Experts debate statin acceleration of type 2 diabetes development

Recently published data has raised the question of statin use and the acceleration of type 2 diabetes. Endocrine Today reached out to leading experts in the field who are a part of this ongoing research for their input.

Helena W. Rodbard, MD, FACP, MACE: The potential association of the use of statins and acceleration of development of diabetes was initially raised by the JUPITER trial. Since then there has been a lot of controversy and confusion in the medical literature as well as in the lay press. Patients, not understanding the benefits of statins in terms of prevention of cardiovascular disease, have discontinued or failed to initiate treatment with statins, which could be otherwise indicated.

A recent study by Paul M. Ridker, MD, and colleagues indicates that the enhanced rate of developing diabetes has the effect of shifting the onset from 89 weeks to 84 weeks following initiation of rosuvastatin. The authors conclude that the enhanced risk of diabetes is so low, that it would be inappropriate to withhold statin therapy in view of its potential beneficial effects in terms of cardiovascular disease.

A comprehensive review of the literature published in Current Opinion in Lipidology in 2011 supported that conclusion. Now a study from David S. H. Bell, MD, FACE, FACP, appearing as an e-pub this month, confirms that conclusion. They note that “the clinical significance of increased glucose levels in patients treated on statins is uncertain,” and “despite the fact that higher statin doses are more likely to lead to new-onset diabetes, for every case of diabetes caused, there are approximately three cardiovascular events reduced with high dose versus moderate dose statin therapy. Overall, the small risk of developing type 2 diabetes with statin therapy is far outweighed by the potential of statins to decrease cardiac events.”

The increased risk of diabetes in a meta-analysis by Swapnil N. Rajpathak, MD, DRPH, and colleagues was small (RR= 1.13; 95% CI, 1.03-1.23). Several hypotheses have been put forward as possible mechanisms, but none have been proven. In conclusion: physicians should continue to utilize statin therapy, as appropriate, for patients with dyslipidemia to reduce risk of cardiovascular events, and should not be deterred by the potential, controversial, and very small increase in risk of acceleration of development of diabetes.

Mark O. Goodarzi, MD, PhD: In my opinion, the balance of the evidence suggests yes. This question has an interesting history. There have been a multitude of mixed results in the large statin clinical trials that have been performed over the years. In fact, one of the earliest trials called the WOSCOPS trial actually suggested a protection against diabetes when using statins. Then, several trials in the following years seemed to be more or less neutral. The issue was dying down, but then a more recent large trial called JUPITER actually found that those who were put on statins had a 25% increased risk for diabetes. This surprising result from JUPITER put this issue back on the radar screen.

Investigators subsequently performed a meta-analysis to clarify the situation. They collected 13 of the large landmark statin trials and assessed new cases of diabetes, finding an overall 9% increase in diabetes risk in patients who were started on a statin. That was pretty convincing data, and people generally agreed with it, but most people’s reaction was that the CV benefit of statins probably outweighed the negative impact of this increased risk for diabetes.

JUPITER investigators performed a sub analysis where they looked at the risk in women vs. men, and found that in women, the increased risk for diabetes was about 50%, whereas in the men it was 15%, suggesting possibly that women might be at significantly higher risk. This is what prompted me and my colleagues examine this issue. The authors of the first published meta-analysis did not examine any gender issues, to discern whether there was a difference in the risk of statin-induced diabetes in women versus men.

We went back to the literature and collected 13 statin trials, 12 of which overlapped those from the initial meta-analysis. We looked at the proportion of women in these trials and plotted that against the risk for diabetes. Sure enough, there was a striking relationship where the more women in the trial overall, the higher the risk for diabetes. In our analysis (not reported in the paper), we found that the overall increased risk for diabetes was 12%; this risk was 10% in men and 17% in women. Then it becomes clear in the older WOSCOPS trial that actually suggested diabetes prevention was 100% men. This reflects a classic problem in cardiovascular trials: the under-representation of women. The newer trials like JUPITER had a greater proportion of women (over 35%). The increasing inclusion of women in statin trials has allowed us to detect the effects of statins on diabetes, particularly as it relates to gender.

The WHI, an observational study, looked at diabetes risk with statins and found that there was almost a 50% increase, and when you plot their diabetes risk it’s right on the trajectory of the regression line from all the trials. We don’t know why women are at higher risk of statin-induced diabetes, which is a very interesting research question: Is there some interaction between statins and estrogens or is it just a matter that women tend to have smaller BMI, such that a fixed dose of a statin actually gives higher drug exposure in women vs. men? We do not know the mechanism at this point. It’s potentially very important.

Currently, there is a lot of controversy with the new AHA/ACC cholesterol risk calculator and guidelines, and a lot of people are suggesting that if health care providers follow these recommendations the number of people prescribed statins may double. Given this possibility, the concept of statins increasing risk of diabetes becomes even more important. What worries me is that in general, women have a lower CV risk, but a lot of practitioners will just measure somebody’s cholesterol, and if it looks elevated, prescribe a statin without calculating the actual cardiac risk. This common practice has a high chance of leading to inappropriate statin prescription, particularly in women. In that scenario, some of those women will get diabetes as a result. Again, it’s not tremendous – just that 17% - but that amounts to thousands of people because statins are among the most prescribed drugs in America.

A particular subgroup where this is extremely relevant is people with prediabetes. A recent sub analysis of the NAVIGATOR trial reported that in patients with prediabetes, the risk for developing diabetes on a statin may be as high as 30%. Compared with the 9% overall, that’s pretty dramatic. I favor the concept of risk calculation because it does take gender into account. Whether or not the calculator is entirely accurate, it at least puts the gender issue into context. Rather than just blindly putting everybody on a statin because of high cholesterol, calculating their risk is a necessary step in my view; not only because of the diabetes risk but also the other potential adverse events of statins.

