Interaction of sleep state and chemical stimuli in sustaining rhythmic ventilation.

The effect of sleep state on ventilatory rhythmicity following graded hypocapnia was determined in two normal subjects and one patient with a chronic tracheostomy. Passive positive-pressure hyperventilation (PHV) was performed for 3 min awake and during nonrapid-eye-movement (NREM) sleep with hyperoxia [fractional inspired O2 concentration (FIO2) = 0.50], normoxia and hypoxia (FIO2 = 0.12). During wakefulness, no immediate posthyperventilation apnea was noted following abrupt cessation of PHV in 27 of 28 trials [mean hyperventilation end-tidal CO2 partial pressure (PETCO2) 29 +/- 2 Torr, range 22-35]. During spontaneous breathing in hyperoxia, PETCO2 rose from 40.4 +/- 0.7 Torr awake to 43.2 +/- 1.4 Torr during NREM sleep. PHV during NREM sleep caused apnea when PETCO2 was reduced to 3-6 Torr below NREM sleep levels and 1-2 Torr below the waking level. In hypoxia, PETCO2 increased from 37.1 +/- 0.1 awake to 39.8 +/- 0.1 Torr during NREM sleep. PHV caused apnea when PETCO2 was reduced to levels 1-2 Torr below NREM sleep levels and 1-2 Torr above awake levels. Apnea duration (5-45 s) was significantly correlated to the magnitude of hypocapnia (range 27-41 Torr). PHV caused no apnea when isocapnia was maintained via increased inspired CO2. Prolonged hypoxia caused periodic breathing, and the abrupt transition from short-term hypoxic-induced hyperventilation to acute hyperoxia caused apnea during NREM sleep when PETCO2 was lowered to or below the subject's apneic threshold as predetermined (passively) by PHV. We concluded that effective ventilatory rhythmogenesis in the absence of stimuli associated with wakefulness is critically dependent on chemoreceptor stimulation secondary to PCO2-[H+].