good group in which to test alcohol’s effects.”

Reviewer Zhang noted that the subjects were not complete “abstainers” prior to the study, as a certain percentage (proportion not given) consumed small amounts of alcohol.  Zhang added: “Based on the results the authors provided, the average number of alcohol beverage drinks at baseline was approximately one/week for the whole study sample.  Anyone who reported more than one drink/week was excluded from the study.  At any rate, it is a well-conducted study, the type of RCT we have been looking for in alcohol research.  Still, it is critical to define precisely the previous drinking habits of the study cohort, as it could affect the potential generalizability of the study findings.”

While some suggested that grape juice (without alcohol) may have been another intervention to test, Reviewer Waterhouse stated: “The comment about the missing grape juice treatment is most surprising.  Grape juice is very high in sugar (150 mL would have contained 22-25 g sugar).  I am not sure this would have been a wise treatment for diabetic subjects.”   He added: “It is great to see a randomized control trial with wine.”  Reviewer Lanzmann-Petithory agreed: “Grape juice would have been too rich in sugar, especially glucose, to serve as a treatment arm for the study.”

It was noted by reviewers that the group randomized to red wine had a tendency for higher HDL-cholesterol and lower triglycerides than other groups at baseline, even though their average intake of alcohol prior to the study tended to be lower and the apolipoprotein (a) levels were identical in all groups.  While this is unlikely to have had any effects on the results of the study, it might warrant further investigation into how to identify subjects who might respond well to intervention with wine.

Forum member de Gaetano commented:  “I understand that in the present study women and men were given the same amount of wine to drink for practical and economical reasons, but it would have been better to advise men to consume one or two drinks of wine per day, and to reduce the amount in women.  Since no clinical outcomes were included in the end point of this study (two years follow up were indeed too short a period for that), I hope that they will continue a longer follow up of the participants in their trial.  On the other hand I believe that even a much shorter trial would have shown the metabolic effects observed here.  For example, in an intervention trial by Estruch, me, and our colleagues about  20 years ago, we found in 40 apparently healthy male volunteers, in a cross-over design, that one month’s consumption of 30 g of wine a day induced metabolic effects comparable to that now reported in the present paper.  The results of that and other studies by our group on the effects of alcohol and of wine polyphenolics are described in Badia et al, Estruch et al 2004, and Estruch et al 2009.  The main interest of the present paper is that it shows that diabetic patients too could benefit  – from a metabolic viewpoint – from a moderate daily dose of wine.”

De Gaetano continued: “In addition, all participants in the present trial were advised to adhere to a Mediterranean Diet.  This might have been of importance in contributing to the observed effects of wine.  It may be of interest to recall that one of our recent observational studies (Bonaccio et al) revealed that the component of the diet with the largest impact on all-cause mortality was wine consumption.”  In an earlier critique of that paper by our Forum, we noted that “The authors’ calculations estimated that, in descending order of importance, moderate alcohol consumption (associated with a reduction of 14.7% in the protection against mortality when it was removed from the total score), cereal intake (12.2% reduction), ratio of monounsaturated to saturated fats (5.8%), and consumption of vegetables (5.8%), fruits and nuts (5.2%) and fish (5.0%) lowered mortality risk.  Lower intakes of dairy products (reduction of 13.4%) and meat and meat products (3.4%) were also associated with lower mortality (”

Forum member Estruch had a number of comments: “In order to achieve the highest level of scientific evidence, large randomized intervention trials are needed; this paper should be the first of a large streaming of this type of trials.  Although the main results are as expected, it is important to underline the relationship between some genetic variants and the outcome variables analyzed (mainly those related to lipid profile).  Accordingly, these effects should be attributed to the ethanol contained in wine.  An additional arm with an alcoholic beverage without polyphenols would help to interpret the results.  Additional effects of red wine should be attributed to non-alcoholic compounds (polyphenols?).  An analysis of changes in polyphenol profile in blood and urine before and after interventions should be performed and an attempt made to correlate changes in polyphenols (due to red wine intake) and clinical and analytical variables.  In this respect, analysis of biomarkers of interventions adds even more quality to this randomized controlled trial.  A final comment on the dose of alcohol administered: 150 ml/d of red/white wine is an adequate dose to test for women, but not for men (who should consume perhaps around 300 ml/d.  With a higher dose, the results of this trial might have been even better.”

The key results of this study were that the administration of 150 ml/day of red wine favorably affected many cardiometabolic factors, and especially led to an increase in HDL-cholesterol and apolipoprotein (a).  A similar amount of white wine was also beneficial, and specifically improved measures of glycemic control.  Forum members agreed with the authors’ conclusion that “initiating moderate wine intake, especially red wine, among well-controlled diabetics as part of a healthy diet is apparently safe and modestly decreases cardiometabolic risk.”

