Causes of Different Force-Velocity Relationships in Sprint Running Depending on Horizontal Loads and Profiling Methods

Med Sci Sports Exerc. 2025 Jan 6. doi: 10.1249/MSS.0000000000003629. Online ahead of print.

Abstract

Purpose: We aimed to clarify how the horizontal force-velocity (Fvh) relationship during over-ground sprint running differs with horizontal resistance loads and profiling methods (multiple- and single-trial methods).

Methods: Twelve males performed sprint running (one unresisted and five resisted) using a motorized loading device. During the trials, the ground reaction forces at every step were obtained using a 50 m force plate system. The step-averaged Fvh relationships were then determined using single- and multiple-trial methods with linear and curvilinear models. The differences in Fvh parameters between loading conditions and between profiling methods, as well as the goodness of fit of the regression models to the measured data, were examined.

Results: We found that Fvh plots in each loading condition almost overlapped during acceleration, whereas, the horizontal forces deviated toward a lower value around maximal velocity; the linear Fvh parameters derived using the single-trial method had a load-dependency; the linear Fvh relationship derived from the multiple-trial method had a bias toward lower force values with less negative slopes compared with the single-trial method; the curvilinear models fitted the pooled data of all loading conditions better than the linear model; and the Fvh relationship within the velocity range of unresisted sprinting was almost linear.

Conclusions: The results of this study indicate that the reported load-dependency of Fvh parameters is mainly due to large horizontal forces at very low velocities in resisted sprinting, and the profiling method-dependency is mainly due to the attenuation of horizontal force around the maximal velocity of each loading condition. Factors of deviations from a linear Fvh relationship in horizontal force and the validity and usefulness of nonlinear models require further investigation.