Dietetic‐Led Cognitive Behavioral Therapy for ARFID: A Novel Approach to Avoidant/Restrictive Food Intake Disorder With Case Studies

ABSTRACT

Background

Cognitive‐behavioral therapy for avoidant/restrictive food intake disorder (CBT‐AR) is an emerging treatment for adults with avoidant/restrictive food intake disorder (ARFID). This article explores how CBT‐AR can be integrated with and enhanced by the dietetics skillset and presents two case studies in which a dietitian primarily provided treatment within a multidisciplinary team.

Methods

The dietetic‐led CBT‐AR treatment journeys are described for two patients. We utilised dietetic and psychological assessments to measure change, including: nutrient pathology; weight; Pica, ARFID and Rumination Disorder Interview scores; Australian Core Food Score; and Structured Assessment of Gastrointestinal Symptoms. We describe the stages of CBT‐AR with examples of how a dietitian's skillset might be utilised within this treatment approach.

Results

Both patients demonstrated significant improvements in ARFID psychopathology and dietetic measures after completing dietitian‐led CBT‐AR. These findings provide preliminary evidence that dietitians with specialised expertise in eating disorders could effectively deliver CBT‐AR as part of a multidisciplinary team.

Conclusions

These case studies highlight the potential for dietitians to act as primary clinicians in delivering CBT‐AR and provide preliminary evidence on how their skillset could lead to clinically meaningful improvements for adults living with ARFID. Future research trials are required to confirm that dietitian‐led CBT‐AR could be a viable treatment option for ARFID.

Summary

• Dietetic‐led Cognitive Behavioral Therapy for avoidant/restrictive food intake disorder (CBT‐AR) is a manualised treatment that can be delivered by dietitians in the role of primary clinician within a multidisciplinary team.

• The additional materials created for dietetic‐led CBT‐AR can be used by all clinicians treating ARFID.

• Two cases outline the improvement of nutrition and psychosocial outcomes, and support further research for dietetic‐led CBT‐AR.

1. Introduction

The introduction of avoidant/restrictive food intake disorder (ARFID) into the fifth edition of the Diagnostic Statistical Manual [1] marked a significant shift in the discourse on eating disorders. In contrast to other eating disorders that centre around weight and shape concerns [2], diagnostic criteria for ARFID include nutrient deficiencies, reliance on enteral nutrition, weight loss or significant psychosocial impairment. While individuals with ARFID historically present as extremely ‘fussy’ or ‘picky’ eaters, three phenotypes of ARFID have been identified: (i) sensory sensitivities (appearance, texture, taste and smell); (ii) lack of interest (e.g., food is a chore, poor hunger cues); and/or (iii) fear of aversive consequences (e.g., fear of gastrointestinal pain or choking).

To treat ARFID, Thomas & Eddy [3] have manualised cognitive‐behavioral therapy for ARFID (CBT‐AR) [4]. Pilot research has demonstrated reductions in ARFID psychopathology, an increased number of foods incorporated into the diet, and high patient satisfaction. Willmott et al. [5] identified CBT‐AR as the prominent evidence base for the treatment of adults with ARFID. However, thus far, research has only examined psychologist‐led CBT‐AR [6, 7], and there is a need to explore alternative options for treatment delivery.

The differing presentations of ARFID compared with anorexia nervosa or bulimia nervosa have created an increased demand for health professionals with specialised skills and knowledge in ARFID treatment [8] and the ability to address complex presentations [9]. Where mental health professionals face increasing clinical demand to treat eating disorders, leading to burnout and longer waitlists [10], an option is to utilise and upskill the multidisciplinary workforce. Dietitians are essential members of the clinical team [11] due to the disorder's focus on nutritional sequalae, which include key diagnostic criteria [1] and components of CBT‐AR [3].

Dietitians possess the knowledge and skills to treat ARFID and extend the scope of practice offered by mental health professionals [11]. Dietitians are uniquely placed to manage malnutrition, nutrient deficiencies or excesses, and oral nutrition supplements (ONS) in individuals with ARFID [12]. Arguably, in some health care settings, dietitians may be the only professionals who, in addition to strong counselling aptitude and communication, are specifically trained in medical nutrition therapy, eating behaviours and behaviour change.

