Eating Disorders Inventory (EDI)

Abstract

The Eating Disorders Inventory (EDI), originally developed by David M. Garner, Marion P. Olmsted, and Janet Polivy in 1983, is one of the most widely utilized and psychometrically validated self-report assessment batteries in clinical psychology, psychiatry, and behavioral medicine. Designed to evaluate the multidimensional psychological and behavioral traits characteristic of anorexia nervosa and bulimia nervosa, the original EDI comprises 64 items divided into eight clinically derived subscales. These subscales assess three core eating-specific cognitive-behavioral domains—Drive for Thinness (DT), Bulimia (B), and Body Dissatisfaction (BD)—and five broad, personality-related psychological constructs central to the maintenance of eating pathology: Ineffectiveness (I), Perfectionism (P), Interpersonal Distrust (ID), Interoceptive Awareness (IA), and Maturity Fears (MF).

Respondents evaluate each statement using a 6-point forced-choice Likert-type scale ranging from 1 (“Always”) to 6 (“Never”). In standard clinical scoring protocols, responses are recoded into a 4-point scale (0–3) to isolate severe pathological presentations, while continuous scoring paradigms (1–6) are frequently deployed in non-clinical epidemiological contexts to maximize variance. Psychometric investigations across diverse international cohorts consistently reveal robust internal consistency (Cronbach’s alpha coefficients typically ranging from α = .80 to .93 for eating-specific scales and .65 to .88 for general psychological traits), substantial test-retest reliability across 1- to 3-week intervals (r = .75–.95), and pronounced convergent and discriminant validity against established criterion instruments, such as the Eating Attitudes Test (EAT-26) and clinical diagnostic interviews. This paper provides an exhaustive review of the psychometric architecture, theoretical foundations, factor analytic structural evidence, scoring protocols, and cross-cultural adaptations of the foundational 64-item EDI.

Keywords

Eating Disorders Inventory, EDI, anorexia nervosa, bulimia nervosa, Drive for Thinness, Body Dissatisfaction, Interoceptive Awareness, perfectionism, psychometrics, self-report inventory, eating pathology assessment

Authors

The original Eating Disorders Inventory was conceptualized, operationalized, and validated by a multidisciplinary team of clinical psychologists and researchers specializing in psychosomatic medicine and eating pathology:

• David M. Garner, Ph.D. — Eminent clinical psychologist, former Director of Research at the Toronto General Hospital Eating Disorder Centre, former Professor in the Departments of Psychiatry and Psychology at the University of Toronto and Michigan State University, and Founder of the River Centre Clinic (Sylvania, Ohio, USA). Dr. Garner is widely recognized for his pioneering contributions to the empirical conceptualization, measurement, and cognitive-behavioral treatment of eating disorders.
• Marion P. Olmsted, Ph.D., C.Psych. — Senior Clinical Psychologist and Director of the Ambulatory Eating Disorders Program at the University Health Network, Toronto General Hospital, and Associate Professor in the Department of Psychiatry at the University of Toronto (Toronto, Ontario, Canada). Dr. Olmsted has conducted extensive psychometric and treatment outcome research in adult anorexia nervosa and bulimia nervosa.
• Janet Polivy, Ph.D. — Professor Emerita of Psychology and Psychiatry at the University of Toronto Mississauga (Mississauga, Ontario, Canada). An internationally recognized expert on dieting, restrained eating, self-regulation, and the behavioral consequences of caloric restriction, Dr. Polivy co-developed seminal constructs in eating behavior research, including the Restraint Scale and the behavioral mechanisms of binge eating.

Purpose

The Eating Disorders Inventory was engineered to address a critical empirical gap in the assessment of eating pathology during the late 1970s and early 1980s. Prior to its publication, available psychometric instruments—such as the original 40-item Eating Attitudes Test (EAT)—focused predominantly on overt behavioral symptoms, dieting habits, and food-related preoccupations. While effective as unidimensional screening devices for identifying potential cases of anorexia nervosa, existing tools failed to capture the multifaceted, heterogeneous psychological disturbances that underpin, sustain, and differentiate eating disorder variants.

