Statistical approach for the estimation of daily physical activity levels using the probability density function of heart rate.

An investigation was undertaken into the statistical properties of heart rate during daily physical activity. The probability density function of the heart rate was estimated using the Gram-Charlier series. In addition, the probability density was separated into two Gaussian distributions: relatively low and relatively high heart rates. The former appeared to correspond to the metabolic rate associated with basic daily living and the latter appeared to be associated with more active physical activity of the type necessary to sustain or elevate the level of physical fitness. The higher heart rate distribution of five subjects occupied 8.72 +/- 2.15% of a period of waking. The validity of the statistical approach was confirmed with fractional estimation error of 1.22 +/- 0.62%.