Potential causal associations of 100 known and suspected influencing factors with risk of overall ovarian cancer and six histotypes: A Mendelian randomization study

Abstract

Background: Observational studies have linked various exposures to ovarian cancer (OC) risk, but the findings are potential subject to reverse causation and confounding. Herein, we performed comprehensive Mendelian randomization (MR) analyses to systematicly evaluate potential causal associations of known and suspected influencing factors with risk of OC and six common histotypes. Methods: Two-sample MR analyses were applied to data from the genome wide association study summary results comprising a total of 25,509 women with epithelial OC and 40,941 controls of European descent in the Ovarian Cancer Association Consortium. Genetic instrumental variables associated with influencing factors were selected. Inverse-variance weighted method was used as the primary analysis, and the MR assumptions were evaluated in sensitivity analyses. MR-PRESSO method was applied for the detection and correction of potential horizontal pleiotropy. Results: OC and six histotypes were considered in this study. Of 100 known and suspected influencing factors, 7 lifestyle factors, 12 dietary factors, 4 reproductive factors, 12 body size factors, 3 comobidities, and 7 biomarkers were significantly associated with risk of OC. Among them, 26, 9, 25, 19, 5, 13, and 22 factors were associated with the risk of OC, clear cell OC, endometrioid OC, high grade serous OC, low grade serous OC, mucinous OC, and low malignant potential OC respectively. Conclusion: Our study adds to current knowledge on the causal effect of known and suspected influencing factors on OC and six histotypes. Further investigation is needed to better understand potential pathways or mechanisms of these factors.