Both hypoxia and carbon dioxide increase cerebral blood flow (CBF), and their effective interaction is currently thought to be additive. Our objective was to test this hypothesis. Eight healthy subjects breathed a series of progressively hypoxic gases at three levels of carbon dioxide. Middle cerebral artery velocity, as an index of CBF; partial pressures of carbon dioxide and oxygen and concentration of oxygen in arterial blood; and mean arterial blood pressure were monitored. The product of middle cerebral artery velocity and arterial concentration of oxygen was used as an index of cerebral oxygen delivery. Two-way repeated measures analyses of variance (rmANOVA) found a significant interaction of carbon dioxide and hypoxia factors for both CBF and cerebral oxygen delivery. Regression models using sigmoidal dependence on carbon dioxide and a rectangular hyperbolic dependence on hypoxia were fitted to the data to illustrate this interaction. We concluded that carbon dioxide and hypoxia act synergistically in their control of CBF so that the delivery of oxygen to the brain is enhanced during hypoxic hypercapnia and, although reduced during normoxic hypocapnia, can be restored to normal levels with progressive hypoxia.

Some animals have adapted to hypoxia by increasing their haemoglobin affinity for oxygen, but in vitro studies have not shown any change of haemoglobin affinity for oxygen in human high altitude natives or lowlanders acutely acclimatized to high altitude. We conducted the first in vivo study of the oxyhaemoglobin dissociation curve by progressively reducing arterial PO 2 while maintaining normocapnia in lowlanders at sea level, lowlanders sojourning at 3600m for two weeks and native Andeans at the same altitude. We found that the in vivo PO 2 at which haemoglobin is half-saturated (P 50 ) is higher in lowlanders at sea level (32 mmHg) than that measured in vitro (27 mmHg) and that lowlanders and highlanders do significantly increase the in vivo affinity of their haemoglobin for oxygen with exposure to high altitude. These results indicate the value of an in vivo approach for studying the oxyhaemoglobin dissociation curve.iii

We used Duffin's isoxic hyperoxic (P O 2 = 150 mmHg) and hypoxic (P O 2 = 50 mmHg) rebreathing tests to compare the control of breathing in eight (7 male) Andean highlanders and six (4 male) acclimatizing Caucasian lowlanders after 10 days at 3850 m. Compared to lowlanders, highlanders had an increased non-chemoreflex drive to breathe, characterized by higher basal ventilation at both hyperoxia (10.5 ± 0.7 vs. 4.9 ± 0.5 l min −1 , P = 0.002) and hypoxia (13.8 ± 1.4 vs. 5.7 ± 0.9 l min −1 , P < 0.001). Highlanders had a single ventilatory sensitivity to CO 2 that was lower than that of the lowlanders (P < 0.001), whose response was characterized by two ventilatory sensitivities (VeS1 and VeS2) separated by a patterning threshold. There was no difference in ventilatory recruitment thresholds (VRTs) between populations (P = 0.209). Hypoxia decreased VRT within both populations (highlanders: 36.4 ± 1.3 to 31.7 ± 0.7 mmHg, P < 0.001; lowlanders: 35.3 ± 1.3 to 28.8 ± 0.9 mmHg, P < 0.001), but it had no effect on basal ventilation (P = 0.12) or on ventilatory sensitivities in either population (P = 0.684). Within lowlanders, VeS2 was substantially greater than VeS1 at both isoxic tensions (hyperoxic: 9.9 ± 1.7 vs. 2.8 ± 0.2, P = 0.005; hypoxic: 13.2 ± 1.9 vs. 2.8 ± 0.5, P < 0.001), although hypoxia had no effect on either of the sensitivities (P = 0.192). We conclude that the control of breathing in Andean highlanders is different from that in acclimatizing lowlanders, although there are some similarities. Specifically, acclimatizing lowlanders have relatively lower non-chemoreflex drives to breathe, increased ventilatory sensitivities to CO 2 , and an altered pattern of ventilatory response to CO 2 with two ventilatory sensitivities separated by a patterning threshold. Similar to highlanders and unlike lowlanders at sea-level, acclimatizing lowlanders respond to hypobaric hypoxia by decreasing their VRT instead of changing their ventilatory sensitivity to CO 2 . Abbreviations HVR, hypoxic ventilatory response; P ET,CO 2 , end-tidal P CO 2 ; P ET,O 2 , end-tidal P O 2 ; T2, second ventilatory threshold; VeB, sub-ventilatory recruitment threshold basal ventilation; VeS, ventilatory sensitivity to CO 2 in highlanders; VeS1, first ventilatory sensitivity to CO 2 in lowlanders; VeS2, second ventilatory sensitivity to CO 2 in lowlanders; VRT, ventilatory recruitment threshold.

Because the skin is an oxygen sensor in amphibians and mice, we thought to confirm this function also in humans. The human upright posture, however, introduces additional functional demands for the maintenance of oxygen homeostasis in which cerebral blood flow and autonomic nervous system (ANS) function may also be involved. We examined nine males and three females. While subjects were breathing ambient air, at sea level, we changed gases in a plastic body-bag during two conditions of the experiment such as to induce skin hypoxia (with pure nitrogen) or skin normoxia (with air). The subjects performed a test of hypoxic ventilatory drive during each condition of the experiment. We found no differences in the hypoxic ventilatory drive tests. However, ANS function and cerebral blood flow velocities were modulated by skin hypoxia and the effect was significantly greater on the left than right middle cerebral arteries. We conclude that skin hypoxia modulates ANS function and cerebral blood flow velocities and this might impact life styles and tolerance to ambient hypoxia at altitude. Thus the skin in normal humans, in addition to its numerous other functions, is also an oxygen sensor.