Causal Effects of Oral Microbiome, Circulating Metabolites, and Inflammatory Proteins on Recurrent Aphthous Stomatitis: A Two-Sample Mendelian Randomization Study
Abstract
Purpose Recurrent aphthous stomatitis (RAS) affects 20% of the global population with unclear etiology and limited treatment options. While observational studies suggest associations with oral microbiome dysbiosis and metabolic alterations, causal relationships have not been established. Using Mendelian randomization, the aim of this study was threefold: 1. to systematically investigate the causal architecture underlying RAS; 2. to partition the total effect of identified microbial taxa into direct and mediator-adjusted components; 3. to determine whether specific circulating metabolites and circulating inflammatory proteins mediate the microbiome–RAS relationship. Materials and Methods We conducted two-sample Mendelian randomization (MR) using oral microbiome, metabolomics, inflammatory proteins, and RAS genome-wide association study summary statistics (GWAS). Univariable MR assessed direct causal effects, while multivariable MR evaluated mediation pathways. Sensitivity analyses included MR-Egger, weighted median, and MR-PRESSO. Results The order Clostridiales demonstrated protective effects against RAS (OR=0.851, 95% CI: 0.759-0.954, p=0.006). Among 36 causally associated metabolites, myo-inositol (OR=1.619), sphingosine 1-phosphate (OR=1.672), and hexadecadienoate (OR=1.825) increased RAS risk, while 4-acetylphenol sulfate (OR=0.606) and acetylcarnitine (OR=0.604) conferred protection. No inflammatory proteins showed statistically significant associations. Two-step mediation analysis identified sphingomyelin (d18:1/22:1, d18:2/22:0, d16:1/24:1) as a metabolic intermediate in the Clostridiales–RAS pathway, with the direct protective effect of Clostridiales remaining statistically significant after adjustment for sphingomyelin levels. Conclusions This study provides genetic evidence supporting causal relationships linking the oral microbiome and metabolites to RAS, and identified Clostridiales as protective against RAS, but found multiple lipid mediators as risk factors. The interplay between Clostridiales and sphingolipid metabolism provides candidate therapeutic targets for RAS management through microbiome modulation or metabolic intervention.