Evidence Summary

Eating Disorders in Adolescents and Adults: Screening

March 15, 2022

Recommendations made by the USPSTF are independent of the U.S. government. They should not be construed as an official position of the Agency for Healthcare Research and Quality or the U.S. Department of Health and Human Services.

By Cynthia Feltner, MD, MPH; Christine Peat, PhD; Shivani Reddy, MD, MS; Sean Riley, MA, MSc; Nancy Berkman, PhD; Jennifer Cook Middleton, PhD; Casey Balio, PhD; Manny Coker-Schwimmer, MPH; Daniel E. Jonas, MD, MP

The information in this article is intended to help clinicians, employers, policymakers, and others make informed decisions about the provision of health care services. This article is intended as a reference and not as a substitute for clinical judgment.

This article may be used, in whole or in part, as the basis for the development of clinical practice guidelines and other quality enhancement tools, or as a basis for reimbursement and coverage policies. AHRQ or U.S. Department of Health and Human Services endorsement of such derivative products may not be stated or implied.

This article was published online in JAMA on March 15, 2022 (JAMA. 2022;327(11):1068-1082. doi:10.1001/jama.2022.1807).

Return to Table of Contents

Importance: Eating disorders are associated with adverse health and social outcomes.

Objective: To review the evidence on screening for eating disorders in adolescents and adults to inform the US Preventive Services Task Force.

Data Sources: MEDLINE, Cochrane Library, PsycINFO, and trial registries through December 19, 2020; surveillance through January 1, 2022.

Study Selection: English-language studies of screening test accuracy, randomized clinical trials (RCTs) of screening or interventions for eating disorders in populations with screen-detected or previously untreated eating disorders (trials limited to populations who are underweight were ineligible).

Data Extraction and Synthesis: Dual review of abstracts, full-text articles, and study quality. Meta-analysis of test accuracy studies and intervention trials.

Main Outcomes and Measures: Test accuracy, eating disorder symptom severity, quality of life, depression, and harms.

Results: Fifty-seven studies were included (N = 10,773); 3 (n = 1073) limited to adolescents (mean or median age, 14-15 years). No study directly evaluated the benefits and harms of screening. Seventeen studies (n = 6804) evaluated screening test accuracy. The SCOFF questionnaire (cut point ≥2) had a pooled sensitivity of 84% (95% CI, 74% to 90%) and pooled specificity of 80% (95% CI, 65% to 89%) in adults (10 studies, n = 3684). Forty RCTs (n = 3969) evaluated interventions for eating disorders; none enrolled a screen-detected population. Lisdexamfetamine for binge eating disorder (4 RCTs; n = 900) was associated with larger reductions in eating disorder symptom severity on the Yale-Brown Obsessive Compulsive Scale modified for binge eating (YBOCS-BE) than placebo (pooled mean difference, −5.75 [95% CI, −8.32 to −3.17]). Two RCTs (n = 465) of topiramate for binge eating disorder found larger reductions in YBOCS-BE scores associated with topiramate than placebo, from −6.40 (95% CI, −8.16 to −4.64) to −2.55 (95% CI, −4.22 to −0.88). Nine pharmacotherapy trials (n = 2006) reported on harms. Compared with placebo, lisdexamfetamine was associated with higher rates of dry mouth, headache, and insomnia, and topiramate was associated with higher rates of paresthesia, taste perversion, confusion, and concentration difficulty. Twenty-four trials (n = 1644) assessed psychological interventions. Guided self-help for binge eating disorder improved eating disorder symptom severity more than control (pooled standardized mean difference, −0.96 [95% CI, −1.26 to −0.67]) (5 studies, n = 391). Evidence on other interventions was limited.

Conclusions and Relevance: No studies directly assessed the benefits and harms of screening. The SCOFF questionnaire had adequate accuracy for detecting eating disorders among adults. No treatment trials enrolled screen-detected populations; guided self-help, lisdexamfetamine, and topiramate were effective for reducing eating disorder symptom severity among referred populations with binge eating disorder, but pharmacotherapies were also associated with harms.

Return to Table of Contents

Eating disorders are conditions marked by a disturbance in eating or eating-related behaviors that impair functioning.1 This review focused on common eating disorders that could be asymptomatic or undetected in routine primary care: anorexia nervosa, avoidant/restrictive food intake disorder, bulimia nervosa, binge eating disorder, and other specified eating or feeding disorder. Estimated lifetime prevalences for anorexia nervosa, bulimia nervosa, and binge eating disorder in adult women are 1.42%, 0.46%, and 1.25%, respectively, and are lower in adult men (anorexia nervosa, 0.12%; bulimia nervosa, 0.08%; binge eating disorder, 0.42%).2 In adolescents aged 12 to 17 years, estimated lifetime prevalence for anorexia nervosa, bulimia nervosa, and binge eating disorder are 0.3%, 1.3%, and 2.3%, respectively, for females and 0.3%, 0.5%, and 1.3% for males.3 Estimated prevalence for some disorders vary by race and ethnicity and age category (eTable 1 in the JAMA Supplement).

Eating disorders are associated with adverse health outcomes which vary by diagnosis, duration, and frequency of certain behaviors. For bulimia nervosa, purging behaviors (eg, self-induced vomiting) can lead to electrolyte disturbances and dental erosion.4 Binge eating disorder can contribute to obesity,5 and anorexia nervosa is associated with morbidity attributed to weight loss and malnutrition.6 Eating disorders are also commonly comorbid with mood and substance abuse disorders.7

Measurement of weight, height, and body mass index is routine in primary care practice and may detect some eating disorders, particularly anorexia nervosa. Disorders without physical symptoms may go unrecognized, and some individuals experiencing symptoms may not seek care. Routine screening could detect eating disorders early, lead to earlier treatment, and reduce future morbidity.

The US Preventive Services Task Force (USPSTF) has not previously made a recommendation on screening for eating disorders. This review evaluated the evidence on screening adolescents and adults for eating disorders for populations and settings relevant to primary care in the US to inform a recommendation by the USPSTF.

Return to Table of Contents

Scope of the Review

Detailed methods are available in the full evidence report.8 Figure 1 shows the analytic framework and key questions (KQs) that guided the review.

Data Sources and Searches

PubMed/MEDLINE, the Cochrane Library, PsycINFO, and ClinicalTrials.gov were searched for English-language articles published through June 23, 2020 (eMethods in the JAMA Supplement). The searches were supplemented with reference lists of pertinent articles and studies suggested by peer reviewers or public comment respondents. Since June 2020, ongoing surveillance was conducted through article alerts and targeted searches of journals to identify major studies published in the interim that may affect the conclusions or understanding of the evidence and the related USPSTF recommendation through January 1, 2022. No relevant studies were identified.

Study Selection

Two investigators independently reviewed titles, abstracts, and full-text articles using prespecified eligibility criteria (eMethods in the JAMA Supplement). Disagreements were resolved by consensus. For all KQs, English-language studies of adolescents and adults 10 years or older conducted in settings generalizable to primary care, including school-based health centers, and in countries categorized as “very high” on the United Nations Human Development Index were included.9 The scope of this review was focused on populations with eating disorders unlikely to be detected in the context of routine primary care. Studies limited to populations with physical signs of eating disorders (eg, populations who are underweight) were ineligible because eating disorders would be part of the diagnostic assessment for individuals presenting with an abnormally low body weight. For KQ1 and KQ3 (direct evidence of benefits and harms of screening), randomized clinical trials (RCTs) comparing screening with no screening in asymptomatic populations were eligible. For KQ2 (screening test accuracy), studies comparing a screening test with a diagnostic reference standard for eating disorders (structured or semistructured diagnostic interview or diagnostic questionnaire) were eligible. Eligible screening tests included those feasible for use in primary care settings (brief, easy to interpret) and designed to detect any eating disorder or specific disorders (eg, binge eating disorder); longer questionnaires (eg, the 26-item Eating Attitudes Test) were excluded.

For KQs on benefits and harms of treatment (KQ5 and KQ6), RCTs enrolling populations with screen-detected eating disorders, or populations from specialty settings or via advertisements who had not been previously treated for eating disorders, were included. Eligible treatments included psychological interventions (eg, cognitive behavioral) delivered in a group, individual, or family-based format, including self-help interventions, or pharmacotherapy with US Food and Drug Administration–approved medications. Eligible RCTs had to compare treatment with an inactive control (ie, no treatment, wait-list, minimal intervention [eg, brief education about eating disorders], or placebo). RCTs evaluating combined psychological and pharmacotherapy interventions were eligible if they included an inactive control group.

Eligible outcomes for KQs on the benefits of screening or treatment included measures of eating disorder symptom severity, health-related quality of life or function, depression, and others. Intermediate outcomes such as mean change in frequency of specific behaviors (eg, change in frequency of binge eating episodes) were excluded. Eligible outcomes for KQ3 (harms of screening) included increased anxiety, labeling, and stigma associated with screening; for KQ5 (harms of interventions), outcomes included any harms attributed to interventions, such as harms associated with medications.

