Research published today uses genetic evidence to link alcohol consumption to a range of cardiovascular diseases, including stroke, peripheral artery disease, and coronary heart disease.
The investigation, published in Circulation: Genomic and Precision Medicine, led by the Karolinska Institute and the MRC Biostatistics Unit, University of Cambridge, considers over 350,000 individuals from UK Biobank, a population-based cohort of UK residents, combined with data from large consortia of cardiovascular studies from the UK and other Western countries. People with a genetic predisposition to drinking more alcohol were shown to have increased risk of several cardiovascular diseases. This suggests that increasing alcohol consumption raises the risk of these diseases.
While several observational studies have shown that moderate drinking correlates with decreased risk of cardiovascular disease, such studies are subject to bias due to confounding factors, as alcohol drinking is associated with other risky behaviours. A further complication is reverse causation – the observed correlation between alcohol drinking and improved cardiovascular health may be driven by individuals with a prior health condition reducing their alcohol intake.
In contrast, by investigating genetic predictors of alcohol consumption, reverse causation is not possible as the genetic code cannot be affected by disease, and links with confounding factors are less strong. Due to the inherent randomness in genetic inheritance, comparisons between different genetic subgroups of the population are similar to comparisons between treatment arms in a randomized trial. When comparing genetically-defined subgroups of the population that only differ with respect to alcohol drinking, researchers consistently show that the subgroup with greater alcohol consumption has higher cardiovascular risk.
Susanna Larsson, senior researcher and associate professor of cardiovascular and nutritional epidemiology at Karolinska Institute in Stockholm, Sweden said: “Higher alcohol consumption is a known cause of death and disability, yet it was previously unclear if alcohol consumption is also a cause of cardiovascular disease. Considering that many people consume alcohol regularly, it is important to disentangle any risks or benefits.”
The approach taken in this paper is referred to as Mendelian randomization. In a randomized controlled trial, we typically compare subgroups of the population: one subgroup of individuals who are randomized to receiving an intervention to reduce alcohol consumption versus another subgroup who are randomized to receiving no intervention (also called the “control” or “placebo” group). In Mendelian randomization, we use genetic variants to define subgroups of the population. This is an example of a naturally-randomized trial – rather than randomizing individuals into subgroups ourselves, we rely on genetic variants to divide the population into subgroups.
Evidence from Mendelian randomization investigations is not as convincing as evidence from randomized trials, but it is more convincing than evidence from a standard observational analysis. In this case, a long-term randomized trial for alcohol consumption is not feasible, and so evidence from naturally-randomized trials is the best available evidence we have. In this paradigm, the genetic variants are equivalent to randomization, but it is alcohol that is the risk factor. Just as in a randomized trial, it is not randomization itself that is the causal factor, but the intervention, in Mendelian randomization it is not the genetic variants that are the causal factor, but alcohol consumption. Hence while being genetically predisposed to drinking more alcohol is associated with increased cardiovascular risk, the take-home message from this research is not that the genetics is the key factor, but consumption of alcohol.
While this study was unable to demonstrate a specific link with cardiovascular disease in light drinkers, frequency of heavy drinking amongst participants in UK Biobank was low, meaning that it is unlikely that the burden of increased disease risk is restricted to heavy drinkers. A one standard deviation increase in alcohol consumption, equivalent to around a three-fold increase in number of units of alcohol consumed per week, was linked with a three-fold increase in risk of peripheral artery disease, a 27% increase in risk of stroke, and a 16% increase in risk of coronary heart disease.
Previous evidence has suggested similar results in a Chinese population for risk of stroke, and in a European-ancestry population for risk of coronary heart disease based on a single genetic variant. This study broadens the evidence base by considering a larger set of genetic predictors of alcohol consumption, and a wider set of cardiovascular diseases. Convincing evidence for a harmful effect of alcohol was demonstrated for haemorrhagic stroke and peripheral artery disease, and suggestive evidence for ischaemic stroke, atrial fibrillation, abdominal aortic aneurysm, and coronary heart disease. This research also showed an association between genetic predictors of alcohol consumption and increased blood pressure, providing a potential mechanism for increased harm of raised alcohol consumption.
Stephen Burgess, group leader at the MRC Biostatistics Unit and Cardiovascular Epidemiology Unit, University of Cambridge, Cambridge, UK, also added: “This research strengthens an emerging narrative dispelling the myth that moderate drinking is beneficial for cardiovascular health. If individuals choose to drink alcohol, they should do this in awareness of the increased cardiovascular risk.”
Image credit: Flickr