Peak oxygen uptake during treadmill cardiopulmonary exercise testing: applicability analysis of the FRIEND registry prediction equation in a Brazilian healthy sample

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES). Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). Background Peak oxygen uptake (VO2peak) measured by cardiopulmonary exercise testing (CPET) is an important marker for health, physical performance and cardiovascular prognosis. Measured values are compared to reference standards for an adequate individual assessment. However, regional and international heterogeneities for VO2peak values have already been demonstrated by Milani, which may limit the external validation of the predictive equations developed in different nationalities. One of these equations was developed in an American population sample (FRIEND Registry) by Myers, but this equation has never been evaluated for portability in a population sample from the Brazilian Midwest region. Purpose To analyze the applicability of the FRIEND registry prediction equation for treadmill VO2peak in a healthy Brazilian sample from the Midwest region. Methods Cross-sectional study of subjects assessed by CPET in the Brazilian Midwest region between January/2011 to March/2020. The inclusion criteria were CPET performed on a treadmill in healthy subjects aged 20 years or older, and with respiratory exchange ratio ≥ 1.0. The exclusion criteria were any history of cardiovascular or pulmonary disease, presence of cardiovascular risk factors (such as hypertension, diabetes mellitus, current smoking, and obesity) and abnormalities on CPET. Variables were described as median and interquartile range. Statistical comparisons between measured and predicted VO2peak were performed using Spearman correlation test, Wilcoxon signed rank test, and Bland-Altman agreement analysis. Results During the study period, 7,843 CPET were performed and included for eligibility. After applying the inclusion and exclusion criteria, 3,544 assessments of healthy individuals were included (1,574 women; 1,970 men; age 20 to 80 years). The measured and predicted VO2peak were, respectively, 29.4 (25.1; 33.9) and 33.1 (29.5; 36.2) mL/kg/min in females, and 39.4 (34.3; 43.9) and 42.2 (39.1; 45.0) mL/kg/min in males, with significant differences (p < 0.0001). Correlations were significant (p < 0.0001) for both sexes (rho = 0.418 in women and 0.432 in men) (Figure 1). The mean error by the Bland-Altman agreement analysis was similar in both sexes, with a bias of -2.96 mL/kg/min (95% limits of agreement: -14.96 to 9.04 ) in females and a bias of -2.78 mL/kg/min (95% limits of agreement: - 15.39 to 9.83) in males (Figure 2). Conclusions The median values of the measured VO2peak were lower than the predicted values for both sexes, with a significant mean error, which limits the applicability of the FRIEND registry prediction equation in the Brazilian Midwest region population and highlights the international heterogeneity of the variable.