Abstract
Avoidant/restrictive food intake disorder (ARFID) was introduced in the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders. Three different subtypes of ARFID are described: individuals who seem disinterested in eating, those who avoid certain foods because of a sensitivity to specific characteristics of the food, and those who are concerned about an aversive experience associated with eating. There is currently no first-line treatment for ARFID. Three case studies are presented of patients with ARFID who participated in a family-based partial hospitalization program/intensive outpatient program for eating disorders. A description of the course of treatment is included, as well as ways in which the eating disorder program adapted treatment to more closely meet the unique needs of these patients. An approach with emphasis on parental involvement seems promising, although research is needed to investigate this more fully.
Keywords
Introduction
Avoidant/restrictive food intake disorder (ARFID) is characterized by an eating or feeding disturbance resulting in one or more of the following: significant weight loss or failure to achieve expected growth, nutritional deficiency, dependence on enteral feeding or nutritional supplements, or interference with psychosocial functioning (American Psychiatric Association (APA), 2013).
Information on this diagnosis, in terms of both research and clinical knowledge, is growing, although much of what is known is based on ARFID patients receiving treatment in eating disorder programs. There is no first-line treatment for patients with ARFID, although a behavioral parent-training intervention has been suggested (Murphy & Zlomke, 2016), as has family-based treatment (FBT) adapted for ARFID (Fitzpatrick, Forsberg, & Colborn, 2015) and cognitive behavioral therapy (Thomas & Eddy, 2018). The behavioral parent-training intervention (Murphy & Zlomke, 2016) integrates approaches from applied behavior analysis, typically used in intensive feeding programs, such as differential reinforcement and contingency management, with parent-training components, such as psychoeducation, in vivo coaching, and parent modeling. Suggested adaptations to FBT for patients with ARFID (Fitzpatrick et al., 2015) include psychoeducation about ARFID, such as the need to present novel foods multiple times before concluding whether a food is truly liked or disliked, and expanding the range of foods eaten by the patient, with a focus on the frequency with which foods are introduced rather than on the quantity of food eaten.
The range of presentations in ARFID is wide, possibly necessitating the availability of a number of effective interventions, which may vary depending on the “subtype” of ARFID. Although not officially recognizing subtypes, the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5) gives three examples of what might qualify as an eating or feeding disturbance: individuals who seem disinterested in eating, those who avoid certain foods because of a sensitivity to specific characteristics of the food, such as texture or color, and those who are concerned about an aversive experience associated with eating, such as choking or difficulty swallowing (APA, 2013). Empirical studies have supported the existence of these subtypes (Norris et al., 2018; Zickgraf, Lane-Loney, Essayli, & Ornstein, 2019), although they are not mutually exclusive, with one study finding that over 50% of patients endorsed symptoms characteristic of more than one subtype (Reilly, Brown, Gray, Kaye, & Menzel, 2019).
Individuals who seem disinterested in eating seem to lack the pleasure that others take in eating and show little interest in initiating eating on their own, perhaps related to low appetite signaling (Thomas et al., 2017). They often need prompting by parents to remember to eat and may be described as “lazy eaters” (Bryant-Waugh, 2013). Anecdotally, these individuals often report that they are easily distracted by other things when it comes time to eat and “forget” to eat, or do not want to take the time necessary to eat, preferring to engage in other more enjoyable activities. The second subtype includes those who avoid certain foods due to the sensory characteristics of the food. These may include taste, texture, color, smell, appearance, or temperature. The third group includes individuals who have had a negative experience associated with eating and thus develop a negative conditioned response to food intake. This could include experiences with choking or vomiting during eating, or may include those who have had a medical procedure, often involving the gastrointestinal system, which they deemed unpleasant or painful, resulting in a fear of eating in an effort to avoid recreating a similar physical sensation. This may also include individuals who had a medical condition that caused pain when eating, and who remain afraid of re-experiencing this pain even after they have recovered from the original cause. It is also possible to have a fear of aversive consequences without having had a negative experience oneself, for example, a fear of vomiting without having a specific conditioning experience, a fear of choking after reading about choking first aid in health class, or a fear of a heart attack from high cholesterol.
