No genetic causal association between periodontitis and ankylosing spondylitis: a bidirectional two-sample mendelian randomization analysis

Abstract

Abstract Background Observational studies that reveal an association between periodontitis (PD) and ankylosing spondylitis (AS) exist. However, observational research is prone to reverse causality and confounding factors, which make it challenging to infer cause-and-effect relationships. We conducted a two-sample Mendelian randomization (MR) study to examine the causal relationship between the genetic prediction of PD and AS. Methods In our study, single-nucleotide polymorphisms (SNPs) were defined as instrumental variables (IVs). The genetic association with PD came from the Gene-Lifestyle Interactions and Dental Endpoints (GLIDE) consortium, wherein 17353 cases of European ancestry and 28210 controls of European ancestry were included in this study. The genetic association with AS from the Neale Laboratory Consortium included 337,159 individuals from the United Kingdom, with 968 cases and 336,191 controls. MR analysis was mainly performed using the inverse-variance weighted (IVW) method. In addition, the robustness of the study findings was assessed using sensitivity, pleiotropy, and heterogeneity analyses. Results Eighteen independent SNPs with P-values significantly smaller than 1 × 10− 5 were used as IV SNPs for PD, while 39 independent SNPs with P-values significantly smaller than 1 × 10− 5 were used as IV SNPs for AS. The results of the IVW method revealed no causal association between PD and AS (odds ratio = 1.00, 95% confidence interval: 0.99953 to 1.00067, P = 0.72). The MR-Egger method did not support the causal association between PD and AS. It is unlikely that horizontal pleiotropy distorts causal estimates based on sensitivity analysis. No significant heterogeneity was observed in the Q test. The ‘’leave-one-out’’ analysis demonstrated that the robustness of our results was unaffected by eliminating any of the IVs. Likewise, no significant causative effect for AS on PD was observed in the inverse MR analysis. Conclusions The study results do not support shared heritability or a causal association between PD and AS.