Data Extraction and Quality Assessment

For each study, 1 investigator extracted information about populations, tests or interventions, comparators, outcomes, settings, and designs, and a second investigator reviewed the information for completeness and accuracy. Two investigators independently assessed each study’s methodological quality, using the Cochrane Risk of Bias Tool (RoB 2.0)10 and the QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies 2) for studies of test accuracy.11 Disagreements in quality ratings were resolved through discussion or independent assessment from a third senior investigator. Risk-of-bias assessments using these instruments were translated into an overall study quality rating of good, fair, or poor using predefined criteria developed by the USPSTF and adapted for this topic (eMethods in the JAMA Supplement). Individual study quality ratings are reported in eTables 4-7 in the JAMA Supplement.

Data Synthesis and Analysis

Findings for each KQ were summarized in tables, figures, and narrative format. For KQ2, pooled sensitivities and specificities for screening tests were calculated using a hierarchical summary receiver operating characteristic curve analysis when at least 4 similar studies were available. For KQ4, random-effects restricted maximum likelihood models were conducted on continuous measures of eating disorder and depression symptom severity (analyzing standardized mean difference or unstandardized mean difference in change between groups) when at least 3 similar studies were available. When studies reported more than 1 continuous outcome for eating disorder symptom severity, the outcome most commonly reported by similar studies in pooled estimates was preferentially selected. Statistical significance was assumed when 95% CIs of pooled results did not cross the null. All testing was 2-sided. Comprehensive Meta-Analysis version 3.4 (Biostat Inc) and Stata version 16 (StataCorp)12 were used to conduct all quantitative analyses.

The overall strength of the evidence for each KQ was assessed as high, moderate, low, or insufficient based on the overall quality of the studies, consistency of results between studies, precision of findings, risk of reporting bias, and limitations of the body of evidence, using methods developed for the USPSTF (and the Evidence-based Practice Center program).13 Additionally, the applicability of the findings to US primary care populations and settings was assessed. Discrepancies were resolved through consensus discussion.

Return to Table of Contents

A total of 57 studies (59 articles) with 10,773 participants were included (Figure 2).

Benefits of Screening

Key Question 1. Does screening for eating disorders in adolescents and adults improve health outcomes, including for specific subgroups of interest?

No eligible studies were identified.

Screening Accuracy

Key Question 2. What is the accuracy of primary care–relevant screening tests for eating disorders in adolescents and adults, including for specific subgroups of interest?

Ten good-quality14-25 and 7 fair-quality20,26-31 studies (18 articles) assessed the accuracy of 9 screening questionnaires; 5 were designed to detect any eating disorder,14,15,17,18,20,21,24-28,31 and 4 were designed to detect eating disorders characterized by binge eating (bulimia nervosa or binge eating disorder).19,23,29,30 Detailed characteristics are reported in Table 1.14-32 Of the studies, 11 assessed the 5-item SCOFF questionnaire. (Some experts recommend not considering SCOFF an acronym since it is based on terminology from signaling questions that may not translate well [eg, “Have you recently lost more than One stone in a 3-month period?”].) Reference standards used to evaluate screening test accuracy included a diagnostic clinical interview or a longer diagnostic questionnaire.

Most studies enrolled participants from university settings17,18,20,22,25,31 and outpatient clinics (primary care,15,19,24-26 psychiatry,14 and obesity clinics).23,28,30 Six studies were set in the US;15,18-20,26,29 others were set in the UK,24,25 Taiwan,14 Malaysia,17 and various European countries.21-23,27,28,30,31 Most studies enrolled only females15-18,22,24,28,31 or predominantly females (>60%);14,20,23,25,29,30 2 enrolled a majority of males.19,21 Two studies limited to adolescents with a mean or median age of 14 years, and all others enrolled adults (mean age, 20-63 years).21,23 In 4 studies evaluating a screening tool for bulimia nervosa or binge eating disorder, prevalence (based on the reference standard) ranged from 8% to 22%;19,23,29,30 the prevalence of any eating disorder ranged from 2% to 46%.

Table 2 summarizes results of screening test accuracy. In studies of adults (10 studies, n = 4348), the SCOFF questionnaire (cut point ≥2) had a pooled sensitivity of 84% (95% CI, 74% to 91%) and pooled specificity of 80% (95% CI, 65% to 89%)(Table 2; eFigure 1 in the JAMA Supplement). Seven studies (n = 3424) assessed the accuracy at a higher cut point (≥3);14,17,22,24,26,28 pooled sensitivity was lower at 69% (95% CI, 56% to 80%), and specificity was higher at 90% (95% CI, 69% to 98%) (Table 2; eFigure 2 in the JAMA Supplement). One study evaluated the SCOFF questionnaire (cut point ≥2) among adolescents (n = 954; mean age, 14 years);21 sensitivity was 73% (95% CI, 63% to 83%), and specificity was 78% (95% CI, 75% to 80%).

Eight other screening questionnaires were assessed across 8 included studies.15,18,19,23,25,29-31 One, the EDS-PC (5 items, developed for use in primary care25)was evaluated in 2 studies (n = 627) enrolling different populations (Table 2); sensitivity ranged from 97% to 100%, and specificity ranged from 40% to 71%.15,25 All other screening questionnaires were assessed by 1 study each; results are summarized in Table 2.19,2

Harms of Screening

Key Question 3. What are the harms of screening for eating disorders in adolescents and adults, including for specific subgroups of interest?

No eligible studies were identified.

Benefits of Treatment

Key Question 4. How effective are interventions for improving health outcomes in screen-detected or previously untreated adolescents and adults with eating disorders, including for specific subgroups of interest?

Forty fair- to good-quality RCTs (n = 3969) of treatment for eating disorders were included—18 (19 publications) assessing pharmacotherapy33-51 and 24 assessing therapy (eTables 9 and 10 in the JAMA Supplement); of these, 2 assessed both pharmacotherapy and therapy interventions compared with a control.50,51 All enrolled populations referred or recruited to treatment; none enrolled populations detected by screening in primary care. In 17 studies describing race or ethnicity, 1 was limited to Latinas only, 522 enrolled a population that was 54% to 55% non-White (from the US),53,54 and all others enrolled a majority of White participants.

Among 18 RCTs evaluating the benefit of pharmacotherapy compared with placebo over 6 to 16 weeks (eTable 9 in the JAMA Supplement), 14 enrolled populations with binge eating disorder (defined by Diagnostic and Statistical Manual of Mental Disorders [Fourth Edition] or Diagnostic and Statistical Manual of Mental Disorders [Fifth Edition] [DSM-5] criteria), and 4 enrolled populations with bulimia nervosa defined by Diagnostic and Statistical Manual of Mental Disorders (Third Edition) criteria.41,43,48,51 All enrolled adults, and 1 trial (n = 50) enrolled both adults and adolescents as young as 16 years (mean age, 25 years).55 Detailed characteristics of populations and pharmacotherapy are reported in eTable 9 in the JAMA Supplement. 

Four RCTs (described in 3 publications) compared lisdexamfetamine with placebo among adults with binge eating disorder.33,35,37 All measured binge eating disorder symptom severity using the Yale-Brown Obsessive Compulsive Scale modified for binge eating (YBOCSBE); for doses ranging from 50 to 60mg/d, the pooled mean difference in change from the baseline score over 11 to 12 weeks (4 studies, n = 900) was −5.75 (95% CI, −8.32 to −3.17) (Figure 3).This difference falls within the range considered a minimum clinically important change on the YBOCS-BE (−4 to −17).56 Other eligible outcomes were reported by only 1 or 2 studies each (eTable 9 in the JAMA Supplement). Two trials of topiramate (n = 465) measured reduction in eating disorder symptom severity using the YBOCS-BE over 14 to 15 weeks (mean or median dose, 212-300 mg/d).44,48 Both found significant improvement favoring topiramate (Figure 3); 1 found a difference between groups in mean change from baseline score (−6.50) within the range considered a minimum clinically important change (−4 to −17),44 and the other found a smaller difference in mean score change (−2.55).48

Five RCTs assessed a selective serotonin reuptake inhibitor (SSRI) for improving binge eating disorder, including fluoxetine (2 studies)38,50 and 1 study each of fluvoxetine,42 sertraline,40 and escitalopram.46 None selected participants based on the presence of comorbid depression; however, in 4 trials prevalence of lifetime depression ranged from 37% to 77%,38,40,46,50 and in 3 trials prevalence of current major depression ranged from 18% to 25%.38,40,46 Only 2 trials measured eating disorder symptom severity (the Eating Disorder Examination–Questionnaire [EDE-Q] and YBOCS-BE); although both found a reduction in symptom scores favoring SSRIs (eFigure 3 in the JAMA Supplement), results were imprecise. All reported on change in depression symptoms (eFigure 3 in the JAMA Supplement); SSRIs were associated with a larger reduction in depression symptom scores than placebo (pooled standardized mean difference [SMD], −0.6 [95% CI, −0.90 to −0.33]) (5 studies; n = 208).36,39,45 Three trials assessed fluoxetine for populations with bulimia nervosa and found inconsistent results for eating disorder symptom severity and depression (eTable 12 in the JAMA Supplement). One trial each evaluated duloxetine,39 bupropion,36 and imipramine45 for populations with binge eating disorder, and 1 evaluated desipramine for bulimia nervosa41 (eTable 9 in the JAMA Supplement); none found a significant differences between groups on measures of eating disorder symptom severity or depression.

