No evidence of genetic causality between diabetes and osteonecrosis: a bidirectional two-sample Mendelian randomization analysis

Abstract

Abstract Objective This study aimed to examine whether diabetes mellitus is causally associated with osteonecrosis. Method Using publicly accessible genome-wide association study statistics, a bidirectional two-sample Mendelian randomization analysis was carried out. In order to determine whether diabetes has a causal effect on osteonecrosis and whether osteonecrosis has a causal effect on diabetes, we extracted six date on diabetes in Europeans from IEU OpenGWAS and GWAS Catalogue and osteonecrosis in Europeans from FinnGen. We then evaluated the data using inverse variance weighting, MR-Egger regression, weighted median, weighted mode, and simple mode. The results’ stability and dependability were then evaluated using sensitivity analysis and heterogeneity analysis. Finally, meta-analysis is used to further confirm if there is a relationship between diabetes and osteonecrosis. Results When diabetes was used as an exposure factor, MR-Egger regression showed that directional fold product was unlikely to bias the results. Cochran’s Q test showed only minor heterogeneity in a few data sets. Multidirectional tests Egger-intercept, MR-PRESSO and funnel plots for most data did not show multidirectional and asymmetry at the gene level. Most of the IVW results showed no causal relationship between diabetes mellitus and osteonecrosis. The results of meta-analysis of IVW methods further confirmed the absence of a causal relationship. Inverse MR analysis also showed no causal relationship between osteonecrosis and diabetes. Conclusion Results of bidirectional MR analysis show no evidence of causal relationship between diabetes and osteonecrosis.