Additional file 1 of Identifying molecular mediators of the relationship between body mass index and endometrial cancer risk: a Mendelian randomization analysis

Additional file 1: Appendix S1. Supplementary methods. Appendix S2. List of unique molecular traits identified from literature search. Appendix S3. PubMed search terms for literature review. Table S4. Additional information on GWAS, including covariates adjusted for. Table S6. Results of female BMI sensitivity analysis MR. Table S7. Results of female-specific SNP fasting insulin sensitivity analysis MR. Table S8. Results of female-specific SNP BMI sensitivity mediation analysis. Table S9. Results of female-specific SNP fasting insulin mediation analysis. Table S10. Results of post-hoc power analysis for MR of all molecular traits on endometrial cancer risk, and BMI on all traits confirmed to have a causal effect on endometrial cancer. Table S11. Conditional F-statistics with different levels of genetic correlation for SHBG and bioavailable testosterone and BMI in multivariable Mendelian randomization analyses. Table S12. Conditional F-statistics for fasting insulin (adjusted for BMI) and BMI with differing thresholds used to construct fasting insulin. Table S13. Conditional F-statistics for multivariable Mendelian randomization of BMI and mediators on endometrial cancer risk. Table S14. Conditional F-statistics for further multivariable Mendelian randomization analyses. Table S15. Sample overlap between GWAS. Figure S17. Leave-one-out analysis for MR examining the effect of adult BMI on overall endometrial cancer risk. Table S18. Results from sensitivity analyses examining the influence of Winner’s curse on GWAS with overlapping samples. Table S19. Results from sensitivity analyses examining the influence of Winner’s curse on GWAS with overlapping samples in analyses determining the mediating role of traits in the relationship between BMI and endometrial cancer risk. Figure S20. Leave-one-out analysis for MR examining the effect of adult BMI on endometrioid endometrial cancer risk. Figure S21. Leave-one-out analysis for MR examining the effect of adult BMI on non-endometrioid endometrial cancer risk. Figure S22. Leave-one-out analysis for MR examining the effect of total testosterone level on endometrial cancer risk. Figure S23. Leave-one-out analysis for MR examining the effect of bioavailable testosterone level on endometrial cancer risk. Figure S24. Leave-one-out analysis for MR examining the effect of fasting insulin level on endometrial cancer risk. Figure S25. Leave-one-out analysis for MR examining the effect of SHBG level on endometrial cancer risk. Figure S26. Leave-one-out analysis for MR examining the effect of total serum cholesterol level on endometrial cancer risk. Table S27. Results of female-specific SNP CRP sensitivity analysis MR. Figure S28. Leave-one-out analysis for MR examining the effect of total testosterone level on endometrioid endometrial cancer risk. Figure S29. Leave-one-out analysis for MR examining the effect of bioavailable testosterone level on endometrioid endometrial cancer risk. Figure S30. Leave-one-out analysis for MR examining the effect of fasting insulin level on endometrioid endometrial cancer risk. Figure S31. Leave-one-out analysis for MR examining the effect of SHBG level on endometrioid endometrial cancer risk. Figure S32. Leave-one-out analysis for MR examining the effect of adult BMI on fasting insulin level. Figure S33. Leave-one-out analysis for MR examining the effect of adult BMI on SHBG level. Figure S34. Leave-one-out analysis for MR examining the effect of adult BMI on bioavailable testosterone level. Figure S35. Leave-one-out analysis for MR examining the effect of adult BMI on total serum cholesterol level. Figure S36. Leave-one-out analysis for MR examining the effect of adult BMI on C-reactive protein level. Table S37. Results of multivariable MR mediation analysis examining the effect of BMI and endometrial cancer with fasting insulin as a potential mediator with BMI-adjusted fasting insulin instrument and 100 SNPs from BMI instrument. Table S38. Results of sensitivity analysis examining the effect of only including 100 SNPs for the BMI instrument in MR analyses. Table S39. Results of multivariable MR mediation analysis examining the effect of BMI and endometrial cancer with fasting insulin as a potential mediator with 100 SNPs from BMI instrument. Table S40. Results of multivariable MR mediation analysis examining the effect of BMI and endometrial cancer with fasting insulin as a potential mediator with BMI-adjusted fasting insulin instrument. Table S41. Results of multivariable MR mediation analysis examining the effect of endometrial cancer with pairs of confirmed mediating molecular traits without BMI. Table S42. Results of multivariable MR mediation analysis examining the effect of BMI and endometrial cancer with all confirmed mediating molecular traits with BMI-adjusted fasting insulin. Table S43. Results of multivariable MR mediation analysis examining the effect of BMI and endometrial cancer with pairs of confirmed mediating molecular traits. Table S44. Results of multivariable MR mediation analysis examining the effect of BMI and endometrial cancer with all confirmed mediating molecular traits including fasting insulin adjusted for BMI with 100 SNPs from BMI instrument. Table S45. Results of multivariable MR mediation analysis examining the effect of BMI and endometrial cancer with pairs of confirmed mediating molecular traits with 100 SNPs from BMI instrument. Table S46. Results of multivariable MR mediation analysis examining the effect of endometrial cancer with all confirmed mediating molecular traits without BMI. Table S47. Results from sensitivity analyses examining the influence of Winner’s curse on GWAS with overlapping samples in analyses determining the interdependent effects of mediators of the relationship between BMI and endometrial cancer risk. Table S48. Results of HEIDI test-filtered low-density lipoprotein (LDL) cholesterol and overall endometrial cancer MR.