Long range correleations of stride intervals in uphill and downhill trail running

Matteo Genitrini^1*Jon Wheat²Hermann Schwameder^3*

• ¹Lietuvos sporto universitetas, Kaunas, Lithuania
• ²Sport and Human Performance Enhancement Research Centre, Nottingham Trent University, Clifton Campus, Nottingham NG11 8NS, UK, University of Nottingham, Nottingham, United Kingdom
• ³Department of Sport and Exercise Science, University of Salzburg, Schlossallee 49, 5400 Hallein-Rif, Austria, Department of Sports and Tourism Management, Lithuanian Sports University, Kaunas, Lithuania

Participation in trail running competitions has steadily increased over the recent years. Relatively few studies have been carried out in ecological conditions. In the last decades, nonlinear analysis has drawn attention as a valuable tool to investigate time series of sports-related data. The goal of the present study was to investigate long range correlations in stride intervals during a full trail running time trial. Adopting an exploratory approach, it was hypo-thesized that the strength of such correlations would differ between uphill and downhill sections and between the initial and final stage of the race (incline and stage as independent variables). Twenty participants were recruited to run a solo all-out time trial equipped with inertial sensors to calculate stride intervals. The strength of long range correlations in stride intervals was quantified by means of Detrended Fluctuations Analysis alpha exponents (DFA-alpha). Differences across conditions were tested by means of linear mixed effect models. A significant main effect for incline was found, with higher values of DFA-alpha in downhill sections (resulting from less tight control) with respect to uphill. This is likely due to the higher technical difficulty running at high speed on an uneven surface. A significant main effect was found for race stage, with stronger correlations in the second race half as compared to the first one, most likely resulting from the difficulty to regulate running cadence in presence of acute fatigue. A significant interaction between incline and race stage was found as well, indicating that the strength of long range correlations in the second half of the race increased in both uphill and downhill sections, but the increase was significantly larger in uphill sections. This is likely due to the increase in physical fatigue which is prevalent in uphill sections, whilst the technical difficulty of downhill section remains constant. The present study shows that DFA-alpha is a sensitive quantity to discriminate between more and less challenging motor control scenarios. Incorporating such DFA-alpha among the metrics provided by wearables may aid runners in choosing a pacing strategy aiming to minimize fall and injury risks.