First drinking ‘protects your heart’. Then it ‘shortens your life’. Who to believe?
Health researchers should tread carefully to avoid the alcohol-guidelines morass
Other well-regarded studies have reported that drinking one to two units of alcohol a day markedly protects against death from cardiovascular disease. Photograph: Johnny Green/PA Wire
A headline in the Guardian newspaper on April 13th exclaimed: “One extra glass of wine a day will shorten your life by 30 minutes’’ – referring to a study by Angela Ward and others published in the Lancet. The story uncritically presents the paper’s conclusions as remarkable new findings. But, on examination, the findings are not very remarkable and are open to criticism.
Irish guidelines on the maximum amount of alcohol that can be consumed per week to ensure low risk to health are 17 standard drinks per week for men and 11 for women. A standard drink contains 10 grams of alcohol and is equivalent to half a pint of beer, a single measure of spirits or a small (125 millilitres) glass of wine. The maximum low-risk intake of alcohol per week for a man is therefore 170 grams and 110 grams for a woman. Some other countries have different guidelines, eg the UK recommends a maximum weekly intake of 112 grams of alcohol for both men and women.
The comprehensive study reported in the Lancet surveyed data from nearly 600,000 drinkers included in 83 studies carried out in 19 countries. The authors conclude drinking more than 100 grams of alcohol per week shortens lifespan. Compared to those who reported drinking 0 to 100 grams of alcohol a week, a 40-year-old man drinking 100 grams to 200 grams per week shortens his life by six months, drinking 200 grams to 300 grams per week shortens his life by one to two years and drinking more than 350 grams per week shortens his life by four to five years.
Starting at about 150 grams of alcohol per week, the Lancet data show an increase in mortality for some cardiovascular diseases; heart failure, fatal hypertensive disease and fatal aortic aneurysm, and for stroke, but not for myocardial infarction (heart attack). The authors conclude drinking alcohol in excess of 100 grams per week is deleterious for health and that the many countries recommending low-risk drinking levels greater than 100 grams per week – such as Ireland, Spain, Italy, Portugal and the US – should reduce their levels.
However, the Lancet paper data also show that all-cause mortality is basically unaffected by alcohol consumption up to 100 grams per week and that people who drink up to 150 grams of alcohol per week have a lower than average risk of all cardiovascular disease. These data are not emphasised, only the dangers of drinking more than 100 grams per week for some types of heart disease.
The Lancet study conclusions can also be criticised on the basis of insufficiently controlling for important confounding variables. Correlating alcohol consumption with various health outcomes doesn’t necessarily imply causation. A famous 1981 study linked pancreatic cancer with coffee drinking. However, that study did not control for smoking. People who drink lots of coffee also tend to smoke and it was actually the smoking, not the coffee drinking, that caused the cancer.
The Lancet study controlled for age, sex, smoking, diabetes and other conditions, but controlled insufficiently for other important confounders associated with cardiovascular disease such as diet, exercise, socioeconomic status and ethnicity. In other words, alcohol may not be fully responsible for the negative impacts on health reported. Also, levels of alcohol consumption were mostly “self-reported “, but people regularly underestimate how much they drink.
There is reason to suspect the Lancet paper overestimates the risk of low-level drinking and that the Irish recommendation of 170 grams per week for men and 110 grams for women is more reflective of the real risk than a recommendation of 100 grams for all. Certainly, the Lancet results do not explain why west Europeans live longer than north Americans, despite drinking more.
Several other well-regarded studies have reported that drinking one to two units of alcohol per day markedly protects against death from cardiovascular disease. The Lancet paper casts doubt on this, but health researchers should tread carefully to avoid falling into the morass nowadays inhabited by general nutritional advice: this month this food is good for you, next month it is causing cancer. After a while the public stops paying attention to all advice; the good as well as the bad.
William Reville is an emeritus professor of biochemistry at UCC