Due to the salience of nutritional concerns in the clinical presentation of ARFID, we hypothesised that dietitians could lead CBT‐AR as the primary clinician. This article is the first to outline the dietetic enhancements of CBT‐AR and the protocol of dietetic‐led CBT‐AR through two cases of adults living with ARFID from a larger ongoing trial.

2. Methods

CBT‐AR is a flexible, manualised treatment for individuals are aged 10 years and older who eat orally. The manual outlines the complete treatment protocol [11], with worksheets collated into a workbook for patient use. Treatment occurs in four stages: Stage 1 focuses on ARFID education and early change, Stage 2 treatment planning, Stage 3 addressing the phenotypic presentations and Stage 4 relapse prevention. The alignment of CBT‐AR to dietetic medical nutrition therapy is outlined in Figure 1. We delivered sessions weekly for 50 min as a mixture of in‐person and virtual.

Figure 1.

The stages of Dietetic‐led CBT‐AR, outlining key areas of medical nutritional therapy addressed in each stage.

Author C.W. (the dietitian) provided dietetic‐led CBT‐AR in the Queensland University of Technology multidisciplinary community eating disorders clinic. C.W. is an Accredited Practising Dietitian (APD) and Credentialed Eating Disorder Clinician, (CEDC) with 1 day of formal training in CBT‐AR, receiving weekly supervision from either J.T., psychologist who manualised CBT‐AR, E.S., a psychologist and CEDC, or L.R., an Advanced APD. Each patient's General Practitioner (GP) oversaw medical safety. Postgraduate psychology interns, supervised by E.S., assessed psychological diagnosis and safety. Queensland University of Technology Eating Disorders Clinic held weekly multidisciplinary team meetings to review patient care.

2.1. Study Procedure

Eligibility criteria included individuals aged over 18 years, residents of Queensland, Australia, eating orally, with no other diagnoses significantly impacting intake (e.g., active Crohn's) or counter‐indicated concerns. Patients needed to have clinically significant scores on at least one scale of the Nine Item ARFID Screen [13] (i.e., ≥ 10 for picky eating, ≥ appetite and/or ≥ 10 for fear [14]. A postgraduate psychology intern confirmed the diagnosis using the Structured Clinical Interview for DSM‐5 Research Version(SCID) [15]. The GP's ascertained medical stability according to the Queensland Eating Disorders Service Guidelines [16] and saw patients at least monthly. The psychology interns were available throughout the study to support and manage psychological risk.

Patients completed the Pica, ARFID and Rumination Disorder Interview (PARDI) [17], and pre/post‐questionnaires the week before commencing CBT‐AR and the week after completing. The questionnaires included: the Structured Assessment of Gastrointestinal Symptoms (SAGIS) [18], a validated self‐report measure to measure gastrointestinal concerns; and the Australian Core Food Score (ACFS), an unvalidated self‐report measure that assesses regularity, adequacy, social eating, and nutrition knowledge by number of foods consumed in a month (Supporting Information S9). The ACFS was used as no validated tools are available. C.W. created this tool to measure change in dietary interventions that were appropriate for use by people with eating disorders. The dietitian and J.T. reviewed the PARDIs to confirm phenotypic presentation. We calculated the reliability of the change index [19], a statistical measure of meaningful change between pre‐ and post‐scores, of the phenotypes present with the PARDI for each case to ascertain reliable change. A score outside the range of −1.96 to 1.96 indicates reliable change, based on the 95% confidence interval norms for z‐score standard deviation.

2.2. Case Study 1

Andrew (pseudonym), a 23‐year‐old Caucasian male, presented with a fear of choking with no co‐diagnoses. At 16, he choked on a vegetable chip, reducing his intake to two foods (condensed milk and plain yoghurt) and one type of ONS (vanilla Resource Plus). Andrew was unable to take any vitamin supplements due to his fear of choking. He avoided eating in social settings and feared most liquids, believing particulates could cause choking. He engaged in numerous accommodations, such as extensive checking and cleaning his lips before meals, fearing germs in leftovers, using only his own utensils, washing all lids, and visually inspecting food before eating.

