Development of University of Tokyo's eating disorders inventory in female athletes

Aims: This study aimed to develop a scale to screen for eating disorders in female athletes. Methods: Preliminary survey: A total of 275 female athletes (mean age: 19.4 +/- 1.0 years) and 7 female athletes diagnosed with eating disorders (mean age: 20.1 +/- 2.5 years) were administered screening items prepared based on an existing scale, followed by exploratory factor analysis. Main survey: Six items, relating to three factors, were extracted, and 201 female athletes (mean age: 22.3 +/- 4.8 years) and 6 female athletes diagnosed with current or a history of eating disorders (mean age: 18.8 +/- 2.9 years) were queried. The diagnostic validity of the scale was then evaluated. Results: Preliminary survey: Questions (alpha = 0.71) were extracted from six items, relating to three factors, and collectively termed the University of Tokyo's eating disorders inventory in female athletes (TEDIFA). To determine the scale cut-off score, ROC analysis was performed with the total score, and the cut-off and gray zone scores were set at 13 and 11, respectively. Main survey: At the cut-off score of 13, AUC, sensitivity, and specificity were 0.83 (p < 0.05), 75%, and 90%, respectively. Conclusions: The scale that was developed, TEDIFA, consisted of six items. The cut-off scores were set at 11 for the gray zone (sensitivity: 75%; specificity: 56%; accurate diagnosis rate: 60%), and 13 for positivity (sensitivity: 75%; specificity: 90%; accurate diagnosis rate: 87%), demonstrating the reliability and validity of the scale.