To score the EDE-Q 6.0, average the items in each of its four subscales (Restraint, Eating Concern, Shape Concern, Weight Concern), rated 0 to 6, then average the four subscale scores for the global score. Items 13 to 18 are behaviour counts, not summed. In Australia, an EDE-Q score of 3 or more is an MBS eating disorder plan criterion.
The Eating Disorder Examination Questionnaire (EDE-Q) is the self-report version of the Eating Disorder Examination interview, developed by Fairburn and Beglin at the Centre for Research on Eating Disorders at Oxford (CREDO). It covers the past 28 days and produces two kinds of data: severity scores and behaviour frequencies. Most scoring errors come from mixing the two.
How do you score the EDE-Q?
The current version, EDE-Q 6.0, has 28 items. Twenty-two of them are rated on a 0 to 6 scale and feed the subscales. The scoring key is published in the EDE 17.0D document, which lists the EDE-Q item numbers for each subscale:
- Restraint (5 items): 1, 2, 3, 4, 5
- Eating Concern (5 items): 7, 9, 19, 20, 21
- Shape Concern (8 items): 6, 8, 10, 11, 23, 26, 27, 28
- Weight Concern (5 items): 8, 12, 22, 24, 25
Item 8 (preoccupation with shape or weight) belongs to both Shape Concern and Weight Concern, so it is counted twice.
For each subscale, add the ratings and divide by the number of items. If some items are missing, CREDO's instructions allow you to divide by the number of items rated, as long as more than half of the subscale's items have a rating. The global score is the sum of the four subscale scores divided by four.
A worked example
Suppose a client's ratings give these subscale totals:
- Restraint: 20 across 5 items = 4.00
- Eating Concern: 12 across 5 items = 2.40
- Shape Concern: 35 across 8 items = 4.38
- Weight Concern: 21 across 5 items = 4.20
The global score is (4.00 + 2.40 + 4.38 + 4.20) / 4 = 3.74.
A common shortcut is to average all 22 items instead. With the same answers that gives 3.82, not 3.74, because the subscales have different lengths and item 8 appears twice. The difference is small here, but it can move a client across a threshold, so always calculate the global score from the subscale scores.
How do you interpret items 13 to 18?
Items 13 to 18 ask for counts over the past 28 days, not ratings:
- 13: episodes of eating an unusually large amount of food
- 14: how many of those involved a sense of loss of control (objective binge episodes)
- 15: days on which those episodes occurred
- 16: self-induced vomiting to control shape or weight
- 17: laxative use to control shape or weight
- 18: driven or compulsive exercise to control weight, shape or body fat
Report these as raw frequencies next to the subscale scores. Do not add them to anything. They often matter more for clinical decisions than the global score: a client with a modest global score and frequent vomiting needs a different response from one with high shape concern and no compensatory behaviour.
Two limits are worth knowing. The EDE-Q asks about objective binge episodes but not subjective ones (loss of control over an amount others would not call unusually large), which Mond and colleagues flagged as a gap. And items 1 to 12 ask about the number of days, while items 13 to 18 ask about the number of times, so a client can misread which one they are answering.
What is a normal EDE-Q score?
The most relevant comparison data for Australian clinicians come from Mond et al. (2006), who gave the EDE-Q to 5,255 women aged 18 to 42 in the ACT region. Their mean scores (with standard deviations) were:
- Restraint: 1.30 (1.40)
- Eating Concern: 0.76 (1.06)
- Weight Concern: 1.79 (1.51)
- Shape Concern: 2.23 (1.65)
- Global: 1.52 (1.25)
Scores tended to fall with age, and the paper reports norms for five age bands, so compare a 20-year-old client against the youngest band rather than the overall mean.
Norms for other groups differ:
- Adolescent girls: Carter, Stewart and Fairburn (2001) published means and percentile ranks for 808 girls aged 12 to 14.
- Adolescent boys: in an ACT school sample of 531 boys aged 12 to 18, Mond et al. (2014) found subscale scores substantially lower than in girls.
- Men in treatment: Smith et al. (2017) published clinical norms for 386 males and found greater severity among females with eating disorders.
The practical point: a score that looks unremarkable against female norms can be high for a male client. Mond and colleagues also noted that eating and weight-control behaviours largely confined to males may not be adequately assessed, so a low score in a young man does not rule out a problem.
What EDE-Q cut-off indicates an eating disorder?
CREDO does not publish a single diagnostic cut-off, and the EDE-Q is not a diagnostic test. Research thresholds vary by sample:
- In a community sample of women, Mond et al. (2004) found a global score of 2.3, combined with any objective binge episodes or exercise for weight control, gave a sensitivity of 0.83 and specificity of 0.96. That analysis rested on only 13 cases.
