Eating Disorder Diagnostic Scale (EDDS)

The Eating Disorder Diagnostic Scale (EDDS; Stice, Telch, & Rizvi, 2000) is a 22-item self-report questionnaire designed to measure Anorexia nervosa, Bulimia nervosa, and Binge-eating disorder symptomatology aligned with the DSM-IV diagnostic criteria.

The scale is comprised of a combination of Likert ratings, dichotomous scores, behavioural frequency scores, and open-ended questions asking for weight and height. The first four questions assess attitudinal symptoms of Anorexia and Bulimia within the past 3 months. The next four items measure the frequency of uncontrollable food consumption, with a focus on the number of days per week over the past 6 months (a criterion for Binge-eating disorder), and number times per week over the last 3 months (a criterion for Bulimia). The following four items measure frequency of compensatory behaviours. Lastly, individuals are asked to record their height, weight, presence of menstrual cycles and birth control pill use.

There are two further scales used in the EDDS that differentiate between eating disorders and deviance from healthy eating pathology. The diagnostic scale may be used to inform diagnosis of Anorexia, Bulimia and Binge-eating disorders. Stice et al. (2000) have developed a scoring algorithm to accompany this scale to determine score cut-offs. The symptom composite scale may be used to create a continuous composite score of disordered eating pathology.

Psychometric Development & Validation

The EDDS went through a rigorous development and validation process with careful adherence to a number of steps. The developers first generated a pool of items to assess DSM-IV eating disorder diagnostic criteria. These items were evaluated by a panel of 14 eating disorder experts, followed by revision to eventually produce the final EDDS to test for reliability and validity against an American female sample aged 13 to 65 years inclusive of those with and without eating disorders.

Results revealed excellent 1-week test-retest reliability for Anorexia (kappa = .95), and adequate test-retest coefficients for Bulimia (kappa = .71) and Binge-eating disorder (kappa = .75). The overall symptom composite test-retest reliability was also strong (kappa = .87). Likewise, internal consistency of the overall symptom composite score was robust (Cronbach’s α = .91). These reliability magnitudes reflect Shrout’s (1998) psychometric rule-of-thumb whereby kappa values above .8 represent high reliability, values between .4 and .8 indicate moderate agreement, and values less than .4 suggest poor reliability.

Content validity results generated by the 14 eating disorder experts revealed that items within the scale adequately reflected the DSM-IV diagnostic criteria for Anorexia, Bulimia, and Binge-eating disorder. Consistently, data also suggested that the EDDS possessed convergent validity by comparing participants with eating disorders with their non-diagnosis control counterparts; with higher scores reported for those with eating disorders than those without.

Strengths & Weaknesses

The EDDS has an abundance of strengths. It is short and quick to complete. With only 22 items, it takes only a few minutes to complete the entire instrument. It is sensitive to change over time; that is, the EDDS has the versatility of being used as a screening tool at the beginning stages of assessment, a diagnostic tool in supporting eating disorder diagnostic criteria, and lastly it may also be used for treatment monitoring and evaluation.

However, the EDDS is not without its limitations. In at least one study, the EDDS has been found to generate a large number of ‘false positives’ (Lee et al., 2007), indicating a weakness in specificity. Conversely, this may not necessarily be a negative drawback considering that when used as a screening tool it is preferable to be able to identify more people as false positives than run a risk of missing out on detecting potential cases of eating disorders. This is because eating disorders, though low in prevalence compared to other clinical disorders, has one the highest mortality rates amongst psychiatric conditions (Arcelus, Mitchell, Wales, & Nielsen, 2011). Additional weaknesses include gender and cultural insensitivity. Different attitudes towards food consumption for gender was found to be reinforced by differing cultural ideals–which were not adequately captured in the EDDS (Lee et al., 2007). Similarly, eating disorder pathology and risk factors were not invariant across Caucasian American women and African American women (Kelly et al., 2012).

Conclusion

Overall, the EDDS is a short and quick to complete self-assessment tool that is versatile to use as a screening measure, diagnostic instrument, and treatment evaluation and monitoring tool for the assessment of Anorexia, Bulimia, and Binge-eating disorder. Its tendency to detect more false positives need not necessarily be a weakness given the vulnerability of the eating disorder population as having some of the most severe mortality and prognosis rates amongst mental conditions. The lack of gender and cultural sensitivity warrants further modifications and refinements by researchers so that this tool can adequately capture the individual nuances that exist both within and between minority groups.

Access

The EDDS is freely available following this link: http://www.ori.org/files/Static%20Page%20Files/EDDS.pdf. Information regarding scoring and interpretation may be found here: http://www.ori.org/files/Static%20Page%20Files/SticeTelch00.pdf.

References