Children’s Eating Attitudes Test (ChEAT)

Abstract

The Children’s Eating Attitudes Test (ChEAT) is a standardized, self-report psychometric instrument designed to evaluate abnormal eating attitudes, weight preoccupation, body dissatisfaction, and disordered eating behaviors in preadolescent and early adolescent populations aged 8 to 15 years. Adapted by Michael J. Maloney and colleagues (1988, 1989) from the adult Eating Attitudes Test (EAT-26) originally established by David M. Garner and Paul E. Garfinkel, the ChEAT modifies complex syntax and clinical terminology to ensure developmental appropriateness, cognitive accessibility, and linguistic comprehension for younger youth. Comprising 26 items, the scale evaluates core transdiagnostic features across four widely recognized factor-analytically derived dimensions: Dieting, Restricting and Purging, Food Preoccupation, and Oral Control. Responses are recorded on a 6-point Likert-type frequency scale and converted into a standard 0–3 scoring matrix (Always = 3, Very Often = 2, Often = 1, Sometimes = 0, Rarely = 0, Never = 0; with Item 19 reverse-scored), yielding a maximum score of 78 points. A diagnostic screening threshold score of 20 or greater serves as an empirically validated cutoff indicating significant maladaptive eating attitudes and heightened risk for clinical eating disorders such as anorexia nervosa and bulimia nervosa. Psychometric evaluations demonstrate robust internal consistency (Cronbach’s alpha typically ranging from .76 to .90 across community and clinical samples), satisfactory test-retest reliability (r = .81 over short intervals), and substantial convergent validity with related constructs such as childhood depression, anxiety, perfectionism, and body image disturbance. The ChEAT remains one of the most widely employed, cross-culturally validated screening instruments in pediatric psychology, epidemiological surveillance, school-based mental health initiatives, and clinical eating disorder research worldwide.

Keywords

Children’s Eating Attitudes Test, ChEAT, eating disorders, pediatric assessment, anorexia nervosa, bulimia nervosa, body dissatisfaction, dieting behavior, psychometrics, child psychopathology, screening instrument, factor analysis

Authors

The Children’s Eating Attitudes Test was developed and psychometrically validated by an interdisciplinary team of pediatric psychiatrists, clinical researchers, and biostatisticians:

• Michael J. Maloney, M.D. — Professor of Child and Adolescent Psychiatry, Division of Child and Adolescent Psychiatry, Cincinnati Children’s Hospital Medical Center and the University of Cincinnati College of Medicine, Cincinnati, Ohio, USA. Dr. Maloney served as the principal investigator in adapting the adult EAT-26 for pediatric and school-aged populations.
• Julie B. McGuire, M.S.W., L.I.S.W. — Clinical Social Worker and Research Specialist, Department of Psychiatry, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, USA.
• Stephen R. Daniels, M.D., Ph.D. — Pediatrician and Epidemiologist, Department of Pediatrics, University of Cincinnati College of Medicine and Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, USA.
• Bonny L. Specker, Ph.D. — Biostatistician and Epidemiologist, Clinical Research Center, Children’s Hospital Medical Center, Cincinnati, Ohio, USA.

Subsequent psychometric refinement and factor-analytic structural validation were prominently advanced by Linda Smolak, Ph.D. (Department of Psychology, Kenyon College, Gambier, Ohio) and Michael P. Levine, Ph.D. (Department of Psychology, Kenyon College), who provided foundational structural validity benchmarks across middle school and preadolescent student cohorts.

Purpose

The Children’s Eating Attitudes Test was constructed to fill a critical diagnostic and epidemiological void in pediatric mental health. Prior to the late 1980s, the vast majority of psychometric assessment tools for disordered eating were normed and calibrated exclusively on late adolescent and adult clinical cohorts. Consequently, clinicians and developmental researchers possessed few validated, age-tailored instruments capable of identifying disordered eating attitudes, emergent food phobias, and subclinical eating pathologies in middle childhood and early adolescence.

During the developmental transition spanning ages 8 through 15, children experience marked somatic, hormonal, and cognitive shifts, including pubertal weight gain, body fat redistribution, and the cognitive emergence of formal operational reasoning. Concurrently, children become increasingly susceptible to sociocultural pressures, peer teasing, parental modeling of dieting, and pervasive media depictions of the thin ideal. Because children often lack the metacognitive sophistication to articulate abstract psychopathological constructs (e.g., “morbid fear of fatness” or “loss of control over consumption”), adult scales frequently introduce measurement artifacts, misinterpretations, and severe construct underreporting.

