Alcohol and Health: An Interview With Curtis Ellison

Henry R. Black, MD; R. Curtis Ellison, MD


October 30, 2014

This feature requires the newest version of Flash. You can download it here.

Addressing Mis-Science About Alcohol

Henry R. Black, MD: I am Dr Henry Black, adjunct professor of medicine at the Langone New York University School of Medicine and a former president of the American Society of Hypertension.

I'm here today with my friend and colleague, Dr Curt Ellison. Our relationship goes back 20 years, when we had a debate about alcohol at an American Society of Hypertension meeting. What is your role now?

Curtis R. Ellison, MD: I am professor of medicine and public health at Boston University School of Medicine and co-director of an international scientific forum that deals with emerging data on alcohol research.

Dr Black: What is the International Scientific Forum on Alcohol Research?

Dr Ellison: For the past 20 years, a group of us at Boston University have read and critiqued emerging papers related to alcohol and health. There is a lot of "mis-science"—mistakes that are preached to the public. If you pick up USA Today, it will tell you that a drink a day makes you live forever, or that a drink a day leads you straight to cancer. We try to look at the papers that have come out, and if it is something that we think is important, we write a critique on it.

Four years ago we expanded our work at Boston University and recruited 41 international scientists from Europe and other parts of the world—the United States, Australia, and New Zealand. These are people who work in alcohol, as epidemiologists, nutritionists, basic scientists, chemists, and so forth. We distribute studies that come out and seem to be important, either in supporting what we thought we knew or contradicting the data that we thought we had or which received a lot of publicity. We review these, usually within 3-4 days, and publish the critique on our website at Boston University, giving our opinion on whether the study is accurate and has been interpreted correctly.

Dr Black: You have said that we can drink alcohol in moderation. What do you mean by "moderation"?

Dr Ellison: Moderation is very difficult because people vary tremendously in how they react to alcohol. A chronic heavy drinker feels no effect with two or three drinks, but in an older woman who weighs 90 pounds, one drink can knock her off her feet, so it is very difficult to give a guideline that applies to everybody.

When people ask me, "How much should I drink?" I say, "Don't ask me. Ask your spouse or significant other, because he or she can tell if you are drinking too much." What our message seems to be now is not to drink too much. We have a lot of data on the evil effects of abusive drinking. What we are saying is that it should not be too little, either.

The Cardiovascular Benefits of Moderate Drinking

Dr Black: There are some biological reasons why alcohol could be beneficial, such as a reduction in platelet aggregation, a modest increase in high-density lipoprotein (HDL) cholesterol, and a reduction in C-reactive protein. How solid are these data?

Dr Ellison: The data on the mechanisms by which moderate drinking relates to cardiovascular disease are extremely strong.[1] We know from clinical trials—the first study was done in college students—that if you drink two beers daily for a week, your HDL cholesterol level goes up, and if you stop the alcohol, it comes back down.[2] We have tremendous data on HDL, other lipids, coagulation, fibrinolysis, inflammation, insulin resistance, and direct effects on endothelial function, so there is no question of the mechanisms that have been identified and verified in clinical trials and in experiments, both in animals and in humans.

We do not have a clinical trial where we take 50,000 people and assign them to have one or two drinks daily, or no alcohol, and see what happens over 50 years. That trial is not feasible, so we do not have the trial that looks at the effects of a given amount of alcohol when administered vs no alcohol over a long period of time in terms of health outcomes.

Dr Black: It's the same with smoking; we couldn't take 500,000 children and teach them to smoke.

Dr Ellison: That is exactly right, and it is the same with diet. We know that if you are on a good diet and body parameters change, you become healthy and you may lose weight, but we don't assign people to eat a certain diet and assign someone else to eat another diet and watch them until they die. These are lifestyle factors. It is not like taking a pill or a placebo. You have to base your best judgment on observational data and experimental studies on mechanisms.

