Time to cut down on the booze?

As of the 23rd May 2022 this website is archived and will receive no further updates.

understandinguncertainty.org was produced by the Winton programme for the public understanding of risk based in the Statistical Laboratory in the University of Cambridge. The aim was to help improve the way that uncertainty and risk are discussed in society, and show how probability and statistics can be both useful and entertaining.

Many of the animations were produced using Flash and will no longer work.

The recent study on alcohol and cancer published in the British Medical Journal is a fine piece of epidemiology and attracted a lot of coverage of the estimate that 10% of male cancers and 3% of female cancers could be attributed to alcohol. But while it is useful as a description of what happens in populations, as usual when translated to an individual it stops looking so impressive.

According to Cancer Research UK, 1 in 3 of us will get cancer in our lives. So out of 100 average drinkers, 33 will get cancer at some point.

According to Table 2 in the BMJ paper, each 12g drink (1.5 units) increases annual cancer risk by 3%. So if another set of 100 people all had a drink less a day, then we would expect 32 would get cancer. If another set of 100 all drank an extra drink a day, 34 would get cancer.

This reinforces Sir Ian Gilmour's comments that we can't expect such evidence to change people's habits - he says we need to have tougher regulation.

Also the attached graphic in the Daily Mail is dreadful - it strongly suggests that 1 in 10 men will get cancer because of drinking.


Free tags: 


Without reading the paper I am still confused: "each 12g drink (1.5 units) increases annual cancer risk by 3%." 1) Should that drink be "per day"? 2) The cancer risk increase from the paper (3%) is given annually and but the population base cancer rate from Cancer Research UK (1 in 3) is given "at some point". Is that right?

I was slightly disturbed by the reporting of this too- If 3333 / 10000 females get cancer, and 3% of those are caused by alcohol then 100/10000 or 1% of females will get cancer caused by alcohol. Only 4% of cases attributable to alcohol in women (17470 / 397043) were caused by drinking above the accepted limit, which they say is 'substantial'. But to get at whether this is substantial I would want to know the number of these women that were actually drinking above the accepted limit (average for both genders was 33%). In any case, I'm sure the main conclusion of this study should be that if you really want to cut out this cause of cancer, we should not drink at all. Alternatively, if you're going to have a drink, don't worry too much about going over the limit?! (In terms of cancer risk anyway) Of course on top of this, this is looking at the incidence of cancer rather than fatalities as a result, which is perhaps what the general population are more interested in? Liz

This comment by Sir Ian Gilmour is ridiculous. If the evidence is there but people choose to ignore it why should there be regulation. It is their choice and drinking alcohol is not an illegal activity. People choose to ignore the fact that cave diving and free climbing may shorten their life expectancy but we don't regulate to stop them doing it. They have made a choice based on risk/benefit to them.