Perspective on alcohol use and cancer risk

ARCHIVED CONTENT: As a service to our readers, Harvard Health Publishing provides access to our library of archived content. Please note the date each article was posted or last reviewed. No content on this site, regardless of date, should ever be used as a substitute for direct medical advice from your doctor or other qualified clinician. 

As a primary care doctor who gets to know and follow patients over many years, I like to think of myself as a trusted health advisor. When it comes to advice about lifestyle issues that affect health, I rarely think in terms that are black or white. There are exceptions — it is hard to argue against the benefits of regular physical activity and nothing good can come from cigarette smoking. However, for most other things, including use of alcohol, things aren’t so straightforward. Now if I was a specialist who mainly saw the negative effects of alcohol misuse — such as liver specialists who care for people with acute and chronic liver injury or trauma surgeons who see the outcomes of drinking and driving — maybe it would be simpler. But for most of my patients, alcohol is a normal part of life, and has both good and bad attributes.

This is the background from which I view a new study on the relationship between alcohol and cancer. In case you don’t have time to read to the end, here is the bottom line: This study isn’t going to become part of my discussion about the pros and cons of alcohol consumption. For those who have the time, here’s why:

First, it isn’t because some of the cancers attributed to alcohol use aren’t serious––they are.

Anyone who has had a loved one with cancer of the esophagus (swallowing pipe) knows this to be true. The problem is in the context in which I counsel about the role of alcohol in my patient’s lives. For a few, alcohol is nothing but bad news, and this study doesn’t add to what we already know. For these individuals, the challenge is that I don’t have especially effective tools to help these patients remain alcohol free.

For others, the problem isn’t chronic abuse but bad judgment when they do drink.

So called binge drinkers can function very well day-to-day, but whether they drink once a week, once a month, or once a year, when they do drink they may not realize their impairment as they get in their car to drive home. For these patients, I assess their risk (for example, are they drinking more than they say they are) and spend a lot of time making them aware of the potential risk and discussing specific strategies to put in place well before the first drink is consumed.

For most of my patients in whom alcohol isn’t misused, the question is how to frame the health effects.

I’ll admit to an occasional glass of wine after a long day at work. Then, there is the so-called J-shaped curve of heart disease-related death saying that the lowest risk is in patients with moderate use, that is 1-2 beers (12 oz.), glasses of wine (5 oz.) or 1.5 oz. mixed drink a day. People who abstain entirely have somewhat higher risk, but that is overshadowed by much higher risk in heavy drinkers. I also caution that alcohol can be a source of unneeded calories and for many of my patients it is a simple way to eliminate some.

Though this new study doesn’t create any new data, it uses existing studies to argue that there is sufficient information to support a role for alcohol in causing cancer despite the problem that were not sure how because alcohol itself (unlike cigarettes) isn’t a known carcinogen. There is also the problem, as noted previously that some alcohol use may decrease risk of heart disease. Finally, if alcohol was causing cancer, wouldn’t we see correlations between population level alcohol consumption and cancer death rates? I’m not aware that such data exists. Without more information about the level of risk for those who drink in moderation and separating that risk from other behaviors that can go along with drinking, such as cigarette use, I don’t find this new evidence to be compelling me to change the discussion I have with patients.

How I’ll talk to my patients about patterns of alcohol use

Instead, I will continue categorize my patients based upon their pattern of alcohol use. Even if moderate alcohol use has some finite risk, there is also the question of how much of a risk it is—and how does that risk compare to other things they do or don’t do in their lives. I see this as a key role for primary care physicians. We want to frame personal choices — not enough activity, safe driving habits, unhealthy eating, and yes, alcohol use — in a way that provides perspective and hopefully motivation. My role is to advise about the things each of us can do to improve the quality and quantity of our lives. We all can do better, but I don’t think that scaring patients with the C word is the way to do it.