He attended a specialist eating disorders GP weekly and had previously worked with a psychologist who identified ARFID and a dietitian who initially prescribed ONS. His GP prescribed 20 mg of escitalopram for his anxiety. He self‐referred to the study's eating disorders clinic after experiencing medical complications of dehydration, electrolyte abnormalities and cardiac concerns, at his lowest weight of 91 kg and 186 cm tall (BMI: 26.3 kg/m²). To avoid hospital admission, he consistently consumed large portions of his preferred foods, along with six to eight ONS daily. He was motivated to change due to a 19 kg weight loss over 2 years, the fear of requiring hospitalisation, and his ARFID limiting full‐time work opportunities.

Andrew began dietetic‐led CBT‐AR at 97 kg and was diagnosed with malnutrition by the dietitian. He presented with weight loss, inadequate oral intake, and micronutrient abnormalities. Given his malnourished state, the dietitian selected the underweight version of CBT‐AR. His weight goal was to return to his premorbid weight of 110 kg, where he reported no signs of malnutrition. Biochemistry results indicated elevated levels of iron and B6 alongside low manganese levels. His reliance on ONS contributed to these imbalances, containing six times the recommended daily intake of iron and five times the recommended intake of B6. The PARDI confirmed the phenotype as fear of aversive consequences. Andrew attended 30 weekly sessions between November 2023 and August 2024. His treatment goals were to improve his medical condition, gain weight, eat solid foods, reduce preoccupation with food and re‐engage in social eating.

2.3. Case Study 2

Lily (pseudonym), a 19‐year‐old multiracial female, presented with limited nutritional variety. Lily had co‐diagnoses of autism, migraines and Disorders of the Gut Brain Interaction (DGBIs) including globus sensation and postprandial distress syndrome. She regularly consulted with a specialist eating disorders GP, had received previous anxiety treatment from a postgraduate psychology intern and had seen an eating disorders dietitian. She and her mother self‐referred to the study's eating disorders clinic on her GP's recommendation.

The assessment determined Lily was undernourished but weight stable at 62 kg and 167 cm (BMI: 22.2 kg/m²). Biochemistry results identified low vitamin D, high vitamin A, copper and iron deficiency. She reported symptoms correlating to the irregular results including fatigue, muscle weakness, dizziness and nausea. The PARDI confirmed all three ARFID phenotypes. Lily's eating pattern involved grazing through the day and dinner. She preferred crisp foods and disliked mushy foods and the mouth feel of fats. Meals could take hours to consume, often eaten cold. She feared gastrointestinal pain. Lily completed family‐based CBT‐AR with her mother. They attended 21 sessions between November 2023 and June 2024. Her goals were to increase food variety and reduce fear of gastrointestinal sensations.

2.4. Dietetic Assessment

In addition to CBT‐AR, the dietitian conducted a dietetic assessment and measured each patient's weight and height at their evaluation session. The dietitian explored medical history, including weight, family and gastrointestinal concerns, using the SGAIS. The patient's dietary intake was assessed using a 24‐h recall in conjunction with exploring dietary patterns, the ACFS, and food insecurity history. Additional evaluation covered childhood intake, culturally significant foods, food allergies and intolerances, plus nutrition beliefs. This questioning revealed that Andrew was food insecure. Both patients completed a malnutrition assessment to determine energy‐protein malnutrition using a visual Subjective Global Assessment, then underwent micronutrient blood biochemistry to identify undernutrition. As defined by the World Health Organisation [20], Andrew was moderately malnourished and Lily was undernourished, but not malnourished.

2.5. Neuro‐Affirming Practices

To support Lily's autism, the dietitian offered visual aids and written documentation and encouraged the use of sensory strategies, such as standing during sessions or using sensory tools. Extended silences occurred as needed for Lily's processing. Her body language and non‐verbal cues differed from typical expectations, such as her eye contact focused on her mother. The dietitian did not conceptualise these variations as challenges or barriers to communication or CBT‐AR, but as reflections of her neurodivergence. Her mother played a critical role, taking notes, offering encouragement and repeating questions to facilitate engagement. They jointly engaged in problem‐solving to help establish regular eating patterns and strategies for family support at home.

