r in

c,f, S g,D.ark

nt C anana

n o cal

Presented at the Society for Adolescent Health and Medicine Annual Meeting,Atlanta, Georgia, March 13e16, 2013; 2013 International Conference on EatingDisorders, Montreal, Canada. May 2e4, 2013; Resident Research Day, Hospital for

* Address correspondence to: Martin M. Fisher, M.D., Division of AdolescentMedicine, Cohen Childrens Medical Center, 410 Lakeville Road, Suite 108, NewHyde Park, NY 11042.

E-mail address: sher@nshs.edu (M.M. Fisher).

www.jahonline.org

Journal of Adolescent Health 55 (2014) 49e52Sick Children and University of Toronto, May 22, 2013. 2014 Society for Adolescent Health and Medicine. All rights reserved.less likely to have a mood disorder (19% vs. 31% vs. 58%).Conclusions: Patients with ARFID were demographically and clinically distinct from those with ANor BN. They were signicantly underweight with a longer duration of illness and had a greaterlikelihood of comorbid medical and/or psychiatric symptoms.

treatment.A B S T R A C T

Purpose: To evaluate the DSM-5 diagnosis of Avoidant/Restrictive Food Intake Disorder (ARFID) inchildren and adolescents with poor eating not associated with body image concerns.Methods: A retrospective case-control study of 8e18-year-olds, using a diagnostic algorithm,compared all cases with ARFID presenting to seven adolescent-medicine eating disorder programsin 2010 to a randomly selected sample with anorexia nervosa (AN) and bulimia nervosa (BN).Demographic and clinical information were recorded.Results: Of 712 individuals studied, 98 (13.8%) met ARFID criteria. Patients with ARFID wereyounger than those with AN (n 98) or BN (n 66), (12.9 vs. 15.6 vs. 16.5 years), had longerdurations of illness (33.3 vs. 14.5 vs. 23.5 months), were more likely to be male (29% vs. 15% vs. 6%),and had a percent median body weight intermediate between those with AN or BN (86.5 vs. 81.0and 107.5). Patients with ARFID included those with selective (picky) eating since early childhood(28.7%); generalized anxiety (21.4%); gastrointestinal symptoms (19.4%); a history of vomiting/choking (13.2%); and food allergies (4.1%). Patients with ARFID were more likely to have a co-morbid medical condition (55% vs. 10% vs. 11%) or anxiety disorder (58% vs. 35% vs. 33%) and were

IMPLICATIONS ANDCONTRIBUTION

AdolescentswithARFIDaredemographically and clini-cally distinct from thosewith AN or BN. They aresignicantly underweight,oftenwith associatedmed-icaland/orpsychiatricsymp-toms. This report supportsthe potential clinical utilityof ARFID and underlinesthe need for additional re-search to clarify course,outcome, and response toManual (DSM-5); Anorexia nervosa (AN); Bulimia nervosa (BN)Original article

Characteristics of Avoidant/Restrictive Food Intake DisordeChildren and Adolescents: A New Disorder in DSM-5

Martin M. Fisher, M.D. a,*, David S. Rosen, M.D. b, Rollyn M. Ornstein, M.D.Kathleen A. Mammel, M.D. d, Debra K. Katzman, M.D. e, Ellen S. Rome, M.D.Joan Malizio, R.N., M.N. a, Sarah Kearney, M.D. e, and B. Timothy Walsh, M.aDivision of Adolescent Medicine, Cohen Childrens Medical Center, North ShoreeLong Island Jewish Health System, New Hyde PbDepartments of Pediatrics, Internal Medicine, and Psychiatry, University of Michigan Medical School, Ann Arbor, MichigancDivision of Adolescent Medicine and Eating Disorders, Penn State Hershey Childrens Hospital, Hershey, PennsylvaniadDepartment of Pediatrics, Oakland University William Beaumont School of Medicine, Division of Adolescent Medicine, BeaumoeDivision of Adolescent Medicine, Department of Pediatrics, Hospital for Sick Children, University of Toronto, Toronto, Ontario, CfCenter for Adolescent Medicine, Cleveland Clinic Childrens Hospital, Cleveland, OhiogDivision of Adolescent and Young Adult Health, Monroe Carell Jr. Childrens Hospital at Vanderbilt, Nashville, TennesseehDepartment of Psychiatry, New York State Psychiatric Institute, Columbia University Medical Center, New York, New York

