We’re often told that statin (cholesterol-reducing) drugs are effective for preventing heart disease and stroke (cardiovascular disease). Whether you buy into this statement or not depends, to a degree, on how you look at the data. It is true that statins reduce the risk of heart attack by about a third. However, this ‘relative risk’ reduction needs to be taken into the context of overall risk to begin with. For example, if the risk of succumbing to a heart attack is, say, 50 per cent over the next five years, then a third reduction in risk (about 17 per cent risk reduction) is perhaps meaningful. But if the underlying risk was, say, 3 per cent, then the overall risk reduction (what is known as the ‘absolute risk’ reduction) is a mere 1 per cent. As it turns out, many individuals on statins have low base risk and are therefore unlikely to see very significant benefits in terms of disease prevention as a result.
Another way to measure the impact of statins is to assess their effect on overall risk of death. Here, we can see if (over a finite period of time) those taking statins are statistically less likely to die than those not taking them. Now, it turns out that in individuals with no prior history of cardiovascular disease, statins do not reduce risk of death.
I was interested to read a study which was published recently in the journal PLoS Medicine which used statistical and mathematical models to calculate the likely increase in life expectancy from taking statins . The predicted benefits appeared to depend on other characteristics such as weight, blood pressure and blood fat levels (those with unfavourable profiles were predicted to benefit more). Also, younger individuals were predicted to benefit more than older individuals.
Mathematical and statistical models are unlikely to accurately predict the true outcomes of a treatment, but they’re probably better than guessing. And it turns out that when all the individuals were thrown into the mix statins, on average, allowed individuals to live without cardiovascular disease for an additional 7 months. When it comes to the all-important life expectancy question, this was increased by an average of just 3 months. I did some maths and calculated that this represents an increase in lifespan of about 0.3 per cent.
Here’s some excerpts from what the journal’s editor had to say about this study:
Current guidelines recommend that asymptomatic (healthy) individuals whose likely CVD risk is high should be encouraged to take statins—cholesterol-lowering drugs—as a preventative measure. Statins help to prevent CVD in healthy people with a high predicted risk of CVD, but, like all medicines, they have some unwanted side effects, so it is important that physicians can communicate both the benefits and drawbacks of statins to their patients in a way that allows them to make an informed decision about taking these drugs. Telling a patient that statins will reduce his or her short-term risk of CVD is not always helpful—patients really need to know the potential lifetime benefits of statin therapy…
…The model estimated that statin therapy increases average life expectancy in the study population by 0.3 years and average CVD-free life expectancy by 0.7 years…
…These findings suggest that statin therapy can lead on average to small gains in total life expectancy and slightly larger gains in CVD-free life expectancy among healthy individuals,…
…Whether communication of personalized outcomes will ultimately result in better clinical outcomes remains to be seen, however, because patients may be less likely to choose statin therapy when provided with more information about its likely benefits.
These last words utterly sum up, I think, how it is in the real world: when individuals are informed about the true facts about the likely benefits, hardly without exception they take a pass. And that’s before we even get to talk about the potential side effects such as fatigue, muscle pain and memory loss.
1. Ferket BS, et al. Personalized prediction of lifetime benefits with statin therapy for asymptomatic individuals: a modeling study. PLoS Med. 2012 Dec;9(12):e1001361. doi: 10.1371/journal.pmed.1001361. Epub 2012 Dec 27.