Periodontitis and non-alcoholic fatty liver disease: a Mendelian randomisation analysis

Abstract

Background Previous observational and epidemiological studies have reported the association between periodontitis (PD) and non-alcoholic fatty liver disease (NAFLD). However, evidence from long-term randomized controlled trials (RCTs) is lacking. Therefore, this study aimed to explore the causal relationship between PD and NAFLD. Methods Genetic information for individuals of European ancestry with PD (17,353 clinically diagnosed cases and 28,210 controls) and NAFLD (8,434 clinically diagnosed cases and 770,180 controls) were obtained using published genome-wide association study statistics, following which we conducted two-sample bivariate Mendelian randomisation (MR) analyses. Various techniques such as inverse-variance weighted (IVW), weighted median, and MR-Egger regression methods were used to calculate the causal relationship between exposure and the result. Results No causal effect of genetically determined PD on NAFLD existed (odds ratio [OR] = 0.99, 95% confidence interval [CI]: 0.90–1.10, P = 0.95). Furthermore, no causal effect of NAFLD on PD was observed in the reverse MR analysis (OR = 1.02, 95% CI: 0.92–1.13, P = 0.63). No heterogeneity was observed between individual single nucleotide polymorphisms pursuant to the heterogeneity assessment (Q > 0.1). Horizontal pleiotropy pursuant to the outcomes of MR-Egger regression and MR Pleiotropy RESidual Sum and Outlier overall tests (P > 0.1) were less likely to distort the causal relationship between PD and NAFLD. Conclusion Collectively, we did not find substantial evidence to support a causal association between PD and NAFLD in this bidirectional MR study. Clinical relevance: Periodontitis does not seem to be a risk factor for worsening of non-alcoholic fatty liver disease.