The Association of Avoidant/Restrictive Food Intake Disorder (ARFID) and Neurogastroenterology Disorders (Including Disorders of Gut–Brain Interaction [DGBI]): A Scoping Review

ABSTRACT

Background

Patients with neurogastroenterology disorders like disorders of gut–brain interaction (DGBI) and gastrointestinal (GI) motility disorders often adopt restrictive diets to manage symptoms. Without professional guidance, these patients may risk developing avoidant/restrictive food intake disorder (ARFID), potentially affecting their physical and mental health.

Purpose

This scoping review aimed to explore the prevalence of ARFID in patients with neurogastroenterology disorders and vice versa, the direction of their association, potential risk factors, and available treatments.

Methods

Following PRISMA‐ScR guidelines, we searched PubMed, Web of Science, and Cochrane. Abstracts were screened for eligibility by two independent reviewers.

Key Results

Eighteen studies met our inclusion criteria. The prevalence of ARFID symptoms in neurogastroenterology patients ranged from 10% to 80%, while the prevalence of neurogastroenterology disorders and related GI symptoms in ARFID patients ranged from 7% to 60%. Findings on the direction of the association between eating difficulties and GI symptom occurrence were conflicting. Patients with ARFID‐neurogastroenterology disorder overlap were more likely to be female, have a lower BMI, higher anxiety and depression levels, and poorer quality of life. Two small studies evaluating treatment for this overlap suggested promising effects of cognitive behavioral therapy (CBT).

Conclusions and Inferences

This review highlights heterogeneity in study designs and questions the suitability of ARFID assessment tools in this context. It also underscores gaps in understanding the underlying pathophysiology and treatment approaches. Future research should prioritize validating ARFID screening tools specific to this population and standardizing study methodologies. Improved understanding of this overlap will help healthcare professionals improve management strategies and patient outcomes.

This scoping review found that avoidant/restrictive food intake disorder (ARFID) symptoms are highly prevalent in neurogastroenterology disorders such as disorders of gut–brain interaction (DGBI), and vice versa. Patients with an ARFID‐neurogastroenterology disorder overlap were found to have higher anxiety and depression levels, and a lower quality of life and body mass index (BMI) compared to those without this overlap.

Summary.

• The overlap between ARFID and neurogastroenterology disorders is prevalent and affects patients' mental health and quality of life.

• Patients with an ARFID‐neurogastroenterology disorder overlap are more likely to be female, to have lower BMI, higher anxiety and depression levels, and poorer quality of life.

• Validated ARFID screening tools in neurogastroenterology and standardized research are needed to improve management.

1. Introduction

Patients with neurogastroenterology disorders including disorders of gut–brain interaction (DGBI) and motility disorders tend to restrict or avoid some food items to improve their gastrointestinal (GI) symptoms [1, 2]. Self‐restrictive diets, such as those low in fermentable oligosaccharides, disaccharides, monosaccharides, and polyols (FODMAP), especially if followed without professional guidance, can have negative effects on both nutritional and psychological aspects of health [3]. These impacts go beyond the risk of nutrient deficiencies, extending to potential harm to emotional well‐being, mental health, and eating behaviors [3]. For instance, severe food avoidance and restriction were found to be associated with more severe symptoms and a reduced quality of life in IBS patients, and with more eating disorders [4, 5].

Avoidant/restrictive food intake disorder (ARFID) is a new entity, formerly known as “feeding disorder of infancy or early childhood” before 2013. It was added to the Diagnostic and Statistical Manual, 5th Edition (DSM‐5) in 2013, broadening its diagnosis to all age groups [6]. Unlike other eating disorders, ARFID does not result from preoccupations with body weight or image, and individuals with ARFID are not necessarily underweight [7]. Instead, it presents with three main characteristics: (1) avoidance of foods with certain sensory characteristics (such as texture, taste, smell); (2) fear of aversive consequences of eating (such as GI symptoms); (3) lack of interest in food or low appetite [7]. To be diagnosed with ARFID, patients must have medical and/or psychosocial impairment due to their restrictive eating, which cannot be explained by other conditions (Table 1) [7, 8].

TABLE 1.

DSM‐5 ARFID diagnostic requirements.

Required features a	Additional clinical features
Medical impairment: significant weight loss; clinically significant nutritional deficiencies; dependence on oral nutritional supplements or tube feeding	Avoidance of foods with certain sensory characteristics (texture, taste, smell)
Psychosocial impairment: in personal, family, social, educational, occupational or other important areas of functioning (avoidance or distress related to participating in social experiences involving eating)	Fear of aversive consequences of eating (choking, vomiting, other gastrointestinal symptoms)
	Lack of interest in eating or low appetite
Exclusion criteria b
Eating patterns motivated by preoccupation with body weight/shape Unavailability of food Other medical conditions (e.g., food allergies, hyperthyroidism) or mental disorders Substance/medication effects on the CNS, including withdrawal effects

Abbreviations: ARFID, avoidant/restrictive food intake disorder; CNS, central nervous system.

^a Required features: Food avoidance or restriction must lead to medical and/or psychosocial impairment.

^b Exclusion criteria: Food avoidance or restriction and its effect on weight and other aspects of health and functioning should not be explained by any of the above conditions.

Source: ICD‐11 for Mortality and Morbidity statistics.

