Is Alcohol in Moderation Helpful or Harmful?
How politics, cherry-picking, and bias made the benefits of moderate alcohol consumption disappear
The skeptical cardiologist has a confession to make. He enjoys drinking alcohol.
It puts him in a happy place where his frenetic brain stops fretting over the past and stops planning for the future. It lowers his inhibitions such that he becomes more loquacious, jocular, and silly around others. Occasionally it prompts chaotic dancing if appropriate music is played.
After some youthful experiences, he thinks he has discovered the dose and manner of consuming alcohol such that it enhances his overall experience of living without untoward acute and chronic effects.
Does my long-standing tendency to enjoy daily one or two glasses of wine or beer (plus an occasional cocktail) create bias when I review the overall health effects of moderate alcohol consumption? Conversely, are teetotalers (whether motivated by religion or philosophy) capable of unbiased analysis of data in this space?
My goal on this website is to present lifestyle, medication, device and surgical recommendations that are unbiased and deeply informed because there are few sources of information out there that are not biased or conflicted by special interest.
If it were clear that alcohol was a carcinogen or killer in any dosage or form I would have stopped drinking and broadcast the news from my office and website pulpits, therefore I feel I can make unbiased recommendations on the topic.
My interpretation of the scientific data until recently was that my level of moderate drinking was associated with lower cardiovascular events and lower all-cause mortality.
I changed my recommendations last year, writing that folks should be moving from moderate to minimal alcohol consumption after reading several recent papers that suggested that the scientific evidence had changed on the benefits and/or safety of moderate alcohol consumption. 1
Multiple media outlets have seized on these perceived changes with headlines trumpeting that “No Level Of Alcohol Consumption Is Safe”, a phrase the World Health Organization began proclaiming in 2022.
An article in the NY Times earlier this month suggested that the evidence was “evolving” such that worldwide scientists were concluding that the risk of alcohol consumption begins with the “first drop.”
The casual reader might conclude that, indeed, there is now a scientific consensus that any level of alcohol consumption is dangerous.
However, after delving more deeply into the data, I realized that there are no recently published earth-shaking original scientific publications on alcohol consumption justifying my 2024 change or buttressing those headlines.
All available data on this topic still comes to us from observational studies (MOOPS) which, as we have discussed ad nauseam on this website often attempt to overstate their findings, conflating correlation with causation.
I now feel that the preponderance of the scientific evidence when not cherry-picked and manipulated by scientists with anti-alcohol agendas still shows that moderate alcohol consumption is associated with less cardiovascular disease and a longer life than abstaining.
While it is clear that heavy alcohol consumption is not good for one’s health the real public health question is whether the previous “OK to drink moderate amounts of alcohol” recommendations should be retracted and replaced by “no alcohol consumption is safe.”
Does the dose make alcohol the poison? Or does the first drop start killing us?
Root Cause of Recent Alcohol Kerfuffle
In the last 10 years, a small group of antialcohol activists at the Canadian Institute for Substance Use Research (CISUR) led by Tim Stockwell have been manipulating data on alcohol and death and publishing papers in an attempt to eliminate the well-known J-Curve.
The J-curve shows that those with zero alcohol consumption have higher all-cause cardiovascular and mortality rates than those consuming up 200 grams of alcohol per week. Those who consume high levels (>/=500 grams/ week) have higher levels of mortality compared to those consuming between zero and 200 grams per week.
Christopher Snowdon has written two excellent Substack articles reviewing the statistical shenanigans (see here and here.) which eliminated the J-shaped curve and replaced it with a curve that shows mortality increasing starting at any consumption of alcohol and progressing monotonically upward as consumption creases. Writing in 2023 following the publication of a Stockwell systematic review and meta-analysis, he noted that:
“Stockwell has waged a one man crusade against the benefits of drinking for over a decade. He is past president of the neo-temperance Kettil Bruun Society and was involved - and heavily cited - in the process that lowered the UK drinking guidelines for no good reason. Along with his colleague Tim Naimi, he has since been at the helm of efforts to reduce Canada’s guidelines to two drinks per week. Despite being slapped down by experts in the field, he continues to produce countless journal articles casting doubt on the benefits of moderate drinking. This is his third attempt at making a meta-analysis that erases the J-Curve.
The Surgeon General Wields His Scalpel Against Alcohol
In early January, 2025 The Surgeon General of the US, Vice Admiral Vivek Murthy, issued his Advisory on Alcohol and Cancer Risk which claims that
Consuming alcohol increases the risk of developing at least 7 types of cancer.
The causal relationship between alcohol consumption and cancer is firmly established
You can download the full “advisory” or just take a look at this terrifying graphic.
