Understanding the Factors that Increase Disordered Eating and Eating Disorder Risk among Female Undergraduates

This study employed a survey design to examine the relationship of numerous eating disorder (ED) risk factors (i.e., living environment, dieting to lose weight, food avoidance, shape concerns, weight concerns, dietary restraint, self-esteem, negative affect, thin-ideal internalization, and social media frequency use) among 211 college-aged women across the United States at ED risk. A multiple regression analysis was used to examine the strength of individual risk factors in predicting ED risk, and a discriminant analysis was used to categorize participants into low- (Eating Attitudes Test-26 [EAT-26] score of ≤ 10) and high-ED risk (EAT-26 score ≥ 11) groups based on the individual risk factors. The regression model was statistically significant, F(10, 206) = 30.13, p \u3c .001, accounting for 61% of the variance in ED risk by the various risk factors. Of all the risk factors, only dietary restraint, shape concerns, and thin-ideal internalization statistically significantly predicted ED risk. The discriminant analysis revealed a statistically significant discriminant function, Wilks’s lambda = .737, χ2(df = 10) = 60.913, p \u3c .001, with an eigenvalue of .36 and a squared canonical correlation of .26. All of the variables except for residence statistically significantly contributed to differentiating between the two ED risk groups. The model correctly classified 73.4% of participants, with more accurate predictions of low ED risk (81%) than high ED risk (62.8%). These findings suggest that multiple risk factors should be used in predicting ED risk with three risk factors (i.e., dietary restraint, shape concerns, and thin-ideal internalization) showing statistical importance in their predictive value, and using these factors seems to be effective in identifying those with low ED risk. Future work should be done to identify risk factors to improve early identification of those at high ED risk. These findings suggest differentiated practices in prevention and treatment efforts