Evaluation of the Pleiotropic Effects of Statins: A Reanalysis of the Randomized Trial Evidence Using Egger RegressionBrief Report

Objective To reanalyze data from recent randomized trials of statins to assess whether the benefits and risks of statins are mediated primarily via their LDL-C (low-density lipoprotein cholesterol) lowering effects or via other mechanisms. Approach and Results We adapted Egger regression, a technique frequently used in Mendelian randomization studies to detect genetic pleiotropy, to reanalyze the available randomized control trial data of statin therapy. For cardiovascular end points, each 1 mmol/L change in LDL-C with statin therapy was associated with a hazard ratio of 0.77 (95% confidence interval, 0.71-0.84) with an intercept that was indistinguishable from zero (intercept, -0.0032; [95% confidence interval, -0.090 to 0.084]; P=0.94), indicating no pleiotropy. For incident diabetes mellitus, a 1 mmol/L change in LDL-C with statin therapy was associated with a hazard ratio of 1.07 (95% confidence interval, 0.99-1.16) and an intercept nondistinguishable from zero (intercept, -0.015; [95% confidence interval, -0.30 to 0.27]; P=0.91), again indicating no pleiotropy. Conclusions Our reanalysis of the randomized control trial data using Egger regression adds to the existing evidence that the cardiovascular benefits of statins and their association with incident diabetes mellitus are mediated primarily, if not entirely, via their LDL-C lowering properties rather than by any pleiotropic effects.