Is Having A Drink A Day Healthy?

New Research Sheds a Ray of Light

Introduction

You might have heard or read somewhere that one to two glasses of wine a day can be good for your health. Some consider it a practice—part of the so-called ‘Mediterranean Diet’, and moderate alcohol consumption, defined by the CDC as two drinks or less daily for men and one or less for women is a well-accepted cultural norm in many parts of the world.

Some studies have shown moderate drinking to be associated with longer life spans and several studies have shown that it helps raise blood levels of HDL (the so-called ‘good cholesterol’), leading some medical experts to hypothesize that it might be heart-protective. Some researchers believe that it is the anti-inflammatory effects of the polyphenol compound resveratrol, which has been described as exerting a wide range of health benefits including protecting the endothelial lining of blood vessels, lowering LDL (the so-called ‘bad cholesterol’), and preventing blood clots, found in low doses in red wine (as well as in foods like chocolate, blueberries, blackberries, and pistachios), that is responsible for the positive association between longevity/health and daily wine consumption. 

More recently, however, you may also have heard or read the opposite—that even modest drinking is strongly associated with an increased risk for heart disease, dementia, and various forms of cancer. And some research has failed to show that resveratrol positively affects cardiovascular health or lowers inflammatory markers or metabolic risk. Last year, the WHO changed its recommendations about safe alcohol use, declaring that no amount of alcohol consumption is safe to consume…

So, what is the truth? As you will see, while it is not yet time to offer an authoritative ruling on the issue, new data are giving us the first solid clues about how much alcohol consumption, if any, may be safe and possibly even healthy. But before we present the new information, let’s get the thirty-thousand-foot view of the science behind drinking over the last couple of decades and try to understand why there has been so much disagreement and confusion on this subject.

The Problem(s) With Many Alcohol (OH) Studies

I. Conflicts of Interest

There have been some recent doubts cast about the validity of studies purporting to show that moderate alcohol (OH) consumption may have positive health effects, based in part on a report showing that more than 13,000 studies in the academic literature on the subject have been funded by the beer, wine, and liquor industries. The manufacturers, distributors, and sellers of alcoholic beverages have a vested interest in promoting the story that drinking is healthy (or at least not harmful), and that represents a clear conflict of interest that could bias how they conduct research and interpret the data from the studies they fund. One could imagine studies showing results favorable to their interests being put forward for publication while those showing otherwise being withheld. To be clear, I am not aware of any credible data to support that suspicion, But there exists at least the appearance of bias when so many studies funded by ‘Big Alcohol’ have resulted in health-favorable findings.

And the potential for bias cuts both ways. As my friend, research scientist Nicolai Worm, has pointed out in private communications, several members of the WHO and the Canadian Panel on Alcohol Consumption Guidelines are members of religious groups and/or temperance organizations that are explicitly, for reasons that extend beyond the question of health, anti-drinking. This, too, represents a significant conflict of interest for those advisors to offer unbiased opinions and create recommendations for safe OH use based exclusively on the scientific evidence…

The institutions of autocracies run on fear and intimidation. But in democracies, they are held together by trust and cooperation. In free societies, public-facing institutions like the National Institutes of Health and the CDC must create and maintain a culture that prioritizes the unending struggle to be corruption-free. And they must be seen by the public whom they exist to serve as succeeding in that endeavor. Societal health and cohesion depend in large part on the health of its institutions, and the steady creep of bias into the fields of scientific and medical research, driven by financial, political, and religious interests, represents a threat to the well-being of Western-style democracies.

It’s a problem that is growing. While funding for basic science research in the US remains robust, over the last two decades, support from the federal government has shrunk significantly while an increasing share of the money driving research has come from private business interests, creating at least the potential for bias through implicit conflicts of interest.

The manufacturers, distributors, and purveyors of alcoholic beverages make up a not insignificant share of that funding. For example, approximately two-thirds of the $100 million that went toward the Worldwide Moderate Alcohol and Cardiovascular Health (MACH) Trial, was supplied by five global alcohol producers (Anheuser-Busch, Carlsberg, Diageo, Heineken, and Pernod Ricard). That study was terminated by NIH before reaching any definitive conclusions about the health effects of moderate drinking based on concerns about the potential for bias and the possibility that, as a result of its design, the study would miss important harms associated with even moderate OH consumption.

In short, those with a vested interest (both positive and negative) in the outcomes of research on the health effects of OH consumption are driving an ever-increasing share of the new science, creating at least the appearance of thumbs being placed upon a scale whose power lies in its trustworthiness, and whose trustworthiness derives from its properly maintained impartiality.

