Feeding and Eating Disorders

Case: Mateo, 15. Five chapters follow: the case, epidemiology, presentation, functional neuroscience, and treatment.

The case: Mateo, 15

Mateo Moreno is fifteen and in tenth grade. His pediatrician referred him after a preseason sports physical. Her note reads: "15-year-old male, approximately 60 lb weight loss over ten months, initially attributed to healthy eating after gastroenteritis last winter. Long-standing selective eater. Resting heart rate 44, rising to 84 on standing, blood pressure 94/58 supine. Query ARFID versus anorexia nervosa." At two earlier visits that year she had told his parents he "looked great."

Mateo's account is short. "I'm eating healthy. I cut out junk. Everyone said I looked better." Food "sits like a rock." He is full after half a sandwich, nauseated if he eats in the morning. He does not want to be thin: "That's not what this is. I just don't want to go back to how I was."

Developmental and family history

Mateo was born at term and met milestones on time. From toddlerhood he ate a narrow range: plain pasta, chicken nuggets, crackers, apples, milk. Mixed textures made him gag. At four he refused yogurt because it was "slimy" and has not eaten it since. His mother, Luz, remembers a feeding clinic being suggested at three; the family "managed by keeping it simple." In elementary school his weight rose. At thirteen his body mass index was near the 93rd percentile. In seventh grade, boys on the bus called him names about his weight for most of a year. He stopped taking the bus and told his parents only the following summer. He has always been careful and a little anxious, a rule-follower who dislikes surprises.

Luz "has dieted her whole life" and uses a calorie-tracking app. Mateo's father, Ray, was treated for panic attacks in his twenties. A maternal aunt was hospitalized for an eating disorder at nineteen. His sister Sofia is twelve.

The year of weight loss

Last January Mateo vomited for two days with gastroenteritis. Afterward food "didn't feel right," and he stopped eating breakfast. Dairy went next ("makes me feel sick"), then fried food. In March his track coach told him he would be faster if he stayed lean. By June he was running alone after practice, and his weight had fallen from 168 lb to 141 lb at 5 ft 8 in. In September he weighed 108 lb.

Observations in session

Mateo keeps his jacket on in a warm room. His hands are cold. He answers briefly and looks at the floor when Luz describes how he "used to look." He becomes animated only about running: "If I don't run it feels wrong. Not guilty exactly. Wrong." Asked what he avoids, he lists textures first ("anything mushy, anything with sauce"), then "anything with oil or fat." Nuts, which he ate by the handful last year, are now "basically pure fat." He weighs himself each morning and checks his stomach in the mirror "to make sure I'm not going backwards." He denies vomiting since the illness, laxatives, or diet pills. He denies eating with loss of control now, but in middle school he ate "a lot, at night, when no one was around," and hated it. He follows several fitness accounts. His mood is "fine, tired." He concentrates poorly, falls asleep late, and calls the referral "an overreaction."

Collateral

Luz says Mateo has always hated the noise of family dinners and calls his eating in his room "his sensory thing." Sofia says he began taking his plate upstairs in April, after their father remarked that he had "barely touched it," and that he cuts food into small pieces and eats after everyone else has finished. His coach calls him "the most dedicated kid on the team" and had not noticed the loss: "everyone gets leaner in season."

Several of these observations look alike. The yogurt refused at four and the nuts refused at fifteen are both foods Mateo will not eat. The fullness after the January illness and the fullness after half a sandwich in September are both reported as "full." Eating upstairs is either a sensory accommodation or a way to eat unobserved. A resting heart rate of 44 belongs to a trained runner or to a starving one. The chapters that follow sort each pair.

Epidemiology

How common

The NCS-A found a lifetime prevalence of eating disorders of 2.7% among US adolescents aged 13 to 18 [53]. The review that anchors this chapter gives disorder-specific figures, mostly from adult women, and shows how much the definition governs the count [99]. Lifetime anorexia nervosa was 2.2% in women in a Finnish cohort, rising to 4.2% under a broad definition, and 0.24% in men; bulimia nervosa 1.7% in women in the same cohort, and 0.9% to 1.5% in women and 0.1% to 0.5% in men in two-stage US and European studies; binge-eating disorder 1.9% in women and 0.3% in men across six European countries, 3.5% and 2.0% in US adults, and among US adolescents aged 13 to 18, 2.3% in girls and 0.8% in boys [99]. The female-to-male ratio for anorexia nervosa in the Finnish twin study was 10:1 [99]. In adult outpatient settings, eating disorder not otherwise specified is the most common diagnosis, about half of samples, with bulimia nervosa about a third and anorexia nervosa the remainder; inpatient samples are mostly underweight EDNOS or anorexia nervosa [100].

