An infographic is available for this article at:https://links.lww.com/MPG/C614.
See “Invited Commentary Re: Prevalence and Characteristics of Avoidant/Restrictive Food Intake Disorder in Pediatric Neurogastroenterology Patients” by Oliveira and Kaul on page 547.
What Is Known/What Is New
What Is Known
- Avoidant/restrictive food intake disorder (ARFID) is a feeding and eating disorder that is not primarily characterized by body shape or weight concerns.
- ARFID symptoms have been reported in 13–40% in adult gastrointestinal functional/motility disorder samples.
What Is New
- In a retrospective study of pediatric patients undergoing neurogastroenterology examinations, we found symptoms of ARFID to be present in 23% of patients and to be most frequently related to fear of gastrointestinal symptoms.
- Pediatric patients undergoing neurogastroenterology or motility examinations should be evaluated for symptoms of ARFID.
- Further research is needed to inform prevention of ARFID in pediatric gastroenterology populations.
Individuals with gastrointestinal functional/motility disorders such as disorders of gut-brain interaction (DGBI) disorders of gut-brain interaction (1) are likely to avoid or restrict food intake in attempt to manage or prevent gastrointestinal symptoms (2). Recent studies show 13–40% (3–6) of adults with gastrointestinal functional/motility disorders can have avoidant/restrictive eating that results in medical impairment (eg, weight loss, nutritional deficiency) and/or psychosocial impairment consistent with avoidant/ restrictive food intake disorder (ARFID) (7). Disparate from shape/ weight-motivated eating disorders, avoidant/restrictive eating in ARFID is primarily motivated by any of three presentations—sensitivity to sensory characteristics, lack of interest in eating/ low appetite, and/or fear of aversive consequences related to food/eating (Supplemental Digital Content 1, https://links.lww.com/MPG/C615) (7,8). Although there is awareness of the need to screen for eating disorders in pediatrics (9), there are no data yet on ARFID in pediatric neurogastroenterology patients.
Among pediatric patients presenting to a tertiary care neuro-gastroenterology center for gastrointestinal functional/motility symptoms, we aimed to identify the frequency and characteristics of ARFID symptoms by Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5) criteria and examine their association withboth specific gastrointestinal symptoms and diagnoses. Based on previous studies of ARFID symptoms in adults with gastrointestinal functional/motility disorders (4,6), we hypothesized that the fear of aversive consequences presentation would be the most common (Supplemental Digital Content 1, https://links.lww.com/MPG/C615) and that both lower weight status and eating/weight-related symptom complaints would be significantly associated with likelihood of having ARFID symptoms. We also explored the most common symptom complaints, diagnoses, and demographics.
METHODS
Procedure
Participants included consecutively referred pediatric patients (N = 129; ages 6–18years; 57% female) who presented for initial evaluation to our tertiary care academic medical center for consultation with one neurogastroenterologist from January 1, 2016—December 31, 2018. We reviewed electronic medical records beginning with the initial clinic visits and including subsequent notes covering hospital and clinic visits (including mental health visits) that occurred through December 31, 2019. We used medical record review because no self-report measure for ARFID had been validated compared to clinician diagnosis at the time. Thus, electronic medical records allowed coders to obtain information provided by clinicians to confer ARFID diagnosis.
Three coders (F.U.R., C.B., C.J.S.) were trained to systematically review medical records for evidence of DSM-5 eating disorder criteria, including ARFID. One coder independently read each of their assigned case's longitudinal medical record notes including demographic and clinical characteristics: age, height, weight, and body mass index (BMI) percentile for age at initial consultation; presenting complaints (Table, Supplemental Digital Content 2, https://links.lww.com/MPG/C616); and gastrointestinal diagnoses (if any) conferred by providers. Coders completed a DSM-5 diagnostic checklist to indicate presence or absence of criteria for ARFID (with presentations) and other eating disorders. Coders conferred “Definite ARFID” when cases met all DSM-5 ARFID criteria. Coders conferred “Potential ARFID” when cases met some criteria for ARFID but not enough information was available to make a full diagnosis; for example, one case had an extremely limited diet (eg, five foods), but Criterion A (psychosocial and/or medical impairment) was unclear (eg, due to lack of documentation). Coders met on a weekly basis with a psychologist (HBM) to discuss ambiguous cases and come to agreement on coding. Other eating disorders (eg, anorexia nervosa, binge-eating disorder) were ruled out before conferring Definite or Potential ARFID diagnoses. “Potential ARFID” and “Definite ARFID” were only conferred when avoidant/restrictive eating symptoms were above and beyond what would be expected for the gastrointestinal condition, indicative that the ARFID symptoms warranted independent attention.
