EAT-26 (Eating Attitudes Test): Clinical Screening Tool for Eating Disorders

The Eating Attitudes Test-26 (EAT-26) is one of the most widely used standardized self-report screening tools for identifying attitudes, behaviors, and concerns characteristic of eating disorders. It was developed by David M. Garner and colleagues in 1982 as a shortened version of the original 40-item Eating Attitudes Test (EAT-40), which was first published by Garner and Garfinkel in 1979.

The EAT-26 is not a diagnostic instrument — it does not confirm or rule out an eating disorder diagnosis. Rather, it functions as a screening measure designed to identify individuals who exhibit patterns of eating-related attitudes and behaviors that warrant further professional evaluation. Think of it as a first-pass filter: it flags potential concerns so that clinicians can then conduct comprehensive diagnostic assessments.

The tool is used across a wide range of settings, including primary care offices, university counseling centers, athletic programs, and large-scale community screening initiatives. Its brevity, strong psychometric properties, and ease of administration have made it a standard instrument in both clinical practice and research on eating disorders worldwide.

The EAT-26 measures a constellation of attitudes and behaviors associated with disordered eating. The 26 items are organized into three subscales, each capturing a distinct dimension of eating pathology:

• Dieting (13 items): This subscale assesses preoccupation with being thinner, avoidance of fattening foods, and a pathological drive to control food intake. Items capture the cognitive and behavioral hallmarks of restrictive eating patterns — for example, awareness of calorie content, cutting food into small pieces, and feeling that food controls one's life.
• Bulimia and Food Preoccupation (6 items): This subscale taps into binge eating behaviors, purging (self-induced vomiting), and an overwhelming preoccupation with food and eating. It reflects the loss-of-control eating and compensatory behaviors described in the DSM-5-TR criteria for bulimia nervosa and binge-eating disorder.
• Oral Control (7 items): This subscale measures self-control around food and the perceived pressure from others to gain weight. It captures features often associated with anorexia nervosa, including deliberate food restriction and social pressure related to eating.

Beyond the 26 scored items, the EAT-26 also includes a behavioral questions section that asks about specific behaviors over the past six months, including binge eating, self-induced vomiting, use of laxatives or diet pills, and exercise patterns. These behavioral items are not included in the total score but provide critical clinical information that can trigger a referral for further evaluation even when the total score falls below the clinical threshold.

The EAT-26 was originally developed and validated with adult females, reflecting the historically higher prevalence of diagnosed eating disorders in this population. However, over the past four decades, its use has expanded considerably:

• Adolescents and adults: The instrument is generally considered appropriate for individuals aged 13 and older. For younger adolescents, reading comprehension and the relevance of certain items should be considered.
• Males: While the EAT-26 was initially normed on female samples, subsequent research has supported its use with males. However, clinicians should be aware that certain forms of disordered eating more prevalent in males — such as muscularity-oriented behaviors — may not be fully captured by the instrument.
• Diverse populations: The EAT-26 has been translated into numerous languages and used in cross-cultural research. Validation studies exist for populations across North America, Europe, Asia, the Middle East, and Latin America. That said, cultural factors can influence how items are interpreted, and norms established in Western populations may not generalize perfectly.
• Athletes: The tool is frequently used in athletic screening programs, where the prevalence of disordered eating is elevated, particularly in sports emphasizing leanness, weight classes, or aesthetic performance.

The DSM-5-TR recognizes several feeding and eating disorders, including anorexia nervosa, bulimia nervosa, binge-eating disorder, avoidant/restrictive food intake disorder (ARFID), and other specified feeding or eating disorders. The EAT-26 is most sensitive to features associated with anorexia nervosa and bulimia nervosa. It was not designed to screen for ARFID or other conditions characterized primarily by sensory-based food avoidance rather than body image disturbance.

One of the EAT-26's greatest practical strengths is its ease of administration. The instrument is a self-report questionnaire that typically takes 10 to 15 minutes to complete. It requires no specialized training to administer, though interpretation should always be conducted by a qualified professional.

Format: Each of the 26 items presents a statement about eating attitudes or behaviors (e.g., "I am terrified about being overweight" or "I find myself preoccupied with food"). Respondents rate each item on a 6-point Likert scale: Always, Usually, Often, Sometimes, Rarely, Never.

Administration settings: The EAT-26 can be administered via paper-and-pencil format or electronically through online platforms. It is suitable for individual clinical use, group screening in schools or universities, and large-scale epidemiological research. The National Eating Disorders Association (NEDA) has historically offered an online version as part of its Eating Disorders Screening Program.