Twenty-four trials (n = 1644) assessed the benefit of a psychological intervention compared with an inactive control (eTable 10 in the JAMA Supplement).50-55,57-74 Most enrolled populations with binge eating, either binge eating disorder or bulimia nervosa with recurrent binge eating behavior; 1 trial enrolled those with bulimia nervosa without mention of binge eating,73 and 3 enrolled women with any DSM-5 eating disorder.55,57,62 One trial (n = 25) was limited to adolescents (mean age, 15 years),74 1 (n = 82) enrolled both adults and adolescents (as young as 14 years),55 and all others enrolled adults only.

Included trials focused on a variety of psychological interventions (eTable 10 in the JAMA Supplement); most evaluated a form of self-help based on cognitive behavioral therapy or other strategies, designed to help participants cope with eating disorder symptoms.51-55,59-61,63,66,68,72,75 Seven trials evaluated a type of group therapy57,65,67,69-71,73 and 4 evaluated a form of individual cognitive behavioral therapy.50,62,64,74 eTable 11 in the JAMA Supplement provides additional detail related to the intervention approach, components, and intensity.

Thirteen trials evaluated a self-help intervention, 7 assessed a form of “guided” self-help,52,58,59,61,63,68,72 and 7 assessed an “unguided” self-help intervention.51,53-55,60,63,66 One trial compared both guided and unguided self-help interventions with a control.63 Guided interventions included ongoing support and guidance, for example, several brief (25- to 30-minute) individually guided sessions, regular email contact for support,52,59,61,63,72 or individual feedback on assignments.58 Unguided interventions involved providing the intervention materials with instructions only.

Guided self-help was associated with a larger reduction in eating disorder severity than control (measured by the EDE or EDE-Q) over 12 to 24 weeks (pooled SMD, −0.96 [95% CI, −1.26 to −0.67] [5 studies; n = 391]) (Figure 4). Results from trials of unguided self-help (6 studies; n = 368) were consistent in favoring self-help, but pooled estimates were not statistically significant (SMD, −0.18 [95% CI, −0.38 to 0.03]) (Figure 4). For measures of depression, pooled results demonstrated larger reductions in mean scores compared with controls for both guided self-help (SMD, −0.73 [95% CI, −1.04 to −0.43]; 4 studies; n = 324) and unguided self-help (SMD, −0.37 [95% CI, −0.68 to −0.05]; 3 studies; n = 156). Few trials of self-help measured other eligible outcomes.

Seven RCTs assessed a group-based psychological intervention for binge eating disorder and bulimia nervosa with recurrent binge eating using different therapeutic approaches and number of sessions (eTable 11 in the JAMA Supplement).57,65,67,69-71,73 Group therapy (7 studies; n = 253) was associated with larger reductions in depression scores from baseline than inactive control (pooled SMD, −0.48 [95% CI, −0.69 to 0.27]). Three trials of group therapy measured eating disorder symptom severity using the EDE-Q and found inconsistent results (eFigure 4 in the JAMA Supplement).65,67 Four trials assessed different forms of individual therapy among adults and found mixed results (eFigure 4 in the JAMA Supplement).50,62,64 One trial of individual cognitive behavioral therapy in adolescents (n = 25; mean age, 15) found no significant differences between groups at 12 or 24 weeks on depression symptoms and psychosocial functioning.74

Harms of Treatment

Key Question 5. What are the harms of interventions for eating disorders, including for specific subgroups of interest?

No included studies of psychological interventions reported on harms. Nine studies of pharmacotherapy reported various harms associated with 4 medications, including lisdexamfetamine (4 studies),33,35,37 topiramate (2 studies),44,48 fluoxetine (2 studies),38,43 and escitalopram.46 Characteristics are described in KQ4 and eTable 9 in the JAMA Supplement.

In 1 trial of lisdexamfetamine (n = 259) over 11 weeks, 1 participant died during the study, and postmortem toxicology analysis found that methamphetamine/amphetamine levels were consistent with a methamphetamine overdose (death was not attributed to the study drug).37 Across all 4 trials of lisdexamfetamine, treatment-emergent harms were higher for the treatment groups than the placebo groups; commonly reported harms were dry mouth, insomnia, and jitteriness (eTable 13 in the JAMA Supplement).33,35,37 Two trials of topiramate (duration, 14-16 weeks) found significantly higher rates of paresthesia and taste perversion44,48 associated with topiramate than placebo. One trial found significantly higher rates of difficulty concentrating44 and the other found significantly higher rates of confusion.48

In 3 trials of SSRIs, 1 found significantly higher rates of several harms in the fluoxetine group than in the placebo group (eTable 13 in the JAMA Supplement), such as insomnia, nausea, and tremor.43 The other 2 trials reported no significant differences between groups for any adverse effects over 6 weeks.38,46

Return to Table of Contents

This systematic review evaluated evidence relevant to screening for eating disorders in adults and adolescents. A summary of findings, including an assessment of the strength of evidence for each KQ, is presented in Table 3. To date, there is no direct evidence from trials comparing the benefits and harms of routine screening vs no screening. Thus, this review answered 2 questions: how well screening detects eating disorders and how effective are interventions at treating eating disorders among populations with screen-detected or previously untreated eating disorders.

Screening tools are available for clinical practice that may reasonably identify adults with eating disorders, primarily the SCOFF questionnaire. Other tools were assessed by only 1 study each, limiting the ability to make stronger conclusions about screening test accuracy. The estimates of SCOFF screening test accuracy were derived from populations with a current prevalence of eating disorders ranging from 4% to 46% based on the reference standard, higher than recent estimates of eating disorders in the US. Potential harms of screening include false-positive screening results that lead to unnecessary referrals or labeling. Based on the pooled estimates accurateness for detecting any eating disorder (Table 2) among adults (10 studies, 4348 participants), the expected rate of false-positive test results would be 20%.

Most RCTs evaluating interventions for eating disorders were limited to adult women with binge eating disorder and bulimia nervosa, enrolled treatment-seeking populations (either respondents to advertisements or referrals), and measured outcomes over a relatively short duration. Some recruited participants using ads that indicated trials of treatment for binge eating and obesity. Both lisdexamfetamine and topiramate were effective in reducing eating disorder severity among adults with binge eating disorder but were also associated with various harms. Few trials of SSRIs reported an eligible health outcome specific to eating disorder symptoms; however, results of 5 trials enrolling adults with binge eating disorder found consistent improvement in depression symptoms associated with various SSRIs. Although trials did not enroll participants based on depression status, lifetime depression rates ranged from 37% to 77% in 4 trials reporting on mental health comorbidity. Whether improvement on depression scores indicates improved eating disorder symptom severity is not clear.

Among the 24 trials assessing psychological interventions, guided self-help improved eating disorder symptom severity and depressive symptoms among adults with binge eating disorder; results for unguided self-help were consistent in direction of effect, but pooled estimates were imprecise and smaller in magnitude than estimates for guided self-help. Evidence for group and individual psychological interventions was heterogeneous, and results were mixed. No trials of psychological interventions reported on potential harms of interventions, including whether some participants experienced increased anxiety or stigma because of the intervention.

The evidence from the current report highlights several important research needs. First, RCTs assessing health outcomes that directly compare routine screening with no screening among populations with no obvious signs or symptoms of eating disorders would inform the potential effectiveness of routine screening. Studies of screening test accuracy that enroll populations from general primary care settings would improve certainty about the accuracy of existing screening tests in these settings. Studies of screening test accuracy in adolescents are needed, given that adolescence is considered a time of risk for eating disorder onset and concern about how social media influences the mental health of adolescents. Similarly, studies of screening test accuracy that enroll a more diverse population with respect to race and ethnicity, gender, and sexual identity would help assess whether findings are broadly representative of the US population. In addition, RCTs of treatment enrolling screen-detected populations, rather than treatment-seeking populations, would inform future recommendations on the benefit of screening followed by referral to treatment. Ideally, these trials would assess treatment specific to the range and severity of eating disorders likely to be detected via routine screening (which may differ from trials of referred or treatment-seeking populations).

Limitations

This review has several limitations. First, because its purpose was to inform a recommendation on routine screening in persons without signs and symptoms of an eating disorder, studies limited to populations who are underweight (defined by body mass index or other criteria) were excluded. In addition, studies evaluating head-to-head comparisons of different interventions were excluded because the scope was designed to provide evidence on benefits of treatments compared with no treatment rather than assess the comparative effectiveness of interventions. For these reasons, no trials of populations with anorexia nervosa met the eligibility criteria; however, a larger body of evidence demonstrates treatment benefits for populations with anorexia nervosa. Although studies of populations with “other specified feeding and eating disorder” were eligible, no included study used this DSM-5 diagnosis to characterize participants; several enrolled populations with “subthreshold” criteria for bulimia nervosa or binge eating disorder (based on heterogeneous definitions), which may include participants who meet criteria for other specified eating or feeding disorder. The scope of this review was limited to studies that reported on health outcomes, including global measures of eating disorder severity; some excluded studies reported intermediate outcomes only (eg, mean changes in the frequency of binge eating episodes over short durations), which do not necessarily indicate long-term benefit. Most included treatment trials enrolled adults via advertisements and focused on specific eating disorders (primarily bulimia nervosa and binge eating disorder), including some focused on obesity and binge eating disorder; the applicability of results to populations who are not seeking care for eating disorder symptoms or who may have a new onset or less severe eating disorder is uncertain.