To complete inter-rater reliability for eating disorder diagnoses, raters completed blinded re-coding for 5% of randomly selected control cases and all cases with either Definite or Potential ARFID. We compared initial diagnostic coding to blinded coding by calculating percent agreement. The resulting Cohen's kappa was 0.60, demonstrating moderate agreement (10). All study data were collected using Research Electronic Data Capture; a secure, web-based application hosted at Massachusetts General Hospital. The Partners Human Research Committee provided scientific review and approval (Protocol #2018P002539).
Statistical Analysis
Using SPSS Statistics (v.24), we calculated frequency of cases satisfying all criteria for ARFID (Definite ARFID), and those for which ARFID diagnosis was likely but more information was needed (Potential ARFID), and the total group with ARFID symptoms (Definite + Potential ARFID). To evaluate specific hypotheses, we used the total group of those with Definite or Potential ARFID. For simplicity, we refer to this group as “ARFID symptoms.” We calculated frequencies of the three ARFID presentations (ie, sensory sensitivity, lack of interest in eating, and fear of aversive consequences).
Then, we conducted two logistic regressions to evaluate likelihood of having ARFID symptoms (0 = no; 1 = yes). In Model 1, we explored relations of likelihood of having ARFID symptoms by age, sex, and BMI percentile (Model 1) and by presenting complaints that were significantly more frequent in the ARFID group on univariate screen (Model 2). To have sufficient power to detect differences, we classified specific presenting complaints and diagnoses into categories (Supplemental Digital Content 2, https://links.lww.com/MPG/C616). We applied a conservative Bonferroni correction for each regression model. For exploratory/descriptive purposes, we also calculated frequencies of presenting complaints among those with ARFID versus without ARFID symptoms; calculated frequencies of specific gastrointestinal diagnoses and frequency of having multiple diagnoses across systems. For all descriptive comparisons, we conducted Mann-Whitney U tests for continuous data (because BMI percentile and age had significant right skew and kurtosis) and chi-square tests for categorical data.
RESULTS
Frequency and Characteristics of Avoidant/ Restrictive Food Intake Disorder
ARFID symptoms were present in 30 cases (23% of total cases), of which 11 cases (37%) had Definite ARFID (met all criteria) and 19 cases (63%) had Potential ARFID. No case was formally diagnosed with ARFID by a clinician before or after initial consultation during the 2-year timeframe of the chart review. In addition, no cases suspected with ARFID had a comorbid other eating disorder; thus, all cases with ARFID were included in our analyses. Compared to patients without ARFID symptoms, patients with ARFID symptoms were older (P < 0.001), had lower BMI percentile (P < 0.012), and were more likely to be female (77% vs 52%, P= 0.015) (Table 1). However, multivariable analysis showed that likelihood of having ARFID symptoms was only associated with older age and lower BMI percentile (Supplemental Digital Content 4, Model 1, https://links.lww.com/MPG/C618).
TABLE 1 -
Characteristics of pediatric neurogastroenterology patients by ARFID symptom presence (N = 129)
|
ARFID symptoms (n = 30) |
No ARFID symptoms (n = 99) |
U or chi∗ |
P value |
| Age, M (SD) |
13.9 (3.6) |
10.8 (3.7) |
815.00 |
<0.001 |
| Sex-female, n (%) |
23 (76.7%) |
51 (51.5%) |
5.96 |
0.015 |
| BMI percentile, M (SD)†
|
43.2 (32.8) |
60.5 (32.7) |
980.50 |
0.012 |
| Symptom complaints, n (%) |
| Eating/weight-related |
10 (33.3%) |
15 (15.2%) |
4.87 |
0.027 |
| Esophageal |
5 (16.7%) |
16 (16.2%) |
0.01 |
0.948 |
| Stomach |
7 (23.3%) |
19 (19.2%) |
0.25 |
0.620 |
| Lower GI |
18 (60.0%) |
80 (80.8%) |
5.46 |
0.019 |
| Abdominal pain |
13 (43.3%) |
23 (23.2%) |
4.62 |
0.032 |
| Gastrointestinal diagnoses, n (%) |
| Esophageal diagnosis |
9 (30.0%) |
36 (36.4%) |
0.41 |
0.522 |
| Stomach-related diagnosis |
4 (13.3%) |
6 (6.1%) |
1.70 |
0.192 |
| Lower GI diagnosis |
20 (66.7%) |
80 (80.8%) |
2.64 |
0.104 |
| Abdominal pain diagnosis |
6 (20.0%) |
15 (15.2%) |
0.40 |
0.529 |
ARFID = avoidant/restrictive food intake disorder; BMI = body mass index; GI = gastrointestinal; SD = standard deviation. ∗Continuous variables analyzed with Mann-Whitney U-test and categorical variables analyzed with chi-square tests. +BMI percentile data missing for n = 5.
