A Ground Truth on Alcohol Intake

My review of the 3 major reports and podcast with Dr. Vivek Murthy

While we’ve largely moved on from the French paradox myth of red wine’s protection from heart disease, there is still considerable controversy as to what level of alcohol intake is safe, potentially beneficial, or harmful. In this edition, I’m going to take you through the three recent reports, how they converge and differ, and then we’ll turn to my recent conversation with Vivek Murthy, who issued the Surgeon General report just before leaving that office in January. After that I’ll pull it together with some concluding remarks.

N.B. The topic of alcohol intake, a major component of our lifestyle factors, is discussed in depth in my new book SUPER AGERS, out May 6th. At the bottom of this long post there’s more information about the book.

The 3 New Major Reports: Brief Description of Purpose and Objectives

1. The National Academies

The report was requested by Congress and is unique among the three because it was a de novo comprehensive review of studies published in the past 15 years. The team raking through all the studies made a key call to address the potential confounder of “abstainer bias”—excluding former drinkers from the never drinkers subgroup. This was intended to provide a more accurate differentiation between moderate drinkers and lifelong non-drinkers. The data review for each endpoint was qualified with a level of certainty. The scope was not just cancer; it addressed neurodegenerative and cardiovascular outcomes, maternal alcohol consumption, all-cause mortality, and effect on body weight. Since there are no randomized controlled trials, no conclusion reached the level of high certainty. The report is 254 pages long with naming of its 14 authors and external review panel members of 10 experts.

2. The Surgeon General Report

The 20-page report was solely related to cancer risks. The evidence presented for risk of cancer presented was drawn from a few peer-reviewed publications, one from the British Journal of Cancer, published in 2014, that included a review of 572 studies, nearly 500,00 cancer cases of 23 different types. Another from the British Journal of Cancer 2021 of an Australian cohort age 45 and older with over 17,000 cases of cancer, and a paper in Lancet Oncology, published in 2021, presenting estimates of global data.

3. The ICCPUD Draft Report

Notably, ICCPUD stands for the Interagency Coordinating Committee for Prevention of Underage Drinking. Its focus was on the broad risks for youth and adults, not just cancer but included cardiovascular, neurologic, infectious diseases, mortality, and injuries. The 81-page report and analysis, prepared by 6 authors, relied on modeling based on previous systematic reviews and national datasets. Like the National Academies, it was externally reviewed, but the names of the experts was not revealed. Unlike the National Academies, it excluded studies using Mendelian randomization or randomized trials and it did not exclude abstainer bias.

Two of these reports (1 and 3) have been attacked for conflict of interest among their authors (a bias for either pro- or anti-alcohol).

Where The 3 Reports Agreed

Breast Cancer

My graph (semi-quantitative based on the 3 reports) or breast cancer risk as linked t alcohol intake

Beyond concordance that high consumption of alcohol poses health risks, and gender specificity (difference in amount of alcohol intake and relationship to risk for men vs women) the only main point of agreement for risk of alcohol intake and outcomes was for breast cancer in women, as seen in the graph above. Even with that you can see there was not agreement for the level of alcohol vs the magnitude of risk. For the National Academies report, the risk was only seen beginning with 1 drink/day among women. The Surgeon General and ICCPUD reports were concordant except for the magnitude of increased risk.

Other Cancers

Both the ICCPUD and Surgeon General reports found increased risk, beginning at low levels of intake, for several cancers: oral, pharynx, larynx, liver, esophagus, colon. But the National Academes did not agree that the evidence for these associations was adequate, exemplified by the graph below for colon cancer, a non-significant ~9% increased risk (95% confidence intervals ranging from 0.96 to 1.22)

Where the Reports Disagreed

The National Academies concluded that moderate alcohol intake (2 drinks /day in men; 1 drink/day in women) was associated with reduced all-cause (16% reduction vs never consumption, see Table below) and cardiovascular mortality (~18% reduction). The other 2 reports focused on risks and did not report any benefits. In contrast, the ICCPUD concluded that 7 drink per week increased the risk of dying in 1 per 1,000 people and >9 drinks per week increased the risk to 1 per 100 people.

Below is the Table for reduction of mortality in the National Academies report which interestingly was using the same meta-analysis data source presented in the Surgeon General’s that did not touch on mortally reduction.

Not Included in the Reports

The International Agency for Cancer Research and WHO label alcohol as a carcinogen and assert that there is no safe level of alcohol intake for risk of cancer. In Ireland there are mandated labels: “There is a direct link between alcohol and fatal cancers.” Other countries such as Canada and Thailand are considering doing the same.

I mentioned that the ICCPUD review exclude studies using Mendelian randomization (MR) which actually are quite important because they have the potential to demonstrate or support causality, unlike observational studies. Notably, not it any of these reports is one such MR study with evidence of cause and effect for high intake of alcohol and hypertension and coronary artery disease (Graphs below, note these are log plots). I reviewed all of this in SUPER AGERS.

From Biddinger et al and SUPER AGERS, Figure p. 33 book...

My Podcast With Dr. Vivek Murthy

I discussed his report and the marked discrepancies with the others, particularly the National Academies.

Audio file (also available at Apple and Spotify)

Full video is on YouTube

Transcript with links

Eric Topol (00:06):

Well, hello. This is Eric Topol from Ground Truths, and I'm delighted to welcome Dr. Vivek Murthy, the 19^th and 21^st US Surgeon General recently, moving on to other things, we'll get to that. And a prolific writer, including the book, Together: The Healing Power of Human Connection in a Sometimes Lonely World. And what we're going to mainly talk about today is the report that was published by him and Office of the Surgeon General from HHS on January 3^rd about alcohol. So before we get to that, Vivek, I just want to mention,

I recently saw a patient and his wife asked me during the clinic visit, she said, is it okay for my husband to continue his moderate alcohol drinking? So I said, well, what is that? And she said, well, he has two tequilas and six beers every night. And I said, are you kidding with me? And she says, no. Does that fulfill moderate? I said, no. No, that's not moderate alcohol.