Different treatment approaches may be needed for each subtype of ARFID. For example, the third subtype described may benefit from cognitive behavioral treatment for anxiety, whereas the second subtype may warrant more intensive behavioral interventions from a multidisciplinary team (Clawson & Elliott, 2014) and the first may improve with FBT. Further research is needed to determine the best treatment approach for the various presentations of ARFID. The purpose of this article is to describe three different presentations of ARFID and how each responded to a family-based partial hospitalization program (PHP) for eating disorders. Identifying information has been removed and/or details have been changed to protect the identities of the patients and families.
Treatment
The University of Michigan Comprehensive Eating Disorders Program includes a PHP based on FBT principles (Hoste, 2015). In FBT, parents are temporarily put in charge of their child’s eating, similar to the role that nurses or dietitians would have on an inpatient eating disorder treatment team (Lock & Le Grange, 2012). The PHP was initially designed to treat patients with anorexia nervosa (AN), bulimia nervosa (BN), or eating disorder not otherwise specified (EDNOS). However, patients with ARFID were presenting to the program in need of help, so they were incorporated into the family-based program with some necessary adjustments, which will be outlined in the case illustrations below. Although ARFID is clinically quite different from AN, BN, or EDNOS, an increasing body of evidence has emerged to support the use of FBT in the renourishment of children and adolescents (Lock et al., 2010), and as such patients with ARFID seemed appropriate for inclusion in a family-based program. For most patients with ARFID, a pediatric feeding psychologist was involved in treatment.
The treatments used by the pediatric feeding psychologist are applied behavior analysis and behavioral parent training. During each session, parents are asked to describe in detail and/or demonstrate how the patient is eating and how they have conducted treatment at home since the last appointment. The psychologist then identifies strategies for the patient and family to work on at home, typically by working directly on eating with the patient. The parents and patient (when age appropriate) are trained to conduct the intervention, and frequently a behavioral contract is signed by the parents and child that specifies the tasks to be completed at home, reinforcement contingencies, and roles the parents and child will play in completing the homework.
In the PHP, patients participated in programming Monday through Friday for 6 hours per day. In addition to group therapy, patients ate two meals and a snack during the program day, with trained staff support. Parents were involved in treatment in numerous ways: parents chose all meals and snacks for their children during their stay in the program (either choosing off of a hospital menu or bringing food from home) and were required to attend breakfast daily with the patient. All families attended “introduction to FBT” sessions prior to joining the PHP to orient them to the treatment philosophy and their role in the therapeutic process, and to have a family meal with the therapist prior to joining the treatment milieu.
Some patients stepped down to the intensive outpatient program (IOP). They participated in the same programming as the PHP patients, but for 3 hours per day, 3 days per week. For all patients, the average length of stay in the PHP is approximately 5 weeks, and the average length of stay in the IOP is 3 weeks. Patients would then step down to weekly outpatient treatment.
Introductory FBT session
The structure of the first session of FBT is fairly consistent across families. In the first session, the entire family is seen together. The therapist reviews the impact that the eating disorder has had on the patient and on the family. Many parents begin to realize that they have unknowingly accommodated the eating disorder over time, by serving their child what he or she was willing to eat rather than what was necessary to meet his or her nutritional and weight-gain needs. The medical and psychological consequences of the eating disorder are reviewed and parents are given the task of refeeding. It is explained that parents will temporarily be making all eating-related decisions for their ill child until the disorder has loosened its grip and the child can start making healthy decisions again.
Case presentation #1
Introduction
C was an 8-year-old African American female referred to the program by her primary physician and pediatric gastroenterologist after inpatient medical hospitalization for recent weight loss after two choking incidents. She lost 13 lbs within 7 months, dropping from between the 50th and 75th percentile for weight to the 5th percentile. At the time of intake, she was 51 lbs (10th percentile), having regained 4 lbs in the hospital, and 50.59 inches (61st percentile) with a body mass index (BMI) of 14.1 (10th percentile).
