In the three decades I spent as an emergency physician practicing in California, I saw tragic accidents caused by drunk driving and treated countless patients for alcohol intoxication and withdrawal, along with many other alcohol-related illnesses. I have seen firsthand the disastrous impact of alcohol abuse.
I also made a point to read every scientific paper that was published about alcohol and health. As a member of a centenarian wine family, as well as a Stanford-trained medical doctor, I wanted to be able to answer questions truthfully and without bias about what constitutes safe levels of alcohol consumption.
Understanding the science behind the impact of alcohol consumption is important to the many drinkers around the world, myself included, who derive great pleasure from having a glass or two of wine with dinner – an 8,000-year-old tradition in some parts of the world.
My current position is shared by many doctors, scientists, and public health researchers with decades of expertise in the field. Take the measured words of Dr. Gregory Marcus, MD, a cardiologist and professor of medicine at University of California, San Francisco, speaking about moderate alcohol consumption on KQED’s Forum in September: “The current evidence for health and harm is fairly equal.”
But while there is clear consensus that heavy and binge drinking are very harmful, a group of scientists, some of whom have received funding from temperance organizations, are making blanket statements, such as the one released by the World Health Organization (WHO) in January 2023 that “no level of alcohol consumption is safe” for our health. Such declarations fail to acknowledge the nuances in the data and dismiss decades of research in reputable journals showing cardiovascular protective effects for moderate drinkers. These nuances are not making it into recent alarmist media articles, which mostly quote anti-alcohol researchers rather than a mix of subject-appropriate experts.
In July, I created In Defense of Wine – the name is inspired by New York Times wine critic Eric Asimov’s article – as an online resource for credible, science-based information on health effects of light to moderate alcohol consumption. I am concerned that misinformation about the true risks of light to moderate alcohol drinking might lead people to either ignore the guidelines, or unnecessarily give up something that provides enjoyment and is not proven to be harmful. Here, I want to set out why we need to dispense with the idea that there is no safe level of alcohol consumption, and give people the tools to make their own personal decision in consultation with their primary care physician.
The cardiovascular effects of light to moderate alcohol consumption
The scientific community uses a metric called impact factor to evaluate the relevance of research journals. The Lancet has an impact factor of 98.4, among the highest in medicine, which is why its 2020 Global Burden of Disease study (published in 2022) is so important.
“For adults over age 40, health risks from alcohol consumption vary by age and region,” state the study’s authors. “Consuming a small amount of alcohol for people in this age group can provide some health benefits, such as reducing the risk of cardiovascular disease, stroke, and diabetes.” This declaration is supported by decades of observational research consistently showing that moderate drinkers experience cardioprotective effects and a reduction in mortality. (The definition of moderate drinking varies somewhat between studies but is usually between one and two drinks per day, with a few alcohol-free days every week.)
These cardioprotective effects on the human body make sense: they follow the common scientific principle that when there is a mechanism to explain a certain result, it is more likely that the result is true. In cell, animal, and human studies it has been demonstrated that alcohol can increase good cholesterol (HDL) and reduce the kind of clot formation that mediates heart attacks and embolic strokes (essentially, alcohol acts as a clot buster. There is also solid data showing that moderate alcohol consumption reduces average blood sugar levels and lowers the risk of diabetes, which is another potential explanation for moderate alcohol’s cardioprotective findings as diabetes is a known risk factor for cardiovascular disease. (Although alcohol consumption has been associated with a decrease in heart attacks, for some people it can increase the risk of certain arrhythmias such as atrial fibrillation, and you must always evaluate personal risks with your doctor.)
Despite this robust catalog of scientific evidence, moderate alcohol’s cardioprotective effects have been dismissed and attacked by groups whose anti-alcohol messaging has appeared in major consumer publications. When examined more closely, however, the studies referenced in these articles often fail to reveal the full picture. For example, an oft-cited JAMA Network 2023 review, referenced in articles such as “How Red Wine Lost Its Health Halo” in The New York Times, does not show statistically significant harm from moderate drinking. Even this study concludes that the higher mortality risk related to alcohol drinking starts for women at 1.8 drinks per day and for men at 3.2 drinks per day, quantities which exceed the French, U.K. and U.S. guidelines.
Another study frequently cited in these articles was published this year in the Journal of Studies on Alcohol and Drugs – which has an impact factor of 2.4 compared to the Lancet’s 98.4 – and presents a rehash of previously published data, while proposing to debunk moderate alcohol’s cardioprotective findings. The researchers – including Tim Stockwell and Tim Naimi, who both acknowledge accepting funding from anti-alcohol organizations – start with 3,248 relevant studies, and then cherry-pick their way down to six by eliminating, for example, important studies that looked at pattern of drinking, in order to reach the conclusion that moderate alcohol is not cardioprotective. Even more problematic is that the reference articles are incorrectly listed, something Stockwell admitted to in an interview, and upon review, three of the six studies chosen as bias-free actually do show the association of lower mortality in moderate drinkers – also known as the J-curve.