Article history: Received September 12, 2013; Accepted November 19, 2013Keywords: Avoidant/Restrictive food intake disorder (ARFID); Children and adolescents; 5th Editio1054-139X/$ e see front matter 2014 Society for Adolescent Health and Medicine. All rights reserved.http://dx.doi.org/10.1016/j.jadohealth.2013.11.013. Todd Callahan, M.D.h

, New York

hildrens Hospital, Royal Oak, Michigda

f the diagnostic and statisti

eating since early childhood, generalized anxiety, fear of vomit-ing or choking, gastrointestinal symptoms, food allergies, andother reasons.

Data were entered into EXCEL, and analyzed in aggregate byone of the authors (B.T.W.). Chi-square analyses were used forcategorical variables and ANOVA for continuous variables. Post-hoc testing was performed to assess statistical signicance be-tween ARFID and AN groups and between ARFID and BN groupswhen there was overall statistical signicance. For post-hoctesting, Tukeys Honestly Signicant Difference (HSD) was usedfor continuous variables and 2-by-2 Chi-square analysis withBonferroni correctionwas used for categorical variables. Statisticalanalysis was conducted using Stata 12.1 (StataCorp LP, CollegeStation, TX).

Institutional review boards/research ethics boards at each ofthe participating study sites approved the study.

Results

A total of 712 new patients with eating disorders presented tothe seven study sites during 2010. Of these, 98 patients (13.8%)

M.M. Fisher et al. / Journal of Adolescent Health 55 (2014) 49e5250A signicant number of children and adolescents who presentwith feeding or eating disorders cannot be classied using theDSM-IV criteria [1e3]. Patients who did not meet the DSM-IVcriteria for anorexia nervosa (AN) or bulimia nervosa (BN) wereoften given a diagnosis of eating disorder not otherwise specied(EDNOS). Using the DSM-IV, over 50% of children and adolescentswith eating disorders were diagnosed with EDNOS [1]. Althoughthese patients did notmeet theDSM-IV criteria for a specic eatingdisorder, they experienced clinical impairment in developmentand function andwere at risk for severemedical complications [1].

EDNOS has been a heterogeneous category that includes asubset of patients who are, in general, younger than those withAN or BN and do not endorse signicant body image distortion ora fear of weight gain [4e9]. Currently, there are no evidence-based studies describing this group of patients. Anecdotally,clinicians report that patients in this subset had experiencedchoking episodes or vomiting followed by the development offear of eating solid foods, had restricted diets since early child-hood, or had reported abdominal pain that had prevented themfrom eating sufciently. The Great Ormond Street criteria forcategorizing eating disorders in younger patients, includingfunctional dysphagia, selective eating, and food-avoidanceemotional disorder, were found to be more reliable, becausethese diagnostic categories captured the presentations of manyof these children and adolescents [10].

The DSM-5 Eating Disorder Work Group was charged withrevising and updating the DSM-IV diagnostic criteria to provideclinical guidelines for clinicians caring for individuals with eatingdisorders, including such patients [11,12]. This meant improvingthe denition of eating disorders in children and adolescents toreect the clinical expression of these disorders across thedevelopmental spectrum and lifespan. A specic goal of this workgroupwas to examine the variety of clinical presentations of thosepatients who had been diagnosed with EDNOS, explore the clin-ical utility of the DSM-IV category of Feeding Disorder of Infancyor Early Childhood, and consider the variety of clinical pre-sentations that did not t into the existing categories in the DSM-IV (e.g., food-avoidance emotional disorder, selective eating). Inkeeping with this goal, the work group revised the DSM-IVdiagnoses and combined the former sections Eating Disordersand Feeding and Eating Disorders of Infancy or Early Childhoodinto a single section, Feeding and Eating Disorders, with Avoi-dant/Restrictive Food Intake Disorder (ARFID) as a newlydescribed diagnosis in this section (Table 1) [13]. It was anticipatedthat the inclusion of ARFID as a diagnosis within Feeding andEating Disorders in the DSM-5 would improve clinical utility andcapture a population of young people who had an eating disorder,experienced medical and psychological morbidities, and whomight otherwise be excluded from the DSM diagnostic criteria.