3. CBT‐AR Stage 1

Andrew completed individual CBT‐AR, declining family or other support people. Lily engaged in family‐supported CBT‐AR, requesting family support and living at home. The dietitian educated both patients on ARFID, their presenting phenotypes, and CBT‐AR treatment. Each patient collaborated with the dietitian to develop their ARFID formulation. In accordance with the CBT‐AR manual, dietary intake is recorded; Andrew used the Recovery Record app, while Lily tracked intake using paper records from the CBT‐AR workbook. Andrew participated in the in‐session meal. The dietitian observed, prompted Andrew to try a novel food, and reflected on the meal in relation to his goals, strategies to increase meal volume and identifying accommodations. Both were encouraged to make small changes to intake, with a focus on regularity and volume.

3.1. Addition of Medical Nutrition Therapy Education

Andrew and Lily incorporated the impact of nutrition into their formulation outlining concerns such as dysregulated appetite hormones, malnutrition, and gastrointestinal symptoms. The dietitian provided Andrew with education on the impacts of malnutrition, noting he met his caloric volume with ONS and preferred foods, but lacked food to support balanced nutrition. Both patients learned about the gastrointestinal system relationship to ARFID and its impact on quality of life. The dietitian shared a visual of the gastrointestinal system to aid Lily's understanding.

3.2. Management of ONS

C.W. managed Andrew's ONS prescription, an essential skill within a dietitian's scope of practice, aimed at meeting nutrition needs to resolve malnutrition. The goal was for Andrew to systematically transition from ONS to food for his nutritional needs.

3.3. RAVES [21]

The RAVES eating model—Regularity, Adequacy, Variety, Eating socially and Spontaneity—is a dietetic protocol describing aspects of eating that build sequentially, added to support the dietetic adaptation of CBT‐AR. Andrew and Lily built momentum and learnt agency in the treatment process alongside their engagement in the stages of the RAVES model. For example, Lily built regularity, resulting in more eating occasions that then supported variety in Stage 3. Andrew recognised that he needed to extend his variety before he was able to eat socially without significant levels of anxiety.

4. CBT‐AR Stage 2

A key aspect of stage 2 of CBT‐AR is treatment planning and nutritional education achieved through education and assessment of nutrient deficiencies, dietary habits and identifying foods patients are ‘willing to learn about’.

4.1. Nutrition‐Specific Inclusions

The dietitian provided detailed nutrition education by updating and creating new worksheets to share with the patients (Supporting Informations [Link], [Link], [Link]). A.A. designed a car visual as an analogy for fuelling the body (e.g., petrol fuels a car analogous to how carbohydrates fuel the body). A.A. also updated the nutrients list to include potential micronutrient deficits in ARFID and created a human body diagram illustrating the impact of specific micronutrients to support visual learners. The dietitian utilised these worksheets with Andrew and Lily to implement purposeful, individualised dietary changes. For Andrew, the focus was on his reliance on ONS and related micronutrient excesses. For Lily, the body visual, combined with biochemistry results, helped her link physical symptoms to the lack of variety in her diet.

4.2. Building Blocks

C.W. and G.H. revised the original United States‐based CBT‐AR building blocks list for the Australian context, incorporating an ‘all foods are equal’ approach (Supporting Information S1). This revision reflected dietetic knowledge of the Australian food environment and the REAL Food Pyramid [22]. The blank rows after each food group were retained for the inclusion of culturally specific foods and the authors added a column for how food is eaten or prepared, to support small novel changes.

Lily and Andrew completed the building blocks list, incorporating their micronutrient requirements and social eating goals. Lily, who was iron deficient, agreed she was willing to learn about red meat, and used the blank rows to include foods from her Indonesian heritage. She consistently ate a limited variety of all food groups. Andrew consistently ate dairy foods and was willing to learn about certain fruit and vegetable juices. Both patients planned to incorporate these foods into Stage 3 exposures.

4.3. Plate Models

The original CBT‐AR ‘MyPlate’ Model, based on United States dietary guidelines, visually divides a plate into food group portions to promote healthy eating [23]. C.W. and G.H. revised the plate diagram to align with international recommendations [23, 24, 25] and eating disorder dietetic resources such as the REAL Food Pyramid [22] and Plate by Plate approach [26]. Renamed the ‘Nourishing Plate’(Supporting Information S5), the schematic reconceptualised the activity's purpose to nourish the body while challenging misconceptions about healthy eating and weight stigma [27].

