Prediction of ground reaction force during water immersion running

Abstract

This study aimed at developing a model to predict ground reaction force during deep-water running. A total of 20 subjects ((9 men, 11 women) ran in water at two immersion levels and three different speeds. Each subject performed six valid trials in each condition, data being captured by an underwater force plate. The force prediction model was build by multiple linear regression. Dependent variables were the vertical and anteroposterior components of the ground reaction force; independent variables were runners' immersion, sex, speed, body mass, body density, and percentage of fat. At the final regression model for the vertical component, only speed remained (p<0.001), while for the anteroposterior component, speed, immersion, and body mass were maintained (all at p<0.001). The obtained model for the anteroposterior component of ground reaction force may be found satisfactory, as adjusted determination coefficient was 0.79. However, the prediction model for the vertical component cannot be recommended for prediction during deep-water running, since that coefficient was 0.18. It must be noted that the proposed prediction model applies to subjects provided that they have similar characteristics to those who took part in this study.