Language requirements: The questionnaire requires approximately a sixth-grade reading level, making it accessible for most adolescent and adult populations. When used with non-English-speaking populations, validated translations should be selected to preserve the psychometric integrity of the instrument.

The EAT-26 uses a specific scoring system that weights responses toward the symptomatic end of the scale:

• "Always" = 3 points
• "Usually" = 2 points
• "Often" = 1 point
• "Sometimes," "Rarely," and "Never" = 0 points

One item (Item 26: "I enjoy trying new rich foods") is reverse-scored, meaning points are assigned in the opposite direction. Total scores range from 0 to 78.

Clinical threshold: A total score of 20 or above is the established cutoff indicating a high level of concern about dieting, body weight, or problematic eating behaviors. This score does not mean a person has an eating disorder — it means the pattern of responses is consistent with significant eating-related distress and warrants a thorough clinical evaluation.

Referral triggers: Importantly, a referral for further assessment is recommended if any of the following conditions are met:

• Total EAT-26 score is 20 or higher
• The individual reports any of the behavioral symptoms in the supplementary questions (binge eating, purging, laxative use, etc.) regardless of total score
• BMI is below expected ranges for age and sex

This multi-criteria referral approach is critical because some individuals with clinically significant eating disorders — particularly those with binge-eating disorder or purging behaviors — may score below 20 on the attitudinal items while endorsing dangerous compensatory behaviors. The behavioral questions serve as a safety net to catch these cases.

Subscale analysis: While the total score drives referral decisions, examining subscale scores can provide useful clinical information. For example, an elevated Dieting subscale with low Bulimia scores suggests restrictive patterns, whereas an elevated Bulimia subscale points toward binge-purge behaviors. This information can help guide the direction of a follow-up assessment.

The EAT-26 has a robust evidence base supporting its psychometric properties, accumulated over more than four decades of research:

Internal consistency: The instrument demonstrates strong internal reliability, with Cronbach's alpha values typically ranging from 0.83 to 0.93 across studies and populations. This means the items consistently measure the same underlying construct of disordered eating attitudes.

Test-retest reliability: Research shows acceptable temporal stability, indicating that scores remain relatively consistent over short time intervals when no intervention has occurred.

Criterion validity: The EAT-26 has demonstrated good sensitivity in identifying individuals who meet diagnostic criteria for eating disorders. In the original validation study by Garner and colleagues (1982), the EAT-26 correctly identified over 90% of cases that were identified by the longer EAT-40. Research generally reports sensitivity values ranging from approximately 77% to 90% for detecting eating disorders, depending on the population studied.

Specificity: The tool performs well at correctly identifying individuals who do not have eating disorders, though false positive rates can be elevated in certain populations — for example, in community samples where prevalence is lower, a greater proportion of those scoring above 20 may not meet full diagnostic criteria upon evaluation.

Convergent validity: EAT-26 scores correlate significantly with other established measures of eating pathology, including the Eating Disorder Inventory (EDI), the Eating Disorder Examination Questionnaire (EDE-Q), and clinician-rated assessments.

Factor structure: The three-factor structure (Dieting, Bulimia and Food Preoccupation, Oral Control) has been replicated across multiple studies, though some cross-cultural research has found variations in factor loadings, underscoring the importance of culturally validated versions.

Despite its widespread use and strong evidence base, the EAT-26 has several important limitations that clinicians and researchers must consider:

• It is a screening tool, not a diagnostic instrument. This distinction is fundamental. A score above 20 does not constitute a diagnosis, and a score below 20 does not rule out an eating disorder. Individuals with significant pathology — particularly those with binge-eating disorder, atypical anorexia nervosa, or well-concealed symptoms — can score in the normal range.
• Limited sensitivity to binge-eating disorder (BED). The EAT-26 was developed before BED was recognized as a distinct diagnosis. While the Bulimia subscale captures some binge-related features, the instrument's emphasis on restriction and weight concern means it may underdetect individuals whose primary symptom is recurrent binge eating without compensatory behaviors.
• Body image and muscularity concerns in males. The EAT-26 was designed around eating disorder presentations more typical in females. Male-specific pathology, such as drive for muscularity, excessive protein supplementation, or muscle dysmorphia, is not assessed.
• Self-report bias. Like all self-report instruments, the EAT-26 is vulnerable to denial, minimization, and social desirability bias. Individuals with eating disorders — particularly anorexia nervosa — often lack insight into the severity of their condition or may deliberately conceal symptoms.
• Cultural limitations. Items referencing specific foods, body ideals, or eating norms may not translate directly across cultures. While many validated translations exist, clinicians should exercise caution when using the EAT-26 in populations for which no local validation study has been conducted.
• Does not assess medical severity. The EAT-26 measures attitudes and behaviors but provides no information about the medical consequences of disordered eating, such as electrolyte imbalances, cardiac complications, or bone density loss. Medical evaluation remains essential regardless of EAT-26 scores.
• Age limitations. The instrument has limited validation for children under 13. For younger populations, the Children's Eating Attitudes Test (ChEAT) is a more appropriate alternative.

