Abstract

BACKGROUND: This systematic review and meta-analysis aims to explore in heart failure (HF) patients with reduced ejection fraction (EF) undergoing exercise-based cardiac rehabilitation: 1) the comparison of temporal changes between peak oxygen uptake (VO2peak) and first ventilatory threshold (VO2VT1), 2) the association of VO2peak and VO2VT1 changes with physiological factors and 3) the differential effects of continuous aerobic exercise (CAE) and interval training (IT) on VO2peak and VO2VT1. METHODS: A systematic literature search was conducted in PubMed, CENTRAL, and Scopus. INCLUSION CRITERIA: 1) original research articles using exercise-based cardiac rehabilitation, 2) stable HF patients with reduced EF, 3) available values of VO2peak and VO2VT1 (in mL/kg/min) for both baseline and after exercise training with comparison between these time points. RESULTS: Among the 30 eligible trials, 24 used CAE, 5 IT and one CAE and IT. Multivariable meta-regression with duration of exercise training and percentage of males as independent variables and the change in VO2peak as dependent variable showed that the change in VO2peak was negatively associated with duration of exercise training (coefficient= -0.061, p=0.027), implying the possible existence of a waning effect of exercise training on VO2peak in the long term. Multivariable meta-regression demonstrated that both age (coefficient= -0.140, p<0.001) and EF (coefficient=0.149, p<0.001) could predict the change in VO2VT1, whereas only age (coefficient= -0.095, p=0.022), but not EF (coefficient=0.082, p=0.100) could predict the change in VO2peak. The post-training peak respiratory exchange ratio, as an index of maximum effort during exercise testing, correlated positively with the change in VO2peak (coefficient=0.021, p=0.044). The exercise-induced changes of VO2peak (p=0.438) and VO2VT1 (p=0.474) did not differ between CAE and IT. CONCLUSIONS: The improvement of endurance capacity during cardiac rehabilitation may be detected more accurately with the assessment of VO2VT1 rather than VO2peak.