Given that restrictive diets are often part of managing neurogastroenterology disorders, it is crucial to assess the risk of eating disorders, including ARFID, and address the coexistence of these conditions to adapt management and avoid restrictive diets that could worsen health [9]. The very few reviews conducted on the association between ARFID and GI disorders primarily aim to provide insights into identifying and managing this overlap [10, 11, 12, 13]. However, to our knowledge, no scoping review has specifically focused on ARFID and neurogastroenterology disorders. For these reasons, a scoping review was conducted in order to systematically map the research done in this area, as well as to identify any existing gaps in knowledge. Therefore, this scoping review aims to provide an overview of the association of ARFID and neurogastroenterology disorders, by (1) examining the prevalence of ARFID in neurogastroenterology disorders and vice versa, (2) examining the direction of this association, (3) exploring the risk factors, and (4) reviewing available treatments.

2. Materials and Methods

The Preferred Reporting Items for Systematic reviews and Meta‐Analyses extension for Scoping Reviews (PRISMA‐ScR) Checklist was used as a guide for this review. The protocol was drafted using the JBI protocol template for scoping reviews and was revised by all co‐authors. The final protocol was registered in FigShare on March 25, 2024 [14].

2.1. Eligibility Criteria

We only included studies conducted on humans, with both adult and pediatric populations being eligible. They had to be published since at least 2013 (year of the inclusion of ARFID in DSM‐5). Included studies were those conducted on patients with neurogastroenterology disorders, such as DGBI and motility disorders like gastroparesis overlapping with ARFID. If conducted on other GI disorders, studies were excluded. Abstracts, but not reviews, were also eligible.

2.2. Search Strategy and Data Extraction

First, an experienced librarian (EL) helped with the search strategy, which was then discussed and refined by the co‐authors. Our search strategy included a broad number of keywords (found in Appendix S1), in order to be as exhaustive as possible, as the subject of this scoping review is not frequently studied. The following databases were searched from February 22, 2024 until May 1, 2024: PubMed, Web of Science, and Cochrane. The final search strategy can be found in our aforementioned scoping review protocol [14]. The final search outcomes were transferred to Rayyan, and we subsequently eliminated duplicates [15].

2.3. Data Assessment of Quality and Selection

Titles and abstracts were initially screened independently by two reviewers (H.M.‐M. and V.B.) using Rayyan. A third reviewer (C.M.) verified and resolved disagreements on study selection. Full‐text article review was then conducted by H.M.‐M. If they did not meet the inclusion criteria, articles were excluded. For instance, we excluded articles that explored general “food avoidance” rather than focusing specifically on ARFID, as defined by the DSM‐5 criteria and the 3 ARFID presentations, using validated questionnaires.

2.4. Data Extraction

An extraction tool developed by the reviewers was used for data extraction [14]. Extracted data included specific details about the participants, aims, study model, methods, and relevant findings and conclusions.

3. Results

In total, 8392 records were identified through our search strategy across the 3 databases, but only 29 were retrieved (Figure 1). During screening, 11 reports were excluded, of which 5 were abstracts that were later published as full articles and were included in our review [4, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]. This left us with 18 studies ultimately included. Regarding our third aim mentioned in the protocol, we were unfortunately unable to achieve it due to a lack of studies on the mechanisms underlying ARFID development [14]. Instead, we replaced this aim with an analysis of the association direction between ARFID and neurogastroenterology disorders, as outlined in Section 3.3.

FIGURE 1.

Preferred reporting items for systematic reviews and meta‐analyses extension for scoping reviews (PRISMA‐ScR) flowchart for data selection and inclusion. Source: Page MJ, et al. BMJ 2021;372:n71. https://doi.org/10.1136/bmj.n71.

3.1. Prevalence of ARFID in Neurogastroenterology Disorders

We found 10 studies exploring ARFID in patients with neurogastroenterology disorders, of which 8 were original articles and 2 were abstracts (presented in scientific congresses) (Table 2). The prevalence of ARFID symptoms was found to range from 10% to 80% in patients with neurogastroenterology disorders, depending on the diagnostic criteria used. For the assessment of ARFID symptoms, 3 studies applied the DSM‐5 ARFID criteria (which involved ruling out other eating disorders and exploring psychosocial and medical impairment), 1 study used a modified ARFID Canadian Pediatric Surveillance Program questionnaire, and the remaining 6 studies used the Nine Item Avoidant/Restrictive Food Intake Disorder Screen (NIAS) [26, 27, 28, 29, 30, 31, 32, 33, 34, 35]. One of these 6 studies using the NIAS also used the Pica, ARFID, Rumination Disorder Interview‐Questionnaire (PARDI‐AR‐Q) for result comparison [31]. Only 3 studies were conducted on/included pediatric patients [26, 31, 33]. The tools used for psychosocial and medical impairment assessment (if mentioned), along with the elimination of other eating disorders in each study, are summarized in Table 2.

TABLE 2.

Summary of studies exploring the prevalence of ARFID symptoms in patients with neurogastroenterology disorders.

