Alcohol – an update
My last blog on alcohol caused somewhat of a stir, as I suspected it would. To those who did not read it, I recommended that, from a cardiovascular health point of view, those who do not drink alcohol should start. I recommended this because there is strong evidence that moderate alcohol consumption significantly reduces the risk of cardiovascular disease – and can also reduce overall mortality/increase life expectancy.
There were many objections, scientific and, in some ways, moral. Because of this, I felt the need to go over the area again, which is a bit unusual for me. I think there were three main objections that were raised:
1) People who do not drink are not drinking because they have illnesses that have stopped them drinking, therefore they are less healthy than moderate drinkers to begin with. Ergo, you are not comparing apples with apples.
2) None of the studies have been randomised controlled studies, they are purely observational.
3) If people who do not drink, are advised to start drinking, a proportion of them will end up drinking too much and will damage their health.
1) People who do not drink are not drinking because they have illnesses that have stopped them drinking, therefore they are less healthy than moderate drinkers.
This is probably the easiest objection to refute. The massive one million patient study in the BMJ, that I quoted in my previous blog, looked at this potential confouder1. By which I mean that the researchers took care to separate out those who had drunk previously, from those who had never drunk.
Whilst the BMJ study looked at all sorts of outcomes, I shall restrict myself to two here. The ones that are most important. Namely, fatal cardiovascular disease (CVD), and all-cause mortality.
Increased risk of fatal CVD vs. moderate drinking
- Non-drinker = 1.32 (32% increased risk)
- Former drinker = 1.44 (44% increased risk)
Increase risk of all-cause mortality vs. moderate drinking
- Non-drinker = 1.24 (24% increased risk)
- Former drinker = 1.38 (38% increased risk)
As you can see, there is some merit to the argument that former drinkers are less healthy than never drinkers. However, if you remove former drinkers from the equation, non-drinkers remain at a significantly increased risk of CVD, and overall mortality, compared to moderate drinkers.
2) None of the studies have been randomised controlled studies, they are purely observational.
I cannot really argue too powerfully against this objection, for it is true. No-one has, to the best of my knowledge, taken a large number of people and split them into two groups. One to drink alcohol, the other to abstain. Then, after ten years or so, find out which group did better. I should point that that whilst such a trial could be randomised, and controlled, there is no way it could be placebo controlled, or double blinded (double blinded means that neither the participant or the researcher would know if the participant was, or was not, drinking alcohol). Thus, no perfect trial could ever be done.
The reality is that, in medicine and medical research you just have to roll with what you have got. In recent years, I have seen a growth in a research fundamentalist belief, which is that the only way you can ever prove anything is through a randomised placebo controlled double blind study, with tens of thousands of people in each arm.
I find this somewhat strange, and more than slightly strange. The vast, vast, majority of things that are done in medicine, have no randomised controlled studies to support. Do you think penicillin was subjected to a controlled study before it was used? Um, no. Do you think hip replacements have ever been studied in a randomised controlled trial? Um, no. Do you think breast cancer screening has ever had a single randomised controlled study? Coronary artery bypass grafts, Um, no. Almost any surgical intervention you think of. Um, no. Vaccines. Um, no.
I could keep going on for a long, long, long time on the interventions that are widely accepted, which have far less evidence to support them, than the benefits of moderate alcohol consumption. I worked with the European Society of Cardiology (ESC) at one time, to develop their educational website. By our estimate, around 13% of cardiology interventions had any evidence at all to support them (let alone randomised controlled studies). This statistic may have improved, but I doubt it.
Much of practice was defined by ‘expert consensus’. Which I also call ‘Eminence Based Medicine’.
My view is that, to dismiss all evidence that does not fit into the ‘gold standard’ of placebo controlled randomised double blind study is easy – of course. But if you are going to do this, you would have to also dismiss all the evidence on smoking and lung cancer – for example.
Certain things will never, can never, be studied in randomised controlled studies. So, we must look at best possible evidence, and make decisions based on that. Otherwise what are we to do? We can chuck all antibiotics, and vaccines, into the dustbin for starters.
3) If people who do not drink, start drinking, a proportion of them will end up drinking too much and damaging their health.
This last point is clearly the most difficult to argue against. What if I do advise people to start drinking and a significant proportion become alcoholics. Will I not have done great harm? Well, of course, this is not impossible. However, I consider it highly unlikely, because non-drinkers are almost certainly a very different group of people from already drinkers. Probably highly health conscious and well controlled people.
