Open questions which can be addressed with human hypoxia models
The animal and human models used to study OSA through intermittent hypoxia induction are inadequate as they do not fully represent all aspects of the disease [77]. Researchers are therefore seeking better models to gain a deeper understanding of the underlying mechanisms. The healthy human model of OSA is continually evolving, with some limitations that can be resolved through technological adjustments. To maximize construct validity, experiments should be conducted overnight on sleeping participants rather than during the waking hours [78]. Secondly, the intermittent hypoxia model must include hypercapnia, which is present during obstruction in OSA. Thirdly, the dose and duration of intermittent hypoxia used in healthy human models should be equivalent to moderate to severe OSA, with mild to moderate OSA and equivalent intermittent hypoxia providing beneficial physiological stimuli [79]. Fourthly, the experimental model’s intermittent hypoxia should mimic the disease’s characteristic slow desaturation and rapid re-saturation of oxygen during obstructive events, as opposed to the current model with a square-wave design of rapid desaturation and re-saturation. Lastly, full polysomnography is necessary to characterize sleep architecture, including frequent brain arousals. The use of dial-down CPAP during sleep can induce upper airway obstruction and negative intrathoracic pressure swings, thus creating an experimental model that closely simulates OSA [80]. However, this technique is labor-intensive, invasive, and risky. The experimental model for OSA should be tailored to the research question at hand and may or may not require the complete simulation of OSA.