The primary clinical and research purposes of the ChEAT encompass:

• Universal and Targeted Screening: Operating as a frontline screening tool within pediatric primary care settings, pediatric endocrinology clinics, school mental health screenings, and community health surveys to identify children exhibiting prodromal or clinical eating disorder symptomatology who require immediate secondary diagnostic evaluation.
• Epidemiological Surveillance: Quantifying the prevalence, onset trajectory, gender distribution, and demographic correlates of body dissatisfaction, caloric restriction, and binge-purge inclinations in nonclinical, community-based youth cohorts.
• Treatment Outcome Monitoring: Tracking symptom mitigation and psychological shifts in young patients undergoing specialized clinical interventions—such as Family-Based Treatment (FBT) or developmentally adapted Cognitive Behavioral Therapy for Eating Disorders (CBT-ED).
• Preventive Intervention Evaluation: Measuring changes in cognitive vulnerability before and after universal or selective body-image and media-literacy prevention curricula in school systems.

The clinical rationale underlying the ChEAT is grounded in evidence demonstrating that the early onset of disordered eating in preadolescence carries profound risks of physiological compromise, including stunted linear growth, impaired bone mineral accrual, neurodevelopmental interruptions, and chronic psychiatric comorbidity. Early, accurate detection using a developmentally sensitive measure markedly optimizes long-term prognostic trajectories.

Psychological Construct

The ChEAT assesses the multidimensional construct of eating disorder psychopathology in youth, defined as an interrelated constellation of cognitive distortions, affective vulnerabilities, and behavioral patterns centered on body shape, body weight, caloric intake, and eating behavior. Rather than evaluating a monolithic entity, the instrument captures both internalizing cognitive dynamics (such as obsessive preoccupation with thinness and acute guilt following nutrient ingestion) and externalizing or compensatory motoric behaviors (such as self-induced emesis, rigorous carbohydrate restriction, and frantic caloric expenditure).

Extensive factor-analytic investigations—most notably by Smolak and Levine (1994)—delineate four distinct psychological sub-dimensions measured by the instrument:

1. Dieting

This dimension reflects an explicit cognitive drive for thinness, active cognitive restriction, and conscious dietary restraint. Children scoring high on this dimension exhibit an exaggerated consciousness regarding dietary composition, hyper-vigilance over caloric values, avoidance of energy-dense foods, and intentional skipping of meals despite somatic sensations of hunger. Psychologically, this construct encapsulates the cognitive internalization of the thin ideal, characterized by persistent fears of becoming overweight, desires to be significantly thinner, and distress upon consuming sugars or carbohydrates. Representative manifestations include actively purchasing or demanding “diet foods,” exercising primarily to incinerate calories, and pervasive feelings of guilt following food consumption.

2. Restricting and Purging

The Restricting and Purging dimension captures severe behavioral manifestations of compensatory pathology and extreme somatic regulation. It measures overt behaviors aimed at preventing weight gain or ridding the body of consumed calories. Items loading on this factor reflect pathological somatic practices, including self-induced vomiting (emesis), the subjective urge to purge after eating, deliberate fasting, prolonged food avoidance, and the sensation of wanting one’s stomach to remain completely empty. Elevated scores on this factor indicate advanced psychopathology, signifying that cognitive body dissatisfaction has transitioned into active, hazardous physiological disruption.

3. Food Preoccupation

This subscale measures the cognitive intrusion and attentional bias associated with nutrition, hunger, and eating. Children with high food preoccupation experience recurrent, uncontrollable thoughts regarding meals, binge-eating urges accompanied by a perceived lack of behavioral control, and the feeling that food dominates or controls their daily cognitive functioning. This dimension directly taps the cognitive dissonance between severe caloric restriction and biological hunger drives, often mirroring the cognitive obsession with food historically documented in physiological starvation states.

4. Oral Control

Oral Control assesses behaviors reflecting rigid environmental, behavioral, and interpersonal management of eating rituals. This construct involves obsessive food manipulation, prolonged mealtime consumption (e.g., cutting food into microscopic pieces, chewing excessively), and perceived interpersonal conflict or pressure from parents and caregivers regarding eating. Elevated scores frequently indicate resistance to parental feeding efforts, subjective feelings that external figures are coercing food consumption, and hyper-controlled behavioral pacing during meals designed to mask dietary restriction.

