Moderate alcohol consumption, alcohol abstinence, and chronic diseases: is there a correlation?
On June 21, 2022, the Irish Government proposed to the European Commission to enact regulations that would put health warnings on wine labels and all alcohol bottles. Dublin did not receive an objection from the European Commission during a six-month moratorium period despite protests from over ten UE member states, including Italy.
Italy is a key exporter of wine. More than half of the Italian wine industry’s € 14 billion of annual revenue comes from abroad. It is understandable why Italy’s government and wine producers’ associations are worried Ireland’s plan risks damaging a leading part of their food and agriculture sector.
The main reason behind Ireland’s labelling decision is the need to address alcohol as a public health issue: the warnings will read as “Drinking alcohol causes liver disease” and “There is a direct link between alcohol and fatal cancers”. Wine-producing countries object that these warnings are disproportionate and discriminatory because they do not distinguish between abuse and moderate consumption.
The debate regarding the effects of alcohol consumption on human health is still open. That said, we would like to share the views of Dr. Riccardo Dalle Grave, Scientific Responsible for Associazione Disturbi Alimentari e del Peso (AIDAP) and Head of Nutritional Rehabilitation Unit – Casa di Cura Villa Garda (VR). Here is a recent article of his.
—
Moderate alcohol consumption, alcohol abstinence and chronic diseases: inconsistent and diverging recommendations based on weak evidence
For the past few years, the scientific community has seen a prolonged and heated debate between those who agree that the moderate consumption of alcohol provides benefits to health or no harm, and those who, on the contrary, claim that any amount of alcohol intake increases the risk of sever chronic noncommunicable diseases, such as cancer and cardiovascular diseases.
This scientific debate has led to inconsistent and divergent recommendations to the public on the consumption of alcohol by health stakeholders. For example, the U.S. Dietary Guidelines for Americans (DGA) 2015-2020 recommended for people over 50 years old to avoid binge-drinking and to drink less than 14 units a week for men and less than 7 units a week for women. One unit corresponds to approximately 150ml of wine with an alcohol content of 12 per cent. In contrast, the International Agency for Research on Cancer (AIRC) has recommended complete alcohol abstinence, emphasizing that alcohol was included in the Group 1 Carcinogens 30 years ago and appears to be implicated in the development of 4,1% of cancers worldwide in 2020.
The divergence in recommendations on alcohol consumption has become extremely relevant, with major implications not only in health policy but also economics, resulting in Ireland’s decision to state on bottles of wine, beer, and other spirits that alcohol inflicts severe harm, in a similar manner to that already found on cigarette packaging. The proposal was communicated to the European Commission on June 21, 2022, and the deadline to object now passed; thus, there are no obstacles to the employment of the new labels.
These initiatives and warnings were strongly opposed by Italy through its Minister of Foreign Affairs, who defined Ireland’s decision to introduce a label on alcoholic beverage, including Italian wine, despite the objection of the European Parliament and the opposition of countries, including Italy, France, Spain and six others, as absurd, because it ignores the difference between moderate consumption and abuse of alcohol.
In Italy this dispute also involved some leaders in scientific communication during the COVID-19 pandemic and doctors who, despite not being experts in nutritional epidemiology, released interviews in which they stated that even moderate alcohol consumption is harmful to health or, on the contrary, that it is not harmful.
Inconsistent and diverging recommendations create consumer disorientation and can have significant negative effects on the economies of wine-producing countries and, as I will explain in this article, are the consequence of methodological limitations that afflict epidemiological research studying the relationship between diet and health in general and alcohol consumption and chronic noncommunicable diseases in particular.
Studies that have evaluated the association between alcohol consumption and disease
Nutritional epidemiology assesses dietary or nutritional factors in relation to the insurgence of diseases in populations. Findings from nutritional epidemiology often contribute to evidence used to guide dietary recommendations for the prevention of cancer and other chronic noncommunicable diseases, such as type 2 diabetes and cardiovascular diseases.