Incidence and its peak

The incidence of anorexia nervosa was stable across two periods, 7.7 versus 7.4 per 100,000 person-years, but concentrated: 109.2 per 100,000 person-years among girls aged 15 to 19 [99]. The incidence of bulimia nervosa fell from 8.6 per 100,000 in 1985 to 1989 to 6.1 in the later period [99]. Mateo is a 15-year-old boy. The incidence peak reported for girls falls at his age; the sex ratio makes him the rarer patient, and the presentation chapter's account of why several adults called his weight loss healthy is the clinical face of that ratio.

Mortality

Anorexia nervosa carries a mortality of 5.1 deaths per 1,000 person-years and a standardized mortality ratio of 5.86 (95% CI 4.17–8.26), with one in five deaths by suicide; the standardized mortality ratios for bulimia nervosa (1.93, CI 1.44–2.59) and EDNOS (1.92, CI 1.46–2.52) are lower but still elevated [99]. The medical assessment in Mateo's presentation chapter, with a resting heart rate of 44, is where those figures enter the room.

Who is missed

Two sources describe populations the stereotype leaves out. Adolescents with a weight history in the overweight or obese range "represent a substantial portion" of those presenting for eating-disorder treatment, and their symptoms often go unrecognized because of their higher weight; the source is a case report with commentary and gives no figure [92]. A content analysis of 66 US film and television characters with eating disorders found them 89.39% women, 84.85% White, 75.56% heterosexual and 84.85% under 30, against population claims the authors cite: approximately one third of US people with an eating disorder are male, one in seven males will experience one by age 40, Black adolescents are 50% more likely to engage in bulimic behaviors, and LGBTQ+ adolescents and adults have higher incidence [98]. Those population claims are the media study's summary of other work and should be sourced to their originals before being cited as figures.

Avoidant/restrictive food intake disorder

ARFID's prevalence has been measured in the clinics where it hides. In a retrospective chart review of 2,231 consecutive referrals aged 8 to 18 to 19 pediatric gastroenterology clinics, 33 (1.5%) met DSM-5 criteria, 67% of them male, with presentations mainly of insufficient intake or little interest in eating (n=19) or a sensory-based limited diet (n=7); a further 2.4% met at least one criterion with insufficient documentation, and the authors call full-criteria ARFID rare in that treatment-seeking sample [102]. In 129 consecutive referrals aged 6 to 18 for pediatric neurogastroenterology evaluation, 8% met full criteria and a further 15% had clinically significant avoidant or restrictive eating without enough information to diagnose, 23% in all; those with ARFID symptoms were older, more often female (79% versus 51%) and more often reported eating or weight concerns, and 67% of those with a documented rationale cited fear of gastrointestinal symptoms [103]. The two clinics disagree on sex and on rate because they see different children; Mateo's lifelong selective eating and the post-gastroenteritis onset make both patterns relevant.

Risk

In the Great Smoky Mountains Study, a prospective cohort of 1,420 children assessed across ten waves from ages 9 to 25, victims of bullying were at increased risk for symptoms of anorexia and bulimia nervosa in childhood and adolescence after adjustment for prior symptoms, psychiatric status and family adversity; bullies were at increased risk of bulimia symptoms and bully-victims of anorexia symptoms; there was little evidence of sex differences; and childhood bullying was not associated with persistent symptoms into adulthood [2]. A meta-analysis of 77 studies found media exposure related to body dissatisfaction, thin-ideal internalization and eating behaviors with small-to-moderate effects (d = −0.28, −0.39, −0.30) in women [97]; whether the finding transfers to an adolescent boy is a claim the presentation chapter marks as needing a source.

Sources not retrieved

The public-facing APA page [91], the NICE guideline [95], the 2006 APA practice guideline [96] and its 2023 revision [299], and the two books [93, 94] carry guidance rather than epidemiological figures; they are cited in the treatment chapter (and #94 in the neuroscience chapter), not here.

What this case cannot establish

A 15-year-old boy sits at the age of the incidence peak reported for girls and on the wrong side of the sex ratio, in the group two sources say is missed. The numbers explain the delay in his referral; they do not decide between anorexia nervosa and ARFID, which is the presentation chapter's work.

