REM-related obstructive sleep apnea: low AHI-high hypoxemia paradox

Area of Science:

• Sleep Medicine
• Respiratory Physiology
• Cardiovascular Health

Background:

• Obstructive sleep apnea (OSA) is often classified by the apnea-hypopnea index (AHI).
• REM-related OSA presents a diagnostic challenge, as low AHI may mask underlying disease severity.
• Understanding different OSA phenotypes is crucial for accurate risk stratification.

Purpose of the Study:

• To compare polysomnography findings, comorbidities, and oxygenation metrics across OSA phenotypes.
• To determine if AHI-based severity classifications underestimate nocturnal hypoxemia in REM-related OSA.
• To investigate the utility of T90 and ODI in assessing physiological strain in REM-related OSA.

Main Methods:

• Retrospective analysis of 378 OSA patients categorized into REM-related, REM-aggravated, and stage-independent phenotypes.
• Analysis of sleep architecture, AHI, oxygen desaturation index (ODI), and T90 (% sleep time with SpO₂<90%).
• Multivariable linear regression to assess phenotype's independent contribution to nocturnal hypoxemia.

Main Results:

• REM-related OSA patients had the lowest total AHI but similar comorbidities to stage-independent OSA.
• Nocturnal SpO₂ levels were comparable across phenotypes.
• T90 was significantly lower in REM-aggravated and stage-independent groups compared to REM-related OSA, indicating greater cumulative hypoxemia.
• ODI was primarily influenced by body weight and total AHI, not phenotype.

Conclusions:

• REM-related OSA is a distinct phenotype where low AHI underestimates risk and obscures comorbidities.
• Time-based metrics (T90) are superior to event-based indices (ODI) for assessing physiological strain in REM-related OSA.
• A phenotype-specific diagnostic approach is needed to improve risk stratification for OSA.