Salivary alpha-amylase as a biomarker for obstructive sleep apnea severity and continuous positive airway pressure adherence: A cluster-based phenotyping approach

Background Obstructive sleep apnea (OSA) is characterized by repetitive apneic events leading to sympathetic nervous system activation. Salivary alpha-amylase (sAA) is a non-invasive biomarker reflecting sympathetic activity; however, its clinical relevance in OSA remains underexplored. Methods This prospective, observational study included 177 adults with suspected OSA who underwent overnight polysomnography and pre-sleep sAA activity measurement. Nasal obstruction was assessed subjectively and objectively. Cluster analysis was performed using sAA activity, the apnea-hypopnea index (AHI), sleep efficiency (SE), and nasal parameters in patients requiring continuous positive airway pressure (CPAP; n = 104). CPAP adherence was evaluated at 3 and 6 months. Results Higher sAA activity was observed in patients with severe OSA (AHI ≥ 30). Elevated sAA activity was also associated with lower SE and greater subjective nasal obstruction. In the multivariate logistic regression analysis, sAA activity remained significantly associated with severe OSA. Cluster analysis identified three phenotypes: (1) low sAA activity with preserved sleep; (2) high sAA activity with nasal obstruction; and (3) high sAA activity with severe OSA, high body mass index, and low SE. Clusters 2 and 3 showed lower CPAP adherence at 3 and/or 6 months than Cluster 1. Conclusion sAA activity was associated with OSA severity, reduced SE, and greater subjective nasal obstruction. Phenotypes incorporating sAA activity and these clinical characteristics showed different patterns of CPAP adherence. These findings suggest that phenotyping incorporating sAA activity may help identify patients at risk of poor CPAP adherence and inform treatment strategies.