Development and Validation of a Screening Tool to Identify Eating Disorders in the Female Athlete

Abstract

Participation in competitive athletics is associated with social pressures and performance anxieties that may increase the risk of eating disorders. Psychometric instruments routinely used in the general population may not accurately estimate eating disorders in the female athlete. Recognizing the need for a screening instrument to distinguish between those behaviors which are essential to athletic performance versus those which maintain eating pathology, we developed the development of the Female Athlete Screening Tool (FAST). The FAST was randomly administered to a diverse sample of 32 Division I university athletes. Concurrent validity, commonly known as comparison to a gold standard, was obtained by asking each subject to also complete the Bulimia Test Revised: BULIT-R and the Eating Disorder Inventory 2: EDI-2. FAST correlations with the BULIT-R inventory (r=0.69,p<0.0001) and four related subscales of the EDI-2: Drive for Thinness (r=0.70,p<0.0001), Bulimia (r=0.035,p<0.04), Body Dissatisfaction (r=0.78,p<0.000 l), and Perfectionism (r=0.48,p<0.006) supported its concurrent validity. To test the internal characteristics of the scale, a reliability analysis was performed. Using Cronbach's alpha, a high internal consistency of the remaining items (α=0.88) was demonstrated. Average item-to-total correlation was 0.51(0.30-0.74). Our results demonstrate the process involved in the initial validation of a new measure, the FAST. These findings support the internal reliability and concurrent validity of this measure.