Altered Salivary Microbiota in Patients with Obstructive Sleep Apnea Comorbid Hypertension

Abstract

Purpose: Microorganisms contribute to the pathogenesis of obstructive sleep apnea (OSA)-associated hypertension (HTN), while more studies focus on intestinal microbiome. However, the relationship between oral microbiota and OSA-associated HTN has yet to be elucidated. This study aimed to identify differences in salivary microbiota between patients with OSA comorbid HTN compared with OSA patients, and furthermore evaluate the relationship between oral microbiome changes and increased blood pressure in patients with OSA.

Patients and methods: This study collected salivary samples from 103 participants, including 27 healthy controls, 27 patients with OSA, 23 patients with HTN, and 26 patients with OSA comorbid HTN, to explore alterations of the oral microbiome using 16S rRNA gene V3-V4 high-throughput sequencing. And ultra-high-performance liquid chromatography was used for metabolomic analysis.

Results: Alpha- and beta-diversity analyses revealed a substantial difference in community structure and diversity in patients with OSA comorbid HTN compared with patients with OSA or HTN. The relative abundance of the genus Actinomyces was significantly decreased in patients with HTN compared with healthy controls, and those with OSA concomitant HTN compared with the patients in OSA, but was not significantly different between patients with OSA and healthy controls. Linear discriminant analysis effect size and variance analysis also indicated that the genera Haemophilus, Neisseria, and Lautropia were enriched in HTN. In addition, Oribacterium was an unique taxa in the OSA comorbid HTN group compared with the control group. Metabolomic analysis of saliva identified compounds associated with cardiovascular disease in patients with OSA comorbid HTN.2-hydroxyadenine, was significantly increased in the group of patients with OSA compared with controls, and L-carnitine was significantly decreased in patients with OSA comorbid HTN compared with OSA patients.

Conclusion: This study highlighted noninvasive biomarkers for patients with OSA comorbid HTN. As the first study to find alterations of the salivary microbiome in patients with OSA comorbid HTN, it may provide a theoretical foundation for clinical diagnosis and treatment of this condition.

Keywords: 16S rRNA; OSA; hypertension; metabolomics; oral microbiome.

Figure 1

The Alpha and Beta diversity. The Alpha diversity was estimated by Chao1 index (A), Shannon index (B) and Richness index (C). Both Chao1 index and Richness index was significantly decreased in HTN group (p<0.001) compared with control group. Compared with complication group (OSA+HTN), Chao1 index and Richness index was significantly decreased in HTN group (p=0.014, p=0.0087, respectively). Alpha diversity by Shannon index did not show a significant difference among these 4 groups (p=0.43). Beta diversity was calculated using bray_curties distance by PCoA to analyze the variation of salivary microbial community structure (D and E). *p<0.05, **p<0.01, ***p<0.001.

Figure 2

Phylogenetic profile of salivary microbes among patients with OSA (n=27), HTN (n=23), OSA with HTN (n=26) and healthy controls (n=27). Composition of Salivary microbiota at the phylum level (A) and genus level (B). The heatmap shows relative abundance of microbiota at the genus level in each group (C).

Figure 3

Comparison of the relative abundance of the top 20 genera with changes in relative abundance. *p<0.05, **p<0.01, ***p<0.001.

Figure 4

Predicted salivary microbiome taxa by linear discriminant analysis effect size (LEfSe), of OSA group (n=27), HTN group (n=23), OSA with HTN group (n=26) and healthy controls (n=27). (A) Histogram of taxonomic. (B) Cladogram of taxonomic. Each dot represents a taxonomic hierarchy, from inner to outer circles represented taxa from phylum to genus level. Linear discriminant analysis (LDA) score>2 indicates significant bacterial differences among four groups. Prefix p_phylum, c_class, o_order, f_family and g_genus.

Figure 5

Receiver operating characteristic (ROC) for ASVs-based markers identified by random forest models. (A) Prediction of the key genera for all-OSA (OSA+complication) group from control group, area under the parametric curve (AUC)=91.9% (95% CI, 85.35–98.51%). (B) AUC value achieved 94.7% (95% CI, 86.57–100%) between HTN group and complication (OSA+HTN) group. (C) AUC value achieved 78.5% (95% CI, 65.85–91.13%) between OSA group and complication (OSA+HTN) group.

Figure 6

Score plots of the orthogonal partial least-squares discriminant analysis (OPLS-DA) model for saliva samples. (A) OPLS-DA model for OSA and control group, the model parameters were: R²X=0.344 R²Y=0.949 Q²=0.352. (B) OPLS-DA model for complication (OSA+HTN) and control group, the model parameters were R²X=0.619 R²Y=0.998 Q²=0.772. (C) OPLS-DA model for complication (OSA+ HTN) and HTN group, the model parameters were: R²X=0.609 R²Y=0.993 Q²=0.587. (D) OPLS-DA model for complication (OSA+HTN) and OSA group, the model parameters were: R²X=0.378 R²Y=0.881 Q²=−0.0792.