- In a primary care sample of young women, Mond et al. (2008) found a global score of 2.80 or more gave the best balance, with sensitivity and specificity both 0.80.
The positive predictive values in both studies were modest (0.56 and 0.44), meaning many people above the threshold did not have an eating disorder. Treat a raised score as a reason for a full assessment, not a conclusion.
How does the EDE-Q fit the MBS eating disorder plan?
Under MBS note AN.36.1, a patient with a clinical diagnosis of anorexia nervosa is eligible for an Eating Disorder Treatment and Management Plan on that basis. Patients with bulimia nervosa, binge-eating disorder or other specified feeding or eating disorder must also meet all of the following:
- an EDE-Q score of 3 or more
- rapid weight loss, or binge eating or compensatory behaviour occurring 3 or more times per week
- at least two of five listed indicators, such as a hospital admission for an eating disorder in the previous 12 months
The note does not say which score it means. The global score is the EDE-Q's summary measure, so record it alongside the four subscales and the behaviour counts so the plan shows exactly what was assessed. On frequency, three times a week works out to roughly 12 or more episodes across the EDE-Q's 28-day window, which items 14, 16, 17 and 18 let you document directly.
The MBS threshold of 3 is an eligibility rule, not a clinical cut-off. It sits above both research thresholds above, so a client with a clinically significant eating disorder can score below it. For how the EDE-Q then anchors plan objectives and reviews, see our guide to eating disorder treatment plan goals and objectives.
Why are EDE-Q scores higher than interview scores?
Self-report tends to run higher. Fairburn and Beglin (1994) found the questionnaire produced higher scores than the interview for complex features such as binge eating and shape concern, while simple behaviours such as vomiting matched well. Mond et al. (2004) found EDE-Q subscale scores higher than EDE interview scores on every subscale, by 0.25 for Restraint up to 0.85 for Shape Concern.
The term "unusually large amount of food" is the main source of disagreement. If items 13 and 14 drive a decision, go through a recent episode with the client in session before relying on the count.
Is there a shorter version of the EDE-Q?
The EDE-QS is a 12-item version with a 4-point response scale covering the past 7 days. Prnjak et al. (2020) found a total of 15 gave the best screening balance. Because the items, scale and time frame differ, EDE-QS totals cannot be compared with EDE-Q global scores, and the MBS criterion refers to the full questionnaire.
Is the EDE-Q free to use?
According to CREDO, the EDE-Q and its items are under copyright and freely available for non-commercial research use without seeking permission. For commercial use, CREDO asks users to contact them directly. If you are unsure how your setting is classified, check with CREDO.
How should you record EDE-Q results in clinical notes?
A useful EDE-Q entry includes:
- version (EDE-Q 6.0) and date completed
- the four subscale scores and the global score, to two decimal places
- raw counts for items 13 to 18
- any missing items and whether each subscale still met the more-than-half rule
- the norm group you compared against
- what changes as a result: further assessment, referral, plan eligibility or a re-test date
Because every item covers the past 28 days, re-administering more often than every four weeks means the windows overlap. Repeating it at plan review points keeps comparisons like-for-like. If you use other measures alongside it, such as the K10, keep them separate: a falling general distress score does not mean eating disorder symptoms have changed. Our guide to writing a psychological assessment report covers how to present scores in a report.
PractaLuma is AI-native practice management software for Australian mental-health practices. Its standardised assessments feature includes the EDE-Q and keeps scores with the client record, so you can compare repeat results alongside your notes. See all features and pricing.
Frequently asked questions
What is a high EDE-Q global score?
There is no official cut-off. Research thresholds of 2.3 and 2.80 have been reported in studies of women, and the MBS uses 3 or more for plan eligibility. The ACT community mean for women aged 18 to 42 was 1.52.
What does an EDE-Q global score of 4 mean?
A global score of 4 is well above community norms for women and meets the MBS score criterion. It does not establish a diagnosis on its own. Review the subscale pattern and the behaviour counts, then complete a full assessment.
Can the EDE-Q diagnose an eating disorder?
No. It measures the severity of eating disorder features and the frequency of key behaviours. Diagnosis needs a clinical assessment against DSM-5 criteria, which the MBS note asks practitioners to consider.
Is the EDE-Q valid for men?
It is used with men, but scores run lower than in women and behaviours more common in males may not be fully captured. Compare male clients with male norms where possible and interpret a low score with caution.
This guide is general information for registered practitioners and does not replace clinical judgement or the current MBS explanatory notes.