Critique 172: Alcohol and the risk of developing diabetes — 7 October 2015
Knott C, Bell S, Britton A. Alcohol Consumption and the Risk of Type 2 Diabetes: A Systematic Review and Dose-Response Meta-analysis of More Than 1.9 Million Individuals From 38 Observational Studies. Diabetes Care 2015;38:1804–1812 | DOI: 10.2337/dc15-0710
OBJECTIVE: Observational studies indicate that moderate levels of alcohol consumption may reduce the risk of type 2 diabetes. In addition to providing an updated summary of the existing literature, this meta-analysis explored whether reductions in risk may be the product of misclassification bias.
RESEARCH DESIGN AND METHODS: A systematic search was undertaken, identifying studies that reported a temporal association between alcohol consumption and the risk of type 2 diabetes. No restrictions were placed upon the language or date of publication. Non-English publications were, where necessary, translated using online translation tools. Models were constructed using fractional polynomial regression to determine the best-fitting dose-response relationship between alcohol intake and type 2 diabetes, with a priori testing of sex and referent group interactions.
RESULTS: Thirty-eight studies met the selection criteria, representing 1,902,605 participants and 125,926 cases of type 2 diabetes. A conventional noncurrent drinking category was reported by 33 studies, while five reported a never-drinking category. Relative to combined abstainers, reductions in the risk of type 2 diabetes were present at all levels of alcohol intake <63 g/day, with risks increasing above this threshold. Peak risk reduction was present between 10–14 g/day at an 18% decrease in hazards. Stratification of available data revealed that reductions in risk may be specific to women only and absent in studies that adopted a never-drinking abstention category or sampled an Asian population region.
CONCLUSIONS: Reductions in risk among moderate alcohol drinkers may be confined to women and non-Asian populations. Although based on a minority of studies, there is also the possibility that reductions in risk may have been overestimated by studies using a referent group contaminated by less healthy former drinkers.
Prospective cohort studies for decades have tended to show that the risk of developing Type II diabetes mellitus (DM) is reduced for moderate drinkers in comparison with non-drinkers (Stampfer et al, Perry et al). In previous meta-analyses, the risk of DM among moderate drinkers has usually been about 30% lower than the risk among abstainers (Howard et al, Koppes et al). A more recent meta-analysis by Baliunas et al reported that “For women, the protective effect was greatest at the 24 g/day level, with a risk reduction of 40% compared with lifetime abstainers (95% CI 0.52– 0.69). Alcohol consumption remained protective until just under 50 g/day. For men, the protective effect of alcohol consumption was greatest at 22 g/day, with the risk of diabetes being 0.87 times that of lifetime abstainers (95% CI 0.76 –1.00), and remained protective until consumption of ~60 g/day.”
The present study carried out a multi-language search for studies on alcohol and DM and conducted a meta-analysis involving almost two million subjects from a final group of 38 cohort, case-cohort, case-control, or nested case-control studies. Thirty-three of these studies compared drinkers with non-drinkers (over varying periods of time), while 5 compared drinkers with never drinkers.
The main conclusions of the authors are that only females may show a significant inverse association between alcohol consumption and the risk of diabetes, and that previous studies may have overestimated the reduction in risk from moderate drinking.
Specific comments on paper by Forum members: Reviewer Ellison had concerns about the analyses and the conclusions of the authors. “It is apparent from the Supplementary Data presented, but not emphasized in the text, that there were huge differences between the associations with alcohol for Asian and non-Asian populations. The estimated relationship between alcohol intake and diabetes for non-Asians shows the u-shaped curve generally reported in the literature; for Asians, however, the curve shows only an increase in risk for moderate drinking. It appears that we have a similar problem with trying to combine very divergent data into one equation here as we did with the previous paper we reviewed by Smyth et al (www.bu.edu/alcohol-forum/critique-171). When the effects among Asians are almost the exact opposite of those in non-Asians, is it appropriate to combine the data and try to present a single overall result applicable to all?
