Nicely et al.

Journal of Eating Disorders 2014, 2:21

http://www.jeatdisord.com/content/2/1/21

RESEARCH ARTICLE Open Access

Prevalence and characteristics of avoidant/

restrictive food intake disorder in a cohort of
young patients in day treatment for eating
disorders
Terri A Nicely1, Susan Lane-Loney2, Emily Masciulli2, Christopher S Hollenbeak3 and Rollyn M Ornstein2*

Abstract
Background: Avoidant/Restrictive Food Intake Disorder (ARFID) is a “new” diagnosis in the recently published DSM-5,
but there is very little literature on patients with ARFID. Our objectives were to determine the prevalence of ARFID in
children and adolescents undergoing day treatment for an eating disorder, and to compare ARFID patients to other
eating disorder patients in the same cohort.
Methods: A retrospective chart review of 7-17 year olds admitted to a day program for younger patients with eating
disorders between 2008 and 2012 was performed. Patients with ARFID were compared to those with anorexia nervosa,
bulimia nervosa, and other specified feeding or eating disorder/unspecified feeding or eating disorder with respect
to demographics, anthropometrics, clinical symptoms, and psychometric testing, using Chi-square, ANOVA, and
post-hoc analysis.
Results: 39/173 (22.5%) patients met ARFID criteria. The ARFID group was younger than the non-ARFID group
and had a greater proportion of males. Similar degrees of weight loss and malnutrition were found between
groups. Patients with ARFID reported greater fears of vomiting and/or choking and food texture issues than
those with other eating disorders, as well as greater dependency on nutritional supplements at intake. Children’s
Eating Attitudes Test scores were lower for children with than without ARFID. A higher comorbidity of anxiety
disorders, pervasive developmental disorder, and learning disorders, and a lower comorbidity of depression,
were found in those with ARFID.
Conclusions: This study demonstrates that there are significant demographic and clinical characteristics that
differentiate children with ARFID from those with other eating disorders in a day treatment program, and helps
substantiate the recognition of ARFID as a distinct eating disorder diagnosis in the DSM-5.
Keywords: Avoidant restrictive food intake disorder, Children and adolescents, Day treatment, DSM-5

Background Disorder Not Otherwise Specified (EDNOS), likely lead-

Historically, children and adolescents have not been eas- ing to missed diagnoses and difficulty obtaining appro-
ily diagnosed with eating disorders (EDs) based on past priate and timely treatment [1-3]. With the preparations
versions of the Diagnostic and Statistical Manual of for publication of the 5th edition of the DSM (DSM-5),
Mental Disorders (DSM), including the 4th edition. In the Eating Disorders Work Group was assigned the tasks
fact, over 50% of these patients met criteria for Eating of improving clinical utility of the diagnostic categories
and reducing the frequency of EDNOS. One of the im-
peratives was to recognize new disorders and eliminate
* Correspondence: rornstein@hmc.psu.edu
2
Division of Adolescent Medicine and Eating Disorders, Penn State Hershey
others by exploring the clinical profiles of patients who
Children’s Hospital, 905 West Governor Road, Suite 250, 17033 Hershey, PA, fell under the heterogeneous EDNOS category. In
USA
Full list of author information is available at the end of the article

© 2014 Nicely et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain
Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,
unless otherwise stated.
Nicely et al. Journal of Eating Disorders 2014, 2:21 Page 2 of 8
http://www.jeatdisord.com/content/2/1/21