(BTW a drink, 14g of alcohol, means 12 ounces of 5% ABV beer of 5 ounces of 12% ABV wine)

Vivek Murthy (01:18):

Oh my goodness.

Vivek Murthy (01:20):

I'm really glad she asked you, Eric, because I think sometimes we assume these terms mean the same to everyone, but when people bring it to thoughtful clinicians like yourself, they can understand perhaps what's moderate, what's not, but wow.

Eric Topol (01:33):

Yeah. I think one of the things you touched on with your report and that we can drill down on is there's just very poor awareness about alcohol, levels of risk, just so much confusion. So it's really good that you kind of brought this to the fore. Now what's fascinating is the timing. So there were three reports within a month. On December 17^th, there was a National Academies report. You issued your report on January 3^rd on that one and we’ll come back to that in a second. And then there was the Interagency Coordinating Committee on the Prevention of Underage Drinking (ICCPUD) of HHS, which put out the report 11 days after yours in January, and all three reports had somewhat differing recommendations and findings. But I was going to ask you first, when you put your report out on January 3^rd, was that just to avoid the holidays and New Year's Eve or was that just when it was ready?

Vivek Murthy (02:37):

Well, the reality is, if it had been ready significantly earlier, I would've wanted it to come out earlier because one of the things I've learned over time with the reports we put out is that there's an initial reaction in terms of press coverage, et cetera, but then it's in the weeks and months after that, you get to have a lot of rich discussion with the public. Unfortunately, this took a bit longer to get ready than we had hoped, and so it ended up coming out a bit later than I had hoped. But I'm happy to see that it is still contributing to conversation on this topic. I still to this day get messages from people around the country who encounter the report or saw news coverage of it and are reassessing perhaps their own relationship with alcohol. And that is one of the most important things I think, to come out of reports like this. Sometimes in health and medicine, as you know better than anyone, Eric, it's not always black and white, but when we have new information, we want to get it to people so that they can use it as they make decisions and balance benefits and risks in their life.

Eric Topol (03:44):

Well, and this is something that's been progressing over recent years, as you know, both in Ireland and South Korea, they have warning labels on all alcohol products and other countries are considering doing that. And in fact, in Canada, they markedly changed the threshold for recommendation recently, and there was a paper that we will link to in Lancet Public Health that showed that warnings can be effective for reducing alcohol consumption just like we saw with tobacco. So there are some reasons, and we know historically the WHO has classified alcohol as a carcinogen because no question, at very high levels of intake, there are links to various cancers that we'll get into. I guess the biggest question is the disparity in the findings among these three reports that are recent because they were looking at data that were published. We'll talk about the flaws potentially in that data that are somewhat difficult, but how do you account for whereby the National Academy says, well, if you're less than two drinks per day for men or less than one per women, you actually have a lower risk of all-cause mortality and cardiovascular mortality with moderate certainty and that only higher than that threshold do you start to see risk.

Eric Topol (05:10):

Like for example, they found the 10% increased risk of breast cancer, whereas your report as well as the subsequent report that I mentioned had very different findings mainly related to cancer risk. How do you process all that?

Vivek Murthy (05:28):

Yeah, it's a good question and I am glad we're having this conversation because I think sometimes the scientific process of looking at data, doing analysis of it, even if it's directionally consistent, sometimes differences in around the details, around the edges can actually really impact how people view these reports and can actually show confusion. So let's talk about this for a moment. I want to start by talking about what is actually consistent between these three reports, which is that all three reports actually do find that there is this link between alcohol and cancer risk, and we'll get into some of the differences there, but they all generally agree particularly for breast cancer, that there is a connection here. This is building on everything you said, Eric, which is other countries already moving forward with changing warning labels to include cancer risk on alcohol containers, and in some cases, like in the case of Canada, changing their recommended upper limits of drinking, if you will.

Vivek Murthy (06:31):

But what's also interesting is that within the United States, for years we've had medical societies trying to ring the alarm bell, if you will, about the risk associated between alcohol and cancer. And for example, the American Society of Clinical Oncology, the American Medical Association, and others for years have been talking about this link with concern. And I say link, but I'm talking about a causal link. And so, this is all in the background. When I was a surgeon general for my first term, Eric, back in 2015-16, we had published a report on alcohol, drugs and health and had actually noted in that report, the link between alcohol and cancer risk. Even the dietary guidelines published in 2020 spoke to that link between alcohol and cancer and recommended that no one should start drinking for health reasons. So that's just some important background on cancer.

Vivek Murthy (07:24):

Finally, let me just talk about what you mentioned, which is some potential areas of discrepancy around mortality, around cardiovascular risk. This story that many of us were given during medical training years ago, decades ago was that alcohol is good for your heart and so a couple of drinks a day is good for your heart and go for it. And in fact, if you don't drink well, maybe you should consider starting. That advice along with some other pieces of advice perhaps we were led to believe during our training, turns out actually not I think to be warranted based on the data that we have now. And again, this part is consistent even with the last set of dietary guidelines, the notion that you should start drinking for health benefit is generally not something that experts in this space and organizations and countries have recommended, including ours.

