The effect of non-sedative/hypnotic treatment of obstructive sleep apnoea on the respiratory arousal threshold: a systematic review and meta-analysis.

Contrary to the view that a lower arousal threshold causes greater susceptibility to obstructive sleep apnoea (OSA), the arousal threshold has been shown to decrease following treatment with continuous positive airway pressure (CPAP) therapy. This effect has not been consistently demonstrated with other non-CPAP therapies. We systematically reviewed all randomised and non-randomised trials (of non-hypnotic therapy) for OSA, in which respiratory arousal threshold was quantified both on- and off-treatment. Random effects meta-analysis models were used for all meta-analyses. Our analysis demonstrates that OSA treatment is associated with a significant reduction in the respiratory arousal threshold (standardised mean difference = -0.36 [CI95 % -0.49, -0.22], p < 0.001), however, this effect was highly heterogenous (I2 = 67.6 %). Greater reductions in the arousal threshold are observed in studies with highly efficacious treatment as quantified by larger absolute reductions in the apnoea-hypopnea index (AHI, r2 = 0.26, p = 0.002) and increases in SpO2 nadir (r2 = 0.74, p < 0.001). These data provide support for the notion that reductions in the respiratory arousal threshold occur as an indirect consequence of treatment. Importantly, reductions in the respiratory arousal threshold should be an expected finding in trials in which a treatment has successfully reduced OSA severity.