Related Information: Harvard Health Letter


  1. dor cohen

    Wonderful website. Lots of useful information here. I’m sending it to a few friends ans also sharing in delicious. And of course, thanks for your sweat!

  2. Cristina

    I think all doctors should only tell what statistics say. You are in no right to ‘moderate’ statistics. Patients are capable of deciding what they want when they know the what scientific studies shows as a risk. The doctors can interpret the statistics but never use personal preferences to easy a statistic truth. Then we are back in the 50’s when smoking doctors recommended smoking ‘in moderation’

  3. John Williams

    Ethanol is metabolized into ethanal which is a hard electrophile. Many hard electrophiles are carcinogens: Diazomethane, Bischloromethylether (Rohm and Haas tragedy), and epoxides formed in the metabolism of polycyclic aromatic hydrocarbons (colon cancer found in chimney sweeps disease) to name a few. The hard electrophilic epoxide in a PAH is attacked by the hard nucleophilic primary amine in adenine in DNA resulting in a permanent insult to the DNA. We need to avoid hard electrophiles in our environment.

  4. Ed

    Here is the crux of human behavior.
    People have differing vices and some of those involve tobacco, alcohol, food, drugs, etc.

    The question is WHY do they use these substances or vices. Some drink or smoke to reduce stress or to cope with other life problems be it work, family, life or health related.

    Life is about choices which we all have and make whether we are aware of them or not!!

    Life is full of risks some of which we can control and some of which we cannot.
    We CAN control what we do to our bodies- diet, lifestyle, exercise , etc.

    Others are out of our control – accidents, floods, fires, other acts of God/Nature, genetics, etc!!

    If you have bad genes you have to work harder to stay healthy versus someone with good genes.

    That being said you still control the ” lion’s share” of your health by the choices you make every day. Sure everyone knows of someone who drinks like a fish and smokes like a chimney and lives a long time but these are the ” exceptions” not the rule and most people I know or have known who are chronic smokers and/ or drinkers have died from cancer or liver failure before their time.

    Some luck out and it doesn’t catch up with them until much later in life but in the end it ALWAYS catches up with you!!

    • Steven Atlas

      I appreciate the many and varied comments. Ed has nicely restated a key point I was making. In discussing health related goals with patients, it is helpful to explicitly frame lifestyle choices. Knowing each patient’s unique history including family risk and personal factors, one can discuss how to get the most out of any lifestyle changes being considered. There are choices that are risky and I need to point them out. But my role is to advise and I continue to care for individuals who make choices I wouldn’t. At the present time, personal genetic information beyond what is available from family history and routine screening hasn’t been shown to improve outcomes. I hope that in the future, such testing may help, but as several have pointed out lifestyle choices we make seem to contribute more to our long-term health than our genes.

  5. Donjo

    We can’t eliminate risk from our lives — skydiving, driving a car, and walking down a flight of stairs all entail some risk. A rational strategy is to identify and quantitate risks, mitigate them where possible, and make appropriate personal risk/benefit choices.

  6. Fred Woolman

    Cutting through all the … stuff … the real question is:

    Why do people even need alcohol?
    The answer is the same as for cigarettes — we don’t.

    So don’t drink it … End of story.

  7. Philo Fido

    Rather than smuggling morals about a way of life void of alcohol, what about having science fixing the health issues related to alcohol so that we can improve our well being by drinking and smoking… not to mention other great consumptions that may lead people to think in a subversive manner… gosh stop being a bunch of puritans !

  8. Aslam

    Our Prophet advised Muslims 1400 years before that do not drink Alcohol as it is not good for health. Now Science has proved and seconded it as well.

    • Philo Fido

      Your prophet is just another moralistic nerd.

    • Patrice Pierre Henry

      However, It does seem that the muslim population is less subject to cancer issues.

      • Abass

        That’s true!!