Drawing on the clinical formulations of Hilde Bruch and contemporary cognitive-behavioral frameworks, Garner, Olmsted, and Polivy (1983) conceptualized eating disorders not merely as aberrant feeding behaviors, but as complex psychosomatic disturbances involving profound cognitive distortions regarding body weight and shape, severe self-regulatory and affective deficits, pervasive feelings of personal inadequacy, and developmental anxieties. The primary objective of the EDI was therefore twofold:

1. To provide a standardized, psychometrically sound instrument capable of measuring specific cognitive and behavioral dimensions directly tied to eating and weight regulation (eating-specific pathology).
2. To quantify broad, fundamental psychological traits and ego-deficits that characterize individuals with severe eating disorders, distinguishing them from individuals engaged in normative dieting or non-pathological body dissatisfaction.

In clinical practice, the EDI serves as a comprehensive diagnostic adjunct, treatment-planning battery, and outcome-monitoring instrument. By profiling an individual across eight distinct subscales, clinicians can identify specific targets for therapeutic intervention. For instance, a patient presenting with marked elevations in Interoceptive Awareness and Ineffectiveness may require intensive emotion-regulation training and schema-focused cognitive restructuring, whereas a patient with isolated elevations in Drive for Thinness and Perfectionism may benefit most from exposure-based behavioral experiments and perfectionism-focused cognitive behavioral therapy (CBT-E). In research domains, the EDI provides an empirically validated standard for subtyping clinical cohorts, assessing treatment efficacy in randomized controlled trials, and exploring developmental etiology in longitudinal epidemiological studies.

Psychological Construct

The EDI operationalizes eating pathology as a multidimensional construct comprising eight discrete, theoretically informed dimensions. Each subscale captures a distinct psychological or behavioral facet, enabling an integrated evaluation of the patient’s cognitive landscape:

1. Drive for Thinness (DT — 8 items)

This subscale assesses an excessive, relentless preoccupation with weight reduction, extreme pursuit of thinness, intense fear of weight gain, and rigid dietary restriction. High scorers exhibit profound distress regarding minor fluctuations in weight and harbor an unyielding obsession with dieting. Exemplar items include: “I am terrified of gaining weight” (Item 16) and “I am preoccupied with the desire to be thinner” (Item 32). This dimension represents the cardinal cognitive hallmark of both restrictive and bulimic spectrum disorders.

2. Bulimia (B — 7 items)

The Bulimia subscale measures the tendency toward episodes of uncontrollable overeating (binge eating), secret consumption of food, emotional eating, and thoughts or impulses to engage in self-induced vomiting as a compensatory mechanism. Rather than assessing objective caloric intake alone, it gauges the subjective loss of control and subsequent affective distress. Exemplar items include: “I have gone on eating binges where I have felt that I could not stop” (Item 28) and “I have the thought of trying to vomit in order to lose weight” (Item 53).

3. Body Dissatisfaction (BD — 9 items)

Body Dissatisfaction quantifies subjective discontent with the overall shape and size of specific body regions most susceptible to post-pubertal adiposity, particularly the stomach, hips, thighs, and buttocks. This subscale captures the somatic dissatisfaction that precipitates and sustains pathological weight-control practices. Exemplar items include: “I think that my stomach is too big” (Item 2) and “I think my hips are too big” (Item 45).

4. Ineffectiveness (I — 10 items)

Ineffectiveness evaluates pervasive feelings of personal inadequacy, worthlessness, low self-esteem, emptiness, and an overarching belief that one lacks control over one’s destiny. Derived directly from Hilde Bruch’s clinical characterization of anorexia nervosa, this construct reflects an ego-deficit wherein dietary restriction or compensatory behaviors serve as desperate attempts to establish an artificial sense of mastery. Exemplar items include: “I feel ineffective as a person” (Item 10) and “I feel empty inside (emotionally)” (Item 56).