addition, the DSM-5 as a whole has attempted to take a failure to gain appropriate weight, may occur. Studies
developmental, or life-span, approach to all disorders. have shown a higher prevalence of boys with selective
Feeding Disorder of Infancy or Early Childhood, a eating, as well as a high degree of co-morbid anxiety
diagnosis in the DSM-IV, delineated a persistent eating [11,12].
dysfunction leading to weight loss or failure to gain Functional dysphagia is a fear of swallowing or an
weight, with the requirement that patients be less than inability to eat or swallow food, especially solid or lumpy
six years of age. This was a non-specific diagnostic cat- foods. There is generally a fear of gagging, choking, or
egory that was rarely used in practice and for which vomiting, often subsequent to actual traumatic episodes
there was insufficient literature [4]. A great number of or witnessed episodes. Sometimes an illogical connection
patients are over six years old at the time of initial ED in the child’s mind leads to development of the phobia.
evaluation, even if some have had symptoms from an Some children present with food refusal specifically
early age, and have been necessarily given the diagnosis out of fears of vomiting, contamination, poisoning, or
EDNOS in the past. Feeding Disorder of Infancy or Early defecation as well. Many cases of acute food refusal due
Childhood also excluded those children with abnormal to specific fears present clinically malnourished and ill,
eating patterns or nutritionally deficient or limited diets, as they often lose weight rapidly. They can easily be mis-
but who were growing normally secondary to sufficient taken with AN on initial presentation due to the severity
caloric intake, possibly due to the use of nutritional supple- of the restriction; however, they are not concerned with
ments. The inability of DSM-IV to capture such patients weight or shape [4,5].
was significant, as they often presented with considerable The DSM-5 has subsumed and expanded Feeding Dis-
impairment, both physically and functionally [5]. order of Infancy or Early Childhood to capture a greater
Clinicians and researchers have long recognized specific number of patients who present with avoidant or re-
types of EDs that fall under the umbrella of EDNOS. The strictive eating, but are clearly different from those with
Great Ormond Street (GOS) classification system cap- AN in that there are no disturbed cognitions about weight
tured a way to describe these types of patients, and was and/or shape, or a wish to lose weight. It has been renamed
often utilized by clinicians for descriptive purposes. These Avoidant/Restrictive Food Intake Disorder (ARFID) and
criteria were actually found to have a higher inter-rater includes those types of patients recognized in the GOS sys-
reliability for younger patients than the DSM-IV [1]. tem. Patients with ARFID may present with clinically sig-
The GOS categories include: Food Avoidant Emotional nificant restrictive eating leading to weight loss or lack of
Disorder (FAED), Selective Eating, and Functional Dys- weight gain, nutritional deficiencies, reliance on tube feed-
phagia, as well as Anorexia Nervosa (AN) and Bulimia ing or oral nutritional supplements and/or disturbances in
Nervosa (BN). psychosocial functioning (see Table 1) [13]. Additionally,
FAED was first described as a combination of inad- they may exhibit similar physical signs and symptoms as
equate food intake and emotional disturbance; these patients with AN due to semi-starvation.
young people knew that they were underweight and Very little has been published on patients with ARFID.
wanted to be heavier, but found this difficult to achieve [6]. Recently, a large multicenter study of children and ado-
The GOS system further clarified this group, and differen- lescents presenting as new patients to adolescent medi-
tiated their presentation by the absence of weight and cine ED programs, revealed a 14% prevalence of ARFID,
shape concerns in the presence of significant food restric- with unique clinical characteristics, including younger
tion. Somatic complaints were frequent as well as more age and a greater number of males [14,15]. An 11-year
general psychopathology, e.g. generalized anxiety [4,5]. retrospective chart review of adolescent ED patients in
Selective eating, also known as “picky eating”, is a Canada reported a 5% prevalence of ARFID [16]. These
common problem of childhood, with anywhere between patients were compared to a matched sample of AN pa-
13 to 22% of children between 3 and 11 years of age tients, and demonstrated a younger age at presentation,
being reported to be picky eaters at any given time [7]. and a higher likelihood of being male. There were spe-
While young children are typically thought to “grow out cific behaviors and symptoms in the ARFID group, in-
of” their pickiness, studies have shown that between 18 cluding food avoidance, decreased appetite, abdominal
and 40% of the rigidity concerning food persists into pain, and emetophobia. Both of these studies included
adolescence [8-10]. Patients with selective eating are all new patients presenting for initial assessments to
usually not underweight, as they take in adequate calo- tertiary care ED programs.
ries from preferred foods, but their diets may be lack- Due to the dearth of literature on ARFID, we sought
ing in micronutrients. Some selective eaters have to determine the prevalence and clinical characteristics
sensory concerns related to the taste, smell, color, or of ARFID in young patients admitted to a day treatment
texture of foods, which may limit their intake to such a program for EDs, and to compare patients with ARFID
narrow range of acceptable foods that weight loss, or to those with AN, BN, and Other Specified Feeding or
Nicely et al. Journal of Eating Disorders 2014, 2:21 Page 3 of 8
http://www.jeatdisord.com/content/2/1/21

Table 1 Diagnostic Criteria for Avoidant/Restrictive Food Intake Disorder

WHAT IS ARFID? WHAT ARFID IS NOT
● A problem with eating or feeding (e.g. seeming disinterest in food or ● The eating problems are not due to body image disturbance, and anorexia
eating; repulsion to certain foods based on their sensory qualities; nervosa or bulimia nervosa cannot be diagnosed instead.
fears about aversive effects of eating) leading to recurrent inability to
take in adequate nutrition and/or energy coupled with one (or more)
of the following:
○ Major nutritional deficiency. ● Feeding or eating problems are not the result of scarcity of food or a
culturally endorsed tradition.
○ Substantial weight loss (or lack of weight gain). ● The disordered eating is not due to a concomitant medical problem or
another psychiatric disorder, so that if the medical or psychiatric disorder is
treated, the eating problems resolves.
○ Reliance on nasogastric or gastric tube feeding or oral nutrition
supplements.
○ Impaired psychosocial function.
Adapted from the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, American Psychiatric Association, 2013.

