Avoidant/Restrictive Food Intake Disorder Symptoms Are Not as Frequent as Other Eating Disorder Symptoms When Ulcerative Colitis Is in Remission

Abstract

1. Introduction

Avoidant/restrictive food intake disorder [ARFID] is an eating disorder characterised by food avoidance/restriction leading to medical and/or functioning impairments. Distinguished from other eating disorders by an absence of body image disturbance motivating food restriction, ARFID motivations include one or more of: fear of aversive consequences [eg, gastrointestinal symptoms], lack of interest in eating/low appetite, and sensitivity to the sensory characteristics of food [eg, taste, texture].¹ There is growing debate about where the line is for reasonable food avoidance/restriction crossing the threshold into ARFID, particularly in the context of gastrointestinal disorders.^2,3

For gastrointestinal disorders classified as organic, such as inflammatory bowel diseases [IBD], diet modifications are commonly prescribed or self-initiated by patients. A high proportion [90%] of patients with IBD report avoiding foods to manage symptoms during active disease,⁴ and management guidelines provide recommendations on diet modifications dependent on disease state.⁵ Food avoidance may be adaptive and non-problematic for IBD patients in flare, but some patients may continue to avoid foods during remission periods.^6,7 Two recently published studies reported ARFID symptoms in 14–53% of adults with IBD using the self-report Nine Item ARFID Screen [NIAS].^2,6 A limitation of both studies is that the NIAS cutoff scores used had not been validated in samples diagnosed with ARFID.² As active disease markers have been significantly associated with ARFID symptom presence,⁶ assessing ARFID symptoms during remission could shed light on more accurate rates of ARFID.

2. Case report

Among adults with ulcerative colitis [UC] in remission, we aimed to evaluate the frequency and characteristics of ARFID symptoms using the NIAS. We used the 8-item Eating Disorder Examination-Questionnaire [EDE-Q-8] both as a rule out for ARFID symptoms and to characterise other eating disorder symptoms.

Participants included adults with UC recruited from the Massachusetts General Hospital [MGH] Gastroenterology Division and participating in an ongoing cohort study with quiescent UC. Inclusion criteria included established UC diagnosis confirmed by colonoscopy or flexible sigmoidoscopy and quiescent disease (Simple Clinical Colitis Index of Activity [SCCAI] ≤2 or faecal calprotectin <150 µg/g with corticosteroid-free clinical remission for ≥3 months) at baseline. Exclusion criteria included Crohn’s disease, indeterminant colitis, current pregnancy/breast-feeding, antibiotic usage within 3 months, active treatment with corticosteroids for UC. We used self-reported data on demographics, gastrointestinal medications, medical comorbidities, NIAS scores, and other eating disorder symptom scores [EDE-Q-8].⁸ We applied the following cutoffs for possible ARFID on any NIAS subscale—≥10 Picky Eating, ≥9 Interest; ≥10 Fear.⁹ We applied the following cutoffs for possible other eating disorder symptoms on the EDE-Q-8—≥2.3 and ≥3.88⁸ Global score. Because one prior study with [non-UC] patients diagnosed with eating disorders showed that use of both the NIAS and another eating disorder measure was needed to correctly classify ARFID from other eating disorders,⁹ we also reported the frequency of participants who screened positively on the NIAS and on the EDE-Q-8 [≥2.3 or ≥3.88 Global score].

We included 101 participants who completed the NIAS at their baseline cohort assessment [age 49.9 ± 16.5; 55% female; Table 1]. Eleven participants [11%] screened positively on at least one NIAS subscale [n = 8 male]. The overall distribution of scores on the NIAS showed most participants scored negatively on all items [‘slightly disagree’, ‘disagree’, or ‘strongly disagree’], with less than 10% of participants endorsing positive responses [Supplementary Figure 1]. Among 11 participants meeting our screening criteria for ARFID symptoms, positive screens by subscale were: n = 3 Picky Eating, n = 4 Interest, n = 2 Fear, n = 2 Picky Eating and Interest. Thirty participants [30%] screened positive for other eating disorder symptoms by EDE-Q-8 ≥2.3 [age range = 24-76 years; n = 9 male], with 12 [12%] screening positive by EDE-Q-8 ≥ 3.88. Overall score distributions on the EDE-Q-8 showed that participants scored highest on the Weight Concern and Shape Concern subscales [Supplementary Figure 2]. Two participants [2%] screened positively for ARFID on the NIAS and for other eating disorder symptoms by EDE-Q8 Global ≥2.3. No participants screened positively for ARFID on the NIAS and on the EDE-Q8 Global ≥3.88.

