Carbone S, Kim Y, Kachur S, Billingsley H, Kenyon J, De Schutter A, Milani RV, Lavie CJ
Aims Cardiac rehabilitation (CR) improves survival in patients with coronary heart disease (CHD), which is largely mediated by the improvements in cardiorespiratory fitness (CRF) defined as peak oxygen consumption (VO2). Therefore, measuring CRF is essential to predict long-term outcomes in this population. It is unclear, however, whether peak VO2 achieved at the end of CR (END-peak VO2) predicts survival or whether the changes of CRF achieved during CR provide a greater prognostic value. To determine whether END-peak VO2 independently predicts long-term survival in patients with CHD undergoing CR. We also aimed at identifying cut-offs for END-peak VO2 that could be used in clinical practice. Methods and results Retrospective analysis of 853 patients with CHD referred to CR who completed a maximal cardiopulmonary exercise test. Survival analysis was performed to examine the risk of all-cause mortality (average follow-up years: 6.65) based on peak VO2. The Contal and O'Quigley's method was used to determine the optimal cut-off of END-peak VO2 based on the log-rank statistic. END-peak VO2 was inversely associated with mortality risk [hazard ratio (HR) = 0.84; 95% confidence interval (CI) = 0.78-0.90], independent of changes in peak VO2 adjusted for the baseline peak VO2. The estimated cut-off of END-peak VO2 at >= 17.6 mL/kg/min best predicted the survival with high predictive accuracy and patients with END-peak VO2 under the cut-off had a greater risk of mortality (HR = 2.93; 95% CI = 1.81-4.74). Conclusions In patient with CHD undergoing CR, END-peak VO2 is an independent predictor for long-term survival. Studies utilizing higher intensity CR programmes, with and without pharmacologic strategies, to increase peak VO2 to a greater degree in those achieving a suboptimal END-peak VO2, are urgently needed.