Speaking of terrifying graphics, Dr. Murphy would like cancer warnings put on bottles of wine, beer, and liquor and suggests everyone stop drinking alcohol.
It is not clear who actually wrote this “advisory” or why it was so focused on cancer risk and ignored equally (or more) important cardiovascular risk and overall mortality.
The skeptic in me wants to point out that Dr. Murthy is likely a teetotaler based on his religious background as a Hindu.2
The giant leap the former Surgeon General made from weak observational data to a statement that “the causal relationship between alcohol consumption and cancer is firmly established” is baffling and must be considered a leap of faith.
NASEM Study Confirms the J-shaped Curve
Interestingly Congress last year contracted with the National Academies of Sciences, Engineering, and Medicine, or NASEM (authorizing) 1.3 million dollars for the purpose) of examining the risk of moderate alcohol consumption
The report, published earlier this year reviewed recent evidence on specific questions posed by the US Department of Agriculture to inform new dietary guidelines.
The report stresses the importance of distinguishing between moderate and excessive consumption. Major health risks, including the seven types of cancer associated with alcohol, are closely linked to abuse, not moderate consumption.
The NASEM review found lots of uncertainty in all aspects of alcohol and health but concluded that moderate drinking (≤14 g/day for women, ≤30 g/day for men) is associated with a 16% reduction in all-cause mortality, a 22% reduction in cardiovascular mortality, and a 10% increase in the risk of breast cancer.
You heard that correctly. They found that moderate alcohol consumption is associated with a lower overall risk of dying.
This table from the NASEM report shows that moderate alcohol consumption compared with never consuming alcohol is significantly associated with lower mortality in both men and women.
This lower mortality was primarily due to a lower risk of cardiovascular mortality.
The NASEM committee was not as impressed with the cancer risk of moderate alcohol consumption and determined that
no conclusion could be drawn regarding the association between moderate alcohol consumption compared with lifetime nonconsumers and risk of colorectal cancer.
No conclusion could be drawn regarding an association between moderate alcohol consumption and oral cavity, pharyngeal, esophageal, or laryngeal cancers.
Three Cheers for Nuance in Public Health Messaging
As we saw during the COVID-19 pandemic, public health organizations are very fond of simple messaging, even at the expense of the truth. They feel nuanced messaging leads to uncertainty and confusion in the public. Unfortunately, when the simple message is revealed to be false, the public loses trust in authorities.
We also see this in dietary recommendations. The AHA and major nutritional guidelines tell everyone to reduce their saturated fat consumption, irrespective of source when it has been clear for decades that saturated fats in dairy and chocolate are not bad for you and that no randomized trials have shown a cardiovascular benefit of lowering saturated fat consumption to less than 10% of daily calories.
I feel that those organizations (WHO, various cancer societies, Dr. Stockwell and colleagues at CSIRU) and the surgeon general are being influenced by their biases and/or a desire to send a simplified message to the public.
The toll that heavy alcohol consumption and reckless alcohol binging take on society in diseases and deaths ranging from motor vehicle accidents to liver cirrhosis to pancreatitis to head and neck cancers is huge. I agree wholeheartedly with public health organizations’ attempts to make this clear to the public and enact sensible and proven measures to lower that toll.
These organizations understandably would like to eliminate alcoholism, and heavy and/or reckless drinking. One approach is to decree that any amount of alcohol is toxic and dangerous and push for warning labels, heavier taxes, and perhaps initiate a total ban on legal sales.3
These approaches would punish those of us who gain satisfaction from alcohol and drink responsibly.
Two researchers from Harvard’s School of Public Health and Medical School (Eric Rimm and Kenneth Mukamal )who have studied alcohol and health for decades have been quite vocal that the simple and dire messaging that no amount of alcohol is safe vastly oversimplifies the existing knowledge about alcohol consumption.
They argue for more nuanced messaging in a recent article entitled “Is alcoholc good or bad for you? Yes:”
we continue to see reductive narratives, in the media and even in science journals, that alcohol in any amount is dangerous. Earlier this month, for instance, the media reported on a new study that found even small amounts of alcohol might be harmful. But the stories failed to give enough context or probe deeply enough to understand the study’s limitations—including that it cherry-picked subgroups of a larger study previously used by researchers, including one of us, who concluded that limited drinking in a recommended pattern correlated with lower mortality risk.
And most importantly we need high-quality randomized trials to answer questions about alcohol consumption and health. Torturing and rehashing observational studies will never get to the truth.
Those who try to correct this simplistic view are disparaged as pawns of the industry, even when no financial conflicts of interest exist. Meanwhile, some authors of studies suggesting alcohol is unhealthy have received money from anti-alcohol organizations.