Return to Table of Contents

No studies directly assessed the benefits and harms of screening. The SCOFF questionnaire had adequate accuracy for detecting eating disorders among adults. No treatment trials enrolled screen-detected populations; guided self-help interventions, lisdexamfetamine, and topiramate were effective for reducing eating disorder symptom severity among referred populations with binge eating disorder, but pharmacotherapies were also associated with harms.

Return to Table of Contents

Source: This article was first published online in the Journal of the American Medical Association on March 15, 2022 (JAMA. 2022;327(11):1068-1082. doi:10.1001/jama.2022.1807).

Conflict of Interest Disclosures: Dr Peat reported serving as a member of a clinical advisory board for Equip Health. No other disclosures were reported.

Funding/Support: This research was funded under contract HHSA-290-2015-00011-I, Task Order 15, from the Agency for Healthcare Research and Quality (AHRQ), US Department of Health and Human Services, under a contract to support the US Preventive Services Task Force (USPSTF).

Role of the Funder/Sponsor: Investigators worked with USPSTF members and AHRQ staff to develop the scope, analytic framework, and KQs for this review. AHRQ had no role in study selection, quality assessment, or synthesis. AHRQ staff provided project oversight, reviewed the report to ensure that the analysis met methodological standards, and distributed the draft for peer review. Otherwise, AHRQ had no role in the conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript findings. The opinions expressed in this document are those of the authors and do not reflect the official position of AHRQ or the US Department of Health and Human Services.

Additional Information: A draft version of the full evidence report underwent external peer review from 4 content experts (Bryn Austin, ScD, Harvard Medical School; Denise Wilfley, PhD, Washington University School of Medicine; Devan Kansagara, MD, Oregon Health & Science University; Susan Kornstein, MD, Virginia Commonwealth University). Comments from reviewers were presented to the USPSTF during its deliberation of the evidence and were considered in preparing the final evidence review.