Eating Disorders/Unspecified Feeding or Eating Disorder Of the 177 eligible subjects, a total of four participants
(OSFED/UFED) in the same cohort. were excluded from the study. Two were excluded for
having medical conditions that were retrospectively
determined to fully account for their disordered eating
Methods behaviors. Two subjects were excluded for having
Participants Binge Eating Disorder and composed too small a distinct
A retrospective chart review was conducted on 177 pa- group for data analysis.
tients admitted to a day program for children and ado-
lescents with EDs between August 4th, 2008 and May
Measures
1st, 2012. This program treats female and male patients,
Demographics, historical and clinical features
ages 7 to 17 years, with EDs and co-morbid psychopath-
Data collected at intake included age, gender, and ethni-
ology. The majority of patients in the program have re-
city. Historical information included past history of ED
strictive EDs, based mostly on the younger average age.
and/or other mental health treatment, other medical disor-
However, patients with purging disorders are treated as
ders and consultations by other medical specialists, pres-
well. While we treat some patients with sensory features
ence of weight loss, percentage of body weight lost, length
related to food, who may or may not also have an autism
of illness, use of nutritional supplements, presence of pur-
spectrum disorder diagnosis, it is important to clarify
ging behaviors, excessive exercise, history of food allergies,
that patients with longstanding feeding issues and aut-
fears of choking and/or vomiting, and sensory issues re-
ism are not typically admitted to our program, and are
lated to food. This information was gathered from the ini-
usually managed in the Feeding Disorders Program at
tial evaluations by the adolescent medicine physician, the
our institution.
psychiatrist, and the psychologist or clinical social worker.
Initial ED and co-morbid psychiatric diagnoses were
made upon admission to the program based on a com-
prehensive diagnostic psychiatric evaluation, by both a Anthropometrics
trained child and adolescent psychiatrist and either an Weight and height were measured by trained staff at
experienced clinical psychologist or a licensed social initial presentation. Gowned weights were obtained on a
worker/clinical psychiatric specialist, using DSM-IV-TR hospital-grade SECA digital scale and recorded to the
criteria. Some of the co-morbid diagnoses were based on nearest tenth of a kg. Heights were measured in bare
history conveyed by the parent to the health care provider. feet using a fixed stadiometer with a right angle head-
DSM-5 ED diagnoses were determined retrospectively and piece and recorded to the nearest tenth of a cm. BMI
agreed upon together through careful discussion by two of was calculated using the standard formula (kg/m2) and
the psychiatric specialists and an adolescent medicine the % Median Body Weight (%MBW) was determined
physician, all of whom were personally involved with the based on the 50th percentile BMI-for-age.
cases, using a checklist based on the proposed DSM-5
diagnostic criteria, which were almost identical to the Psychometric measures
published criteria. Therefore, these diagnoses were not The Children’s Eating Attitudes Test (ChEAT) [17] The
made in a blinded fashion. ChEAT is a 26-item scale assessing attitudes and behaviors
Nicely et al. Journal of Eating Disorders 2014, 2:21 Page 4 of 8
http://www.jeatdisord.com/content/2/1/21