The purpose of this study was to describe the characteristicsof children and adolescents presenting to seven adolescent-medicine eating disorder programs who met the DSM-5criteria for ARFID and to compare them to patients meeting theDSM-5 criteria for AN and BN. The same seven eating disorderprograms have recently published a paper demonstrating thegeneral distribution of eating disorder diagnoses in children andadolescents using the DSM-5 criteria [14].

Methods

A retrospective chart review was completed on all newpatients between 8 and 18 years of age who presented to sevenadolescent-medicine eating disorder programs across the UnitedStates and Canada between January and December 2010. Patientswith a diagnosis of ARFID were identied at each site using adiagnostic checklist based on the proposed DSM-5 diagnosticcriteria. A table of random numbers was used to select an equalnumber of patients with AN and BN. Cases of AN and BN wereidentied based on DSM-5 criteria. At some sites, there werefewer patients seen with BN than with ARFID. Under these cir-cumstances, all patients with BN were enrolled, resulting infewer patients with BN than those with ARFID and AN.

Data collected on all patients included age, gender, ethnicity,weight, and height. Additional clinical information includedduration of illness, highest and lowest weights and body massindex (BMI) percentile, intake setting (out-patient vs. other),referral source, and presence of a medical condition, mood dis-order, anxiety disorder, autism spectrum disorder, cognitiveimpairment, food allergies, a choking episode, difculty swal-lowing, or sensory issues.

The documented reason for poor nutritional intake in theARFID group was also delineated, and included selective (picky)

Table 1DSM-5 diagnosis of Avoidant/Restrictive Food Intake Disorder [13]

A. An eating or feeding disturbance (e.g., apparent lack of interest in eatingor food; avoidance based on the sensory characteristics of food; concernabout aversive consequences of eating) as manifested by persistent fail-ure to meet appropriate nutritional and/or energy needs associated withone (or more) of the following:1. Signicant weight loss (or failure to achieve expected weight gain or

faltering growth in children).2. Signicant nutritional deciency.3. Dependence on enteral feeding or oral nutritional supplements.4. Marked interference with psychosocial functioning.

B. The disturbance is not better explained by lack of available food or byassociated culturally sanctioned practice.

C. The eating disturbance does not occur exclusively during the course ofanorexia nervosa or bulimia nervosa, and there is no evidence of adisturbance in the way in which ones body weight or shape isexperienced.

D. Theeatingdisturbance isnot attributable toa concurrentmedical conditionor not better explained by another mental disorder. When the eatingdisturbance occurs in the context of another condition or disorder, theseverity of the eating disturbance exceeds that routinely associated withthe condition or disorder and warrants additional clinical attention.

met criteria for ARFID using DSM-5 criteria. The diagnosis ofARFIDwas fairly consistent across sites, with the exception of onesite (University of Michigan). At six of the seven sites, the fre-quency of patients with ARFID ranged between 7.2% and 17.4%;however, the incidence at the remaining site was 41.0%. Thenumber of patients who met DSM-5 criteria for AN and BN usedin the analysis to compare with the patients with ARFID were 98and 66, respectively.

