P008 Ventilatory burden predicts change in self‑reported sleepiness following continuous positive airway pressure treatment in obstructive sleep apnoea

Abstract Introduction Excessive daytime sleepiness is an important symptom of obstructive sleep apnoea (OSA) and commonly quantified using the Epworth Sleepiness Scale (ESS). Measures of ventilatory burden, flow limitation, and hypoxaemia could be used to predict sleepiness change (i.e. ΔESS) following continuous positive airway pressure (CPAP) treatment. We hypothesised that ΔESS can be predicted from baseline polysomnography. Methods Associations between OSA severity measures and ESS were evaluated in three cross‑sectional cohorts (n=3117), adjusting for age, sex, BMI, total sleep time, and depression. ΔESS prediction was evaluated using CPAP treatment studies (HomePAP, BestAIR, and ABC) in three steps: severity measures were compared (adjusted regression predicting ΔESS, n=64), a prediction model was developed using baseline ventilatory burden and baseline ESS (n=139), then evaluated in holdout participants (n=74). Results In cross‑sectional analyses, increased ESS was associated with increased ventilatory burden (0.27 points/SD; 95% CI 0.07−0.48), the apnoea‑hypopnoea index (AHI) (0.25; 0.05−0.45), flow limitation severity (0.24; 0.05−0.43), and hypoxic burden (0.21; 0.00−0.42). Comparison analysis revealed that ΔESS was most strongly associated with baseline ventilatory burden (‑1.07 points/SD; ‑2.12 to ‑0.05) and baseline ESS (‑2.75; ‑3.83 to ‑1.69); the AHI association was weaker (‑0.97; ‑2.01−0.05). Predicted and actual ΔESS were associated in holdout participants (adjusted R²=0.315); median [IQR] actual ΔESS of predicted responders (≥2 ESS reduction) was ‑5.0 [‑10.0 to ‑2.0] and non‑responders was 0.0 [‑1.0−1.0] (p<0.001). Discussion Baseline ventilatory burden and baseline ESS were independently associated with ΔESS and could be used together to inform clinicians whether CPAP treatment will likely improve a patient’s sleepiness.