But we should be careful to avoid throwing the baby out with the bathwater. As is true of the arts, with less of the US federal budget going toward scientific research, private funding is needed to fill the gap. The costs for vaccine and drug-related research today are borne substantially by the pharmaceutical industry. Without the research funded and conducted by Big Pharma, the FDA would not have the data it needs to approve new medicines and vaccines in a timely way or create the safety and efficacy standards needed for new pharmaceuticals. The more interconnected our world becomes, the greater the risk of epidemics and pandemics. If we are going to keep up medically with that world, we need more research, not less…

II. Not All Science is Good Science

Allowing Big Pharma to foot the bill for clinical trials to test new drugs and vaccines is an accepted practice despite the inherent conflict of interest—one that has not received too much criticism. Why not?

One reason is that we haven’t much choice. Novel viruses and other pathogens like M-Pox that were once confined to small regions of the world are spreading. Meanwhile, an ever-increasing share of US government funds earmarked for healthcare is being funneled into a byzantine for-profit system that is ballooning in an unsustainable way. Epidemic levels of obesity, addiction, and chronic metabolic illness have made America arguably the sickest country in the world, and as the budget needed to manage the societal burden of chronic illnesses has expanded, the budget for new research has shrunk.

Those who advocate for less government oversight of big corporations are quick to point out that there exists an implicit check on corruption that is inherent in the for-profit, market-based healthcare system. Should a pharmaceutical company be caught manipulating or fabricating data, the consequences would be disastrous—financially and reputationally—for the bad actor. Of course, that is not always the case. The Sackler family, for example, owners of Purdue Pharma, the company that developed and marketed OxyContin on the back of corrupt science falsely showing that their drug was non-addictive, have not been imprisoned or even charged with any crime and remain extremely wealthy.

Could we say that the same ‘check’ on abuse inherent in the for-profit, market-based healthcare system that should, in theory, act as a hedge against corruption in ‘Big Pharma’ exists for ‘Big Alcohol’? That the experts empaneled to make public health recommendations would face sufficient market-driven repercussions to keep them honest? Not with a straight face. However, that has not diminished the passion of many conservative ideologues advancing an extremist agenda of governmental deregulation that has contributed, in my view, substantially to the degradation of public trust in industries that impact public health.

So, there is not enough public money to go around, and corruption, driven by private interests, is steadily encroaching on the world of science, supported by an irrational faith in the magic of free markets. What else might be playing a role in creating so much conflicting data? Unlike drug or vaccine studies, which are typically controlled and (often double) blinded (A.K.A. good science), most of the studies purported to evaluate the health risks and benefits of drinking consist mainly of observational data. Observational studies often fail to control for important covariates that can drive both positive and negative outcomes.

For example, it is well established that people who drink regularly are three times more likely to smoke than non-drinkers. One study found that 58% of people who were at risk of becoming alcohol dependent were current smokers compared to just 15% of the general population. If drinking is that strongly tied to smoking, then it could very well be the effects of cigarettes, not OH, that explain the association of poor health outcomes among those who drink moderately.  

Similarly, an analysis of more than eighty prospective studies showed that moderate OH consumption is associated with up to two years of shorter lifespan. However, according to the National Highway Traffic Safety Administration, drunk driving accounts for about 11K deaths each year in the US—about one-third of all traffic fatalities. How much of the reduction in lifespan seen among moderate drinkers compared to non-drinkers was due to fatal drunk driving incidents among young persons rather than the ill health effects of OH itself?

Conversely, we know that people who eat a Mediterranean diet tend to have longer lifespans and that moderate drinking is a common feature among people who eat a Mediterranean-style diet. Those associations have been used to promote the narrative that moderate OH consumption can improve lifespan. However, Mediterranean-style diets are rich in fiber and low in processed foods and sugar—things that are known to promote good health and longevity. Is it the diet or the OH that is driving the longer lifespan? Those who follow a Mediterranean-style diet but do not drink may very well be healthier and living longer than those in their dietary cohort who have a glass or two of wine each day. We would need studies comparing those two groups to know whether OH plays a positive, negative, or neutral role in health outcomes…

So, most of the data on moderate drinking comes from observational studies that do not control for critical independent variables (factors that could potentially influence the outcome(s) being tested).

Another major problem with observational studies on OH consumption is that they often rely on self-reported data. Subjects participating in studies by reporting what/how much they are eating or drinking are notoriously inaccurate. A so-called social desirability bias in which people routinely over or under-report what they are eating and drinking to conform to what they believe to be socially appropriate or ‘good’ behavior, as well as the mood of respondents at the time of reporting, memory errors, misunderstanding the task of reporting (does a light beer count the same as a gin and tonic?), and several other factors all contribute to making self-reported data highly inaccurate.

How much confidence should we place in a study that demonstrates an association between moderate drinking and increased risk of heart attack when self-reporting is inaccurate, smoking and diet are not controlled for, and the population studied was mostly obese and elderly?