To establish the existence of a definite relationship between alcohol intake and disease, evidence has been sought by means of three main study methodologies. Unfortunately, each of these study methods has significant methodological limitations and has produced diverging results.
Preclinical studies
These studies investigate the mechanisms through which a specific dietary factor may be implicated in the development of a disease (e.g. cancer). The studies typically include cell culture experiments and research on laboratory animals.
These studies have highlighted that high concentrations of alcohol unleash oxidative and cardiotoxic mechanisms in cells that cause cardiomyopathy, arrhythmia, and heart failure. In contrast, moderate concentrations of alcohol have been shown to increase the level of HDL cholesterol (the good cholesterol) and improve the energy metabolism profile of the heart. Resveratrol has also been suggested as an anti-atrial fibrillation agent through the modulation of reactive oxygen species (ROS), also known as free radicals, and oxidative stress.
The mechanisms involved in the association between alcohol and cancer have been widely studied, in particular for breast cancer, highlighting the detrimental effects of ethanol on breast cancer cells, such as the promotion of their growth and angiogenesis. The studies that have also found that cancer cells repeatedly treated with alcohol increase their capacity to migrate and metastasize other bodily tissues.
However, it should be underlined that the preclinical studies “cannot” prove that a particular dietary factor causes or prevents cancer or other diseases in humans. The proof can only be established by studies performed on humans.
Prospective cohort studies on humans
These are observation studies that periodically examine the dietary habits of a specific population and then assess whether an association, rather than a causal link, exists between the development of certain diseases and the intake of certain aliments. These studies, from which population dietary guideline recommendations are generally obtained, have many methodological biases (e.g. the distortion of measurements due to systematic errors) and often produce conflicting results, which are the main source of the nutritional recommendations, such as those on alcohol consumption.
The main bias in these studies is the use of food frequency questionnaires. In fact, often people do not remember accurately the frequency of foods they ate and of beverages they consumed and fill out the questionnaires quickly and approximately. Regarding alcohol consumption, in particular, there are three specific additional methodological biases that significantly affect the interpretation of the results:
The “sick quitter” effect that includes among non-drinkers former drinkers who have quit drinking because of an illness. This effect can lead to erroneously conclude that non-drinkers have a higher risk of developing certain diseases that are instead linked to the presence of a pre-existing diseases.
The “healthy user” effect that includes among people who consume alcohol moderately more educated people who adopt a healthy lifestyle and drink responsibly. This effect can lead to wrongly crediting moderate drinking with beneficial health effects, which can instead be the consequence of an overall healthy lifestyle.
The effect of underestimation of amount of alcohol intake, which can reach up to 45-60% among heavy drinkers. This causes some heavy drinkers to be classified as moderate drinkers with the consequence of health risks associated to the moderate consumption of alcohol being accentuated.
In any case, results from large prospective cohort studies have concluded that moderate alcohol consumption reduces the development of cardiovascular diseases, type 2 diabetes and even all-cause mortality, particularly in subjects over 50 years old. However, these cohorts have included highly selected and educated upper middle-class subjects, and the results are therefore influenced by the healthy user bias (see above).
The high alcohol intake has resulted in being associated with the development of several digestive diseases, such as alcoholic hepatic steatosis, chronic hepatitis and cirrhosis. Moreover, although the liver is the main effected organ, high alcohol consumption can lead to pancreatitis or gastritis. However, it is important to stress that cohort studies have found that moderate alcohol intake does not appear to be associated to these diseases.
Regarding cancer, due to the oxidative compounds caused by alcohol metabolism, cells, as marked by pre-clinical studies, are more inclined to mutations, creating the impression that alcohol has a direct link to cancer. Most organisations, as a matter of fact, recommend zero consumption of alcohol to prevent breast, oropharynx, larynx, esophagus, liver, colon, and rectum cancer. However, analogous to the diseases described in the previous paragraphs, cohort studies have found that moderate consumption of low alcohol content is associated with lower cancer risk compared to complete abstinence. Again, however, the results may be influenced by the effects of some of the biases described above.
