Maurizio Fanchini1,2, Ermanno Rampinini3, Marco Riggio2, Aaron Coutts4, Claudio Pecci3, and Alan Mccall5

1, University of Verona (Italy);2, US Sassuolo Calcio  (Italy);3, MAPEI Sport Research Centre  (Italy);4, University of Technology, Sydney (UTS) (Australia);5, Edinburgh Napier University (United Kingdom)

Introduction Monitoring load is considered a useful tool for injury reduction in elite soccer players. Session-RPE is a valid method to assess load in soccer and its association with injuries has previously been described (Malone et al., 2016). However, this study was conducted over a single season, and therefore, generalisability of these findings is limited. The aim of this study was to examine the association and predictive ability of monitoring load and non-contact injuries over multiple seasons. Methods Thirty-four players (age 265 y, height 1825 cm, body mass 784 kg) from an Italian serie A team participated in a prospective study including 3 in-season periods. Predictors examined were: weekly load (WL), week-to-week load change (W-WL), cumulative 2, 3, 4 WL, acute:chronic 1:3 (AC3) and 1:4 (AC4) WL ratios. Multicollinearity was checked between predictors. Generalized Estimating Equation analysis examined association between predictors and injury risk (IR) in the subsequent week. Signicant load predictors were split into 4 groups based on 15th, 50th, 85th percentile to compare IR in different zones. ROC curve analysis was performed to determine predictive ability. Results & Discussion Cumulative WLs were excluded from analysis for multicollinearity. IR increased when WL of 1086-1542, 1542-1985, > 1985 was compared with < 1086 au (OR, 90%CI 3.4, 1.4-8.3; 3.1, 1.3-7.5; 2.3, 0.8-6.2). An unclear di difference was found comparing WL> 1985 vs 1542-1985 and 1086-1542 au (OR, 90%CI 0.7, 0.4-1.4 and 0.7, 0.3-1.3). IR increased with W-WL change of -572-1, 1-614, > 614 au compared to < -572 au (OR, 90%CI 1.2, 0.6-2.5; 1.5, 0.8-2.9; 1.7, 0.8-3.5). IR increased when AC3 of 1.01-1.23, > 1.23 was compared to < 0.80 (OR, 90%CI 1.9, 0.9-3.8; 2.5, 1.2-5.4). IR increased when comparing AC4 of 0.78-1.02, 1.02-1.26, > 1.26 vs < 0.78 (OR, 90%CI 2.4, 1.4-3.9; 3.3, 1.6-6.6; 3.5, 1.7-7.4). An unclear IR difference was found comparing AC4> 1.26 vs 1.02-1.26 (OR, 90%CI 1.1, 1.0-1.1). Area Under ROC Curve values were 0.60 for all predictors. Conclusion IR was associated with an increase of WL. Lower IR (OR 0.7) was observed with higher WL (> 1985 au) compared with intermediate levels but the effect was deemed unclear. IR in W-WL change and AC4 were similar to recent reports using similar ranges (Malone et al., 2016). In contrast, no protective effect was found with AC4 between 1.00 and 1.25. Coaches should gradually achieve relatively high WL. However, the ability of the model to predict non-contact injuries is limited. References Malone, S et al., (2016). J Sci Med Sport; in press.