P088 Presence versus absence of flow limitation during stable breathing in patients with obstructive sleep apnoea

INTRODUCTION: Flow limitation is the distinguishing characteristic of obstructive sleep apnoea. Critically, periods of flow limitation can occur without overt reductions in airflow (e.g. disproportionate increase in ventilatory drive vs. achieved ventilation), however, such periods are ignored by clinical scoring. Here we investigate flow limitation during so-called “stable breathing”, i.e. periods of sleep without scored events, by applying our recently-validated model to estimate flow limitation from the airflow signal. METHODS: Flow limitation was visually-scored (N=117,871 breaths) from N=40 participants attending an overnight sleep study for suspected sleep apnoea. Scoring was aided by physiological signals (e.g. intra-oesophageal diaphragm EMG). Model flow limitation classification used features extracted from the pneumotach signal (cross-validated accuracy=92.4%). We applied this method to investigate the occurrence of flow limitation during stable breathing, defined as periods of sleep >3 min duration without scored arousals or respiratory events. RESULTS: Model predicted flow limitation frequency was strongly correlated with visual scoring (R²=0.84 p<0.001). The median flow limitation frequency during stable breathing ranged from 8–91%, with an overall median of 59% (IQR 37%-75%). Flow limitation frequency during stable breathing was only modestly associated with the apnoea-hypopnea index (R²=0.12 p<0.05). DISCUSSION: Flow limitation occurs surprisingly frequently during stable breathing. While some individuals achieve stable breathing with minimal flow limitation, others demonstrate substantial flow limitation. Heterogeneity in frequency of flow limitation (within and between individuals) may provide further insights into emergent phenotypic variability within sleep disordered breathing. Finally, this model performed similarly in nasal pressure (88.2% accuracy), indicating potential application to clinical studies.