5. Perfectionism (P — 6 items)

This dimension measures the cognitive conviction that only flawless, superior personal performance is acceptable, alongside the perception that parents or authority figures demand perfection. It taps into both self-oriented perfectionism and socially prescribed perfectionistic standards. Exemplar items include: “Only outstanding performance is good enough in my family” (Item 13) and “I feel that I must do things perfectly or not do them at all” (Item 52).

6. Interpersonal Distrust (ID — 7 items)

Interpersonal Distrust assesses an individual’s reluctance to form intimate relationships, an inability to express emotional states to others, and a general suspicion regarding the motives of others. Individuals scoring high on this scale experience emotional detachment and social alienation, which severely hampers therapeutic alliance formation and interpersonal functioning. Exemplar items include: “I need to keep people at a certain distance” (Item 54) and “I have trouble expressing my emotions to others” (Item 34).

7. Interoceptive Awareness (IA — 7 items)

Interoceptive Awareness captures the capacity to accurately identify, label, and respond to internal visceral sensations, distinguishing between primary emotional arousal states and physiological hunger or satiety cues. Marked deficits in interoceptive awareness manifest as profound alexithymia and confusion regarding somatic vs. affective states, often leading individuals to consume food when emotionally distressed. Exemplar items include: “I get confused about what emotion I am feeling” (Item 21) and “I get confused as to whether or not I am hungry” (Item 40).

8. Maturity Fears (MF — 7 items)

Maturity Fears evaluates the psychological yearning to retreat into the perceived security of childhood and an active dread of adult physical, psychological, and social demands. Garner and colleagues posited that adolescent-onset anorexia nervosa often functions as an unconscious defensive strategy to arrest pubertal maturation and avoid psychosexual development. Exemplar items include: “I wish that I could return to the security of childhood” (Item 3) and “The demands of adulthood are too great” (Item 35).

Theoretical Framework

The conceptual architecture of the EDI is anchored in the integration of psychodynamic ego psychology, cognitive-behavioral theory, and developmental psychopathology. Most notably, the instrument operationalizes the foundational clinical insights of psychoanalyst Hilde Bruch (1973), who postulated that the central pathognomonic feature of anorexia nervosa is not food avoidance per se, but rather an underlying disturbance in self-concept characterized by three cardinal deficits:

• Severe perceptual distortion: An inability to accurately perceive one’s body size, resulting in relentless body dissatisfaction and pursuit of emaciation.
• Interoceptive deficits: A constitutional or developmental incapacity to correctly recognize visceral bodily stimuli, including hunger, satiety, and visceral emotional experiences.
• Paralyzing sense of ineffectiveness: An overarching experience of oneself as completely powerless, passive, and reactive to environmental demands, leading the individual to utilize strict bodily control as the sole arena in which autonomy can be exercised.

Garner, Olmsted, and Polivy synthesized Bruch’s developmental ego-deficit framework with contemporary cognitive-behavioral principles formulated by Aaron T. Beck and Christopher Fairburn. Under this cognitive-behavioral paradigm, dysfunctional core beliefs—such as the equation of personal worth with thinness and perfection—interact with interoceptive confusion and social alienation to trigger self-reinforcing vicious cycles. Dietary restriction provides temporary relief from feelings of ineffectiveness; however, severe restriction intensifies neurobiological and psychological starvation effects, triggering bulimic episodes, which subsequently heighten guilt, body dissatisfaction, and perceived loss of mastery.

Furthermore, the EDI incorporates developmental theories of adolescent transition. Puberty presents unprecedented psychological, interpersonal, and biological shifts. For adolescents predisposed to high perfectionism and low self-efficacy, the physiological onset of female secondary sex characteristics (specifically adiposity in the hips, thighs, and abdomen) is perceived as an intolerable crisis. The Maturity Fears construct reflects this developmental conflict, framing the starvation response as an attempt to biologically reverse somatic maturation and psychologically evade adult autonomy.