We believe it’s worth trying, again, to set the record straight. We need more high-quality evidence to assess the health impacts of moderate alcohol consumption. And we need the media to treat the subject with the nuance it requires. Newer studies are not necessarily better than older research.
Until we get clear answers from good RCT data I will continue to advise patients that there are no compelling reasons to reduce their mild to moderate alcohol consumption if they are on balance noticing more benefits from such consumption than adverse effects.
All of us are capable of N=1 experiments to determine whether alcohol is interfering with sleep, mood, weight, or blood pressure. I highly recommend such trials. In my recent N=1 trials I have concluded that those four variables are improved at approximately one glass of alcohol per day versus either zero or two glasses.
Feel free to do your own N=1 alcohol trial and please report the results to us!
Nonabstemiously Yours,
-ACP
I am retracting the sentence in bold from “What is the optimal diet for a long and happy life?”
Ten years ago I felt the evidence showed that alcohol in moderation lowered the risk of dying, primarily by reducing cardiovascular death. Almost exclusively, the data for this came from observational studies which are inherently limited. More recent observational studies have suggested that any benefit on CV disease is minimal and only at lower amounts of alcohol intake.
In addition, higher intakes of alcohol are associated with an increased risk of certain cancers, predominantly nasopharyngeal ones, and breast cancer. Again, these are associations and don’t prove causality.
Higher intakes of alcohol are associated with increased recurrence of atrial fibrillation,
Keep in mind that moderate consumption is up to one drink per day for women, and two drinks for men (my apologies to women in general and the Eternal Fiancee’ of the Skeptical Cardiologist in particular) and be aware of what constitutes “one drink.”
Also, keep in mind that any alcohol consumption raises the risk of atrial fibrillation and that if you have a cardiomyopathy caused by alcohol you should avoid it altogether. In addition, excess alcohol consumption definitely causes liver cirrhosis, pancreatitis, and lots of lethal motor vehicle accidents.
Murthy was “sworn in as the 19th Surgeon General using his great great grandfather's Bhagavad Gita akin to when Congresswoman Tulsi Gabbard (D-HI) was sworn in as the first Hindu American elected to Congress.” (Hindu readers inform me that alcohol abstinence although commonly practiced is not a mandate of their religion.)
Google Prohibition to see how well this worked in the US previously
Love it. I also love a glass of wine maybe two. Thank you for your skepticism. A correction the name Murthy is changed to Murphy in a few places probably AI. The province of India Murthy comes from Alcohol is totally banned not because of religion but the poor people spend the earnings on alcohol. The negative side is moonshine has gone up & regularly scores of people die in villages. In India the politicians who ban alcohol in specific states get elected. No science behind it but populism. There is nothing in Hinduism against intoxicants except a large part of the population in India is Muslim & there religion does not approve.
This is very similar to Salt & Hypertension debate (By the way I am a Cardiologist& I do enjoy your skepticism most of the time).
Interesting post. Thanks!
A couple of comments:
I think it is worth remembering that the reason for drinking alcohol was to kill the potentially lethal organisms in water. Here in France it was wine, and even young children traditionally had a small amount of wine added to the water they drank. Continued in many households to this day.
In Germanic countries, including Anglo Saxon Britain, it was mead, ale and beer, as much as 8 pints a day consumed with breakfast, midday meal and dinner. No one dare drink water!
Much French cuisine includes marinading meat in wine or brandy, which was a good idea before refrigeration! And most meals would begin with an aperitif (strong alcohol) with various wines suitably marched with each course, and a brandy afterwards, all of which would be good protection from any slightly off food! So the health benefits might be considerable!
I saw some news report a few years ago that in America, cardiac emergencies soar every time there is a snowfall. Apparently it is directly related to unfit and overweight American men clearing snow from their driveway to get their car out - the consequence of cold weather, thickened blood and excess exercise.
On the same thread, when I lived in the French Alps, in winter the local workmen would meet in the cafe before work, around 8am, and take a cognac with their café, sometimes two. They'd then go off clearing snow and avalanches, cutting up fallen trees and dealing with the winter's problems. I later realised that the morning alcohol may be saving their lives!
In summary, I think the circumstances and wider picture of the use of alcohol is very relevant to the outcomes. I could add situations like stress (a known killer) and drinking to relax, and drinking to be more sociable (loneliness also a known killer), and drinking to help bonding, such as maintain marriages and relationships (married people live longer - never understood that one! 😬).
I too am sceptical of medical research, so often with an agenda, so often later proven false, especially that which ignores the complications and circumstances. There should be full disclosure on funding and corporate ties for research, and heavy fines for biased and deliberately false reports.
Just my view.