Return to Table of Contents

1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. American Psychiatric Association; 2013.
2. Udo T, Grilo CM. Prevalence and correlates of DSM-5-defined eating disorders in a nationally representative sample of US adults. Biol Psychiatry. 2018;84(5):345-354. doi:10.1016/j.biopsych.2018.03.014
3. Kessler RC, Avenevoli S, Costello EJ, et al. Design and field procedures in the US National Comorbidity Survey Replication Adolescent Supplement (NCS-A). Int J Methods Psychiatr Res. 2009;18(2):69-83. doi:10.1002/mpr.279
4. Forney KJ, Buchman-Schmitt JM, Keel PK, Frank GK. The medical complications associated with purging. Int J Eat Disord. 2016;49(3):249-259. doi:10.1002/eat.22504
5. Kessler RC, Berglund PA, Chiu WT, et al. The prevalence and correlates of binge eating disorder in theWorld Health OrganizationWorld Mental Health Surveys. Biol Psychiatry. 2013;73(9):904-914. doi:10.1016/j.biopsych.2012.11.020
6. Faje AT, Fazeli PK, Miller KK, et al. Fracture risk and areal bone mineral density in adolescent females with anorexia nervosa. Int J Eat Disord. 2014;47(5):458-466. doi:10.1002/eat.22248
7. Udo T, Grilo CM. Psychiatric and medical correlates of DSM-5 eating disorders in a nationally representative sample of adults in the United States. Int J Eat Disord. 2019;52(1):42-50. doi:10.1002/eat.23004
8. Feltner C, Peat C, Reddy S, et al. Screening for Eating Disorders in Adolescents and Adults: An Evidence Review for the U.S. Preventive Services Task Force. Evidence Synthesis No. 212. Agency for Healthcare Research and Quality; 2022. AHRQ publication 21-05284-EF-1.
9. United Nations Development Programme. Human development index. Published 2016. Accessed August 20, 2018. http://hdr.undp.org/en/composite/HDI
10. Sterne JAC, Savović J, Page MJ, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:l4898. doi:10.1136/bmj.l4898
11. Whiting PF, Rutjes AW, Westwood ME, et al; QUADAS-2 Group. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529-536. doi:10.7326/0003-4819-155-8-201110180-00009
12. Stata release 16.0 [computer software]. StataCorp LP; 2019.
13. US Preventive Services Task Force. US Preventive Services Task Force Procedure Manual. Published May 2021. Accessed February 17, 2022. https://www.uspreventiveservicestaskforce.org/uspstf/procedure-manual
14. Liu CY, Tseng MC, Chen KY, Chang CH, Liao SC, Chen HC. Sex difference in using the SCOFF questionnaire to identify eating disorder patients at a psychiatric outpatient clinic. Compr Psychiatry. 2015;57:160-166. doi:10.1016/j.comppsych.2014.11.014
15. Maguen S, Hebenstreit C, Li Y, et al. Screen for Disordered Eating: improving the accuracy of eating disorder screening in primary care. Gen Hosp Psychiatry. 2018;50:20-25. doi:10.1016/j.genhosppsych.2017.09.004
16. Hill LS, Reid F, Morgan JF, Lacey JH. SCOFF, the development of an eating disorder screening questionnaire. Int J Eat Disord. 2010;43(4):344-351. doi:10.1002/eat.20679
17. Wan Wahida WMZ, Lai PSM, Abdul Hadi H. Validity and reliability of the English version of the Sick, Control, One stone, Fat, Food (SCOFF) in Malaysia. Clin Nutr ESPEN. 2017;18:55-58. doi:10.1016/j.clnesp.2017.02.001
18. Graham AK, Trockel M,Weisman H, et al. A screening tool for detecting eating disorder risk and diagnostic symptoms among college-age women. J Am Coll Health. 2019;67(4):357-366. doi:10.1080/07448481.2018.1483936
19. Dorflinger LM, Ruser CB, Masheb RM. A brief screening measure for binge eating in primary care. Eat Behav. 2017;26:163-166. doi:10.1016/j.eatbeh.2017.03.009
20. Parker SC, Lyons J, Bonner J. Eating disorders in graduate students: exploring the SCOFF questionnaire as a simple screening tool. J Am Coll Health. 2005;54(2):103-107. doi:10.3200/JACH.54.2.103-107
21. Muro-Sans P, Amador-Campos JA, Morgan JF. The SCOFF-c: psychometric properties of the Catalan version in a Spanish adolescent sample. J Psychosom Res. 2008;64(1):81-86. doi:10.1016/j.jpsychores.2007.06.011
22. Garcia FD, Grigioni S, Chelali S, Meyrignac G, Thibaut F, Dechelotte P. Validation of the French version of SCOFF questionnaire for screening of eating disorders among adults. World J Biol Psychiatry. 2010;11(7):888-893. doi:10.3109/15622975.2010.483251
23. Chamay-Weber C, Combescure C, Lanza L, Carrard I, Haller DM. Screening obese adolescents for binge eating disorder in primary care: the Adolescent Binge Eating Scale. J Pediatr. 2017;185:68-72. doi:10.1016/j.jpeds.2017.02.038
24. Luck AJ, Morgan JF, Reid F, et al. The SCOFF questionnaire and clinical interview for eating disorders in general practice: comparative study. BMJ. 2002;325(7367):755-756. doi:10.1136/bmj.325.7367.755
25. Cotton M-A, Ball C, Robinson P. Four simple questions can help screen for eating disorders. J Gen Intern Med. 2003;18(1):53-56. doi:10.1046/j.1525-1497.2003.20374.x
26. Mond JM, Myers TC, Crosby RD, et al. Screening for eating disorders in primary care: EDE-Q versus SCOFF. Behav Res Ther. 2008;46(5):612-622. doi:10.1016/j.brat.2008.02.003
27. Lähteenmäki S, Aalto-Setälä T, Suokas JT, et al. Validation of the Finnish version of the SCOFF questionnaire among young adults aged 20 to 35 years. BMC Psychiatry. 2009;9:5. doi:10.1186/1471-244X-9-5
28. Siervo M, Boschi V, Papa A, Bellini O, Falconi C. Application of the SCOFF, Eating Attitude Test 26 (EAT 26) and Eating Inventory (TFEQ) questionnaires in young women seeking diet-therapy. Eat Weight Disord. 2005;10(2):76-82. doi:10.1007/BF03327528
29. Striegel-Moore RH, Perrin N, DeBar L, Wilson GT, Rosselli F, Kraemer HC. Screening for binge eating disorders using the Patient Health Questionnaire in a community sample. Int J Eat Disord. 2010;43(4):337-343. doi:10.1002/eat.20694
30. Ricca V, Mannucci E, Moretti S, et al. Screening for binge eating disorder in obese outpatients. Compr Psychiatry. 2000;41(2):111-115. doi:10.1016/S0010-440X(00)90143-3
31. Rosenvinge JH, Perry JA, Bjørgum L, Bergersen TD, Silvera DH, Holte A. A new instrument measuring disturbed eating patterns in community populations: development and initial validation of a five-item scale (EDS-5). Eur Eat Disord Rev. 2001;9(2):123-132. doi:10.1002/erv.371
32. Beaulieu K, Casanova N, Oustric P, et al. Matched weight loss through intermittent or continuous energy restriction does not lead to compensatory increases in appetite and eating behavior in a randomized controlled trial in women with overweight and obesity. J Nutr. 2020;150(3):623-633. doi:10.1093/jn/nxz296
33. Guerdjikova AI, Mori N, Blom TJ, et al. Lisdexamfetamine dimesylate in binge eating disorder: a placebo controlled trial. Hum Psychopharmacol. 2016;31(5):382-391. doi:10.1002/hup.2547
34. McElroy SL, Mitchell JE, Wilfley D, et al. Lisdexamfetamine dimesylate effects on binge eating behaviour and obsessive-compulsive and impulsive features in adults with binge eating disorder. Eur Eat Disord Rev. 2016;24(3):223-231. doi:10.1002/erv.2418
35. McElroy SL, Hudson J, Ferreira-Cornwell MC, Radewonuk J, Whitaker T, Gasior M. Lisdexamfetamine dimesylate for adults with moderate to severe binge eating disorder: results of two pivotal phase 3 randomized controlled trials. Neuropsychopharmacology. 2016;41(5):1251-1260. doi:10.1038/npp.2015.275
36. White MA, Grilo CM. Bupropion for overweight women with binge-eating disorder: a randomized, double-blind, placebo-controlled trial. J Clin Psychiatry. 2013;74(4):400-406. doi:10.4088/JCP.12m08071
37. McElroy SL, Hudson JI, Mitchell JE, et al. Efficacy and safety of lisdexamfetamine for treatment of adults with moderate to severe binge-eating disorder: a randomized clinical trial. JAMA Psychiatry. 2015;72(3):235-246. doi:10.1001/jamapsychiatry.2014.2162
38. Arnold LM, McElroy SL, Hudson JI, Welge JA, Bennett AJ, Keck PE. A placebo-controlled, randomized trial of fluoxetine in the treatment of binge-eating disorder. J Clin Psychiatry. 2002;63(11):1028-1033. doi:10.4088/JCP.v63n1113
39. Guerdjikova AI, McElroy SL, Winstanley EL, et al. Duloxetine in the treatment of binge eating disorder with depressive disorders: a placebo-controlled trial. Int J Eat Disord. 2012;45(2):281-289. doi:10.1002/eat.20946
40. McElroy SL, Casuto LS, Nelson EB, et al. Placebo-controlled trial of sertraline in the treatment of binge eating disorder. Am J Psychiatry. 2000;157(6):1004-1006. doi:10.1176/appi.ajp.157.6.1004
41. Walsh BT, Hadigan CM, Devlin MJ, Gladis M, Roose SP. Long-term outcome of antidepressant treatment for bulimia nervosa. Am J Psychiatry. 1991;148(9):1206-1212. doi:10.1176/ajp.148.9.1206
42. Pearlstein T, Spurell E, Hohlstein LA, et al. A double-blind, placebo-controlled trial of fluvoxamine in binge eating disorder: a high placebo response. Arch Womens Ment Health. 2003;6(2):147-151. doi:10.1007/s00737-003-0172-8
43. Fluoxetine Bulimia Nervosa Collaborative Study Group. Fluoxetine in the treatment of bulimia nervosa: a multicenter, placebo-controlled, double-blind trial. Arch Gen Psychiatry. 1992;49(2):139-147. doi:10.1001/archpsyc.1992. 01820020059008
44. McElroy SL, Hudson JI, Capece JA, Beyers K, Fisher AC, Rosenthal NR; Topiramate Binge Eating Disorder Research Group. Topiramate for the treatment of binge eating disorder associated with obesity: a placebo-controlled study. Biol Psychiatry. 2007;61(9):1039-1048. doi:10.1016/j.biopsych.2006.08.008
45. Laederach-Hofmann K, Graf C, Horber F, et al. Imipramine and diet counseling with psychological support in the treatment of obese binge eaters: a randomized, placebo-controlled double-blind study. Int J Eat Disord. 1999;26(3):231-244. doi:10.1002/(SICI)1098-108X(199911)26:3<231::AIDEAT1> 3.0.CO;2-6
46. Guerdjikova AI, McElroy SL, Kotwal R, et al. High-dose escitalopram in the treatment of binge-eating disorder with obesity: a placebo-controlled monotherapy trial. Hum Psychopharmacol. 2008;23(1):1-11. doi:10.1002/hup.899
47. Sheehan DV, Gasior M, McElroy SL, Radewonuk J, Herman BK, Hudson J. Effects of lisdexamfetamine dimesylate on functional impairment measured on the Sheehan Disability Scale in adults with moderate-to-severe binge eating disorder: results from two randomized, placebo-controlled trials. Innov Clin Neurosci. 2018;15(5-6):22-29.
48. McElroy SL, Arnold LM, Shapira NA, et al. Topiramate in the treatment of binge eating disorder associated with obesity: a randomized, placebo-controlled trial. Am J Psychiatry. 2003;160 (2):255-261. doi:10.1176/appi.ajp.160.2.255
49. Kaneva R, Rissanen A, Sarna S. Fluoxetine in the treatment of anxiety, depressive symptoms, and eating-related symptoms in bulimia nervosa. Nord J Psychiatry. 1995;49(4):237-242. doi:10.3109/08039489509011912
50. Grilo CM, Masheb RM, Wilson GT. Efficacy of cognitive behavioral therapy and fluoxetine for the treatment of binge eating disorder: a randomized double-blind placebo-controlled comparison. Biol Psychiatry. 2005;57(3):301-309. doi:10.1016/j.biopsych.2004.11.002
51. Mitchell JE, Fletcher L, Hanson K, et al. The relative efficacy of fluoxetine and manual-based self-help in the treatment of outpatients with bulimia nervosa. J Clin Psychopharmacol. 2001;21(3):298-304. doi:10.1097/00004714-200106000-00008
52. Cachelin FM, Gil-Rivas V, Palmer B, et al. Randomized controlled trial of a culturally-adapted program for Latinas with binge eating. Psychol Serv. 2019;16(3):504-512. doi:10.1037/ser0000182
53. Grilo CM, White MA, Gueorguieva R, Barnes RD, Masheb RM. Self-help for binge eating disorder in primary care: a randomized controlled trial with ethnically and racially diverse obese patients. Behav Res Ther. 2013;51(12):855-861. doi:10.1016/j.brat.2013.10.002
54. Grilo CM, Masheb RM, White MA, et al. Treatment of binge eating disorder in racially and ethnically diverse obese patients in primary care: randomized placebo-controlled clinical trial of self-help and medication. Behav Res Ther. 2014;58:1-9. doi:10.1016/j.brat.2014.04.002
55. Green MA, Kroska A, Herrick A, et al. A preliminary trial of an online dissonance-based eating disorder intervention. Eat Behav. 2018;31:88-98. doi:10.1016/j.eatbeh.2018.08.007
56. Deal LS, Wirth RJ, Gasior M, Herman BK, McElroy SL. Validation of the Yale-Brown Obsessive Compulsive Scale modified for binge eating. Int J Eat Disord. 2015;48(7):994-1004. doi:10.1002/eat.22407
57. Wade S, Byrne S, Allen K. Enhanced cognitive behavioral therapy for eating disorders adapted for a group setting. Int J Eat Disord. 2017;50(8):863-872. doi:10.1002/eat.22723
58. Wagner B, Nagl M, Dölemeyer R, et al. Randomized controlled trial of an internet-based cognitive-behavioral treatment program for binge-eating disorder. Behav Ther. 2016;47(4):500-514. doi:10.1016/j.beth.2016.01.006
59. Sánchez-Ortiz VC, Munro C, Stahl D, et al. A randomized controlled trial of internet-based cognitive-behavioural therapy for bulimia nervosa or related disorders in a student population. Psychol Med. 2011;41(2):407-417. doi:10.1017/S0033291710000711
60. Kelly AC, Carter JC. Self-compassion training for binge eating disorder: a pilot randomized controlled trial. Psychol Psychother. 2015;88(3):285-303. doi:10.1111/papt.12044
61. Ljotsson B, Lundin C, Mitsell K, Carlbring P, Ramklint M, Ghaderi A. Remote treatment of bulimia nervosa and binge eating disorder: a randomized trial of Internet-assisted cognitive behavioural therapy. Behav Res Ther. 2007;45(4): 649-661. doi:10.1016/j.brat.2006.06.010
62. Fairburn CG, Cooper Z, Doll HA, et al. Transdiagnostic cognitive-behavioral therapy for patients with eating disorders: a two-site trial with 60-week follow-up. Am J Psychiatry. 2009;166(3): 311-319. doi:10.1176/appi.ajp.2008.08040608
63. Carter JC, Kenny TE, Singleton C, Van Wijk M, Heath O. Dialectical behavior therapy self-help for binge-eating disorder: a randomized controlled study. Int J Eat Disord. 2020;53(3):451-460. doi:10.1002/eat.23208
64. Hill DM, Craighead LW, Safer DL. Appetite-focused dialectical behavior therapy for the treatment of binge eating with purging: a preliminary trial. Int J Eat Disord. 2011;44(3):249- 261. doi:10.1002/eat.20812
65. Alfonsson S, Parling T, Ghaderi A. Group behavioral activation for patients with severe obesity and binge eating disorder: a randomized controlled trial. Behav Modif. 2015;39(2):270-294. doi:10.1177/0145445514553093
66. Carter JC, Olmsted MP, Kaplan AS, McCabe RE, Mills JS, Aimé A. Self-help for bulimia nervosa: a randomized controlled trial. Am J Psychiatry. 2003;160(5):973-978. doi:10.1176/appi.ajp.160.5.973
67. Schag K, Rennhak SK, Leehr EJ, et al. IMPULS: impulsivity-focused group intervention to reduce binge eating episodes in patients with binge eating disorder—a randomised controlled trial. Psychother Psychosom. 2019;88(3):141-153. doi:10.1159/000499696
68. Schmidt U, Andiappan M, Grover M, et al. Randomised controlled trial of CD-ROM–based cognitive-behavioural self-care for bulimia nervosa. Br J Psychiatry. 2008;193(6):493-500. doi:10.1192/bjp.bp.107.046607
69. Wilfley DE, Agras WS, Telch CF, et al. Group cognitive-behavioral therapy and group interpersonal psychotherapy for the nonpurging bulimic individual: a controlled comparison. J Consult Clin Psychol. 1993;61(2):296-305. doi:10.1037/0022-006X.61.2.296
70. Schlup B, Munsch S, Meyer AH, Margraf J, Wilhelm FH. The efficacy of a short version of a cognitive-behavioral treatment followed by booster sessions for binge eating disorder. Behav Res Ther. 2009;47(7):628-635. doi:10.1016/j.brat.2009.04.003
71. Telch CF, Agras WS, Rossiter EM, Wilfley D, Kenardy J. Group cognitive-behavioral treatment for the nonpurging bulimic: an initial evaluation. J Consult Clin Psychol. 1990;58(5):629-635. doi:10.1037/0022-006X.58.5.629
72. Masson PC, von Ranson KM, Wallace LM, Safer DL. A randomized wait-list controlled pilot study of dialectical behaviour therapy guided self-help for binge eating disorder. Behav Res Ther. 2013;51(11):723-728. doi:10.1016/j.brat.2013.08.001
73. Laessle RG, Waadt S, Pirke KM. A structured behaviorally oriented group treatment for bulimia nervosa. Psychother Psychosom. 1987;48(1-4):141-145. doi:10.1159/000288044
74. Debar LL, Wilson GT, Yarborough BJ, et al. Cognitive behavioral treatment for recurrent binge eating in adolescent girls: a pilot trial. Cogn Behav Pract. 2013;20(2):147-161. doi:10.1016/j.cbpra.2012.04.001
75. Palmer RL, Birchall H, McGrain L, Sullivan V. Self-help for bulimic disorders: a randomised controlled trial comparing minimal guidance with face-to-face or telephone guidance. Br J Psychiatry. 2002;181:230-235. doi:10.1192/bjp.181.3.230