Vivek Murthy (08:15):

Where it comes some mortality, there are actually multiple studies including the HHS study that you referenced that was put out in mid-January, which have showed actually that there is an increase in mortality with increased consumption of alcohol including below the two drinks a day, which we think of as “moderate”, but which I think we have to reconsider. Now, the National Academies report had perhaps a different take on that, but I actually think the National Academies report is the outlier here. And when you get into the details of how that report was done, it actually considered a narrower dataset both for looking at cardiovascular benefit, cancer risk and mortality benefit compared to many other broader based studies that have been done, including reviews and meta-analysis. And some of those decisions had real consequences for diminishing the risks that was found. And particularly when I think about the cardiovascular risk, which you mentioned, I think it's fair to say that the scientific community is reassessing significantly that many of us were taught that alcohol is unconditionally good for your heart.

Vivek Murthy (09:26):

I think what has come about is there's a real chance that alcohol consumption at lower levels may and I emphasize may reduce the risk of ischemic events, so specifically heart attacks and perhaps ischemic stroke, but the magnitude of that benefit is likely lower than what we thought before. And also, what's important to note, and this is so important for public communication, and this is a place where I do not think we have done a good job in medicine and public health, is that as you know better than anyone, the heart and heart disease is more than just about ischemic disease. We care about hypertension, we care about heart failure, we care about arrhythmias like atrial fibrillation. And in all of those situations we see the risk increase of hypertension, heart failure, and atrial fibrillation with alcohol consumption. So if we want to talk to the public about this, I think it's fair to say that the effects on the heart are mixed. We do see that some types of heart disease and factors that contribute to heart disease get worse with drinking. It's possible that other forms of heart disease like ischemic heart disease, there may be a modest benefit, but how that all washes out together is unclear. And so, the notion that you should start drinking for heart benefit I think is not supported by the data and I think is a dangerous piece of advice to give people.

Eric Topol (10:55):

No, I think it's important because these reports largely were centered around cancer, yours and the subsequent HHS, and didn't really emphasize the cardiovascular risks that I'm in touch with. As you mentioned, atrial fibrillation, alcohol cardiomyopathy, an important form of heart failure and hypertension. So no question about that. The curves for cancer, as you pointed out, we will link to the nice figure you have of the seven types of cancer associated with alcohol in it. And in the HHS report, they actually had the kind of quantification curves where 14 grams is one drink approximately, and they showed how these relationships were across the different studies, and it wasn't any question that with oral cavity cancer, larynx cancer, esophageal cancer, throat cancer, these take off very quickly, even one drink a day, 14 grams a day, but you didn't see much in colon cancer for that and for liver cancer among women, it was there.

Eric Topol (12:11):

But what was interesting is the breast cancer wasn't really seen in premenopausal. It was only seen in postmenopausal women, which is interesting. And it was not nearly as accentuated as some of these other cancer and drink quantitative relationships. Well post that, those curves. But I wonder if you could point out, because the threshold here for the HHS study, they were talking about over seven drinks per week, whereas in your report you were comparing these other increased groups of risk to one drink per week, which is getting down there to almost no drinks per week. So it doesn't seem like the relationship holds when you get well below, unless you start to get to one drink per day or seven drinks per week. Could you comment about that?

Vivek Murthy (13:05):

Yeah. So this is a place where I do wish from a data perspective, Eric, that we had, we always want better and richer data. The kind of data I would like is really granular data about where I can compare people who have one drink a week, two drinks a week, three drinks a week, four drinks a week. So you can really start to understand where does the risk take off? And from the data that we aggregated and looked at and the other studies that we pulled from the United States and around the world, what we saw is that the increased risk or where risk started arising was somewhere between one drink a week and one drink a day, right? Now, one could say, and this is a very reasonable question, does that mean I should aim for one a week? Does that mean three a week is okay?

Vivek Murthy (13:51):

What about five a week? Is that all right for me? And this is where I think it gets tricky, and I think about this as you do as a clinician, you explain meeting a patient at the bedside, you're talking to them trying to give them practical advice on their life. And this is I think where there's a challenge between population and individual data. We know that these risks are real. We know at a population level that when you move from everyone drinking, let's say one drink a week to everyone drinking one drink a day, you will see meaningful rises in cancer cases across the country. But what does that mean for an individual? I think about this a bit differently and the way I've talked to friends about this as well is if you're somebody who's particularly concerned about cancer risk, if you've got a family history of cancer, if you have a personal history of cancer, if you have a history of exposure to carcinogens and hence are worried about increased risk now or in the future, those all might be reasons to be particularly cautious with your level of alcohol consumption.

Vivek Murthy (14:47):

And the one thing we do see in the data is more alcohol means more cancer risk. And as you pointed out with, we see, especially when you're consuming at higher levels, a real increase in the risk of cancer. So that's how I would advise people just to think about this. And I think for somebody like, let's say like my wife who might have a drink on a birthday or on a special event from here and there over the course of the year, do I want her to be scared that somehow she's now in imminent danger of developing cancer if she drinks on a birthday or on a special holiday? No, I don't want her to be concerned about that. But I do think that the people should know the numbers and they should know what we're seeing in the trends. But I finally also want to draw attention to subpopulations.

Vivek Murthy (15:34):

I mentioned people who have a personal history of cancer or risk factors like the BRCA gene or others, but let's also talk about Asian Americans for a moment. Many Asian Americans may experience that flushing sensation when they drink, and we now understand more about why that comes about. But it turns out that for populations that experience those symptoms, they often aren't able to break down the sort of byproducts of alcohol as it goes through our body as well as somebody else might be able to. And what that means is that products like acetaldehyde, which we know are carcinogenic, may hang out in their system for longer and increase their risk significantly of cancer. And that's one of the reasons why not only in our report, but in others, there's particular attention drawn to Asian American populations who may experience that flushing sensation. I remember when I was in college, people thought that that was a source of entertainment. It was interesting, and they thought, we'll just take some over the counter medications and it'll get rid of that, and we'll be fine. But we know it's not that simple. It's not just about what you feel on the outside. It's about what's happening on the inside that could be contributing to increased risk.