Of those with ARFID symptoms, the fear of aversive consequences presentation was most common (n = 20, 67%). Patients reported fears of abdominal pain (n = 11), nausea (n = 8), bloating/ distention (n = 4), constipation (n = 3), vomiting (n = 3), chest pain (n = 1), throat pain (n = 1), and globus sensation (n = 1), including 10 (50%) who expressed fear of two or more gastrointestinal symptoms. Fourteen (47%) cases had the lack of interest in eating presentation. Five (17%) cases had the sensory sensitivity presentation. Some overlap was noted among presentations; 7 (23%) had both the fear of aversive consequences and lack of interest presentation and 1 (3%) with all three presentations (Supplemental Digital Content 3, https://links.lww.com/MPG/C617).
Of the four DSM-5 medical and psychosocial sequelae under Criterion A (of which presence of one or more of which is required for ARFID diagnosis), weight loss or inability to gain weight was the most common (n = 17 had explicit evidence, n = 4 did not have enough information). For nutritional deficiencies, n= 5 had explicit evidence (eg, iron deficiency anemia, vitamin D deficiency), n = 5 had an extremely restricted diet (eg, diet limited to chicken nuggets, pasta, ice cream), and n = 5 did not have enough information. For dependency on supplemental feeding, or oral or nutritional supplements, n = 4 had explicit evidence (eg, Ensure, Boost Plus). For psychosocial impairment, n = 2 had explicit evidence and n = 9 did not have enough information.
Exploratory Associations Between Gastrointestinal Symptoms and Avoidant/ Restrictive Food Intake Disorder
We explored the relation between presence of ARFID with gastrointestinal presenting complaints. Compared to patients without ARFID symptoms, patients with ARFID symptoms more frequently presented with eating/weight-related (15% vs 33%, P = 0.026) and abdominal pain complaints (23% vs 43%, P = 0.032), and less frequently presented with lower gastrointestinal complaints (81% vs 60%, P= 0.019) (Table 1); however, multivariable analysis showed no significant relative contribution of abdominal pain complaints and eating/weight-related complaints to likelihood of ARFID symptoms above and beyond age and BMI percentile (Supplemental Digital Content 4, https://links.lww.com/MPG/C618). Frequency of categorized gastrointestinal diagnoses did not significantly differ between groups (Table 2). For descriptive purposes, we explored the frequencies of specific gastrointestinal complaints and diagnoses within those with ARFID (we did not conduct statistical comparisons). In the ARFID group, the most common complaints were abdominal pain (n = 13, 43%) and lower gastrointestinal related symptoms including constipation (n = 16, 53%) (Table 2); the most common diagnosis was chronic constipation (60%).
TABLE 2 -
Presenting GI complaints and diagnosis of patients with ARFID symptoms (
n= 30)
| Presenting GI complaint |
n (%) |
| Constipation |
16 (53%) |
| Abdominal pain |
13 (43%) |
| Nausea |
7 (23%) |
| Low weight/underweight |
6 (20%) |
| Weight loss |
5 (17%) |
| Vomiting |
4 (13%) |
| Bloating/distention |
4 (13%) |
| Fecal incontinence |
3 (10%) |
| Diarrhea/loose stools |
3 (10%) |
| Dysphagia |
2 (7%) |
| Poor appetite |
1 (3%) |
| Feeding intolerance |
1 (3%) |
| Globus |
1 (3%) |
| Heartburn |
1 (3%) |
| Chest pain |
1 (3%) |
| Regurgitation |
1 (3%) |
| GI diagnosis |
n (%) |
| Chronic constipation |
18 (60%) |
| Functional abdominal pain |
7 (23%) |
| Chronic nausea |
4 (13%) |
| Gastroesophageal reflux disease |
4 (10%) |
| Chronic vomiting |
3 (13%) |
| Irritable bowel syndrome |
2 (7%) |
| Constipation predominant |
1 |
| Diarrhea predominant |
1 |
| Dysphagia |
2 (7%) |
| Rumination syndrome |
2 (7%) |
| Gastroparesis |
1 (3%) |
Percentages do not add up to 100% since some patients had multiple presenting complaints and diagnoses. ARFID = avoidant/restrictive food intake disorder; GI = gastrointestinal.
