Lots of chat on the internet today about a study that has just been published in the 0pen-access version of the British Medical Journal. (Article is here.) It made front-page headlines in The Daily Mail, a British tabloid newspaper targeted at right-leaning women, and designed to keep its readership in a state of heightened fear about the horrible world we all live in.
My interest in alcohol and health is a personal one. I really, really want to know how much wine I can drink without damaging my health. As a man in the over-40 segment of the population (and thus at elevated risk of cardiovascular disease), I am one of the people for whom moderate drinking has been shown to have a protective effect. But I enjoy wine a great deal and drink it every day, so I don’t want to take things too far and end up dying younger than I should.
This study in BMJ Open therefore sounds right up my street. Titled, ‘What is the optimal level of population alcohol consumption for chronic disease prevention in England? Modelling the impact of changes in average consumption levels’, it is authored by British and Australian scientists Melanie Nichols, Peter Scarborough, Steven Allender and Mike Rayner. They’ve used a mathematical model based on data trawled from published meta-analyses (studies summing up other studies) to jiggle around with the figures in order to see the likely impact of higher or lower drinking levels on mortality in the UK.
Some of these studies have shown a benefit from moderate drinking because of reduced risk of heart and circulation diseases. But drinking also increases the risk of death through liver cirrhosis, cardiomyopathy, cancers and accidents. When you dial all these figures into the model and play around a bit, the optimum level of alcohol consumption comes out at half a unit a day. Hence the Daily Mail headlines.
Doing a study like this is problematic in the extreme. Ultimately, the results will only be as good as the meta-analyses themselves in the first place. This study involved feeding in imperfect data: for example, for some of the studies on the incidences of cancer caused by different levels of drinking, smoking was not controlled for. And there will be problems both ways because of confounding. For example, is there something unhealthy about populations who drink more, that could be skewing the data? Do people who drink more eat fewer vegetables? Or exercise less?
Cancers are diseases of ageing. Older people tend to drink more than younger people. Was this controlled for?
And then there’s the reliance in these studies on self-reported drinking behaviour. How many people tell the truth when they are asked by a medical profession about their drinking habits? All these factors will influence the result.
Overall, the tone of the article seems to be profoundly patrician in its anti-alcohol message. This paragraph is telling:
“The model showed no additional beneﬁt to chronic disease mortality if the proportion of the population abstaining from alcohol were to be increased. Results indicated that increasing the proportion of alcohol consumers in the population (drinking moderately) would result in reduced CVD mortality; however, this is of little practical relevance given that there are safer and more socially acceptable means of reducing CVD risk, and there are a number of reasons why it would be imprudent to encourage current non-drinkers to start drinking. These include the following: encouraging abstainers to start drinking while encouraging drinkers to reduce their alcohol consumption is a mixed message that may be difﬁcult to communicate and promote and reducing the number of non-drinkers may have an adverse impact on non-chronic disease health (eg, accidents and injuries).”
It seems to me to be dishonest not to tell people the truth because we are afraid that they might misunderstand it. I acknowledge that this is a difficult discussion on a number of levels, but as a wine lover, and a scientist who has followed the alcohol and health literature for many years, the recommendation of half a unit a day seems at odds with almost every other study I have seen.
However, to their credit, the authors acknowledge the limitations of the study, which is clearly and logically written up. I admire their approach and the thinking behind the study. I just hope that people can see past the newspaper headlines to the detail, because the headlines spell bad news for the wine industry if they are widely accepted by the increasingly powerful anti-alcohol lobby.
It may be that in decades to come, fine wine lovers will be driven to underground tasting clubs where illicit fine wines, a memory of a healthy, culturally rich aspect of society, are consumed in secret, away from the prying eyes of the state.