[Internal validity and diagnostic utility of the Eating Disorder Inventory in Mexican women]

To assess the diagnostic utility and internal validity of the Eating Disorder Inventory (EDI-2) in Mexican women.

An observational, cross-sectional study was conducted between October 2000 and January 2001. The sample population consisted of two groups of women: patients diagnosed with anorexia nervosa or bulimia, but no other psychiatric pathology, seen by psychiatrists at the Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán (National Institute of Medical Sciences and Nutrition Salvador Zubirán), and a randomly selected sample of women from a school in Mexico City. A structured interview was conducted using SCID/DSM-IV as the gold standard. Finally the EDI-2 was administered. Statistical analysis was performed using Student's t test for independent samples, Cronbach's alpha test, and calculation of sensitivity, specificity, and positive and negative predictive values.

No significant differences were found for weight, height, and body mass index. Women with no eating disorders were younger than those with eating disorders (16.7 +/- 2.8 vs. 19.9 +/- 4.3, p = 0.004). The EDI-2 total score was clearly different between the groups (53.8 +/- 32.4 vs. 146.3 +/- 45.6 respectively, p < 0.001). All Cronbach's alpha coefficients for all subscales were over 0.85. For the total score, a sensitive cutoff point of 80 (sensitivity 90.9, 95% confidence interval [CI] 69.4 to 98.4; specificity 80, 95% CI 58.7 to 92.4) and a specific cutoff point of 105 (sensitivity 81.8, 95% CI 59 to 94; specificity 89.3, 95% CI 70.6 to 97.2) are recommended.

The EDI-2 has a high internal validity and should be a useful tool to assess eating disorders.