Statins and diabetes: is there any difference between the different statins?

We read with great interest the paper from Thakker et al 1 discussing the risk of developing diabetes during statin exposure. After a systematic literature review and network meta‐analysis of 29 randomized trials including 163,039 participants, the authors found a significant association for statins in general, atorvastatin or rosuvastatin in particular, but not with other statins. 1 Since these results were obtained from clinical trials, we decided to investigate this safety signal in real conditions of life, using reports of adverse drug reactions (ADRs) in a pharmacovigilance database. Using Vigibase®, the World Health Organization (WHO) Global Individual Case Safety Report (ICSR) database , which included until March 2017 over 14 million reports, we performed a disproportionality analysis for the signal of diabetes with statins using the case/noncase method. ICSRs were included until March 16, 2017, whatever the country of origin and only if age (≥18 y) and gender were known. Doses were not taken into account since they are not exhaustively recorded in Vigibase®. Cases were ICSRs with diabetes and noncases all other ICSRs reports registered during the same period in Vigibase®. Diabetes cases were defined as reports registered under the 2 HLT MedDRA terms “diabetes mellitus” or “hyperglycemic conditions” in the SOC (system organ class) “Metabolism and Nutrition Disorders.” Drug exposure to statins was identified using Anatomical Therapeutic and Clinical (ATC) code C10AA (HMG CoA reductase inhibitors defined as “suspected” or “concomitant”). Statins included were atorvastatin, cerivastatin, fluvastatin, lovastatin, pitavastatin (alias itavastatin), pravastatin, rosuvastatin, and simvastatin. Strength of the link between exposure to statins