Human oral microbiota dysbiosis may be a novel non-invasive indicator of erectile dysfunction.

Background/purpose: Dysbiosis of the oral microbiota has been associated with various human diseases, however, the relationship between the oral microbiota and erectile dysfunction (ED) remains unexplored. This study aimed to investigate the connection between the oral microbiota and ED, ultimately identifying a novel non-invasive indicator for ED. Materials and methods: We recruited 74 male patients diagnosed with erectile dysfunction and 38 men with normal sexual function to serve as the control group. We compared the overall conditions and clinical data between the two groups. Oral samples were collected from the subjects, followed by 16S rRNA sequencing. We assessed the differences in oral microbiota between the two groups. We analyzed the correlation between oral microbiota and clinical indicators, ultimately constructing ten machine learning prediction models based on the oral microbiota data. Results: There were distinct differences in the oral microbiota at both the phylum and genus levels between the two groups. Furthermore, significant differences in both α-diversity and β-diversity were observed between the two groups. The different oral microbiota between the groups were correlated with certain clinical indicators. We constructed ten machine learning prediction models, among which the Light Gradient Boosting Machine (LGBM) model demonstrated superior performance, achieving a relatively higher area under the curve (0.711), accuracy (0.733), sensitivity (0.784), specificity (0.636) and F1-score (0.795). Conclusion: An dysbiosis in oral microbiota may have a potential connection to ED. This dysbiosis could serve as an indicator for assessing the risks associated with the occurrence and progression of ED.