CBT for ARFID

1. Introduction

Version followed: CBT-AR as published by Jennifer J. Thomas and Kamryn T. Eddy (Massachusetts General Hospital), Thomas & Eddy (2019) Cambridge manual, reconstructed from the authors' CBT-AR Patient and Family Workbook (version dated 12.4.18), MGH training slides (Eddy 2021; Thomas, IAEDP 2021) and the MGH Academy course outline. The manual itself is not drawn on. The full texts of the open trials (Thomas et al., 2020, children and adolescents; Thomas et al., 2021, adults) are not drawn on, apart from the adult trial's abstract. [uncertain: session-by-session detail that appears only in the manual]

CBT-AR (cognitive-behavioral therapy for avoidant/restrictive food intake disorder) is a flexible, modular outpatient CBT for ARFID in patients aged about 10 and older. It runs about 20–30 sessions over four stages and comes in a family-supported format and an individual format. [source: Eddy 2021 slides; Thomas IAEDP 2021; CBT-AR workbook; MGH course page]

Core idea of change: the model sees ARFID as maintained by three overlapping mechanisms, each kept going by avoidance:

• Sensory sensitivity: limited variety means little exposure to new tastes and textures.
• Fear of aversive consequences: choking, vomiting, allergic reaction, pain. Safety behaviours and avoidance keep the fear alive.
• Lack of interest in food or eating: limited volume blunts hunger and fullness cues.

Nutritional compromise, a biological predisposition, food-related trauma and limited opportunities for exposure feed into the model. Change comes from early behavioural change in how much and how regularly the patient eats, followed by exposure targeted at the maintaining mechanism(s) that apply. Treatment does not try to make patients "foodies". It aims to restore weight and growth, correct nutritional deficiencies, broaden intake across the five food groups and reduce psychosocial impairment. [source: CBT-AR workbook, "What is ARFID" and maintenance pages; Thomas IAEDP 2021; Eddy 2021 slides]

2. Framework

Formats [source: Eddy 2021 slides; Thomas IAEDP 2021; UTHealth CBT-AR page]

• Family-supported CBT-AR: children and early adolescents (about 10–15), and older adolescents or young adults living at home who have significant weight to gain. Parents attend and are coached to support eating.
• Individual CBT-AR: older adolescents (16+) and adults who are not underweight.
• Age cut-offs vary across sources. Eddy 2021 gives 10–15 for family-supported. Thomas IAEDP 2021 says "under 16 and/or underweight". The UTHealth clinic uses 12–16, plus 16–17 if underweight. Interventions are the same across formats; who attends differs. [uncertain: exact cut-offs; follow the manual]

Eligibility / readiness [source: Thomas IAEDP 2021; CBT-AR workbook readiness checklist]

• Aged 10 or older and able to engage cognitively (at most a mild developmental disorder).
• Eats at least some food by mouth: liquids or soft foods are enough, but not tube-dependent.
• Under a physician's monitoring and medically stable enough for outpatient care.
• Acknowledges an eating problem, sees some benefit in changing, and can attend weekly and do home practice.

Length and format

• About 20–30 weekly outpatient sessions. The workbook, UTHealth and the adult trial give 20–30; Thomas IAEDP 2021 says "approximately 20". UTHealth links the longer end to the need for weight restoration. [source: CBT-AR workbook; UTHealth; Thomas et al. 2021 abstract]
• Session length and the number of sessions per stage are not given in the sources. [uncertain]
• A training-course listing for health educators mentions "three stages". Every Thomas/Eddy source, including the manual's table of contents, describes four. Four is followed here. [source: Helm CHES page; Cambridge chapter list]

Typical session structure [source: Thomas IAEDP 2021]

1. Set the agenda together.
2. Weigh the patient.
3. Review home practice and monitoring records.
4. Deliver the intervention for the current stage.
5. Address the patient's or family's agenda items.
6. Plan at-home practice.