Validity

Extensive psychometric investigations conducted across four decades have established the rigorous validity of the Eating Disorders Inventory across diverse clinical and non-clinical populations:

Construct and Criterion-Related Validity

In the seminal validation study by Garner, Olmsted, and Polivy (1983), the EDI was administered to female patients diagnosed with anorexia nervosa (N = 113), cross-validated in a second clinical sample (N = 146), and contrasted with a large normative control group of female university students (N = 577). The results demonstrated profound discriminant criterion validity: patients with anorexia nervosa scored significantly higher than non-clinical female controls across all eight subscales (all p < .001). Specifically, Drive for Thinness yielded mean scores of 14.8 (±5.7) for patients versus 5.1 (±5.7) for controls, and Ineffectiveness yielded 12.4 (±7.5) versus 2.3 (±3.6), representing large effect sizes (Cohen’s d > 1.50).

Subgroup Discrimination

The EDI successfully discriminates between clinical subtypes of eating pathology. Patients with bulimia nervosa and the binge-eating/purging subtype of anorexia nervosa score markedly higher on the Bulimia subscale compared to restricting-type anorexia nervosa patients (mean B score of 10.8 vs. 1.2, p < .001). Conversely, the Ineffectiveness and Perfectionism subscales remain elevated across both restricting and bulimic presentations, confirming that these personality dimensions represent shared transdiagnostic vulnerabilities.

Convergent and Discriminant Validity

Convergent validity has been established through robust correlations with established psychometric measures. The eating-specific subscales correlate strongly with the Eating Attitudes Test (EAT-26) (r = .68 to .82). Ineffectiveness demonstrates high convergent validity with the Beck Depression Inventory (BDI) (r = .65 to .75) and negative correlations with the Rosenberg Self-Esteem Scale (r = -.70). Interoceptive Awareness correlates strongly with measures of alexithymia, such as the Toronto Alexithymia Scale (TAS-20; r = .58–.67). Discriminant validity is evidenced by weak associations between EDI subscales and measures of general non-pathological intelligence or unrelated personality traits (e.g., extraversion, spatial reasoning; r < .15).

Reliability

The reliability of the EDI has been extensively documented in clinical cohorts, university samples, and epidemiological surveys worldwide:

Internal Consistency

Internal consistency estimates (Cronbach’s alpha) across the eight subscales demonstrate exceptional reliability in clinical samples. In Garner et al.’s (1983) original validation study, the alpha coefficients for clinical anorexia nervosa patients were:

• Drive for Thinness: α = .85
• Bulimia: α = .90
• Body Dissatisfaction: α = .91
• Ineffectiveness: α = .90
• Perfectionism: α = .82
• Interpersonal Distrust: α = .82
• Interoceptive Awareness: α = .83
• Maturity Fears: α = .83

In non-clinical populations, internal consistency remains high to acceptable, typically ranging from α = .80 to .92 for the eating-specific subscales (DT, B, BD) and α = .68 to .84 for the general psychological subscales (P, ID, IA, MF). In the Portuguese validation by Machado, Gonçalves, Martins, and Soares (2001), Cronbach’s alpha values ranged from α = .71 (Maturity Fears) to α = .91 (Body Dissatisfaction), corroborating cross-cultural structural consistency.

Test-Retest Reliability and Stability

Studies evaluating temporal stability demonstrate substantial test-retest reliability across clinical and non-clinical samples. Over a 1- to 3-week test-retest interval, Pearson correlation coefficients routinely range between r = .75 and r = .94 across all subscales. In longitudinal cohorts monitored over 1-year follow-up periods, the trait-like subscales (Perfectionism, Ineffectiveness, Maturity Fears) demonstrate higher baseline-to-follow-up stability (r = .65–.78) compared to symptom-state subscales (Drive for Thinness, Bulimia), which show greater responsiveness to clinical intervention, reflecting sensitivity to therapeutic change.

Factor Analysis

The structural validity of the EDI has been scrutinized through extensive exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) across numerous international investigations.