“There are always problems when the basic study data have a marked degree of heterogeneity; as the authors state, ‘Heterogeneity between sampled studies was high, complicating interpretation.’ Other problems included the lack of data on the pattern of drinking (regular versus binge), type of beverage consumed, or potential changes in alcohol intake over time, as well as the inclusion of case-control and cohort data in the same analysis.”
Forum members consider it unfortunate that 86% of the males in their main analyses were Asian, which would apparently make it difficult for the authors to make conclusions that there is “no effect among males,” especially when being applied to western populations. Thus, while the analyses were all carried out appropriately, attempting to present advice regarding alcohol and diabetes that would apply globally, as done by the authors, is somewhat precarious.
Forum member Barrett-Connor stated: “I too am concerned about the differences by ethnicity and also about the sex differences. The differences in Asians could be explained by their totally different diets with higher fish oils (presumably more protective than diets of other ethnicities). I am more concerned about the total omission of controlling for tobacco use by sex and age and ethnicity; in fact I didn’t see tobacco as a covariate in any of the analyses. Surely it must have been reported in some of the studies included.” Others pointed out the large differences in BMI between most Asian populations and western ones; obesity is a key factor in the development of type II diabetes.
Reviewer Finkel commented: “The observations of other Forum members square with my own general view of meta-analyses, and of this one in particular. At best, a meta-study can give clues to what might be true or untrue, leading to performance of a ‘real’ study. At worst: Would you like to live in a house built of 1.9 million dissimilar bricks?” Added member Van Velden: “There are just too many environmental influences in these diverse ethnic groups that were not taken into consideration. We cannot assume that all people are equal.”
Forum member de Gaetano was also concerned about the predominance of Asians in these analyses: “The curves of Asians and of males are very similar, which might imply that their message is: male Asians are apparently not protected by alcohol against diabetes. I could not find whether the five studies that included in the control group lifetime abstainers only were from Asia and how many males were included. I would not be surprised if these 5 studies mainly included male Asians. In our experience (Di Castelnuovo et al 2002; Di Castelnuovo et al 2006) the benefit of alcohol against CV risk and total mortality was only marginally affected by differences in the selection of control groups.”
Failure to evaluate effects according to type of beverage: Reviewer Mattivi noted: “I am rather concerned about the lack of separation between alcohol source and pattern of drinking. In particular, red wine, which is the result of the long hydroalcoholic maceration of grape pomace in the fermenting fruit juice, contains several phytochemicals which may act on diabetes via known biochemical mechanisms. Just to mention a few, inhibition of salivary, intestinal and pancreatic enzymes, PPARgamma ligand-binding activity, modulation of oxidative stress, etc. It is expected that a meta-analysis of beverage-specific studies could in future shed light on the putative role of the non-alcohol fraction of red wine, not unexpected in light of its content of bioactive compounds such as tannins, anthocyanins, stilbenes, oleanolic acid, etc.” Added Forum member Ursini: “Wine is much more than an ethanol solution.”
Reviewer Estruch stated: “Meta-analysis provides the highest level of scientific evidence in evidence-based medicine only when randomized clinical trials are included in the analysis. Maybe I am wrong, but meta-analysis from cohort studies provides a ‘medium-to-high’ level of evidence, not the ‘highest level’. Again, large multicenter randomized clinical trials are needed to achieve the highest level of scientific evidence and meta-analysis of these studies achieve also the highest level of scientific evidence. “On the other hand, according to the results of the clinical trials performed in our group (Chiva-Blanch et al) with 67 men, we observed a beneficial effect of the non-alcoholic fraction of red wine (namely polyphenols) on insulin resistance; positive effects were observed only after alcoholic and dealcoholized red wine interventions, not after gin intervention. Thus, the type of alcoholic beverage seems to play a significant role in the protection on disorders related on glucose metabolism.”