Study	Population and sample size	Study design	ARFID diagnostic criteria	Cut‐offs used	Elimination of other ED	Evaluation of medical/psychosocial impairments	ARFID prevalence
Atkins et al. Neurogastroenterol Motil. 2022 [26]	495 (119 pediatric and 376 adult) neurogastroenterology outpatients and inpatients	Retrospective chart review	DSM‐5 criteria	None	Yes (DSM‐5 criteria)	Yes (documentation/chart information)	25% of pediatric and 23% of adult patients had ARFID symptoms
Burton Murray et al. Neurogastroenterol Motil. 2020 [27]	288 adult outpatients with Gp/dyspepsia symptoms	Retrospective chart review	NIAS	Cut‐off score of ≥ 12 for each of three subscales to indicate participants “agreed” or “strongly agreed” with all 3 items in each subscale	Yes (EDDS)	Yes (unclear method)	39.9% of patients had ARFID symptoms. 23.3% had “definite ARFID”
Burton Murray et al. Neurogastroenterol Motil. 2022 [28]	93 adult outpatients with DGBI	Prospective	Modified ARFID Canadian Paediatric Surveillance Program Questionnaire	Unknown	Yes (EDE‐Q)	Yes (self‐report survey)	39.8% of patients had ARFID symptoms
Fink et al. Clin Gastroenterol Hepatol. 2022 [29]	289 adult outpatients with GI disorders	Prospective	NIAS	Scores > 23 on the total scale or > 12 on an individual subscale	No	No	75.7% of achalasia patients had ARFID symptoms
Hollis et al. Neurogastroenterol Motil. 2024 [30]	177 adult outpatients with Gp	Prospective	NIAS	≥ 10 on the picky subscale, ≥ 9 on the appetite subscale, and ≥ 10 on the fear subscale	No	Yes (ARFID checklist: questions about impairments scored from 0 to 2)	80.4% of patients had ARFID symptoms. 76.6% of patients had ARFID symptoms with medical/psychosocial impairment
Kaul et al. Neurogastroenterol Motil. 2024 [31]	171 pediatric outpatients: 33 with Gp, 66 with FD, 72 HC	Prospective longitudinal	NIAS and PARDI‐AR‐Q	≥ 10 on the picky subscale, ≥ 9 on the appetite subscale, and ≥ 10 on the fear subscale	Yes (EDE‐Q)	Yes (ARFID checklist: questions about impairments scored from 0 to 2)	At baseline: 48.5%–63.6% of Gp patients; 65.1%–66.7% of FD patients. At 2‐month follow‐up: 53.3%–71.9% of patients positive at baseline
Burton Murray et al. Clin Gastroenterol Hepatol. 2020 [32]	410 adult neurogastroenterology outpatients and inpatients	Retrospective chart review	DMS‐5 criteria	None	Yes (DSM‐5 criteria)	Yes (documentation/chart information), but unknown results	23.7% of patients had ARFID symptoms, 6.3% had definite ARFID
Burton Murray et al. J Pediatr Gastroenterol Nutr. 2022 [33]	129 pediatric neurogastroenterology outpatients and inpatients	Retrospective chart review	DSM‐5 criteria	None	Yes (DSM‐5 criteria)	Yes (documentation/chart information)	23% of patients had ARFID symptoms, 8.5% had definite ARFID
Silva et al. J Gastroenterol Hepatol. 2022. Abstract nr 381 [34]	74 adult IBS patients	Prospective	NIAS	Unknown	Unknown	Unknown	30% of IBS patients (educated on a low‐FODMAP diet) had ARFID symptoms
Blomsten et al. Neurogastroenterol Motil. 2023. Abstract nr NGS18129‐93 [35]	536 adults meeting frequency threshold for at least one FBD	Prospective	NIAS	≥ 10 on the picky subscale, ≥ 9 on the appetite subscale, and ≥ 10 on the fear subscale	Yes (SCOFF)	Unknown	9.9% of patients with FBD and 27.2% of patients with FD + FBD had ARFID symptoms

Note: ARFID symptoms represent “definite ARFID” (when all DSM‐5 criteria for ARFID are met), and “potential ARFID” diagnoses (when patients meet at least one ARFID presentation but either lack sufficient data or do not meet all DSM‐5 criteria for ARFID).

Abbreviations: ARFID, avoidant/restrictive food intake disorder; DGBI, disorders of gut–brain interaction; DSM5, diagnostic and statistical manual of mental disorders, fifth edition; EDDS, eating disorder diagnostic scale; EDE‐Q, eating disorder examination questionnaire; FBD, functional bowel disorder; FD, functional dyspepsia; FODMAP, fermentable oligosaccharides, disaccharides, monosaccharides, and polyols; Gp, gastroparesis; IBS, irritable bowel syndrome; NIAS, nine item avoidant/restrictive food intake disorder screen; SCOFF, sick control one fat food questionnaire.

3.1.1. Studies in Unspecified Neurogastroenterology Disorders

Three of the included articles were conducted on neurogastroenterology populations using DSM‐5 ARFID criteria for ARFID symptom assessment [26, 32, 33]. In the pediatric neurogastroenterology population, ARFID symptom prevalence ranged from 23% to 25% and in the adult neurogastroenterology population it was around 23% [26, 32, 33]. In pediatric neurogastroenterology patients, 23% had ARFID symptoms, and 8.5% had definite ARFID (meeting all DSM‐5 ARFID criteria) [33]. Whereas in adult neurogastroenterology patients, 23.7% had ARFID symptoms, but only 6.3% had definite ARFID [32]. Both adult and pediatric neurogastroenterology patients had “fear of aversive consequences” as the most frequent ARFID presentation (92%–93% in adults and 47%–67% in children), and the “lack of interest/low appetite” presentation was less frequent (21.6%–23% in adults and 33%–47% in children) [26, 32, 33]. However, none of the adult neurogastroenterology population was found to have the “sensory sensitivity” presentation, whereas it was found in 10%–17% of the pediatric neurogastroenterology population [26, 32, 33]. In the study conducted on adult neurogastroenterology patients, medical and psychosocial impairments were assessed, but their prevalence was not reported in the results [32]. However, in the pediatric neurogastroenterology population, researchers gave information on the impairment: 57% had weight loss or inability to gain weight, 17% had nutritional deficiencies and an extremely restricted diet, 13.3% were dependent on supplemental feeding or nutritional supplements, and 7% had psychosocial impairment [33].