Other effects of the wine intervention:  Reviewer Lanzmann-Petithory commented: “The effect of wine on sleep quality has previously been described in an old intervention trial in mental institutions in which giving wine to elderly subjects resulted in a dramatic reduction in the use of chloral hydrate to induce sleep (Mishara & Kastenbaum).”  She also commented on other important aspects of this study: “This study could open the minds in the future to facilitate obtaining the permission from ethical committees for trials involving alcohol.  This kind of intervention could be applied also to the elderly in retirement homes or geriatric hospitals because they are at high risk of cardiovascular disease; at the moment, at least in my hospital, alcohol is completely forbidden.  (Perhaps in the future wine could be reimbursed by insurance companies when administered to certain patients with diabetes!?)”  Reviewer Finkel added: “A few (too few) hospitals in the US wisely offer wine to those patients thought suitable, and to good effect.”

Should certain non-drinkers be advised to begin alcohol consumption?  Forum member de Gaetano had some problems with advising non-drinkers to consume alcohol for their health: “We believe that no total abstainer should be asked to start drinking for health reason, and our Ethical Committee would not have accepted the inclusion criteria of this otherwise excellent trial.”  Reviewer Ellison disagreed: “For older adults with diabetes, their key adverse health outcomes are retinopathy, neuropathy, kidney disease, and cardiovascular disease, with the latter being the leading cause of death.  Unless there is a contraindication to alcohol use (e.g., previous abuse, a small number of medical conditions, religious prohibitions), a physician may well decide that certain individuals would benefit from being advised to include a glass of wine with dinner every evening; the risk would be extremely small, and the health benefits could be large.”

References from Forum critique

Ahmed AT, Karter AJ, Warton EM, et al: The relationship between alcohol consumption and glycemic control among patients with diabetes: the Kaiser Permanente Northern California Diabetes Registry.  J Gen Intern Med 2008;23:275-282.

Badía E, Sacanella E, Fernández-Solá J, Nicolás JM, Antúnez E, Rotilio D, de Gaetano G, Urbano-Márquez A, Estruch R.  Decreased tumor necrosis factor-induced adhesion of human monocytes to endothelial cells after moderate alcohol consumption.  Am J Clin Nutr 2004;80:225-230.

Beulens JW, Kruidhof JS, Grobbee DE, et al: Alcohol consumption and risk of microvascular complications in type 1 diabetes patients: the EURODIAB Prospective Complications Study.  Diabetologia 2008;51:1631-1838.

Bonaccio M, Di Castelnuovo A, Costanzo S, Persichillo M, De Curtis A, Donati MB, de Gaetano G, Iacoviello L, on behalf of the MOLI-SANI StudyInvestigators.  Adherence to the traditional Mediterranean diet and mortality in subjects with diabetes.  Prospective results from the MOLI-SANI study.  European Journal of Preventive Cardiology 2015.  Pre-publication.  DOI: 10.1177/2047487315569409.

Estruch R, Sacanella E, Badia E, Antúnez E, Nicolás JM, Fernández-Solá J, Rotilio D, de Gaetano G, Rubin E, Urbano-Márquez A.  Different effects of red wine and gin consumption on inflammatory biomarkers of atherosclerosis: a prospective randomized crossover trial. Effects of wine on inflammatory markers.  Atherosclerosis 2004;175:117-123.

Estruch R, Sacanella E, Mota F, Chiva-Blanch G, Antúnez E, Casals E, Deulofeu R, Rotilio D, Andres-Lacueva C, Lamuela-Raventos RM, de Gaetano G, Urbano-Marquez A.  Moderate consumption of red wine, but not gin, decreases erythrocyte superoxide dismutase activity: a randomised cross-over trial.  Nutr Metab Cardiovasc Dis 2011;21:46-53. doi: 10.1016/j.numecd.2009.07.006.

Hansen AS, Marckmann P, Dragsted LO, Finné Nielsen IL, Nielsen SE, Grønbaek M.  Effect of red wine and red grape extract on blood lipids, haemostatic factors, and other risk factors for cardiovascular disease.  Eur J Clin Nutr 2005;59:449-455.

Joosten MM, Beulens JWJ, Kersten S, Hendriks HFJ. Moderate alcohol consumption increases insulin sensitivity and ADIPOQ expression in postmenopausal women: a randomised, crossover trial.  Diabetologia 2008;51:1375–1381.