Source status

#99 and #100 from open-access full text; #53 from the article text; #2 from the Crossref abstract; #92, #97, #102, #103 from published abstracts; #98 from open-access full text. #91 web text and #95 recommendations 1.2.8, 1.3.10 and 1.3.15 verified in the citation check of 2026-10-05; #96 located (2006 third edition), with a 2023 fourth edition [299] now the current guidance; #94 confirmed as a bibliographic record only; #93 confirmed as the 2017 third edition (Mehler and Andersen), text not inspected.

Figures for this chapter

Presentation

What was observed and what was concluded

For ten months, several adults watched Mateo lose weight and concluded different things: healthy eating, dedication, a boy who looked great. The observations were shared; the interpretations were not. Assessment begins by separating them again.

What can be observed directly is intake, its timing and setting, the food categories that have disappeared, the weight trajectory, the running, and the checking. What Mateo perceives is fullness, nausea, cold, and a feeling that is "wrong" when he does not run. What he interprets is that fullness means food sitting like a rock, that fat is a danger, and that any move toward his former weight is "going backwards." What he participates in is a set of daily practices: weighing, checking, running, cutting food small, eating last and alone, and rejecting the referral's premise. Each layer can be described before deciding which disorder it belongs to.

The same behavior can have different sources. Refusing yogurt at four and refusing nuts at fifteen are both refusals; the first followed a gag, the second follows a calculation about fat. Eating upstairs is, in Luz's account, a lifelong sensitivity to the noise of the table; in Sofia's, it began the week after a comment about how little he ate. Neither has to be wrong. The question is which function each behavior serves now.

Three labels, one pattern

Anorexia nervosa is defined by restriction leading to a significantly low weight, an intense fear of gaining weight or persistent behavior that interferes with weight gain, and a disturbance in how body weight or shape is experienced, with restricting and binge-eating/purging subtypes [91]. Two features of the definition matter for Mateo. The fear need not be spoken; behavior that interferes with weight gain counts. And "significantly low" is judged against age, sex, and the individual's own trajectory, not a single cutoff. Mateo has fallen from near the 93rd percentile to below the 10th in a year when he should have been growing. He denies wanting to be thin, but he selectively removes fat, weighs and checks daily, and names the reference point he is guarding against.

Atypical anorexia nervosa is the same pattern with weight still in or above the normal range. Adolescents with a history of obesity can develop the full psychological and medical picture of a restrictive eating disorder while their weight reads as normal, and recognition is delayed because the loss is praised [92]. This is Mateo's January-to-June window: at 141 lb he was at a healthy weight for the first time in years, and nobody asked how he had gotten there.

Avoidant/restrictive food intake disorder describes avoidance or restriction producing weight loss, nutritional deficiency, supplement dependence, or psychosocial impairment, without the body-image disturbance that defines anorexia nervosa; the avoidance is based on sensory characteristics, apparent lack of interest, or concern about aversive consequences such as vomiting [91]. Mateo's early childhood is a sensory pattern: narrow range, gagging on mixed textures, a feeding clinic nearly consulted. The January gastroenteritis supplied a second ARFID-type route: an aversive experience followed by avoidance of the sensations that preceded it. Adolescents with this picture present in pediatric gastroenterology settings, where restriction sits alongside functional gastrointestinal symptoms and a measurable minority of patients meet ARFID criteria [102, 103]. Referred for nausea in February, Mateo would have gone there.

The hinge between the labels is body-image disturbance, and in a fifteen-year-old boy who denies it, the hinge is found in behavior and specific language rather than in a confession. The words to listen for are not "I want to be thin" but "going backwards," "pure fat," and the mirror each morning. Those are present. So are the sensory history and the fear of fullness. One label does not erase the other pattern; what must be recorded is which processes operate now, because treatment will need to address each of them.

Bulimia nervosa and binge-eating disorder stay live until asked about directly. Both involve recurrent episodes of eating an objectively large amount with loss of control; bulimia nervosa adds compensation by vomiting, laxatives, fasting, or exercise, and binge-eating disorder does not [91]. Mateo denies current episodes but describes secret night eating in middle school that he "hated," a loss-of-control history and a reason to ask again once refeeding begins. His running is training in his account and compensation in the pattern; the test is what happens when it is blocked. Purging leaves evidence, including hypokalemia, parotid enlargement, dental erosion, and knuckle abrasions, and the examination looks for it even when it is denied [93].