Alcohol and diabetes: An overview by Forum member Puddey: “Gender related differences in the association of alcohol consumption with type II diabetes mellitus have been consistently reported for over 2 decades. These differences remain largely unexplained, but Kao et al offered 3 possibilities. Firstly that women are more likely to under-report the total amount of alcohol consumed; secondly that the preferred alcoholic beverage for women is wine rather than beer or spirits, with the attendant confounding of other healthier diet and lifestyle choices; and thirdly that there are often much smaller numbers of women in the heavier drinking categories in any epidemiological study. However, a recent meta-analysis of intervention studies that have measured the effects of alcohol consumption on insulin sensitivity and glycemic status (Schrieks et al) lends weight to a genuine gender related difference with the conclusion that among women but not men, alcohol consumption reduces fasting insulin levels and tends to improve insulin sensitivity.
“Less than half of the studies included in the current meta-analysis (17 of the 38 studies) adjusted for any measure of adiposity or body fat distribution when ascertaining overall risk of type II diabetes mellitus. Given the substantial differences between men and women in body fat distribution in particular, this may have been an important unmeasured confounder in the finding of an association of alcohol consumption with reduced diabetic risk in females compared to males. When the authors utilised a model based on crude or age-adjusted data was contrasted to multivariable-adjusted data they showed less reduction in risk at moderate levels of alcohol consumption but with the reduction in risk present across a broader range of exposure. In this regard, however, they did not model men and women separately and did not consider presence or absence of adiposity adjustment alone.
“In a similar vein, the cut-off for abdominal circumference in defining abdominal obesity is lower for Asians than for Europeans and both BMI and waist hip ratios are associated with Type II diabetes mellitus at lower cut points in Asians (Tan et al). The question remains as to whether pooling the Asian and European populations together in this meta-analysis potentially diluted any effects of alcohol to reduce diabetic risk in men vs women when there was adjustment for measures of adiposity. Simultaneously, such an approach could have decreased any potential increase in diabetic risk in men vs women with heavier consumption.
“Finally, Asian populations tend to drink fermented rice wine and spirits rather than grape wine. Any effects of alcohol on type II DM may be related to beverage type, with Kao et al finding in a US population that an increased risk of diabetes with heavy alcohol consumption was related in men to those who drank spirits rather than wine or beer. This provides a further basis that any meta-analysis of alcohol and type II diabetes mellitus may need to consider effects in Asian populations separately.”
References from Forum Critique
Baliunas DO, Taylor BJ, Irving H, Roerecke M, Patra J, Mohapatra S, Rehm J. Alcohol as a Risk Factor for Type 2 Diabetes: A systematic review and meta-analysis. Diabetes Care 2009;32:2123–2132.
Chiva-Blanch G, Urpi-Sarda M, Ros Em , Valderas-Martinez P, Casas R, Arranz S, Guillén M, Lamuela-Raventós RM, Llorach R, Andres-Lacueva C, Estruch R. Effects of red wine polyphenols and alcohol on glucose metabolism and the lipid profile: a randomized clinical trial. Clin Nutr 2013;32:200-206.
Di Castelnuovo A, Rotondo S, Iacoviello L, Donati MB, De Gaetano G. Meta-analysis of wine and beer consumption in relation to vascular risk. Circulation 2002;105:2836-2844.
Di Castelnuovo A, Costanzo S, Bagnardi V, Donati MB, Iacoviello L, de Gaetano G. Alcohol dosing and total mortality in men and women: an updated meta-analysis of 34 prospective studies. Arch Intern Med 2006;166:2437–2445
Howard AA, Arnsten JH, Gourevitch MN: Effect of alcohol consumption on diabetes mellitus: a systematic review. Ann Intern Med 2004;140:211-219.
Kao WH, Puddey IB, Boland LL, Watson RL, Brancati FL. Alcohol consumption and the risk of type 2 diabetes mellitus: Atherosclerosis Risk In Communities study. Am J Epidemiol 2001; 154:748-757.
Koppes LL, Dekker JM, Hendriks HF, et al: Moderate alcohol consumption lowers the risk of type 2 diabetes: a meta-analysis of prospective observational studies. Diabetes Care 2005;28:719-725.
Perry IJ, Wannamethee SG, Walker MK, et al: Prospective study of risk factors for development of non–insulin dependent diabetes in middle aged British men. BMJ 1995;310:560-564.
