An Appetite Stimulant for ARFID? Mirtazapine Against a Purely Behavioral Approach
This eight-year-old isn't restricting because of any fear of weight gain — his aversion is sensory, and it predates any concept of dieting he could even have. Whether an appetite-stimulating antidepressant belongs anywhere near a problem this specific, on evidence this thin, is what's actually being decided.
The mother of T.J., an 8-year-old boy, keeps a running list on her phone of the six foods he will reliably eat — plain pasta, a specific brand of chicken nugget, white bread, plain rice, apple slices, one particular cracker — and says the list hasn't meaningfully changed in over a year despite everything the family has tried. T.J. was born at 34 weeks and spent nine days in the NICU, and while nothing in his early feeding records points to a single dramatic cause, his mother has always suspected those first weeks of tube feeding and unfamiliar textures set something in motion that ordinary childhood pickiness never quite explains. He is an enthusiastic Cub Scout, genuinely social, articulate for his age about almost everything except food — he can describe exactly why a food is "wrong" (texture, almost always, rarely taste) but cannot talk himself past it, and gags reflexively, not performatively, when a new food touches his tongue.
What distinguishes his case from a weight-and-shape-driven eating disorder, and what makes the diagnosis avoidant/restrictive food intake disorder rather than anorexia nervosa, is that he has never expressed any concern about weight gain or body image at all — his avoidance is definitionally sensory and anxiety-based, not fear-of-fatness-based, and the distinction matters because it changes what pharmacology, if any, is even plausible here. His weight-for-age has fallen from the 40th percentile at age six to the 8th percentile now, a real and accelerating trajectory, not a stable low baseline, and six months of a structured pediatric feeding program — graded food exposure, a behavioral psychologist, no forced eating — has produced steady engagement in sessions but almost no change in his actual list of accepted foods. The medication on the table is mirtazapine, an antidepressant whose appetite-stimulating side effect — driven by its antihistamine and serotonin-receptor-blocking activity — is well documented in adult populations and has a small but real case-series literature specifically in pediatric feeding difficulty and ARFID — Gray and colleagues, 2018, describing weight gain in children and adolescents given mirtazapine for exactly this indication. That evidence is real but genuinely thin — case-series level, not a randomized trial — and the honest question in front of the team is whether "thin evidence, low apparent risk" is reason enough to add a drug to an eight-year-old's regimen for a problem that isn't, by definition, primarily about appetite at all.
Pediatric feeding team review
I think we should start mirtazapine. His growth curve isn't just low, it's actively falling — 40th percentile to 8th over two years, and that's accelerating, not stabilizing. Six months of good behavioral engagement with almost no dietary expansion tells me the current approach, on its own, isn't moving fast enough for what his growth trajectory actually needs. Mirtazapine's appetite-stimulating effect is well documented in adults, and there's a real, if small, case-series literature — Gray and colleagues' 2018 series among them — specifically describing weight gain in children and adolescents given mirtazapine for feeding difficulty. It's generally low-risk in pediatric use. I'd rather add a low-risk lever now than wait for the growth decline to become a more urgent problem.
I want to name directly what mirtazapine is actually built to fix, versus what's actually wrong with T.J. It's an appetite stimulant — its mechanism increases hunger drive. T.J. isn't refusing food because he isn't hungry. He gags reflexively at new textures and can tell you exactly why a food is "wrong" — that's a sensory and anxiety-driven avoidance, not an appetite deficit, which is the whole reason his diagnosis is ARFID and not anorexia. Making him hungrier doesn't touch the actual mechanism stopping him from eating new foods; it's solving an adjacent problem he doesn't have. And the evidence for using it here specifically is case-series level — small, uncontrolled, heterogeneous populations — not something I'd want to lean on over continuing a behavioral process he's genuinely engaged with.
I understand the urgency behind the growth curve, but I don't think a mismatched mechanism becomes the right answer just because the current approach feels too slow.
I think you're both right about different halves of the same problem, and I want to point at a target neither of you has actually proposed. The mechanism mismatch argument is correct — mirtazapine treats appetite, and appetite isn't T.J.'s problem. But the growth urgency is also real. What I'd ask is: what's actually stopping the exposure sessions from working faster? By his mother's own description, it's the anticipatory gag response and anxiety at the moment of contact with a new texture — not a lack of hunger. That's a genuinely different pharmacologic target than "make him hungrier." A brief, low-dose anxiolytic used specifically around exposure sessions — the same logic already used adjunctively in other anxiety-driven avoidance — is worth discussing with the feeding team as a mechanism-matched option, rather than either adopting or rejecting mirtazapine as the only medication question on the table tonight.
Agreed: hold mirtazapine given the mechanism mismatch with T.J.'s sensory- and anxiety-based avoidance, continue the structured feeding program unchanged, and refer the anxiolytic-around- exposure proposal to the full multidisciplinary feeding team for discussion rather than deciding it unilaterally in this consultation. Iron deficiency will be addressed with supplementation and rechecked in six weeks.
Not agreed: how long to continue the current behavioral-only approach before revisiting pharmacologic options if the growth trajectory doesn't stabilize. The Pediatrician wanted an explicit growth-based trigger — a defined further percentile drop — that would prompt reconsidering mirtazapine specifically regardless of the mechanism-mismatch concern, given how much weight the accelerating trajectory itself carries. The Feeding Program Psychologist preferred not pre-committing to a specific drug as the fallback, wanting any future medication discussion to start fresh from whatever T.J.'s actual presentation looks like at that point rather than defaulting back to mirtazapine because it was the first option raised. No specific threshold was set.
Educational content only — a composite teaching case, not a real patient encounter or a substitute for clinical guidance. About These Cases →