      • Ak2016

        Please share the stat or source of data, if you know. In general, the average life span among the middle eastern Muslims is lower than in Western Europe or China or Japan where people do use alcohol. Cancer is related to issue of life style. All societies/ population groups have some risky life/food style behaviors.

    • AK2016

      The drinks made 1400 years ago in the Saudi Arabian desert were primitive, unhygienic, and unsafe. For the same reasons, Before Muslims, the Catholics, the Buddhists, the Hindu’s had the same injunctions. The issue here is of moderation and safe use – rather than no use.

  9. JBernardes-Correia

    I think that talk about alcohol is not the same as speaking of drinks containing alcohol, as well the message is unfair in relation to fermented beverages such as red wine and beer contain both substances (eg. polyphenols). There is enough scientific evidence that points to benefits for their customers responsibly.

  10. S. Angevine

    Many published studies link sexually transmitted forms of human papillomavirus with a growing number of cancers.

    People who drink excessive quantities of alcohol tend to engage in risk taking behavior – out of control sexual intimacy with numerous partners.

  11. Sesame

    From what I’ve read over the years, nothing much has changed when it comes to what causes cancer. A mutating rogue gene that infects other healthy cells around it. Causes are apparently just poor individual health management and genetic defects. Bad diets, poor choices, where to live, pollution, chemicals, and on and on and on. We have better and more ways to treat cancer but the sources are still somewhat of a mystery. And, then there’s the business of cancer cures. We have all heard about large organizations that turn bad and spend the money on themselves, just plain thieves (much like our politics) need the money! I’m pretty sure I’ll get some form of cancer before I die. I hardly worry about why but more about “then what”?

  12. Kelley

    I believe that many insults to any body part can reduce the chances of remaining healthy, and that alcohol is one factor that decreases certain body parts from remaining healthy. If a person drinks alcohol in excess, alcohol alone may be enough. But what if a person who drinks in moderation also eats pickled foods, smokes cigarettes or gas GERD, combined this is a person asking for trouble. Add in some stress, chronic lack of sleep or a compromised immune system and something is most likely to go wrong unless that person’s homeostasis is so responsive that they are unaffected. I have great respect for primary care doctors, and for Harvard University, however I am happy that you are not my PCP. If people don’t know all the facts , they will not make the changes. Chances are, your patients are NOT telling you the truth , the whole truth and nothing but the truth anyhow; we as patients don’t stroll in to our annual appointments and say , ” yes Doc, I went on several benders this year. I do it quite frequently.” . All active alcoholics deny alcohol abuse , and most family members are trained to deny any drinking problems within their families. And I would also like to know why the word cancer is being called the “C” word by anyone in the medical field.

  13. Debbie

    Alcohol and cancer is a subject worth writing on and bringing to the public attention and conversation. My sister got breast cancer and on the advice of her doctor did genetic testing (although we have a 4-generation genealogy with zero breast cancer) and amid a flurry of testing there was absolute silence on the subject of her 30+ years of daily alcohol consumption (currently up to a pint or more of vodka per day). I’m sure smokers used to live in the same comfortable state of denial until public awareness would no longer allow it.

    • Mary

      I totally agree with you. I tell my children that anything you do to abuse your body will catch up with you at some point in your life. I work in healthcare and I see it all the time.

  14. Wacho

    Is there anything left to condemn?

  15. William D. Carr

    What a useless article; it only states what any reasonable person already knows! A waste of time to read it, and, I suspect, to write it!

  16. Donald Campion

    Better a few beers than a few cokes.

  17. Fred

    I think that the danger is the association with cigarette. Because cigarette burns the throat and enables the cigarette to penetrate in the tissue. Alcohol increases the rate of insulin and sugar could be the engine of inflammation.

