Abstract
BACKGROUND: Smoking increases mortality risk and alters the oral microbiome, but its mediating role in the smoking-survival relationship remains unclear. This study examined whether oral microbiome diversity mediates the association between smoking and all-cause mortality. METHODS: We included 8,223 participants from the National Health and Nutrition Examination Survey with linked mortality data through 2019. Oral microbiome diversity was assessed using alpha and beta diversity metrics. Associations between smoking, diversity, and mortality were assessed using Weibull Accelerated Failure Time models. Multivariable linear regression evaluated the relationship between smoking and oral microbiome diversity. Mediation analysis estimated the Natural Direct Effect (NDE) and Natural Indirect Effect (NIE). Sensitivity analyses assessed effect heterogeneity. RESULTS: Among participants, 429 were deceased. Current smoking was associated with a 42.3% shorter survival time (TR = 0.577). Greater ln-transformed observed Operational Taxonomic Units (OTU) richness was associated with 33.2% longer survival time (TR = 1.332). Smoking was associated with survival time through NIE = 1.013 (95% CI: 1.003, 1.033) and NDE = 0.577 (95% CI: 0.474, 0.697). Sensitivity analyses supported the findings. DISCUSSION: Oral microbiome diversity partially mediated the association between smoking and mortality. Although smoking shortened survival, its effect on increasing OTU richness modestly suppressed this risk. These results highlight a complex microbial pathway and support further investigation into species-level mechanisms and potential microbiome-targeted interventions.