No causal effects between rosiglitazone and cardiovascular disease or risk factors: a Mendelian randomization study.

ABSTRACT

OBJECTIVE: Although many observational studies have shown an association between rosiglitazone and cardiovascular disease (CVD) or risk factors, controversy remains. We conducted a Mendelian randomized (MR) study to explore whether rosiglitazone is causally related to CVDs and risk factors. PATIENTS AND METHODS: Single-nucleotide polymorphisms associated with rosiglitazone at genome-wide significance were identified from a genome-wide association study of 337,159 European-ancestry individuals. Four treatments with rosiglitazone-associated single-nucleotide polymorphisms associated with a higher risk of CVDs were used as an instrumental variable (IV). Summary-level data for 7 CVDs and 7 risk factors were obtained from UK Biobank and consortia. RESULTS: We found no causal effects of rosiglitazone, either on CVDs or risk factors. The results were consistent in sensitivity analyses using Cochran's Q test, MR-PRESSO method, leave-one-out analysis and Mendelian randomization-Egger method (MR-Egger), and no directional pleiotropy was observed. Sensitivity analyses confirmed that rosiglitazone was not significantly associated with CVDs and risk factors. CONCLUSIONS: The findings from this MR study indicate no causal relationship between rosiglitazone and CVDs or risk factors. Hence, previous observational studies may have been biased.