History and initial presentation
C was a previously healthy child who was growing at approximately the 50th to 75th percentile for height and weight when she choked on a chicken nugget and, although she was able to cough it out, had a “panic attack.” Subsequently, she reported trouble with swallowing solids and liquids, feeling like there was something stuck in her throat. She then choked on a hamburger 4 months after the initial incident and had a panic attack, after which her eating diminished and fluctuated. She saw pediatric gastroenterology for difficulty swallowing after losing another 8 pounds, falling from a high of 60 to 47 lbs. At that visit, it was reported that C was “cheeking” food and liquids, not swallowing her saliva, and refusing supplements. An inpatient admission was arranged given the severe weight loss. By the time of discharge, she was eating ~2200 kcals/day with a goal of 2500–3000. Her labs were normal. She was started on omeprazole for reflux noted on imaging study and Miralax for constipation. Her discharge weight was 49.9 lbs.
After discharge from the inpatient unit, her mother reported great difficulty in getting C to eat all that was required, especially as she started to eat extremely slowly while in the hospital, taking 2–3 hours for meals at times. She also reported an inability to help C gain further weight, although C stated that she wanted to gain more weight. C endorsed symptoms consistent with a diagnosis of ARFID.
Treatment and goals
The aim of treatment for C was to reduce anxiety around eating, eliminate fear of choking, increase food consumption, and increase her weight.
Feeding consult
In addition to attending the PHP program, C and her family consulted with a pediatric feeding psychologist during program hours. C was observed engaging in several dysfunctional oral-motor behaviors that she believed would prevent her from choking, but which also dramatically slowed the pace of her eating. Behaviors observed included pocketing food in her cheek, chewing only a tiny portion of the bite at a time, chewing each portion of the bite for extended periods of time, taking and pocketing another bite while still pocketing a portion of the previous bite, and rolling masticated food around in her mouth for several seconds before attempting to swallow. In addition, C was observed swallowing many foods passively by transferring very small boluses of masticated food laterally (rather than onto her tongue, as is typical) and tilting her head slightly backward to allow the bolus to move toward the posterior portion of the oral cavity with her saliva. She was frequently tearful during the initial visit and expressed significant anticipatory anxiety before meal times and while eating.
PHP considerations and adaptations recommended by pediatric feeding psychologist
Following recommendations of the pediatric psychologist, staff and family prompted C to take small, consistently sized bites that she could chew and swallow within a typical period of time. Bite size was increased gradually. Staff also prompted C during program meals to move the food in her mouth and reminded her to “swallow from her tongue” when they observed her rolling food around in her mouth or attempting to swallow passively. C continued to attend weekly sessions with the pediatric feeding psychologist while in the PHP to learn relaxation strategies that would allow her to reduce physiological symptoms of anxiety prior to and during meals and continue to train her parents to conduct the intervention at home.
Specifically, parents were present while C learned the relaxation strategies to be able to help her practice them correctly at home. They were also asked to require her to practice a relaxation strategy for 10–20 minutes every day Modeling and practice with immediate verbal feedback were used to train her parents to identify subtle signs that she was engaging in the dysfunctional oral-motor behaviors described above and immediately provide the necessary prompts. Parents were taught that this would not only help her learn new habits more quickly by repeating only correct behaviors, but would also assist her in learning to self-monitor her oral-motor behavior. They were also taught that immediate, consistent prompting would function as negative reinforcement because C made it clear that frequent prompting from her parents was irksome, and she would learn that she could terminate or avoid the prompting if she identified dysfunctional oral-motor behavior and corrected it independently.
Treatment progress
C was in the PHP program for 22 treatment days. She gained 4.9 lbs from intake to the end of PHP. At program discharge, she was 55.0 lbs with a BMI of 14.99. She was at the 22nd percentile BMI for her age and height. Biofeedback demonstrated that C could effectively remain relaxed while eating by utilizing deep breathing, and C reported increased comfort when swallowing while relaxed. Latency to swallow decreased from 19–38 to 5–8 seconds following implementation of the behavioral feeding strategies. C filled out the Children’s Depression Inventory (CDI; Kovacs, 1985) and the Multidimensional Anxiety Scale for Children (MASC; March, Parker, Sullivan, Stallings, & Conners, 1997) at intake and at end of treatment in the PHP. Her CDI score went from 4 to 5, and her MASC score went from 45 to 24.
