Respiratory-hypoxic burden and multimorbidity-related vulnerability are distinct dimensions of obstructive sleep apnea: A retrospective cohort analysis.

Background

Obstructive sleep apnea (OSA) is a heterogeneous disorder in which conventional apnea-hypopnea index (AHI)-based severity classification incompletely captures respiratory burden, nocturnal hypoxemia, obesity, and multimorbidity. Whether OSA severity is organized along distinct respiratory and multimorbidity-related dimensions remains unclear.

Objectives

To determine whether OSA severity is organized along distinct respiratory-hypoxic and multimorbidity-related dimensions and whether these dimensions define clinically meaningful patient phenotypes.

Methods

We retrospectively analyzed prospectively collected data from 306 untreated adults with OSA. Demographic, anthropometric, respiratory polygraphy, oximetry, daytime sleepiness, and Cumulative Illness Rating Scale (CIRS)-derived measures were assessed. Principal component analysis (PCA) identified latent severity dimensions, followed by hierarchical and k-means clustering to derive clinical phenotypes.

Results

PCA identified two partially dissociated dimensions explaining 57.4% of total variance: a respiratory-hypoxic burden dimension characterized by AHI, nocturnal hypoxemia, and obesity, and an age- and multimorbidity-related vulnerability dimension characterized by age and CIRS-derived measures. Four clinically distinct phenotypes emerged: older-multimorbid (n = 101), hypoxic-metabolic (n = 65), young-preserved (n = 92), and severe obese-hypoxic (n = 48). Multimorbidity-related vulnerability showed only weak correlations with conventional respiratory severity metrics.

Conclusion

OSA severity appears to be organized along complementary respiratory-hypoxic and multimorbidity-related dimensions. This multidimensional framework provides a clinically interpretable approach to understanding patient heterogeneity beyond conventional respiratory severity metrics.