The oral microbiome and all-cause mortality in the US population.

Abstract

Importance: Poor oral health, including periodontal disease, is associated with oral microbiome changes and increased mortality risk. However, no large studies have evaluated whether the oral microbiome is directly associated with mortality.

Objective: To evaluate whether measures of the oral microbiome is prospectively associated with all-cause mortality.

Design: A cross-sectional survey with samples collected from 2009-2012 and mortality linkage to the restricted-use National Death Index (NDI) through 2019.

Setting: The National Health and Nutrition Examination Survey (NHANES) 2009-2012, a multistage probability sample of the US population.

Participants: NHANES participants 20- to 69-years-old who were eligible for linkage to the NDI and provided oral rinse specimens (N=7,721, representing approximately 194 million individuals).

Exposure: Oral microbiome ascertained by sequencing the V4 region of the 16S rRNA gene of extracted DNA from oral rinse specimens. Alpha diversity, beta diversity, and genus-level data were generated using DADA2 and QIIME.

Main outcome and measure: All-cause mortality.

Results: After an average of 8.8 years, a total of 426 participants died. Using Cox proportional hazards regression and after controlling for multiple comparisons where appropriate, continuous alpha diversity was inversely associated with all-cause mortality, but only the association for the Shannon-Weiner index was significant with full adjustment for major risk factors (hazard ratio [HR] per standard deviation [SD]=0.85; 95% confidence interval [CI]=0.74-0.98). The principal coordinate analysis (PCoA) vector 2 from the Bray-Curtis dissimilarity matrix (HR per SD=0.83; 95% CI=0.73-0.93) and PCoA1 from weighted UniFrac (HR per SD=0.86; 95% CI=0.75-0.98) were significantly associated with all-cause mortality after full adjustment. Few associations were observed at the genus-level after Bonferroni correction, but an increase in 1 SD of the relative abundance of Granulicatella and Lactobacillus were associated with a 17% (95% CI=1.11-1.24) and 11% (95% CI=1.06-1.16) increase in mortality risk, respectively. Compared to participants with no detectable Bacteroides, participants in the highest tertile of Bacteroides had decreased mortality risk (HR=0.54; 95% CI=0.40-0.74).

Conclusions and relevance: Some measures of the oral microbiome were associated with all-cause mortality in this representative population cohort. These results suggest that oral bacterial communities may be important contributors to health and disease.