Case presentation #2
Introduction
J was a 10-year-old Caucasian male who was referred to the program by his primary care provider for selective eating patterns, early satiety, and failure to gain weight over the previous 1.5 years, despite his height increasing from his typical 3rd–5th percentile to the 10th–15th percentile. At the time of intake, he was 64.4 lbs (21st percentile) and 52.17 inches (11th percentile) with a BMI of 16.6 (46th percentile).
History and initial presentation
J endorsed extreme sensitivity to the taste and texture of food and significant anxiety around trying new foods. He also experienced early satiety attributed to dysmotility from undernutrition. The foods J would eat at the time of intake included garlic bread, the breading of fried chicken nuggets, popcorn, vanilla shakes, and macaroni and cheese. He had chronic issues with eating since infancy when he had reflux, vomiting, and colic, as well as pica at the age of 2 years. He had not seen a gastroenterologist, hence a referral was made for further assessment while in PHP, which included an abdominal CT that ruled out bezoar; no other gastrointestinal concerns were identified. J had also been diagnosed with sensory integration disorder, attention-deficit/hyperactivity disorder (ADHD), and anxiety, which were managed with clonidine, dextroamphetamine/amphetamine, and melatonin.
J’s endorsed symptoms met criteria for an ARFID diagnosis. J described anxiety related to eating and fear that he would vomit or that food would taste bad. His parents also reported that J could only tolerate one food on the plate at a time, foods could not be of mixed type or touching, and he avoided vegetables, fruits, and most proteins. J denied low mood and wanted to learn to eat “everything I need.”
Treatment and goals
PHP level of care was recommended to address J’s selective food choices and subsequently promote normal growth and development. The goal of treatment was to broaden J’s range of accepted foods and to increase caloric consumption.
Feeding consult
J and his parents met with the pediatric feeding psychologist to further address his sensory disturbances with food. J’s mother reported that he would only eat specific brands of his preferred foods and inspected these foods “constantly” while eating them, which indicated anxiety even when eating his preferred foods. He permitted very little variation in his preferred foods (e.g. garlic toast had to be lightly toasted and dark spots were picked off). J was observed swallowing whole all bites of nonpreferred food by either chasing each bite with large amounts of fluid or filling his mouth with fluid before putting the bite into his mouth. When his family and the PHP staff attempted to require him to chew by withholding the fluid chaser, his refusal behaviors escalated significantly to include hissing and swinging his arms at people, throwing his plate, running away from the table, and hiding food.
Many novel foods were served to J during his first 10 days in the PHP, and many of those foods varied in their presentation each time they were served (e.g. different brands, cuts, and flavors of ham; canned, frozen, straight cut, and French-cut green beans). The pediatric psychologist concluded that J’s severe food selectivity was driven by a desire for complete predictability and consistency in the foods that he ate. He became highly anxious when unable to predict exactly how the food would taste and feel in his mouth. His anxiety was maintained by escape, delay, and avoidance of all foods that were not highly familiar. Swallowing foods whole was one of J’s strategies to avoid tasting or experiencing the texture of unfamiliar foods. In addition, J had learned that delaying eating unfamiliar foods by, for example, arguing or running away from the table, usually resulted in decreased requirements. Positive incentives, such as toys or fishing trips, which were offered contingent on cooperation with PHP requirements, were rarely effective because escape was a far more powerful incentive than were any other types of rewards.
PHP considerations and adaptations recommended by pediatric feeding psychologist
Staff and family provided support to J during program meals by encouraging him to take small bites of a new, challenging food before eating items he found acceptable. The number of challenge foods was decreased, and no additional foods were introduced until J was eating age-appropriate portions of previously introduced foods. Consistency and predictability of the taste and texture of the challenge foods were increased by holding the brand, presentation, and preparation of those foods constant. Staff increased bite size and the number of bites required based on behavioral indicators of low anxiety (e.g. easy cooperation with acceptance of bites, sufficient mastication) rather than increasing the number of bites required every day. By introducing unfamiliar foods more gradually, staff and family decreased J’s anxiety and therefore the intensity of his desire to escape, delay, and avoid, which decreased the intensity of his refusal behaviors and increased his cooperation with eating requirements. J and his parents were also told that he would not like a food when he first tried it. He was required to use correct oral-motor skills despite his disgust reactions, as unusual oral-motor movements increase the frequency of gagging and difficulty swallowing.