DISCUSSION
Among pediatric neurogastroenterology patients ages 6–18 years, DSM-5 ARFID symptoms were common (up to 23%) and most frequently motivated by fear of aversive consequences, similar to findings in adults with gastrointestinal functional/motility disorders (4–6). Contrary to our hypothesis that eating/weight-related complaints would be most associated with ARFID symptom presence, we found that only older age and lower weight status (by BMI percentile) were significantly associated with ARFID presence. While future research is needed with prospective studies and larger samples, we believe these data add to the growing literature on ARFID symptoms in patients with gastrointestinal functional/ motility disorders.
The relatively high frequency of ARFID symptoms in our pediatric neurogastroenterology sample could be critical information for treatment decision-making. Patients with DGBI are likely to use food avoidance in an attempt to prevent symptoms (2) and despite limited data on the efficacy of diet elimination approaches in pediatrics (eg, low FODMAP diet) in pediatrics (11,12), some providers may recommend dietary eliminations based on efficacy in adults (13). At face value, dietary elimination approaches contradict the current first-line treatment for ARFID, and may put some patients at risk for developing ARFID. Exposure-based cognitive-behavioral therapy (CBT) for ARFID supports exposure to avoided foods or food amounts to test feared predictions around eating (14,15). Although the re-introduction phase of some elimination diets like the low FODMAP diet (16) may act as a form of exposure, it frequently fails in the absence of close dietitian supervision. For example, even if a child and their family attempt food reintroduction after a dietary elimination phase, some children may develop fear of what will happen when they try foods again or will have difficulty tolerating gastrointestinal symptoms that naturally occur during re-introduction. Providers should not prescribe restrictive exclusion diets to pediatric patients who have ARFID symptoms (17), and future research is needed to identify for whom these approaches pose risk for ARFID development.
We also explored the frequency of specific gastrointestinal presentations in those with versus those without ARFID symptoms. In contrast to our adult study in which lower gastrointestinal diagnoses were associated with a higher odds of ARFID symptoms (6), lower gastrointestinal (GI) complaints were less frequent among our ARFID versus non-ARFID pediatric group; however, the most common presenting complaints and diagnoses among patients with ARFID symptoms were related to constipation and abdominal pain. This might explain the high frequency of the lack of interest presentation in our sample (47%). In fact, decreased appetite can be an accompanying symptom of functional constipation (1); however, further research is needed to identify the association of ARFID symptoms with specific gastrointestinal presentations, particularly with patients who have confirmed ARFID diagnoses.
Also in contrast to a higher frequency of males with ARFID in a general pediatric gastroenterology sample (18), we found ARFID symptoms had a trend association with female sex (77% in ARFID group vs 51% in no-ARFID group). Similar to our adult study (6), the sex disparity is possibly due to the lower representation of the ARFID sensory sensitivity presentation in neurogastroenterology patients (10% in this study vs 21% in the general pediatric sample) (18).
This study had several limitations related to the retrospective chart review design, specifically the reliance on free text information and lack of discrete data collection leading to possible over- or under-identified cases and symptoms, moderate (rather than substantial or almost perfect) original rater agreement with blinded coding, and inability to identify other factors that could contribute to ARFID symptoms. Our sample size also limited our ability to adjust for additional potential confounds. Finally, while we used DSM-5 ARFID criteria, there are calls to improve ARFID criteria standards that may alter its operationalization in the future (19).
Our pilot findings provide novel information on characteristics associated with ARFID symptoms among pediatric neurogastroenterology patients, which we hope will inform future prospective work. We suggest providers consider screening for ARFID using both an ARFID self-report measure and a shape/weight-motivated eating disorder measure (to screen for and in the case of ARFID, rule-out significant other eating disorders) (20). However, providers should partner with behavioral health specialists (eg, psychologists) for case consultation and evaluation when available to identify appropriate diagnostic and therapeutic steps. Providers may also consider discussing the possibility of developing problematic avoidant/restrictive eating to support patients in maintaining adequate nutrition and a flexible relationship with eating.
