Each glass shortens lifespan as much as a cigarette

We have a new study out, in the Lancet no less, telling us that the new, lower, limits for reasonable alcohol consumption are just right. Well, of course the report says that, right? The problem being that it’s entirely contrary to the more general experience we’ve got of booze consumption. For, yes indeedy, there’s a level of drinking which will – as always, on average – shorten life. But our experience to date is that it’s several times what is the current measure of safe consumption. This basic understanding of ours being that no booze lowers lifespan, too much lowers it, a modicum increases it. The argument being the definition of modicum of course.

Observation of large populations being that modicum is anything from some up to perhaps 40 or even 50 units a week. This isn’t what the current study shows at all:

Drinking will shorten your life, according to a major new study that suggests every glass of wine or pint of beer over the daily recommended limit will cut half an hour from the expected lifespan of a 40 year old.

Those who think a glass of red wine every evening will help keep the heart healthy will be dismayed. The paper published in the Lancet medical journal says five standard 175ml glasses of wine or five pints a week is the upper safe limit – about 100g of alcohol, or 12.5 units in total. More than that raises the risk of stroke, fatal aneurysm (a ruptured artery in the chest), heart failure and death.

A result so at odds with our other experience does rather need substantial proof really, doesn’t it? Our best option here is to wait for Chris Snowdon and see what he thinks the gaps in the methodology are. But it’s fun to speculate ourselves while we wait.

This, for example, is great big hairy dangly ones:

“This study makes clear that on balance there are no health benefits from drinking alcohol, which is usually the case when things sound too good to be true.”

Nope, that’s not what the study says at all. It, even with whatever might be wrong with it, states that there’re benefits up to those 12.5 units a week.

So, to speculate:

In the 599 912 current drinkers included in the analysis, we recorded 40 310 deaths and 39 018 incident cardiovascular disease events during 5·4 million person-years of follow-up. For all-cause mortality, we recorded a positive and curvilinear association with the level of alcohol consumption, with the minimum mortality risk around or below 100 g per week. Alcohol consumption was roughly linearly associated with a higher risk of stroke (HR per 100 g per week higher consumption 1·14, 95% CI, 1·10–1·17), coronary disease excluding myocardial infarction (1·06, 1·00–1·11), heart failure (1·09, 1·03–1·15), fatal hypertensive disease (1·24, 1·15–1·33); and fatal aortic aneurysm (1·15, 1·03–1·28). By contrast, increased alcohol consumption was log-linearly associated with a lower risk of myocardial infarction (HR 0·94, 0·91–0·97). In comparison to those who reported drinking >0–≤100 g per week, those who reported drinking >100–≤200 g per week, >200–≤350 g per week, or >350 g per week had lower life expectancy at age 40 years of approximately 6 months, 1–2 years, or 4–5 years, respectively.

I’m not in any manner a medical expert but that does look odd. 5 million observation years on half a million people, looks like 10 years on average per person. They’re using this to predict lifespan at age 40? When lifespan at 40 is, these days, a further 40 to 50 years or so? OK, maybe there’s some sekkrit decoder ring for epidemiologists here but anyone want to try and explain it?


We focused our study on current alcohol drinkers

So the comparison doesn’t include those who don’t drink. We’re not therefore getting a baseline of no alcohol consumption to compare with. That is, by design, the study excludes the known to be higher death rates (or lower lifespans) of the temperance types. No, really:

Third, never-drinkers might differ systematically from drinkers in ways that are difficult to measure, but which might be relevant to disease causation.

Our more general stats do indeed say that heavy drinkers (that 40 to 50 unit level perhaps) and never drinkers have about the same lifespans. Quick, gotta exclude that information, eh?

As far as we’re concerned that’s probably enough. We’ll see what Snowdon has to say about it, shall we? Because this finding is contrary to pretty much everything else we know about booze consumption. Explaining why it is will be important.

Subscribe to The CT Mailer!

Leave a Reply

Please Login to comment
3 Comment threads
0 Thread replies
Most reacted comment
Hottest comment thread
3 Comment authors
Ian ReidBaron_JackfieldSpike Recent comment authors

This site uses Akismet to reduce spam. Learn how your comment data is processed.

newest oldest most voted
Notify of

This is such a perfect call for government to manage-by-the-numbers the entire human herd that it won’t matter that there are gaping errors in the numbers. All we have to do is have the state enforce this “modicum,” as government beer stores in Sweden used to try to track individual consumption and predict health problems. Lots of secure jobs!

PS – It’s impressive how easy it was for the authors to wordsmith away the fatal flaws in their study.


That’s a load of statistical orchitis… Look at the 95% C.I.s, nothing better than about 1.33, or 0.91 for negative correlations. These figures wouldn’t be acceptable for a double-blind trial, let alone this type of study. Previous editors of The Lancet have demanded CIs of >3 before considering publishing.

It’s basically a rubbish study, the sort of thing you’d expect from ASH or “Alcohol Concern”.

Ian Reid
Ian Reid

The other commentators have pretty much nailed it. But you could write a book with the flaws in this. Indeed John Brignell did that with his “The Epidemiologists. Have They Got Scares For You!”. In no particular order some things I can think of are:- – How did they select the studies to include in this meta-study? Unless they were all conducted with the same methodology, adding them all together simply isn’t valid. – Unless the studies all had similar populations again it isn’t valid e.g. is comparing outcomes between migrating Mongolian herdsmen with Dutch dairy eating farmers valid. There… Read more »