Return to Table of Contents

Figure 1 depicts the key questions within the context of the eligible populations, screenings/interventions, comparisons, outcomes, and settings. On the left, the population of interest is adults and adolescents without recognized signs or symptoms of eating disorders. Moving from left to right, the figure illustrates the overarching question: Does screening for eating disorders in adults and adolescents improve health outcomes (KQ 1)? The figure depicts the pathway from screening to reduction in the morbidity and mortality of eating disorders to illustrate the question: What is the accuracy of primary care-relevant screening tools for eating disorders in adolescents and adults (KQ 2). Screening may result in harms (KQ 3). After early detection of eating disorders, the figure illustrates the following questions: How effective are interventions for improving health outcomes for screen-detected or previously untreated adults and adolescents with eating disorders and do interventions improve health outcomes compared with no interventions, usual care, or different treatment targets (KQ 4)? Treatment may result in harms (KQ 5).

Evidence reviews for the US Preventive Services Task Force (USPSTF) use an analytic framework to visually display the key questions that the review will address to allow the USPSTF to evaluate the effectiveness and safety of a preventive service. The questions are depicted by linkages that relate interventions and outcomes. For additional information see the USPSTF Procedure Manual.13

Return to Table of Contents

Figure 3 displays a forest plot reporting the mean differences comparing lisdexamfetamine or topiramate to control on the outcomes of eating disorder symptom severity, depression, and anxiety (KQ 4). Lisdexamfetamine was associated with a reduction in eating disorder symptom severity measured by the YCBOCS-BE (pooled standardized mean difference, -5.75[95% CI, -8.32 to -3.17]; I2 = 82.75; 4 trials). Two trials of  topiramate yielded mean differences ranging between -2.55 (95% CI, -4.29 to -0.81) and -6.4 (95% CI, -8.14 to -4.66) in eating disorder symptom severity measured by the YCBOCS-BE. One trial of topiramate yielded a mean difference of -0.55 (95% CI, -1.57 to 0.46) on the HAM-D.  One trial of topiramate yielded a mean difference of -0.50 (95% CI, -1.79 to 0.79) on the MADRS.  One trial of topiramate yielded a mean difference of -0.60 (95% CI, -1.52 to 0.32) on the HAM-A.

HAM-A indicates Hamilton Anxiety Rating Scale; HAM-D, Hamilton Rating Scale for Depression; KQ, key question; MADRS, Montgomery-Åsberg Depression Rating Scale; YBOCS-BE, Yale-Brown Obsessive Compulsive Scale modified for binge eating.

Return to Table of Contents

Figure 4 displays a forest plot reporting the standardized mean differences comparing self-help interventions to control on the outcomes of eating disorder symptom severity and depression (KQ 4). Guided self-help was associated with a reduction in the EDE-Q and EDE (pooled standardized mean difference, -0.96 [95% CI, -1.26 to -0.67]; I2 = 46.37; 5 trials). Six trials of unguided self-help yielded a pooled standardized mean difference of  -0.18 (95% CI, -0.38 to 0.03; I2 = 0.00). Guided self-help was associated with a reduction in depression symptoms measured by the BDI-II, BDI, MADRS, or HADS-D (pooled standardized mean difference, -0.73 [95% CI, -1.04 to -0.43]; I2=41.88; 4 trials). Unguided self-help was associated with a reduction in depression symptoms measured by BDI (pooled standardized mean difference, -0.37 [95% CI, -0.68 to -0.05]; I2 = 0.00; 3 trials).

BDI indicates Beck Depression Inventory; CBT, cognitive behavioral therapy; CFT, compassion-focused therapy-based self-help; DBT, dialectical behavioral therapy; EDE, Eating Disorder Examination; EDE-Q, EDE Questionnaire; HADS-Dep, Hospital Anxiety and Depression Scale–Depression; KQ, key question; MADRS, Montgomery-Åsberg Depression Rating Scale.
a Cognitive-dissonance-based intervention (http://www.bodyprojectsupport.org/).

Return to Table of Contents

Source, country Quality Screener Reference standard: eating disorder diagnoses assessed Recruitment
setting
Population Eating disorder prevalence, %a Age, mean (SD), y Female, % Non-White, % BMI, mean (SD)
Cohort studies
Lui et al,14 2015
Taiwan
Good SCOFF SCID (DSM-IV): anorexia nervosa, bulimia nervosa, BED, EDNOS Outpatient psychiatric clinics 1541 Adults (18-45 y) recruited at their first outpatient psychiatric visit Any eating disorder: 16 31 (7.9) 61 NRb 22.2 (5.4)
Graham et al,18,32 2019
US
Good SWED EDE (DSM-5): anorexia nervosa, bulimia nervosa, BED University campuses 549 College-age women (18-25 y) responding to recruitment ads and flyers for an eating disorder prevention trial Any eating disorder: 19cd 21 (1.97) 100 44d 24.5 (5.02)
Dorflinger et al,19 2017
US
Good VA-BES QEWP-R: BED VHA medical center 116 Veterans recruited at primary care–based weight management group BED: 8 62 (8.73) 11 26 37.9 (7.35)
Rosenvinge et al,31 2001
Norway
Fair EDS-5 SCID (DSM-III-R): any eating disorder University campuses 51 College-age women (20-42 y) recruited at their teaching and nursing colleges CED: 20d 25.2 (5.33) 100 NR NR
Mond et al,26 2008
US
Fair SCOFF EDE Primary care practices 147 Adult women (18-40 y) recruited at their primary care visit Any eating disorder: 17 28 (6.50) 100 12 28.10 (7.20)
Cotton et al,25 2003
UK
Good SCOFF, EDS-PC QEDD University campuses and primary care 225 Students (18-65 y) recruited from posters and lecture announcements and adults (18-65 y) recruited at a primary care visit Any eating disorder: 12 29 77 NR 22
Lähteenmäki et al,27 2009
Finland
Fair SCOFF SCID (DSM-IV): anorexia nervosa, bulimia nervosa, EDNOS Households 541 Young adults recruited via mail Current anorexia nervosa, bulimia nervosa, EDNOS: 1d