REFERENCES
1. DrossmanD, Chang L, Chey WD (Eds):et al. ROME IV: Functional Gastrointestinal Disorders – Disorders of Gut–Brain Interaction. 4th ed. Raleigh, NC: Rome Foundation; 2016. No. 2.
2. Reed-Knight B, Squires M, Chitkara DK, et al. Adolescents with irritable bowel syndrome report increased eating-associated symptoms, changes in dietary composition, and altered eating behaviors: a pilot comparison study to healthy adolescents.
Neurogastroenterol Motil 2016; 28:1915–1920.
3. Burton Murray H, Kuo B, Eddy KT, et al. Disorders of gut-brain interaction common among outpatients with eating disorders including avoidant/restrictive food intake disorder.
Int J Eat Disord 2021; 54:952–958.
4. Murray HB, Jehangir A, Silvernale CJ, et al. Avoidant/restrictive food intake disorder symptoms are frequent in patients presenting for symptoms of gastroparesis.
Neurogastroenterol Motil 2020; 32:e13931.
5. Harer K, Baker J, Reister N, et al. Avoidant/restrictive food intake disorder in the adult gastroenterology population: an under-recognized diagnosis? Philadelphia, PA: American College of Gastroenterology Annual Meeting; 2018.
6. Murray HB, Bailey AP, Keshishian AC, et al. Prevalence and characteristics of avoidant/restrictive food intake disorder in adult neurogastroenterology patients.
Clin Gastroenterol Hepatol 2019; 18: 1995.e1–2002.e1.
7. APA. Diagnostic and Statistical Manual of Mental Disorders (DSM-5). 5th ed.Arlington, VA: American Psychiatric Publishing; 2013.
8. Thomas JJ, Lawson EA, Micali N, et al. Avoidant/restrictive food intake disorder: a three-dimensional model of neurobiology with implications for etiology and treatment.
Curr Psychiatry Rep 2017; 19:54.
9. Vriesman MH, Koppen IJN, Camilleri M, et al. Management of functional constipation in children and adults.
Nat Rev Gastroenterol Hepatol 2020; 17:21–39.
10. Landis JR, Koch GG. The measurement of observer agreement for categorical data.
Biometrics 1977; 33:159–174.
11. Chumpitazi BP, Cope JL, Hollister EB, et al. Randomised clinical trial: gut microbiome biomarkers are associated with clinical response to a low FODMAP diet in children with the irritable bowel syndrome.
Aliment Pharmacol Ther 2015; 42:418–427.
12. Brown SC, Whelan K, Gearry RB, et al. Low FODMAP diet in children and adolescents with functional bowel disorder: a clinical case note review.
JGH Open 2020; 4:153–159.
13. Dionne J, Ford AC, Yuan Y, et al. A systematic review and meta-analysis evaluating the efficacy of a gluten-free diet and a low FODMAPs diet in treating symptoms of irritable bowel syndrome.
Am J Gastroenterol 2018; 113:1290–1300.
14. Thomas JJ, Eddy K. Cognitive-Behavioral Therapy for Avoidant/Restrictive Food Intake Disorder: Children, Adolescents, and Adults Cambridge. UK: Cambridge University Press; 2019.
15. Thomas JJ, Becker KR, Kuhnle MC, et al. Cognitive-behavioral therapy for avoidant/restrictive food intake disorder: feasibility, acceptability, and proof-of-concept for children and adolescents.
Int J Eat Disord 2020; 53:1636–1646.
16. Whelan K, Martin L, Staudacher H, et al. The low FODMAP diet in the management of irritable bowel syndrome: an evidence-based review of FODMAP restriction, reintroduction and personalisation in clinical practice.
J Hum Nutr Diet 2018; 31:239–255.
17. Chey WD. Elimination diets for irritable bowel syndrome: approaching the end of the beginning.
Am J Gastroenterol 2019; 114:201–203.
18. Eddy KT, Thomas JJ, Hastings E, et al. Prevalence of DSM-5 avoidant/restrictive food intake disorder in a pediatric gastroenterology healthcare network.
Int J Eat Disord 2015; 48:464–470.
19. Harshman SG, Jo JH, Kuhnle MC, et al. A moving target: how we define avoidant/restrictive food intake disorder can double its prevalence.
J Clin Psychiatry 2021; 82:20m13831.
20. Murray HB, Dreier MJ, Zickgraf HF, et al. Validation the nine item ARFID Screen (NIAS) subscales for distinguishing ARFID presentations and screening for ARFID.
Int J Eat Disord 2021; 54:1782–1792.