Koppes LL, Dekker JM, Hendriks HF, et al: Meta-analysis of the relationship between alcohol consumption and coronary heart disease and mortality in type 2 diabetic patients. Diabetologia 2006;49:648-652.

Mishara BL, Kastenbaum R.  Wine in the treatment of long-term geriatric patients in mental institutions.  J Am Geriatr Soc 1974;22:88-94.

Shai I, Wainstein J, Harman-Boehm I, Raz I, Fraser D, Rudich A, Stampfer MJ.  Glycemic Effects of Moderate Alcohol Intake Among Patients With Type 2 Diabetes. A multicenter, randomized, clinical intervention trial.  Diabetes Care 2007;30:3011-3016.

Forum Summary

Observational epidemiologic studies have consistently found that moderate drinkers are at lower risk of cardiovascular disease (CVD); such alcohol consumption also lowers both the risk of diabetes as well as of CVD among diabetics.  However, there have been few clinical trials of the administration of alcohol or wine among diabetics, and most have been for relatively short periods of time.  Forum members welcomed the publication of the results of the present study, and consider that it provides valuable information on the relation between wine and cardiometabolic risk.

In this study, among 224 subjects with diabetes aged 40 to 75 years who were abstainers or very light drinkers (previously reported none or < 1 drink/week), the investigators carried out a randomized clinical trial of the effects of wine on cardiometabolic factors.  The subjects were randomly assigned to 150 mL of mineral water, white wine, or red wine with dinner for 2 years; all groups followed a Mediterranean diet without caloric restriction.  The study was described to subjects as a dietary intervention, and there were group sessions led by clinical dietitians each month for the first three months and then at 3-monthly intervals thereafter.  The focus was on the Mediterranean diet, and wine was not discussed at these meetings.  There was excellent compliance with the provided beverages (>80%) and excellent continued participation of subjects (94% for one year and 87% for the full two years).

The key results of this study were that the administration of 150 ml of red wine with dinner each evening, in comparison with mineral water, favorably affected many cardiometabolic factors, and especially led to an increase in HDL-cholesterol and apolipoprotein (a).  The administration of similar amounts of white wine was also beneficial, and specifically improved measures of glycemic control.

Further, in 203 of the subjects, effects according to the genetic factors determining ADH were evaluated.  In the study, 35.6% of subjects were homozygous for the polymorphism for “slow ethanol metabolism,” 21.3% were homozygous for that for “fast ethanol metabolism,” and 43% were heterozygous.  Forum members agreed with the summary statement of the authors: “We found that diabetic patients who were slow alcohol metabolizers had improved glycemic control by initiating moderate wine consumption, which suggests that alcohol may play a role in glucose metabolism.  In contrast, diabetic patients who were fast ethanol metabolizers benefited the most from the wine-induced BP-lowering effect, which suggests a mediatory role for ethanol metabolites.”

The study has important implications for advice to be given to diabetics.  Most, but not all, Forum members agreed with the authors’ conclusion that “Initiating moderate wine intake, especially red wine, among well-controlled diabetics as part of a healthy diet is apparently safe and modestly decreases cardiometabolic risk.”


Comments on this critique by the International Scientific Forum on Alcohol Research were provided by the following members:

Giovanni de Gaetano, MD, PhD, Department of Epidemiology and Prevention, IRCCS Istituto Neurologico Mediterraneo NEUROMED, Pozzilli, Italy

R. Curtis Ellison, MD, Professor of Medicine & Public Health, Boston University School of Medicine, Boston, MA, USA

Ramon Estruch, MD, PhD.  Associate Professor of Medicine, University of Barcelona, Spain

Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA

Dominique Lanzmann-Petithory,MD, PhD, Nutrition/Cardiology, Praticien Hospitalier Hôpital Emile Roux, Paris, France

Fulvio Mattivi, MSc, Head of the Department of Food Quality and Nutrition, Research and Innovation Centre, Fondazione Edmund Mach, in San Michele all’Adige, Italy

Erik Skovenborg, MD, specialized in family medicine, member of the Scandinavian Medical Alcohol Board, Aarhus, Denmark

Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway

Dag S. Thelle, MD, PhD, Senior Professor of Cardiovascular Epidemiology and Prevention, University of Gothenburg, Sweden; Senior Professor of Quantitative Medicine at the University of Oslo, Norway

Fulvio Ursini, MD, Dept. of Biological Chemistry, Universityof Padova, Padova, Italy

Andrew L. Waterhouse, PhD, Department of Viticulture and Enology, University of California, Davis, USA

Yuqing Zhang, MD, DSc, Clinical Epidemiology, Boston University School of Medicine, Boston, MA, USA