3.1.2. Studies in Disorders of Gut‐Brain Interaction

In 2021 Murray et al. found 39.8% of adult DGBI patients having ARFID symptoms using the modified ARFID Canadian pediatric surveillance program questionnaire, after ruling out other eating disorders with the Eating Disorder Examination Questionnaire (EDE‐Q) [28]. Similar to the aforementioned studies, the most frequent ARFID presentation in DGBI patients was “fear of aversive consequences” (70%–90% of patients endorsed items related to this presentation) [28]. However, “sensory sensitivity” was also common in this population (43%), while “lack of interest/appetite” was the least frequent (22%–35%) [28]. As for the ARFID symptom criteria, 57% had psychosocial impairment, 3% were on enteral feeding and/or oral nutritional supplements, 60% had nutritional deficiency, and 30% had weight loss [28].

In a cross‐sectional study of IBS patients educated on a low‐FODMAP diet by a GI specialist dietitian between 2008 and 2018, 30% were identified as being at risk of ARFID [34]. However, the authors did not clarify whether they assessed the prevalence of ARFID prior to the dietary education to understand if the diet itself contributed to the ARFID risk [34]. Interestingly, strict FODMAP restriction was reported by only a minority of patients having ARFID symptoms (18%) [34]. It is unknown whether the authors evaluated medical and psychosocial impairment in these patients [34].

Adult patients with an overlap of functional bowel disorders and dyspepsia were found to have significantly more ARFID symptoms (defined as meeting the cut‐off for any NIAS subscale, excluding other eating disorders) compared to those with only functional dyspepsia (27.2% vs. 9.9%) [35]. This result was consistent when comparing each NIAS subscale in patients with and without the overlap, respectively: “appetite” 38.3% vs. 10.3%; “picky eating” (sensory sensitivity), 25.9% vs. 12.3%; and interestingly, the “fear” subscale was the least frequent, 11.1% vs. 3.3% [35]. However, in this study, the evaluation of medical and psychosocial impairment was not mentioned [35].

In pediatric patients with functional dyspepsia symptoms, 66.7% had ARFID (meeting cut‐off for at least one subscale with the presence of at least 1 medical/psychosocial impairment and the exclusion of other eating disorders) using the NIAS and 65.1% using PARDI‐AR‐Q [31]. After 2 months, among those who initially tested positive and completed follow‐up, 71.4% remained positive using the NIAS and 47.4% using PARDI‐AR‐Q [31]. Both tests indicated that patients scored highest on the “lack of interest” subscale at both baseline and follow‐up [31].

3.1.3. Studies in GI Motility Disorders

With the use of the NIAS, a study suggests that 78.4% of adult patients with achalasia meet diagnostic criteria for ARFID (70.3% of them having fear of GI symptoms, and 32.4% meeting criterion for lack of interest in food) [29]. However, it is important to note that this study did not mention ruling out other eating disorders nor evaluating medical/psychosocial impairment.

In adult patients with confirmed gastroparesis, 80.4% scored positively on at least one NIAS subscale (having potential ARFID, but other eating disorders were not excluded), and 76.6% had at least one medical or psychosocial impairment, (having clinically significant screen for ARFID) [30]. The most frequent ARFID presentations in these patients were “lack of appetite/low interest in eating” (84.2%) and “fear of aversive consequences” (75.6%), and the “sensory sensitivity” presentation was the least frequent (45.1%) [30]. In patients with definite ARFID, psychosocial impairment was reported in: academics/work (72%), social life (85%), relationships (66%), and distress/anxiety/embarrassment (89%); and medical impairment was reported as follows: weight loss (80%), nutritional deficiencies (60%); need for supplementation (63%) [30].

However, in adult patients with gastroparesis symptoms, 40% met the NIAS cut‐off for ARFID symptoms (excluding other eating disorders), but only 23.3% had “definite ARFID” (having medical or psychosocial impairment) [27]. Participants had more frequently endorsed questionnaire items in relation to the “lack of appetite/low interest in eating” (35%–57.7%) and “fear of aversive consequences” (42.4%–57.6%) presentations than those of the “sensory sensitivity” presentation [27]. As for ARFID criteria in patients with definite ARFID: psychosocial impairment: 27% reported impairment in academics/work; 34% in social life; and 27% in family; medical impairment: 24% reported weight loss; 18% a current low weight (BMI < 18.5 kg/m²); 31% nutritional deficiencies; 34% reliance on vitamins/supplements; and 13% tube feeding [27]. Interestingly, only gastroparesis symptoms and not gastric emptying were linked to ARFID [27].