Following the CBT‐AR manual, the dietitian and patient reviewed their typical meals. The dietitian supported Andrew and Lily to assess meal volume and macronutrient ratios. Andrew, with his high dairy intake, reflected he needed to diversify protein and fats while adding carbohydrates, fruits and vegetables. For Lily, overall meal volume was low despite containing all food groups.

4.4. Food Neutrality

The dietitian incorporated an ‘all foods are equal’ approach to alleviate societal pressures regarding a ‘healthy’ diet and challenge diet culture [28]. This appeared to foster motivation for sustainable and reasonable dietary changes while addressing stigma Lily and Andrew felt regarding their eating habits. The dietitian employed neutral language describing Andrew and Lily's intake as ‘consistent’ rather than focusing on its nutritional content or sensory profile. The term ‘consistent’ denotes uniform food choices, such as Lily's preference for crackers, always the same in texture and packaging, and Andrew's consistent choice of yoghurt, always the same brand, packaging and flavour.

Lily, influenced by diet culture, felt guilty about not meeting her definition of healthy eating. After discussions on these topics, Lily and her mother reported less pressure to make ‘healthier’ choices and progressed at her own pace. Andrew, who discreetly consumed ONS at work, described greater motivation to include foods in his food hierarchy when food was reframed as equal, and this coincided with reported reduced societal pressures.

4.5. Hydration & Fluids

The dietitian educated patients about hydration, adequate water intake and caffeine's impact on appetite. Andrew learned about the caffeine in energy drinks. Lily saw from her monitoring records that she drank inadequate fluids and linked this to her migraines. The dietitian explained the possible connection [29]. Starting treatment, Lily had only coffee for breakfast, the dietitian provided education on its impact on hunger signals, and she was encouraged to add food upon waking.

5. CBT‐AR Stage 3

Stage 3 of CBT‐AR focuses on targeted exposures to address presenting phenotypes. Lily began with the lack of interest phenotype, then sensory sensitivities and finally fear of aversive consequences, following the CBT‐AR manual.

Lily and Andrew worked through the fear of aversive consequences module. Lily moved quickly through the hierarchy in six sessions, ultimately confronting her most feared food‐‐ deep‐fried chicken. Andrew had difficulty engaging in exposures that did not directly involve the consumption of feared foods. He declined activities of touching a banana or watching videos of others eating his feared foods. He also reported difficulties in identifying mid‐level fear foods, creating a gap between low and high fear foods. Further, Andrew reported much lower levels of anxiety than his body language of muscle tension, hand fidgeting and stooped posture suggested during exposures. He progressed rapidly through his fear hierarchy but experienced a setback when an unexpected lump in tinned pumpkin soup triggered a panic attack; he reflected this reduced his confidence and motivation. Momentum appeared to return by addressing his cognitions in the exposure hypotheses, for example, ‘Sarsaparilla (soft drink) is made from solids, various chunks, seeds, roots which could make me choke after one sip’. The dietitian offered psychological support during this period; however, he declined.

5.1. Sensory Fatigue

Some patients with the lack of interest phenotype struggle completing meals due to ‘sensory fatigue’—a term describing the sensory boredom, particularly with flavour or texture, that prevents them from eating an adequate portion. Lily identified with this as a reason for grazing, extensive mealtime and incomplete meals. To address this, the dietitian recommended separating flavours in meals or incorporating varied sensory components to maintain interest. For example, serving dinner and dessert together allowed Lily to switch between different sensory experiences.

5.2. Choice Paralysis

During the sensory sensitivity module, Lily described feeling overwhelmed when selecting foods for exposures, leading to avoidance. This difficulty, identified as ‘choice paralysis’, prompted the dietitian to create a worksheet to help patients explore alternative ways to choose food beyond the building blocks activity (Supporting Information S6). The worksheet included chained food options, where each food was linked, such as by flavours or meals. Lily and her mother used it to select foods, such as choosing weekly themes, (e.g. breakfast foods) to guide novel exposures.