associated with food and eating, validated in patients Results

as young as 8 years old, adapted from the original Demographics and anthropometrics
EAT-26 [18]. A score of ≥ 20 is considered clinically Using the proposed DSM-5 criteria, 39 (22.5%) patients
significant relative to the normative population. The met criteria for ARFID, 93 (53.8%) for AN, 20 (11.6%)
three subscales reflecting varying types of eating path- for BN, and 21 (12.1%) for OSFED/UFED. Notably, all
ology include: Dieting, Bulimia/Food Preoccupation, patients diagnosed with ARFID carried a DSM-IV diag-
and Oral Control [18]. nosis of EDNOS. None were diagnosed with DSM-IV
Children’s Depression Inventory (CDI) [19]. The CDI is Feeding Disorder of Infancy or Early Childhood, as all
a 27-item self-report inventory for assessing depression were over six years old at intake. Of the 173 participants
in children between the ages of 7 and 17 years. The included, 92% were female with a mean age of 13.5 years
measure yields a Total score (M = 50; SD = 10) and five (SD = 2.03) (range 7.2 -16.9 years). The cohort was pre-
factors: Negative Mood, Interpersonal Problems, Ineffect- dominantly Caucasian (95%), reflecting the ethnic/racial
iveness, Anhedonia, and Negative Self-Esteem (M = 10; makeup of the geographic area. There was no significant
SD = 3). difference in duration of illness between those patients
Revised Children’s Manifest Anxiety Scale (RCMAS) with ARFID and the other ED groups.
[20]. The RCMAS is a 37 item self-report instrument de- Patients with ARFID were found to be younger than
signed to measure anxiety for children and adolescents those with other EDs (11.1 years, SD = 1.7 vs. 14.2 years,
ages 6 to 17 years. The measure yields a Total Anxiety SD = 1.5; p < 0.0001) and to have a greater percentage of
score based upon 28 items, with 9 items comprising the males (20.5% vs 4.5%; p = 0.008). Of the patients who
Lie Scale which is designed to detect responses that are had lost weight as part of their ED, those with AN lost a
socially desirable. The Total Anxiety Score is expressed as greater percentage of their premorbid weight than the
a T-score (M = 50, SD = 10) and there are three factor- other ED groups, including those with ARFID (Table 2).
based subscales, expressed as scaled scores (M = 10, There was a significant difference found in %MBW
SD = 3): Physiological Anxiety, Worry/Oversensitivity, between those with ARFID and BN, but not between
and Social Concerns/Concentration. ARFID and AN, or OSFED/UFED (Table 2). While the
The Child Behavior Checklist (CBCL) [21]. The CBCL degree of malnutrition was similar to that of patients
provides three global measurements which are expressed with AN, those with ARFID were found to have a
as a T- score (M = 50; SD = 10) including Total Score; greater dependence on nutritional supplements, fears of
Internalizing; and Externalizing Scales. In addition, the vomiting and/or choking, and texture/sensory issues
measure includes 14 Syndrome Scores which reflect pertaining to food (all p < 0.0001).
clusters of psychiatric symptoms. These scales are also
expressed as T-scores (M = 50; SD = 10) and include Psychometric assessment and psychiatric co-morbidities
the following scales: Anxious/Depressed, Withdrawn/ Patients with ARFID were less likely to report typical ED
Depressed, Somatic Complaints, Social Problems, Thought symptoms, e.g. purging behaviors and excessive exercise,
Problems, Attention Problems, Rule-Breaking Problems, during intake interview (all p <0.0001). In addition, they
Aggressive Behavior, Affective Problems, Anxiety Problems, had significantly lower total scores on the ChEAT (14.86,
Somatic Complaints, ADHD Problems, Oppositional Defi- SD = 2.10) than those of the remaining patients overall
ant Problems, and Conduct Problems. It is completed by (27.51, SD = 17.28) (p < 0.0001) (Figure 1). Post-hoc ana-
parents and/or other caregivers. lysis revealed significant differences among patients with
ARFID and all other groups for the total ChEAT score.
While patients with ARFID also had significantly lower
Statistical analysis scores on both the Dieting and Bulimia Nervosa/Food
Analysis included descriptive statistics, chi-square, ana- Preoccupation subscales (p < 0.0001), there was no signifi-
lysis of variance (ANOVA), and Pearson’s correlation. cant difference between groups on the Oral Control sub-
Bonferroni correction was used to adjust for Type I error, scale. An interesting finding on chart review was that
with thresholds set at p < 0.01 for patient characteristics, while patients with ARFID did not have true body image
p < 0.007 for ED symptoms and features, and p < 0.008 for distortion, as seen in AN, 21% exhibited body preoccupa-
psychiatric co-morbidities. Post-hoc testing to examine tion with somatic concerns. For example, some children
between-groups effects was performed with the Hochberg were fixated on fears of physical illness due to issues re-
GT2 test. Data were entered and analyzed using SPSS lated to shape/weight, e.g. high cholesterol and/or obesity
(version17.0, SPSS Inc., Chicago, Illinois). leading to heart disease, either because of personal experi-
This study was approved by the Institutional Review ences with relatives or information in their school curricu-
Board of the Penn State Hershey Medical Center/College lum. Others who were chronically underweight due to
of Medicine. their feeding and eating disturbance had suffered teasing
Nicely et al. Journal of Eating Disorders 2014, 2:21 Page 5 of 8
http://www.jeatdisord.com/content/2/1/21