In pediatric patients with gastroparesis symptoms, 48.5% had ARFID (meeting cut‐off for at least one subscale with the presence of at least 1 medical/psychosocial impairment and the exclusion of other eating disorders) using the NIAS and 63.6% using PARDI‐AR‐Q [31]. After 2 months, among those who initially screened positive and completed follow‐up, all patients remained positive on NIAS and 66.7% on PARDI‐AR‐Q [31]. Both tests indicated that patients scored highest on the “lack of interest” subscale at both baseline and follow‐up [31]. This study also found that there was no difference in the proportion of patients screening positive for ARFID whether they had normal gastric emptying or not [31].

3.2. Prevalence of Neurogastroenterology Disorders in ARFID

For this part of our scoping review, we identified 10 articles that evaluate the prevalence of neurogastroenterology disorders or related GI symptoms in patients with ARFID (Table 3). Of these 10 articles, 3 were also included in Section 3.1, while their study was conducted on GI patient samples, it also evaluated the prevalence of GI disorders in patients with ARFID [28, 32, 33]. The prevalence of neurogastroenterology disorders and related GI symptoms in patients with ARFID symptoms was found to range from 3% to 60%, depending on the diagnostic criteria used.

TABLE 3.

Summary of studies exploring the prevalence of neurogastroenterology disorder symptoms in patients with ARFID.

Study	Population and sample size	Study design	GI diagnosis	Elimination of organic GI disorders	GI disorder prevalence in ARFID patients
Studies from ED patient samples
a Almeida et al. Neurogastroenterol Motil. 2024 [36]	344 ED outpatients and inpatients (ARFID n = 26)	Retrospective chart review	Chart review	Yes (GI diagnoses categorized by coders)	84.6% were diagnosed with a functional/motility disorder
Cooper et al. Int J Eat Disord. 2021 [37]	129 ED inpatients (ARFID n = 22)	Retrospective chart review	Chart review	Yes (GI diagnostic tests like endoscopy, colonoscopy, etc.)	59.1% had GERD; 18.2% Gp; 22.7% IBS
Burton Murray et al. Int J Eat Disord. 2021 [38]	168 ED outpatients (including ARFID, n not specified)	Prospective	Modified Rome IV Questionnaire for Functional GI Disorders	Yes (evaluation by a medical professional)	30% had DGBI
Nicholas et al. Int J Eat Disord. 2021 [39]	2610 adults self‐identifying as “picky eaters”	Prospective	Self‐report	No	11.5% self‐reported GI disorders including functional GI disorders
Nitsch et al. Int J Eat Disord. 2023 [40]	122 adult ARFID inpatients	Retrospective chart review	Chart review	Yes (GI diagnoses categorized by coders)	41% had DGBI; 26% GERD; 20% Gp
Zickgraf et al. Int J Eat Disord. 2019 [41]	83 pediatric ARFID outpatients and inpatients	Retrospective chart review	Chart review	No	Reported medical comorbidities (gastroesophageal reflux, early satiety, nausea, Gp, abdominal pain)—prevalence not mentioned
Macdonald et al. J Eat Disord. 2024 [42]	42 adult ARFID patients (inpatient n = 27, intensive outpatient n = 3; outpatient individual therapy n = 12)	Retrospective chart review	Chart review	—	7.1% had GERD
Studies from GI patient samples
Burton Murray et al. Neurogastroenterol Motil. 2022 [28]	93 adult DGBI outpatients (ARFID n = 37)	Prospective	Self‐report of a provider diagnosis + extraction of provider‐assigned DGBI diagnoses from medical records	Yes (exclusion of patients having history of co‐existing GI pathology)	22% had FD; 51% IBS
Burton Murray et al. Clin Gastroenterol Hepatol. 2020 [32]	410 adult neurogastroenterology outpatients and inpatients (97 with ARFID symptoms)	Retrospective chart review	Chart review	Yes (GI diagnoses categorized by coders)	22% had FD; 12.4% Gp; 13.4% GERD; 28.8% IBS; 34% chronic constipation; 9.3% AB/distension; 4.1% functional diarrhea
Burton Murray et al. J Pediatr Gastroenterol Nutr. 2022 [33]	129 pediatric neurogastroenterology outpatients and inpatients (30 with ARFID symptoms)	Retrospective chart review	Chart review	Yes (GI diagnoses categorized by coders)	60% had chronic constipation; 23% FAP; 10% GERD; 7% IBS; 3% Gp

Abbreviations: AB, abdominal bloating; ARFID, avoidant/restrictive food intake disorder; DGBI, disorders of gut–brain interaction; ED, eating disorder; FAP, functional abdominal pain; FD, functional dyspepsia; GERD, gastroesophageal reflux disease; GI, gastrointestinal; Gp, gastroparesis; IBS, irritable bowel syndrome.

^a Almeida et al. Neurogastroenterol Motil. 2024 [36]: this study included patients who consulted for both ED and GI disorders; it was included in this section as it aimed to evaluate GI disorders among patients with ED.

3.2.1. Studies in Mixed‐Age Populations

The prevalence of DGBI in adult ARFID patients ranges from 30% to 41% [38, 40]. Sixty percent of ARFID patients with DGBI had the “lack of interest/low appetite” presentation, 55% “sensory sensitivity,” and 50% “fear of aversive consequences” [38]. In addition, 60% of these patients were found to have weight loss or failure to gain weight, 60% had psychosocial interference, 15% had nutritional deficiency, and 10% were dependent on enteral feeding/oral supplements [38]. Almeida et al. found that 84.6% of patients with ARFID were diagnosed with a functional/motility GI disorder [36]. But in a study done through an online survey on adults who self‐identify as “picky eaters,” 11.5% of participants who were “likely to have ARFID” also had self‐reported GI disorders including functional GI disorders [39].

