The reliability of VO2(peak) determination in healthy females during an incremental arm ergometry test.

This study examined the reliability of an incremental arm ergometry (AE) protocol during peak aerobic power (VO2(peak)) determination in healthy females. Fifteen females completed two incremental AE tests to exhaustion, seven to eleven days apart, using a mechanically braked arm ergometer. The initial work rate was 16W and increased by 16W every two minutes until exhaustion. Significant differences between tests were determined by repeated measures ANOVA, and paired t-tests or Wilcoxon signed-rank tests, where appropriate. Reliability was determined by intraclass correlation coefficients (ICC), typical error, coefficient of variation (CV) and measurement bias/ratio and 95% limits of agreement (LOA). Peak cardio-respiratory responses were similar between tests, except for tidal volume (1.95 +/-0.47 vs. 1.81 +/- 0.41 L, P<0.05). Reliability for peak variables was moderate to high (ICC=0.659-0.941; CV< or =10%) while LOA were considerable for most variables including VO2(peak) (LOA=0.57 L.min(-1)). Similar peak cardio-respiratory responses, low CV and moderate-high ICC confirmed the reliability of the current incremental AE protocol to be similar to that of prior reported protocols for VO2(peak) determination. Substantial within-participant variability (LOA) for respiratory rate and tidal volume was common during the AE protocol and possibly reflects the influence of respiratory entrainment on reliability that requires further investigation.