Development of a Clinically Applicable Risk Score for Knee Injuries in Recreational Athletes: A Probabilistic Model for Preventive Decision-Making
Horacio Rivarola, Camilo Helito, Gonzalo Arteaga, Carlos Peñaherrera-Carrillo, Alejandro Barros Castro, Francisco Endara UrrestaBackground:
Knee injuries are highly prevalent among recreational athletes, who often lack access to structured prevention programs or professional supervision. While several predictive models exist for elite athletes, there is a critical need for accessible, clinically usable tools that can estimate injury risk in nonprofessional populations.
Purpose:
To develop and internally validate a probabilistic model for predicting knee injury risk in recreational athletes and to derive a simplified clinical scoring system for practical use in preventive care settings.
Study Design:
Cohort study (Diagnosis); Level of evidence, 2.
Methods:
This study was designed as a prospective multicenter cohort study including 628 recreational athletes aged 18 to 50 years followed for up to 12 months to monitor the incidence of clinically confirmed knee injuries. Baseline data included demographic, clinical, functional, and behavioral variables. A Least Absolute Shrinkage and Selection Operator–penalized logistic regression model was used to identify independent predictors and construct a probabilistic model, which was then converted into a point-based risk score. Model performance was assessed using receiver operating characteristic curve analysis, calibration metrics, and internal validation with Monte Carlo simulation.
Results:
During the follow-up, 92 participants (14.6%) sustained a knee injury. Nine variables were retained in the final model: age ≥35 years, body mass index ≥28 kg/m
2
, contact sport participation, previous knee injury, malalignment, single-leg squat failure, fatigue ≥4 out of 5, lack of warm-up, and metabolic disorder. The model demonstrated strong discriminative performance (area under the curve [AUC] = 0.83; 95% CI, 0.79-0.87) and good calibration (Hosmer-Lemeshow;
Conclusion:
This study introduces an internally validated, clinically applicable model for predicting knee injury risk in recreational athletes, based on routine clinical and functional variables. The derived score facilitates early risk identification and supports personalized preventive strategies, including physical therapy referral, training modification, and metabolic evaluation. Future studies are warranted to externally validate this model in larger and more diverse athletic populations to enhance generalizability.