In patients with ARFID symptoms, the prevalence of IBS ranges from 6.86% to 51%, gastroesophageal reflux disease from 7.1% to 59.1%, gastroparesis from 12.4% to 20%, and functional dyspepsia from 22% to 40% [28, 32, 37, 40, 42]. In a study previously cited in 3.1.1., 34% of adult neurogastroenterology patients with ARFID symptoms had chronic constipation, 9.3% had abdominal bloating/distension, and 4.1% had functional diarrhea [32]. However, it is important to note that some of these studies did not use validated tools on patients for measuring GI symptomatology, such as the Rome diagnostic questionnaire, nor did they have a definite ARFID diagnosis.

3.2.2. Studies in Pediatric Populations

To our knowledge, there was no study evaluating the prevalence of DGBI in the pediatric ARFID population. However, in a study previously cited in Section 3.1.1., 60% of pediatric neurogastroenterology patients with ARFID symptoms had chronic constipation, 23% had functional abdominal pain, 10% had GERD, 7% had IBS, and 3% had gastroparesis [33].

In a retrospective chart review of pediatric patients with ARFID in an eating disorder partial hospitalization program, they found that medical comorbidities like gastroesophageal reflux, gastroparesis, and abdominal pain were most commonly reported [41]. These symptoms were reported in 63.6% of patients with a low appetite presentation, 47.4% of patients with appetite disturbance and food selectivity, 20% of patients with food selectivity, and 18.75% of patients with fear of aversive consequences [41].

3.3. Direction of the Association Between ARFID and GI Symptom Occurrence

Regarding the timeline of eating difficulties in relation to GI symptoms, one study retrospectively reviewed medical charts of patients with an eating disorder diagnosis and found that 74.2% of them had a pre‐existing eating disorder (including ARFID) before their GI consult, while 23.8% developed an eating disorder afterward [36]. Unfortunately, this study did not specify how many of the ARFID patients received their diagnosis before or after their GI consult. Conversely, in another study where gastroparesis patients were asked about their perception of the timeline between gastroparesis and ARFID symptoms, 17.1% reported experiencing eating difficulties before their gastroparesis diagnosis, and 37.8% reported developing them afterward [30]. No difference in these responses was found between patients with and without ARFID [30].

3.4. Characteristics of Patients With an Overlap of ARFID and Neurogastroenterology Disorders

Both adult and pediatric neurogastroenterology patients with ARFID symptoms were more likely to be female than those without ARFID [32, 33]. Adult neurogastroenterology patients with ARFID symptoms were younger than those without ARFID, whereas pediatric neurogastroenterology patients with ARFID symptoms were found to be older than those without ARFID [32, 33]. Each included study evaluating body mass index (BMI) had found that neurogastroenterology (including DGBI and gastroparesis) patients with ARFID symptoms had lower BMI than those without ARFID [28, 30, 32, 33].

Moreover, neurogastroenterology patients with ARFID symptoms were found to have higher anxiety and depression levels (as assessed by self‐report severity scales) and poorer quality of life compared to those without ARFID [28, 29, 34].

3.5. Treatment for Patients With an Overlap of ARFID and Neurogastroenterology Disorders

We identified only two studies that addressed the topic of ARFID treatment in neurogastroenterology disorders (Table 4) [42, 43]. The first study was conducted on ARFID patients, some of whom had GI comorbidities (7.1% had gastroesophageal reflux disease), though it was not exactly conducted on patients with overlapping ARFID and neurogastroenterology disorders [42]. The treatment involved admission to a program offering three levels of care based on clinical indication: inpatient treatment, intensive outpatient treatment, and outpatient individual therapy [42]. Admission lengths varied by malnutrition severity (based on BMI): 6 weeks for patients who were severely underweight (BMI < 16) and 3 weeks for others. Treatment included 5–6 supervised meals/snacks per day, several weekly individual cognitive behavioral therapy (CBT) sessions, group psychotherapy, occupational therapy, social work, and medical and psychiatric care [42]. Overall, the study found positive outcomes, with weight restoration and improvements in impairments caused by the eating disorder across all three levels of care [42].

TABLE 4.

Summary of studies exploring treatment outcomes in patients with an ARFID‐neurogastroenterology disorder overlap.

Study	Population and sample size	Study design	Treatment	Outcomes
Macdonald et al. J Eat Disord. 2024 [42]	42 adult ARFID patients (7.1% with GERD; inpatient n = 27, intensive outpatient n = 3; outpatient individual therapy n = 12)	Retrospective chart review	Three levels of care involving individual CBT	Significant improvement and weight restoration in the 3 levels of care
Burton Murray et al. Int J Eat Disord. 2023 [43]	14 adult DGBI + ARFID outpatients	Prospective	Eight sessions of CBT	Medium‐high improvements in clinical outcomes

Abbreviations: ARFID, avoidant/restrictive food intake disorder; CBT, cognitive behavioral therapy; DGBI, disorders of gut–brain interaction; GERD, gastroesophageal reflux disease.

In the second study, researchers developed and tested a CBT protocol for ARFID within the context of DGBI [43]. The treatment consisted of eight CBT sessions, including education about DGBI, regularizing eating patterns, behavioral exposures, and post‐treatment maintenance planning [43]. The study reported promising outcomes for the eight‐session exposure‐based treatment, with high patient satisfaction and medium to high improvements in clinical outcomes [43]. However, it is worth noting that the sample size was small (n = 14), and the researchers plan to test this treatment more rigorously through a randomized controlled trial [43].

