Major Finding: Over an average follow-up of 4 years, 4.9% of the statin group developed diabetes, compared with 4.5% of the control patients.
Data Source: A meta-analysis of 13 placebo-controlled trials.
Disclosures: Although there was no funding for the meta-analysis, Dr. Sattar reported that the trials included in it were supported by the pharmaceutical industry. Many of the authors of the current study have financial ties to AstraZeneca, Merck & Co, Pfizer, Sanofi-Aventis, Roche, Bristol-Myers Squibb, and other drug manufacturers.
For every 255 patients on statins for 4 years, there is 1 additional case of diabetes—a risk too small to change current recommendations for statins in patients with cardiovascular disease, or with moderate to high risk of developing heart disease.
However, the increased risk “should be taken into account if statin therapy is considered for patients at low cardiovascular risk,” Naveed Sattar, Ph.D., and colleagues wrote (Lancet 2010 Feb. 16 [10.1016/S0140-6736(09)61965-6
Dr. Sattar of the Glasgow Cardiovascular Research Centre at the University of Glasgow, Scotland, searched Medline, Embase, and Cochrane Central Register of Controlled Trials for studies with the term “statin” as a title word and keyword between 1994 and 2009. The investigators included in their analysis English-language studies with at least 1,000 patients and a mean follow-up of at least 1 year, and excluded trials in patients with organ transplants, with diabetes, or on hemodialysis.
Overall, the researchers looked at 13 placebo-controlled and standard care-controlled trials with a total of 91,140 nondiabetic patients, including JUPITER (Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin), CORONA (Controlled Rosuvastatin Multinational Study in Heart Failure), and the MEGA trial (Management of Elevated Cholesterol in the Primary Prevention Group of Adult Japanese).
There were a total of 45,521 patients taking statin therapy, and 45,619 controls. Over an average follow-up of 4 years, 2,226 patients (4.9%) in the statin group developed diabetes, as did 2,052 control patients (4.5%).
The 174 extra cases amounted to a 9% increase in diabetes likelihood, or “one additional case of diabetes per 255 (95% [confidence interval] 150–852) patients taking statin therapy for 4 years,” wrote Dr. Sattar and colleagues, for an incidence of 12.2 cases per 1,000 patient-years with statin treatment, and 11.3 cases per 1,000 patient-years for controls.
The risk was similarly elevated even in a subanalysis of only those patients in placebo-controlled trials (n = 75,507, odds ratio 1.10, 95% CI 1.01–1.20), though “the association weakened slightly … when we analysed only trials that used fasting glucose measurements [to identify diabetes patients]—possibly because of a loss of statistical power” (n = 75,033, OR 1.07, 0.97–1.17).
No mechanism has yet been implicated for the increased risk of diabetes among statin users. The authors of the current analysis postulate that the risk could be due to confounders—for example, patients not on statin therapy suffer cardiovascular events and subsequently are likely to change their diet and exercise habits.
In an accompanying editorial, Dr. Christopher P. Cannon of Brigham and Women's Hospital and Harvard University, Boston, pointed to a 2005 article by Dr. Sattar, which “estimated that 5.4 deaths or myocardial infarctions would be avoided over those 4 years [taking statins], and nearly the same number of strokes or coronary revascularisation procedures would also be avoided.
“Therefore the benefit in preventing total vascular events to the risk of diabetes is a ratio of about 9:1 in favour of the cardiovascular benefit,” he wrote.