In clinical settings, EAT-26 results serve several practical functions:

Step 1: Initial screening. The EAT-26 is typically administered as part of a broader intake process or targeted screening program. In primary care, it can be used when a patient presents with weight-related concerns, amenorrhea, gastrointestinal complaints, or other symptoms that raise clinical suspicion. In university counseling centers, it may be offered as part of routine mental health screening.

Step 2: Identifying referral needs. When scores meet or exceed the cutoff of 20 — or when behavioral items are endorsed — the standard of care is to refer the individual for a comprehensive eating disorder evaluation. This evaluation typically involves a structured clinical interview (such as the Eating Disorder Examination, or EDE), a medical assessment, and a thorough psychosocial history.

Step 3: Informing clinical assessment. Subscale scores and specific item endorsements can help clinicians focus their diagnostic interviews. For instance, high Oral Control scores may prompt closer exploration of restrictive eating and social eating patterns, while endorsement of vomiting or laxative use in the behavioral section signals the need for immediate medical assessment.

Step 4: Monitoring treatment progress. Although the EAT-26 was designed primarily for screening, some clinicians and researchers use it longitudinally to track changes in eating attitudes over the course of treatment. Declining scores can reflect meaningful improvement in disordered cognitions and behaviors, though the instrument should be supplemented with more detailed outcome measures for this purpose.

Step 5: Population-level screening and research. The EAT-26 is extensively used in epidemiological research to estimate the prevalence of eating disorder risk across populations. It has been a cornerstone of NEDA's National Eating Disorders Screening Program and is used in school-based, college-based, and community health screening initiatives.

According to NIMH estimates, eating disorders affect approximately 2.7% of adolescents aged 13–18 in the United States, and lifetime prevalence for anorexia nervosa, bulimia nervosa, and binge-eating disorder combined is estimated at 2–5% of the general adult population. Tools like the EAT-26 play an essential role in identifying the substantial number of affected individuals who never seek treatment on their own.

The EAT-26 is available for use by qualified professionals and in approved screening programs. Key access points include:

• Official website: The EAT-26 is available through the official resource at eat-26.com, maintained by the instrument's authors. This site provides the questionnaire, scoring instructions, and guidelines for appropriate use.
• National Eating Disorders Association (NEDA): NEDA has historically offered the EAT-26 as part of its online screening tools at nationaleatingdisorders.org. These platforms often combine the EAT-26 with supplementary questions and automated scoring.
• Published literature: The original validation article — Garner, D.M., Olmsted, M.P., Bohr, Y., & Garfinkel, P.E. (1982), "The Eating Attitudes Test: Psychometric features and clinical correlates," published in Psychological Medicine — contains the full instrument and scoring key.
• Permission for use: The EAT-26 is generally available for clinical and research use without charge, but users are typically asked to cite the original source and adhere to the administration and scoring guidelines provided by the authors. Commercial or large-scale use may require written permission.

Notably, the EAT-26 should always be used within the context of a broader clinical or screening framework. Administering the questionnaire without a clear plan for follow-up evaluation and referral when positive results emerge is considered clinically inappropriate and potentially harmful.

If you or someone you know is experiencing persistent concerns about food, weight, body shape, or eating behaviors, professional evaluation is strongly recommended — regardless of any screening score. Eating disorders are serious psychiatric conditions with significant medical consequences and the highest mortality rate of any mental illness.

Seek immediate evaluation if any of the following are present:

• Significant unintentional weight loss or rapid weight fluctuations
• Recurrent binge eating or feeling unable to stop eating
• Self-induced vomiting, laxative misuse, or excessive exercise used to compensate for eating
• Fainting, dizziness, chest pain, or other physical symptoms related to eating patterns
• Preoccupation with food, calories, or body weight that interferes with daily functioning
• Withdrawal from social activities, particularly those involving food

A qualified mental health professional — such as a psychologist, psychiatrist, or licensed therapist with expertise in eating disorders — can conduct a comprehensive evaluation and develop an appropriate treatment plan. Primary care physicians and pediatricians can also initiate evaluations and coordinate care.

Crisis resources: If you are in crisis, contact the 988 Suicide and Crisis Lifeline (call or text 988) or the NEDA helpline at 1-800-931-2237 for support and referral information. The Crisis Text Line is also available by texting "NEDA" to 741741.