4. Discussion

Our scoping review revealed that ARFID symptoms are substantially prevalent in neurogastroenterology disorders, with rates ranging from 10% to 80%. Similarly, neurogastroenterology disorders or related GI symptoms are common among patients with ARFID symptoms, with prevalence rates ranging from 7% to 60%. Factors such as higher levels of anxiety and depression, lower BMI, and reduced quality of life were observed in patients with an ARFID‐neurogastroenterology disorder overlap. However, our review also highlighted a significant gap of studies exploring the pathophysiology underlying this overlap. In addition, few studies explored the directionality of the association between ARFID and neurogastroenterology disorders, or proposed and tested treatments for patients with this overlap.

This wide range and variation in prevalence between studies highlights the heterogeneity among them, particularly in terms of methods and assessment tools used. For instance, not all studies used the same validated questionnaires, cut‐offs, or applied the complete DSM‐5 criteria for ARFID diagnosis. ARFID prevalence may be overestimated in cases where other eating disorders were not ruled out, or when medical/psychosocial impairments were not considered in the diagnosis. In such cases, authors were using terms like “ARFID symptoms” or “potential ARFID” instead of “definite ARFID” [27, 30, 32, 33]. Additionally, the majority of the included studies used the NIAS for ARFID assessment. Although the NIAS is a validated questionnaire, it has only been validated in a sample of patients with ARFID, not in those with GI symptoms [44]. When using NIAS and PARDI‐AR‐Q for ARFID assessment, Kaul et al. found a small proportion of healthy controls screening positive [31]. Consequently, the accuracy of the results obtained using these questionnaires may be limited, and ARFID prevalence might be overestimated. Therefore, improving ARFID assessment methods in GI patients warrants further exploration, and input from mental health practitioners in the diagnostic process should be considered. This variation in the methods was also seen in the studies evaluating the prevalence of neurogastroenterology disorders in patients with ARFID, where some relied on non‐validated questionnaires or patient self‐reports rather than established diagnostic criteria, such as the Rome criteria [37, 39, 40]. Hence, the use of standardized, validated tools and the integration of all diagnostic criteria are crucial for ensuring accurate results, avoiding overestimation of prevalence, and facilitating the comparison and combination of findings across studies.

The prevalence of ARFID symptoms among DGBI patients (39.8%), and the prevalence of DGBI symptoms among patients with ARFID symptoms (30%–41%) were notably similar [28, 38, 40]. However, patients with GI motility disorder symptoms, such as achalasia and gastroparesis, showed a higher prevalence of ARFID symptoms (75% and 80% respectively) [29, 30]. This could be explained by the symptom overlap in ARFID and GI motility disorders: for instance, patients with achalasia and gastroparesis often experience weight loss and require parenteral or enteral feeding and nutritional supplements—factors that align with the ARFID DSM‐5 criteria (medical impairment) [7, 8, 45, 46]. Therefore, paradoxically, patients with not only GI motility disorder symptoms might be more prone to having ARFID symptoms, but also ARFID symptom prevalence in these patients might be overestimated due to the important symptom overlap in these two conditions.

Studies conducted on pediatric and adult populations with the same disorders were too few to allow for meaningful comparisons, and their methods were heterogeneous. However, in neurogastroenterology populations, the prevalence of “definite ARFID” was similar in pediatric (8.5%) and adult (6.3%) patients [32, 33]. Pediatric patients with gastroparesis showed higher ARFID prevalence (48.5%–63.6%) compared to adult patients with gastroparesis symptoms (23.3%) [27, 31]. This discrepancy could be explained by the variability between the studies, as a meta‐analysis published recently on the impact of study quality on ARFID prevalence rates found that ARFID prevalence ranged from 5% to 11% depending on the quality of studies, with a higher prevalence in children than in adults when using a random‐effects model [47]. However, with their quality‐effects analysis (adjusting for the between‐study variability), ARFID prevalence rates were similar for adults and children [47]. This suggests that study quality influences ARFID prevalence estimates and that ARFID prevalence might be overestimated in studies among children [47]. Despite this, a common finding in both pediatric and adult patients with gastroparesis symptoms was the absence of an association between gastric emptying and ARFID [27, 31]. Instead, ARFID was found to be linked to gastroparesis symptom severity [27].

