Optimal Methodology for Peak Oxygen Uptake (VO2 Peak) Sampling Interval in Patients on Hemodialysis

Background: Cardiopulmonary exercise testing (CPET) is the gold-standard diagnostic tool for determining oxygen uptake at peak exercise (VO2Peak). VO2Peak is an index of cardiovascular functional capacity and is significantly impaired in patients on hemodialysis (HD). VO2Peak is typically averaged using intervals between 10s-60s. However, there are no universally accepted methods for averaging VO2Peak in patients on dialysis. Herein, we sought to determine the optimal VO2Peak sampling interval in patients on HD compared to controls. Methods: We conducted a cross-sectional, secondary analysis using the Comparison of the Self-Paced versus Ramp Incremental Exercise Protocols on Patients with Kidney Failure (SPARK) cohort. Patients on HD (n=12) and healthy controls (CON) (n=24) completed two CPET visits. VO2Peak sampling intervals were averaged during the last 10, 20, 30, and 60 seconds of exercise. Kruskal-Wallis test comparisons were performed for continuous data. Reliability of each sampling interval was assessed by coefficient of variation (CV) between the two CPET visits. Results: Groups were matched in age and gender (P>0.05). BMI was higher in HD group (28.6±4.9 kg/m2) compared to the CON group (25.4±2.7 kg/m2; P=0.04). The median VO2Peak for the 60s sampling interval (11.2 [10.2, 13.9] mL/kg/min) was significantly lower (P=0.007) than the 10s (11.6 [10.0, 14.6] mL/kg/min), 20s (11.8 [10.0, 14.4] mL/kg/min), and 30s (11.5 [10.8, 14.7] mL/kg/min). The same trend was observed in the CON group with significantly lower VO2Peak values at the 60s interval (28.3 [24.4, 34.7] mL/kg/min) compared to shorter intervals (P<0.001). VO2Peak measurements exhibited low variability between CPET visits in both HD and CON groups. The median CV in the HD group ranged from 4.2% to 5.8% across the different intervals, with no significant differences observed (P=0.59). Similarly, the median CV for the control group showed no differences between the intervals, ranging from 1.9% to 2.9% (P=0.52). Conclusion: Longer sampling intervals (60s) significantly reduce VO2Peak in both HD and CON participants, while 10s, 20s, and 30s intervals showed no significant difference. The consistently low CV across all intervals in both groups supports the standardization of shorter VO2Peak sampling intervals. Funding: Other NIH Support - This project was funded, in part, with support by the Board of Directors of the Indiana University Health Vales Fund for Research Award and the Indiana Clinical and Translational Sciences Institute funded, in part by Grant # UL1TR001108 from the National Institutes of Health, National Center for Advancing Translational Sciences, Clinical and Translational Sciences Award