Table 2 Clinical characteristics of patients by eating disorder diagnosis

ARFID AN BN OSFED/UFED p-value
(N = 39) (N = 93) (N = 20) (N = 21)
Patient Characteristics (mean or %)
Age (years) (SD) 11.1 (1.7)* 14.0 (1.5) 14.9 (1.1) 14.2 (1.7) <0.0001
% MBW (SD) 87.1 (13.0) 82.6 (9.2) 108.1 (19.5)* 93.2 (6.8) <0.0001
% Body Weight Lost (SD) 10.5 (8.4) 18.5 (10.2)* 6.4 (6.5) 14.8 (12.2) <0.0001
Length of illness (months) (SD) 9.8 (13.2) 8.6 (7.9) 15.9 (11.9) 9.8 (4.9) N.S.
% Female 79.5 95.7 100 90.5 0.008
% Male 20.5* 4.3 0 9.5
Symptoms & Features (%)
Enteral Supplement Use 46* 20 0 0 <0.0001
Purge-vomit 0 6 95* 38 <0.0001
Excessive exercise 15* 68 65 52 <0.0001
Food allergy 20 5 10 5 N.S.
Fear of choking or vomiting 44* 1 0 0 <0.0001
Sensory issues 26* 1 0 0 <0.0001
Recent medical specialist consult 46 19 20 33 N.S.
Psychiatric comorbidities (%)
Mood disorder 33* 48 80 76 <0.0001
Anxiety disorder 72* 37 25 14 <0.0001
Autism Spectrum Disorder 13* 0 0 0 0.001
Attention Deficit Disorder 4* 0 1 1 N.S.
Learning Disorder 10* 2 2 0 <0.0001
Cognitive impairment 26* 2 10 0 <0.0001
*Significant finding on post-hoc analysis using Hochberg GT2 test.
AN = Anorexia Nervosa ARFID = Avoidant/Restrictive Food Intake Disorder.
BN = Bulimia Nervosa OSFED/UFED = Other Specified Feeding or Eating Disorder/Unspecified Feeding or Eating Disorder % MBW = % Median Body Weight.

Figure 1 Total and subscale ChEAT scores by DSM-5 Diagnosis. ChEAT = Children’s Eating Attitudes Test. All differences between groups
significant at p < 0.0001 except Oral Control = N.S.
Nicely et al. Journal of Eating Disorders 2014, 2:21 Page 6 of 8
http://www.jeatdisord.com/content/2/1/21

by their peers because of their low weight, which may found to be 5% [16]. There is no mention of age range
have led to body image concerns, although of a different in that study, only that the patients were adolescent
nature than typically seen in AN and BN. ED patients assessed in a pediatric tertiary care hospital
There was a significantly higher comorbidity of anxiety program. Our study included children and adolescents
disorders in patients with ARFID (72%) than the other between 7 and 17 years, which may have been a slightly
ED groups (31%), as determined by clinician diagnosis lower range than the Canadian study; this might also
(p < 0.0001). Furthermore, this was supported by paren- justify the higher prevalence of ARFID found in our
tal report on the CBCL (p =0.005). However, there were cohort. Another possible explanation for the discrepancy
no significant differences between groups on the total in prevalence rates across studies is that younger pa-
RCMAS score. Autism spectrum disorder (p = 0.001), tients with atypical EDs, like ARFID, may be increasingly
learning disorders (p < 0.0001), and cognitive impair- referred to adolescent medicine ED programs in more
ment (p < 0.0001) were also seen more frequently in the recent years, as there has been greater recognition of
patients with ARFID, based on past history reported at these presentations as true EDs. The Canadian study
initial assessment (Table 2). On the CBCL, children with reviewed records starting in 2000 and it would be inter-
ARFID had significantly more social problems (p = 0.001) esting to know whether the prevalence increased annu-
and attention problems (p < 0.0001) than those with AN. ally over the 11 years. In our experience, referrals from
There was a lower comorbidity of depression diagnosed in primary care providers tend to generate more referrals
children with ARFID (23%) than the other EDs (57%) once they are successfully managed. Lastly, the higher
(p < 0.0001), and total CDI scores were lower in this group prevalence in our cohort may reflect the fact that many
as well (54.4 vs. 60.0, p = 0.05). Additionally, children with children and adolescents with ARFID present acutely
ARFID were found to have significantly lower scores on and significantly malnourished, requiring a higher level
the CDI subscales Negative Mood (p =0.02) and Negative of care, such as day treatment.
Self Esteem (p < 0.0001). There were no significant differ- Similar to the multicenter and Canadian studies [15,16],
ences between the groups on the Interpersonal Problems, our results demonstrate that there are significant demo-
Ineffectiveness, or Anhedonia subscales, however. graphic and diagnostic characteristics that differentiate
A smaller percentage of children with ARFID (35%) children with ARFID from those with other EDs. First,
sought outpatient psychotherapy before coming to the while female patients remain the majority, there was a
program, compared to patients in the other ED groups higher preponderance of male patients in the ARFID
(AN = 60.22%, EDNOS = 75%, BN = 80%; p = 0.002). How- group than in the other ED groups. Children and adoles-
ever, there were no differences in the past history of cents with ARFID were more likely to present at a youn-
higher levels of psychiatric care, e.g. inpatient, residential, ger age with significant weight loss or failure to gain
or day treatment. In contrast, more children with ARFID appropriate weight, were more dependent on oral or
(46.2%) had seen other medical specialists for consultation enteral nutritional supplementation, and had significantly
(e.g. gastroenterology, endocrinology) before coming to more fears of choking and/or vomiting, and texture and/
program than those with other EDs (26.1%), although this or sensitivity issues regarding food. These findings are
did not reach statistical significance with the Bonferroni consistent with those in studies of early-onset EDs
correction (p = 0.02). [2,22,23], as well as in the recent multicenter study [15],
and many are relevant and important features in making
Discussion the diagnosis of ARFID [24].
This study adds to the literature on ARFID by compar- Based on DSM-5 criteria, a patient cannot have body
ing a cohort of children and adolescents undergoing day image distortion and be diagnosed with ARFID. How-
treatment for EDs, including patients with this “new” ever, our data revealed that 21% of patients diagnosed
diagnosis. Notably, almost a quarter of our patients were with ARFID had body preoccupation with somatic
diagnosed with ARFID, which illustrates the significant concerns. It is important to reiterate that none of the
prevalence of this disorder amongst children and adoles- patients with ARFID had been diagnosed with AN using
cents requiring an intensive level of ED treatment in a DSM-IV criteria, which underscores the absence of true
tertiary care setting. This was a higher prevalence than body image distortion. During evaluation of a young
that found in the multicenter studies [14,15], which patient with possible ARFID versus AN, it is critical to
might be accounted for by the fact that our patients probe about body concerns that need to be distinguished
were encountered over four years in a day treatment set- from body image distortion. For example, if a patient
ting, as opposed to all ED patients presenting for initial has worries about becoming fat, this may have some-
evaluation over a one-year period. The prevalence rate thing to do with events in the family’s medical history,
was in even starker contrast to the 11-year retrospective e.g. an overweight parent or grandparent with a recent
review from Canada, where the prevalence was only myocardial infarction or diabetes diagnosis. Children
Nicely et al. Journal of Eating Disorders 2014, 2:21 Page 7 of 8
http://www.jeatdisord.com/content/2/1/21

