Rebuttal from J. E. Orr, B. A. Edwards and A. Malhotra.

Abstract

Younes and others have pioneered our understanding of loop gain (LG) in obstructive sleep apnoea (Younes et al. 2001). On his side of the argument, Younes eloquently differentiates between the ‘chemical’ LG of the respiratory control system and the ‘net LG’, which is an upper airway modulated chemical LG. We agree with this differentiation, as well as most of the arguments provided by our colleague. In fact, the only thing we disagree with is how the (limited) data addressing chemical LG as a cause is interpreted. At sleep onset, upper airway resistance increases due to upper airway collapse yielding CO2 increases as determined by plant gain. Hypercapnia will increase ventilatory drive as determined by controller gain. Increasing drive will eventually trigger arousal, and the cycle repeats. Thus, chemical LG (defined here as the product of controller and plant gain) must be a critical factor responsible for causing an individual's obstructive sleep apnoea (OSA). We currently treat OSA by preventing the initial increase in upper airway resistance with continuous positive airway pressure (CPAP). Alternatively, we could attenuate CO2 increases during respiratory events (i.e. acetazolamide) or increases in respiratory drive (i.e. oxygen). Both would be predicted to reduce an individual's propensity towards developing OSA. For a patient to develop OSA in the first place, they must have some degree of anatomical susceptibility; therefore, we view OSA as caused by ‘vulnerable’ upper airway anatomy/collapsibility, with the non-anatomical traits being thought of as ‘effect modifiers’ that dictate whether or not an individual develops OSA (Fig. ​(Fig.1).1). Considering LG as an effect modifier helps explain why treatments to reduce LG (i.e. oxygen or acetazolamide) often improve OSA in some but not all patients. The fact that OSA is not completely resolved with such therapies suggests that an elevated LG is not the only factor causing OSA. Recent data suggest that 23% of OSA patients have a highly collapsible upper airway, such that a high or low LG may be irrelevant (Eckert et al. 2013). However, the majority of patients have an airway in the intermediate range, where factors such as LG will affect the development of OSA. Therefore an elevated chemical LG is important in OSA pathogenesis in some patients. We hope that the current debate has produced a large enough disturbance in the research community to stimulate the appropriate response, i.e. future research in this area. Figure 1 Role of the non-anatomical traits in the pathogenesis of obstructive sleep apnoea