Exploratory Factor Structure

In the original scale development study, Garner et al. (1983) performed principal components analysis with varimax rotation on clinical and non-clinical cohorts. The analyses confirmed an eight-factor solution that accounted for the majority of the common variance, closely mirroring the theoretical clinical constructs. All 64 items demonstrated dominant primary factor loadings (≥ .40) on their designated conceptual dimensions, with minimal cross-loadings (≤ .25), demonstrating clear simple structure.

Confirmatory Factor Analysis (CFA) and Model Fit

Subsequent confirmatory psychometric evaluations have rigorously tested the fit of the theoretical eight-factor model. While classical independent clusters models sometimes encounter moderate fit indices when evaluating all 64 items simultaneously—owing to the complex semantic interplay between Ineffectiveness, Low Self-Esteem, and Interoceptive deficits—refined multi-group CFA studies demonstrate acceptable to excellent fit indices when residual covariances between semantically adjacent items are accounted for:

• Comparative Fit Index (CFI): .90 – .94
• Tucker-Lewis Index (TLI): .89 – .93
• Root Mean Square Error of Approximation (RMSEA): .045 – .058 (90% CI [.042, .061])
• Standardized Root Mean Square Residual (SRMR): .051 – .064

Furthermore, second-order factor analytic investigations consistently extract two higher-order latent dimensions: a Pathological Eating / Weight Preoccupation factor (subsuming Drive for Thinness, Bulimia, and Body Dissatisfaction) and a General Psychological Maladjustment / Ego-Deficit factor (subsuming Ineffectiveness, Perfectionism, Interpersonal Distrust, Interoceptive Awareness, and Maturity Fears). This hierarchical architecture provides empirical justification for examining both granular subscale profiles and broad composite indicators.

Instrument / Measurement Tool

The operational characteristics and administration guidelines of the original Eating Disorders Inventory are detailed below:

• Instrument Name: Eating Disorders Inventory (EDI)
• Authors: David M. Garner, Ph.D., Marion P. Olmsted, Ph.D., and Janet Polivy, Ph.D. (1983)
• Assessment Type: Multidimensional self-report psychological inventory
• Target Population: Adolescents and adults (ages 12 years and older) suspected of having eating disorders, as well as non-clinical populations screened for disordered eating patterns
• Administration Format: Individual or group pencil-and-paper or computerized administration; completion time is typically 15 to 25 minutes
• Total Item Count: 64 items
• Subscale Structure (8 Dimensions):
  • Drive for Thinness (DT): 8 items (1, 7, 11, 16, 25, 32, 49, 64)
  • Bulimia (B): 7 items (4, 5, 28, 38, 46, 53, 61)
  • Body Dissatisfaction (BD): 9 items (2, 9, 12, 19, 31, 45, 55, 59, 62)
  • Ineffectiveness (I): 10 items (10, 18, 20, 24, 27, 37, 41, 42, 50, 56)
  • Perfectionism (P): 6 items (13, 29, 36, 43, 52, 63)
  • Interpersonal Distrust (ID): 7 items (15, 17, 23, 30, 34, 54, 57)
  • Interoceptive Awareness (IA): 7 items (8, 21, 26, 33, 40, 51, 60)
  • Maturity Fears (MF): 7 items (3, 6, 14, 22, 39, 48, 58)
• Raw Response Format: 6-point forced-choice Likert scale:
  • 1 = Always
  • 2 = Usually
  • 3 = Often
  • 4 = Sometimes
  • 5 = Rarely
  • 6 = Never
• Scoring Protocols:
  • Standard Clinical Scoring (0–3 point method): Developed by Garner et al. to eliminate mild, normative body dissatisfaction and isolate severe, clinically significant pathology. For positively keyed (pathological) items: “Always” = 3, “Usually” = 2, “Often” = 1, and the three non-pathological options (“Sometimes”, “Rarely”, “Never”) = 0. For negatively keyed (reversed) items: “Never” = 3, “Rarely” = 2, “Often” = 1, and “Sometimes”, “Usually”, “Always” = 0. Subscale scores are calculated by summing the recoded item points.
  • Continuous Linear Scoring (1–6 point method): Commonly employed in non-clinical, epidemiological, and college samples to preserve variance, mitigate floor effects, and ensure statistical normality for correlational and structural equation modeling. In this system, raw responses are scored 1 to 6 (with reversed items scored 6 to 1).
• Reverse-Keyed Items: Items 1, 12, 15, 17, 19, 20, 22, 23, 26, 30, 31, 37, 39, 42, 50, 55, 57, 58, 62 reflect psychological health or low pathology and must be inverted prior to subscale summation.

