Nocturnal hypoxia is a major pathological factor associated with cardio-respiratory diseases including obstructive sleep apnea and congestive heart failure. During wakefulness, a decrease in arterial oxygen tension results in a decrease in cerebral vascular tone and a consequent increase in cortical blood flow; however, the cerebral vascular response to hypoxia during sleep is unknown. We recently reported that cerebral vascular reactivity to CO₂ is markedly reduced during sleep (Meadows et al. 2003). From this we hypothesised that, in normal human subjects, isocapnic hypoxic cerebral vascular reactivity is decreased during stage III/IV, non-rapid eye movement (NREM) sleep compared to wakefulness.

In 13 healthy male individuals (mean age ± sd: 22 ± 4), left middle cerebral artery velocity (MCAV) was measured using transcranial Doppler ultrasound. In each subject, the cortical blood flow responses to four separate conditions (eucapnic euoxia, isocapnic euoxia, isocapnic hypoxia -5 % arterial oxygen saturation (S_aJ{special}), and isocapnic hypoxia -10 % S_aJ{special}), were tested during wakefulness and during the first 90-minute cycle of stage III/IV, NREM sleep. Isocapnic hypoxia was achieved by regulating the inspired fraction of oxygen, whilst maintaining the end-tidal partial pressure of carbon dioxide (P_ETCJ{special}) within ± 2 mmHg of a predetermined level, independent of changes in ventilation (Banzett et al. 2000). NREM sleep was determined using electroencephalograms and electro-oculograms.

During wakefulness, in response to isocapnic hypoxia (-5 and -10 % S_aJ{special}), the mean (± S.E.M.) MCAV increased by 8.8 ± 1.7 % and 12.9 ± 2.2 %, respectively (P < 0.001, ANOVA); during NREM sleep, the same levels of isocapnic hypoxia was associated with a -6.97 ± 1.6 % and -7.0 ± 1.6 % reduction in MCAV (P < 0.001). Mean arterial blood pressure was unaffected by isocapnic hypoxia (P > 0.05); R-R interval decreased similarly in response to isocapnic hypoxia during wakefulness -21.9 ± 10.4 %, P < 0.001) and sleep -20.5 ± 8.5 %, P < 0.001).

In summary, during wakefulness, in response to isocapnic hypoxia, cortical blood flow increased; in contrast, during sleep, in response to the same degree of isocapnic hypoxia, cortical blood flow decreased. The inability of the cerebral vasculature to react to hypoxia during sleep suggests a major state-dependent vulnerability associated with the control of the cerebral circulation and may contribute to the pathophysiologies of stroke and sleep apnea.

This work was supported by The Wellcome Trust.