In the wake of the surgeon general’s guidance on alcohol, what advice do you give your patients about drinking?

John Mandrola, Adam Cifu and Vinay Prasad answer

Last weekend, I asked John and Adam to put together 400 words on what advice they would give a patient who asked about drinking alcohol. I joined them in this exercise.

For the sake of discussion, we imagined a 47 year old man, no medications or past medical history, married with 2 kids, who enjoys 2 generous glasses of red wine a night. He asks (John/Adam or me) “Is that healthy? Should I cut back? or Should I Drink more?” We prepared answers without sharing, and we offer those to subscribers of Sensible Medicine as a thank you for your support.

If you don’t subscribe already, I hope you consider it. Finally, if you are interested in my review of the Surgeon Generals’ report, check out this post.

-Vinay Prasad MD MPH

Here is:

John Mandrola

The first words I say when discussing alcohol in the clinic is that I am not a preacher. My advice is from the perspective of a neutral health adviser….

What I would tell this man about his two glasses of wine per night depends on why he is seeing me. If he is overweight or obese and is seeking to lose weight, I would tell him that wine is full of carbohydrates and calories. Two fewer glasses per day would help him lose weight.

I would then ask about dose. Many people who cite their dose as 2 drinks per day underestimate their intake. If he had high blood pressure dose might be important. This meta-analysis of trials found that people who drink more than 2 drinks per night and reduce their intake, achieved a robust reduction of blood pressure.

I am a heart rhythm doctor, and if this man was seeing me for atrial fibrillation, I would have a lot to say about alcohol—because there is strong data.

• First, population-level data shows an association of AF and ETOH intake. Multiple meta-analyses have found that habitual alcohol consumption associates with a higher risk of AF. The association persists after correcting for binge drinking. And it occurs in a dose-dependent manner.

• Second, UCSF researchers showed in an RCT that alcohol infusions induce electrical changes in the atria that predispose to AF. This supports biologic plausibility.

• Third, Alex Voskoboinik, and colleagues in Melbourne, conducted an RCT in patients with AF who consumed alcohol. One group was randomized to abstinence and the other to continue their standard intake. Though it was hard to recruit for this trial, the abstinence group had significantly fewer AF episodes and lower AF burden.

Taken together, these data strongly support a causal role for alcohol in patients with AF. I tell patients about the data and they can choose what to do.

I know of no such data for moderate doses of alcohol in other conditions.

If this patient asked me about his wine intake from a general health perspective, I would first ask if he was drinking solely for health. If yes, I would say that the French paradox was chock full of biases and there is no evidence to support wine as a health enhancer.

If he was drinking wine for pleasure and asked about its health consequences, I would simply say that the observational data on general health outcomes are far too confounded to make any recommendations. When there is no data, I strive to give no specific recommendations.

---

Adam Cifu

I don’t envy the Surgeon General.

He or she is supposed to be “America’s doctor.” But what does that even mean? Should he be making public health recommendations or speaking to individuals? Sometimes these two tasks overlap, and when they do, the surgeon general can have an important voice. Luther Terry spoke out against smoking. C. Everett Koop educated Americans about AIDS.

Usually, the proclamations of Surgeons General lean more towards public than individual health. Their advice might save lives for the country, while doing little for the individual. Surgeon General Vivek Murthy’s recent recommendations regarding alcohol use and cancer fell into this category.

The Science

Excessive drinking is terrible for your health. It causes weight gain, leads to liver disease, raises cancer risk, increases the likelihood of motor vehicle crashes and other accidents, and can have profoundly negative social effects. The health effects of light drinking — up to a drink a day for the average woman, up to 2 for the average man — cannot really be known; there is simply too much confounding in the observational research. My take is that the overall effect is slightly negative – nearly meaningless on an individual level, measurable in 262 million American adults.

What I Tell Patients

For the small number of my patients who have had problems with heavy drinking, and/or have liver disease, I tell them they should not drink at all. The risk is just too high. Most of these people do not need to be told this. We spend our time discussing the best way to achieve abstinence: AA, acamprosate, disulfiram, naltrexone.

For the overwhelming majority of my patients — those who drink only a little and have no tendency toward alcohol use disorder — the decision about whether or not they drink is of little consequence. It barely affects their risk of illness. I discuss alcohol if they want to lose weight, sleep better, or improve their diabetes control. If a patient asks me if she should stop drinking, I tell her that it might lead to some small benefits, and I ask why she drinks. What do you enjoy about drinking? I then help the patient weigh the benefits of abstinence/reduction with the harms of giving up something they enjoy. Most people end up feeling like they can drink less.

A problem I find in most people who begin drinking less? They start eating more sweets.

---

Vinay Prasad

When doctors don’t have strong evidence, and don’t know the right answer, they should be silent. I don’t tell my patients whether to divorce their spouse, what school to send their kids to, or how to knit a sweater.

About alcohol, I say, “If it interferes with your work, how you parent or your relationships, it is a problem. If you have an underlying condition (like cirrhosis or active hepatitis), you might want to cut back, but if you enjoy it, we have no strong data to advise you.”

Alcohol has been intertwined with humans for thousands of years, and might affect 4 things: how long we live, how well we live (morbidity), our mental health, how we interact with others. Alcohol might have negative effects, but it may also have positive effects on these endpoints. Thinking only of mouth cancer, for example, misses the forest for the trees.

The available literature is largely observational and it compares people who choose to drink vs those who choose not to — more accurately, it compares people who report that they drink vs report they don’t. This is hopeless in several ways: measurement error, confounding, time-zero bias, multiple hypothesis testing and more.

It is profoundly American to wonder about the health effects of alcohol. Italians just know what wine pairs well with dinner, and which drinks are best before or after meals. Alcohol is best with friends and family. On my website, I provide 25 rules about alcohol that you may find entertaining.

There is a way to settle the alcohol debate. That is a simple randomized trial that enrolls people where they are (0,1,2 or 3+ drinks), and asks them to drink more, less, stay the course, and a few other options. The NIH almost ran such a study; until it came out they were courting Jose Cuervo to fund it. I would love to see such a study run under Jay B, without conflicted funding.

The Surgeon General is not chosen based on merit, and his swan song is a report that relies on low level evidence to make a controversial recommendation. The likelihood it will be a medical reversal — like eggs or butter — is quite high. He should have taken the honorable road and remained silent, and I think he is best ignored.

If you find 2 glasses of red wine is pleasurable, relaxing, helps lubricate conversation, and pairs well with steak, friends and laughter, then I say cheers, and I hope you are available for my next dinner party.