Eric Topol (16:40):

Yeah, those are important points. I mean, I think the individualized assessment is critical. We'll get to that in a minute. Regarding things like Mendelian randomization studies, and also as you said, people that have already embedded risk, at least known risk we’ll know more about that in the future, I think as we get a better handle on individual risk of cancer and other conditions. But I wanted to get to the problems with these studies in general. As you know very well, these are observational studies, some of them, very large, lots of them of course, but we don't have, and we won't have a randomized trial of people randomized to drinking or not drinking. So this is kind of all we're going to have. And there's lots of confounders because people in higher socioeconomic strata might drink more because they can afford it and because social issues or whatever, who knows?

Eric Topol (17:43):

There's all these confounders. And the problem also is as we've seen the reports of pretty important conflicts of interest. So for example, in the National Academy study, there was concerns that there were people who were favoring alcohol intake that were part of that and connected the big alcohol industry lobbying force, whereas there was just the opposite with the HHS report of people having potential conflicts because they were tied to the anti-alcohol entities. So it's hard to interpret. And then you add to this, and yes, you highlighted in your report there's a big difference, particularly here between relative risk, which looks alarming with absolute risk, which is oftentimes quite small. So with all of that, those factors, many people have interpreted these studies with their priors, their cognitive biases. I like my drinks and with all those uncertainties, I'm just going on as I have and not really being convinced that there's anything new here because of the fuzziness. Could you comment about that?

Vivek Murthy (19:11):

Yeah, as usual, you're bringing up really thoughtful and important points here, and you're right that the data is not as perfect as we'd want it to be. You're right that it's really tricky to conduct prospective trials. And I would even say that given what we know about some of the potential harms of alcohol, whether it's contributing to injury, intentional and unintentional, whether it's contributing to other forms of disease, I think it would be challenging for an IRB also to consider how you would do this in a way that would be ethical. So for a bunch of reasons, I think it's challenging to get prospective trials. So we are left with having to use human data that's observational or retrospective, having to use animal data and having to use laboratory based data to help us on mechanisms of harm. And that's not perfect. But I think it's also worth noting that decades ago when my predecessor as Surgeon General Luther Terry issued the first Surgeon General's report on tobacco, he was actually faced with a similar circumstance where there had not been large perspective trials of tobacco consumption.

Vivek Murthy (20:21):

And in that time, they put the data together, used criteria like Bradford Hill criteria and others to help do everything they could to increase their confidence about the conclusion they were coming to. And it was on that basis that they actually came out and said, there's a connection here between alcohol and in that case lung cancer specifically. So I do think it's reasonable for people to say we need better data here. I think it's reasonable for people to speak to some of the limitations of the data that we have, but I think it's also important that we look at this data in aggregate and over time, and it's in that respect that I think we see that the data has been continuing to build over time, pointing to the risks associated between alcohol and cancer consumption in particular. But also, when you look at the preponderance of the data, specifically when it comes to certain cardiovascular conditions like arrhythmias or hypertension or heart failure there, it has been largely consistent.

Vivek Murthy (21:22):

There is not data that I have come across in these studies that seems to indicate that drinking is good for your heart failure or that it reduces the chance of arrhythmias. So this is not easy to say, but I think in medicine in particular, one of the things that you and I and other clinicians are used to is having to make decisions and having to make recommendations based on imperfect data when there are real concerns at hand. So for example, when somebody is coding in the hospital, we use the best data that we have and we augment that with our best judgment and we make a decision because a patient needs that. They can't wait five years for a clinical trial to take place. And here in this case, I think when you looking at the preponderance of data, it leads me to, particularly when it comes to cancer, say there is in fact to my best judgment, and it turns out the best judgment of many of our leading medical organizations in the United States and around the world, a legitimate cause for concern that there is in fact a causal link here.

Vivek Murthy (22:18):

Finally. Look, I think the point about bias is real. Just because people may have a lot of education or may have degrees or experience or whatever it might be, doesn't mean that they're not human beings who are subject to bias. And this is played out in medicine too, where if somebody tells doctors something that they're doing is unhealthy and they'll have perhaps an initial counter reaction to it. And my hope is that they'll pause and then look at the data closely afterward. But the truth is we're all subject to some bias here, which is why it's really important that we have these kind of public conversations and ultimately that we look at the data, but also continue to gather data over time. Because the cardiovascular story, I think is a great example of why we need to continue studying these issues, we now recognize that there were methodological flaws, right?

Vivek Murthy (23:08):

In many of those cardiovascular studies done early on where the control group included people who had an alcohol use disorder and quit and hence made the control group seem unhealthier and may have contributed to some of the “benefit” from a cardiovascular perspective that we were seeing among people who were again, “moderate” drinkers. When you start to root that out and have different control groups, then some of those results it turns out have actually changed and the effect or the benefit has actually diminished. So all this to say, I recognize the limitations, but the reason I put out this report, Eric, is because even though the evidence has been building around alcohol and cancer risk for years, even though medical societies had said it, even though it was the subject of a 2016 larger report that I had, what I realized is that public awareness is really low, that people knew that smoking cause cancer, they knew that asbestos was a problem when it came to cancer. They knew that radiation was a problem when it came to cancer, but the majority of people did not know about the link between alcohol and cancer.

