Study objectives: Obstructive sleep apnea hypopnea syndrome (OSAHS) poses a significant health hazard, as intermittent hypoxia inflicts damage throughout the body and is considered a critical risk factor for metabolic disorders. The aim of this study was to establish a metabolic profile for patients with OSAHS using nontargeted metabolomics detection techniques, providing a basis for OSAHS diagnosis and novel biological marker identification.
Methods: Forty-five patients with OSAHS composed the OSAHS group, and 44 healthy volunteers composed the control group. Nontargeted metabolomics technology was used to analyze participants' urinary metabolites. Differentially abundant metabolites were screened and correlated through hierarchical clustering analysis. We constructed a composite metabolite diagnostic model using a random forest model. Simultaneously, we analyzed the relationships between 20 metabolites involved in model construction and OSAHS severity.
Results: The urinary metabolomics pattern of the OSAHS group exhibited significant changes, demonstrating noticeable differences in metabolic products. Urinary metabolite analysis revealed differences between the mild-moderate OSAHS and severe OSAHS groups. The composite metabolite model constructed in this study demonstrated excellent diagnostic performance not only in distinguishing healthy control participants from patients with mild-moderate OSAHS (AUC = 0.78) and patients with severe OSAHS (AUC = 0.78), but also in discriminating between patients with mild-moderate and severe OSAHS (AUC = 0.71).
Conclusions: This study comprehensively analyzed the urinary metabolomic characteristics of patients with OSAHS. The established composite metabolite model provides robust support for OSAHS diagnosis and severity assessment. Twenty metabolites associated with OSAHS disease severity offer a new perspective for diagnosis.
Keywords: UHPLC–MS/MS; metabolomics; obstructive sleep apnea hypopnea syndrome; random forest model.
© 2024 American Academy of Sleep Medicine.