Theoretical Framework

The conceptual architecture of the ChEAT is anchored in the convergence of the Transdiagnostic Cognitive Behavioral Model of Eating Disorders (Fairburn et al.), Developmental Psychopathology, and the Sociocultural Tripartite Influence Model (Thompson et al.).

Transdiagnostic Cognitive Behavioral Model

According to Fairburn’s cognitive-behavioral formulation, the core psychopathological engine driving eating disorders is the over-evaluation of eating, shape, and weight. While healthy individuals evaluate their self-worth across diverse psychosocial domains (such as academic mastery, interpersonal relationships, athletic skills, and artistic pursuits), individuals with eating disorders judge their personal value almost exclusively in terms of their body weight, shape, and capacity to maintain strict dietary control. The ChEAT operationalizes this core cognitive psychopathology for children, capturing how cognitive over-evaluation translates directly into dietary restraint, caloric avoidance, somatic checking behaviors, and compensatory purging.

Developmental Psychopathology Perspective

From a developmental psychopathology perspective (Cicchetti & Cohen), disordered eating in middle childhood represents a maladaptive developmental pathway emerging from transactional interactions between biological vulnerabilities, emerging cognitive structures, and family environment. During the preadolescent transition, children are developing operational autonomy and bodily awareness. When biological pubertal changes occur concurrently with heightened stress, emotional dysregulation, or familial conflict, bodily control often becomes a primary conduit for establishing personal agency. The ChEAT specifically targets this developmental interface by measuring both internal cognitive distress (fear of fatness, guilt) and systemic interpersonal dynamics (feeling pressured by parents to consume food).

Sociocultural Tripartite Influence Model

The Tripartite Influence Model posits that sociocultural pressures emanating from three primary socialization agents—parents, peers, and mass/digital media—foster the internalization of unrealistic appearance standards and promote appearance-based social comparison. In children as young as 8 to 11 years, peer teasing regarding weight, parental dietary modeling, and exposure to idealized body imagery accelerate the internalization of the “thin ideal.” This internalization produces body dissatisfaction, which subsequently triggers restrictive dieting and compensatory behaviors. The ChEAT’s items explicitly measure these sociocultural consequences, evaluating items such as fear of being overweight, awareness of caloric content, and behavioral dieting rituals.

Validity

The psychometric validity of the ChEAT has been rigorously investigated across clinical, non-clinical, community, and cross-cultural cohorts since its inception in 1988.

Construct and Structural Validity

Construct validity was initially established by Maloney et al. (1988, 1989) across 318 public school students in grades 3 through 6. The instrument successfully identified that eating concerns and dieting behaviors were not confined to clinical or post-pubertal adolescents, demonstrating that nearly 45% of surveyed children wanted to be thinner and 37% had already attempted caloric restriction. Smolak and Levine (1994) confirmed the construct validity of the ChEAT in an independent cohort of 308 middle school girls, establishing that the scale distinctly operationalizes multidimensional facets of disordered eating through clear factor structures that cleanly separate cognitive restraint from behavioral purging.

Convergent and Concurrent Validity

The ChEAT demonstrates statistically robust convergent validity with numerous validated pediatric psychiatric measures:

• Body Image Disturbance: ChEAT scores correlate strongly with visual figure rating scales and the Body Shape Questionnaire, showing correlation coefficients ranging from r = .55 to r = .72, indicating that children scoring higher on the ChEAT exhibit pronounced perceptual and affective body dissatisfaction.
• Affective Psychopathology: Moderate to strong positive correlations exist between ChEAT total scores and measures of childhood depression, such as the Children’s Depression Inventory (CDI; r = .38 to .52), as well as generalized childhood anxiety inventories (r = .34 to .46).
• General Eating Disorder Inventories: High concurrent validity is documented between the ChEAT and pediatric adaptations of the Eating Disorder Examination-Questionnaire (ChEDE-Q), yielding convergent correlation coefficients exceeding r = .70.