Eric Topol (24:15):

And one thing, just to back up what you're saying, the totality of evidence. Agree, there's a level of alcohol that's in the population tied to an excess of several cancers. I mean, I think there could be some debate about where that risk starts, but the Mendelian randomization study that I mentioned is another perhaps even more objective way because here you're looking at the UK Biobank, we'll link to that report, I know you're familiar with it, where people with variants I should say that have issues with metabolizing alcohol, they have an increased risk of cancer. So the tie-in between alcohol intake and cancer has been established in many ways. That's certainly one that you couldn't argue that's as good as we can get right now because of the other lack of being able to get to other ways to tackle this or to zoom in.

Eric Topol (25:20):

Now, there's been no shortage of backlash because of the people's current practices. And I thought it would be at least somewhat fun to run through those quickly with you. You may have seen the cartoon by Michael Ramirez where there's a caricature of you, alcohol, raising kids, mental illness, work…this is my prescription for everything. And it's with a patient asking government? You probably saw that one. It's pretty funny. And then there was a crazy one by that woman at the Wall Street Journal, Allysia Finley, ‘No, moderate drinking won’t give you cancer’, but the ones that I thought were particularly interesting, you may have seen in the New York Times, Rachael Bedard who's a columnist there, ‘The new alcohol warning is not a prescription.’ Basically, she's making the case that this is good to have the advice, but it shouldn't be a prescription.

Eric Topol (26:21):

One that I thought was really interesting is Eric Asimov, the person who writes about food and at the New York Times. So he wrote about how he has three glasses of wine every night with his dinner, which I thought was interesting, and also defending his practice and latching onto the National Academies of Medicine report as a definitive report. So that's interesting. But the one that I noted was Derek Thompson in the Atlantic and he wrote about that every drink takes five minutes off your life. And he said, well, I can counter that because one minute of exercise adds five minutes of life. It's kind of amusing. Anyway, I'm sure there were many more. I'm just giving you a kind of cross section, and I just wonder, you probably expected backlash because you're trying to present data that is unfavorable to people who enjoy their whatever few drinks per week or whatever that is. And what are your thoughts? Is that something you anticipated, or did you think that it was excessive?

Vivek Murthy (27:42):

Actually, no. I actually thought it was quite within the expectations that I had. And in fact, what we found when I looked at both the media coverage, the discussion among various groups that we were privy to as well as the feedback we received, is that actually the majority of responses were positive or responses are like, oh, I didn't know this. I'm glad I do. Right? There was a minority of responses that we found that were actually critical, but look, and in fact, less than I thought there would be, but this is actually not unexpected because again, in 1964 when Luther Terry published a report about cancer and smoking, he actually received a very similar response from people who said, what are you talking about? I've been smoking for years and I'm fine. Or my mother smoked until she was 90 and she was fine. Or this is just part of the fabric of American culture, why are you trying to change that?

Vivek Murthy (28:36):

So these are understandable responses. And look, I'd be the first to admit that this data that alcohol is causally linked to cancer is highly inconvenient, very inconvenient. And I wish that wasn't the case because I know for many people, including people in my life that alcohol is a part of their life and an enjoyable part of their life. But my point in putting this out was that there were too many people, I believe, who did not know about this link and hence could not use that information in their decision making. The point of our advisor is not to say that overnight everybody should stop drinking any quantity of alcohol. That was not the point of the report. The point was to do something more nuanced, so let people know about a risk so they could factor that into their decision making, especially people who were predisposed to cancer or who were more concerned about cancer.

Vivek Murthy (29:32):

Finally, I'll say one other, just to share one more line of critique that we got, which I thought was very interesting, is our office and me in particular, have worked a lot on the issue of loneliness and isolation and talking about the profound impacts that has on mental health, physical health, and societal health. And I did receive one message from a friend somewhat tongue in cheek, which said, you can either put out this report on alcohol and cancer risk, or you can say we need to be more socially connected, but both can't be true because many of us use alcohol as a way to connect with each other. And I recognize that too, right? And we have built a lot of our social interaction around alcohol, not all of it, but a lot of it, and this was actually true back in the fifties and sixties too, people would gather and smoke together.

Vivek Murthy (30:20):

And so, when you said, hey, smoking might be bad for your health, part of how some people interpreted that was, wait, you're telling me that something, not only do I do that's enjoyable, but I do with friends is now going to potentially harm me. I don't like the sound of that. So look, I recognize that this is a process. People need to hear information, they need to assess it, they need to then come to terms with it and then decide whether they're going to use that to change their behavior in any meaningful way. And our job, I think, as people in public health and in medicine, is to be with people on that journey and to listen to their feedback, to provide the data as it continues to evolve and to have real conversations so people can interpret this when people say, oh, government shouldn't tell me what to do.

Vivek Murthy (31:05):

This isn't a government telling you what to do. This is many bodies of science outside of government who have come to similar conclusion and are offering data so people can make decisions for themselves. So I do think we have to be a part of this ongoing conversation. And finally, just as where we began, just to come back to full circle to where we began with your anecdote about the patient who came to you about her husband who is “drinking at moderate levels” with two tequilas and six beers a day. The whole notion of what constitutes moderate is something that we need to have a conversation about in the country. Because what has happened, I worry, is that the dietary guidelines over the years, and these are for those who don't know, these are guidelines that are put out every five years from the federal government, usually a joint effort between the US Department of Agriculture and the Department of Health and Human Services.

Vivek Murthy (31:57):

There's often an independent advisory group that provides recommendations, but then ultimately the departments make the final decision on these recommendations. Notably the advisory group, the independent advisory group was instructed this time not to provide their input on alcohol for the 2025 guidelines in a move that I thought was to say interesting, to say the least. But what they had recommended five years ago for the 2020 guidelines was that the upper limits of alcohol consumption, which are currently two a day for men, one a day for women, that those be downward revised. But that advice was not taken, and it was kept at two and one for men and women per day. I actually think we need to reassess that because I do think that in light of some of this data, we do need to have a conversation about whether that's actually appropriate. Other countries have taken steps to reduce what are considered the upper limits of drinking. I think we should too. And you don't have to wait for government to do that. I think that's something that individuals can consider in their own lives.