In the sessions with the pediatric psychologist, J’s parents were taught a specific progression of bolus sizes and practiced cutting food into these sizes while feedback was provided. Parents were also taught how to effectively negotiate and write concrete behavioral contracts that specified the challenge food(s) that they would be working on during a specified time period, the bolus size and number of bites required for each food, and the frequency of challenge food presentations. They included in each behavioral contract clear expectations as to which specific tasks J and his parents would be responsible for during food challenge presentations (e.g. J will decide which days of the week he is served challenge foods, parents will cut the food into the agreed upon bolus sizes, J will chew each bite at least three times, and his parents will count the number of chews). Finally, J’s parents were taught how to effectively use positive and negative reinforcement, response cost, and negative punishment to increase the likelihood that J would engage in appropriate eating behaviors and decrease the likelihood that he would engage in inappropriate mealtime behaviors.
Treatment progress
J gained 3.6 lbs during his 19 days enrolled in PHP. With a BMI of 17.5 at discharge, he finished the PHP at the 35th percentile BMI for his age and height. He was pleased with his weight gain and ability to eat more foods. His mother made changes in adjusting her own expectations at meal times, learned to repeatedly provide small exposures to the same food for a stretch before exposing to another new item, and not to “change the rules”/“back down” once seated at the table. J completed the CDI and MASC at intake but not at end of treatment so it was not possible to assess change in scores.
Case presentation #3
Introduction
P was a 14-year-old Caucasian male referred to the program by pediatric gastroenterology for his selective eating. At 105.2 lbs (11th percentile) and 69.1 inches (70th percentile), P had a BMI of 15.5 (0 percentile) at the time of intake.
History and initial presentation
P endorsed a general disinterest in food and eating, as well as limited variety, consistent with empirical findings of overlap among the different ARFID subtypes (Reilly et al., 2019). His dietary intake was limited to a small number of foods including sandwiches, frozen food items, and the item of the current food jag (corndogs at time of presentation). He denied purposely restricting calories or avoiding any food textures; however, he had a long history of abdominal pain and early satiety that interfered with more sufficient intake of the items he accepted. P’s weight was below what was expected for his age and stature, and P believed that his weight was too low.
Growth records were not available until the age of 8.2 years when he was first evaluated by nutrition and pediatric gastroenterology, with height at the 50th percentile, weight at the 7th percentile, and BMI below the 1st percentile; his mother reported he had been at the 25th percentile for weight previously. He had negative celiac screens with normal immunoglobulin A (IgA), complete blood count (CBC), and erythrocyte sedimentation rate (ESR) at 8 and 11 years of age. P was not on psychotropic medication at the time of presentation to PHP. Based on both patient and parent reports, P’s symptoms met criteria for an ARFID diagnosis. In addition to ARFID, P met criteria for anxiety disorder and depression.
Treatment and goals
The goals for P during PHP were to diversify his diet and promote weight gain.
PHP considerations and adaptations
P was not able to see the pediatric feeding psychologist during his stay in the PHP due to an extremely long waitlist. The eating disorder program psychologist met regularly with P’s mother to encourage her to work on increasing food variety, calories, and consistency in her interactions with P, similar to what would be done in outpatient FBT for AN.
Treatment progress
P completed 19 days in the PHP before stepping down to IOP for 12 days. P gained 14.0 lbs during his PHP stay, putting him at a BMI of 17.6. In IOP, he gained an additional 4.2 lbs bringing him to a total of 123.4 lbs with a BMI of 18.2 at the 18th percentile for his age and stature. P’s mother was successful in increasing the variety of foods he ate and not “backing down” in the face of resistance from the eating disorder. However, P’s interest in food did not seem to change, as he preferred to play video games than to eat. P’s CDI score went from 3 at intake to 1 at end of treatment, and his MASC score was 19 at intake and 18 at end of treatment.
