Abstract
PURPOSE: We used Mendelian randomization (MR) analyses to examine the potential causal effects of type 2 diabetes and glycemic traits on bladder cancer risk. METHODS: Two-sample MR analyses were conducted using summary data from genome-wide association studies (GWAS). Exposures included type 2 diabetes, fasting glucose, glycosylated hemoglobin (HbA1c), fasting insulin, and proinsulin levels, with bladder cancer as the outcome. Four methods-inverse variance weighted, MR-Egger, weighted median, and weighted mode-were used to assess the causal effects. Sensitivity analyses were conducted to ensure that the results were robust. RESULTS: In the inverse variance weighted model, a weak positive effect was detected between genetically predicted HbA1c and bladder cancer (OR = 1.003, 95% CI = 1.0001 to 1.0052, P = 0.043). Other MR methods produced results with the same trend, although not all were statistically significant. However, there was no evidence to support the effect of type 2 diabetes, fasting insulin, or proinsulin levels on bladder cancer. No significant heterogeneity or pleiotropy was detected. CONCLUSION: Mendelian randomization analysis indicated a mild promoting effect of increased HbA1c levels on bladder cancer risk. Further studies with larger sample sizes are needed to confirm this hypothesis.