ARFID, A Childhood Eating Disorder With No Proven Treatments Now Has Two. Only One Adds Weight

Avocado. Pomegranate. Chia seeds. Edamame, and yogurt with three sorts of berry in it, and eggs, which for a long while were the sticking point. That is a partial inventory of what Julia Ceresnak eats now and would not eat at 10, when she joined a clinical trial with a notebook, a sheet of stickers and a deal: fifty new foods and she would earn a painting kit, the mosaic sort.

She kept lists. Always foods, sometimes foods, not yet foods, and the small satisfaction of watching an item shift columns.

“That was very motivational for me, as a little kid,” says Julia, who is 15 now.

She was one of 98 children in the first adequately powered randomized trial ever run for avoidant restrictive food intake disorder, the condition that everyone involved simply calls ARFID. It entered the psychiatric diagnostic manual in 2013 and has been accumulating patients ever since, somewhere between two and six per cent of children and adolescents, without accumulating much in the way of evidence about what to do for them. Clinicians had guidelines borrowed from anorexia. They did not have a trial. “This is the first study, worldwide, to take a systematic, randomized, adequately powered approach to testing treatments for this disorder,” says James Lock at Stanford Medicine, who led it.

The count of proven treatments for children this age was zero. It is now two, which sounds tidy but isn’t, because the two treatments turned out to work in ways that don’t quite line up.

Not Fussy, Ill

What separates ARFID from anorexia is that nobody with ARFID is trying to be thin. Body image doesn’t come into it at all. There are instead three broad routes to eating almost nothing: a flat absence of appetite; a sensory recoil at texture or color or smell; and fear, usually the residue of a choking incident or an allergic reaction that taught the child eating was dangerous. Most children have some combination; in this trial roughly 90 per cent presented with mixed profiles. “Another group of patients have had a traumatic experience like a choking incident or allergic reaction and became afraid to eat,” Lock says.

The consequences are not subtle. Vitamin A can fall low enough to endanger vision; vitamin C low enough to produce scurvy, an affliction one associates with 18th-century sailors rather than 8-year-olds in California. Long term, children with ARFID may face stunted growth, short stature, impaired fertility.

And yet parents who raise it with a pediatrician are often told that picky eating is normal and the child will grow out of it. Which is true of picky eating, and not true of this.

Lock and his colleague Brittany Matheson recruited nationally between April 2021 and July 2025, randomly assigning each family to one of two approaches, 14 hour-long sessions over four months, everything by telehealth. The first, family-based treatment, hands the wheel to the parents. They take charge of what and how much the child eats, and only later does the child resume age-appropriate control. Siblings sit in. Everybody attends. “The therapist guides and consults them, but parents are the experts on their child, their family’s food culture and their family systems,” says Matheson, a clinical associate professor of psychiatry and behavioral sciences. Therapists work to separate the child from the disorder, which for a 6-year-old means finding an image that works: Matheson says a therapist might tell them “It’s like an alien is coming to your brain.”

The second approach, psychoeducational motivational therapy, inverts the arrangement. The child gets nine sessions, the parents get five separately, and parents are specifically asked not to push for behavioral change. Instead the therapist plays: inventing restaurant menus, picking a country the child wants to visit and investigating what people eat there. The starting point is that young children often have no framework for their own eating at all. “A 6-year-old doesn’t always have that insight,” Matheson says.

Two Answers, Not Quite the Same Question

Children in the family arm gained weight, moving from roughly 85.7 per cent of expected body weight to about 90.5 by the end of treatment, a large effect. Children in the individual arm went essentially nowhere, 86.7 to 86.5. On weight, in other words, family-based treatment won outright and the between-group difference was highly significant. But severity of ARFID symptoms fell substantially in both arms, by almost identical margins, and there the two treatments were statistically indistinguishable. Something loosened in these children regardless of which therapy they got. Parental confidence, measured on a scale adapted from anorexia research, rose sharply in the family arm only, a plausible mechanism for the weight difference that the team plans to test directly rather than assume.

An exploratory analysis found the most severely symptomatic children did better with family treatment. So did children whose parents reported a more authoritarian style, who fared poorly when handed the individual approach.

Lock offers an observation that cuts against most intuitions about eating disorders. “In one way, this disorder is easier to treat than other eating disorders, because the kids don’t really want it,” he says.

The caveats are real. Weight and height were reported by parents, not measured in a clinic. Only one standardized behavioral measure was used. Every child in the trial was underweight, so nothing here speaks to the many children with ARFID who are not: 42 per cent of otherwise eligible candidates, over two hundred kids, were turned away for that reason alone. And the PARDI scores across the sample sat below the suggested clinical thresholds, raising the possibility that this was a comparatively mild group. Durability wasn’t tested. Nobody yet knows what these children look like a year on.

Camp, Sleepovers, Birthday Parties

What sends families to the clinic is rarely the vitamin panel. “Lots of kids come to see us for treatment because they want to go to summer camp, a sleepover or a birthday party,” says Matheson. Restricted diets can stop a family taking a holiday. A child who eats nothing all day is a child who is tired and short-tempered by fifth period.

Julia was randomized to the individual arm, the one that didn’t move the weight needle across the group. She still sees Matheson. The prizes are gone; she now negotiates food for her own reasons, having planned strategies for camp and for school trips to Disneyland and to Costa Rica. Her mother Karen, listing the foods that have migrated to the always column, remembers the plainer version of the same ambition. “I was like, ‘I just want her to eat eggs!'” she says.

Julia eats eggs. “I like eggs a lot,” she says, which is the sort of sentence that means nothing at all unless you know what it cost, and there are perhaps two million children in America for whom it would mean something. The trial that produced it ran for four years and answered one question about one age group at one end of the weight spectrum. The next ones will need to ask about adolescents, about children who aren’t underweight, about whether any of it holds.

Frequently Asked Questions