Development and validation of a pre-screening tool for early detection of low bone mineral density in female athletes (PreST-LBMD)

The prevalence of low bone mineral density (BMD) among female athletes is higher than that among their male counterparts up to 40%. Low BMD increases the risk of stress fractures.
To examine factors associated with low BMD in female athletes, and to develop, test, and evaluate a Pre-screening Tool for Early Detection of Low BMD in Female Athletes (PreST-LBMD).
The development and validation groups (381 and 233 subjects, respectively) underwent a medical interview and a BMD test at the lumbar spine using dual-energy X-ray absorptiometry (DXA). A Z-score < −1 was classified as low BMD. After extracting factors associated with low BMD from univariate logistic regression and multivariate least absolute shrinkage and selection operator (LASSO) regression analyses, each variable was scored based on the β coefficients from LASSO regression to create the PreST-LBMD. The development model was validated using receiver operating characteristic (ROC) curves. Furthermore, using external validation data, sensitivity, negative predictive value (NPV), and area under the curve (AUC) were calculated to evaluate the performance of PreST-LBMD.
Factors associated with low BMD included age at menarche, amenorrhea for more than 1 year during the teenage years, current duration of amenorrhea, body mass index, history of stress fractures, and exercise load. The calculated cutoff value for a high likelihood of low BMD was 3 points, with a sensitivity of 0.83 (0.79–0.88), a negative predictive value of 0.85 (0.81–0.90), and an area under the curve of 0.74 (0.67–0.81).
 We successfully developed the PreST-LBMD, which showed promising results identifying female athletes with low BMD. Those scoring ≥3 should be assessed by DXA and evaluated for the likelihood of low BMD by the athlete’s health and performance team.