March 4, 2015

expert reaction to new study on statins and diabetes

A paper published in the journal Diabetologia has attempted to analyse the effect of taking statins on incidence of type 2 diabetes in men, reporting a small increase in absolute risk. These comments were accompanied by a Before the Headlines analysis.

Prof. Stephen Evans, Professor of Pharmacoepidemiology at the London School of Hygiene & Tropical Medicine (LSHTM), said:

“This research adds little to our current knowledge and should not be used by patients to decide whether they should take, or continue to take, a statin.

“The study, slightly confusingly presented, compares people who were taking a statin at an initial measurement (when they were aged between 45 and 73) with those who were not taking any statin at that time. They were followed up for an average of nearly six years.

“The users of statins were generally at higher risk of heart disease and differed in a number of ways, probably including several ways that were not measured (no direct dietary measurements are reported). As expected, those who took a statin had a slightly increased risk of increasing their blood sugar. (Diabetes is defined by measurements of blood sugar or similar factors).

“The paper does not properly discuss the absolute increase in the risk of having a diagnosis of diabetes. About 1 in 100 people per year of follow-up would have such a diagnosis made on the basis of blood tests. Taking a statin might increase that risk to about 3 people in 200 (1.5%).

“While the introduction mentions that statins are effective in treating and preventing cardiovascular disease in those with as well as in those without diabetes, the magnitude of this reduction is not set in context with the headline figure of 46% increase in diabetes.  Their cited reference 8 (Sattar et al) suggests that for every extra case of diabetes in the trials of statins, about 5 extra cardiovascular events (CVD) are prevented.

“The risks of CVD in the trials may be higher than in those treated with statins in the population, but this paper unnecessarily raises a scare about diabetes and statins, using relative percentage increases that make for headlines.

“The adjustments made for the differences between the statin users and non-users are necessarily incomplete, so the effect in reality is likely to be less than the authors state. A strength of their study is that they have used a variety of methods to ascertain diabetes (as defined by the various tests), but this changes the absolute risks and not necessarily the relative risks, so that their results are not as different from the high-quality randomised trials as they suggest.

“The authors note that most of those they studied used either simvastatin or rosuvastatin.  Consequently, when they say that “other statins did not increase the risk of diabetes” this may at least be partially explained by the small numbers using other statins.”

Prof. Keith Frayn, Emeritus Professor of Human Metabolism at the University of Oxford, said:

“This paper confirms a well-known side-effect of statin treatment, the increased risk of development of type 2 diabetes. The figure of 46% increased risk sounds alarming and it is important to put this into perspective.

“The figure of 46% relates to relative risk. In fact 11% of those taking a statin developed diabetes, compared with 6% of those not taking a statin – and that apparent difference was considerably reduced when other factors such as BMI and family history of diabetes were taken into account.

“It should be remembered that this was a not a randomized controlled trial (RCT) of statin therapy: some patients were taking a statin at the beginning of the observation period, some not. Whilst the authors have tried to control for confounding factors (other differences between the statin-treated and non-treated groups) we cannot know for certain that such differences did not contribute towards the observed effect.

“In contrast, other studies including meta-analyses of RCTs of statin therapy show the same effect, but to a lesser degree (closer to 10% increased risk of diabetes). The authors of the current study suggest that in these earlier studies the development of diabetes was under-estimated because the measurements were not made so carefully as in this study.  We would only know the true answer by conducting RCTs in larger numbers of people and using strict criteria for diagnosing diabetes.

“The evidence is still clear that statins reduce overall mortality. That’s true even in studies with extended follow-up beyond the period of randomisation, and with no increased cancer risk as had been feared at one time.”

Prof. Tom Sanders, Professor emeritus of Nutrition and Dietetics at King’s College London (KCL), said:

“The benefit from taking statins is that it reduces the risk of both fatal and non-fatal cardiovascular disease which is also the major health risk of type 2 diabetes.

“The risk of developing type 2 diabetes increases enormously even with small increases body weight compared with taking statins. For people prescribed statins, risk of developing type 2 diabetes can be reduced by regular physical activity (30 min brisk walking daily) and maintaining a healthy weight.”

Prof. Naveed Sattar, Professor of Metabolic Medicine at the University of Glasgow, said:

“This is an interesting study that broadly supports what we already understand – namely that statins are associated with a slightly raised risk of diabetes.

“However, one must be careful to suggest this study is more meaningful or more accurate in estimating magnitude of risk than the data from the clinical trials since here the data are observational in nature which means important biases – e.g. why do some patients get statins and other did not? – are simply impossible to ever fully account for.

“The trials suggest overall risk increases around 12% for modest dose statins and a further 12% once you move to more intensive (so ~25% higher risk on high dose versus placebo).  It does, however, suggest at mechanisms for more diabetes but, again due to study design, cannot definitively prove these.

“In terms of clinical guidance, the present report does not alter anything as the statin-diabetes link is already in clinical guidelines and checking for diabetes risk in patients about to commence statins is now also recommended in some guidelines. Equally, clinicians should already be warning their patients about any diabetes risk and use this as an incentive to get them to take lifestyle changes more, not less, seriously.”

‘Increased risk of diabetes with statin treatment is associated with impaired insulin sensitivity and insulin secretion: a 6 year follow-up study of the METSIM cohort’ by Henna Cederberg et al. published in Diabetologia on Wednesday 4 March 2015. 

All our previous output on statins can be seen at this weblink: http://www.sciencemediacentre.org/tag/statins/

Declared interests

Prof. Sattar:

I was lead or senior author on some of the papers that led to FDA change of label

Have consulted for Astrazeneca and Kowa and Amgen and Sanofi on lipid-lowering agents

Prof. Frayn:

Prof Frayn has no conflict of interest with the findings of the paper nor with its authors.  He is a scientist working the area of fat metabolism and author of ‘Lipid Biochemistry: an Introduction’ (edn 5, Blackwell Science) and ‘Metabolic Regulation: a Human Perspective’ (edn 3, Wiley).

Prof. Evans:

I have no financial conflicts of interest. I was on the BMJ international panel convened to decide if two papers that presented interpretations of data suggesting risks of statins were greater than their benefits in patients at low risk. Both papers published a major correction but, given the corrections, the panel agreed unanimously that the papers should not be retracted (as was requested by many proponents of statins), so I am not in the “pro-statin” camp, but nor am I in the “anti-statin” camp- I am at relatively low risk of CVD given my age but I do take simvastatin.