Abstract
OBJECTIVES: Reports suggest an association between periodontitis and oral cancer. Therefore, this study used a Mendelian randomisation analysis to investigate whether a causal relationship exists between periodontitis and oral cancer and whether periodontitis is a reliable early indicator of oral cancer. METHODS: Publicly available genome-wide association study data of a European population were used to perform a two-sample, two-way Mendelian randomisation (MR) analysis primarily via inverse variance weighting (IVW). Complementary methods were used to detect and correct the effects of horizontal pleiotropy. RESULTS: Acute periodontitis (IVW [odds ratio (OR) = 0.999, 95% confidence interval (CI) = 0.999-1.000, P = 0.972]; MR-Egger [OR = 1.000, 95% CI = 0.999-1.000, P = 0.843]; and weighted median [OR = 1.000, 95%CI = 0.999-1.000, P = 0.947]) and chronic periodontitis (IVW [OR = 0.999, 95% CI = 0.999 -1.000, P = 0.725; MR-Egger [OR = 1.000, 95% CI = 0.998-1.000, P = 0.245); and weighted median [OR = 1.000, 95% CI = 0.999-1.000, P = 0.834]) did not affect oral cancer. CONCLUSIONS: Our MR analysis did not support a causal relationship between periodontitis and oral cancer. Preventing and treating acute or chronic periodontitis may not reduce the oral cancer risk, and the advantages of screening patients for periodontitis for early cancer detection are limited, which may provide assistance in reducing clinical inputs. This study bears some limitations including inapplicable method if the relationship is non-linear. Analyzing statistical data in a database stratified by sex or age is difficult, potentially leading to biased results.