Statins and Diabetes: Should we be more concerned?

It’s time for a statin battle again. Steve has his annual review and the computer pop ups are telling us to consider a statin due to his high CVD risk. He doesn’t mind a blood pressure pill but statins? Forget it. He’s heard all about the side effects – “Bob got muscle pain and couldn’t play golf” – and he knows the potential complications – “They can give you diabetes. Why would you give me that?” Does he have a point? We’ve known for years that statins are associated with an increased risk of developing diabetes, even if reported risks show wide variation.

Should this be more of a concern? It raises a number of questions yet to be satisfactorily answered. Is the type of statin important? Does the dose make a difference? And crucially how much do statins really alter blood glucose? 

Good news, then, that the most up to date data on this subject was recently published in The Lancet Diabetes and Endocrinology journal.

The authors included 19 RCTs of statin therapy vs placebo for at least 2 years duration with a total of 124000 patients included and a mean 4.3 years follow up, plus 4 RCTs comparing low to high intensity statins, with a total of 30724 patients.

Firstly, this data confirms that statins do increase the rate of new-onset diabetes compared to placebo. The risk of depends on whether you go on a low – moderate intensity statin (e.g. simvastatin 20-40mg, atorvastatin 10-20mg) or a high intensity statin (e.g. atorvastatin 40-80mg, rosuvastatin 20-40mg). 

With the former the risk is very small – while the reported 10% proportional increase sounds quite significant in reality this means going from 1.2% of patients on placebo developing diabetes to 1.3% on statin, a 0.1% absolute risk increase.

Higher intensity statins have a greater risk, with a relative risk increase of 36%, or an absolute risk increase of 1.3% (from 3.5% of patients on placebo to 4.8% on statins). 

Statins, then, do increase the risk of developing diabetes – the higher the dose or the stronger the intensity the greater the risk. But what does this actually mean in terms of glycaemia?

For low to moderate intensity statins mean HbA1c increase by just 0.06% and in the high intensity statin group 0.08%, or <1mmol/mol.

The key driver for cardiovascular disease from T2DM is not the presence of diabetes per se but the duration of exposure to hyperglycaemia and the degree of that hyperglycaemia. Statins may push some patients over the diagnostic threshold, but we appreciate glycaemia exists as a spectrum not a binary result, and so a small increase in glycaemia only leads to a small increase in risk, outweighed by the benefits of statins on major cardiovascular events. It was also confirmed that most patients subsequently diagnosed with diabetes were already close to the diagnostic threshold. 

What are the clinical implications? 

This new data allows us to counsel Steve on his concerns around diabetes and reassure him the risks are low and remain outweighed by the benefits for cardiovascular disease.

For clinicians this data also shows that statins shouldn’t be considered as a standalone treatment, but, as the linked editorial explains, part of wider holistic care: encouraging patients on proven strategies to reduce diabetes risk such as modest weight reduction and regular exercise will help mitigate potential glycaemic increases from statins while also having a positive effect on CVD risk and other areas of health.

Finally, would patients be better on a lower intensity or lower dose statin? This is a complex area. Usually, patients offered higher dose statins are the ones at highest risk of CVD with the greatest to gain, so as the authors of the paper suggest, it is very likely the benefits still outweigh the risks. In a world where international guidelines encourage clinicians to push cholesterol levels ever lower necessitating higher intensity statins, it is reasonable to ask how favourable this risk:benefit remains in primary prevention, especially when statins may be given for a lifetime but this data only captures a handful of years. Ultimately, as is so often the case with medicine, greater understanding leads to more questions.