Effects of exercise training during phase II and phase III cardiac rehabilitation on heart rate performance curve (HRPC) deflection (K
HR). A and B, Exemplary HRPCs. Time indicates the duration of an incremental exercise test. Blood lactate concentration after each step is used to determine LTP1 and LTP2. The region between LTP1 and the end of the exercise test (max) is used to determine K
HR by fitting a quadratic function to the heart rate data and relating the slopes of tangents at LTP2 and max (dotted lines) to each other (A) Upward‐deflected HRPC indicated by positive K
HR. B, Downward‐deflected HRPC indicated by negative K
HR. C, Descriptive statistics. K
HR values of each patient of the training group (n = 96) and the control group (n = 32) shown by thin, gray lines. Symbols indicate group means, and error bars show standard deviations. Horizontal arrows indicate the period in which regular exercise training was performed in each group. D, Inferential statistics. Estimated marginal K
HR value means of both groups with 95% confidence intervals after adjustment for the potential confounders age, baseline body weight, baseline power output in watts, smoking status (yes/no), and the use of β‐blockers (yes/no). The model is also adjusted for changes in β‐blocker intake over time. Symbols of each time point are slightly separated in x‐axis direction to avoid overlapping error bars. Note the adjusted y‐axis scaling compared to A. ***P < 0.0001 and the vertical bracket indicate the group difference at the end of phase III rehabilitation