Is Hyperoxic Hyperventilation Caused by Reduced Carbon Dioxide Washout or Disturbed Brain Redox Homeostasis?

Abstract

Under oxygen‐enriched conditions, humans present a biphasic ventilatory response that is initially characterized by reduced ventilation (V̇E) as hyperoxia (HX) inhibits the peripheral chemoreceptors. This short‐lived response precedes a paradoxical increase in V̇E that has been termed hyperoxic hyperventilation. Proposed but untested mechanisms for this paradoxical response include central chemoreceptor stimulation via either an increased cerebral partial pressure of carbon dioxide (CO2), which would result from reductions in cerebral blood flow (CBF) and brain CO2–[H+] washout, or (2) an HX‐induced exaggerated production of reactive oxygen species (ROS). To determine the mechanism behind the paradoxical increase in V̇E during hyperoxemia, we submitted 8 healthy young men (24±1 yrs) to 10‐min trials of poikilocapnic HX under (1) saline or (2) ascorbic acid (AA) infusion. V̇E, arterial (a) and internal right jugular venous (Jv) PO2, PCO2, pH, oxidant (8‐isoprostane), and antioxidant (AA) markers, as well as volumetric CBF (Doppler ultrasound) responses, were quantified at the last minute of each trial. Under saline infusion, HX evoked an increase in PaO2 followed by a hyperventilatory response (+3.0 ± 1.2 L/min, p = 0.003) and arterial hypocapnia (−2.4 ± 0.9 mmHg, p = 0.03). HX also evoked a small but significant reduction in CBF (−219.6 ± 36.8 mL/min, p = 0.001) that was followed by unchanged PjvCO2 (−0.1 ± 0.7 mmHg, p = 0.619) and pH (0.001 ± 0.005, p = 0.957). HX under saline infusion also provoked an increase in arterial 8‐isoprostane (18.6 ± 7.5 pg/mL, p = 0.04). On the other hand, intravenous AA augmented arterial antioxidant markers (+584.1 ± 64.6, p = 0.0001 vs. saline) and blunted not only the increase in arterial 8‐isoprostane (−10.7 ± 9.5 pg/mL, p = 0.204) but also the HX‐induced hyperventilation (+0.87 ± 0.64 L/min, p = 0.232) and the arterial hypocapnia (−1.7 ± 0.5 mmHg, p = 0.04 vs. saline) observed during infusion of saline. HX under AA infusion also blunted CBF hypoperfusion (−41.8 ± 26.3 mL/min, p = 0.001 vs. saline) that, in turn, was followed by increases in PjvO2 (+5.3 ± 0.7 mmHg, p = 0.006 vs. saline), PjvCO2 (+1.1 ± 0.50 mmHg, p = 0.03 vs. saline) and a reduction in pH (−0.012 ± 0.004, p = 0.01) that may be explained by the Haldane effect. These findings suggest that hyperoxic hyperventilation is not a consequence of an increased cerebral CO2 partial pressure and a reduced brain CO2–[H+] washout but mediated by central chemoreceptor stimulation likely evoked by excessive production of ROS.Support or Funding InformationSupport: FaperRJ, CNPq, CAPES