5.3. Neutral Words Worksheet

In the CBT‐AR manual, the sensory sensitivity module, asks patients to describe foods neutrally using the five senses. Initially, Lily struggled to articulate her experience or remain neutral and described foods she felt were negative descriptors, such as mushy. The dietitian created a worksheet with neutral descriptors for each sense (Supporting Information S3). Lily used it extensively to build confidence in describing the novel foods.

5.4. Food Safety

ARFID can lead to distorted beliefs about food preparation and safety. A dietitian provides accurate food safety information and advice on safe cooking and food preparation. In Andrew's case, the dietitian addressed his concerns by educating him on safe food preparation practices and discouraged the use of unnecessary accommodations (e.g., excessive cleaning of utensils before use). He demonstrated increased confidence in cooking, storing, and handling food observed by him engaging in these activities more frequently, which increased home exposure frequency and meal regularity.

5.5. Hunger, Fullness and Extrinsic Cues

The lack of interest module addresses hunger and fullness. Some neurodivergent patients may not be able to engage in identifying these signals [30], in addition to the impacts of malnutrition and variations in appetite‐regulating hormones [31]. The dietitian used an individualised approach. For Lily, she created a personalised hunger and fullness scale incorporating emotional, physical and external prompts. Lily identified physical symptoms such as energy levels or stomach sensations, which the dietitian emphasised in the scale.

5.6. Food List

To track novel foods exposures across Stage 3 modules, the dietitian developed a worksheet (Supporting Information S8). It categorises food as ‘Learning about/Tasting’, ‘Incorporating’, and ‘Incorporated’, avoiding labels of ‘safe or unsafe’. Foods currently consumed are defined as ‘preferred’. The list encouraged Lily and Andrew to be consistent with exposures. For example, Lily first identified bananas as ‘not for me right now’ but later reintroduced them at her own pace.

6. CBT‐AR Stage 4 and Dietetic Integration

In Stage 4, relapse prevention, the dietitian integrated pre/post‐anthropometric data and micronutrient pathologies with repeated building blocks and nourish plate activities to evaluate progress. A key focus was fostering autonomy, equipping Lily and Andrew with the skills and knowledge necessary for long‐term management and continued progress. For his relapse plan, Andrew aimed to continue to expand his variety of foods and updated his fear hierarchy accordingly, with ONS as a contingency plan. Lily planned to continue exploring new foods and engage in new social eating situations. The relapse plans for Lily and Andrew reinforced progress, encouraged motivation, and outlined their commitment to ongoing and future dietary changes.

7. Results

7.1. Case 1

At the conclusion of CBT‐AR, Andrew experienced a decrease in self‐reported anxiety in food and eating, his malnutrition resolved and his quality of life improved. He achieved his treatment goals by dining with colleagues and friends, improving his medical status, expanded his diet to 21 new foods with no ONS, and less occupied with food. He independently planned and conducted exposures from his fear hierarchy, and his accommodations significantly decreased—he rarely engaged in lip checking, ceased abnormal utensil use, and could eat leftovers. Although he still met ARFID criteria, due to micronutrient deficiencies and psychosocial impairment, he no longer met the weight loss or enteral nutrition dependence criteria, indicating that overall his ARFID severity had decreased. The reliable change index indicated significant positive improvement across all PARDI subscales. His iron levels normalised, and B6, folate and manganese trended toward normal levels. His weight increased above the goal of 110 kg, and it was agreed that resuming physical activity and introducing nutrient‐dense foods would support long‐term weight stabilisation and health (Table 1).

Table 1.

Case 1: Measures of change in psychopathology, weight and nutrients.

Measure	Pre	Post	RC
Weight (kg)	97.0	126.4
Body mass index (BMI) (kg/m2)	28.0	36.5
PARDI—Severity	4.56	2.0
PARDI—Sensory sensitivity	0.9	0.5	16.94*
PARDI—Lack of interest	2.0	0.9	3.50*
PARDI—Fear of aversive consequences	3.3	2.1	13.11*
ACFS (number of foods)	1	15
SAGIS	2	2
B6 (µg/L)	340 (H)	210 (H)
Total iron (µmol/L)	321 (H)	17
Transferrin (µmol/L)	3.3 (H)	2.7
Total iron‐binding capacity (µmol/L)	82 (H)	69
Manganese (µg/L)	7 (L)	9 (L)
Serum folate (nmol/L)	4.3 (L)	5.4 (L)

Note: RC, reliable change [19] and *p < 0.05.