Lifetime anorexia nervosa, bulimia nervosa, EDNOS: 4d

NR NR NR NR
Cross-sectional studies
Maugen et al,15 2018
US
Good EDS-PC, SCOFF, SDE EDE-Q (DSM-5e): anorexia nervosa, bulimia nervosa, BED VHA medical center 402 Female veterans (18-70 y) responding to mailed questionnaires Any eating disorder: 16d 49 (NR)d 100 52f NR
Chamay-Weber et al,23 2017
Switzerland
Good ADO-BED SCID (DSM-IV): BED Outpatient pediatric obesity center 94 Adolescents (12-18 y) recruited at their outpatient pediatric visit BED:
   Sub: 28
   Full: 22
   Overall: 50
Median (range): 14 (11-18) 60 NR NR
Luck et al,24 2002; Hill et al,16 2010
UK
Good SCOFF Clinical interview (DSM-IV): anorexia nervosa, bulimia nervosa, EDNOS Primary care practices 341 Women (18-50 y) attending primary care practices Any eating disorder: 4d NR 100 NR NR
Siervo et al,28 2005
Italy
Fair SCOFF “Clinical diagnosis” (DSM-IV): bulimia nervosa, BED Outpatient diet clinics 162 Women (16-35 y) recruited at an outpatient dietetic clinic Any eating disorder: 46d 24d 100 NR 29.6d
Parker et al,20 2005
US
Good SCOFF EDE-Q (DSM-IV): anorexia nervosa, bulimia nervosa, EDNOS University health center 297 Adults (20-51 y) recruited at their campus health visit Any eating disorder: 20 <23 y: 10
23-26 y: 66
>26 y: 23
72 33 Range: 16-44
Ricca et al,30 2000
Italy
Fair BES SCID (DSM-IV): bulimia nervosa, BED Outpatient clinic for metabolic diseases 344 Patients recruited at an outpatient clinic for metabolic diseases including obesity BED: 8 43.5 (13.6) 83 NR 35.8 (6.1)
Wan Wahida et al,17 2017
Malaysia
Good SCOFF EAT-26 University 292 Undergraduate students (18-22 y) who understood English Any eating disorder: 11 20 (0.5) 65 Malay: 44
Chinese: 42
Indian: 14
NR
Garcia et al,22 2010
France
Good SCOFF MINI (DSM-IV-TR): any eating disorder, anorexia nervosa, bulimia nervosa University clinic 400 Female undergraduate students (18-35 y) Any eating disorder: 9 21 (2.5) 100 14f 21.98 (3.5)
Muro-Sans et al,21 2008
Spain
Good SCOFF EDI-2 Primary and secondary schools 954 Adolescents (10-17 y) recruited from schools Any eating disorder: 8d 14 (1.31) 49 NR NR
Striegel-Moore et al,29 2010
US
Fair PHQ-ED EDE (DSM-IV): bulimia nervosa, BED Health maintenance organization 348 Adults (18-35 y) selected from the EHR of an HMO via letter Bulimia nervosa or BED: 8d,e 28 (5.38) 82 13 NR

Abbreviations: ADO-BED, Adolescent Binge-Eating Disorder Questionnaire; BED, binge eating disorder; BES, Binge Eating Scale; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CED, clinical eating disorder; DSM-III-R, Diagnostic and Statistical Manual of Mental Disorders (Third Edition Revised); DSM-IV, Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition); DSM-IV-TR, Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition, Text Revision); DSM-5, Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition); EAT-26, Eating Attitudes Test; EDE, Eating Disorder Examination; EDE-Q, Eating Disorder Examination Questionnaire; EDI-2, Eating Disorder Inventory 2; EDNOS, eating disorder not otherwise specified; EDS-5, Eating Disturbance Scale 5; EDS-PC, Eating Disorder Screen for Primary Care; EHR, electronic health record; HMO, health maintenance organization; KQ, key question; MINI, Mini International Neuropsychiatric Interview; NR, not reported; PHQ-ED, eating disorder module of the Patient Health Questionnaire; QEDD, Questionnaire for Eating Disorder Diagnoses; QEWP-R, Questionnaire of Eating and Weight Patterns–Revised; SCID, Structured Clinical Interview for DSM Disorders; SDE, Screen for Disordered Eating; SWED, Stanford-Washington University Eating Disorder screen; VA-BES, Veterans Affairs Binge Eating Screener; VHA, Veterans Health Administration.
a Full refers to meeting the full diagnostic criteria for a given eating disorder; sub refers to a subthreshold condition definition.
b Conducted in Taiwan and required to understand Mandarin.
c Also conducted analysis including subthreshold bulimia nervosa, BED, and purging disorder (in addition to threshold anorexia nervosa, bulimia nervosa, and BED).
d Computed by data abstractors.
e BED defined as an average of 1 or more objective binge episodes per week without compensatory or purging behaviors.
f Enrolled all screen-positive participants and a random sample of those who screened negative.

Return to Table of Contents

Screener (cut point) Eating disorder diagnosis No. of studies
(No. of participants)
% (95% CI) LR+ (95% CI) LR- (95% CI)
Sensitivity Specificity
SCOFF
   ≥3 Any 7 (2749) Pooled: 69 (56-80) Pooled: 90 (69-98) Pooled: 7.3 (2.2-24.0) Pooled: 0.34 (0.25-0.46)
   ≥2 Any 10 (3684) Pooled: 84 (74-90) Pooled: 80 (65-89) Pooled: 4.1 (2.3-7.3) Pooled: 0.20 (0.12-0.33)
SWED (>59)18 Any 1 (549) 80 (NR) 82 (NR) NR NR
EDS-PC (≥2)15, 25 Any 2 (627) 97 (88-100) 40 (35-46) NR NR
100 (90-100) 71 (64-77)
SDE (≥2)15 Any 1 (402) 91 (80-96) 58 (80-96) NR NR
EDS-5 (≥16)31 Any 1 (51) 90 (NR) 88 (NR) NR NR
PHQ-ED (NA)29 BN, BED 1 (348) 100 (NR) 30 (NR)a    
ADO-BED (NA)23 BED 1 (94 adolescents) 100 (NR) 27 (NR) NR NR
VA-BES (≥1)19 BED 1 (162) 89 (NR) 65 (NR) NR NR
BES (≥17)30 BED 1 (344) 85 (NR) 75 (NR) NR NR

Abbreviations: ADO-BED, Adolescent Binge-Eating Disorder Questionnaire; BED, binge eating disorder; BES, Binge Eating Scale; BN, bulimia nervosa; EDS-5, Eating Disturbance Scale 5; EDS-PC, Eating Disorder Screen for Primary Care; KQ, key question; LR, likelihood ratio; NR, not reported; PHQ-ED, eating disorder module of the Patient Health Questionnaire; SDE, Screen for Disordered Eating; SWED, Stanford-Washington University Eating Disorder screen; VA-BES, Veterans Affairs Binge Eating Screener.
a Value calculated based on individual cell frequencies differs from reported specificity value reported in study (91.7% vs 27.7%, respectively).

Return to Table of Contents

Topic No. of studies/study designs
(No. of participants)
Summary of findings Consistency and precision Study quality Limitations (including reporting bias) Overall strength of evidence Applicability
KQ1: Benefits of screening
  0 No eligible studies NA NA NA Insufficient NA
KQ2: Accuracy of screening tests for detecting eating disorders
  SCOFF (≥2): 10 (3684) Pooled sensitivity, 84% (95% CI, 74%-90%); specificity, 80% (95% CI, 65%-89%) Consistent and precise for sensitivity; inconsistent and imprecise for specificitya 7 Good; 3 Fair Potential bias related to participant selection

Reference standards varied across studies

Moderate for adequate accuracy Studies enrolled adults and either limited to women or enrolled a majority of women

Several studies enrolled from specialty clinics or college campuses

  SCOFF (≥3): 7 (2749) Pooled sensitivity, 69% (95% CI, 56%-80%); specificity, 90% (95% CI, 69%-98%) Inconsistent and imprecise for both sensitivity and specificityb 4 Good; 3 Fair Potential bias related to participant selection

Reference standards varied
across studies

Low for adequate accuracy All studies enrolled adults and either
limited to women or enrolled a majority of women

Several studies enrolled from specialty clinics or college campuses

  EDS-PC (≥2): 2 (627) Sensitivity, 97% (95% CI, 88%-100%); 100% (95% CI, 90%-100%)

Specificity, 40% (95% CI, 35%-46%); 71% (95% CI, 64%-77%)

