The time constant (T) obtained by fitting post-exercise heart rate (HR) rec
overy to a first order exponential decay curve has been promoted as an inde
x of parasympathetic activity. However, acceptance has been limited because
reported data are inadequate to assess goodness of fit for the model, dete
rmine the best exercise protocol, or optimize the duration of post exercise
monitoring. Consequently, we evaluated T for nine healthy volunteers (age
24-46) following treadmill exercise at maximal (max) and two stages sub-max
exercise (Bruce protocol). T stabilized only after 3 min of post-exercise
monitoring. With max exercise, T varied unacceptably with small changes in
onset of monitoring, e.g. -16.7+/-16.6 (-13.2%) in the first 5 s, and resid
uals of the fitted curve were non-random. In contrast, sub-max exercise pro
duced consistent T values, e.g. -1.9+/-3.2 (-4.2%) in the first 5 s, and re
siduals were more nearly random. In conclusion, first order decay is an ina
dequate model for HR recovery following max exercise, but may be reasonable
for sub-max levels. (C) 2000 Elsevier Science B.V. All rights reserved.