The minute ventilation falls (1.23) and the arterial Pco2 rises by 2–3mmHg. The stage of NREM sleep oscillates frequently at sleep onset and the threshold for carbon dioxide to act as a respiratory stimulus fluctuates correspondingly. This may lead either to frequent central sleep apnoeas or to Cheyne–Stokes respiration. During the apnoeic phases the Pco2 rises progressively at a rate determined by the metabolic rate, and when it exceeds the apnoeic threshold, respiration restarts, often with a phase of hyperventilation until the Pco2 again falls below the threshold and the next apnoea begins.
The threshold for arousal from NREM sleep rises progressively from stage 1 to stage 4, but remains lower than in REM sleep  (
The reduction in cerebral cortical influence on the respiratory centres and the reduction in somatic reflexes have important effects on respiration in NREM sleep.
The respiratory drive is reduced, although it remains greater than in REM sleep. Within each NREM sleep stage the respiratory pattern is regular, but it varies between stages as the threshold for responding to Pco2 (apnoeic threshold) alters. The threshold for the ventilatory response to Pco2 increases from wakefulness to stage 1 NREM sleep and progressively into the deeper stages of NREM sleep with the effect that a
Pco2 which stimulates respiration during wakefulness may lead to apnoeas during sleep.
The activity of the chest wall muscles is globally reduced, unlike REM sleep in which the diaphragm is selectively spared, so that the ratio of rib cage to abdominal movement is greater in NREM than in REM sleep (
Fig. 2.1). The reduction in respiratory activity almost parallels the reduced ventilatory requirements needed to cope with the lower metabolic rate during NREM sleep.
The action of the upper airway muscles during sleep is complex. The dilators and constrictors are reciprocally inhibited and the dilators and the diaphragm are activated together. There is a sequence of activation from the alae nasi to the diaphragm so that the upper airway is stabilized a few milliseconds before the negative pressure is developed by chest wall muscle contraction. Negative pressure itself is sensed by receptors in the upper airway and this leads to further activation of the dilator muscles.
This response is less during NREM sleep than in wakefulness.
The upper airway resistance increases in NREM sleep compared to wakefulness, but to a lesser degree than during REM sleep. It increases the work of breathing and predisposes to obstructive sleep apnoeas.