Observation	Reads as anorexia nervosa	Reads as ARFID	Distinguishing question
Avoids nuts, oil, dairy	Fat avoided to prevent gain	Dairy avoided after post-illness nausea	What is each food expected to do?
Full after half a sandwich	Starvation-slowed emptying	Fullness feared since January	Is fullness dreaded or welcomed?
Eats alone upstairs	Avoids observation of intake	Avoids sensory load	When did it begin, and what changed?
Runs after practice	Compensatory; distress if blocked	Not an ARFID feature	What happens when a run is blocked?
Weighs and checks daily	Body-image disturbance	Not an ARFID feature	Present in Mateo

The medical examination is part of the phenomenology

Mateo's resting heart rate of 44, its rise to 84 on standing, his blood pressure of 94/58, and his cold hands in a warm room are observations in the same sense as his words. Bradycardia, orthostatic instability, hypothermia, and cold extremities are among the most consistent physical findings of restrictive eating disorders and reflect an adaptive slowing that reverses with nutritional rehabilitation [93]. The same heart rate in a trained distance runner is a finding of fitness; on the monitor the two are identical. What distinguishes them is the weight trajectory, the orthostatic change, the temperature, and the response to food over the following weeks.

Laboratory assessment looks for the consequences of restriction and the risk of correcting it: electrolytes, phosphate, magnesium, glucose, a blood count, liver enzymes, thyroid function, and an electrocardiogram [93, 96]. Refeeding syndrome, in which returning carbohydrate drives an insulin surge that shifts phosphate, potassium, and magnesium into cells, is the hazard of the first days of nutritional rehabilitation, and its risk rises with the degree and speed of weight loss [93]. Mateo has lost more than a third of his body weight in ten months. The examination establishes not only how ill he is but how carefully the first week of treatment must be run. Guidance warns against using body mass index, illness duration, or any other single measure to decide whether to offer treatment [95]; the growth chart with its three unflagged weights is the most informative document in the referral.

One finding sits on both sides of the differential. Starvation slows gastric emptying, and early satiety, bloating, and constipation are consequences of restriction that improve with refeeding [93]. Mateo's fullness in September may be a starvation symptom, a learned fear from January, or both. It cannot be assigned to either label on its own.

Why it was read as health

Mateo began from a higher weight, and weight loss in adolescents with a history of obesity is praised until it is far advanced [92]. He is a boy who runs, and media portrayals of eating disorders concentrate on a demographic he does not resemble, so nobody around him had a template for what was happening [98]. His reasons were plausible: a stomach bug, a coach's advice.

Two pieces of history belong in the formulation without being mistaken for causes. Being bullied in childhood prospectively predicts later eating-disorder symptoms [2]; the bus in seventh grade is relevant to what "how I was" means to Mateo, not proof of how the disorder began. Media exposure is associated with body dissatisfaction and disordered eating, though that meta-analytic evidence concerns women and thin ideals [97]. A meta-analysis in men links media images to body-image concerns [251]; whether the lean, muscular content Mateo follows has that effect in adolescent boys, or drives restriction, is a separate question. Comorbidity across classes of disorder is common in adolescent samples [53]; his anxious temperament, poor sleep, and low concentration need their own assessment once starvation is no longer producing them. Eating disorders carry elevated mortality [99], which is why the medical findings rather than the label set the pace of what follows.

What this case cannot establish

The first assessment cannot establish whether Mateo's body-image disturbance preceded his restriction or emerged from it, whether his early selective eating is a separate condition or a vulnerability the disorder used, or which of his current symptoms are the starvation state rather than the person. It cannot tell, from the heart rate alone, whether he is fit or unwell, or, from the fullness alone, whether he is afraid or slowed. It records that all of these are present and that the physiology requires action before the label is settled.

Figures for this chapter

Functional neuroscience

Part 1. The functional layer

A signal and its appraisal are two operations

Mateo reports fullness after half a sandwich. The report bundles two operations. Interoceptive signaling is the process by which the body's internal state, including gastric distension, hunger, temperature, and heart rate, becomes available to the brain. Interoceptive appraisal is the process that assigns that signal a meaning: safe or dangerous, progress or failure. The sentence "I'm full" does not say which has changed.

