This study aims to explore the relationship between blood biochemical indexes and injury risk for elite male athletes in racing ice sports. The study compared the demographic indexes, monthly injuries, and longitudinal tracking data for the athletes. The non-linear relationship was analyzed using an unrestricted cubic spline. Generalized estimating equations were used to estimate the relative risk (OR) of injury occurrence. Receiver operating characteristics and the area under the curve determined diagnostic accuracy. In the snow sledding group, when creatine kinase rises to 489.46 u/L or testosterone decreases to 41.32 ng/ml, the risk increases by 1.70 times (OR=1.70, p<0.001) and 1.69 times (OR=1.69, p<0.001) with statistical significance. The creatine kinase (OR=1.01, P=0.007) and testosterone (OR=1.00, P<0.001) were included in the injury prediction model. The model exhibits excellent discrimination, with sensitivity and specificity of 82.8% and 86.5%, respectively. In the ice skating group, when creatine kinase rises to 467.00 u/L, the risk increases by 2.56 times with statistical significance (OR=2.56, p<0.001). Creatine kinase (OR=1.01, P<0.001) was included in the predictive model. The model demonstrates good discrimination, with sensitivity and specificity of 90.5% and 66.7%, respectively. Creatine kinase and testosterone are the risk predictors of injury in elite snowmobile male athletes. Creatine kinase is an independent risk factor for injury in elite speed skaters.
© Copyright 2025 International Journal of Sports Medicine. Thieme. All rights reserved.