Schrieks IC, Heil ALJ, Hendriks HFJ, Mukamal KJ, Beulens JWJ. The effect of alcohol consumption on insulin sensitivity and glycaemic staus: A systematic review and meta-analysis of intervention studies. Diabetes Care 2015;38:723-732.
Smyth A, Teo KK, Rangarajan S, O’Donnell M, Zhang X, Rana P, Leong DP, et al. Alcohol consumption and cardiovascular disease, cancer, injury, admission to hospital, and mortality: a prospective cohort study. Lancet 2015. Pre-publication. http://dx.doi.org/10.1016/ S0140-6736(15)00235-4. Online/Comment http://dx.doi.org/10.1016/ S0140-6736(15)00236-6.
Stampfer MJ, Colditz GA, Willett WC, et al: A prospective study of moderate alcohol drinking and risk of diabetes in women. Am J Epidemiol 1988;128:549-558.
Tan CE, Ma S, Wai D, Chew SK, Tai ES. Can we apply the National Cholesterol Education Program Adult Treatment Panel definition of the metabolic syndrome to Asians? Diabetes Care 2004;27:1182-1186.
Prospective cohort studies for decades have tended to show that the risk of developing Type II diabetes mellitus is reduced among moderate drinkers in comparison with non-drinkers, with previous meta-analyses suggesting that moderate drinkers may have about 30% lower risk than that of abstainers. The present study carried out a multi-language search for studies on alcohol and diabetes and conducted a meta-analysis involving almost two million subjects from a final group of 38 cohort, case-cohort, case-control, or nested case-control studies. The main conclusions of the authors are that only females may show a significant inverse association between alcohol consumption and the risk of diabetes, and that previous studies may have overestimated the reduction in risk from moderate drinking.
Forum members had concerns about the analyses and the conclusions of the authors, as the males in the study were primarily Asian (86%), and there were huge differences between the associations with alcohol for Asian and non-Asian populations. There are always problems when the basic study data have a marked degree of heterogeneity. In general, many of the factors that relate to diabetes (diet, body size and adiposity, type of beverages consumed, etc.) are quite different between Asians and non-Asians; combining such groups when their analyses show opposite effects of alcohol on diabetes risk may not be a reasonable way of trying to develop results that apply globally. Forum members do not believe that the analyses presented in this paper can support the conclusion of the authors that there is “no effect among males.” It could be unwise if the results of this study were used to develop alcohol guidelines for western populations.
The authors could not take the pattern of drinking (regular versus binge) or the type of alcohol into consideration. Clinical trials have suggested that wine and its polyphenols may have effects above and beyond those seen with alcohol alone, and beneficial effects of such polyphenols have been demonstrated both for women and for men. Thus, while the analyses in this paper were all carried out accurately, attempting to use these data to develop guidelines that would apply globally, as was done by the authors, may not be appropriate.
Comments included in this critique by the International Scientific Forum on Alcohol Research were provided by the following members:
Elizabeth Barrett-Connor, MD, Distinguished Professor, Division of Epidemiology, Department of Family Medicine and Public Health and Department of Medicine, University of California, San Diego, La Jolla, CA USA
Giovanni de Gaetano, MD, PhD, Department of Epidemiology and Prevention, IRCCS Istituto Neurologico Mediterraneo NEUROMED, Pozzilli, Italy
R. Curtis Ellison, MD. Section of Preventive Medicine & Epidemiology, Department of Medicine, 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
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
Ross McCormick, PhD, MSc, MBChB, Professor Emeritus, The University of Auckland; former Associate Dean, Faculty of Medical and Health Sciences, The University of Auckland, Auckland, New Zealand
Ian Puddey, MD, Dean, Faculty of Medicine, Dentistry & Health Sciences, University of Western Australia, Nedlands, Australia
Creina Stockley, PhD, MSc Clinical Pharmacology, MBA; Health and Regulatory Information Manager, Australian Wine Research Institute, Glen Osmond, South Australia, Australia
Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway
Fulvio Ursini, MD, Dept. of Biological Chemistry, Universityof Padova, Padova, Italy
David Van Velden, MD, Dept. of Pathology, Stellenbosch University, Stellenbosch, South Africa