A novel method for estimating VO2peak using seismocardiography (SCG eVO2peak) at rest was developed some years ago, and this study investigates an updated version using data from a trained population. Forty five subelite female and Sixty seven elite male football players were tested and included in a cross-sectional accuracy comparison between SCG eVO2peak and VO2peak obtained from an incremental cardiopulmonary exercise test (CPET) performed on a treadmill. In addition, the ability of the method to detect changes over time was assessed by comparison of changes in VO2peak over 3 months (female and male, n = 60), 6 months (female, n = 17), and 1 year (male, n = 18) for CPET and SCG, respectively. No difference and a fair agreement were found between combined sex cross-sectional SCG eVO2peak and CPET VO2peak (bias ± 95% CI: -1.0 ± 0.9 mL·min-1·kg-1 and limits of agreement [95%]: ± 9.1 mL·min-1·kg-1), with similar agreement between sexes. The standard error of estimation and mean absolute percentage error were 4.8 mL·min-1·kg-1 and 7.4%, respectively. VO2peak was only increased over 6 months for females (CPET: 2.2 ± 1.1 mL·min-1·kg-1, SCG: 1.2 ± 1.1 mL·min-1·kg-1, and p < 0.05) with no difference between methods (p = 0.170). A correlation between VO2peak for CPET and SCG was found for 1-year analyses (male, R2 = 0.62, and P < 0.05) but no other correlations were significant. The SCG method is accurate in estimating VO2peak in both subelite and elite football players, with fair agreement and low estimation errors. The SCG method was able to detect an association of 1-year changes but not for changes over shorter times; however, these results require more investigation due to small sample sizes.
© Copyright 2025 European Journal of Sport Science. Wiley. Alle Rechte vorbehalten.