Light exercise heart rate on-kinetics: a comparison of data fitted with sigmoidal and exponential functions and the impact of fitness and exercise intensity

This study examined the suitability of sigmoidal (SIG) and exponential (EXP) functions for modeling HR kinetics at the onset of a 5-min low-intensity cycling ergometer exercise test (5MT). The effects of training status, absolute and relative workloads, and high versus low workloads on the accuracy and reliability of these functions were also examined. Untrained participants (UTabs; n = 13) performed 5MTs at 100W. One group of trained participants (n = 10) also performed 5MTs at 100W (ETabs). Another group of trained participants (n = 9) performed 5MTs at 45% and 60% max (ET45 and ET60, respectively). SIG and EXP functions were fitted to HR data from 5MTs. A 30-s lead-in time was included when fitting SIG functions. Functions were compared using the standard error of the regression (SER), and test-retest reliability of curve parameters. SER for EXP functions was significantly lower than for SIG functions across all groups. When residuals from the 30-s lead-in time were omitted, EXP functions only outperformed SIG functions in ET60 (EXP, 2.7 ± 1.2 beats·min−1; SIG, 3.1 ± 1.1 beats·min−1: P < 0.05). Goodness of fit and test–retest reliability of curve parameters were best in ET60 and comparatively poor in UTabs. Overall, goodness of fit and test–retest reliability of curve parameters favored functions fitted to 5MTs performed by trained participants at a high and relative workload, while functions fitted to data from untrained participants exercising at a low and absolute workload were less accurate and reliable.