Borghi-Silva, A., Cardiopulmonary Physiotherapy Laboratory,  São Carlos, Brazil;
Santos-de-Araujo, A.D., Montalvo, S., Christle, J.W. & Myers, J.

Prog Cardiovasc Dis 2026

BACKGROUND: Peak oxygen consumption (pVO2) has been widely recognized to identify potential pathophysiological mechanisms of exercise intolerance and to stratify risk in patients with heart failure (HF).
AIMS: To develop a non-exercise equation to estimate pVO2 (esVO2) and to determine its association with all-cause mortality.
METHODS: A total of 5176 HF patients (age 48 +/- 17 years) who underwent a CPET for clinical reasons, and were followed between May 2011, and December 1, 2023. Clinical and historical variables and forced expiratory volume in 1-s (FEV(1)) were used in a stepwise multivariate regression analysis to predict pVO2 measured from CPET. Kaplan-Meier curves and Cox proportional hazards analyses were used to determine the association between esVO2 and all-cause mortality.
RESULTS: The first esVO2 model was significantly associated with pVO2 (multiple R = 0.73, P < 0.001). Two other models considered gender separately (male: multiple R = 0.72, P < 0.001 and female (multiple R = 0.70, P < 0.001). A fourth model was proposed when spirometry is not available (age, BMI, resting SBP, HR and gender (multiple R = 0.69, P < 0.001)). The fifth proposed model included history of atrial fibrillation and previous stroke (multiple R = 0.70, P < 0.001). Kaplan-Meier survival curves differentiated patients according to esVO2 and pVO2 </= and > 14 mL/kg(-1) per min(-1). Univariate Cox proportional hazard analyses indicated significant prognostic value when esVO2 and pVO2 were </= 14 mL/kg(-1) per min(-1) (P = 0.001).
CONCLUSIONS: Multivariable nonexercise models combining spirometry and/or clinical variables commonly obtained during a clinical assessment predict pVO2 reasonably accurately in HF.