Both may have. Starvation slows gastric emptying, so the same half sandwich produces more distension for longer than it did a year ago, and early satiety and bloating are ordinary consequences of restriction that improve with refeeding [93]. That is a change in the signal, and it belongs to the starvation state rather than to Mateo's beliefs. The appraisal has changed as well. In January fullness was appraised as illness. By September the same sensation means food "sitting like a rock" and, beneath that, the beginning of "going backwards." A meal can be aversive for two reasons that produce the same complaint. Treatment will have to change the signal by restoring nutrition and change the appraisal by supervised experience of fullness that ends without the feared outcome.

Reward can be reassigned

The Depressive Disorders module (Jonah) distinguished anticipatory reward responsiveness, the motivational response before an outcome, from hedonic generation, the pleasure during it. In a restricting adolescent the more striking change is not that food has lost value but that something else has gained it. The number on the scale, the completed run, the plate left full, the coach's "leaner," the pediatrician's "you look great": each is an outcome Mateo's behavior reliably produces. Restriction has acquired reward.

Part of that reward is relief. When a feared event is expected and an action prevents it, the reduction in anxiety reinforces the action. This is negative reinforcement, and it explains why restriction can feel like accomplishment without pleasure. Mateo's "not guilty exactly, wrong" describes the state that the run removes. The run is selected because it takes the state away, not because it is enjoyed.

Reactive versus instrumental sorts the first of the case's look-alike pairs. Refusing yogurt at four was reactive: a texture produced a gag, and the food was avoided thereafter. Refusing nuts at fifteen is instrumental: the food is avoided because of what it is calculated to do. A sensory refusal is addressed by graded exposure to the texture; an instrumental refusal by changing the contingency, so that eating the nuts neither produces the feared outcome nor removes a feared state. A plan that treats both as "picky eating" will succeed with the yogurt, fail with the nuts, and may read the failure as Mateo not trying.

This describes contingencies in his history, not what his brain computes.

Threat, avoidance, and the missing prediction error

A threat representation is a held expectation that a specified event will be followed by harm. Mateo holds at least two. One attaches to fullness and dates from January: fullness was followed by vomiting, and the expectation has not been tested since. The other attaches to weight gain: gaining weight was followed by the bus. These are different representations, belonging to different diagnostic categories, maintained by the same process.

Avoidance learning is that process. The learning rule from the Depressive Disorders module (Jonah) applies here in reverse:

New expected value = old expected value + learning rate × (experienced outcome − expected outcome).

An avoided meal produces no experienced outcome. The parenthesis is empty, and the expectation cannot update. Every skipped breakfast preserves the belief that breakfast would have been unbearable. Avoidance is not merely a symptom; it is the mechanism by which the expectation survives contact with a world that would otherwise correct it. The content of the expectation therefore matters for treatment. Exposure to fullness tests the January representation; exposure to fat-containing food at a restored weight tests the other. Mateo needs both experiments.

Habit under starvation

Habit formation is the process by which an action, with repetition, shifts from being selected for its expected outcome to being triggered by its context. From the inside, habit does not feel like a decision; its omission produces discomfort without a reason attached. "If I don't run it feels wrong. Not guilty exactly. Wrong." is a precise description of cue-triggered action with distress on omission.

One hypothesis holds that in anorexia nervosa restriction undergoes this shift, so that goal-directed dieting becomes stimulus-driven ritual that persists after the goal is reached or forgotten [252]. A later experimental study did not find a generalized habit bias in the disorder [253], and neither study establishes that starvation accelerates the shift, so the account remains a hypothesis. Whether it applies to Mateo cannot be established from his report, because a habitual action, when questioned, is supplied with a reason after the fact. The informative observation is what happens when the run is blocked. If distress rises and subsides without the run, the action was maintaining a state; if another activity produces the same relief, it was serving a goal. Either is more useful than asking him why he runs.

Set-shifting and rigidity: capacity versus access

Set-shifting is the capacity to change the rule or goal that is currently governing behavior. People with anorexia nervosa show reduced set-shifting on neuropsychological tasks during illness; after weight restoration the findings vary by task, and difficulties reported in unaffected siblings suggest a possible familial trait [254, 255]. Mateo's parents describe a rule-follower who dislikes surprises. Temperament and starvation may both contribute to the rigidity visible now.

A test of set-shifting administered this week would measure what Mateo can do at 108 lb after ten months of restriction, not a trait. The same holds for his concentration, sleep, flat mood, and conviction that the referral is an overreaction. Each is a reading under a state that constrains access to capacities he may possess. The rigidity that makes it hard for him to consider that his eating is not healthy is, in part, a product of the eating. The formulation has to be written twice: now, and after nutrition is restored, when what remains can be attributed to Mateo.

The body as a representation

A body representation is a stored model of one's own body, including its size, its history, and its meaning. It is not a photograph and it updates slowly. Mateo's model is anchored to a body from seventh grade, with the bus attached. When he checks his stomach each morning, the comparison is between the image and that anchor, not his present body. This is what "disturbance in the way body weight or shape is experienced" describes. It need not be a perceptual illusion; Mateo may see his size accurately and still experience every increase as a move toward the remembered body.

The cognitive model of eating disorders adds self-evaluation weighting: the extent to which shape, weight, and control over eating determine a person's judgment of their own worth, the core representation that restriction, checking, and avoidance serve [100]. Mateo does not use these words. What the assessment observes is checking, comparison, and a fixed anchor; what it cannot establish is whether he misperceives his size or how much of his self-worth rides on the answer.

Starvation as a state variable

Everything above runs under a state that changes it. Prolonged semi-starvation in healthy volunteers has been reported to produce preoccupation with food, ritualized eating, irritability, low mood, social withdrawal, and loss of interest, resolving with refeeding [256]. These are not traits and not, by themselves, an eating disorder. They are what an energy deficit does to a person.

The starvation state therefore appears in the formulation as a variable in its own right. It alters interoceptive signals [93], constrains access to set-shifting, lowers mood, disrupts sleep, and generates the rituals Sofia notices: cutting food into pieces, eating last. It sits underneath the body representation and the avoidance learning, making each harder to change. "Cognition" in a starved adolescent is not a description of the adolescent; it is a description of the adolescent under that variable. Some of what looks like Mateo will turn out to be starvation, and only weight restoration can show which.

Part 2. Circuits, development, and genes

Distributed systems for interoception, value, and habit

The functions above are implemented by interacting systems, not a single region. The insular cortex participates in representing the body's internal state; orbitofrontal and ventral striatal regions participate in assigning value to food and to outcomes; dorsal striatal circuits participate in the shift from goal-directed to habitual control; and hypothalamic circuits, informed by circulating signals such as leptin that fall during starvation, regulate energy balance and modulate the value of food [257]. Imaging studies of anorexia nervosa have reported altered responses in these systems during taste, food-cue, and interoceptive tasks, in samples that are mostly female, mostly adult, and studied during or after illness, so the findings cannot separate a predisposing difference from a consequence of starvation [257]. None of this locates Mateo's difficulty in a region; it identifies where the operations of Part 1 would have to be implemented.

Reversibility is the finding most relevant to his formulation. In a longitudinal study of adults with anorexia nervosa, gray matter volume reduced during acute illness recovered substantially with weight restoration, though residual deficits remained [258]. If so, structural imaging during illness, like testing during illness, is a reading of a state.

Development: a window that opens in adolescence

Eating disorders characteristically begin in adolescence, a period in which puberty changes body composition, endocrine signaling, and the social evaluation of the body all at once [94]. A fifteen-year-old boy is expected to be gaining weight and height; a sixty-pound loss in that window is a larger deviation from trajectory than the same loss in an adult, and its effects on growth and bone are a reason for urgency the number alone does not show [93]. The bus, the coach's comment, and the fitness accounts all landed on a body in transition, at the age when the body representation is being rebuilt.

Genes: liability without a mechanism

A review of twin studies reports substantial heritability for anorexia nervosa, and genome-wide association studies have identified common-variant loci with genetic correlations to psychiatric and metabolic traits [259, 260]. The aunt hospitalized at nineteen is consistent with familial liability. It is not a mechanism, does not identify what was inherited, and does not explain why Mateo rather than Sofia. For ARFID the evidence is thinner. A twin study has estimated heritability for a broad ARFID phenotype in Swedish children, and clinical samples show overlap with autism and anxiety disorders, but no specific genetic architecture has been established [261, 262, 263]. Whether Mateo's early selectivity belongs to that variation, the first assessment cannot say.

Moving between levels of evidence. A gastric-emptying study establishes what starvation does to a signal; a group comparison describes a distribution of scores during illness; a twin study estimates a population parameter. Connecting any of these to Mateo requires a further inference, made harder by the state he is in when the measurements are taken.

