Hamstring Strain Injury Risk in Soccer: An Exploratory, Hypothesis-Generating Prediction Model.

Area of Science:

• Sports Medicine
• Biomechanics
• Injury Prevention

Background:

• Hamstring strain injuries (HSI) are prevalent in soccer, posing significant challenges for prediction due to their multifactorial nature.
• Traditional risk factors often fail to adequately capture individual susceptibility to HSI.
• Developing accurate predictive models is crucial for mitigating injury rates in athletes.

Purpose of the Study:

• To develop and internally validate a machine learning-assisted logistic regression model for predicting HSI in amateur soccer players.
• To identify key preseason clinical and strength-related predictors of HSI.
• To assess the model's performance using established metrics.

Main Methods:

• A prospective cohort study followed 120 male amateur soccer players for one season.
• Baseline data included age, BMI, injury history, and hip/knee strength measured by dynamometry.
• A logistic regression model with elastic-net regularization was trained and validated using nested cross-validation and feature ranking.

Main Results:

• The final model achieved moderate discrimination (AUC 0.68) and mild overfitting on the test set.
• Dominant-leg hip abduction strength was the sole significant predictor (OR=0.82).
• Previous hamstring injury was identified as the most stable predictor, while age and hamstring strength showed no predictive value.

Conclusions:

• The developed model offers preliminary, hypothesis-generating evidence for predicting HSI.
• Prior injury history and reduced hip abduction strength emerge as important prognostic markers.
• Further large-scale research with external validation is warranted to refine injury prediction models in soccer.