Conference Coverage

Statins' Diabetogenic Impact Influenced by Underlying Risks

Major Finding: High-dose atorvastatin raised the relative risk for incident diabetes 24% in patients with two or more diabetes risk factors.

Data Source: An analysis of 15,056 patients enrolled in two large trials of treatment with 80 mg/day atorvastatin.

Disclosures: The TNT and IDEAL studies were funded by Pfizer. Dr. Waters said that he had received honoraria from and/or has been a consultant or adviser to Genentech, Roche, and Pfizer, Aegerion, Cerenis, Servier, Anthera, Sanofi-Aventis, and Shire. Dr. Cooper-Dehoff said that she had no disclosures.

View on the News

Statins Remain Beneficial Despite Raising Diabetes Incidence

The risk that patients will develop new-onset diabetes on statin treatment appears to be proportional to a patient’s pretreatment risk for diabetes, as well as the statin dose they receive and other drugs they may also be on.

But despite growing evidence that statin treatment seems capable of causing diabetes in a small fraction of patients who get these drugs, study results also consistently show that the reduction in cardiovascular events that statins produce in secondary prevention outweigh the small risk for incident diabetes they pose. That’s the bottom line that has come through over and over again.

Especially for secondary prevention, the diabetes risk from statins is more of a nuisance than anything else. I think you can ignore the risk; avoiding statin treatment just because of the diabetes risk is wrong. If you stop using statins, what else can you use in a high-risk patient who needs a 30%-50% reduction in low-density lipoprotein cholesterol? That level of reduction is very hard to achieve without a statin. I’m not happy about the risk, but currently there are no good alternatives.

Ideally, we should develop new strategies that could reduce the risk, such as possibly earlier treatment with a bile-acid sequestrant or niacin. Another important goal is to try to figure out how statins trigger diabetes. It would be spectacular if the mechanism was identified and someone came up with a statin that did not cause diabetes.

Regarding other drug classes that also seem capable of triggering diabetes, judgment must also be used and alternative drug classes considered. The diabetes-causing potential of beta-blockers and thiazide diuretics can in some cases be avoided by using alternative drugs. But, as with statins, if a patient has depression and needs treatment with an antidepressant, I think it would be a huge mistake to not use that treatment out of concern that it might provoke diabetes.

Dr. P. Barton Duell is an endocrinologist and director of the lipid disorders clinic at the Oregon Health & Science University in Portland. He said that he had been a speaker on behalf of Merck, a consultant to Pfizer, Genzyme, Merck, Aegerion, and Amarin, and that he has received research grants from Genzyme, Pfizer, Bristol Meyers Squibb, and Cerenis. He made these comments as a discussant at the meeting and in an interview.



LOS ANGELES – A patient's risk for developing diabetes because of statin treatment seems to depend in part on two factors: the patient's underlying risk for diabetes, and the total number of diabetogenic drugs they receive, based on results from two separate studies reported at the meeting.

But the findings did not seem to persuade at least some physicians that the risk of new-onset diabetes posed by higher-dose statin should deter physicians from prescribing the drugs to patients who need them.

"You could say that new-onset diabetes is one strike against a patient, but sudden death from a cardiovascular event is 100 strikes. I think patients would rather risk developing new-onset diabetes than having a myocardial infarction, stroke, cardiac arrest, or cardiovascular death," said Dr. David D. Waters, who presented one of the two reports at the annual scientific sessions of the American Heart Association.

"All the patients we looked at had established cardiovascular disease, so they probably won’t live for another 30 or 40 years, and new-onset diabetes in these people is probably not as bad for at least the first 10 years as a cardiac event. Also, patients can only have new-onset diabetes once, but they can have multiple cardiovascular events. I think one edge of the ‘double-edged sword’ of statins is a lot sharper than the other," said Dr. Waters, a cardiologist and emeritus professor of medicine at the University of California, San Francisco.

His study examined the rate of incident type 2 diabetes in two large trials that both compared treatment with 80 mg atorvastatin daily against treatment with a lower statin dosage in patients with established cardiovascular disease: the TNT (Treating to New Targets) trial (N. Engl. J. Med. 2005;352:1425-35), which used treatment with 10 mg/day atorvastatin as the comparator regimen, and the IDEAL (Incremental Decrease in Endpoints Through Aggressive Lipid Lowering) trial (JAMA 2005;294:2437-45), which used treatment with 20 mg or 40 mg/day of simvastatin as the control group. The analysis excluded patients who entered the studies with preexisting diabetes or a fasting plasma glucose of 126 mg/dL or higher, as well as those with a missing fasting plasma glucose measurement at baseline or fewer than two measures during the study. After Dr. Waters and his associates applied these exclusions they had a total of 15,056 patients from the two studies, which split nearly equally between those who received 80 mg/day atorvastatin and those who received a lower-dose statin regimen.

The researchers used four criteria to define a patient’s risk for developing new-onset diabetes: a fasting plasma glucose of more than 100 mg/dL, fasting triglycerides of more than 150 mg/dL, a history of hypertension, and a body mass index of more than 30 kg/m2. They designated patients with no or one risk factor low risk, which included 8,825 (59%). Patients with two or more diabetes risk factors were considered high risk, and 6,231 patients (41%) fell into this subgroup.

In general, the greater the number of risk factors above one, the greater the rate of developing incident diabetes among patients treated with 80 mg atorvastatin, compared with those treated with a lower stating dose. For example, patients with no or one risk factor had no increased rate of incident diabetes in the higher-dose subgroup, but patients with all four risk factors on high-dose atorvastatin had a relative 36% higher rate of new-onset diabetes that just missed reaching statistically significance.

To simplify the analysis and add statistical power, Dr. Waters and his associates divided the database into patients with low or high risk. Patients at high risk on high-dose atorvastatin had a 2.4 percentage point increased rate of incident diabetes during follow-up, compared with those on lower-dose treatment, a 24% relative risk elevation that was statistically significant. In contrast, low-risk patients on high-dose atorvastatin had an incident diabetes rate identical to that of patients on lower-dose regimens.

The analysis also highlighted the increased potency of the 80 mg/day atorvastatin regimen for cutting the rate of cardiovascular events, compared with lower-dosage statin regimens. The higher dosage reduced cardiovascular events by 1.5 percentage points, a relative risk reduction of 15% that was statistically significant. The amount of risk reduction with higher-dose statin was roughly uniform across all strata of diabetes risk, Dr. Waters said.

The second report used data collected from 161,808 postmenopausal women enrolled in the Women’s Health Initiative clinical trials and observational study to examine the impact of treatment with any of four drug classes on incident diabetes. The researchers who ran this analysis excluded women with diabetes and those on a glucose-modifying drug at baseline, which left 139,554 women for the analysis. The drug classes they included as potential diabetes inducers were statins, antidepressants, beta-blockers, and thiazide diuretics. The study group included 108,096 (77%) on none of these drugs, 26,255 (19%) on one drug, 4,768 (3%) on two of these drug classes, and 435 (fewer than 1%) on drugs from three or all four classes.

   Comments ()

Recommended for You

News & Commentary

Quizzes from MD-IQ

Research Summaries from ClinicalEdge

Next Article: