TL;DR: Somatic approaches to ARFID target nervous-system regulation and interoceptive awareness, meaning learning to notice hunger, fullness and other body cues. As of November 2026, a search found no published trials or case reports of somatic, polyvagal-informed or somatic experiencing therapy for ARFID; support rests on clinical reasoning and adjacent research on interoception. CBT-AR is the most-studied ARFID-specific therapy, tested so far in open trials, and family-based treatment produced more weight gain than an individual therapy in a 2026 randomized trial of 98 underweight children. In clinical practice, somatic work is positioned as an adjunct that does not substitute for evidence-based care or medical monitoring, and progress is read in foods added and volume eaten.
A patient in her thirties, a composite drawn from several patients, months into treatment for a narrow diet, brings in a podcast episode about polyvagal theory. She wants to know whether calming her nervous system could do what exposure has been doing slowly and uncomfortably. Body-based work may help her stay present through an exposure step, though no published study has tested it as an ARFID treatment, and a provider can easily promise more than the research supports.
The evidence
Avoidant/Restrictive Food Intake Disorder entered the DSM-5 in 2013, and clinical research on it is still young. The most-studied ARFID-specific therapy is cognitive behavioral therapy for ARFID (CBT-AR), and it has been tested so far in open trials rather than randomized ones. Thomas and colleagues (2021, Journal of Behavioral and Cognitive Therapy) treated adults in an open trial and called for randomized trials to confirm what the early data suggest. For children, family-based treatment moved from a pilot randomized trial of 28 families (Lock, Sadeh-Sharvit and L’Insalata, 2019, International Journal of Eating Disorders) to a randomized trial of 98 underweight children aged 6 to 12, in which it produced more weight gain than an individual therapy, though both reduced ARFID severity (Lock and colleagues, 2026, Journal of the American Academy of Child and Adolescent Psychiatry).
Somatic approaches have less support than that: as of November 2026, a search turned up no published trial or case report testing somatic therapy, polyvagal-informed therapy or somatic experiencing for ARFID. What supports the work is clinical reasoning and research on interoception in ARFID that overlaps with body-based methods without testing them as a treatment. A provider who describes somatic therapy as a proven ARFID treatment overstates the research, and although the work may still be useful, a patient should hear the evidence tier before agreeing to it.
What somatic frameworks offer
The DSM-5-TR (American Psychiatric Association, 2022) describes ARFID through three presentations that can overlap: sensory sensitivity to taste, texture, smell or appearance; low appetite or low interest in eating; and fear of an aversive consequence such as choking or vomiting. Somatic frameworks address the nervous-system side of these presentations.
Polyvagal-informed work treats the body’s threat response as the first target, on the reasoning that a person in a high state of arousal cannot approach a feared food. Polyvagal theory’s physiological claims are contested, with Grossman (2023, Biological Psychology) reviewing the evidence and disputing the theory’s physiological premises, and the practices built on it have not been tested in ARFID. Somatic experiencing principles applied to eating focus on the physical tension that gathers around a meal; in practice that means noticing where the tension sits, often the jaw, shoulders or stomach, and letting that area loosen before a bite. Interoceptive awareness training, which means learning to notice and interpret internal signals such as hunger, fullness and the physical signs of anxiety, targets a documented difference: a 2023 study of 30 adolescents with ARFID (Datta and Lock, Journal of Eating Disorders) found poorer interoceptive accuracy than in comparison participants on a heartbeat-detection task, in which participants count their own heartbeats without touching their pulse and the count is compared with the actual rate.
Interoceptive exposure asks a patient to produce a body sensation on purpose, such as fullness or a racing heart, and stay with it until it loses its alarm. For ARFID it has been described in a single case study: Zucker and colleagues (2019, International Journal of Eating Disorders) reported the Feeling and Body Investigators approach with a 4-year-old child; a randomized trial of the approach (NCT05105308) is listed as completed on ClinicalTrials.gov, with no results posted there or found in print as of November 2026. It is also a component of CBT-AR, so a somatic provider who offers interoceptive work is often using a technique the evidence-based model already contains.
Where it helps as an adjunct
The clearest place for body-based work, in the clinical view, is with patients whose arousal interferes with the treatment that has evidence. Someone who freezes at the table or goes blank mid-meal, whose fear climbs so high that a planned exposure stalls, may do better with a short regulating step first. Orienting, in practice, means looking around the room and naming what is there: the window, the clock, the color of the wall. Grounding means feeling the feet on the floor and the chair under the body, sometimes pressing the hands flat on the table or noticing the temperature of a glass, sometimes lengthening the exhale. A patient spends a minute or two on these steps so that she can stay present at the table for the exposure step the plan calls for. That sequence rests on clinical judgment, since no study has tested whether adding it improves CBT-AR outcomes. The same tools may fit adults who have spent years bracing against meals, a pattern described in ARFID in adults, for whom lowering the physical dread may make the graded steps of ARFID treatment tolerable enough to sustain.
For a patient with the sensory presentation, one who gags at a texture or cannot stay in a room with a particular smell, body-based work may lower the arousal around that texture or smell, but the change mechanism with evidence is sensory exposure, meaning repeated and graded contact with the taste, texture or smell itself, which is how the CBT-AR manual (Thomas and Eddy, 2019) treats this presentation; no study tests somatic work for sensory ARFID. For a child or teen, questions about somatic methods are best raised with the family-based treatment team, because that team runs the meals and any regulating step has to fit the plan for them.
Where it should not substitute for CBT-AR
Somatic work fails a patient when it replaces exposure, since she can feel calmer at meals and still eat the same narrow set of foods a year later. ARFID occurs at every body size, so recovery is not read from a scale; it is measured by the range and volume of food a person eats, by nutritional adequacy on lab work, and in children by growth. In clinical judgment, regulation that never leads to more eating leaves the ARFID itself untreated.
CBT-AR, as laid out in the Thomas and Eddy manual (2019), matches a module to each maintaining mechanism. For fear of aversive consequences, the module is a graded fear and avoidance hierarchy worked through exposure. For sensory sensitivity, it is sensory exposure. For low interest or appetite, it is preferred foods, hunger monitoring, and interoceptive exposure to fullness. The four stages and their mechanisms are described in CBT-AR for adults. Clinicians may use somatic work to prepare a patient for those modules and to steady a patient inside them, but it has no evidence as a replacement for any of them. It also cannot replace medical monitoring, since the DSM-5-TR (American Psychiatric Association, 2022) defines ARFID partly by its consequences: weight loss or faltering growth, nutritional deficiency, dependence on supplements or tube feeding, and interference with daily functioning.
How to evaluate a somatic provider
Ask about CBT-AR training first. In the clinical view, a clinician trained in the evidence-based model who adds body-based tools is on firmer footing than a somatic-only practitioner who has not worked with ARFID specifically. Ask who is watching the medical side, since weight, growth and labs need a physician or dietitian and regulation exercises do not supply that. Ask how progress is measured, and listen for foods added and volume increased. Ask, too, what the provider thinks the somatic evidence does not yet show.
The composite patient with her podcast leaves with a plan for grounding at the start of the meal and, after it, the same exposure step she had been dreading.
Symptoms that need same-day care
Some symptoms cannot wait for a therapy appointment at any body size. Fainting or near-fainting, chest pain, a very slow or irregular heartbeat, dizziness on standing, signs of dehydration, or going a day or more without eating or drinking call for contact with a physician the same day. For chest pain or fainting, call 911 or go to an emergency room. In a mental-health emergency, call or text 988.
This is general educational information, not individualized medical advice. ARFID can carry medical risk, and treatment decisions belong with a qualified clinician who knows your history.