Sleep architecture disruption and coordinated patterns of inflammatory and extracellular matrix biomarkers in obstructive sleep apnea: a cross-sectional study

This study aimed to investigate the associations between circulating inflammatory, extracellular matrix (ECM), and oxidative stress biomarkers and polysomnographic parameters across different severity levels of obstructive sleep apnea (OSA), with adjustment for body mass index (BMI).

Seventy adults with confirmed OSA were categorized as mild-to-moderate or severe based on apnea-hypopnea index (AHI). Serum levels of TNF-α, IL-6, MMP-2, MMP-9, desmosine, cellular fibronectin (cFN), syndecan-1 (SDC-1), glutathione (GSH), and malondialdehyde (MDA) were measured using ELISA or spectrophotometric methods. Biomarker concentrations were natural-log transformed to satisfy normality assumptions; between-group comparisons were adjusted for BMI using ANCOVA. Associations among biomarkers and clinical variables were assessed using Spearman correlation, BMI-adjusted partial correlation, and network analysis.

Severe OSA was associated with higher levels of TNF-α, IL-6, MMP-2, MMP-9, desmosine, and cFN compared with mild-to-moderate OSA (all p < 0.05). SDC-1 reached statistical significance after BMI adjustment (p = 0.009). These biomarkers were positively associated with AHI, oxygen desaturation index (ODI), desaturation count, and time spent with SpO₂ <90% (T90), and inversely associated with N3 sleep duration. After adjustment for BMI, strong inter-biomarker partial correlations (r = 0.831-0.921, p < 0.001) formed a densely connected inflammatory-ECM cluster; this pattern remained significant, albeit more modest in magnitude, in sensitivity analyses using unadjusted Spearman correlations (ρ = 0.427-0.655, p < 0.001). GSH and MDA showed weaker and more isolated associations.

OSA severity is associated with coordinated patterns among inflammatory and ECM-related biomarkers that parallel hypoxic burden and sleep architecture changes. In contrast, systemic oxidative stress markers appear less integrated within these patterns. These findings support the potential utility of multi-biomarker approaches in reflecting disease-related physiological alterations, potentially arising from both systemic and local upper airway tissue processes.