What this case cannot establish

The functional account cannot establish, from Mateo's report, whether his fullness is a changed signal or a changed appraisal, whether his running is a habit or a goal, or whether his rigidity is temperament or weight. It cannot establish the state of any circuit in him, or whether his aunt's illness and his own share anything beyond a family. What it can do is name the variables that a restored nutritional state will make visible, and specify the observations (a blocked run, a supervised meal, a retest after weight gain) that will replace inference with evidence.

Figures for this chapter

Treatment

The order of operations is set by physiology

Mateo's care begins with his heart rate, not his diagnosis. A resting pulse in the mid-forties with a forty-beat rise on standing, hypotension, and cold extremities after losing more than a third of his body weight in ten months places him at or beyond the thresholds that practice guidelines use to indicate a higher level of care: a heart rate approaching 40, orthostatic changes, hypothermia, electrolyte disturbance, and rapid loss with food refusal [96]. The practice guideline cited in this chapter is the 2006 third edition, which a 2023 fourth edition supersedes [299]. Guidance for children and young people adds that a single measure, such as body mass index or duration of illness, should not decide whether treatment is offered [95]. Mateo is admitted for medical stabilization before the label is settled.

The specific hazard of the first week is refeeding syndrome. When carbohydrate returns after prolonged deficit, insulin rises and drives phosphate, potassium, and magnesium into cells; the resulting hypophosphatemia can produce cardiac, respiratory, and neurological complications, and the risk is highest in those who have lost the most weight fastest [93]. Management is monitoring rather than caution for its own sake: daily electrolytes and phosphate in the early days, repletion when levels fall, cardiac monitoring while bradycardic, and a nutritional plan that advances steadily under that surveillance [93]. These practices rest on the physiology of complications and clinical experience, not on randomized trials of outcomes, and that is an honest statement of the evidence rather than a weakness in it.

Running stops during stabilization. This is a medical decision about a bradycardic heart, and also the first test of the functional account: what Mateo does when the run is blocked is an observation the team can record.

Renourishment is the treatment that changes the state

The neuroscience chapter placed starvation in the formulation as a variable that alters interoceptive signals, constrains set-shifting, lowers mood, and prevents the updating of feared expectations. Nutritional rehabilitation acts on that variable directly. Guidelines specify weight-gain targets by setting, on the order of two to three pounds per week in the hospital and a half to one pound per week as an outpatient [96]. For an adolescent the target is not a number but a return to his own growth trajectory, judged against the chart rather than a population cutoff.

Mechanism and evidence line up here more closely than anywhere else in the module. Gastric emptying normalizes with refeeding, so the fullness Mateo appraises as danger becomes, over weeks, a smaller signal [93]. Mood, sleep, and concentration are expected to improve as the state resolves. What renourishment cannot do is unlearn the threat representations or dissolve the habits. An adolescent restored to weight who still believes that fat means going backward has had his state changed and his representations left intact, and relapse is the expected consequence. Weight restoration is therefore necessary and not sufficient, and the second formulation is scheduled for the point at which it has been achieved.

Family-based treatment: managing the state variable through the parents

Family-based treatment for adolescent anorexia nervosa is the approach that most directly operationalizes the state argument. In its first phase the parents take charge of their child's eating, planning and supervising every meal and snack, without requiring the adolescent to agree that anything is wrong. The illness is treated as something that has happened to the adolescent, and the family as the resource rather than the cause. In the second phase, as weight is restored and eating is reliable, control is handed back gradually. The third phase addresses the developmental tasks the illness interrupted [264]. Randomized trials in adolescents have compared this approach with individual therapy and reported an advantage in remission at follow-up [265]. Guidance for children and young people names anorexia-nervosa-focused family therapy as the psychological treatment to consider first [95], and the practice guideline supports family involvement as a central element of adolescent care [96].

The functional rationale is that the treatment does not depend on the capacity the disorder has most degraded. Mateo does not need to believe the referral is anything but an overreaction for Luz and Ray to serve dinner and stay at the table until it is eaten. The supervised meal is also the experiment that avoidance has been preventing: it produces an experienced outcome, and the missing prediction error is finally generated. Nuts are eaten and the slide backward does not begin. Fullness arrives and vomiting does not follow. Such learning requires repetition under conditions in which the outcome can be attributed to the meal, which is what the phase-one structure supplies.