Consistent and precise for sensitivity; inconsistent and imprecise for specificity Good Studies used different reference standards and enrolled diverse populations Insufficient One study recruited females and males from primary care and college campuses in the UK (77% females), and the other recruited female US veterans
KQ3: Harms of screening
  0 No eligible studies NA NA NA Insufficient NA
KQ4: Benefits and harms of interventions for screen-detected or previously untreated ED
Benefits of pharmacotherapy Lisdexamfetamine (BED): 4 (900) Pooled mean difference for reduction in YBOCS-BE scores larger in lisdexamfetamine group vs placebo (−5.75 [95% CI, −9.32 to −3.17])

Other outcomes assessed by 1 trial each (depression, anxiety, QOL, and function)

YBOCS-BE: consistent, precise

Other health outcomes: unknown consistency and imprecise

Fair Outcomes assessed over relatively short duration (11-12 wk) Moderate for benefit in eating disorder symptom severity; insufficient for other health outcomes Studies enrolled adults with BED and obesity recruited via study advertisements
Topiramate (BED): 2 (465) Larger reduction in YBOCS-BE in topiramate groups vs placebo; difference between groups in score change, −6.40 (P < .001) and −2.55 (P = .004)

Other outcomes assessed by 1 trial each (depression, anxiety)

YBOCS-BE; consistent, imprecisec

Other outcomes: unknown consistency, imprecise

Fair Outcomes assessed over a relatively short duration (14-16 wk) Low for benefit in eating disorder symptom severity; insufficient for other outcomes Studies enrolled adults with BED and obesity recruited via study advertisements
SSRIs (BED) 5 (208) Two reported on eating disorder symptom severity: fluoxetine (EDE-Q) SMD, −0.69 (95% CI, −1.30 to −0.08) and escitalopram (YBOCS-BE) SMD, −0.29 (95% CI, −0.83 to −0.24)

Larger reduction in depression symptoms among SSRI groups vs placebo (5 trials): pooled SMD, −0.61 (95% CI, −0.90 to −0.33)

Eating disorder symptom severity: unknown consistency, imprecised

Depression: consistent, imprecise

Fair Studies assessed different SSRIs and reported outcomes over 6-16 wk

Study eligibility criteria varied in terms of body weight and duration/frequency of binge eating episodes

Insufficient for eating disorder symptom severity

Low for benefit in depression symptom severity

Studies enrolled adults with BED, most recruited via advertisements. Two limited to populations that were obese, and 1 limited to those with concurrent depression
  Fluoxetine (bulimia nervosa): 3 (528) Two found larger reduction in EAT scores among fluoxetine group vs placebo; difference was statistically significant in 1 trial

Two found larger reductions in HAM-D scores among fluoxetine vs placebo; difference was statistically significant in 1 triale

Eating disorder symptom severity: consistent; imprecise

Depression symptom severity: consistent; imprecise

Fair Studies reported outcomes at different durations (8 and 16 wk) Low for benefit (eating disorder and depression symptom severity) All enrolled populations with bulimia nervosa recruited via advertisements; 1 limited to those with bulimia nervosa and recurrent binge eating
Benefits of therapy interventions Guided self-help: 7 (431) Guided self-help reduced eating disorder symptom severity more than control (5 studies; n = 391): pooled SMD, −0.96 (95% CI, −1.26 to −0.67)

Guided self-help reduced depression symptoms more than control (4 studies; n = 324): pooled SMD, −0.73 (95% CI, −1.04 to −0.43)

Eating disorder symptom severity: consistent; precise

Depression symptom severity: consistent; precise

Fair Frequency and mode of delivering guidance varied (eg, emails, individual sessions); studies assessed eating disorder and depression symptoms using different measures over a relatively short duration (8-16 wk) Moderate for benefit (eating disorder and depression symptom severity) All enrolled adults with BED recruited primarily via advertisements; several limited to populations that were obese
  Unguided self-help: 7 (421) Pooled results (6 studies; n = 368) favored self-help for reduction in eating disorder symptom severity but difference was not statistically significant (SMD, −0.18 [95% CI, −0.38 to 0.03])

Unguided self-help reduced depression symptoms more than control (3 studies; n = 156) (SMD, 0.37 [95% CI, −0.68 to −0.05])

Eating disorder symptom severity: consistent; imprecise

Depression symptom severity: consistent; imprecise

Fair Studies assessed eating disorder and depression symptoms using various measures over a relatively short duration (8-16 wk)

Content and underlying theory of some interventions varied

Low for benefit (eating disorder and depression symptom severity) All enrolled adults with BED recruited primarily via advertisements; several limited to populations that were obese
  Group interventions: 7 (253) Group therapy reduced depression symptoms more than control (7 studies; n = 253) (pooled SMD, −0.48 [95% CI, −0.69 to −0.27])

Three measured eating disorder symptom severity using the EDE-Q; 1 found a statistically significant benefit vs control (SMD, −1.01) and 2 found no significant differences between groups (SMD, −0.10 and −0.30)

Eating disorder symptom severity: consistent; precise

Depression symptom severity: consistent; precise

Fair Type of group therapy differed across studies (eg, CBT-based, interpersonal therapy)

Outcomes were measured over a relatively short duration (8-16 wk)

Number, length, and frequency of sessions varied

Moderate for benefit in depression symptom severity; insufficient for eating disorder symptom severity All enrolled adults with BED recruited primarily via advertisements; several limited to populations that were obese
  Individual interventions: 4 (319) One trial assessing 2 forms of individual CBT found no significant differences between groups on EDE-Q scores; 1 trial of AF-DBT found significant improvement in EDE-Q scores (SMD, −1.18 [95% CI, −1.94 to −0.43])

One trial of CBT found improvement in BDI scores (SMD, −0.60 [95% CI, −1.14 to −0.06]) and a trial of DBT-AF found significant improvement in BDI-II scores (SMD, −0.92 [95% CI, −1.65 to −0.19])

One trial limited to adolescents found no significant improvement in depression (BDI scores) or psychosocial function (SCARED scores)

Unknown consistency; imprecise Fair Trials addressed different types of individual therapy (eg, CBT, DBT) and reported on different outcomes over a relatively short duration (6 to 16 wk) Insufficient All enrolled adults with BED (or BED and bulimia nervosa) referred or recruited via trial advertisements
Harms of pharmacotherapy 9 (2006) Lisdexamfetamine (4 studies) is associated with higher rates of dry mouth, headache, and insomnia vs placebo

Topiramate (2 studies) was associated with significantly higher rates of paresthesia, taste, and difficulty with concentration/confusion vs placebo

Trials of other medications were assessed by only 1 study each, and results were imprecise

Lisdexamfetamine: consistent; imprecise

Topiramate; consistent, imprecise

Other medications: unknown consistency; imprecise

Fair Some trials did not prespecify adverse events or describe how they were ascertained; trials assessed adverse events over a relatively short duration Topiramate and lisdexamfetamine: Moderate for increased rates of various adverse effects

Other medications: insufficient

All studies enrolled adults with BED and obesity, recruited via referrals or study advertisements

Most studies of lisdexamfetamine were limited to populations without ADHD, substance abuse, or other psychiatric comorbidity

Harms of therapy interventions 0 No eligible studies NA NA NA Insufficient NA

Abbreviations: ADHD, attention-deficit/hyperactivity disorder; AF-DBT, appetite focused-dialectical behavior therapy; BDI, Beck Depression Inventory; BED, binge eating disorder; CBT, cognitive behavioral therapy; DBT, dialectical behavior therapy; DBT-AF, dialectical behavior therapy, appetite focused; EAT, Eating Attitudes Test; EDE-Q, Eating Disorder Examination Questionnaire; EDS-PC,  Eating Disorder Screen for Primary Care; HAM-D, Hamilton Depression Rating Scale–Depression; KQ, key question; NA, not applicable; QOL, quality of life; SCARED, Screen for Child Anxiety-Related Emotional Disorders; SMD, standardized mean difference; SSRI, selective serotonin reuptake inhibitor; YBOCS-BE, Yale-Brown Obsessive Compulsive Scale modified for binge eating.
a Based on eFigure 1 in the JAMA Supplement, the 95% prediction region indicates the results are mostly consistent for sensitivity and somewhat inconsistent for specificity; based on the 95% confidence region, estimates are precise for sensitivity and somewhat imprecise for specificity.
b Based on eFigure 2 in the JAMA Supplement, the 95% prediction region indicates results are inconsistent; based on the 95% confidence region, estimates are imprecise.
c Difference between groups in mean YBOCS-BE met threshold considered to be a minimum clinically important change in only 1 of 2 studies.
d Although results were in same direction of effect (favoring SSRI), only 2 studies assessed change in eating disorder symptom reduction. Each assessed a different medication using different measures of eating disorder symptom burden (YBOCS-BE vs EDE-Q) and reported outcomes at slightly different durations (12 vs 16 weeks), limiting ability to assess consistency for this outcome.
e One additional trial (n = 42) assessed fluoxetine 60 mg/d for bulimia nervosa and reported no significant difference between groups for eating disorder symptom severity (EDI) and depression (HAM-D), P > .05.

Return to Table of Contents