Factors Associated With Concussion Diagnosis In Middle School (MS) Athletes

Research has established differences in the risk of sport-related concussion (SRC) attributed to individual factors, including sex, injury mechanism, and sport type. However, this research has focused on late adolescent and adult athletes, leaving the MS population underrepresented. To date, few studies have investigated if these individual factors are associated with the risk of SRC in MS athletes. PURPOSE: To evaluate the interaction between sex, grade level, sport, and mechanism in MS athletes diagnosed with SRC. METHODS: De-identified demographic and injury data was collected as a part of the ACHIEVES Project between the 2015-2022 school years were retrospectively examined. Four predictive models were evaluated using this dataset using Jupyter Notebook (V. 3.11.5), Python 3: Logistic Regression (LR), Decision Tree (DT), Naïve Bayes (NB), and Random Forest (RF). RESULTS: Across the six school years 10,458 athletes participated in MS sports. A total of 7,668 injuries including 238 SRCs were diagnosed. The models were trained on 80% and tested on 20% of the data. All models performed well: Area Under the Curve (AUC) =0.74 (LR), 0.75 (NB), 0.97 (DT) [See Figure]. The best performer was RF with AUC = 0.99. Within these models, Girls were more likely to be diagnosed with SRC than boys (LR:1.3). Factors with the highest likelihood ratios included soccer (2.2), football (2.4), and contact mechanism (2.6). CONCLUSIONS: Our findings indicate the individual factors are associated with the risk of SRC in MS athletes. Similar to previous research, MS female athletes were 1.3 times more likely to sustain a SRC compared to MS males. While our findings suggest that sex, sport, and mechanism are strong predictors of SRC, further research into causality and additional variables is needed.