Eric Topol (33:09):

Yeah, I do think that the data that we review would suggest that two per day that is in men or fourteen per week. I mean, there's certainly a relationship there as we reviewed regarding cancer that starts at the seven per week, so irrespective of sex. So yeah, I agree with you completely about that. Now, before we wrap up, you brought us to the other key topic I wanted to touch on among many others that we could, but the one that is very interesting is loneliness and social isolation, which aren't the same thing. So maybe you could just first differentiate those two important areas.

Vivek Murthy (33:51):

Yeah, thanks, Eric. Loneliness is the subjective feeling that the social connections I need in my life are greater than the ones I actually have. And in that gap we experience that sense of loneliness. Loneliness can feel emotionally painful. It can actually even feel physically painful because we now know about the way our brain interprets emotional and physical pain. Isolation on the other hand is an objective term that is based on the number of people you have around you. Now, people can be not isolated. They could be, for example, like the college students I've encountered on campuses where they're surrounded by thousands and thousands of other students. But if they don't feel connected to those individuals, if they don't feel they can be themselves like anybody understands them, then they actually may feel quite lonely. And so, that is a distinction between the terms. And the reason I came to work on this was actually not because I was taught about it in medical school or in residency training.

Vivek Murthy (34:46):

It was because when I began my first term as Surgeon General, I traveled around the country and asked people a fairly simple question which is, how can I help? And I tried to just close my mouth and listen to what they said, which is something that as you remember, they always taught us in medical school, which is walk in the room, introduce yourself to the patient, and then try to close your mouth and listen. Which is not easy for doctors, but we try and it is in through those conversations in big cities and small towns all across America that I started hearing these stories of loneliness, and I heard a lot of them from college students who talked about feeling like there was nobody they could be themselves with on campus, but I also heard them from moms and dads who were surrounded by people all day long, but felt like all the burdens that they were carrying, they had to carry alone.

Vivek Murthy (35:31):

And they felt a sense of shame and guilt in even speaking about the struggles they had raising their kids. I heard this from CEOs who felt very alone from doctors who were struggling and who said after residency training, I felt like I just didn't have a tribe anymore. I wasn't working side by side. It was such a lonely experience. And I even heard it frankly from members of Congress who told me behind closed doors when there were no cameras, that they struggled with loneliness because they could never really say what they thought or be themselves. There are always cameras around, people who are recording conversations, people who are tweeting out what they said and what they thought was a private conversation. So the more I heard that, it motivated me to dig into this subject, Eric, and that's when I realized that there are profound health consequences, not just mental health, not just increased risk of anxiety, depression, and suicide that we see with social disconnection, but also an increased risk of cardiovascular disease, of premature mortality, of dementia, and not at small levels either. The increase in dementia was like 50% among older people, talking about a 31% increase in cardiac disease, a significant increase in premature death as well. And so, all of this is what led me to work on this issue and to recognize if we're not working to foster stronger social connection, repair the social fabric in our communities, we're not going to be able to get people to be as healthy and as fulfilled as they want.

Eric Topol (36:57):

Yeah, no, I certainly think that your emphasis on this and the book you put together is noteworthy and calling out something that particularly exacerbated by the pandemic is so vital. Now, like the alcohol, but not nearly as fervent there has been some countering. You probably saw earlier this year, last month in the Atlantic, Faith Hill wrote the Myth of a Loneliness Epidemic, and then there was the Nature Human Behaviour paper last fall, September published. It said loneliness is not causal, it's just an association. So obviously in their titles they're talking about loneliness, which as you aptly point out is a feeling. It's not social isolation objective metric. How would you respond to these pieces, whether it's the Nature paper or the Atlantic about that this is a myth?

Vivek Murthy (38:01):

Yeah, it's interesting. I mean, at this point, I've learned over nearly 10 years of serving in various capacities that if you put something out, and there are no critical takes on it, that means that you did something wrong, you didn't push the envelope enough. And so, what I have been surprised by is that is actually how remarkably few critical takes there are on the issue of loneliness on the work that we've done in this space, there are going to be a few, no question. And again, I think this falls into the bucket of things that are inconvenient. We don't want necessarily believe that we have become a lonely nation or increasingly a lonely world, but the data is actually very strong. And also it tells us very clearly, increasingly in multiple datasets across multiple countries over time, is that the population actually that seems to be struggling the most with loneliness is actually young people.

Vivek Murthy (38:55):

Now, this is often surprising to people and counterintuitive, right? People say folks who are young adults, teenagers, they're on tech all the time, they're talking to their friends all the time. They're texting their friends all the time. Actually, they're texting is how they talk primarily, but they say, how can they be lonely as a result of that? When you understand that this is a subjective feeling, then you understand why the data tells us what it does. I'll tell you one of the schools that we went to, a large public university on the college tour that I did on social connection. This university, which has more than a hundred thousand students, said that they wanted to do a survey before we came of their student body to understand just how big a deal loneliness actually was. They thought maybe it was a big deal, maybe it wasn't.

Vivek Murthy (39:36):

They weren't quite sure. And before we came, actually, on the day that I arrived, they presented the data to me and they said, to our surprise, more than 80% of our students said that they were actually struggling with loneliness. And that really surprised us. Another school that I went to halfway, the other side of the country said that when they surveyed their students who go to the mental health clinics, that 92% of the students who seek out mental health care on campus say that loneliness is one of the key reasons that they're coming. So we can sort of hope for what we want. We can define terms in the ways we want, but what people are telling us very directly and young people in particular is that they're struggling with loneliness. And the data that's been accruing actually has now I think in aggregate helped us see that this is more than just an association.