3. Discussion

Our low positive screening rate for ARFID symptoms [11%] contrasts with 53% positive screens reported among adults with IBD with even more conservative [higher] cutoffs on the NIAS.^2,6 However, previous findings show that active symptoms and active inflammation predicted higher positive ARFID screening rates.⁶ Our findings suggest that ARFID screening rates using the NIAS may be lower when UC is in remission.

The NIAS is intended to measure severity of the three primary motivations behind restrictive eating in ARFID [sensory sensitivity, lack of interest/low appetite, and fear of aversive consequences].⁹ Most participants in our sample scored negatively on the NIAS items, which may indicate that ARFID symptoms are overall low during remission. The Fear subscale had the lowest average score and lowest positive screening rate [2%], contrasting with two prior IBD studies showing the highest score on the Fear subscale.^2,6 Fear around eating may be more common among individuals with active symptoms. High fear around eating can become problematic, perpetuating anxiety and avoidance behaviours around food intake,¹⁰ which ARFID symptoms in UC reflect. Given that the medical consequences of avoidant/restrictive eating in ARFID [eg, weight loss, malnutrition] may occur in the context of active disease, it can be difficult to determine when a diagnosis of ARFID is appropriate. Importantly, ARFID cannot definitively be diagnosed using self-report tools, because clinical judgment is required to apply psychiatric criteria [eg, the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, DSM-5].¹ Future research using semi-structured diagnostic interviews [eg, the Pica, ARFID, and Rumination Disorder Interview, PARDI]¹¹ should evaluate ARFID prevalence in UC more rigorously.

Positive screening rates for other eating disorder symptoms were relatively high [up to 30%], with the highest scores reflected in items assessing concerns about body shape/weight. Notably, about one-third of participants with a positive screen were male. These rates are higher than previously reported, with a recent systematic review reporting 6–13% of adults with IBD with a positive screen on a different self-report measure for non-ARFID eating disorder symptoms.¹² We were not able to evaluate the temporal relation between non-ARFID eating disorder symptoms and UC development, but prior research has shown that UC is associated with prospective risk for developing an eating disorder in females.¹³ Further longitudinal research is needed, including to identify risk mechanisms.

In this exploratory study, among adults with UC in remission, we found a low rate of ARFID symptoms by the NIAS [11%] but a high rate of positive screens for other eating disorder symptoms [up to 30%]. Other eating disorder positive screens appeared driven by symptoms of body image disturbance. Clinicians should consider screening patients with UC in remission for eating disorder symptoms, with consideration that dietary restriction may be motivated by shape/weight concerns or ARFID-related motivations [eg, fear of gastrointestinal symptoms]. Future research is needed to identify biological and behavioural risk factors that could contribute to comorbidity between UC and eating disorders.

Supplementary Data

Supplementary data are available at ECCO-JCC online.

Funding

This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases—K23 DK131334 [HBM]. Our funding sources played no role in study design, data collection, data analysis and interpretation, writing of the report, or in the decision to submit the article for publication.

Conflict of Interest

HBM and JJT receive royalties from Oxford University Press for their forthcoming book on rumination syndrome. HK has received consulting fees from Takeda and Aditum Bio and has received grant funding from Takeda and Pfizer. JJT receives royalties from Cambridge University Press for the sale of her books, Cognitive-Behavioral Therapy for Avoidant/Restrictive Food Intake Disorder: Children, Adolescents, and Adults and The Picky Eater’s Recovery Book: Overcoming Avoidant/Restrictive Food Intake Disorder.

Author Contributions

HBM: concept, data collection, data analysis, funding acquisition, original draft writing, final draft reviewing, and editing. KK: data analysis and final draft reviewing and editing. JG: patient recruitment, data collection, and final draft reviewing and editing. KW: patient recruitment, data collection, and final draft reviewing and editing. JJT: concept and final draft reviewing and editing. HK: concept, investigation, and final draft reviewing and editing.

Acknowledgements

We thank Tiffany Taft, Psy.D. for her contribution as a co-author for a previous conference abstract presentation [at the 2022 Association for Behavioral and Cognitive Therapies annual meeting] using a subset of data in the current manuscript.

Data Availability

The data underlying this article are available in the article and in its online Supplementary material.

References