Reproducibility of Respiratory Chemoreflex Characterization by Modified Rebreathing.

Modified hypoxic and hyperoxic rebreathing is used to characterize central and peripheral respiratory chemoreflexes in humans. This study examined within- and between-day reproducibility of the key parameters that are identified by this method: ventilatory recruitment threshold (VRT) and chemoreflex sensitivity. Seven healthy males (mean ± SD; age: 27±5 years) visited the laboratory on 4 occasions to performrepetitions of modified rebreathing tests in isoxic-hypoxic (end-tidal PO2 (PETO2) =50 mmHg) and -hyperoxic conditions (PETO2=150 mmHg). Days 1 and 2 consisted of three repeatedhyperoxic and three hypoxic trials, respectively; Days 3 and 4 included both a hyperoxic and a hypoxic trial completed in a counterbalanced order. Ventilation, end-tidal PCO2 (PETCO2) and PETO2 were measured breath-by-breath by pneumotach and dual gas analyser. For each rebreathing trial, the PETCO2 at which ventilation began to rise was identified as the VRT (mmHg) and the slope of the response above VRT provided respiratory chemoreflex sensitivity (L∙min-1∙mmHg-1). Within- and between-day reproducibility was assessed by one-way ANOVA and intraclass correlation (ICC) analyses. The VRT was not different between trials performed on a single day for bothhyperoxic (44±2, 45±3, vs 44±3 mmHg; for Trials 1, 2, vs 3, respectively; p=0.50; ICC=0.926) and hypoxic rebreathing (40±3, 40±4, vs 40±4 mmHg; p=0.66; ICC=0.994). Chemoreflex sensitivity also was not different between trials performed on the same day in hyperoxic (4.3±2.2, 5.0±3.1, vs4.8±2.3 L∙min-1∙mmHg-1; p=0.33; ICC=0.955) or hypoxic conditions (5.6±2.2, 6.2±2.7, 5.7±2.0 L∙min-1∙mmHg-1; p=0.33; ICC=0.953). Between days, the VRT was not different for hyperoxic rebreathing (44±3 vs 44±2 vs 45±2 mmHg for Days 1, 3, vs 4, respectively; p=0.10; ICC=0.813) but was different for hypoxic rebreathing (40±3 vs 40±2 vs 41±3 mmHg for Days 2-4, respectively; p=0.01; ICC=0.931). Chemoreflex sensitivity was not different between days for either hyperoxic (4.8±2.5,4.4±2.7, vs 4.9±2.0 L∙min-1∙mmHg-1; p=0.01; ICC=0.906) or hypoxic rebreathing (5.8±2.2, 5.8±2.6, vs 7.8±2.8 L∙min-1∙mmHg-1; p=0.01; ICC=0.750). Collectively, these data indicate that model parameters derived from hyperoxic and hypoxic modified rebreathing have good-to-excellent test-retest reproducibility when repeated on the same or different days.