Younger age (ORs 0.90-0.94) and male sex (OR 4.4) significantly predicted discordance between raw VO2peak thresholds for advanced heart failure therapies and predicted normal reference values.
Observational (n=81)
Do reference equations for VO2peak and O2-pulse show discordance with raw thresholds used to recommend advanced heart failure therapies?
Reference equations for normal physiologic ranges on CPET are not always concordant with raw thresholds used to recommend advanced heart failure therapies, highlighting the need for careful interpretation.
Abstract Introduction Multiple societies recommend patients with heart failure undergo cardiopulmonary exercise testing (CPET). Raw values for peak oxygen uptake (VO2peak), measured in cc/kg/min, are used to predict mortality and determine the need for advanced interventions. However, raw values do not account for the myriad of factors that determine VO2peak variability. We sought to compute reference values, using guideline recommended equations, to see how much they vary at the raw VO2peak thresholds that are used to identify candidates for advanced therapies. Methods This was a retrospective, observational study of heart failure patients referred for CPET. We abstracted baseline demographic data, echocardiography, and CPET data at peak exercise. We applied reference values published by Jones, Hansen, and the CanCold group. Using thresholds recommended to guide advanced therapies (VO2 cc/kg/min ≤ 14 for BMI 30, and 19 (adjusted to ideal body weight) for BMI 30) we compared percent predicted values for VO2peak and O2-pulse across reference equations. Results There were 81 patients with data available for analysis. Mean age was 58±11.9, 53 (65.4%) were male, and the majority was Black (41 (50.6%)) and White (30 (37.0%)). Mean ejection fraction (EF) was 26.2±10.9%, 46 (58.8%) had a non-ischemic cardiomyopathy (NICM), and 22 (27.2%) were diagnosed as ischemic (ICM). Mean VO2peak in L/min and cc/kg/min for the group were 1.4±0.5 and 15.8±4.3 respectively. There were 20 (24.7%) patients who fell below the thresholds used to select candidates for advanced therapies, and 61 (75.3%) who did not. There was notable discordance between advanced therapy thresholds and reference values. There were two (10%) patients who qualified for advanced therapies but had a normal VO2peak as predicted by Jones and CanCold. There were also two (10%) and 11 (55%) who had a normal O2-pulse by Jones and CanCold respectively, despite qualifying for advanced therapies. In logistic regression modeling, age (ORs 0.91 (0.86-9.97), 0.94 (0.90-0.99), and 0.90 (0.83-0.97) for Jones, CanCold, and Hansen respectively) and male sex (OR 4.4 (1.2-16.2) for Hansen only) were significant predictors of discordance between qualification for advanced therapy and VO2peak normality. Conclusion Reference equations used to establish normal physiologic ranges on CPET are not always concordant with thresholds used to recommend advanced heart failure therapies. Younger age and male sex were predictors of discordance. CPET results require careful interpretation and contextualization before decisions are made. This abstract is funded by: None
Holley et al. (2026) conducted an observational in Heart failure (n=81). CPET reference equations (Jones, Hansen, CanCold) vs. Raw VO2peak thresholds was evaluated on Discordance between qualification for advanced therapy and VO2peak normality. Younger age (ORs 0.90-0.94) and male sex (OR 4.4) significantly predicted discordance between raw VO2peak thresholds for advanced heart failure therapies and predicted normal reference values.