Stages [source: Cambridge chapter list; CBT-AR workbook; Eddy 2021; Thomas IAEDP 2021; MGH course page]

Team: a physician sets the goal weight range with the therapist, monitors refeeding risk in underweight patients and is consulted before high-allergy foods are tried. The workbook's nutrient table was prepared by a dietitian. The slides do not describe the dietitian's role in sessions. [source: Thomas IAEDP 2021; CBT-AR workbook]

3. Guiding principles

• Modular and formulation-driven. Pick and order Stage 3 modules according to the patient's own maintaining mechanisms, not a fixed sequence. [source: MGH course page; Cambridge ch. 8 title]
• Expect early change. Behaviour change starts in session 1; psychoeducation is never a stand-alone phase. [source: Thomas IAEDP 2021]
• Weight and medical safety first. In underweight patients, increase the volume of preferred foods before adding variety. Weigh at every session. Set the goal weight with the physician. [source: Thomas IAEDP 2021; Eddy 2021]
• Preferred foods are allies. Preferred foods are allowed and used to restore weight; nothing is taken away. [source: CBT-AR workbook]
• No force. The workbook says treatment is not about changing personality, forcing unusual foods or force-feeding. In practice: invite, set graded steps, and let the patient choose targets. [source: CBT-AR workbook]
• Nutritionally meaningful targets. Choose new foods that fill gaps in the five food groups, correct deficiencies and reduce social impairment, not foods that are merely new. [source: Thomas IAEDP 2021; Eddy 2021]
• Repetition beats one-off tasting. Comfort with a food builds over many exposures, often 10 or more, so plan repeats rather than once-only tastes. [source: CBT-AR workbook, Five Steps handout]
• Neutral, descriptive language about food. Describe what the food looks, smells and feels like; avoid "yuck/yum" judgments. [source: CBT-AR workbook]
• Finish the hierarchy. For feared outcomes, keep going until the most distressing task has been done. [source: Eddy 2021; Thomas IAEDP 2021]
• Home practice carries the change. Each session ends with planned at-home practice. Exposures use foods available at home so they can't be "forgotten". [source: Thomas IAEDP 2021]
• Parents as coaches in the family-supported format. Parents monitor and give specific instructions and praise at meals. Their involvement matters most when weight is low. [source: Eddy 2021; Thomas IAEDP 2021]
• Realistic endpoint. Patients are "unlikely to become 'foodies'" [source: Eddy 2021; Thomas IAEDP 2021]. Success means adequate, varied-enough eating with little impairment, not loving food.
• End on schedule. Ending therapy on time is a trained skill. Remaining goals go into the relapse prevention plan rather than extending treatment indefinitely. [source: MGH course page; synthesis on the rationale]
• Cognitive targets come from experience. Beliefs such as "Some new foods actually taste OK" and "Eating can be pleasant" are targeted through behavioural experiments, not debate. [source: Thomas IAEDP 2021]
• Collaborative stance. Set the agenda together, share the formulation and hand back decisions. [source: Thomas IAEDP 2021 on agenda and shared formulation; synthesis]

4. Intervention inventory

The inventory follows the stage order. Stage 3 modules are used only where the formulation shows that mechanism. Steps are paraphrased from the workbook and slides; example wording is the author's own unless quoted. [source for structure: CBT-AR workbook; Eddy 2021; Thomas IAEDP 2021; MGH course page]

4.1 Pre-treatment readiness and medical check (before Stage 1)

• Purpose / when: confirm the patient is suitable for outpatient CBT-AR and engaged, at intake. [source: CBT-AR workbook readiness checklist; Thomas IAEDP 2021]
• Steps:
  1. Confirm the ARFID diagnosis and assess severity. The MGH course uses interviews and self-report measures; the trials used the PARDI interview, and the NIAS is a brief screen.
  2. Confirm a physician is monitoring, and that the patient is medically stable and eats at least some food by mouth.
  3. Go through the workbook's readiness questions with the patient or family: is there an eating problem, is there a benefit to changing, can they attend weekly, will they do home practice.
  4. Choose the format (family-supported or individual) by age, living situation and weight status.
  5. Agree the treatment commitments: weekly attendance, parents' role, food logs, trying new foods.
• Example language (illustrative): "Before we start, I want to check this is the right time. It works best when you can come every week and practise at home in between. How does that fit right now?"
• Watch for: recent weight loss or medical instability (refer for medical assessment or a higher level of care). Also watch for tube dependence, and for shape or weight concerns that suggest a different eating disorder. [synthesis on the last point; ARFID criteria exclude body-image disturbance per Eddy 2021]

4.2 ARFID psychoeducation (Stage 1)

• Purpose / when: give the patient or family a non-blaming explanation of ARFID and how CBT-AR works. Sessions 1–2. [source: CBT-AR workbook; Eddy 2021]
• Steps:
  1. Describe the three presentations (sensory sensitivity, fear of aversive consequences, lack of interest) and ask which apply. More than one can.
  2. Distinguish ARFID from picky eating and from eating disorders driven by weight or shape. ARFID is a recognised psychiatric condition.
  3. Explain the consequences of limited variety and limited volume: nutritional compromise, health effects, blunted appetite.
  4. Present the four treatment goals: healthy weight, corrected deficiencies, eating from all five food groups, comfort eating with others. Say what the treatment is not (no personality change, no forcing unusual foods, no force-feeding).
  5. Outline the four stages.
• Example language (illustrative): "This isn't stubbornness or bad parenting. Some people's senses are more alert to taste and texture, some have had a frightening experience with food, and some just don't feel much hunger. Avoiding food makes all three of those worse over time, and that's what we'll work on."
• Watch for: parents blaming themselves or the child. The patient should recognise which presentation(s) fit them; that agreement is a criterion for leaving Stage 1.

4.3 Personalised formulation (Stage 1)

• Purpose / when: a shared map of what keeps this patient's eating restricted, which later guides module selection. Built in Stage 1 and shared on screen in telehealth. [source: Eddy 2021; Thomas IAEDP 2021; CBT-AR workbook maintenance pages]
• Steps:
  1. Draw the CBT-AR maintenance diagram together. Possible elements: biological predisposition; negative feelings and predictions about eating; food restriction (limited variety and/or volume); nutritional compromise; food-related trauma or negative experiences and the safety behaviours that followed; limited opportunities for exposure.
  2. Fill in each element with the patient's own examples (e.g. "I gagged on rice at age 6, so I cut food into tiny pieces").
  3. Mark the loops where avoidance keeps things going.
  4. Note which mechanism looks primary. This is the provisional priority for Stage 3.
• Example language (illustrative): "So when you see a new food, your body predicts it'll feel wrong, you skip it, and you never get to learn it might be OK. That's the loop we'll target."
• Watch for: a formulation that stays generic. Revise it as new information emerges. [synthesis]

4.4 Self-monitoring / parent-monitoring (Stage 1 onward)

• Purpose / when: an accurate record of intake, timing, and the thoughts, feelings and body sensations around eating. Starts in Stage 1. Parents monitor for younger patients. [source: CBT-AR workbook; Eddy 2021]
• Steps:
  1. Introduce the monitoring record: time, food or drink, thoughts and feelings, physical sensations. In the lack-of-interest module, add a hunger and fullness rating column (see 4.17).
  2. Ask for recording in real time. An app is fine and useful in telehealth.
  3. Review the record at the start of every session. Praise completion first, then look for patterns: long gaps, the same few foods, sensations before stopping.
• Example language (illustrative): "Let's look at Tuesday. Between breakfast and dinner there's a long gap. What was going on around lunchtime?"
• Watch for: blank or retrospectively filled records. Problem-solve the barriers rather than lecture (see Section 5). Patterns in the records should feed the formulation.

4.5 Regular eating (Stage 1)

• Purpose / when: a predictable pattern of meals and snacks, with preferred foods at each one. Usually 3 meals and 3 snacks, every 3–4 hours. From Stage 1 for everyone. [source: Thomas IAEDP 2021; CBT-AR workbook]
• Steps:
  1. Map the current eating pattern from the monitoring record.
  2. Agree fixed times for meals and snacks.
  3. Put a preferred food at every eating occasion, without adding new foods yet.
  4. Review adherence weekly and adjust the times to fit school or work.
• Example language (illustrative): "For now I'm not asking you to eat anything new, just to eat something you like at each of these six times."
• Watch for: skipped snacks, grazing instead of meals, and parents adding new foods too early.

4.6 Weight restoration through more preferred food (Stage 1, underweight patients)

• Purpose / when: restore weight quickly and safely in underweight patients, using foods they already accept. [source: CBT-AR workbook; Thomas IAEDP 2021; Eddy 2021]
• Steps:
  1. With the physician, set a goal weight range. The manual specifies a weekly rate of gain and a daily energy increase (numbers omitted here by house rule; set them with the physician and, where available, a dietitian).
  2. Weigh at every session and plot the weight with the patient or family.
  3. Increase the volume and energy density of preferred foods before adding variety. The workbook gives examples of energy-dense combinations of preferred foods to add on top of usual intake.
  4. Reduce physical activity as advised medically. For minors, parents supervise eating.
• Example language (illustrative): "Your body needs more fuel before we ask it to learn new foods. So this week the job is simply more of the foods you already like."
• Watch for: refeeding risk (physician monitors), a flat weight curve over 2–3 weeks (see Section 5), and patients restricting other foods to make up for it.

4.7 Therapeutic meal (Stage 1, underweight patients)

• Purpose / when: in-session coaching on increasing volume, and for parents on how to support it. For underweight patients only. [source: Thomas IAEDP 2021; Eddy 2021; MGH course page]
• Steps:
  1. Ask the family or patient to bring an energy-dense meal of preferred foods plus one small novel item.
  2. While they eat, coach parents (or the patient, in individual format) to increase volume. Use specific bite requests, persistent and reasonable demands, and specific praise. The slides give the example prompt "Don't put down your fork" [source: Thomas IAEDP 2021].
  3. Once adequate volume has been eaten, encourage one bite of the novel item.
  4. Debrief what helped, and turn it into home practice for family meals.
• Example language (illustrative, coaching a parent): "Try being really specific: 'Take two more bites of the pasta.' Then praise the exact thing: 'Nice job finishing that bite.'"
• Watch for: vague or pleading parental prompts, mealtime arguments, and therapists taking over instead of coaching. Success shows as the patient eating more than usual in session and parents reproducing the coaching at home.

4.8 Early change in variety (Stage 1, patients not underweight)

• Purpose / when: small, quick wins on flexibility for patients without significant weight to gain. [source: Thomas IAEDP 2021; Eddy 2021]
• Steps:
  1. Make small changes in how preferred foods are presented: a different brand, shape, plate or temperature.
  2. Reintroduce foods that were recently dropped from the diet.
  3. Rotate preferred meals rather than eating the identical meal daily.
  4. Drop minor safety behaviours, such as eating only from one plate or cutting food in a particular way.
• Example language (illustrative): "You used to eat apple slices last year. Could you try having them twice this week, with peanut butter like before?"
• Watch for: too-large first steps. Choose changes the patient rates as mildly uncomfortable. [synthesis]

4.9 Nutrition education: five food groups and deficiencies (Stage 2)

• Purpose / when: link the treatment targets to nutritional need. [source: Thomas IAEDP 2021; CBT-AR workbook]
• Steps:
  1. Review the five food groups (fruits, vegetables, protein foods, dairy and alternatives, grains) using a MyPlate-style exercise: sketch a typical plate and see which groups are missing.
  2. Review the patient's labs and diet against the workbook's nutrient table. The table covers signs, possible medical treatment and food sources for calcium, folate, iron, protein, vitamins A, B12, C, D and K, and zinc.
  3. Identify specific foods that could correct the patient's deficiencies.
• Example language (illustrative): "Your doctor mentioned low iron. Looking at this list, which of these iron-containing foods is closest to something you already eat?"
• Watch for: turning this into a lecture. Keep it brief and tied to the patient's own labs. Coordinate supplements with the physician. [synthesis]

4.10 Primary Food Group Building Blocks and target selection (Stage 2)

• Purpose / when: a concrete list of target foods for Stage 3, organised by food group. [source: CBT-AR workbook; Thomas IAEDP 2021]
• Steps:
  1. Go through the food-group lists on the Building Blocks worksheet. For each food, mark whether the patient eats it consistently and whether they are willing to learn about it.
  2. Prioritise foods that fill missing groups, correct deficiencies or reduce social impairment, e.g. a food served at school lunches or family dinners.
  3. Select the first new foods to learn about in Stage 3. Leave room to add mixed foods.
  4. Use the worksheet's tally boxes to count tastes of each food across Stage 3.
• Example language (illustrative): "You have no vegetables you eat regularly. Of the ones you'd be willing to learn about, which feel least scary?"
• Watch for: the patient choosing only foods similar to safe foods. That is a fine start, but make sure gap-filling foods follow. [synthesis]

4.11 Selecting and ordering Stage 3 modules (Stage 2)

• Purpose / when: decide which maintaining mechanisms to target and in what order. [source: MGH course page; Cambridge ch. 8 title "Maintaining Mechanisms in Order of Priority"]
• Steps:
  1. Return to the formulation and rate how much each mechanism drives the current restriction.
  2. Start with the mechanism that most limits nutrition or functioning, then move to the others as needed.
• Example language (illustrative): "Fear of choking seems to be what stops you most, so we'll start there and work on texture later."
• Watch for: the manual's own prioritisation rules were not available. [uncertain]

Exposure safety and consent (applies to 4.12–4.19). Before any food exposure, do the following:

• Confirm that a physician has ruled out a structural or medical swallowing problem, allergy or GI disease that would make the feared outcome likely. [synthesis]
• Check with the physician before trying high-allergy foods such as peanuts. [source: Thomas IAEDP 2021]
• Make sure every exposure is voluntary and graded. Explain the rationale, and agree that the patient may slow down, though not escape the task. [synthesis, consistent with the workbook's "no forcing" principle]
• Before interoceptive exercises (spinning, rapid drinking), check for medical contraindications such as seizure disorders, vestibular or cardiac problems, and electrolyte or fluid-restriction concerns. [synthesis]

4.12 Sensory sensitivity module: the Five Steps food exploration (Stage 3)

• Purpose / when: for patients whose restriction is driven by taste, texture, smell or appearance; learning about a new food through systematic, neutral exploration. [source: CBT-AR workbook Five Steps handout; MGH course page; Eddy 2021]
• Steps:
  1. Pick a target food from the Building Blocks list, preferably one available at home so it can be practised there.
  2. In early sessions, serve a very small portion.
  3. Work through five sensory steps in order: look at it (appearance), touch it (feel), smell it, taste it, then chew to experience its texture. At each step the patient describes the food in neutral, descriptive words rather than judgments.
  4. Ask for a rating, or simply a description, after each step, and move to the next step when ready.
  5. Log each taste on the Building Blocks tally. Repeat the same food across sessions and at home; comfort typically builds over 10 or more exposures.
  6. In later sessions, increase portions and fold the food into meals and snacks so it contributes to intake.
• Example language (illustrative): "Describe it like a scientist. What colour is it? Is it smooth or bumpy? Now smell it: is the smell strong or faint?"
• Watch for: pairing exposure with escape, such as spitting out immediately or washing it down at once (negotiate a gradual reduction), and judgment words creeping back. It is working when the patient tastes with less hesitation over repeated trials and rates or describes the food more neutrally. The cognitive target cited by the authors is "Some new foods actually taste OK" [source: Thomas IAEDP 2021].

4.13 Sensory module: incorporating new foods at home (Stage 3)

• Purpose / when: move a food from in-session tasting to regular eating at home. [source: CBT-AR workbook; MGH course page]
• Steps (choose one or more workbook strategies):
  1. Fade it in: gradually swap part of a preferred food for the new one, increasing the share over days.
  2. Add a flavour: a preferred condiment or spice makes the new food more familiar.
  3. Chain to a goal: step from a preferred food through similar foods toward a target food, e.g. one brand of crackers, then a similar cracker, then crackers with a new topping. Examples are illustrative.
  4. Switch it up: change how an accepted food is prepared so variations become tolerable.
  5. Deconstruct: break a mixed dish into its parts or layers, practise the parts, then recombine.
  6. Agree specific home tasks, e.g. which food, how many times, and at which meal.
• Example language (illustrative): "You're OK with plain pasta. This week, what if a quarter of the bowl is the whole-wheat kind, mixed in?"
• Watch for: foods that stay tasting-only and never reach meals. Move to bigger portions once tastes are tolerable.

4.14 Fear of aversive consequences module: psychoeducation on avoidance (Stage 3)

• Purpose / when: for patients who avoid foods for fear of choking, vomiting, allergic reaction or pain; makes the case for exposure. [source: Eddy 2021; Thomas IAEDP 2021; CBT-AR workbook exposure page]
• Steps:
  1. Draw the anxiety curve. Anxiety rises on approach and falls if the person stays with the situation. Escape gives short-term relief but strengthens the fear.
  2. Link it to the patient's own avoidance and safety behaviours, e.g. tiny bites, drinking after every bite, eating only soft foods, avoiding eating in front of others.
• Example language (illustrative): "Every time you skip the meat because you might choke, your brain files that as 'avoiding meat kept me safe', and the fear stays. Practising in small steps is how it learns otherwise."
• Watch for: agreement in words without willingness. Check the patient's readiness to start the hierarchy.

