Richard Smith: Battling over safe alcohol limits

Richard SmithAdvice on smoking is simple: don’t smoke. But what should be the advice on alcohol? It can’t be “don’t drink,” nor can it be “drink less.” Doctors and governments think that they need to give guidance to people on alcohol—and mostly they do that by suggesting “safe limits” based on units of alcohol. But is this advice scientifically sound and beneficial? These were the issues debated with considerable feeling at the recent Battle of Ideas meeting in London.

I was a member of the Royal College of Physician’s working party that in the early 80s proposed safe limits for the United Kingdom of 21 units a week for men and 14 for women with one unit being 8 grams of alcohol. We advised that a unit was half a pint of beer or a standard glass of wine. I achieved some notoriety about five years ago for telling a journalist that these limits were “plucked out of the air.” Now whenever there is a debate about the validity of the safe limits—as there often is—I’m rung by journalists for a quote. My clumsy statement has not made me popular with the Royal College of Physicians.

Many of the thinkers attending the Battle of Ideas, a good few of them libertarians, don’t like being “told what to drink” by doctors and governments. Dolan Cummings, associate fellow of the Institute of Ideas and editor of Culture Wars, argued that the safe limits are scientifically weak, few people use them, most of those who drink alcohol drink more than the safe limits, and there is a level of “healthy, responsible, controlled” drinking that is well above the limits. “Children,” he said, “are being lied to” about what’s safe, and “unit drinking is not a healthy way to drink” just as calorie counting is not a healthy way to eat. Drinking alcohol is a social not a medical issue, and society not  doctors and governments should decide what is responsible drinking.

Kristin Wolfe, head of alcohol policy at SAB Miller, one of the world’s largest brewers, said that 80-90% of people are responsible drinkers and that the harm results from the other 10-20%. She accepted that people should be given information on safe limits, but she pointed out that there is considerable variation in advice around the world: in Australia the advised limit is 20 grams a day, whereas in Portugal its 37 grams; in the US a unit  is 14 grams, contrasting with the 8 grams in Britain.

I agreed that it’s difficult to set safe limits. I remember the debate at the working party nearly 30 years ago when the epidemiologist said it was impossible to set limits because the evidence was poor. Then there are the problems that the possible consequences of alcohol are hugely varied, including both social and medical problems, people are all different in size, body composition, and responses to alcohol, and the amount of alcohol even within the same category of drinks (beer, wine, or spirits) varies considerably. Despite these difficulties the working party agreed that it was better to offer some advice rather than none and that it needed to be simple. Although there is variation among countries on what they advise they are all, said Ray Tallis, physician, philosopher, and writer, “in the same ballpark.” People are unlikely to come to much harm drinking below the safe limits.

More than a quarter of men (28%) and nearly a fifth of women in England (19%) drink more than the recommended limits in an average week, and more than a third of men (38%) drink more than four units in one day and nearly a third of women (29%) drink more than three units in one night—making them binge drinkers by one definition. Most of those at the meeting liked to think that these drinkers, although above recommended limits, are “responsible” or “healthy” drinkers. In reality many of these people are likely to damage themselves or their families because the risk of medical and social damage rises steadily with consumption and because alcohol is an addictive substance, meaning people tend to increase consumption and find it steadily more difficult to cut down.

We did debate providing people with much more complex information that gives data on the risk of various consequences as consumption increases—perhaps through a smart phone and combined with information on how any given individual reacts to alcohol. There was scepticism, however, that more than a handful of people would use the information. Many people, including possibly me, prefer to fool themselves that their drinking is not harming them.

My main argument, which led to sharper disagreement than over units, was that it was best to respond to alcohol through public health rather than individual measures. Overwhelming data shows that the more alcohol a country or community consumes the more harm it will experience, and we know that increasing price will reduce consumption. So a steady increase in the price of alcohol is the best way to reduce harm—just as with smoking. This is a message that is very unattractive to the drink trade and to many of those at the Battle of Ideas. I left understanding more clearly how politically difficult it can be to do what seems sensible and obvious to doctors.

Competing interest: RS was speaking at the meeting and was offered expenses, but as he cycled there and back he didn’t have any. He did, however, get a free cup of (not very good) coffee.

RS was the editor of the BMJ until 2004 and is director of the United Health Group’s chronic disease initiative.

  • Martin Lam

    Where amongst all these stats referred to on the harm alcohol causes are the ones outlining beneficial effects of “healthy, responsible, controlled” drinking” ?, or don't they exist outside the benefits of polyphenols in certain higher altitude red wines etc. As with much talk about diet, where food is seen as the enemy rather than  the source of sustenance, perhaps encouraging a view  that allows healthy, responsible, controlled drinking might lead to some social changes in that direction.

  • Martin Lam

    Moderate consumption of red wine is beneficial, even over a short periodResearch has shown for the first time that moderate consumption of red wine, even over a short period, combined with a Mediterranean style diet, improves the blood lipid levels and the antioxidant status of patients having suffered a coronary event.

    The research was carried out by the metabolic and nutritional biochemistry team at the INSERM research centre at the University of Burgundy, led by Professor Norbert Latruffe.

    A clinical intervention trial was undertaken amongst a group of patients, all of whom had suffered a heart attack. The study was conducted at the Hôpital de Montbard during the period of cardiac rehabilitation. For two weeks, the patients followed a diet similar to the Mediterranean diet. One group consumed one glass of red Burgundy wine at each of the two meals (250 ml/day) whilst the other drank water.

    The clinical, physiological and blood parameters were measured on the first and last day of the study. The results show the positive effect of the consumption of red wine: a reduction in overall and LDL cholesterol (so-called bad cholesterol) and an increase in the membrane fluidity of the red blood cells and the antioxidant potential of the blood.

    The original aspect of this research was in the exploration of the effects of the consumption of red wine on post-operative patients, with the aim of evaluating the prospects for secondary prevention of a heart attack.

  • Paddy


    The trouble, as I understand it, is that the healthy ideal for a cardioprotective effect and minimum harm is about 5 units a week (e.g. a small glass of not-very-strong wine about 5 times a week).  And this is leagues away from what most people drink.  Whereas the 21 or 14 unit figure is really meant as a (rough) upper limit, not a recommended average, even though this is how many people interpret it.  And then allow themselves some more, of course.

    Personally, I'm cutting down by only drinking on the weekend, and then not drinking heavily.  And if I do slip up a little and drink once in the week, I plan to then take Sunday off instead.  Not sure if it'll work, but I'm vaguely hopeful.

  • Richard Smith

    A valid criticism, Martin, and there was talk at the meeting on the many good things about alcohol. (I'm drinking a glass of Nero d'Avola as I'm writing this.) We doctors tend to concentrate on the harm, although many of us drink over the safe limits. I do urge you, however, to be sceptical about the evidence that a little alcohol makes you live longer. There are many problems with that evidence. Drink because you love the taste, the whole aesthetic experience, and the glow rather than to lengthen your life.

  • DaveAtherton20

    Hi Richard, I hope you are well.

    With respect there are epidemiological medical papers that are cohorts, that do have statistically significant results for the benefit of drinking alcohol, please read my blog post.

    A total of 4,833 participants died. Of these, 1,075 died of coronary heart disease and 1,552 died of cancer.”The conclusion was that drinker’s mortality if consumed in moderation  was far lower than non drinkers.0 drinks/wk 1.00 (referent)1–7 drinks/wk 0.82 (0.76–0.88)8–21 drinks/wk 0.82 (0.75–0.89)22–35 drinks/wk 1.00 (0.89–1.12)35 drinks/wk 1.10 (0.95–1.26)