Prior to each test, the zero-load setting on the ergometer was checked, and a calibration was performed with a 4-kg weight. Ventilatory and pulmonary gas exchange variables were measured breath-by-breath in all tests using a portable system (K4b 2 , COSMED) which has been previously validated at HA ; the accuracy of the telemetric system has been previously established [step 18, 19]. The system comprised a face mask, analyzer unit (containing O2 and CO2 gas analyzers), heart rate monitor, and battery. The analyzer unit with battery pack, face mask, and tubing (weight 0.8 kg) was attached to the subject with a harness and connected to a personal computer by an Ethernet cable connection. The face mask contained a turbine for measurement of volume and flow; calibration was performed with a 3-l syringe (Hans Rudolph, Kansas City, MO, USA) over a range of different flow profiles. Respired gas, sampled continuously from a port within the turbine via a Nafion polymer capillary (PermaPure©, Toms River, NJ, USA), was analyzed at 100 Hz using rapid-response O2 (polarographic) and CO2 (infrared) analyzers (mean response time 120 ms) which were automatically thermostated and compensated for ambient variations in barometric pressure, humidity, and environmental temperature. Analyzers’ calibration was performed using two precision-analyzed gas mixtures spanning the respired range. The volume and gas concentration signals were sampled and digitized every 10 ms, and time-aligned, i.e., correcting for the transport delay between the turbine and gas analyzers and for the analyzer rise time . HR was measured from a chest strip and recorded every breath. SpO2 was monitored non-invasively by finger pulse oximetry (Masimo Rad-5, Masimo Corporation, Irvine, CA, USA).
Statistical analyses
Differences among measured responses were determined by a Student’s paired t test. Pearson’s product–moment correlation coefficient (R 2 ) was used to identify correlations between criterion variables. The level of statistical significance was set at P < 0.05. Group data are presented as mean ± SD. The limits of agreement between the V ? E / V ? CO 2 and the ? V ? E / ?HR methods for VCP estimation were evaluated by the Bland-Altman analysis , where the individual differences are plotted against their respective means. We proceeded with such a type of analysis if a significant linear correlation between methods was previously observed. The same statistical approach was also performed to compare the correspondence between ?L and the breakpoint between S1” and S1‘ in the ? V ? E / ?HR relationship, when the latter was detectable.
Efficiency
The main results of the incremental tests are illustrated in Table 2. At HA, compared to SL, a significant reduction in V ? O 2 peaks , WRpeak, HRpeak, SpO2peak, and ?L was observed; on the contrary, V ? Epeak at HA was appreciably higher.
Figures 1 and 2 show ? V ? E / ? O 2 , V ? E / V ? CO 2 , PMais aussiO2, PAinsi queCO2 vs. V ? CO 2 , and ? V ? E / ?HR relationship in two representative subjects at SL (upper panels) and HA (lower panels). As shown in Figure 3, in all subjects, a breakpoint in ? V ? E / ?HR , which occurred at the VCP estimated by the V ? E / V ? CO 2 method, how to delete blk account was clearly discernible both at SL and HA. No significant differences were found in V ? O 2 measured at VCP ( V ? CO 2 -VCP; Table 3) between methods utilized to identify the threshold (VCP- ? V ? E / ?HR vs. VCP- V ? E / V ? CO 2 ).