and adolescents are often privy to this information, but health arena, as a trend in our data revealed, although it
may make illogical associations based on their cognitive was not significant.
developmental stage. This knowledge may then trigger There were several strengths to this study, including
restrictive eating behaviors. Thorough history-taking can the large sample size and the use of both clinical and
often elicit this information. standardized psychometric measures for patient assess-
As has been documented in other studies of patients ment. Additionally, the use of multiple informants (pa-
with acute food avoidance without weight/shape con- tients, parents, and clinicians) adds to the validity of the
cerns [2,15,22,25,26], there were no significant differ- findings. Furthermore, experienced clinicians completed
ences in our study between % MBW in patients with all assessments and the adolescent medicine physician
ARFID and AN; however, patients with AN lost a signifi- involved in deciding on the retrospective DSM-5 diagno-
cantly greater percentage of their premorbid body weight. ses was integrally involved in the efforts leading up to
This may be explained by the fact that our patients with the inclusion of ARFID in the DSM-5. As ARFID is still
ARFID, notably those with the acute food refusal seen in a relatively “new” diagnosis, there are no formalized
functional dysphagia, may have presented sooner after the assessment tools available yet. However, instruments will
onset of illness than those with AN. The data may not likely be developed, capturing the clinical features and
fully bear this out due to the heterogeneity of the ARFID diagnostic criteria which will help standardize diagnosis.
category (e.g. more chronic selective eaters vs more There are some available resources to help guide the
acute food refusal), which might balance out the length clinician in evaluation [24,27].
of illness data. Furthermore, young patients may present However, there are several limitations that deserve
relatively early in the course of their illness, based on mention. The retrospective nature of this study, and the
their age alone. fact that diagnoses were made on DSM-5 criteria that
Based on both clinician and parental report, patients had not yet been formalized by the time of its comple-
with ARFID had significantly more anxiety and less de- tion, need to be taken into consideration. However,
pression than patients with other EDs, which is similar as previously mentioned, the published DSM-5 criteria
to findings in the large multicenter study on ARFID were essentially the same as the proposed criteria used
[15]. However, our study is the first of patients with for this study. Careful discussion amongst experienced
ARFID to use standardized measures obtained from par- clinicians very familiar with all of the cases was under-
ents to aid in evaluation. There were no self-reported taken to decide upon the appropriate DSM-5 diagnosis
significant differences found between children with ARFID for each patient; this did not allow for direct assessment
and those with other EDs on the RCMAS or any of its of inter-rater reliability. The absence of blinding of the
subscales, which could be due to the generally high comor- clinicians may have introduced bias to the outcome of
bidity of anxiety symptoms in EDs. Alternatively, younger the study, possibly leading to a higher prevalence of
patients (those more likely to be diagnosed with ARFID) ARFID than previously seen in other studies. Lastly, our
may have had a harder time filling out the questionnaire patients were undergoing day treatment, which im-
than older subjects, perhaps in understanding the ques- plies a certain severity of illness, and may limit the
tions or acknowledging symptoms of anxiety, due to cog- generalizability to patients in other settings, or non-
nitive developmental stage. It is important to clarify that clinical populations. Despite these limitations, this study
ARFID is not simply a type of anxiety disorder, as the provides support for ARFID as a separate diagnostic
severity of the eating disturbance exceeds that which category.
might be seen in an anxiety disorder and necessitates
further clinical attention (see Table 1) [13]. Conclusions
Other than the use of outpatient psychotherapy, there This is the first study to examine patients with the diag-
were no significant differences between the groups in nosis of ARFID in a cohort of patients undergoing day
terms of prior mental health treatment, including hospi- treatment and adds to the limited literature available on
talizations for EDs or other mental health issues, admis- this new diagnosis. The inclusion of psychometric mea-
sions to day treatment programs, intensive outpatient sures from both patients and parents has not been docu-
programs, or residential treatment facilities. It should be mented to date. Children and adolescents with ARFID
taken into consideration, however, that ours is a young, are clearly distinct from those with other EDs and can
relatively treatment-naïve population, and that the rate now be identified and labeled more specifically and ac-
of past mental health admissions would be very different curately. Ideally, this will enable more timely recognition
when looking at an older population of patients. Add- and access to care. The degree of both physical and psy-
itionally, children with ARFID may be more likely seen chosocial dysfunction with which these patients present
as medically ill initially, and the early referrals may tend indicates the need for prompt and appropriate treat-
to gravitate toward the medical as opposed to mental ment. The relatively high prevalence of patients with
Nicely et al. Journal of Eating Disorders 2014, 2:21 Page 8 of 8
http://www.jeatdisord.com/content/2/1/21