The Impact of Alcohol Intake "Fibbers"

Dr Black: You have talked about the problem of underreporting, and apparently there is a way to judge that. What exactly is that?

Dr Ellison: Arthur Klatsky[3] at Kaiser Permanente in California has a huge database, more than 1 million people in the Kaiser Permanente studies. They looked at moderate drinking (one to two drinks daily) among their patients in relation to hypertension. Throughout their data they found numerous episodes of alcohol abuse in people who were asked, "How much do you drink?" and they said, "Oh, one or two drinks a day." The investigators then found elsewhere in the patients' records that they had been treated for alcoholism, alcohol abuse, liver cirrhosis, or another alcohol-related condition. So they said that these people are likely to be underreporters; they are not telling the truth.

They had other participants who reported one or two drinks daily, but who had no evidence of alcohol misuse. So they divided people into those who were likely to be underreporters and those who were not likely to be underreporters, and found big differences.

Hypertension, which we know is related to heavy alcohol use, showed up in those who were underreporting their alcohol intake but not among those who were likely to be giving a truthful response about their alcohol intake being only one to two drinks daily.[3] They have recently done this for liver dysfunction, and earlier this year they found that overall cancer rates were increased in people who underreported their alcohol intake but not in those who were truthful about their alcohol intake being one to two drinks daily.[4]

Is It True That "No Amount of Alcohol Is Safe"?

Dr Black: This is important, because some physicians feel that no amount of alcohol is safe with respect to some cancers. Some studies have focused on this, especially in breast cancer. Could you review that data for us?

Dr Ellison: The largest cancer study is the European EPIC study.[5] They continue to report that alcohol is a cause of cancer, and it is if you drink too much. You have problems with upper aerodigestive cancers—cancers of the mouth, throat, larynx, and esophagus. Indeed, heavy drinkers have an increased risk for those cancers.

But they also asked, "What about moderate drinkers?" In their last big report,[6] they concluded that of more than 100,000 people followed, the percentage of cancers caused by light to moderate drinking is far less than 1%. It is there because there are some data suggesting that women who have even one drink daily may have a slight increase in breast cancer incidence.[7] There is some evidence that one or two drinks could relate to colon cancer, and these are very common diseases.[8]

Most alcohol-related cancer is from heavy drinking. They have said that 1 of 1000 women and 3 of 100 men who are moderate drinkers are at risk, but they aren't able to determine whether these people are underreporting their alcohol intake. This is a new aspect that we have to apply to our large studies, and we can do that in large studies at Harvard and in such studies as EPIC, the Framingham Heart Study, and others that have repeated records and exams over many decades. We are beginning to do that now. We are finding the same thing in the Framingham study.

If you ever have evidence of alcohol abuse in any form, you have higher rates of death from cancer and total mortality. If not, we can take what you tell us and use the J-shaped curve to determine whether you have a certain level of risk. If you drink a little, you have a low risk, and if you drink a lot, the risk rises.

Dr Black: It is very important not to value studies that only have one observation compared with those that have serial observations. I hope we don't get stuck looking at studies that only have one observation.

What was considered one drink? Is it 12 oz of beer, 5 oz of wine, or 1.5 oz of 80-proof spirits? These supposedly contain 15 g of alcohol. Is that still the amount you use?

Dr Ellison: One drink is 12-15 g of alcohol, which is one glass of 4-5 oz of table wine, or 12 oz of American beer (or less, if it is Belgian beer). The problem is that if we take an animal and pour the drink in, we know exactly what we are giving, but people have great difficulty in estimating how much constitutes a drink. A mixed drink, a cocktail, can vary from half a drink to two typical drinks' worth of alcohol in a single glass. It can be difficult to know exactly how much alcohol is in a drink. These are general guidelines.

Still in Love With Red Wine?

Dr Black: Could you comment on what I think of as "a little less of a romance with red wine" than there used to be? What is the current feeling among the alcohol researchers about whether there was anything magical about red wine compared with other types of wine or alcohol?

