Associations between oral microbiome diversity and rheumatoid arthritis in U.S. adults: NHANES 2009–2012

OBJECTIVE: Although links exist between periodontitis and rheumatoid arthritis (RA), and the gut microbiome has been implicated in RA pathogenesis, the role of oral microbiome diversity in RA remains insufficiently characterized. This study aimed to explore the association between oral microbiome diversity (including alpha and beta diversity) and RA status through a cross-sectional analysis of the National Health and Nutrition Examination Survey (NHANES). MATERIAL AND METHODS: This cross-sectional study analyzed data from 1,544 participants aged ≥ 20 years derived from the 2009-2012 NHANES cycles. We employed multivariate logistic regression, restricted cubic splines (RCS), receiver operating characteristic (ROC) curve analysis, SHapley Additive exPlanations (SHAP) method and beta diversity assessment (Principal Coordinate Analysis [PCoA] and Permutational Multivariate Analysis of Variance [PERMANOVA]) to examine associations between oral microbiome diversity metrics and RA. RESULTS: After adjustments, four alpha diversity metrics (observed amplicon sequence variants [ASVs]: OR [95% CI] = 0.996 [0.992, 0.999], P = 0.043; Faith's PD: OR [95% CI] = 0.943 [0.900, 0.988], P = 0.014; Shannon-Wiener index: OR [95% CI] = 0.786 [0.641, 0.965], P = 0.021; Simpson index: OR [95% CI] = 0.127 [0.018, 0.915], P = 0.037) were significantly inversely associated with the presence of RA. This relationship was approximately linear (P for nonlinear > 0.05) and moderated by socioeconomic factors (P for interaction < 0.05). ROC and SHAP analyses revealed that the Simpson index had the highest explanatory capacity for RA. However, beta diversity (Bray-Curtis, UniFrac distances) revealed no significant differences between RA and non-RA groups (all P > 0.05). CONCLUSIONS: Higher oral microbiome alpha diversity is significantly associated with lower prevalence of RA. Oral microbial diversity may serve as a potential indicator associated with RA status. However, given the cross-sectional nature of this study, longitudinal and interventional studies are warranted to further elucidate causal relationships.