ARFID in this treatment setting may indicate the need 8. Mascola AJ, Bryson SW, Agras WS: Picky eating during childhood: a
for an intensive level of care for many of these children longitudinal study to age 11 years. Eat Behav 2010, 11(4):253–257.
9. Nicklaus S: Development of food variety in children. Appetite 2009,
and adolescents, depending on their initial presentation. 52(1):253–255.
Future research on ARFID, with respect to course, prog- 10. Jacobi C, Schmitz G, Agras WS: Is picky eating an eating disorder? Int J Eat
nosis and treatment is warranted. Disord 2008, 41(7):626–634.
11. Timimi S, Douglas J, Tsiftsopoulou K: Selective eaters: a retrospective case
note study. Child Care Health Dev 1997, 23(3):265–278.
Competing interests 12. Nicholls D, Christie D, Randall L, Lask B: Selective eating: symptom,
None of the authors have any competing interests relative to the publication disorder or normal variant. Clin Child Psychol Psychiatry 2001, 6(2):257–270.
of this manuscript. 13. APA: Diagnostic and Statistical Manual of Mental Disorders. 5th edition.
Washington, DC: American Psychiatric Association; 2013.
Authors’ contributions 14. Ornstein RM, Rosen DS, Mammel KA, Callahan ST, Forman S, Jay MS, Fisher
TN was involved with the study design, revised the database, carried out all M, Rome E, Walsh BT: Distribution of eating disorders in children and
data analysis, and was involved with drafting the manuscript. SLL was involved adolescents using the proposed DSM-5 criteria for feeding and eating
with the study design, was part of the group of providers determining the disorders. J Adolesc Health 2013, 53(2):303–305.
DSM-5 diagnoses, reviewed the data analysis, and was involved with drafting of 15. Fisher MM, Rosen DS, Ornstein RM, Mammel KA, Katzman DK, Rome ES,
the manuscript. EM was involved with the study design, was part of the group Callahan ST, Malizio J, Kearney S, Walsh BT: Characteristics of avoidant/
of providers determining the DSM-5 diagnoses, reviewed the data analysis, and restrictive food intake disorder in children and adolescents: a “new
reviewed the manuscript. CH supervised the statistical analysis, reviewed the disorder” in DSM-5. J Adolesc Health 2014, 55(1):49–52.
data analysis, created the tables, and reviewed the manuscript. RO conceived 16. Norris ML, Robinson A, Obeid N, Harrison M, Spettigue W, Henderson K:
of the study, was involved with the study design, was part of the group of Exploring avoidant/restrictive food intake disorder in eating disordered
providers determining the DSM-5 diagnoses, reviewed the data analysis, patients: a descriptive study. Int J Eat Disord 2013, 47(5):495–499.
and drafted and reviewed the manuscript. All authors read and approved 17. Maloney MJ, McGuire JB, Daniels SR: Reliability testing of a children’s
the final manuscript. version of the eating attitude test. J Am Acad Child Adolesc Psychiatry
1988, 27(5):541–543.
Author information 18. Garner DM, Olmsted MP, Bohr Y, Garfinkel PE: The eating attitudes test:
SL is a clinical psychologist who has specialized in the care of patients with psychometric features and clinical correlates. Psychol Med 1982,
eating disorders for almost 20 years. She is the Program Director for the day 12(4):871–878.
program described in the manuscript. 19. Kovacs M: Children’s Depression Inventory (CDI). New York, NY: Multi Health
RO is an Adolescent Medicine physician with clinical and research expertise Systems, Inc.; 1992.
