Familiarisation of novice and experienced treadmill users during a running session: Group specific evidence, time and...

The elimination of familiarisation effects is a recurring topic in biomechanical testing during treadmill running among different experience levels. The two aims of this study were (i) to calculate familiarisation times of novice and experienced treadmill runners on a group level and (ii) to examine individual familiarisation patterns in order to classify those with similar characteristics. Twenty runners participated in this study by performing a treadmill running session with 3D motion capture. Familiarisation times for 9 kinematic variables among both groups (novices and experienced treadmill runners) were statistically analysed. Additionally a qualitative clustering process (supported by quantitative criteria) provided individual familiarisation patterns for all participants and variables. Group mean familiarisation times were inconsistent across variables (ranging from 3 to 14 min), with no general tendency for decreased familiarisation time in experienced compared to novice treadmill runners. The analysis of individual familiarisation patterns revealed that 30.5% were not stable after 15 min. Substantial changes compared to the initial state were observed in data sets with detected familiarisation pattern. Treadmill running experience does not affect familiarisation patterns since this process is highly individual and variable-specific. Consequently, no generalised familiarisation time can be provided and the elimination of familiarisation in biomechanical testing a priori does not seem to be possible in the first 15 min for approximately one third of the individual patterns studied. In conclusion, the common practice of collecting data subsequent to a pre-defined generalised familiarisation time ought to be replaced by measurements at several points in time during trials. This required procedure would allow for checking familiarisation patterns and fluctuations in order to exclude inappropriate data sets in future treadmill studies.