Randomised and controlled studies
These are experimental studies in which participants are randomly assigned to two groups, the test group receiving the treatment or the control/comparison group. These studies are the only ones that theoretically allow an assessment of whether a particular nutrient and alcohol is the cause of a disease.
Unfortunately, although potentially feasible, randomised and controlled trials in the research field on the effects of alcohol intake in the development of diseases have encountered obstacles that have not been overcome. Indeed, the time span between a dietary factor and the development of certain diseases is uncertain, thus it is necessary for people to follow the dietary prescription to which they have been allocated for many years, something that has not been put into practice by the majority of individuals.
This explains why in the field of nutrition, and of alcohol in particular, only short term randomised and controlled studies have been performed, in small samples, that tested the effects of alcohol on intermediate outcomes, such as blood lipids, inflammation and glycemia, but did not evaluate its effects on the development of cancer, cardiovascular diseases and other chronic non communicable diseases. In summary, these studies have found that alcohol intake has positive effects because it increases the levels of HDL cholesterol and of its lipoprotein (apolipoprotein A1), of adiponectin, an adiponectin that increases the insulin sensibility and appears to reduce the risk of diabetes, and it decreases fasting levels of insulin, glycosylated hemoglobin, and fibrinogen (a marker of inflammation). In contrast, some studies have found that alcohol increases sex steroid hormones, which could increase the risk of breast cancer.
In Italy, a randomised, controlled study on 131 patients with myocardial infarction and diabetes allocated to the Mediterranean diet with or without adding 120ml of red wine found that after one year those who consumed red wine had, compared to the control group, beneficial effects reaching significantly higher levels of HDL and lower levels of oxidation markers and fasting insulin, reductions in several inflammatory biomarkers, and improved left ventricular function.
The Mediterranean model of alcohol (red wine) consumption and the reduced mortality from all causes
The Mediterranean model has been defined as the moderate consumption of red wine with meals throughout the week and avoidance of binge drinking. Studies have found that this model of alcohol intake is inversely associated to all-cause mortality compared with alcohol abstinence. This is consistent with the so called “French paradox”, a term coined three decades ago to explain the low incidence of cardiac disease among the French population, which adopts a diet rich in saturated fats. This paradox has been attributed to the habitual consumption of wine during meals.
The Greek cohort of the EPIC (European Prospective Investigation into Cancer and Nutrition) study also reported that none of the other nine elements used to define the Mediterranean diet model exerted as strong a benefit as moderate intake of alcohol consumed at meals on all-cause mortality. A finding that could indicate that alcohol assumed moderately has an independent positive effect on all-cause mortality.
However, a large number of prospective studies have reported more benefits from the adherence to the Mediterranean diet model rather than from the intake of a particular nutrient. Therefore, the Mediterranean model of alcohol consumption associated with the Mediterranean diet seems to act through multiple coexistent related mechanisms rather than by the effects of a single nutrient. For example, the antioxidant components of diet and wine could reduce or overpower the carcinogenic effect of ethanol.
Conclusion
Current research does not provide an answer to the controversy over the potential effects (detrimental, neutral or beneficial) of moderate alcohol consumption in people over the age of 50. Doctors, the scientific community and health stakeholders need solid, evidence-based data to provide recommendations to their patients and the public.
It is clear that the complex relationship between diet and alcohol requires new lines of research to provide more precise recommendations that do not rely on the influence of economic lobbies and studies that fail to demonstrate a causal relationship between alcohol intake and disease or that have major biases that influence the results and their interpretation.
Only if randomised, controlled long duration studies, with a high number of a participants are conducted we will be able to provide reliable indications on the models of alcohol consumption to be recommended. On the basis of our current understanding, we can only recommend to those who do not drink alcohol to not start doing so, and to those who do drink, to do so following the Mediterranean model of alcohol consumption in combination with the Mediterranean diet, an eating lifestyle that is associated with reduced mortality from all causes.
