Microbiome and Mycobiome Analyses of Continuous Positive Airway Pressure Devices

Abstract

Objectives: Microorganisms are likely present in continuous positive airway pressure (CPAP) devices in daily use. Given the potential risk of infection among CPAP users, we aimed to compare the microbiomes of CPAP devices with those of nasal mucosa samples obtained from patients using these devices.

Methods: We conducted a prospective cohort study at multiple tertiary medical institutions. Samples were collected from the tubes and filters of CPAP devices and the nasal mucosa of device users. Microbiomes and mycobiomes were analyzed using 16S ribosomal RNA and internal transcribed spacer region sequencing. The results were compared according to sampling site and usage duration for each patient.

Results: Overall, 27 paired samples of human nasal mucosa and CPAP components were analyzed. Bacteria were detected in 7 of the 27 tubes (25.9%) and in 22 of the 27 filters (81.5%). Fungi were found in 2 tubes (7.4%) and 16 filters (59.3%). The most prevalent bacterial phyla across all samples were Actinobacteria and Firmicutes. Fungi were not detected in any nasal mucosa samples. However, fungi were identified in the CPAP filters and tubes, with the Basidiomycota and Ascomycota phyla predominating. No significant associations were identified according to sampling site or duration of CPAP use.

Conclusion: Some bacteria or fungi are detectable in CPAP samples, even after a short period of CPAP usage. However, the association between respiratory infections and these microbiomes or mycobiomes was not investigated. Further research is required to clarify the risk posed by CPAP devices as a microbial contamination source.

Keywords: Continuous Positive Airway Pressure Device; Microbiome; Mycobiome; Obstructive Sleep Apnea.

Fig. 1.

Bar graph of taxa, illustrating the bacterial composition by sampling site. The graph is divided into continuous positive airway pressure (CPAP) filter, CPAP tube, and nasal mucosa samples. (A) At the phylum level, the graph displays the 11 most abundant bacteria. (B) At the genus level, the graph presents the 10 most abundant bacteria.

Fig. 2.

Bar graph of taxa, illustrating the fungal composition by sampling site. The graph is divided into continuous positive airway pressure (CPAP) filter, CPAP tube, and nasal mucosa samples. (A) At the phylum level, the graph displays the 7 most abundant fungi. (B) At the species level, the graph presents the 10 most abundant fungi.

Fig. 3.

Comparison of bacterial alpha and beta diversity levels according to continuous positive airway pressure (CPAP) sampling site. (A-C) Alpha diversity: (A) Chao1 index, (B) Shannon index, and (C) phylogenetic diversity (^**P<0.01 and ^***P<0.001). (D, E) Principal coordinate analysis plots according to CPAP sampling site, based on weighted and unweighted UniFrac distances. (D) Weighted UniFrac distance. Significant differences were observed among the groups (P=0.001; analysis of similarities [ANOSIM]). (E) Unweighted UniFrac distance. Significant differences were also noted among the groups (P=0.001; ANOSIM).

Fig. 4.

Comparison of bacterial alpha and beta diversity levels according to filter usage duration and correlations between duration and microbial taxa (long, >3 months; short, ≤3 months). (A-C) Alpha diversity: (A) Chao1, (B) Shannon, and (C) phylogenetic diversity. (D, E) Principal coordinate analysis plots according to filter usage duration, based on weighted and unweighted UniFrac distances. (D) Weighted UniFrac distance. (E) Unweighted UniFrac distance. No significant difference was observed in either distance metric (P=0.217, P=0.612; analysis of similarities). (F) Amplicon sequence variants (ASVs) significantly correlated with filter usage duration, as determined using MaAsLin2 (false discovery rate [FDR] <0.05).

Fig. 5.

Comparison of bacterial alpha diversity levels according to tube usage duration and correlations between tube usage duration and microbial taxa (long, >3 months; short, ≤3 months). (A-C) Alpha diversity: (A) Chao1, (B) Shannon, and (C) phylogenetic diversity. (D) Associations between tube usage duration and microbiome composition data, as determined using MaAsLin2. FDR, false discovery rate.

Fig. 6.

Comparison of fungal alpha and beta diversity levels according to filter usage duration (long, >3 months; short, ≤3 months). (A-C) Alpha diversity: (A) Chao1, (B) Shannon, and (C) phylogenetic diversity. (D, E) Principal coordinate analysis plots according to filter usage duration, based on weighted and unweighted UniFrac distances. (D) Weighted UniFrac distance. (E) Unweighted UniFrac distance. No significant difference was observed in either distance metric (P=0.134, P=0.206; analysis of similarities).