Comparison of Steady‐state and Rebreathing Methods for Measuring CO2 Chemoreflexes Under Hyperoxic and Hypoxic Conditions

Abstract

Two techniques for measuring ventilatory chemoreflexes are the steady‐state (SS) and rebreathing (RB) methods. In SS tests, a constant inspired PO2 and end‐tidal PCO2 are maintained to allow gas pressure equilibration at the tissues over several minutes. Ventilation is measured at two or more set points to determine CO2 chemosensitivity at a constant PO2. Alternatively, RB methods allow PCO2 to increase gradually as a participant rebreathes from a bag while PO2 remains constant. RB studies reveal a dog‐leg response, where PCO2 must rise above a subject‐specific threshold before ventilation increases. It is argued that measurements made below this threshold with SS may artificially lower the gain. Furthermore, it is unclear how these methods compare when conducted in hypoxia. This study aimed to compare ventilatory responses to CO2 during sustained hyperoxia and hypoxia within the same individuals, to determine (1) whether the difference in gain between methods is due to an inability to detect the recruitment threshold in the SS test, and (2) to determine how these methods compare at multiple PO2 levels. We hypothesized that the SS method would produce lower CO2 sensitivity responses at both oxygen levels due, in part, to measurements taken below the recruitment threshold. Healthy participants (N = 9 men, 10 women) between 18 and 65 years of age completed SS and RB tests under similar environmental conditions on the same day in a randomized order with a >30‐minute rest period between tests. The SS method targeted two stable end‐tidal PCO2 levels (45 and 55 mmHg) under inspired PO2 levels of 170–180 mmHg (hyperoxic) or 70 mmHg (hypoxic). We also conducted a modified Duffin’s RB method in 2 sessions with the same stable inspired PO2 levels. Data were analyzed with Friedman’s two‐way ANOVA with Bonferroni correction and data are presented as Median [Q1,Q3]. CO2 chemosensitivity measured with the SS method was lower than the RB method in both the hyperoxic (SS slope = 0.65 [0.48,0.94] L/min/mmHg CO2; RB slope = 1.70 [1.11,2.29]; p<0.001) and hypoxic (SS slope = 1.16 [0.89,1.46] L/min/mmHg CO2; RB slope = 1.87 [1.17,2.79]; p=0.034) tests. Additionally, chemosensitivity was higher during hypoxia compared to hyperoxia in the SS method (p=0.034). However, although ventilation was higher at a given PCO2 during hypoxia versus hyperoxia in the RB method due to a lower recruitment threshold (hyperoxia RT = 47.4 [45.4,49.2]; hypoxia RT = 43.3 [41.6,44.5], p=0.005), the slope above the threshold was not different (p=1.000). While the longer exposure time during SS likely alters hemodynamics and therefore gas tensions at both the central and peripheral chemoreceptors, this data also supports the idea that, without verifying that measurements are made above the recruitment threshold, the SS method produces lower sensitivity during hyperoxia. Furthermore, the SS method may produce higher gains during hypoxia due to the inability to detect the left shift along the metabolic hyperbola that occurs during hypoxia.Support or Funding InformationThis work was supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under awards to AM and FLP (R01HL081823).