Dr Ellison: The studies from the United States, particularly the large studies from physicians, nurses, and so forth from Harvard find little difference between the people who report drinking wine and those who report drinking beer or even spirits.[9] But in Europe or other parts of the world, where wine is more of an accepted type of beverage, studies show additional protection. However, the best data come from experiments in animals. If you give an animal the equivalent of gin or some other alcohol that has no additional polyphenols, you get a certain effect. If you give them the equivalent amount of alcohol plus the polyphenols in wine, you get better effects.[10] It is very clear from animal experiments that the polyphenols in wine, and to some extent in beer, tend to have additional protection.

When it comes to people, we are not measuring red wine, white wine, beer, or spirits; we are measuring drinkers, and the typical beer drinker and the typical wine drinker may be different in many ways. This is why it is so difficult to pin down the differences. In Europe, however, the data are quite clear: If you drink wine, your risk for coronary heart disease, stroke, and even dementia is considerably lower. For other conditions, such as diabetes, it seems that the alcohol is the main effect, so it is a combination that you do get in wine that you don't get in spirits.

Dr Black: So is there still a French paradox?

Dr Ellison: If you define a paradox as being that the French do a lot of things wrong, like smoke more and eat a lot of fat and yet have lower rates of heart disease, then yes. They still have a considerably lower risk for coronary disease than in the United States. The amount of wine being consumed in France is declining as the younger generation likes fancy beer or cocktails, but the same relation between people who consume some wine is not seen in France. But you can compare countries on many differences. It is much more important to compare physicians or nurses or people in the Framingham study, where they are all similar in many ways except that some of them drink and some do not. That is where we see confirmed differences by moderate amounts of alcohol.

Blaming the Puritans

Dr Black: You told me a quote from Abraham Lincoln. Could you share that with us?

Dr Ellison: I wrote it down because I wanted to quote him correctly. Abraham Lincoln was speaking to the Washington Temperance Society of Springfield, Illinois, in 1842 and he said, "It has long been recognized that the problems with alcohol relate not to the use of a bad thing but to the abuse of a good thing." I think that is true. Our whole focus on the evils of alcohol is the result of the Puritans getting to Plymouth Rock before the Italians got to America. In Northern Europe and in the United States, alcohol has always been viewed as evil, and the focus has been on—and a huge industry is built on—anti-alcoholism, anti-abuse, and the people working to prevent it, which we all agree should be done.

But the attitude has affected a lot of our research. I cannot send a grant application to the National Institutes of Health and say that I would like to study the benefits of moderate alcohol. They will say that they don't want to fund anything that encourages alcohol. In this country, it is very difficult to discuss it on a scientific basis and with a balanced view, because there is this underlying anti-alcohol feeling that persists. I blame it on the Puritans. I grew up as a Southern Baptist; alcohol was a sin then, and it still is in many parts of the country. This makes it difficult to get a balanced and scientifically sound message out to the public.

Dr Black: We even tried it with the Constitutional Amendment, but it didn't work.

Dr Ellison: I am taking part in a debate called "Should We Encourage Moderate Drinking for Its Health Benefits?" The question is, can you encourage anything that goes against the standard "just say no to drugs and alcohol" that we all grew up with over the past 30 years? It is difficult to send a balanced message. We are not going to get that wonderful clinical trial with 100,000 people and follow them for a long time. We would have trouble doing that anyway. The Tea Party would shut down the government because this trial is giving demon alcohol to half the people, and then the Liberal Democrats would shut it down because the trial is depriving one half of the people of this life-enhancing beverage. It is difficult to get balanced views. This is what our forum is trying to do.

Dr Black: Moderation is the way to go. We all think binge drinking is bad. We are all sure that excess alcohol intake is bad, but one or two drinks daily for a man, maybe, and one drink for women is certainly reasonable. It may well have cardiovascular benefits and doesn't seem to be associated with much of an increase in cancer risk.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.