A test of the interaction between central and peripheral respiratory chemoreflexes in humans.

How central and peripheral chemoreceptor drives to breathe interact in humans remains contentious. We measured the peripheral chemoreflex sensitivity to hypoxia (PChS) at various isocapnic CO 2 tensions ( P C O 2 ${P_{{\mathrm{C}}{{\mathrm{O}}_{\mathrm{2}}}}}$ ) to determine the form of the relationship between PChS and central P C O 2 ${P_{{\mathrm{C}}{{\mathrm{O}}_{\mathrm{2}}}}}$ . Twenty participants (10F) completed three repetitions of modified rebreathing tests with end-tidal P O 2 ${P_{{{\mathrm{O}}_{\mathrm{2}}}}}$ ( P ET O 2 ${P_{{\mathrm{ET}}{{\mathrm{O}}_{\mathrm{2}}}}}$ ) clamped at 150, 70, 60 and 45 mmHg. End-tidal P C O 2 ${P_{{\mathrm{C}}{{\mathrm{O}}_{\mathrm{2}}}}}$ ( P ETC O 2 ${P_{{\mathrm{ETC}}{{\mathrm{O}}_{\mathrm{2}}}}}$ ), P ET O 2 ${P_{{\mathrm{ET}}{{\mathrm{O}}_{\mathrm{2}}}}}$ , ventilation ( V ̇ $\dot{V}$ E ) and calculated oxygen saturation (S C O 2 ) were measured breath-by-breath by gas-analyser and pneumotach. The V ̇ $\dot{V}$ E - P ETC O 2 ${P_{{\mathrm{ETC}}{{\mathrm{O}}_{\mathrm{2}}}}}$ relationship of repeat-trials were linear-interpolated, combined, averaged into 1 mmHg bins, and fitted with a double-linear function ( V ̇ $\dot{V}$ E S, L min -1 mmHg -1 ). PChS was computed at intervals of 1 mmHg of P ETC O 2 ${P_{{\mathrm{ETC}}{{\mathrm{O}}_{\mathrm{2}}}}}$ as follows: the difference in V ̇ $\dot{V}$ E between the three hypoxic profiles and the hyperoxic profile (∆ V ̇ $\dot{V}$ E ) was calculated; three ∆ V ̇ $\dot{V}$ E values were plotted against corresponding S C O 2 ; and linear regression determined PChS (Lmin -1 mmHg -1 %S C O 2 -1 ). These processing steps were repeated at each P ETC O 2 ${P_{{\mathrm{ETC}}{{\mathrm{O}}_{\mathrm{2}}}}}$ to produce the PChS vs. isocapnic P C O 2 ${P_{{\mathrm{C}}{{\mathrm{O}}_{\mathrm{2}}}}}$ relationship. These were fitted with linear and polynomial functions, and Akaike information criterion identified the best-fit model. One-way repeated measures analysis of variance assessed between-condition differences. V ̇ $\dot{V}$ E S increased (P < 0.0001) with isoxic P ET O 2 ${P_{{\mathrm{ET}}{{\mathrm{O}}_{\mathrm{2}}}}}$ from 3.7 ± 1.5 L min -1 mmHg -1 at 150 mmHg to 4.4 ± 1.8, 5.0 ± 1.6 and 6.0 ± 2.2 Lmin -1 mmHg -1 at 70, 60 and 45 mmHg, respectively. Mean S C O 2 fell progressively (99.3 ± 0%, 93.7 ± 0.1%, 90.4 ± 0.1% and 80.5 ± 0.1%; P < 0.0001). In all individuals, PChS increased with P ETC O 2 ${P_{{\mathrm{ETC}}{{\mathrm{O}}_{\mathrm{2}}}}}$ , and this relationship was best described by a linear model in 75%. Despite increasing central chemoreflex activation, PChS increased linearly with P ETC O 2 ${P_{{\mathrm{ETC}}{{\mathrm{O}}_{\mathrm{2}}}}}$ indicative of an additive central-peripheral chemoreflex response. KEY POINTS: How central and peripheral chemoreceptor drives to breathe interact in humans remains contentious. We measured peripheral chemoreflex sensitivity to hypoxia (PChS) at various isocapnic carbon dioxide tensions ( P C O 2 ${P_{{\mathrm{C}}{{\mathrm{O}}_{\mathrm{2}}}}}$ ) to determine the form of the relationship between PChS and central P C O 2 ${P_{{\mathrm{C}}{{\mathrm{O}}_{\mathrm{2}}}}}$ . Participants performed three repetitions of modified rebreathing with end-tidal P O 2 ${P_{{{\mathrm{O}}_{\mathrm{2}}}}}$ fixed at 150, 70, 60 and 45 mmHg. PChS was computed at intervals of 1 mmHg of end-tidal P C O 2 ${P_{{\mathrm{C}}{{\mathrm{O}}_{\mathrm{2}}}}}$ ( P ETC O 2 ${P_{{\mathrm{ETC}}{{\mathrm{O}}_{\mathrm{2}}}}}$ ) as follows: the difference in V ̇ $\dot{V}$ E between the three hypoxic profiles and the hyperoxic profile (∆ V ̇ $\dot{V}$ E ) was calculated; three ∆ V ̇ $\dot{V}$ E values were plotted against corresponding calculated oxygen saturation (S C O 2 ); and linear regression determined PChS (Lmin -1 mmHg -1 %S C O 2 -1 ). In all individuals, PChS increased with P ETC O 2 ${P_{{\mathrm{ETC}}{{\mathrm{O}}_{\mathrm{2}}}}}$ , and this relationship was best described by a linear (rather than polynomial) model in 15 of 20. Most participants did not exhibit a hypo- or hyper-additive effect of central chemoreceptors on the peripheral chemoreflex indicating that the interaction was additive.