Discriminant and Known-Groups Validity

The ChEAT exhibits strong known-groups discriminant validity, reliably differentiating between children diagnosed with clinical eating disorders (such as anorexia nervosa or ARFID with weight concern) and age-matched healthy control subjects. Clinical cohorts score substantially above the established threshold of 20, with mean scores often exceeding 32.5 (SD = 11.2), whereas nonclinical community school populations demonstrate mean scores between 6.5 and 9.8 (SD = 6.4). Furthermore, the scale effectively discriminates between children participating in weight-sensitive aesthetic or weight-class activities (e.g., competitive gymnastics, ballet, wrestling) and non-athletic controls.

Predictive Validity

Longitudinal prospective cohort studies have demonstrated that elevated baseline ChEAT scores in preadolescent children (ages 9–11) significantly predict the subsequent onset of partial or full-syndrome eating disorders, excessive compensatory exercise, and extreme unhealthful weight control behaviors during middle to late adolescence (ages 14–17), establishing its utility as an early predictive prognostic marker.

Reliability

The reliability of the ChEAT has been substantiated across diverse demographic segments, grade levels, and international adaptations.

Internal Consistency

In initial validation cohorts by Maloney et al. (1988), the full 26-item scale demonstrated high internal consistency, with a standardized Cronbach’s alpha of .76 in elementary school children. Subsequent investigations by Smolak and Levine (1994) documented an overall Cronbach’s alpha of .87 among middle school girls. Across widespread global validations, internal consistency coefficients for the ChEAT total score routinely fall between .83 and .91, reflecting high item homogeneity.

Internal consistency metrics for the individual subscales, as derived from community samples of girls and boys, typically demonstrate:

• Dieting: Cronbach’s α = .81 to .88 (reflecting high internal reliability due to strong item covariance around weight and calorie restriction)
• Restricting and Purging: Cronbach’s α = .68 to .78 (satisfactory, considering the low base rate of severe purging behaviors in community youth)
• Food Preoccupation: Cronbach’s α = .65 to .75
• Oral Control: Cronbach’s α = .60 to .71 (lower reliability frequently attributable to the inclusion of reverse-scored or socially driven items)

Test-Retest Stability

Temporal stability assessments reveal acceptable to high test-retest reliability across short- and medium-term evaluation intervals:

• Maloney et al. (1988) administered the ChEAT twice over a 3-week interval to a pediatric cohort, yielding an overall test-retest correlation coefficient of r = .81 (p < .001), indicating robust measurement stability in young children.
• Smolak and Levine (1994) documented a test-retest coefficient of r = .74 across a 6-month longitudinal follow-up, confirming that while developmental fluctuation occurs, relative rank-order stability remains moderately durable over time.

Factor Analysis

The latent dimensionality of the ChEAT has been extensively evaluated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) to determine whether the original three-factor model of the adult EAT-26 or novel pediatric configurations best represent the data.

Foundational Exploratory Factor Analyses

In the seminal psychometric re-evaluation by Smolak and Levine (1994), principal components analysis with varimax and oblimin rotations on 308 middle school females identified a robust four-factor structure accounting for approximately 42.4% of the total variance:

• Factor 1: Dieting (accounting for 22.8% of variance; item loadings ranging from .42 to .78 on items such as 1, 2, 6, 10, 11, 12, 14, 16, 17, 22, 23, and 24). This factor captures conscious caloric avoidance, body fat preoccupation, and postprandial guilt.
• Factor 2: Restricting and Purging (accounting for 7.9% of variance; items 2, 7, 9, 10, 17, 22, 23, 24, and 26). Cross-loading on dietary restraint items occurs because behavioral purging is inextricably linked with chronic caloric restriction in youth.
• Factor 3: Food Preoccupation (accounting for 6.2% of variance; items 3, 4, 18, 20, and 21). This factor is characterized by intrusive ruminations regarding eating and subjective loss of control.
• Factor 4: Oral Control (accounting for 5.5% of variance; items 8, 13, 15, and 20). Reflecting externalized pressures and mealtime behavioral rituals.

Confirmatory Factor Analysis and Alternative Models

Subsequent confirmatory factor analyses across culturally diverse populations (e.g., Spanish, Brazilian, Turkish, and Australian cohorts) have compared single-factor, three-factor (mirroring adult EAT-26: Dieting, Bulimia/Food Preoccupation, Oral Control), and four-factor specifications. Generally, the four-factor and modified bifactor models show superior goodness-of-fit indices:

• Comparative Fit Index (CFI): Values typically range between .90 and .95 in adequately powered adolescent samples.
• Root Mean Square Error of Approximation (RMSEA): Values typically range between .042 and .062, indicative of acceptable to close fit.
• Standardized Root Mean Square Residual (SRMR): Consistently below .06.