To be frank, I suspect many non-drinkers do not drink for moral and religious reasons, and would not start drinking even if the evidence for benefit was utterly overwhelming. [Nor would I expect them to, some things are not up for discussion].
There is also the counterargument that if many were to benefit from moderate drinking, this would counterbalance the possible harm of a smaller number becoming alcoholics. The greater benefit for the greater number? Yes, I know, this is one definition of fascism, but hey…
I shall move to the example of sunbathing here. Yes, it is true that sun exposure can cause various skin cancers (probably not malignant melanoma). Doctors urge everyone to avoid the sun, almost at any cost. In doing so, we will prevent a certain amount of skin damage, and certain skin cancers. This is, of course, good.
However, as a study from Sweden demonstrated, the trade-off is that you are far more likely to die from CVD, and you will also reduce your life expectancy by about the same amount as if you smoke ‘Avoidance of sun exposure as a risk factor for major causes of death: a competing risk analysis of the Melanoma in Southern Sweden cohort.’
Women with active sunlight exposure habits experience a lower mortality rate than women who avoid sun exposure; however, they are at an increased risk of skin cancer. We aimed to explore the differences in main causes of death according to sun exposure.
We assessed the differences in sun exposure as a risk factor for all-cause mortality in a competing risk scenario for 29,518 Swedish women in a prospective 20-year follow-up of the Melanoma in Southern Sweden (MISS) cohort. Women were recruited from 1990 to 1992 (aged 25-64 years at the start of the study). We obtained detailed information at baseline on sun exposure habits and potential confounders. The data were analysed using modern survival statistics.
Women with active sun exposure habits were mainly at a lower risk of cardiovascular disease (CVD) and non-cancer/non-CVD death as compared to those who avoided sun exposure. As a result of their increased survival, the relative contribution of cancer death increased in these women. Non-smokers who avoided sun exposure had a life expectancy similar to smokers in the highest sun exposure group, indicating that avoidance of sun exposure is a risk factor for death of a similar magnitude as smoking. Compared to the highest sun exposure group, life expectancy of avoiders of sun exposure was reduced by 0.6-2.1 years.
The longer life expectancy amongst women with active sun exposure habits was related to a decrease in CVD and non-cancer/non-CVD mortality, causing the relative contribution of death due to cancer to increase.2
Why do medics write in such a convoluted way? Is there a course on ‘complete obfuscation of the reader’ that I missed somewhere along the line? Anyway, the point here is that sun exposure meant you very significantly avoid CVD death and live longer (good). But, if you die, you have more chance of dying of cancer (bad?). Of course, if you reduce death from CVD you will, by default, increase the risk of dying of cancer – well you have to die of something. Less of A means more of B.
As a cardiologist once said to me. ‘My job is to keep people alive for long enough for them to die of cancer.’ Sorry, but I do love black humour.
The general point here is that you must look at the greatest benefit to the greatest number. Could I tell a lot of people to avoid drinking alcohol because some people may, I repeat may, turn into alcoholics.
I shall leave you with a quote from an article ‘Ethanol and cardiovascular diseases: epidemiological, biochemical and clinical aspects.’
Conclusion: to drink or not to drink?
‘It is not easy to answer this Hamlet’s question, because alcohol consumption is like a razor-sharp double-edged sword. Current guidelines of the American Heart Association (AHA) state that moderate alcohol consumption is beneficial for cardiovascular health, but the AHA clearly states that non-drinkers should not begin drinking alcohol in middle age due to possible counter-balancing ill consequences of alcohol consumption. Before the definitive decision prospective randomized blinded trials would be important: engage a large pool of non-drinkers, half of whom would commence a moderate drinking regimen, whilst the other half remained abstainers.
The two groups would be followed for years in a search for eventual differences in cardiovascular disease and heart-related deaths. First possible data were available in 2008. King et al observed that of 7697 participants who had no history of cardiovascular disease and were non-drinkers at baseline 6.0% began moderate alcohol consumption and 0.4% began heavier drinking.
After 4 years, new moderate drinkers had a 38% lower chance of developing CVD than did their persistently nondrinking counterparts. Those who began drinking moderately experience a relatively prompt benefit of lower rate of CVD morbidity with no change in mortality rates after 4 years. The collected data make a strong case of the cardiac benefit of controlled drinking.’3
Thank you and cheers. Not that I expect I will have convinced anyone who objected to my last article.
3: Ginter E, Simko V. ‘Ethanol and cardiovascular disease: epidemiological biochemical and clinical aspects.’ Bratisl Lek Listy 2008: 109(12) 590-4