When I was at medical school I remember being lectured on the wonders of hormone replacement therapy (HRT). I was distinctly taught, and without reservation, that women taking HRT had a lower risk of heart disease compared to women ‘going without’. This ‘fact’ turned out to be complete rubbish. Subsequent evidence revealed that HRT actually has the capacity to increase risk of heart disease.
How could we have got it so wrong?
The initial ‘evidence’ on HRT was epidemiological in nature, which meant that it looked at the relationship between HRT/non-HRT use in a population and risk of cardiovascular disease. One fundamental potential problem with these studies relates to what is known as the ‘healthy user’ effect. In short, what this means is that healthier individuals are more likely to be prescribed a drug than sicker people who may already be on multiple medications and prone to side-effects and interactions. So, any ‘benefit’ seen to be associated with a drug may have nothing to do with the drug, and everything to do with the fact that people taking it are inherently healthier.
To untangle all of this, what we need is randomised controlled trials. These trials give essentially equivalent groups the treatment or placebo to assess any potential benefits or harms of the treatment. It’s when these studies were done that we realised that HRT actually increased the risk of heart disease.
The healthy-user effect, though, has not gone away, and is still alive and well in medical research. Here’s another apparent example that concerns cholesterol-reducing drugs known as statins…
In the past, statins have said to help prevent pneumonia (infection in the lung) on the basis of epidemiological studies. However, it is generally the case that frail, elderly individuals, with perhaps complicated health histories are less likely to be prescribed or take statins than healthier individuals. Because of this, when we see lower incidence of infection in those taking statins, we have no idea if it’s the statins, or the fact that these people are generally healthier, or both, that accounts for the reduced infection risk.
One way to get clarity here is to attempt to take into account health status of individuals when performing this sort of analysis. That’s exactly what a team of doctors based in the US did when analysing the relationship between statin use and risk of pneumonia in a study published in 2009 [1]. This more careful analysis revealed that statin use was actually associated with a 26 per cent increased risk of pneumonia. For pneumonia severe enough to require hospitalisation, statin use was associated with a 61 per cent increased risk.
Now, we should not forget that these studies are epidemiological in nature, and cannot be used to prove that statins cause enhanced susceptibility to pneumonia. However, the evidence as it stands is incriminating nonetheless. Further suspicion is raised in the form of evidence which shows that statins have the ability to directly impair the immune system and its ability to resist bacteria [2].
The most comprehensive account of statin side-effects I can find was published last year in the British Medical Journal [3]. Known side-effects of statins include muscle weakness and/or pain (myopathy), liver damage , kidney failure and cataracts. Here, in summary, are the findings of this review:
For every 10,000 women at high risk of CVD [cardiovascular disease] treated with statins, we would expect approximately 271 fewer cases of cardiovascular disease, 8 fewer cases of oesophageal cancer, 23 extra patients with kidney failure, 307 extra patients with cataracts; 74 extra patients with liver dysfunction; and 39 extra patients with myopathy.
For every 10,000 men at high risk of CVD treated with statins, we would expect approximately 301 fewer cases of cardiovascular disease, 9 fewer cases of oesophageal cancer, 29 extra patients with kidney failure, 191 extra patients with cataracts; 71 extra patients with liver dysfunction; and 110 extra patients with myopathy.
This study focused specifically on data relating to individuals deemed to be at high risk of cardiovascular disease. Many individuals who take statins are actually not at high risk of cardiovascular disease. For these, benefits are likely to be significantly lower than those elucidated in this study (while risks are likely to be about the same).
But look at those figures for a moment. Two things jump out to me:
Of 10,000 high-risk individuals, only about 300 will benefit – that’s 3 per cent. That means, of course, 97 per cent will not benefit. The number of people who benefit is roughly matched by those who will get a serious adverse effect. Hands up who wants to take a statin now?
References:
1. Dublin S, et al. Statin use and risk of community acquired pneumonia in older people: population based case-control study. BMJ 2009;338:b2137
2. Benati D, et al. Opposite effects of simvastatin on the bactericidal and inflammatory response of macrophages to opsonized S. aureus. J Leukoc Biol. 2010;87(3):433-42
3. Hippisley-Cox J, et al. Unintended effects of statins in men and women in England and Wales: population based cohort study using the QResearch database BMJ 2010;340:c2197
