Eating Pathology Symptoms Inventory (EPSI)
- 2k Downloads
Purpose and Content
The Eating Pathology Symptoms Inventory (EPSI) is a 45-item self-report measure that is designed to assess the psychopathology of eating disorders. The EPSI contains eight scales that assess: body dissatisfaction (dissatisfaction with body weight and/or shape), binge eating (ingestion of large amounts of food and accompanying cognitive symptoms), cognitive restraint (cognitive efforts to limit or avoid eating, whether or not such attempts are successful), purging (self-induced vomiting, laxative use, diuretic use, and diet pill use), excessive exercise (physical exercise that is intense and/or compulsive), restricting (concrete efforts to avoid or reduce food consumption), muscle building (desire for increased muscularity and muscle building supplement use), and negative attitudes toward obesity (negative attitudes toward individuals who are overweight or obese).
The EPSI is free for noncommercial research or clinical use, and individuals may administer all scales or a subset of scales depending on their specific needs. Adaptations from the self-report format, changes to the instructions, or derivative works (e.g., language translations, adaptation for younger populations, etc.) must be approved in writing from the copyright holder (Dr. Kelsie Forbush). The EPSI has been tested in nonclinical and clinical samples and in men and women. For normative data, please refer to papers by Forbush and colleagues (2013, 2014). The EPSI is designed for use among individuals 14 years of age and older and has a Flesch-Kincaid reading grade level of six. A downloadable copy of the EPSI may be obtained at https://psych.ku.edu/kelsie-t-forbush
Traditional self-report measures have a variety of important strengths (such as good internal consistency and convergent validity), but also have problems that negatively impact their clinical utility. The EPSI was developed to address certain limitations of existing multidimensional eating disorder self-report measures. For example, many existing measures of eating disorder symptoms were developed using purely “rational” methods of scale development and, therefore, did not use statistical procedures to eliminate poorly performing items. This created issues related to poor discriminant validity (modest to high correlations among constructs that are not supposed to relate strongly to each other); indeed, research has shown that some measures of eating disorder symptoms (e.g., the Eating Disorders Inventory Bulimia scale) correlated more strongly with measures of depression than with other similar eating disorder symptoms (e.g., the Multifactorial Assessment of Eating Disorder Symptoms Purging scale) (Anderson et al. 1999). Another problem with traditional self-report measures of eating disorder symptoms is that the majority were developed on the basis of small samples that were exclusively (or almost entirely) comprised of young adult females, which may be why some studies indicate that certain eating disorder measures have reduced psychometric properties in men and in overweight and obese persons.
To address these concerns, Forbush and colleagues created a set of 20 dimensions (assessed with 160 items) that were thought to reflect important theoretical and clinical features of eating pathology [including, but not limited to, all of the aspects of eating disorder symptoms from the Diagnostic and Statistical Manual of Mental Disorders; DSM (APA 2013)]. These items were administered to several large samples of college students (N = 433) and community residents (N = 407). The measure was revised using a variety of statistical methods to eliminate poorly performing items and then administered to new samples of general outpatient psychiatric patients (N = 303) and patients recruited from a specialty eating disorder clinic (N = 158) to determine the final 45-item version of the measure. The EPSI was then administered to a sample of college students (N = 227) to test the short-term reliability of scale scores. The final version of the measure included eight (factor analytically derived) scales that were highly replicable across all samples and reflected dimensions with clear clinical relevance to eating disorders (see below for more information).
Although the EPSI is a relatively new measure, there are a few large-scale studies that describe its psychometric properties (see Forbush et al. under review, 2013, 2014; Tang et al. 2015). Reliability data indicate that the majority of scale scores are internally consistent (median coefficient alpha ranges from 0.84 to 0.89) and reliable over a 2- to 4-week period. Retest reliability for scale scores is above the minimum benchmark of 0.70 for “good” short-term reliability for all scales, except for cognitive restraint, which has a retest reliability of 0.61. Structural analyses indicate that the eight-factor structure of the measure is robust across gender (with the exception that the Muscle Building scale does not perform as well in women) and weight status (e.g., normal weight or overweight/obese) and has been shown to replicate across a number of independent samples. Recently, the EPSI was translated into Mandarin and Cantonese and tested in a large sample of native Chinese-speaking persons. It is noteworthy that the eight-factor structure of the EPSI was replicated in this sample, indicating that the EPSI may have potential utility for cross-cultural research. In all of the samples tested so far, the EPSI has shown evidence for good convergent validity with other measures of similar constructs and good discriminant validity from measures of non-eating disorder constructs (such as depression and anxiety). The EPSI scores have superior criterion-related validity (compared to the Eating Disorders Inventory-3 and Eating Disorder Examination-Questionnaire) for discriminating between anorexia nervosa and bulimia nervosa diagnoses and between clinical and nonclinical samples (Forbush et al. 2013).
Limitations and Strengths
The EPSI demonstrates good psychometric properties in normal samples of men and women, but needs to be tested in larger samples of males with diagnosed eating disorders. The EPSI has also not been tested in younger children or in a sizeable number of individuals in ethnic and racial minorities. It is unclear whether the EPSI can be used as a screening tool or what the optimal cutoff might be across scales for indicating the potential presence of an eating disorder diagnosis.
Despite these limitations, the EPSI scales have demonstrated clear evidence for reliability and a stable factor structure. This means that changes in scores likely represent actual changes in eating pathology across time rather than measurement error. The EPSI also provides excellent differentiation of various types of eating disorder behaviors, which is important for tracking clinical outcomes. For example, the EPSI differentiates between purging behaviors, restricting, and binge eating to enable maximal information on specific forms of disordered eating behaviors (in contrast to other measures that combine these behaviors within a single scale). Finally, the EPSI has been shown to enable greater criterion-related validity for distinguishing among anorexia nervosa and bulimia nervosa compared to certain other commonly used measures of eating pathology.
Given that the EPSI is a relatively new measure of eating pathology, it will be important for future studies to continue to accumulate evidence on the psychometric properties of the instrument, particularly its predictive validity. Other future areas for research include adaptations for younger samples and development of an interview-based version, both of which are currently underway. More research needs to be done to determine the psychometric properties of the EPSI scales in individuals who are in ethnic and racial minorities. Additional work is also needed to establish potential “cut scores” with maximal sensitivity and specificity for identifying eating disorder diagnoses so that the EPSI can be used for screening purposes (in addition to research and routine clinical care).
References and Further Reading
- APA. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: American Psychiatric Publishing.Google Scholar
- Forbush, K.T., Wildes, J.E., Pollack, L.O., Dunbar, D., Luo, J., Patterson, K., … Watson, D. (2013). Development and validation of the Eating Pathology Symptoms Inventory (EPSI). Psychological Assessment, 25, 859–878.Google Scholar
- Forbush, K. T., Hilderbrand, L., & Bohrer, B. K. (under review). Test-retest reliability of common measures of eating disorder symptoms in men versus women.Google Scholar
- Tang, X., Forbush, K. T., & Liu, P. (2015). Development and validation of a Chinese-language version of the Eating Pathology Symptoms Inventory (C-EPSI). International Journal of Eating Disorders, 48, 1016–1023.Google Scholar