Discussion
These three case studies represent patients with the three subtypes of ARFID who were successfully initially treated in a PHP/IOP for eating disorders, two of whom also met with the pediatric feeding psychologist. All patients responded to treatment, despite differences in their presentations.
As the program was originally developed for the treatment of AN and BN, there were important adjustments that were made in the treatment of patients with ARFID. For example, parents of children with AN are encouraged to quickly increase the amount of food eaten by the patient. In contrast, patients with ARFID are not prompted to immediately eat large portions of difficult-to-swallow, completely new, or highly non-preferred foods, but to work up to this gradually, while continuing to receive most of their caloric needs through preferred foods initially. In addition, exposure to increasingly difficult textures happens gradually for patients with ARFID, whereas in treatment for AN, exposure to foods that are being avoided by the patient but were eaten prior to the onset of the eating disorder are reintroduced into the patient’s diet fairly quickly.
A multidisciplinary team is necessary when treating low weight or nutritionally compromised patients with ARFID, although less severe cases may be successfully treated by a single practitioner with expertise in ARFID (Eddy et al., 2019). For the program in the current study, the pediatric feeding psychologist played an important role in treatment by teaching techniques to the patient, parents, and PHP/IOP staff, such as encouraging chewing on both sides of the mouth, avoiding pocketing food, gradually reducing excessive fluid intake, and anxiety management strategies. Interestingly, P improved despite not being able to meet with the pediatric feeding psychologist, although his mother did meet with the eating disorder psychologist. It is possible that patients in the “disinterested” subgroup of ARFID may benefit the most from a family-based approach. However, phase 2 of FBT, during which parents give responsibility over eating back to the patient in an age-appropriate manner, may prove challenging for patients who simply are not that interested in eating and do not find food to have the same reinforcing qualities that many people do. Future research will be needed to investigate this.
Educating parents around consistency and ways to avoid accommodating the eating disorder proved to be important. Prior to starting treatment, parents of patients with AN often endorse feeding their children a gradually more restrictive diet, eventually giving them foods that are not sufficient for maintaining a healthy weight but are all that their children are willing to eat. Similarly, parents of patients with ARFID reported giving their children what they were willing to eat rather than challenging the eating disorder and giving them what they need to be healthy. Parents often do this because they are unsure what to do in these situations and are afraid of challenging the eating disorder and making things worse, perhaps fearing that their children will stop eating altogether. J’s response to the initial attempts to increase his amount and range of food highlights this logical concern. Psychoeducation and coaching for parents of both disorders is crucial to making the necessary changes to restore their children to health, and parents may benefit from the expertise of a specialist who can analyze the unique driving factors that perpetuate an unhelpful eating behavior in children with ARFID, and who can provide recommendations specific to these factors. It is important to point out that unlike parents of children with AN, parents of children with the selective eating and chronic low appetite presentations of ARFID may never have been able to feed their child a normal or healthy diet, and may not have a healthy baseline to which to return. These families may still be candidates for a family-based approach, but may need a different kind of support.
It is important to note that the PHP/IOP was based on FBT principles but did not offer manualized FBT. Manualized FBT is an outpatient treatment (Lock & Le Grange, 2012) that cannot be replicated in higher levels of care. Thus, the current case examples provide some support for a multidisciplinary approach with parental involvement for patients with ARFID, but do not reflect the utility of manualized FBT. A limitation to the study is the lack of ARFID-specific assessment measures, such as the Eating Disorders in Youth—Questionnaire (EDY-Q; Kurz, van Dyck, Dremmel, Munsch, & Hilbert, 2015) or the Nine-Item ARFID Screen (NIAS; Zickgraf & Ellis, 2018) and the reliance primarily on weight as an outcome measure. Few changes in depression and anxiety were found, other than C’s decrease in anxiety on the MASC. The treatment of ARFID represents a challenge as well as an opportunity for the eating disorders field to develop and test effective interventions for this population.
Footnotes
Authors’ Note
Renee D Rienecke receives consulting fees from the Training Institute for Child and Adolescent Eating Disorders, LLC.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