Abbreviations: ACFS, Australian Core Food Score; (H), excess; (L), deficit; PARDI, Pica ARFID Rumination Disorder Interview; SAGIS, Structured Assessment of GastroIntestinal Symptoms.

7.2. Case 2

At the end of treatment, Lily's ARFID symptoms had significantly reduced, and she no longer met the diagnostic criteria. She incorporated 29 new foods, prepared meals at home, ate socially, and consumed more regular, adequate meals. Most newly introduced foods came from the protein and dairy categories. Her sensory preferences remained, but she became more flexible, understanding she could meet her nutritional requirements with alternatives, reducing the pressure to enjoy all foods. Lily learned that her anticipated negative gastrointestinal sensations were less intense or did not occur. She acknowledged that eating remained challenging but felt equipped to manage it independently using dietetic‐led CBT‐AR skills. Lily maintained a stable weight, and her micronutrient levels were clinically stable, except for vitamin A and vitamin D (which was related to her limited sun exposure). The RC showed (Table 2) significant positive improvements across all ARFID phenotypes. She achieved her treatment goals of increased variety and improved gastrointestinal symptom awareness and management.

Table 2.

Case 2: Measures of Change in psychopathology, weight and nutrients.

Measure	Pre	Post	RC
Weight (kg)	60.2	60.4
BMI (kg/m2)	21.6	21.7
PARDI—Severity	4.0	0.2
PARDI—Sensory Sensitivity	4.7	0.7	90.40*
PARDI—Lack of Interest	5.0	2.0	8.99*
PARDI—Fear of Aversive Consequences	4.0	0.3	18.13*
ACFS (number of foods)	51	77
SAGIS	37	18
Vitamin D (nmol/L)	26 (L)	34 (L)
Vitamin A (retinol) (µmol/L)	2.6 (H)	2.6 (H)
Total Iron (µmol/L)	20	12
Ferritin (µg/L)	13 (L)	19
Total Iron Binding Capacity (µmol/L)	91 (H)	75
Copper (µmol/L)	31.4 (H)	27
Ceruloplasmin (mg/dL)	0.46 (H)	0.41

Note: RC, reliable change [19] and *p < 0.05.

Abbreviations: ACFS, Australian Core Food Score; (H), excess; (L), deficit; PARDI, Pica ARFID Rumination Disorder Interview; SAGIS, Structured Assessment of GastroIntestinal Symptoms.

8. Discussion

The two cases reported here illustrate the proof‐of‐concept for dietitian‐led CBT‐AR and provide guidance for the provision of dietetic‐adapted CBT‐AR. Following treatment, Lily no longer met the diagnostic criteria for ARFID, and Andrew met two of five criteria, while improving psychosocial functioning, demonstrating that dietitians possess skills that could be integrated into CBT‐AR to address nutritional complexities. According to the CBT‐AR manual and Fitzpatrick et al. [32], treatment may not fully eradicate all disordered behaviours but aims to minimise diagnostically significant issues, such as reliance on ONS and micronutrient deficiencies, while enhancing the quality of life. Both patients expressed ongoing concerns about eating, though to a lesser degree, as reflected in reductions in their PARDI scores.

These findings suggest that a dietitian could effectively provide CBT‐AR and support the complex, heterogeneous presentations of ARFID. Their medical nutrition expertise allows for tailored nutrition recommendations, as demonstrated with Andrew, whose BMI was classified as ‘overweight’, yet dietetic assessment revealed malnutrition. For Lily, integrating gastrointestinal education made treatment more client‐focused and aligned with her goals. While a dietitian's expertise has possibilities to enhance treatment, it is crucial that they are trained, experienced, and supervised in CBT‐AR to maintain treatment integrity, including session structure, frequency, agenda setting, and range of exposures. Additionally, clear delineation of the scope of practice and psychological risk is essential, with psychological support and diagnostic confirmation, as provided in the current study.