  18. Kathy

    As a toxicologist I would argue that we know enough to help many patients make wiser choices. Alcohol is metabolized by an enzyme family that substantially determines peoples’ reactions to alcohol. Genetically-determined variations in alcohol dehydrogenase help clarify why some populations have lower alcohol tolerance (e.g., people of Asian ancestry). Moving beyond expensive genetic testing, the advent of “23 and Me” and other low cost genetic testing options can inform people about their genetic profile and some related susceptibilities (based on cited peer-reviewed literature). I hope clinicians will soon access these resources and work with their patients to better understand their risk factors and strategies for risk reduction. It requires additional work, but a wealth of information is now available that can inform our approach to alcohol and many other hazards, in concert with the traditional medical tools.
    Note – I have no financial interest in genetic testing companies but have used genetic data in public health investigations.

  19. akash

    Thanks for this informative article. Happy rakhi

  20. Matt

    Alcohol is a group 1 carcinogen – the same category as asbestos, tobacco, and other known cancer causing agents. Add to that the acetaldehyde (another group 1 carginogen) which is thought to be a product of alcohol processing in the body. For people with a family history of cancers of the mouth, esophagus, larynx, breast, stomach, liver and colon (which alcohol is a proven causative factor) patients should have good information on how to reduce their risk – by drinking less or not drinking, for example. Nobody wants to associate America’s favorite pastime with cancer, but it took decades for consumers to learn the truth about another significant cause of cancer – tobacco. Alcohol companies, like tobacco companies, do not want consumers to know the truth – that alcohol related cancers cause about 5.8% of all cancer deaths each year. It may not sound like much, but do the math – it is significant. While it is not as clear as tobacco (being definitively harmful in every way), patients who consume alcohol deserve to know the facts so they can make informed decisions about their health.

    • Donald

      The problem with most people in the FDA and the DEA is that they are alcoholics who consume more than 5 ethanol-based drinks per week. Their judgement on other substances, like Medical Marijuana, is skewed by their lack of ethics.

      • Kelley

        Does the FDA have anything to do with alcohol? Alcohol falls under the government department known as “alcohol, firearms & tobacco”. If the government really did indeed keep products off the market that are known carcinogens, believe it that cigarettes would be illegal.

      • Philo Fido

        The lack of ethics is in your scientific grounds/assumptions to your ethics… Science doesn’the say that you should not drink and smoke… Science doesn’t exclude that it is OK to drink and that related illnesses could be fixed by science while admitting that well being is a fact derived from drinking and smoking. There are no moral imperative from science!

    • Bossy

      Well said. No wonder the religion of the Muslims completely forbid the use of alcoholic drinks for its followers.

    • Dave


      So you can look up Group 1 carcinogens. But your comment proves you can’t think about what you read.

      Estrogen therapy is also group 1. So are birth control pills – but it’s noted that they also have a strong cancer-preventing effect against other cancers (ovarian, uterine).

      Oh, don’t look now, but mineral oil is a Group 1 – and we still put that in baby lotions and cosmetics.

      Hey, what about diesel exhaust and outdoor air pollution? Processed meat? Do you ever eat meat, Matt? If you are a vegetarian, than why not rant about the dangers of meat.

      Sunshine – damn, hard to avoid that one.

      Point is, given the NCI’s analysis of risk factors, alcohol risk, like others, is strongly correlated with volume of use. So the OW opinion that he will nuance his discussion about total lifestyle and recognizing how little he can change some things in his patients only make sense.

      Painters are in Group 1 – would you advise them to give up their job and go on welfare because they are exposed to a Group 1 substance? N0 – you would counsel them based upon what is possible, what their options realistically are, and maybe advise them how to limit exposure within real terms for their lives.

      The fact remains that the vast majority of smokers still do not get cancer (though there are other impacts) and the vast majority of people who drink do not get cancer.

      Get real is putting this into perspective. If we banned all known carcinogens, we still wouldn’t have eliminated natural and other carcinogens and we’d still have to work on overall quality of life.

      Lighten up on the Doc!

  21. Wadiyo

    thank for your shared.
    How to preventif cancer impact?

Commenting has been closed for this post.