Reply: Arousal Threshold in Obstructive Sleep Apnea

From the Authors:

We thank Professor Marcus for her interest in our study (1), and for highlighting these key arousal threshold studies in children with and without obstructive sleep apnea (OSA) (2). Our paper focused on adults. The pathophysiology of OSA in children may be quite different. Nonetheless, comparing potential differences and similarities in the varying causes of OSA between adults and children is of interest. In considering the role of arousal in sleep-disordered breathing pathogenesis across the lifespan, it is noteworthy that termination of obstructive respiratory events are rarely associated with cortical arousal in infants (<10%) (3), occur occasionally in children (<50%) (3, 4), and are present more frequently in adults (∼80%) (5). These divergences may reflect differences in arousal mechanisms, neuromuscular responses, or a combination of both. The timing of EEG arousal also often does not precisely coincide with airway opening in OSA (5, 6). Thus, EEG arousals are not required for airway opening in OSA and the upper airway muscles are capable of restoring airflow via noncortical arousal mechanisms, especially in infants.

On average, our data indicate that respiratory arousal thresholds are marginally higher (harder to wake up) in adult patients with OSA than in healthy control subjects, even after at least 3 months of continuous positive airway pressure (CPAP) therapy for more than 4 hours per night (1). However, most patients with OSA have arousal thresholds that are within the normal range after therapy (1, 6). In those who have high arousal thresholds despite CPAP treatment (∼25% of the patoents with OSA we studied), OSA is typically very severe (6, 7). Increased arousal thresholds in these patients may be inherent or slowly/incompletely reversible. However, caution is warranted, as current treatments for OSA in both children and adults are often incompletely effective in mitigating sleep-disordered breathing and its consequences, particularly when CPAP compliance is considered. Elevated arousal thresholds in OSA may also be protective via preservation of sleep and facilitation of upper airway muscle activity (8–10). Thus, it is unclear if elevated arousal thresholds in some patients with OSA are a consequence/adaptive response or a contributor to the disorder. This may differ between children and adults. A combined pediatric and adult study using standardized techniques to characterize respiratory and nonrespiratory arousal, ideally longitudinally, would likely be insightful in determining the potential role of high arousal thresholds in OSA pathogenesis.

In contrast, CPAP does not appear to decrease the arousal threshold in patients with OSA with low arousal thresholds, who likely represent at least 30% of the current adult OSA population (6). Thus, while there is a group of patients with OSA in whom arousal thresholds are high, others may have inherently low arousal thresholds.

From a phenotyping perspective, it is essential to consider inter-individual variability in arousal thresholds in OSA. The level of hypercapnia associated with cortical arousal in children with OSA varies widely (from ∼53 to ∼65 mm Hg [7]), as does the negative esophageal pressure to obstructive respiratory events (approximately −8 to −65 cm H₂O [4]). Similarly, while the existing adult arousal threshold literature primarily includes patients with severe OSA, the degree of negative epiglottic or esophageal pressure required to elicit cortical arousal to respiratory stimuli ranges from −8 to −147 cm H₂O (6). The pathophysiological contribution and consequences of arousal for a patient with OSA who wakes up repetitively at −10 cm H₂O compared with another patient who arouses at highly negative pressures will be quite different. Waking up too easily (a low arousal threshold) is likely to contribute to repetitive respiratory events. Certain sedatives increase the arousal threshold (11, 12) and can promote breathing stability in patients with OSA with a low arousal threshold (12). On the other hand, quite a high arousal threshold may contribute to more severe hypoxemia and its accompanying consequences (6). Ultimately, the arousal threshold in combination with key anatomical and other nonanatomical phenotypic traits need to be considered on a per-patient basis to improve our understanding of this heterogeneous disorder. While the pathophysiological heterogeneity of OSA is complex, this observed heterogeneity also provides opportunities to better tailor therapies for patients with OSA (1).