What about alcohol and cancer?
There is no doubt that at high doses or with binge drinking, alcohol increases the risk of several types of cancer, cardiovascular disease, mental illness, dementia, and accidents.
Alcohol is a known carcinogen at high doses. When you drink in excess, alcohol can exceed your liver’s natural processing capacity. In this state, your liver is unable to quickly clear harmful by-products such as acetaldehyde. In the case of breast cancer, even at one unit of alcohol per day, there is a potential 1.1x increase in cancer risk, possibly related to a higher estrogen production resulting from alcohol. This would increase the lifetime risk of breast cancer from about 12 percent for a typical woman to about 13.2 percent. For a woman aged 60, whose chances of dying from cardiovascular causes are approximately 10 times those of dying from breast cancer, the balance of health effects leads most doctors not to recommend complete abstention.
Often left unmentioned is that many studies examining alcohol and cancer have shown a reduction in some cancers associated with moderate drinking (for example, thyroid, kidney, and some hematological malignancies), although the mechanism is unknown. When it comes to the 40 percent of cancer risk that is related to modifiable risk factors, according to a recent U.S. study, nearly 20 percent was related to tobacco, nearly 15 percent to excess body weight, unhealthy dietary patterns, and physical inactivity, eight percent to ultraviolet radiation exposure plus certain infections, and a little over five percent to alcohol, which includes the heaviest drinkers. Among the many things that an alcohol consumer can do to reduce cancer risk would be to adhere to the Mediterranean diet (which includes drinking wine in moderation), as it has been associated with a reduction in the incidence of cancer and cardiovascular disease. Women who drink alcohol might be able to mitigate their breast cancer risk by increasing the folate in their diets (in amounts found in a daily multivitamin).
Alcohol’s effects on the human body are broad and complex
Alcohol’s complex effects on the human body have been investigated in many studies – and the results are complex and varied. For example, The Journal of Prevention of Alzheimer's Disease found that moderate regular alcohol intake was associated with better global executive and visual memory function among older adults, even with control for a number of potentially related health and lifestyle variables. Studies of patients with rheumatoid arthritis have reported potential benefits of light to moderate alcohol consumption.
When it comes to stress reduction, a 2023 study out of Harvard found that light to moderate alcohol drinkers had a marked reduction in amygdala activity – the fight or flight stress center of the brain – and the associated cardioprotective effects were even more pronounced in those with a diagnosis of anxiety.
Given the prevalence of alcohol consumption around the world, and the many communities who drink as part of their cultural heritage, it is important that research continues to work toward a better understanding of the overall effects of alcohol on the human body – and for that message to be communicated to the general public.
Our alcohol-health communication is flawed
None of this nuanced messaging is getting through, however, in the barrage of recent articles about alcohol and health. On reviewing a 2023 WHO publication “Reporting About Alcohol: A Guide For Journalists,” the first thing that surprises me is that among the document’s contributors is a person working for an acknowledged temperance organization, Movendi International.
A thorough analysis of the document reveals referenced sources that do not support the document’s statements. The report reads: “While several past studies did suggest that moderate consumption could, on average, promote health benefits, newer research shows that those studies used limited methodologies and that many of them were funded by the alcohol industry.” It references a source which directly contradicts this: a 2020 analysis of 386 observational studies evaluating moderate alcohol consumption and health, showed that nearly 95 percent of the studies were funded independently. And for the few that received alcohol-industry backing, the analysis found no association between the source of funding and different health outcomes, concluding that there was no bias related to funding source.
The WHO guide’s statement negating alcohol’s cardioprotective qualities is not supported by its own references. “There is no good evidence for the pervasive myth that consuming red wine helps prevent heart attacks,” the report states. The referenced 2018 paper in The Lancet directly contradicts this: “…increased alcohol consumption was log-linearly associated with a lower risk of myocardial infarction [heart attack].” That same study concluded that “the threshold for lowest risk of all-cause mortality [death from any cause]” was about 100 grams per week, which is the equivalent of one five-ounce glass of 12% ABV wine per day. Even this analysis would suggest that daily drinking limited to one serving poses no overall health risk.
