March 22, 2017

expert reaction to research on moderate drinking and heart conditions

A new population-based cohort study published in the The BMJ investigates the association between alcohol consumption and cardiovascular disease.

Dr James Nicholls, Director of Research and Policy Development at Alcohol Research UK, said:

“This large-scale study provides strong evidence that the so-called ‘J-curve’ exists: meaning that, in most cases, moderate drinkers are less likely to suffer a heart condition than either heavy drinkers or people who don’t drink at all.  It is an important contribution to the evidence on a controversial subject and its findings should be taken seriously.

“A large body of existing research suggests that moderate drinkers have a reduced risk of heart conditions; however, there have been a number of recent challenges.  It is, for instance, possible that ‘sick quitters’ (i.e. people who have stopped drinking due to illness) skew results, that older drinkers who take part in such surveys tend to be relatively fit (so-called ‘healthy survivors’), or that data misses drinking patterns in people’s teens and twenties that may affect health outcomes in middle age.

“This study addresses a number of those concerns.  The authors are careful to separate out people who have never drunk, those who have quit, and occasional drinkers from moderate and heavy drinkers.  Their findings support the ‘sick-quitter’ hypothesis by showing generally poorer outcomes for ex-drinkers, but they still show increased risks for non-drinkers compared to those who drink moderately.

“The use of a large dataset of anonymised medical records means the study participants are not self-selecting and the authors adjust carefully for a range of potential confounding factors.  The sensitivity analysis also supports the general findings.

“A particular strength of this study is that, because of the numbers involved, the researchers are able to separate out different risks for different heart conditions.  The idea that ‘alcohol is good for the heart’ overlooks the fact that there are a wide range of heart conditions and that, knowing what we do about the biological effects of alcohol, the effect of drinking on these conditions should vary.  This study confirms that alcohol produces different patterns of risk depending whether we are looking at, say, myocardial infarction or angina.

“The study has inevitable limitations: the data does not provide records going back to the age at which most people start drinking, so can’t fully account for changing patterns of consumption over time.  It relies on consumption levels as reported to medical professionals, so those levels may (as is the case with most self-reported consumption) be lower than is actually the case.  The study also can’t analyse the full range of ways in which drinking interacts with other socioeconomic conditions or lifestyle factors that may shape overall risks.  We still need to know more about the precise mechanisms through which apparently protective effects emerge, and we still can’t entirely rule out other confounders.

“While the findings provide convincing evidence for protective effects, the authors sensibly point out that this doesn’t mean it would be wise to take up drinking in order to lengthen one’s life – not least because any protective effects tend to be cancelled out by even occasional bouts of heavier drinking.  There are better ways to strengthen the heart such as exercise and good diet.  All things being equal (and given the increased risk of suffering other health conditions linked to any amount of alcohol consumption), if you drink within the existing guidelines it is unlikely that alcohol will either lengthen or shorten your life.  Nevertheless, as the authors point out, the more we know about the patterns of risks around alcohol and the heart, the better the advice medical professionals can give where needed.”

Prof Aedin Cassidy, Professor of Nutrition at UEA, said:

“It has been known for over forty years that alcohol, within recommended limits, is associated with a reduced risk of cardiovascular disease. This new large study, using clinical records to estimate alcohol intake and determine cardiovascular outcomes from around 2 million people from general practice, supports these long standing views.

“The thing that sets this study apart is the use of the NHS registers to allow researchers access to extensive datasets which can be used to deliver results that can refine public health advice, improve clinical care and help with future research.”

* ‘Association between clinically recorded alcohol consumption and initial presentation of 12 cardiovascular diseases: population based cohort study using linked health records’ by Steven Bell et al. published in The BMJ on Wednesday 22 March 2017.

Declared interests

None declared