Measurement invariance analyses across biological sex demonstrate metric and scalar invariance between boys and girls, validating the use of the ChEAT for cross-gender epidemiological comparisons, although boys systematically endorse lower absolute latent mean levels of Dieting and Purging than girls.

Instrument / Measurement Tool

The structured parameters of the Children’s Eating Attitudes Test are summarized below:

• Test Type: Standardized self-report screening questionnaire (can be administered via oral proctoring/interviewer reading for younger children aged 8–10 to ensure reading comprehension).
• Target Population: Children and early adolescents aged 8 to 15 years (elementary through middle school).
• Total Item Count: 26 items.
• Response Format: 6-point Likert-type frequency scale consisting of: Always, Very Often, Often, Sometimes, Rarely, Never.
• Scoring Paradigm:
  • Standard scoring converts responses into a 4-level categorical weight: Always = 3, Very Often = 2, Often = 1, Sometimes = 0, Rarely = 0, Never = 0.
  • Reverse-Scored Item: Item 19 (“I can show self-control around food”) is keyed inversely: Never = 3, Rarely = 2, Sometimes = 1, Often = 0, Very Often = 0, Always = 0.
  • Note on Item 25: In standard validation literature and scoring protocols, non-symptomatic engagement with food is monitored, but Item 19 remains the primary reverse-scored item in the total score computation.
• Total Score Range: 0 to 78 points.
• Clinical Cutoff Score: A cumulative score of 20 or higher indicates significant disturbed eating attitudes, marked weight preoccupation, and the clinical need for comprehensive psychiatric and physical diagnostic assessment. (A score of 20 corresponds to a mean item rating of ~0.77).
• Subscale Composition (Smolak & Levine, 1994):
  • Dieting: Items 1, 2, 6, 10, 11, 12, 14, 16, 17, 22, 23, 24
  • Restricting and Purging: Items 2, 7, 9, 10, 17, 22, 23, 24, 26
  • Food Preoccupation: Items 3, 4, 18, 20, 21
  • Oral Control: Items 8, 13, 15, 20
• Administration Time: Approximately 10 to 15 minutes.

Permissions & Fee and Test Year

The Children’s Eating Attitudes Test was originally published in 1988 (preliminary reliability testing) and fully operationalized in 1989 by Michael J. Maloney and colleagues. In accordance with the precedent established by David Garner for the adult EAT-26, the ChEAT was placed into the public domain for academic, non-commercial clinical, and scientific research endeavors.

• Permissions: The scale may be utilized, reproduced, and translated without royalty fees by qualified mental health clinicians, academic researchers, and healthcare professionals for non-commercial research and diagnostic screening purposes.
• Commercial Use: Any incorporation into proprietary commercial diagnostic software, fee-for-service digital health platforms, or published books/test batteries requires formal written authorization from the primary authors and copyright holders.
• Citation Requirement: Users are ethically and academically obligated to cite the foundational validation publications (Maloney et al., 1988, 1989; Smolak & Levine, 1994) in any resulting scientific papers, reports, or clinical documentation.

References

• Garner, D. M., & Garfinkel, P. E. (1979). The Eating Attitudes Test: An index of the symptoms of anorexia nervosa. Psychological Medicine, 9(2), 273–279. https://doi.org/10.1017/s0033291700030762
• Garner, D. M., Olmsted, 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
• Maloney, M. J., McGuire, J. B., & Daniels, S. R. (1988). Reliability testing of a children’s version of the Eating Attitude Test. Journal of the American Academy of Child & Adolescent Psychiatry, 27(5), 541–543. https://doi.org/10.1097/00004583-198809000-00004
• Maloney, M. J., McGuire, J., Daniels, S. R., & Specker, B. (1989). Dieting behavior and eating attitudes in children. Pediatrics, 84(3), 482–489. https://doi.org/10.1542/peds.84.3.482
• Smolak, L., & Levine, M. P. (1994). Psychometric properties of the Children’s Eating Attitudes Test. International Journal of Eating Disorders, 16(3), 275–282. Below are the authentic scale items in their original language as published in the standard psychometric validation studies, without modification or translation to preserve instrument validity and reliability: