Direct measurement of peak oxygen uptake in cardiopulmonary exercise testing as a survival predictor in octogenarians

Abstract Background The direct measurement of peak oxygen uptake (VO2peak) through cardiopulmonary exercise testing (CPET) serves as the gold standard for assessing cardiorespiratory fitness. This measurement is a firmly established indicator of health and holds substantial prognostic value concerning all-cause mortality in both populations with and without cardiovascular diseases. However, due to the scarcity of older individuals in studies employing CPET, its significance in this phase of life is less explored. Existing data on octogenarians primarily relies on the indirect measurement of VO2peak through the estimation of peak metabolic equivalents achieved (MET). It is worth noting that indirect estimation is generally less precise compared to direct measurement through CPET. This gap raises questions about the enduring validity of VO2peak as a robust predictor of mortality and morbidity in this specific demographic, akin to its predictive efficacy among younger individuals. Purpose To analyse the role of VO2peak as a predictor of survival in octogenarians. Methods Patients aged 80 years or older who underwent CPET assessment at a single centre from Brazil were included in this study. The variable VO2peak was determined as the highest 30-second average sample obtained during the final minute of the effort. The measured value was subsequently compared to predicted values by the Wasserman algorithm to calculate the percent-predicted VO2peak achieved (%VO2peak). The optimal cut-off point for the %VO2peak was determined by receiver operating characteristic (ROC) curve analysis, and this value was used as a cut-off to categorise individuals with %VO2peak higher or lower. Then, Cox regression models were employed to evaluate this dichotomisation to predict survival outcomes. Results Among 188 patients assessed, 22 (11.7%) died during the median follow-up of 494 days, with non-survivors being older, having lower VO2peak, and lower %VO2peak (Figure 1). Figure 2A illustrates the ROC curve for mortality prediction by %VO2peak. The area under the ROC curve was 0.699 (95% CI: 0.627 to 0.764; p 0.001), and the optimal cut-off point was determined to be %pVO2 ≤ 80%. Cox regression models indicated associations between %VO2peak ≤ 80% and mortality: HR=3.19 (95% CI: 1.30 to 7.86, p=0.011) for the unadjusted model (Figure 2B); and HR=3.124 (95% CI: 1.26 to 7.74, p=0.013) for the model with inclusion of age as a covariate to evaluate the impact of %VO2peak while accounting for age-related effects. Conclusion Amidst the backdrop of a rapidly aging population, our study highlights the enduring significance of VO2peak as a survival predictor among the elderly. These findings hold profound implications for tailoring healthcare strategies to address the evolving demographic landscape.