Is there an association between periodontitis and breast cancer?

Abstract

Design

This study uses bidirectional, two-sample Mendelian randomization (MR) analysis through genome-wide association studies (GWAS) statistics to assess the causal association between periodontitis and breast cancer. Periodontitis data from the FinnGen project and breast cancer data from OpenGWAS was used, with all data subjects being of European ancestry. Periodontitis cases were categorized by probing depths or self-reporting according to the Centers for Disease Control and Prevention (CDC)/American Academy of Periodontology definition.

Case selection

3046 periodontitis cases and 195,395 control cases along with 76,192 cases of breast cancer and 63,082 control cases were obtained from GWAS data.

Data analysis

R (version 4.2.1) through TwoSampleMR and MRPRESSO were used for data analysis. Primary analysis was conducted using the inverse-variance weighted method. Horizontal pleiotrophy was corrected and causal effects were examined through detection methods of the weighted median method, weighted mode method, simple mode, MR-Egger regression method, and MR-pleiosis residual and outlier method (MR-PRESSO). A heterogeneity test was adopted for the inverse-variance weighted (IVW) analysis method and MR-Egger regression (P > 0.05). Pleiotropy was evaluated using the MR-Egger intercept value. The P-value of the pleiotropy test was then used to analyze the existence of pleiotropy. Where P > 0.05, the possibility of pleiotropy in the causal analysis was considered to be less or non-existent. Leave-one-out analysis was used to tested the consistency of results.

Results

171 single nucleotide polymorphisms were extracted for MR analysis where breast cancer was the exposure and periodontitis the outcome. The total sample size of periodontitis was 198,441, and 139,274 for breast cancer. Overall results showed breast cancer had no effect on periodontitis (IVW P = 0.1408, MR-egger P = 0.1785, weighted median P = 0.1885), with Cochran’s Q analysis showing no heterogenicity amongst these instrumental variables (P > 0.05). 7 single nucleotide polymorphisms were extracted for MR analysis where periodontitis was the exposure and breast cancer the outcome. No significant correlation was found between periodontitis and breast cancer (IVW P = 0.8251, MR-egger P = 0.6072, weighted median P = 0.6848).

Conclusions

Based on the use of different MR analysis methods, there is no evidence to support a causal relationship between periodontitis and breast cancer.