The 48-page supplementary appendix of this 2018 Lancet study reveals an even more favorable picture for moderate drinking. Lifetime abstainers had substantially greater all-cause mortality than consumers of one or two drinks a day, and among those who spaced their consumption over several days per week, mortality was actually lower for individuals who consumed up to 200 grams of alcohol per week (equivalent to two glasses per day) than the rarest drinkers. The appendix data also revealed a healthier outcome for wine drinkers over spirits, which was not reflected in the main paper. Researcher Eva Schernhammer of the Medical University of Vienna submitted a comment that the message would have changed had the results in the appendix been included in the main paper: “The headlines might have then resembled those of just a year ago: ‘A glass of wine or pint of beer a day can help people to live longer, according to new research.’”
When a respected institution such as the WHO issues reports containing misleading information, it’s not surprising that we see a glut of headlines with the “no safe level” message. Throughout the WHO’s guide and in several consumer media publications’ recent reporting on the subject, there is a tendency to mix statements about risks of light to moderate drinking with blanket statements about alcohol’s harms. The reader is left with a one-sided impression that all drinking poses significant health risks. If we make a comparison to driving, for example, it would be like looking at death statistics from speeding and aggressive driving and reaching conclusions about safe driving.
A question of methodology
Over the last two decades, researchers have appropriately questioned the methodologies of older studies on alcohol and health. Some of these older studies failed to separate never-drinkers from ex-heavy-drinkers and sick-quitters (people who stopped drinking due to an illness) in their control groups. This important learning resulted in improved research protocols for studies conducted over the last two decades. These newer, better studies still show cardioprotective effects for limited alcohol consumption and that is why reputable, high impact publications such as The Lancet (2022), Journal of the American College of Cardiology (2017, 2023), Circulation (2017) and Mayo Clinic Proceedings (2021) continue to publish articles supporting cardioprotective effects for light to moderate drinking, especially with meals.
Part of the problem is that many of these studies are observational, and with observational data there can be many confounders (variables that provide alternate explanations for the results); therefore, an association can not be interpreted as causation, whether it be positive or negative for health. The same applies to the headline-grabbing Mendelian Randomization studies (using variation in genes to examine the causal effect of an exposure on an outcome) where confounders can also make the results questionable. A large-scale randomized controlled trial – the gold standard of studies, which is better at determining causality and used to evaluate medications – has never been conducted for alcohol and health.
The good news for wine drinkers around the world is that answers are coming. A large, 10,000-patient, randomized controlled study is currently being conducted in Spain with results expected in five years. Half the patients, who normally follow the Mediterranean diet, are being randomized to continue drinking wine in moderation, and half are abstaining from alcohol completely. While this is progress, it will be difficult to extrapolate the results to other diets unless a randomized controlled study is undertaken in a diverse group of countries with diverse populations and cultural practices around alcohol and food.
It is important to underscore that all binge drinking is harmful to health and eliminates any potential benefits of light to moderate consumption. In the U.S., according to the Centers of Disease Control, six percent of Americans drink heavily and 17 percent binge drink. Among Americans who drink alcohol (greater than half the drinking age population), many are drinking in a pattern that is harmful, which needs to be addressed. There is also very solid data that underage and youth drinkers have a much higher risk of developing an alcohol use disorder later in life, so it is important that public health officials, governments, health personnel, and the alcohol industry do everything possible to prevent youth alcohol drinking.
The alcohol industry as a whole needs to help reduce the harm that alcohol can cause. We need to become comfortable not serving people who are intoxicated, pouring water in abundance – like my father taught me, one glass of water for every glass of wine – and always having personal spittoons available, even in fancy restaurants and tastings. There are many reasons for someone to choose not to drink and we need to create a warm and welcoming place for non-drinkers within our social and work communities. We need to provide alcohol-free alternatives, so that people can take a few days off when they find it necessary to cut down, and embrace our colleagues who have stopped drinking for whatever reason.
The WHO calls for increased education on the cancer risks associated with alcohol consumption, perhaps including health warnings on alcohol labels, and few would argue against better informing the public with regard to health. But when we balance cardioprotective effects (positive) with potential cancer risks (negative) we are left with a nuanced answer to the moderate alcohol-health question, at least for those over 40, who are most at risk for cardiovascular disease. It seems reasonable then that the quality of life gained from light to moderate drinking might be deemed greater than the potential harm. And blanket alcohol harm statements that don’t clarify the difference between moderate and excess drinking would inaccurately inform the consumer.
I believe the “no safe level” or “two drinks a week” positions are based on inconclusive science and prohibitionist ideology, and if they become entrenched, then governmental ethics committees and research institutions will potentially stop approving alcohol research. This could result in the public not getting a definitive answer to the question of alcohol’s impact on health, which is so important to the many people who drink in moderation.
Wine helps bring people together and maximizes enjoyment of meals; there is a long history of wine as inspiration to lovers and artists, and as an enhancer of celebrations and social moments. Its positive impact should not be overlooked.