Treating individuals with psychiatric and medical comorbidities, such as DGBIs or autism, requires a holistic approach [33]. While CBT‐AR targets ARFID, it remains flexible in addressing the impact of co‐diagnoses. Treatment should acknowledge these conditions and consider their influence on ARFID based on individual experiences. For example, Lily had DGBIs, and Andrew experienced anxiety. The goal was not to treat these conditions directly but to recognise how ARFID affected them. As Andrew engaged in ARFID treatment, he reported decreased anxiety symptoms across his life. Initial concerns existed about his medical stability, but regular dietary monitoring by a dietitian helped prevent hospitalisation. For Lily, understanding the connection between her gastrointestinal symptoms and ARFID helped reduce symptoms through improved nutrition.

For neurodivergent individuals with ARFID, neuro‐affirming practices can be incorporated into the implementation of CBT‐AR [34], and dietetic‐led CBT‐AR as seen in Lily's case. A dietetic, neuro‐affirming lens encouraged Lily and her mother to reconsider expectations about ARFID recovery and ‘normal’ eating. CBT‐AR's flexibility aligned with Lily's learning style, enabling engagement and to rejection of neurotypical eating norms.

Cultural factors were considered by adapting CBT‐AR worksheets to the Australian context, as the manual was developed in the United States. Personal cultural beliefs and practices, such as Lily's Indonesian heritage, were also incorporated. Additionally, food insecurity was addressed for Andrew, who had only recently started full‐time work and was concerned about food costs.

Dietetic‐led CBT‐AR has the potential to improve access to care, particularly for adults with ARFID, an underserved population. The treatment could provide an opportunity to support individuals who have lived with undiagnosed ARFID for years and struggle to access treatment. Several strengths of the inclusion of a dietitian include addressing malnutrition and undernutrition by directly treating ARFID's diagnostic criteria. It could help alleviate the burden on mental health professionals, offering an alternative pathway for care. Given that the ARFID diagnosis was only introduced in 2013, many adults have experienced chronic, untreated ARFID and significant medical implications [35], making this approach particularly valuable. However, these results have limited generalisability due to a small sample size. Further implementation of dietitian‐led CBT‐AR may not be feasible for all dietitians without CBT‐AR training, due to limitations such as lack of confidence, insufficient ARFID knowledge, limited eating disorders experience, and challenges in maintaining treatment frequency. For some patients, a shared interdisciplinary care model involving both dietitians and psychologists may be more practical [36].

Expanding skillsets and utilising dietitians has come at an advantageous time to support building clinically informed, evidence‐based treatments for ARFID. Lily and Andrew received dietetic‐led CBT‐AR, which provided flexible and manualised treatment, reducing ARFID symptomology and improving physical and psychosocial health. Further research is necessary to strengthen the evidence base and explore in adolescents. A pilot trial is currently being conducted to explore dietetic‐led CBT‐AR in a larger sample assessing psychological, physical and nutritional outcomes, with long‐term follow‐up.

Author Contributions

Copeland G. Winten, Gabriella Heruc and Jennifer J. Thomas: Conceptualization. Copeland G. Winten: methodology. Copeland G. Winten, Esben Strodl: validation. Copeland G. Winten, Gabriella Heruc: resources. Copeland G. Winten: data curation. Copeland G. Winten: writing – original draft preparation. Copeland G. Winten, Esben Strodl, Lynda J. Ross, Jennifer J. Thomas and Gabriella Heruc: writing – review and editing. Copeland G. Winten: visualization. Jennifer J. Thomas, Esben Strodl, Lynda J. Ross and Gabriella Heruc: supervision. Copeland G. Winten: project administration. All authors have read and agreed to the published version of the manuscript.

Ethics Statement

The study was conducted in accordance with the Declaration of Helsinki, and approved by the QUT Human Ethics Committee (HREA 2024‐8273‐18314).

Consent

Written informed consent has been obtained from the patients to publish this paper, with their feedback.

Conflicts of Interest

C.W., G.H., E.S. and L.R. declare no conflicts of interest. J.T. receive consulting fees from Equip Health and royalties for the sale of her books on feeding and eating disorders from Cambridge University Press and Oxford University Press.

1. Peer Review

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/jhn.70087.

Supporting information

JHND S1.

JHND S2.

JHND S3.

JHND S4.

JHND S5.

JHND S6.

JHND S7.

JHND S8.

JHND S9.

Data Availability Statement

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