Permissions & Fee and Test Year

The original Eating Disorders Inventory was first published in 1983 in the International Journal of Eating Disorders by David M. Garner, Marion P. Olmsted, and Janet Polivy. Subsequent expanded revisions, including the Eating Disorder Inventory-2 (EDI-2, 1991) and Eating Disorder Inventory-3 (EDI-3, 2004), were authored by David M. Garner and are commercially published and distributed exclusively by Psychological Assessment Resources (PAR, Inc.).

The official, standardized testing kits, comprehensive administration manuals, normed scoring sheets, and software scoring keys are proprietary instruments subject to commercial purchase fees and strict clinical user qualification requirements (typically Qualification Level B or C, requiring advanced graduate training in psychology, psychiatry, or allied mental health disciplines). However, the original 1983 64-item questionnaire text and various translated adaptations (such as the Portuguese version validated by Machado et al., 2001) have been published in peer-reviewed scientific journals and institutional repositories for academic, non-commercial research, and educational purposes. Researchers intending to utilize the EDI in clinical diagnostic settings, clinical trials, or commercial applications should contact PAR, Inc. to secure appropriate licensing and copyrighted administration materials.

References

• Bruch, H. (1973). Eating disorders: Obesity, anorexia nervosa, and the person within. Basic Books. https://doi.org/10.1080/00029157.1974.10403713
• Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitive behaviour therapy for eating disorders: A “transdiagnostic” theory and treatment strategy. Behaviour Research and Therapy, 41(5), 509–528. https://doi.org/10.1016/S0005-7967(02)00088-8
• Garner, D. M. (1990). Eating Disorder Inventory – 2: Professional manual. Western Psychological Services.
• Garner, D. M. (1991). Eating Disorder Inventory – 2: Professional manual. Psychological Assessment Resources.
• Garner, D. M. (2004). Eating Disorder Inventory – 3: Professional manual. Psychological Assessment Resources.
• Garner, D. M., Olmsted, M. P., & Polivy, J. (1983). Development and validation of a multidimensional Eating Disorder Inventory for anorexia nervosa and bulimia. International Journal of Eating Disorders, 2(2), 15–34. https://doi.org/10.1002/1098-108X(198321)2:2<15::AID-EAT2260020203>3.0.CO;2-6
• Garner, D. M., Olmstead, M. P., Bohr, Y., & Garfinkel, P. E. (1982). The Eating Attitudes Test: Psychometric features and clinical correlates. Psychological Medicine, 12(4), 871–878. https://doi.org/10.1017/S0033291700049163
• Machado, P. P. P., Gonçalves, S., Martins, C., & Soares, I. C. (2001). The Portuguese version of the Eating Disorders Inventory: Evaluation of its psychometric properties. European Eating Disorders Review, 9(1), 43–52. https://doi.org/10.1002/erv.372
• Schoemaker, C., van Strien, T., & van der Staak, C. (1994). Validation of the Eating Disorders Inventory in a nonclinical population using transformed and untransformed responses. International Journal of Eating Disorders, 15(4), 387–393. https://doi.org/10.1002/1098-108X(199405)15:4<387::AID-EAT2260150409>3.0.CO;2-P
• Wear, P. V., & Pratz, O. (1987). Test-retest reliability for the Eating Disorder Inventory. International Journal of Eating Disorders, 6(6), 767–769. https://doi.org/10.1002/1098-108X(198711)6:6<767::AID-EAT2260060611>3.0.CO;2-V

Items of the Scale