in the area of eating disorders for 16 years. She was a Special Interest Group 20. Reynolds CR: Manual, Revised Children’s Manifest Anxiety Scale. Los Angeles,
leader at both the Society for Adolescent Health and Medicine Annual CA: Pro-Ed, Inc. 1985.
Meeting and the International Conference on Eating Disorders. She just 21. Achenbach TM: Manual for the Child Behavior Checklist/4-18 and 1991 Profile.
completed her tenure on the Scientific Program Committee for the 2014 Burlington, Vermont: University of Vermont Department of Psychiatry; 1991.
ICED. She serves as the Adolescent Medicine Director for the day program 22. Pinhas L, Morris A, Crosby RD, Katzman DK: Incidence and age-specific
described in the manuscript. presentation of restrictive eating disorders in children: a Canadian
paediatric surveillance program study. Arch Pediatr Adolesc Med 2011,
Acknowledgements 165(10):895–899.
No funding sources. 23. Peebles R, Wilson JL, Lock JD: How do children with eating disorders
The authors would like to acknowledge Debra K. Katzman, MD for her differ from adolescents with eating disorders at initial evaluation?
careful and critical review of the manuscript. J Adolesc Health 2006, 39(6):800–805.
24. Bryant-Waugh R: Avoidant restrictive food intake disorder: an illustrative
Author details case example. Int J Eat Disord 2013, 46(5):420–423.
1
Penn State College of Medicine, 500 University Drive, 17033 Hershey, PA, 25. Rhodes P, Prunty M, Madden S: Life-threatening food refusal in two
USA. 2Division of Adolescent Medicine and Eating Disorders, Penn State nine-year-old girls: re-thinking the Maudsley model. Clin Child Psychol
Hershey Children’s Hospital, 905 West Governor Road, Suite 250, 17033 Psychiatry 2009, 14(1):63–70.
Hershey, PA, USA. 3Departments of Surgery and Public Health Sciences, Penn 26. Nicholls DE, Lynn R, Viner RM: Childhood eating disorders: British national
State College of Medicine, 500 University Drive, 17033 Hershey, PA, USA. surveillance study. Br J Psychiatry 2011, 198(4):295–301.
27. Bryant-Waugh R, Kreipe RE: Avoidant/restrictive food intake disorder.
Received: 16 May 2014 Accepted: 13 July 2014 Psychiatr Ann 2012, 42(11):402–405.
Published: 2 August 2014
doi:10.1186/s40337-014-0021-3
Cite this article as: Nicely et al.: Prevalence and characteristics of
References avoidant/restrictive food intake disorder in a cohort of young patients in
1. Nicholls D, Chater R, Lask B: Children into DSM don’t go: a comparison of day treatment for eating disorders. Journal of Eating Disorders 2014 2:21.
classification systems for eating disorders in childhood and early
adolescence. Int J Eat Disord 2000, 28(3):317–324.
2. Madden S, Morris A, Zurynski YA, Kohn M, Elliot EJ: Burden of eating
disorders in 5-13-year-old children in Australia. Med J Aust 2009,
190(8):410–414.
3. Peebles R, Hardy KK, Wilson JL, Lock JD: Are diagnostic criteria for eating
disorders markers of medical severity? Pediatrics 2010, 125(5):e1193–e1201.
4. Bryant-Waugh R, Markham L, Kreipe RE, Walsh BT: Feeding and eating
disorders in childhood. Int J Eat Disord 2010, 43(2):98–111.
5. Nicholls D, Bryant-Waugh R: Eating disorders of infancy and childhood:
definition, symptomatology, epidemiology, and comorbidity. Child Adolesc
Psychiatr Clin N Am 2009, 18(1):17–30.
6. Higgs JF, Goodyer IM, Birch J: Anorexia nervosa and food avoidance
emotional disorder. Arch Dis Child 1989, 64(3):346–351.
7. Chatoor I: Feeding disorders in infants and toddlers: diagnosis and
treatment. Child Adolesc Psychiatr Clin N Am 2002, 11(2):163–183.