Chemoreceptor drives and short sleep-wake cycles during hypoxia: a simulation study.

We used a Grodins-type mathematical model of the cardio-pulmonary system to investigate the cycling behaviour of the respiratory system during hypoxia in response to changes of state between waking and sleeping, namely a diminution of respiratory chemosensitivity during sleep. Shifts between waking and sleeping were triggered by various combinations of threshold values for PAO2. Mild or moderate hypoxia was simulated by values of inspired O2 concentration between 13% and 16% (normal 21%). In mild or moderate hypoxia, reductions of overall respiratory gain from about 2 to 0.8 l.min-1 mmHg-1 at sleep onset will produce falls in PAO2 likely to cause sleep-wake cycles with oscillations in PAO2. The higher the arousal threshold (in relation to steady-state PAO2 during sleep), the shorter and more stable the sleep-wake cycles. As the arousal threshold is raised, and as hypoxia is exacerbated from mild (FIO2 = 16%) to moderate (FIO2 = 13%), the sleep-wake cycle length tends to converge to a value around one minute. The level and determinants of the "back-to-sleep" threshold are hard to define from presently available experimental data, but the level is not important in determining the length of the sleep-wake cycle compared to the arousal threshold. Alinearities in chemoreceptor feedback were introduced first by incorporating "drive" thresholds, to simulate central or obstructive apnoea. This produced larger oscillations in respiratory variables, but no change in cycle length. Chemoreceptor thresholds for PCO2 at the level of 38-39 mmHg did produce shorter ventilation cycles, down to about 20 s in length, but these were not related in any simple way to the resulting sleep-wake cycles. The combination of sleep state changes and chemoreceptor feedback alinearities can produce short sleep-wake cycles despite the diminution in chemoreceptor gain occurring in sleep.