We thank Roy et al.  for their relevant comments on our article.  As Roy et al. correctly pointed out, head-of-bed elevation by 30 degrees did not significantly improve oxygenation variables such as mean nadir oxygen saturation measured by pulse oximetry (Spo2), lowest Spo2, and percent time Spo2 < 90% in our study although, Souza’s previous study demonstrated improvement of both apnea hypopnea index (15.7 to 10.7 events/h) and lowest Spo2 (83.5 to 87%) in symptomatic obstructive sleep apnea (OSA) patients with only 7.5-degree head-of-bed elevation.  It would be easy to comment that different patient populations and study design are the cause of the difference. However, we consider that Roy’s question has an important pathophysiologic background calling the readers’ attention. First, Souza’s patients are more obese than ours (body mass index: 29.6 ± 4.8 vs. 26.3 ± 4.5 kg/m2), and, therefore, head-of-bed elevation is expected to increase functional residual capacity and improve oxygenation more effectively in such obese patients. Furthermore, head-of-bed elevation improves pharyngeal airway collapsibility as many previous studies have reported. Our research group demonstrated approximately 6 cm H2O improvement of pharyngeal closing pressure by 60-degree sitting position in anesthetized and paralyzed patients with OSA.  Notably, the improvement of pharyngeal closing pressure was indirectly associated with the severity of OSA. Accordingly, apnea hypopnea index is expected to decrease more effectively in Souza’s patients with less severe OSA (15.7 vs. 59.6 events/h), and, in fact, application of only 7.5-degree head-of-bed elevation did achieve successful improvement. Considering both Souza’s and our findings, head-of-bed elevation does provide better nocturnal oxygenation and breathing pattern, but the optimal degree of head-of-bed elevation may depend on the severity of obesity and OSA.

Second, Roy et al. raised optimal oxygen therapy for patients with the overlap syndrome, in which two diseases, chronic obstructive pulmonary disease and OSA, coexist in a single patient. In our study, four participants had both diseases. Two of them did not improve apnea hypopnea index with the combination therapy of head-of-bed elevation and high-flow nasal cannula with 40% oxygen concentration. Currently, the overlap syndrome receives special attention in the fields of pulmonology and sleep medicine because of its greater degrees of nocturnal oxygen desaturation and cardiovascular consequences than those with either condition in isolation.  Although this is not the original scope of our study, the overlap syndrome also needs to be paid more attention by anesthesiologists because of the possible development of severe sustained and episodic hypoxemia after surgery, leading to poor postoperative outcome. To date, we do not know whether continuous or bilevel positive airway pressure therapy is effective and what level of oxygen concentration is appropriate for postoperative respiratory management in patients with the overlap syndrome. Optimal continuous positive pressure for OSA may negate the auto–positive end-expiratory pressure in the overlap syndrome patients with emphysema phenotype. In contrast, bilevel positive pressure ventilation may be advantageous in emphysema-predominant patients because continuous positive pressure may exacerbate mechanical disadvantage of the flattened diaphragm contraction leading to severe hypoventilation during sleep. We would like to close our comments by sharing the most recent clinical study demonstrating the effectiveness of high-flow nasal oxygen therapy (30 to 60 l/min with oxygen concentration titrated to increase awake Spo2 by 3 to 4%) for treatment of the overlap syndrome.  Yes, all issues raised by Roy et al. are clinically relevant and to be fully answered in the near future.