Regarding ARFID presentations, both “fear of aversive consequences” and “lack of interest in eating/low appetite” were frequent among patients with GI motility disorders, whereas “fear of aversive consequences” was the most frequent, and “lack of interest in eating/low appetite” was the least frequent among the neurogastroenterology population [26, 27, 28, 29, 30, 32, 33]. These findings are logical, as patients with GI motility disorders often experience severe symptoms, including frequent nausea and vomiting, and the “fear” subtype is characterized by food avoidance/restriction due to the fear of triggering such symptoms [45, 46]. Additionally, GI motility disorder patients, especially those with gastroparesis, often experience early satiety, which explains the relatively high prevalence of the “lack of interest in eating/low appetite” presentation in these patients [27, 29, 30, 46]. As for the “sensory sensitivity” or “picky eating” presentation, its prevalence varies across studies, making it difficult to draw conclusions. However, in studies involving neurogastroenterology populations, none of the adult patients exhibited this presentation, while its prevalence in pediatric patients ranged from 10% to 17% [26, 32, 33]. Pediatric patients might be more prone to this presentation than adults, as a small‐scale study found that ARFID patients with a selective/neophobic presentation were predominantly school‐aged [48]. Factors such as neurodevelopmental disorders may influence the prevalence of this presentation in pediatric populations. For instance, in a study conducted on children and adolescents with full and subthreshold ARFID, the severity in the “sensory sensitivity” presentation was found to be associated with a higher likelihood of comorbid neurodevelopmental disorders, such as autism spectrum disorder (ASD) and attention‐deficit/hyperactive disorder (ADHD) [49]. Similarly, in another study, children and adolescents with ARFID and a comorbid ASD exhibited a heightened food‐related sensory sensitivity compared to those without ASD [50]. Nevertheless, neurodevelopmental disorders were not considered in the overlap between ARFID and neurogastroenterology disorders in the studies included in this review. Therefore, we lack information about their association with this overlap.

Only two studies were identified that evaluated the direction of the association between eating difficulties and GI symptoms, and their results were conflicting [30, 36]. However, these studies were conducted on different populations and used varying methods, including self‐reported data. Therefore, further longitudinal cohort studies are necessary to obtain more accurate findings on this association. Notably, such bi‐directionality had also been previously discussed in relation to other eating disorders, as eating disorders can lead to GI symptoms, and GI disorders can be a risk factor for eating disorder development [51].

In addition, this scoping review found that neurogastroenterology patients with ARFID symptoms were more likely to be female [32, 33]. While ARFID prevalence is typically similar between males and females, the higher prevalence of neurogastroenterology disorders in females may explain the high prevalence of ARFID symptoms observed in females with these disorders [7, 52]. Neurogastroenterology patients with ARFID symptoms were also found to have lower BMI than those without ARFID [28, 30, 33]. However, it is important to note that it does not necessarily imply these patients are underweight, as the mean BMI was within the normal or even overweight/obese ranges [28, 30, 33]. Neurogastroenterology patients with ARFID symptoms were also found to have a lower quality of life and higher levels of anxiety and depression compared to those without ARFID symptoms [18, 28, 29, 34]. Psychiatric comorbidities such as anxiety and depressive disorders, as well as poor quality of life, are common in neurogastroenterology disorders [53, 54]. Similarly, anxiety disorders and poor mental health‐related quality of life are prevalent in ARFID patients [55, 56]. Based on these findings, we hypothesize that the coexistence of ARFID symptoms and neurogastroenterology disorders may result in an additive effect, further exacerbating the psychological and quality‐of‐life challenges faced by these patients. Additionally, IBS patients with eating disorders were found to have a poorer quality of life compared to those without eating disorders, as well as a higher prevalence of anxiety and depression compared to healthy volunteers [57]. Among young patients in day treatment for eating disorders, researchers observed a higher comorbidity of anxiety disorders in ARFID patients compared to those with other eating disorders [58]. Interestingly, though, one study found that adult DGBI patients with ARFID symptoms had lower depression, anxiety, and pain interference scores compared to those with eating disorder symptoms motivated by shape/weight concerns [28]. This suggests that while DGBI patients with ARFID may have psychological comorbidities, these might be less severe than those seen in DGBI patients with other eating disorders. Future comparative studies are needed to confirm this hypothesis.

Finally, our review identified only two studies on the treatment of patients with overlapping ARFID and neurogastroenterology disorders [42, 43]. Although based on small sample sizes, these studies reported promising effects of CBT [42, 43]. In ARFID patients, treatment varies depending on individual needs and may also include psychological treatment such as CBT, which is increasingly being tested in this population and has shown promising results [59, 60, 61, 62]. Similarly, CBT has been shown to be effective in patients with neurogastroenterology disorders, improving both GI symptoms and psychological outcomes [63, 64]. Multiple randomized controlled trials are needed to confirm these findings and establish effective treatments for this overlap. To date, ClinicalTrials.gov—an online database providing detailed information about clinical studies conducted worldwide—lists 28 reports of clinical studies on ARFID treatment, with one study testing exposure‐based CBT for ARFID in functional dyspepsia [65]. Treatments can be particularly complex when neurogastroenterology disorders require dietary management that conflicts with ARFID interventions, such as exposure to certain foods that may be restricted. Consequently, multidisciplinary approaches should also be tested to address all factors comprehensively and provide holistic care.

5. Conclusion

This is the first scoping review to examine the association between ARFID and neurogastroenterology disorders. While we identified interesting findings regarding the prevalence of overlaps between these conditions, our review also highlights a lack of studies on the mechanisms underlying this overlap, as well as on its treatment. Additionally, the heterogeneity in methods used across studies makes it challenging to obtain accurate and generalizable data. The suitability of current assessment tools for this population also requires further evaluation.

Author Contributions

C.M. initiated the study. E.L. designed the search strategy. H.M.‐M. and V.B. screened titles and abstracts, and selected relevant abstracts. C.M. resolved disagreements in abstract selection. H.M.‐M. and C.M. drafted the manuscript. Each author has approved the final draft submitted.

Conflicts of Interest

C.M. has served as a consultant/advisory board member for Kyowa Kirin, Norgine, Biocodex, Mayoly Spindler, Tillots, Ipsen, and Nestlé HealthScience.

Supporting information

Appendix S1

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.