Diabetes is a serious condition associated with an increased risk of heart and vascular disease, blindness and kidney failure. To put it in perspective your risk of having a heart attack if you are diabetic is the same as the risk of a person who has already had one heart attack having another one. We take diabetes seriously and spend a lot of time trying to prevent complications from condition. We also try to prevent diabetes by telling people about the lifestyle changes which will reduce their risk.
In England every adult aged 40 to 74 is offered a NHS Health check designed to detect obesity, diabetes, high blood pressure and raised cholesterol. Doctors are told to calculate a patients cardiovascular risk and decide whether to prescribe statins. Originally the threshold of risk for statin prescription was 30%, then a few years later, when branded statins came off patent, it was lowered to 20% and soon, if new draft NICE guidelines go through, it will be lowered to 10%.
As the prescription of statins has moved from people with established medical problems to those simply at risk concerns have been raised about the threat of medicalising the population. With this there has been growing worry about the unintended effects of statins. What has emerged is that statins definitely cause muscle problems and abnormalities of liver function tests but more worryingly there is growing evidence that statins cause diabetes.
Randomized controlled trials prove that statins reduce risk of cardiovascular disease by 30%. This is an impressive number but remember the absolute benefit for an individual patient is absolutely determined by their actual risk of heart disease. If this number is low then the absolute benefit will also be low, 30% of a very small number is still a very small number. So in lower risk people the clinical benefit may be balanced or even outweighed by potential harm from unintended effects.
A recent meta-analysis suggested that statins are associated with a 9% increased relative risk of diabetes and a systematic review reported an odds ratio of 1.31 for the development of diabetes. Another analysis of 5 statin trials with 32,752 participants without diabetes at baseline showed that 2749 developed diabetes (2.0 extra cases/1000 patient-years) with an odds ratio of 1.12 for new-onset diabetes. In women where the beneficial effects of statins are less clear, the Women’s Health Initiative study reported evidence of increased risk of diabetes with a number needed to harm of 44 patients for an additional case of diabetes over three years of treatment.
We think the diabetes epidemic is related to the rising tide of obesity and slothfulness but could it be that the increase is being driven by the prescription of statins which have reached epidemic proportions over the last 10 years.
In 2013 doctors wrote 63 million prescriptions for over 2.13 billion statins tablets in England. If being on a statin increases the risk of diabetes and we use the conservative estimate of 2 cases per 1000 patient year then because the number of people statins are prescribed to is so large (5.84 million people taking statins) we could expect at least 12,000 extra cases of diabetes to occur as a direct result of the statin treatment. A more pessimistic scenario based on the Women's Health Initiative number needed to harm of 44 would predict an extra 133,000 cases of diabetes as a direct result of statin treatment.
The number of people with diabetes is rising by about 140,000 per year. Now association does not prove causation but the trend rise in diabetes and statin prescribing is worrying and should make doctors thick twice when recommending statin treatment to people at lower cardiovascular risk where the potential for harm from the treatment might outweigh the potential for benefit.
Diabetes is now very common so doctors will never be surprised when new cases occur especially in people who are taking statins, after all these are the people at risk of cardiovascular disease in whom diabetes is likely to be prevalent anyway. Individual doctors will miss the association between diabetes and statins because the background level of the diabetes is very high. It is much easier to identify unwanted effects of a drug if it causes a rare or unusual effect because these events stand out.
So we need to think carefully about the balance of risk and benefit when we prescribe statins, after all we need to ensure that overall the patient is likely to benefit from the treatment rather than be harmed by it especially if their underling cardiovascular risk is low.