Model free endurance markers based on the second derivative of blood lactate curves.

Abstract

Blood lactate assessment is used regularly by exercise physiologists as a predictor of endurance performance. Typically fingertip blood samples are collected at selected running speeds on a treadmill and a plot of lactate concentration against workload (e.g. treadmill speed) is presented. Several features of the lactate curve have been suggested as markers for endurance. Typically these features, or endurance markers, are used to monitor changes in aerobic fitness, set training regimes and predict endurance performance. Determination of these markers however can be problematic (Weltman, 1995). In order to explore the structure of lactate curves, Functional Data Analysis (FDA) techniques are applied to provide graphical summaries. A new nonparametric endurance marker is then presented which corresponds to the point of maximum acceleration in the lactate curve. Using smoothing splines it is easily calculated by examining the second derivative of the smoothed lactate response. In addition we propose a simple discrete approximation to this marker.