4.15 Fear module: fear hierarchy with SUDS (Stage 3)

• Purpose / when: a graded list of feared foods or eating situations. [source: CBT-AR workbook fear hierarchy; MGH course page; Eddy 2021]
• Steps:
  1. Introduce a 0–100 fear thermometer (SUDS).
  2. Brainstorm feared foods, textures, bite sizes and situations, e.g. eating quickly, eating in a restaurant, or eating without a drink nearby.
  3. Rate each item and order them from least to most distressing, with small steps between items.
  4. Choose a starting item in the low-moderate range. [synthesis on starting level]
• Example language (illustrative): "If 100 is the most scared you could be, where does one bite of chicken go? What about a whole piece?"
• Watch for: gaps between items that are too large; add intermediate steps. Also watch for hidden safety behaviours that need to become items of their own.

4.16 Fear module: in-session and home exposure with prediction testing (Stage 3)

• Purpose / when: work up the hierarchy to the most distressing item. [source: Eddy 2021; Thomas IAEDP 2021]
• Steps:
  1. Before the exposure, rate SUDS and the patient's certainty that the feared outcome will happen (e.g. "how sure are you that you'll vomit, 0–100%?").
  2. Do the exposure in session. Stay with it, and drop safety behaviours as agreed.
  3. Afterwards, re-rate SUDS and certainty, and ask what actually happened compared with the prediction. In the slides' example, a patient's certainty of vomiting after pizza dropped from 90% to 50% after one in-session exposure, and SUDS from 50 to 30.
  4. Assign the same or a similar exposure for home practice to generalise.
  5. Move up the hierarchy as distress at each step reduces, until the most distressing task has been completed.
• Example language (illustrative): "Before you take the bite, how likely do you think it is you'll choke? … You did it. What happened compared with what you predicted?"
• Watch for: subtle escape (tiny bites, keeping a drink at the lips) and stopping before the top of the hierarchy. It is working when predicted likelihood and SUDS fall across repeats and the patient starts eating the feared foods outside session.

4.17 Lack of interest module: hunger and fullness monitoring (Stage 3)

• Purpose / when: for patients with low appetite and early fullness; rebuilds awareness of body cues. [source: CBT-AR workbook; Eddy 2021; MGH course page]
• Steps:
  1. Add a 1–7 hunger/fullness rating to the monitoring record (low = very hungry, high = very full).
  2. Rate before and after each eating occasion.
  3. Use the workbook's guidance: start eating at moderate hunger, before cues fade, and continue to comfortable fullness rather than stopping at the first sign of fullness.
  4. Review patterns, e.g. whether cues are noticed at all and how early eating stops.
• Example language (illustrative): "Before lunch, where were you on the scale? And when you stopped?"
• Watch for: patients who never notice hunger. Regular eating (4.5) continues regardless, because cues may return only once intake is regular. [synthesis]

4.18 Lack of interest module: interoceptive exposure (Stage 3)

• Purpose / when: reduce fear or intolerance of the fullness, bloating or nausea that stop the patient eating enough. [source: CBT-AR workbook; Eddy 2021; Thomas IAEDP 2021]
• Steps:
  1. Identify which sensations end meals early.
  2. Choose an exercise that brings on a similar sensation: rapidly drinking water (fullness), pushing the belly out (bloating), spinning in a chair (nausea).
  3. Do it in session after the safety check. Rate the sensation and distress, stay with the sensation, and notice it pass without action.
  4. Plan home practice, and then eating practice in which the patient continues eating despite mild fullness.
• Example language (illustrative): "We're going to make your stomach feel full on purpose, so you can see that the feeling is uncomfortable but not dangerous, and it fades."
• Watch for: the safety screen above. It is working when distress about the sensations falls and meal volume rises.

4.19 Lack of interest module: more preferred food and increasing enjoyment (Stage 3)

• Purpose / when: increase intake of highly preferred foods and build positive associations with eating. [source: Eddy 2021; Thomas IAEDP 2021; CBT-AR workbook; MGH course page]
• Steps:
  1. In session, practise eating larger portions of highly preferred foods.
  2. Ask the patient to recall happy food memories.
  3. Have the patient describe several preferred foods using the Five Steps, to notice what they enjoy.
  4. Build in variety among preferred foods to keep interest up. [uncertain: the related training course lists "sensory-specific satiety" as a topic, but the procedure was not seen]
• Example language (illustrative): "Tell me about a meal you remember enjoying. What made it good?"
• Watch for: the cognitive target is "Eating can be pleasant" [source: Thomas IAEDP 2021]. Look for increased volume and spontaneous eating.

