Coffee Consumption and Periodontitis: a Mendelian Randomization Study

Abstract

Abstract Background: Coffee is one of the most consumed beverages in the world, coffee consumption has been growing in the United States over the past 20 years. Periodontitis is defined by the pathologic loss of the periodontal ligament and destruction of the connective tissue attachment and alveolar bone loss and is related to different systemic diseases and conditions. However, the causality has remained unclarified, thus we regarded discovering the causal relationship between coffee consumption and the infection risk of periodontitis as the objective of the study. Methods: Coffee consumption was subdivided into binary coffee consumption and continuous coffee consumption to refine the study design. Genetic instruments were stretched from the MRC-IEU (MRC Integrative Epidemiology Unit)analysis on the UK Biobank, the Gene-Lifestyle Interactions in Dental Endpoints (GLIDE) project, and the joint meta-analysis of a recent GWAS. The IVW (Inverse Variance Weighted) method, MR-Egger regression, and Weighted Median method were introduced to estimate the causality, a scatter plot revealed the intuitive result, and a Leave-One-Out plot was used to test the stability. Heterogeneity and pleiotropy analyses were also conducted to test outlier SNPs (Single Nucleotide Polymorphisms). Results: An effect of continuous coffee consumption on the risk of periodontitis was found (IVW: OR = 1.0104, 95%CI: 1.0016-1.0193, P = 0.0212; MR-Egger: OR = 1.0004, 95%CI: 0.9753-1.0303, P = 0.98; Weighted Median: OR = 1.0077, 95%CI: 0.9963-1.0191, P = 0.19), while the effect of binary coffee consumption on periodontitis did not meet the requirement of indicating a strong causal association, neither were the reverse causality analyses. Conclusions: The study indicated the causality of continuous coffee consumption to the risk of periodontitis with no strong evidence for an effect of binary coffee-consuming behavior on periodontitis. There was also no intensive evidence suggesting reverse causality.