Author(s): DOMINGUEZ, D., O'COYNE, C., ZAK, K., LOCKIE, R.G., ALVAR, B.A., SQUILLANTE, A., Institution: POINT LOMA NAZARENE UNIVERSITY, Country: UNITED STATES, Abstract-ID: 822

INTRODUCTION:
Repeated sprint ability (RSA) has the potential to be influenced by several key performance indicators in soccer. However, these have not been clearly identified in Division II collegiate soccer players. Therefore, the purpose of this study was to identify the key physical determinants of RSA in Division II collegiate players.
METHODS:
24 NCAA Division II men's and women's soccer players completed a two-visit testing battery. During the first visit, isometric mid-thigh pull and countermovement jump were performed using force plates sampling at 1000 Hz, measuring absolute and relative peak isometric force (STRabs; STRrel) and peak concentric power (POWabs; POWrel), respectively. The second visit comprised field-based tests including a RSA test consisting of six 20-m sprints with 15s recovery, a 30-m sprint with a 10-m split and a Yo-Yo Intermittent Recovery Test Level 1 (YYIRT1). RSA was evaluated by total sprint time (RSATT) and percent performance decrement (RSAPD). Aerobic capacity (VO₂max) was estimated from the YYIRT1. Univariate ordinary least squares (OLS) regression models were used to assess the independent predictive ability of key physical qualities for RSATT and RSAPD, with significance set at α = 0.05. Post-hoc multivariable OLS regressions tested whether combined predictors improved RSA prediction beyond univariable models.
RESULTS:
30-m sprint time explained the greatest variance in RSATT (R² = 0.773, β = 4.708, p = 1.55e-08), followed by 10-m sprint time (R² = 0.542, β = 8.662, p = 4.14e-05), POWabs (R² = 0.431, β = -0.00147, p = 0.0005), POWrel (R² = 0.391, β = -0.1246, p = 0.0011), STRabs (R² = 0.299, β = -0.00192, p = 0.0057), VO₂max (R² = 0.283, β = -0.3193, p = 0.0074), and STRrel (R² = 0.266, β = -0.1753, p = 0.0099). Multivariable OLS revealed a model with high fit for RSATT including 30-m sprint, 10-m sprint, and POWabs (R² = 0.795, adj. R² = 0.764), but substantial multicollinearity was present between linear speed metrics (VIFs > 1600). Within this model, only 30-m sprint remained a significant predictor (β = 6.661, p = 0.00052), whereas 10-m sprint (p = 0.163) and POWabs (p = 0.767) were insignificant. For RSAPD, the multivariable model was reduced to VO₂max alone (R² = 0.189, p = 0.0340).
CONCLUSION:
Faster 10- and 30m sprint times demonstrated the strongest relationships with RSATT, indicating that sprint speed is a primary determinant of RSA in Division II collegiate soccer players. Greater POWab, STRabs, and VO₂max were also significantly related to faster RSATT, highlighting meaningful contributions from both neuromuscular qualities and aerobic fitness. Multivariable modeling revealed the 30-m sprint to be the dominant predictor of RSATT, reflecting substantial shared variance with sprint speed and POWabs measures. In contrast, RSAPD was associated only with VO₂max, suggesting a greater reliance on aerobic capacity for maintaining performance across repeated sprint efforts.