4.20 Stage 4: progress evaluation

• Purpose / when: a structured review of gains against the completion criteria, in the final sessions. [source: CBT-AR workbook; Eddy 2021]
• Steps:
  1. Review each criterion together: ARFID symptoms reduced or resolved; healthy weight and growth; regular eating across food groups; nutrients obtained mainly from food rather than supplements; eating no longer interferes socially.
  2. Compare the new-food list and Building Blocks tallies with the start of treatment.
  3. Restate that the goal is adequate, flexible-enough eating, not becoming a "foodie".
• Example language (illustrative): "When we started, you ate 8 foods. Now there are 25, and you went to the class trip lunch. Let's look at what's still on the list."
• Watch for: discounting progress because the patient still dislikes some foods.

4.21 Stage 4: personalised relapse prevention plan

• Purpose / when: keep progress going after treatment ends. [source: CBT-AR workbook; Eddy 2021; Thomas IAEDP 2021]
• Steps:
  1. Write down the improvements made.
  2. Identify likely triggers, e.g. illness, a choking or vomiting episode, stress, transitions such as starting university.
  3. Identify early warning signs, e.g. dropping foods, skipping snacks, weight loss.
  4. List which CBT-AR techniques to keep using.
  5. Set goals for continued change after treatment, such as specific foods still to learn.
• Example language (illustrative): "If after a stomach bug you notice you've stopped eating chicken again, what's your plan?"
• Watch for: vague plans. Make each warning sign link to a specific action. Finish on schedule. [source: MGH course page on ending on schedule]

4.22 Parent coaching (family-supported format, throughout)

• Purpose / when: equip parents to monitor, structure meals and coach eating. [source: Eddy 2021; Thomas IAEDP 2021]
• Steps:
  1. Parents complete monitoring and supervise regular eating and, if underweight, increased volume.
  2. Coach parents to give specific instructions and specific praise at meals, as in the therapeutic meal (4.7).
  3. Parents help set up home exposures and the food-incorporation strategies.
• Example language (illustrative): "Instead of 'just try it', try 'Touch the broccoli with your finger, then tell me how it feels.'"
• Watch for: pressure or bribing, or giving up and serving only safe foods. [synthesis]

4.23 Telehealth delivery adaptations

• Purpose / when: deliver CBT-AR remotely. [source: Thomas IAEDP 2021]
• Steps: arrange an in-person medical evaluation first. Have a scale at home for weighing at the start of each session. Share handouts and the formulation on screen. Use a self-monitoring app. Run tasting exposures with foods from the patient's own fridge or kitchen.
• Watch for: weights that can't be verified; coordinate with the physician. [synthesis]

5. Handling problems

The sources give little troubleshooting. The MGH course covers "troubleshooting common clinical challenges" in week 6, but its content is not drawn on. Items without a source tag are [synthesis], drawn from the CBT-AR structure and general exposure and eating-disorder CBT practice.

