There are four key endotypic traits influencing OSA:

  1. Pharyngeal collapsibility (airway obstruction).
  2. Upper airway muscle compensation (muscle response to maintain airflow).
  3. Arousal threshold (stimulus level required to wake a person).
  4. Ventilatory loop gain (stability of breathing control).

There are three primary approaches for quantifying endotypic traits associated with obstructive sleep apnea (OSA). These methods vary in complexity and applicability for clinical or research settings:

1. Direct Measures from Standard Polysomnography (PSG):

  • Uses routine PSG data to estimate traits indirectly.
  • Examples include:
    • Apnea index (AI) to assess pharyngeal collapsibility.
    • Ratios like REM-AHI to NREM-AHI for ventilatory instability (loop gain).
    • Arousal thresholds derived from oxygen saturation or short respiratory event durations.
  • Advantages: Minimal additional equipment or calculations.
  • Limitations: Lacks precision and excludes mechanistic insights.

2. Physiological Laboratory Measurements:

  • Gold standard methods involve experimental manipulation of CPAP (Younes, 2003, 2004; McGinley et al., 2008; Wellman et al., 2011, 2013; Edwards et al., 2012, 2014b; Eckert et al., 2013; Sands et al., 2014; Messineo et al., 2018) or direct invasive measures:
    • CPAP “drop” techniques to measure pharyngeal collapsibility, arousal threshold, and compensatory responses.
    • Ventilation-vs.-drive models derived from invasive data, such as esophageal pressure or diaphragm EMG.
  • Advantages: Provides detailed mechanistic insights.
  • Limitations: Requires specialized equipment, invasive methods, and expert training, making it impractical for widespread clinical use.

3. Non-invasive Clinical Tools Using PSG Data:

  • Methods like the PUP (Endo-Phenotyping Using Polysomnography) (https://pmc.ncbi.nlm.nih.gov/articles/PMC6019932/pdf/rccm.201707-1435OC.pdf) method analyze ventilation and ventilatory drive without invasive techniques:
    • Estimates traits such as loop gain, collapsibility, and arousal threshold by integrating nasal flow and chemical drive signals.
    • PUPpy, a cloud-based implementation, automates analysis and simplifies accessibility.(https://doi.org/10.1093/sleep/zsaa168)
  • Advantages: Scalable, non-invasive, and applicable in routine clinical settings.
  • Limitations: Requires refinement for higher accuracy and validation against gold-standard methods. To maintain the alignment of the PUPpy method with the original validation of the PUP method, it would be helpful to validate it against gold standard methods (e.g., CPAP drop method/gold standard ventilation and drive signals) to provide an opportunity for ongoing enhancement and development.

While gold-standard laboratory methods provide the most accurate data, non-invasive and scalable tools like PUP and PUPpy are key to integrating endotypic trait analysis into broader clinical practice. Researchers who developed PUP and PUPpy methods should make the code available in public domain for the greater benefit of sleep science.

The repeatability and physiological variability of endotypic traits in obstructive sleep apnea (OSA) needs to be understood well for better clinical use:

Repeatability

  • Within-night repeatability: Moderate to good, with correlation coefficients (Pearson) ranging from 0.69 to 0.83 for measurements taken on the same night.
  • Night-to-night repeatability: Similar results, with intra-class correlation coefficients ranging from:
  • 0.72 to 0.83 in one study.
  • 0.67 to 0.91 in another study.
  • Long-term repeatability: More modest, with correlations between 0.36 and 0.61 observed over 6–7 years in an elderly population.
  • Trait-specific variability:
  • Collapsibility, loop gain, and arousal threshold tend to have better repeatability (>0.8 correlation).
  • Compensation repeatability is lower, likely due to its calculation as a difference between two collapsibility measures, amplifying measurement errors.

Physiological Variability

  • Postural effects: Collapsibility is greater in the supine position compared to lateral positions.
  • Sleep stage effects:
  • Loop gain is lower during REM sleep than NREM sleep.
  • Upper airway muscle compensation is relatively consistent across different states.
  • Incorporating variability:
  • Trait estimates reported are often medians during NREM sleep to account for physiological changes throughout the night.
  • Adjustments for sleep state, position, and other factors could enhance repeatability but require further research.

Implications

The findings suggest that while endotypic traits are moderately repeatable and consistent, further efforts to standardize measurements and account for variability are necessary to improve clinical reliability.