Author(s): GRANGE, T., BROCHERIE, F., CHESY, R., SAMOZINO, P., Institution: UNIVERSITÉ SAVOIE MONT BLANC, Country: FRANCE, Abstract-ID: 2687

INTRODUCTION:
High-intensity displacement (HID) has been characterized through high-velocity running and intense acceleration, analyzed independently. This does not account for the interplay between acceleration and running velocity and has been identified as a drawback since maximal acceleration capacity decreases continuously with velocity. Recently, HID relative threshold has been defined over the velocity spectrum as a fraction of maximal capacities considering running velocity effects on either maximal acceleration (acceleration-speed relationship: %AS) or external power output (%Pext-V). Considering both external and internal power (%Ptot-V) has not been tested yet. The aim was to compare the amount and distribution along the entire velocity continuum (from null to maximal running velocity S0) of HID detected by traditional method (TRAD, fraction of maximal acceleration and velocity) with recent %AS, %Pext-V and %Ptot-V methods.
METHODS:
61 youth rugby union players (29 backs, 32 forwards from U18 and U21 French Division 1) were enrolled. Maximal capacities according to each method (maximal acceleration and velocity, and AS, Pext-V, Ptot-V relationships) were determined from all season training sessions and games. Average game events above HID threshold were detected for each method as 75% of maximal capacities, and corresponding high-intensity work (HIW) was calculated. Correlations in HIW between methods were tested at both team and position (forwards vs. backs) levels. Distribution and magnitude of HIW along the velocity axis (%S0) were analyzed continuously with kernel density estimation. Between-method density curves were compared with statistical parametric mapping (SPM).
RESULTS:
At team level, TRAD displays high to very high correlation with %AS (r=0.84), %Pext-V (r=0.89), %Ptot-V (r=0.91). Correlations were slightly altered at position level with r ranging 0.80-0.88 for forwards and 0.82-0.87 for backs, associated with differences in HIW up to 207% between methods for a player. %AS showed very high correlation at both team and position level with %Pext-V (r=0.97) and %Ptot-V (r=0.97). %Pext-V and %Ptot-V showed nearly perfect correlation at any level (r=1.00). SPM analysis revealed that TRAD detected more HIW than %AS, %Pext-V and %Ptot-V at approximately 65-90% S0, whereas the opposite was observed at 20-50% S0. Moreover, %Pext-V and %Ptot-V detected more HIW than %AS at 2-92% S0, and %Ptot-V detected more HIW than %Pext-V at 8-95% S0.
CONCLUSION:
Even if TRAD correlates with recent %AS, %Pext-V and %Ptot-V methods, distribution of HID is shifted toward moderate to high velocities, potentially including low acceleration and high velocity events far from maximal capacity. TRAD could thus be limited in homogeneous populations or acceleration-oriented activity analysis. Recent methods are very highly correlated, with differences in HID being homogeneously distributed throughout the velocity spectrum but important differences in total HIW detected, %AS being the most restrictive.