III. Not All Journalism is Good Journalism

Standing alongside science, a free and unfettered press is another of the crown jewels of Western-style democracy. The health and cohesion of a society, it could be argued, depends as much on the ability to accurately report the truth as on the ability to find it. But journalism, like healthcare and science, are today mainly profit-driven endeavors upon which special business, political, and religious interests exert or seek to exert a significant measure of control.

Consider the following 2023 headline from Fox News: “Face masks made ‘little to no difference’ in preventing spread of COVID, scientific review finds.” The review study referenced was from the Cochrane Collaboration, a not-for-profit group of scientists and healthcare professionals who conduct independent, systematic reviews of healthcare interventions. Those reviews are subject to infamously rigorous quality standards, so when news broke across a wide swath of the for-profit world of journalism declaring that the Cochrane Collaboration had found masking “may have done little to nothing to curb the transmission of COVID,” many of us familiar with that trustworthy organization who had been following the COVID science for three years were stunned.

Of course, that was not at all the group’s conclusion. The Cochrane review described instead how many of the (mostly observational) population-based studies of outcomes following mask mandates were so flawed that no conclusions could reasonably be drawn about the effectiveness of masking policies, one way or the other. That is very different from concluding that high-quality masks, properly worn, do not protect the wearer against airborne respiratory infections. The authors made a point of noting a high risk of bias in the trials, a wild variation in outcomes that typically speak to sloppy science, and relatively low adherence to the interventions (masking) by subjects, all of which contributed to the inability to determine, based on the available data, whether or not mask mandates have any value...

Did the Fox News journalist understand the critical distinctions between masking and mask mandates, and between bad science providing insufficient evidence to support mask mandates and the conclusion that masking is ineffective? Did she corruptly misrepresent the findings of the Cochrane Collaboration to write an anti-mask piece on behalf of a media outlet that had spent two years spreading disinformation to discredit the CDC, attack the integrity of its leader, Dr. Fauci, and promote a narrative that the federal government, by imposing masking mandates, was infringing upon Americans’ civil liberties (for the record, the federal government did not at any time impose any mask mandate; that too was disinformation)? Did she and her editor(s) simply lack the scientific training to understand the review paper? I often ask similar questions when I read in the New York Times that plant-based protein promotes longevity while red meat has been linked to shorter lifespans…

In short, there is scant data on the health effects of moderate alcohol consumption that would qualify as ‘good science’ and a mountain of scientifically weak studies pointing strongly in opposite directions that make for catchy but misleading headlines.

IV. A Glimpse of Light

The associations between heavy drinking and all manner of chronic diseases, from cirrhosis of the liver to heart disease, diabetes, various forms of cancer, depression, sleep apnea, and sexual dysfunction, are so strong and so consistent as to be widely understood by experts as cause-and-effect. However, when it comes to occasional or moderate OH consumption, we have comparatively little high-quality data to go on.

Why do some studies suggest that moderate drinking is harmful while others suggest it has health benefits? Corruption, whether in the form of bias, sloppy science, misunderstanding data, or intentionally misreporting data (disinformation), is part of the problem. But might there also be a more nuanced understanding of drinking and its effects on health that could make both opposing narratives sometimes true?

A recent analysis of over 300,000 drinkers in the U.K. found that the same total amount of OH consumed per week appeared to increase the chances of premature death if consumed on fewer occasions during the week and outside of meals, but to decrease mortality if spaced out across the week and consumed with meals. 

These findings add a subtle but critical distinction that might help clarify the vastly conflicting observational data and ongoing debate about the relative health effects of moderate drinking and provide the first strong evidence-based clue as to why one or two glasses of wine per day has frequently been associated with better health and longevity outcomes. If replicated by subsequent studies, clinicians will for the first time have the ability to make valuable, evidence-based recommendations to patients about moderate drinking.  

It is noteworthy that the study showed that regular (>3/week), moderate OH consumption (1-2 drinks), regardless of the beverage, when paired with meals, appears to be the formula driving the better health outcomes, thereby negating the resveratrol hypothesis. Why would this be true? My own theory (not supported by scientific evidence) is that alcohol is a stressor that, if taken in small doses over a long period of time, can have a hormetic effect. Hormesis is an adaptive response to a moderate stressor whereby a low dose can have a stimulatory or strengthening effect, while a higher dose can have an inhibitory effect or weakening effect. Physical exercise, sauna bathing, fasting, and exposure to low doses of certain toxic phytochemicals have all been shown to have positive health effects if engaged in regularly over long periods by inducing specific (often antioxidant) responses within the stressed cells, making them better able to resist similar challenges in the future.

I am not ready to start recommending that patients begin drinking for hormetic health benefits. But for those who do drink, advising them to space their drinking out, limit the amount in each sitting to 1-2 drinks, and always pair drinking with a meal makes sense.  That is, unless and until new data tell us a different story…