For Mateo's family the frame itself has to change. Luz's tracking app, Ray's remark at dinner, and the coach's advice were offered in good faith inside a frame in which weight loss was health. Family treatment begins by replacing that frame, and the coach is part of the plan because he is part of the environment in which Mateo's eating recovers or does not.

The ARFID-type maintenance needs its own targets

Whichever label is recorded, two of Mateo's maintaining processes are not addressed by weight restoration or parental supervision alone: the sensory selectivity that predates the illness by a decade, and the fear of fullness learned in January. Cognitive-behavioral approaches developed for ARFID use graded exposure organized by the function of the avoidance: new textures for sensory refusal, tolerated fullness for aversive-consequence fear, structured eating for low appetite. The outcome evidence is early: an open proof-of-concept trial of cognitive-behavioral therapy [266] and a small randomized feasibility study of family-based treatment [267], neither of which establishes comparative efficacy. The neuroscience chapter's distinction becomes a schedule. Exposure to yogurt addresses a texture. Exposure to fullness addresses the January representation. Exposure to fat-containing food at a restored weight addresses the representation that belongs to anorexia nervosa. Three exposures, three functions, whichever diagnosis the chart carries.

CBT-E, bulimia nervosa, and what to watch for during refeeding

Enhanced cognitive behavior therapy treats eating disorders transdiagnostically, organizing treatment around the overevaluation of shape and weight, establishing regular eating, weighing in session, and addressing the checking, avoidance, and mood intolerance that maintain restriction and bingeing [100]. In adults with bulimia nervosa, a multicenter trial found cognitive-behavioral therapy superior to interpersonal psychotherapy at the end of treatment, with the difference narrowing by follow-up as the interpersonal group continued to improve [101]. The adolescent evidence for these individual approaches is thinner. For young people with anorexia nervosa, the guideline advises considering individual eating-disorder-focused CBT or adolescent-focused psychotherapy when family therapy is unacceptable, contraindicated, or ineffective [95].

For Mateo these approaches matter now for a specific reason. A binge-eating/purging subtype exists within anorexia nervosa [91], he has a history of secret night eating, and refeeding is a period in which loss-of-control eating can emerge. The team asks again, directly and without alarm, as intake rises. If it emerges, the regular-eating structure and the work on overevaluation that CBT-E specifies join the plan, and the running, if it returns, is watched for its function.

Medication: what it can and cannot do

No medication reliably restores weight in anorexia nervosa, and the practice guideline is explicit that psychotropic medication should not be the sole or primary treatment [96]. Antidepressants have not shown benefit for core symptoms in the underweight state, and the guideline places their consideration after weight restoration, when residual depression or anxiety can be assessed as the person's rather than the starvation's [96]. In a placebo-controlled trial in adult outpatients, olanzapine produced a modest increase in body mass index without significant benefit on psychological symptoms [268]. That result does not establish a benefit for other second-generation antipsychotics or efficacy in children and adolescents. Here mechanism and evidence do not line up: a drug that increases appetite and weight in other populations acts on a system that, in Mateo, is overridden by the threat representation and the habit, and a molecular action does not translate into an outcome for him.

For bulimia nervosa the evidence is different: fluoxetine at higher doses reduces binge and purge frequency and has regulatory approval for that indication, on the basis of adult trials [96]. Whatever is used, cardiac status governs the choice: agents that prolong the QT interval or affect heart rate are avoided while Mateo is bradycardic and his electrolytes are correcting [93].

Measuring what changes

Progress is recorded on tracks that need not move together: weight and vital signs; meals completed under supervision; avoided foods reintroduced, tracked by function; what happens when running is blocked and, later, permitted; frequency of weighing and checking; and, after weight restoration, a repeat of the cognitive and mood assessment that will separate Mateo from the state he was in. Return to running is conditional on weight, vital signs, and the observed function of the running, not on the season.

What this case cannot establish

The plan cannot establish in advance which of Mateo's symptoms will resolve with weight alone, whether family treatment will succeed in a family whose own frame produced the problem, or whether the sensory and fear-of-fullness processes will yield to exposure at the same rate as the fat avoidance. It cannot establish from an outcome literature that mostly concerns girls and women how closely the trial results apply to a fifteen-year-old boy who began from a higher weight. What it can do is act on the physiology first, sequence interventions by the function each targets, and measure each function separately so that the next formulation is written from evidence rather than inference.

Figures for this chapter

---