Vivek Murthy (40:26):

In fact, one of the reports that's going to be coming out hopefully later this year is a large report from the World Health Organization Commission on Social Connection. This is a commission that I co-chair, but it's gathering it has an independent group called the Technical Advisory Group on Social Connection (TAG-SC) that is drawing from members from across the world scientific experts. We're putting together a report which is actually laying out both the extent of loneliness and the causal link between loneliness and these health outcomes. Finally, I think it's worth just remembering, Eric, that we're talking about health for obvious reasons. That's the field that you and I operate in. But the impact actually of social disconnection actually goes far beyond health. We have seen in the CDC put out a study to this effect a few years ago that kids who actually feel socially disconnected from others in school that seems to impact their academic performance.

Vivek Murthy (41:22):

We see that communities that are more connected with one another are more economically prosperous and have lower rates of violence. At an intuitive level, this sort of makes sense, right? If you know your neighbors and if you are connected to people in your community, when there's a hurricane or a tornado, you're more likely to pitch in and help each other out, which is why more connected communities, it turns out to be more resilient also after natural disasters. This is something also from a public safety perspective that mayors interestingly have realized over time. Mayor Tom Tait at Anaheim, California told me years ago that he saw social connection as a safety factor. And one of the things he knew working with his police force is that a lot of folks who might be looking to case a neighborhood or rob a house, they tend to know when neighbors know each other, when they have relationships or connected to one another.

Vivek Murthy (42:14):

And it's the neighborhoods where people are isolated, where they don't know each other that are easiest to target because even before police or firefighters come onto the scene, who's there in the case of an emergency? It's potentially your neighbor. So the neighborhoods where people knew each other, where they were more socially connected, it turned out had lower rates of crime. So this is one of the reasons why I think we have to look at social connection as bigger than a health issue, but as one of the core pillars of society that helps us enhance the economy, the education, the health and wellbeing of individuals, and when it is weak as it has been disintegrating around the world, as we have shifted more of our offline connection to online, as we spend less and less time with one another, as our ability to dialogue has disintegrated, then that just has profound consequences for society. And we see that now, not just in health, but with the increasing division and polarization in America and around the world.

Eric Topol (43:10):

No, I think for there, you really have your finger on the pulse of a critical issue, Vivek, and I think we're indebted to you for really spotlighting this. So that gets me to the next chapter. You're young and you've already contributed immense amount for public health. What are you going to do now, now that you're not involved with the current administration, of course, and you've got the world out in front of you?

Vivek Murthy (43:38):

Well, first of all, thank you for calling me young because my kids just yesterday told me that I'm old, so this feels reaffirming, but I feel incredibly grateful, Eric to have had the opportunity to serve twice. The second time, in particular, a time of a great crisis from a public health perspective. And I've learned a lot, but I've learned most from the people I met across the country. And as I think about the future, I think about a lot of those stories of people who just sat with me in their homes, in town halls, in gyms, in high schools across the country, and just openly shared about their lives, about what they were worried about, about their struggles. And that's what’s guiding me now as I think about what to work on next. I put out in the last few weeks of my tenure, my final prescription, if you will, I called it My Parting Prescription to America.

Vivek Murthy (44:34):

And that is actually still available on my website, www.vivekmurthy.com. Although as I've come to understand from the public, it is no longer available through the HHS website, at least not temporarily, but it can be found on my website. And what it was aimed to do, Eric, was to synthesize all of these learnings and conversations from the last two terms and to really get at what is it the root cause of the deeper unhappiness that I encountered so often, and I encountered it not just from, we know, we hear and read about in the paper that when you're struggling with poverty, when you're struggling with safety issues, when you're worried about your kid's health, when you can't access healthcare, these all are profoundly impactful in terms of your wellbeing and your happiness level. But what I was finding interestingly is that even when those needs were met, that there was still this deeper sense of anxiety and discontent and unhappiness.

Vivek Murthy (45:26):

And so, that parting prescription was really aimed at trying to understand that. And one of the things I laid out, and this gets to what I really want to work on going forward, is that we find ourselves caught in this battle between two triads that direct our life. One is a triad of success, and the other is a triad of fulfillment. The triad of success is what I heard about from students each and everyday when I traveled around the country. When I asked them, how do you define success? What I was really asking them is, how does society define success for you? And they would tell me some version of money, fame, and power. If you had all three of those, then you really hit the lottery and people will write books about you, make documentaries about you, you'll have made it. But what I realized in all of these conversations and through looking at data over the years and really analyzing that data with experts from across fields, not just health, was that what we really want is to be fulfilled in life, and that requires a different triad.

Vivek Murthy (46:25):

The triad to fulfillment is rooted in relationships, service, and purpose, right? When we have people in our lives we're deeply connected with who understand us, who we care about, when our life involves helping people on a day-to-day basis, even in small ways, and when the purpose that we experience is rooted in contributing to the lives of others, that is a recipe for fulfillment. And what I want to dedicate my time to and my heart to and my energy to in the years ahead is figuring out how we can swing this pendulum toward the triad of fulfillment. How we can help people build relationships, engage more in service, build a sense of purpose in their life, and elevate people who have these elements as role models for our children. The number of kids I talk to, I say kids, but they're really young adults in college and in high school, the number who told me that their goal was to build their brand and to be an influencer.