Stephen A. Brietzke, MD: The 25 year history of statin therapy has been one of the most universally good news stories in the history of drug treatment. Every clinical trial of drugs in the statin class has shown significant outcomes benefit with regard to CV events, and several studies have shown reduced risk of CV deaths and all-cause mortality. Even better, aside from occasional myalgias/myositis (in about 3% of trial participants; arguably, a slightly larger percentage in mainstream clinical practice), the statin drug class has been as free of major treatment-limiting adverse effects as any medication, ever.

However, there is a possible downside of the universal use paradigm. Recent studies, notably including the JUPITER trial, have identified a significantly increased risk for incident type 2 diabetes among users of statins, versus users of placebo in the trials. The relative risk for incident type 2 diabetes was 1.25 for statin users in JUPITER; a separate meta-analysis of 13 large clinical trials of statins identified a relative risk of 1.09 for incident diabetes among statin users. Patients in the Women’s Health Initiative who took statins self-identified diabetes 1.48 times as often as non-statin users.

Rationally examining the data regarding new onset diabetes and statin use suggests that diabetes is precipitated in patients already at high risk for diabetes, based on multiple features of the metabolic syndrome (ie, prediabetes). My belief is that most of these patients would have developed diabetes sooner or later. Moreover, our major incentive to treat patients—especially those with prediabetes and diabetes—with statins, is to prevent CV events and CV deaths, and the data are entirely confluent and unarguable on that point.

The greater good still tips in favor of statin use in patients with a significant risk for CVD. We should keep our eye on the ball: the goal is to prevent CV deaths, and statins clearly do that very well. However, we should hold off on putting statins in the water supply.

David S.H. Bell, MD, FACE, FACP: There is evidence of a definite but small increase in the development of Type 2 diabetes with statin therapy in those who have the characteristics of the metabolic syndrome.The two mechanisms which most likely lead to statin-induced diabetes are reductions in adiponectin and ubiquinone (CoQ10) levels.Lowering of these levels lead to beta cell dysfunction and with adiponectin insulin resistance.Interestingly pitvastatin (the only statin which may decrease the risk of developing diabetes) increases adiponectin levels.However,the risk for developing statin-induced diabetes is dwarfed by the ability of statins to decrease cardiac events.

Silvio E. Inzucchi, MD: Yes, it is reasonably clear that statin therapy increases the risk for new-onset diabetes, but just modestly. A recent meta-analysis of 13 randomized clinical trials involving more than 90,000 patients found the odds ratio for incident diabetes in those assigned to statin therapy to be 1.09 (95% CI 1.02-1.17) - so only a 9% increase in the relative risk. Moreover, it was estimated that treating 255 patients for 4 years would result in 1 additional case of diabetes. To put this in perspective, another meta-analysis had already demonstrated that statin treatment over this same period of time would be expected to prevent at least 5 times as many major coronary events. It is also known that the development of diabetes on statins does not substantially mitigate the benefits of these agents on cardiovascular outcomes. As with any therapy in medicine, the clinician must weigh the risks versus the benefits at the individual patient level. The benefits of statins are clear in high-risk cardiovascular patients - any effect on diabetes risk in my view pales in comparison. Nonetheless, in the patient at high risk for developing diabetes (eg, those with impaired fasting glucose) but whose cardiovascular risk profile appears otherwise benign, the recommendation to begin a statin should no longer be reflexive.

Laurence Sperling, MD, FACC, FAHA, FACP: There is a highly suggestive association between statin use and the onset of diabetes. However, nobody has yet designed a randomized clinical trial to specifically answer that question. All the available data ,primarily through secondary trial analyses, and meta-analyses do suggest this relationship. The relationship seems to be greatest with those of older age, those who already have a predisposition for diabetes and qualities of the metabolic syndrome and the relationship does appear to be most highly associated with the high intensity statins.

For every individual patient for whom you’re considering statin therapy, you should weigh the indications keeping in mind the benefits of statins, which is anywhere between a 25% to 40% relative risk reduction of combined cardiovascular endpoints. Even though you should be attuned to the potential for diabetes and those patients should be monitored for the development of new incidence of diabetes, in no way would you ever stop the statins if someone should cross the threshold of diabetes.

If you feel that a statin is beneficial in that individual, you would focus your efforts more effectively on the diabetes prevention program, which includes a Mediterranean diet, regular exercise, moderate weight reduction and maintenance of that comprehensive plan.

To put it into perspective, in a population based upon the data, there’s anywhere from one to two excess cases of incident diabetes for 1,000 people treated for the year with a moderate to high intensity statin. There would be anywhere from a 6 to 10 event reduction for those 1,000 people for a year. The benefits in the individual where a statin is indicated would outweigh any of the risks of diabetes multiple fold.

With cardiovascular event reduction, you are reducing the likelihood of an overt event. You’re weighing event reduction vs. crossing the threshold for disease (diabetes) , but that disease can still be aggressively and appropriately managed.

Disclosures: Bell, Brietzke, Goodarzi and Sperling report no relevant financial disclosures. Inzucchi reports having served as a consultant for BMS, Boehringer Ingelheim, Janssen, Merck and Novo Nordisk. Rodbard reports financial ties with AstraZeneca, Biodel, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly & Co, Halozyme, Merck & Co, Novartis, Novo Nordisk, Roche Diagnostics, Sanofi and Takeda.