• Weight not rising in Stage 1.
  • Re-check the monitoring record for skipped snacks and for compensatory drops in other foods.
  • Raise the volume and energy density of preferred foods further, and add or repeat a therapeutic meal to coach parents. [source: Thomas IAEDP 2021 on volume-first and the meal]
  • Ask the physician to review medical causes and refeeding.
  • If weight keeps falling or the patient becomes medically unstable, step up to a higher level of care. UTHealth describes referral to residential or higher-level programmes when outpatient progress stalls. [source: UTHealth]
• Stuck in Stage 3 (tastes but no incorporation).
  • Check whether the target foods fill real nutritional or social gaps.
  • Increase repetition (10 or more tastes of the same food).
  • Move from tasting to portions in meals using the fade-in, chaining or deconstruction strategies (4.13).
  • Set home tasks with named foods and meals.
  • Revisit the formulation: an unrecognised fear or a lack of interest may be what is blocking progress.
• Overwhelm during exposure.
  • Do not end the session on escape. Step down to an easier item or a smaller portion (e.g. touch or smell only), stay until distress eases somewhat, and record what happened.
  • Add intermediate steps to the hierarchy.
  • With choking or vomiting fears, a panic-like reaction can follow; use the anxiety-curve psychoeducation (4.14). If true gagging or vomiting happens, pause, re-check medical factors, and resume at a lower step.
• Resistance / "I'll never like vegetables".
  • Use the authors' framing: the goal is adequate and flexible eating, not becoming a foodie. [source: Eddy 2021]
  • Let the patient choose among target foods, and test predictions behaviourally rather than arguing.
  • Link targets to the patient's own goals, such as eating with friends or having more energy.
• Non-adherence to monitoring or home practice.
  • Revisit the treatment commitment (4.1). Simplify the record or use an app.
  • Schedule home practice at a fixed meal and give parents a specific role.
  • Review the record at the very start of each session so it matters. [source on review: Thomas IAEDP 2021 session structure]
• Parental accommodation or conflict.
  • Parents may cook separate safe meals, avoid restaurants, or pressure and bribe.
  • Coach specific, calm prompts and specific praise (4.7, 4.22). Agree a home plan with both parents together.
  • Turn arguments over food into problem-solving outside meals.
• Ruptures (feeling forced, criticised or not believed).
  • Name it, and validate that ARFID is not a choice.
  • Return control: the patient chooses the pace within agreed steps. Restate what treatment is not, using the workbook's "no forcing" point.
  • Repair before the next exposure.
• Treatment running past 30 sessions.
  • Review progress against the Stage 4 criteria. Move remaining targets into the relapse prevention plan and end on schedule. [source: MGH course page on ending on schedule]
• Comorbid autism, ADHD, anxiety or developmental delay. The authors' eligibility requires enough cognitive capacity to engage; only mild developmental disorder is accepted. [source: Thomas IAEDP 2021]
  • For more impaired patients, refer to feeding or behavioural programmes.
  • For autistic patients, use visual supports, very small steps and predictable routines.
  • Treat co-occurring anxiety if it blocks exposure.
• Mixed presentations. Most patients have more than one mechanism. Sequence the modules by priority (4.11), and carry earlier gains forward with home practice while starting the next module.
• Allergy or medical fears that are realistic. Never expose to a medically contraindicated food. Coordinate with the allergist or physician, and set exposures that test the inaccurate part of the fear (e.g. the belief that any new food will cause a reaction). [source on checking high-allergy foods with the PCP: Thomas IAEDP 2021]
• Emerging weight or shape concerns during weight gain. Assess for a different eating disorder. CBT-AR assumes no body-image disturbance. [source: Eddy 2021 ARFID criteria]
• Self-harm or suicidality. This is not addressed in the sources. Follow standard risk assessment and safety planning, and coordinate with the medical team.

Sources

Sources:

• CBT-AR Patient and Family Workbook (Thomas & Eddy), version 12.4.18, hosted by Kelty Eating Disorders
• Eddy, K. T. (2021). Eating disorders, MGH Child Psychopharmacology course slides
• Thomas, J. J., & Chakoff, E. (2021). Assessing ARFID and delivering CBT-AR via telehealth, IAEDP slides
• MGH Academy: CBT for ARFID, An Online Introduction (February 2026), course page
• Cambridge Core: Overview of CBT-AR (Chapter 5), with the book's table of contents
• Thomas et al. (2021). Cognitive behavioral therapy for adults with ARFID, abstract (em-consulte)
• UTHealth Houston: CBT-AR clinic page
• Helm Publishing: CBT for ARFID (CHES course listing)

Books and papers cited (not drawn on in full):

• Thomas, J. J., & Eddy, K. T. (2019). Cognitive-Behavioral Therapy for Avoidant/Restrictive Food Intake Disorder: Children, Adolescents, and Adults. Cambridge University Press. The primary manual; not drawn on.
• Thomas, J. J., Wons, O. B., & Eddy, K. T. (2021). The Picky Eater's Recovery Book. Cambridge University Press. Self-help companion; not drawn on.
• Thomas, J. J., Becker, K. R., Kuhnle, M. C., et al. (2020). Cognitive-behavioral therapy for avoidant/restrictive food intake disorder: feasibility, acceptability, and proof-of-concept for children and adolescents. International Journal of Eating Disorders. Not drawn on. [uncertain]
• Thomas, J. J., Becker, K. R., Breithaupt, L., et al. (2021). Cognitive-behavioral therapy for adults with avoidant/restrictive food intake disorder. Journal of Behavioral and Cognitive Therapy, 31(1), 47–55. Abstract only.