Vivek Murthy (47:24):

I lost track of how many students said that, but every time they did, there was a small part of me that died honestly because, not because I faulted them in any way, but because I worried that when we chase fame, the data is pretty clear that while there's nothing wrong with fame in and of itself, if we rely on it as our source of fulfillment, then we're just begging to be disappointed as so many have been. So I think about my children who are seven and eight growing up in the world. They're growing up, and I think about all of our kids. I want them to grow up in a world where their lives are anchored in relationships, purpose and service, because that is what's going to give them the best shot to lead a fulfilling life. And that's not just about policy and programs, by the way, that's primarily about culture. It's about how do we shift our culture to ground ourselves in this triad of fulfillment so that when kids look around them, when they look for role models, they're seeing people who are dedicating their lives to each other, people who are investing in their family and their friends in their neighborhoods, and people who talk about purpose the way that so many of our colleagues in medicine did over the years, the way I see dedicated teachers and other educators do, talking about wanting to live a life that makes the world just a little bit better.

Eric Topol (48:35):

Yeah. Wow. That's a lot of wisdom there, Vivek. Thanks for that, and we'll certainly link to your website to make sure that people can access it. So I can't thank you enough for joining today. I think it's been great to review all this barrage of new data reports regarding the alcohol and its links, and also to get a bit into the social isolation and these other critical cultural matter. So thank you, Vivek. We'll continue to follow your career because I know you're going to keep making major contributions, and you are really young, just reflected by the age of your children. That tells you you're young too. Thanks so much for joining.

Concluding Remarks

For something that is so relevant to our health, it is especially striking to see how 3 reports yield such different conclusions. That’s not just the difference in amount of alcohol intake and its relationship to risk of breast cancer or other cancers, but also the potential benefit of reduction in all-cause and cardiovascular mortality. How could this have happened?

The purpose of the reports was quite different, no less their methodology and interpretation of the data. The National Academies was the only one with a de novo review of all relevant data and attention to the “never consumed” alcohol comparator. It is also the only one that concluded with moderate certainty there is a mortality benefit which could potentially be seen as titrating the risk of cancer.

There is not doubt that alcohol should be regarded as a carcinogen (nicely and simply reviewed in the Surgeon General’s report). However, my review of all of these reports only comes up with agreement on breast cancer risk. The alcohol related risk associated with the other 7 types of cancer is likely real, but the amount of intake and certainty of the assertion related to the “dose” is ambiguous.

Dr. Murthy’s report was notable for raising public awareness of cancer risk with alcohol intake. The absolute lifetime risk of breast cancer increased from 11.3% for women with < 1 drink per week to 13.1% for 1 drink per day. The absolute increased lifetime risk of cancer, citing only the Australian cohort study, is shown below for women and men. But that associated lifetime risk, higher in women, seems to represent the worst case scenario in context of all the studies reviewed in the National Academies report.

I need to emphasize that all of these reports are concluding at the population level, just like recommendations for diet and exercise, not taking into consideration that each of us is unique, such as our susceptibility to cancer from our genomics, other lifestyle factors (like smoking or physical activity) or environmental exposures. Our genes (genomic variants) for metabolizing alcohol vary widely, especially inter-ancestry, and that is pivotal to understanding the “dose” of intake (e.g. one drink in one person may be equivalent to 3 in another by virtue of their difference in metabolism). Dr. Murthy highlighted these points in our conversation.

With a paucity of any randomized trials for major outcomes (the only decent one I could find was a 2-year trial from 2015 that showed wine reduction of cardiometabolic risk factors) the level of uncertainty is an issue, at best moderate and, for the most part, low. It’s all the more reason to do Mendelian randomization (MR) studies based on genomic data to get to cause and effect, and probing alcohol consumption at the individual level. For this reason, I cannot understand why the ICCPUD decided to exclude MR studies from their analysis.

Where does this leave us?

You can pick which report and recommendation you like to support your confirmation bias and lifestyle preference—at least up to a certain level of alcohol intake (less than 14 drinks per week in men, 7 drinks per week in women). A compromise upper threshold, until we know more, might be 7 drinks per week (and somewhat less in women). Binge drinking needs to be avoided (4 drinks for women, 5 for men, in a single sitting).

Of course, a draconian view of all of this would be to eliminate alcohol consumption or take it down to less than 1 drink per week. But that doesn’t take into account all the factors I’ve tied to highlight—the potential benefit, the lack of certainty, and lack of any individualized guidance. Nor the importance of social interactions for promoting our health, which frequently incorporate alcohol intake.

Overall, it seems prudent to be aware of some risk for alcohol intake but I hope this review of the 3 reports helps put the question in context for you.(And sorry for it being so long!)

A Quick Poll:

**************************

As you can imagine, I’m excited to get my new book out. I've worked with my publisher, Simon & Schuster, to share an exclusive offer just for Ground Truths subscribers. Through May 5, you can pre-order a hardcover copy of the book at Bookshop.org and get 15% off with code SUPERAGERS15. It’s notable that Bookshop.org supports independent bookstores across the country. Of course there’s also Amazon which currently offers a discount on the book, but it’s certainly not helping these bookstores. My op-ed preview of the book was published in The NY Times this week. Here’s a gift link.

In this post, I’ve barely scratched the surface about the content of the book. Here’s the back cover to give you an idea of what some people had to say about it.

Thanks for reading and subscribing to Ground Truths.

If you found this interesting please share it!

That makes the work involved in putting these together especially worthwhile.

All content on Ground Truths—its newsletters, analyses, and podcasts, are free, open-access.

Paid subscriptions are voluntary and all proceeds from them go to support Scripps Research. They do allow for posting comments and questions, which I do my best to respond to. Please don't hesitate to post comments and give me feedback. Many thanks to those who have contributed—they have greatly helped fund our summer internship programs for the past two years.