Characteristics of patients with ARFID, AN, and BN are sum-marized in Table 2. Almost 30% of patients with ARFID were male,a far higher percentage than those with either AN or BN. Thepercent median body weight (% MBW) for patients with ARFID(86.5) was between those with AN (81.0) and those with BN(107.5). Patients with ARFID tended to be younger and had asignicantly longer duration of illness. The majority of all patientswere evaluated as out-patients. Patients with ARFID were lesslikely to be self-referred. Although patients with ARFIDweremorelikely to have a medical condition or an anxiety disorder thanthosewith ANor BN, theywere less likely to have amood disorder.

The authors grouped the ARFID patients according to specicsymptoms documented in the medical record. The groupings ofthe ARFID patients included 28 (28.7%) with selective (picky)eating since early childhood; 21 (21.4%) with generalized anxiety;19 (19.4%) with gastrointestinal symptoms; 13 (13.1%) with fearsof eating secondary to fears of choking or vomiting; 4 (4.1%) withfood allergies; and 13 (13.2%) with restrictive eating for otherreasons. There were no statistically signicant differences by age

or gender among the patients in these groups. Further, unlike thepatients with ARFID, patients with AN and BN had almost none ofthe above associated symptoms.

Discussion

This is the rst study to describe a large cohort of children andadolescents meeting the DSM-5 diagnostic criteria for ARFID.Approximately 14% of all new eating disorder patients whopresented to seven adolescent-medicine eating disorders pro-grams between January and December 2010 met these criteria.

The diagnostic criteria for ARFID are broad. Bryant-Waugh hasoutlined a diagnostic checklist for criterion A of ARFID to facilitategathering the appropriate information [15]. Despite the novelty ofthese criteria, clinicians who staffed programs focusing on eatingdisorders were able to utilize information in the medical recordsto make a diagnosis of ARFID in a substantial number of childrenand adolescents.

This multicenter study revealed that all but one research sitehad a similar incidence of patients with ARFID, suggesting that thediagnosis of ARFID across sites was reliable and stable. The reasonsfor the difference in the frequency of ARFID at the one site areunclear. This site was known as a center that specialized in thecare of younger patients with eating disorders. Therefore, a higherproportion of younger patients may have been directed to the site.However, all datawere analyzed both including and excluding thissite and no differences in statistical analyses were found.

Table 2Clinical characteristics of patients with ARFID, anorexia nervosa, or bulimia nervosa

ARFID (n 98) Anorexia nervosa (n 98) Bulimia nervosa (n 66)Age (years) 12.9 2.5 15.6 1.9a 16.5 1.3a F[2,259] 71.2, p < .001% Median body weight 86.5 15.1 81.0 9.2a 107.5 16a F[2,259] 80.8, p < .001Lowest weight (kg) 35.0 11.9 41.4 7.3a 53.3 9.5a F[2,247] 64.7, p < .001

iso

fere.05.05

M.M. Fisher et al. / Journal of Adolescent Health 55 (2014) 49e52 51Highest weight (kg) 40.8 15.0 54.0 12.9aDuration (months) 33.3 41.3 14.5 12.2aGenderb,c

Female (%) 71.3 85.7Male (%) 28.6 14.3

Intake settingOPD (%) 87.7 85.7Other (%) 12.3 14.3

Referral sourcec

Self (%) 6.2 10.2PCP (%) 51.6 50.0Mental health (%) 11.3 16.3Emergency department (%) 10.3 11.2Social service (%) 1.0 4.1Other (%) 0 3.0

Medical condition or symptomb,c

Yes, related (%) 34.6 8.2Yes, unrelated (%) 16.3 2.0None (%) 49.1 89.8

Mood disorderc

MDD/dysthymia (%) 7.2 19.4Other (%) 11.3 11.2None (%) 81.5 69.4

Anxiety disorderb,c

GAD (%) 28.6 14.3OCD (%) 6.1 8.2Other (%) 23.5 13.3None (%) 41.8 64.2

ARFID Avoidant/Restrictive Food Intake Disorder; GAD generalized anxiety dOPD out-patient department; PCP primary care physician.

a Signicant difference from ARFID, p .05 by Tukeys Honestly Signicant Difb Signicant difference between anorexia nervosa and ARFID by Chi-square, p