Attention-Deficit/Hyperactivity Disorder

Case: Tomas, 9. Five chapters follow: the case, epidemiology, presentation, functional neuroscience, and treatment.

The case: Tomas, 9

Presenting concern

Tomas is nine and in fourth grade. His pediatrician's referral reads: "ADHD, combined presentation, diagnosed at 8 on history and parent and teacher rating scales; symptoms since preschool. Started extended-release methylphenidate three months ago with clear morning improvement per teacher. Now irritable and tearful after school, eating little at lunch, weight down about a kilogram. Two-week gap in supply last month when the pharmacy could not fill it. Family asks whether he needs a different medication and whether the school should do more. Mother also reports 'defiance' at homework."

Tomas, seen alone: "I know what I'm supposed to do. Then something else is happening and I'm doing that instead." About afternoons: "Everything is loud and my body is tired but it won't stop." About homework: "Mom sits there and points. I just do the one she points at."

Developmental and family history

Born at term after an uncomplicated pregnancy; his mother smoked until she learned she was pregnant at about eight weeks and then stopped. Milestones were on time; he walked at eleven months and spoke in sentences by two. Preschool teachers called him "always moving" and he could not stay for circle time. His kindergarten teacher raised concerns; his parents "hoped he'd grow out of it." He reads at grade level. Math facts are inconsistent under time pressure, and handwriting is laborious and often erased. He falls asleep around 10:30 pm on school nights and is woken at 6:45; on weekends he sleeps until 9.

He lives with both parents and a sister, twelve, who is doing well. His mother works in a school cafeteria; his father drives for a delivery company, "was the same as a kid," left school at sixteen, and was never assessed. No family history of tics, seizures, or sudden cardiac death. No head injury, hearing or vision problems, or substance exposure. Tomas draws detailed maps of his neighborhood and knows the bus routes by heart.

Observations in session

Seen in the morning, on medication. He sat, then stood, handled the objects on the desk, and answered before questions were finished. Given paper, he drew the route from school to his grandmother's house for twenty minutes with sustained care. Interrupted mid-map with a question about school, he answered, looked at the map, and started a new drawing rather than resuming. Asked to do three things ("put the pen in the cup, close the folder, bring me the paper"), he did the first two, stopped, and said, "What was the last one?" Asked again with the three instructions written on a card, he completed all three.

While his mother spoke he stood up, returned when asked, and interrupted her once to correct a date, accurately. Asked what the pointing at homework does, he said: "When she points I don't have to think about which one, so it's fast."

Collateral

His teacher's log recorded "out of seat" eleven times in one morning. Asked to distinguish them, she noted that most were trips to the sharpener or the window in the middle of an instruction, while two were standing at his own desk to finish a math sheet, which he completed. "Doesn't listen" appears in both parent and teacher reports. His mother means that he does not respond when she calls him from a video game. His teacher means that he looks at her through a three-part instruction and then does only the first part.

During the two weeks without medication, the teacher recorded more call-outs and unfinished work in the mornings; afternoons were "the same as always." His mother reports that the tearfulness occurs on school days around 3:30 and not on weekends, when he is simply restless. Arguments occur only at homework; he is cooperative with chores, at his grandmother's, and with his soccer coach. Parent and teacher rating scales remain elevated on both inattentive and hyperactive-impulsive items. Oppositional items were elevated by his mother only, and only those concerning homework. No aggression, no cruelty to animals, no fire-setting, no stealing. Tomas denies low mood and worries, and says the afternoons are "the worst part of the day."

Epidemiology

How common

A 2023 meta-analysis pooled 61 cross-sectional studies published to October 2020 and estimated the prevalence of ADHD at 7.6% (95% CI 6.1–9.4) among 96,907 children aged 3 to 12, and 5.6% (95% CI 4.8–7.0) among adolescents aged 12 to 18 [54]. The adolescent estimate was higher in studies that applied DSM-5 criteria than in those using earlier criteria, which is a reminder that the diagnostic rule moves the count [54]. In the United States, the NCS-A found a lifetime prevalence of 8.7% among 13- to 18-year-olds by structured interview [53]. These figures bracket Tomas's classroom: in a class of 25, one or two children are likely to meet criteria at some point, and the figure depends on who is asked and by what rule.

When it begins and who is affected

Behavior disorders as a class had a median age of onset of 11 in the NCS-A, later than anxiety and earlier than mood disorders [53]; ADHD itself, by definition, requires several inattentive or hyperactive-impulsive symptoms to have been present before age 12, though the diagnosis need not have been made by then [111, 135], and the sources retrieved for this book do not give a separate onset distribution for it. Boys were more often affected than girls in the NCS-A behavior-disorder class [53]. The meta-analysis does not report a sex ratio in its abstract [54].

The diagnosis is not evenly distributed across populations that reach care. In a health-system cohort of 1,347 transgender and gender-nonconforming youth, attention-deficit disorders had been diagnosed in 15% of transfeminine and 16% of transmasculine 3- to 9-year-olds, three to seven times the rate in cisgender comparison youth [13]. Among children with fetal alcohol spectrum disorders, a review reports ADHD in 50% in a recent study and in 63% to 95% in earlier ones [79]. Neither figure says what proportion of children with ADHD carry those histories; each says that ADHD is common within those groups.

Risk and what the study designs allow

ADHD is highly heritable; a review of prenatal risk factors puts the share of variance attributable to environmental factors at an estimated 10% to 40% [56]. The same review, which concentrated on designs able to account for genetic and familial confounding, concluded that none of the proposed prenatal risk factors could be confirmed as causal and that "the stronger the study design, the less likely it is to support an association" [56]. The often-cited link between maternal smoking in pregnancy and ADHD is a case in point: earlier reviews found the association but insufficient evidence that it is causal [56].

Cohort data illustrate how a risk factor can look different in a different population. In a cohort of children born extremely preterm, elevated maternal pre-pregnancy BMI and pregestational or gestational diabetes were each associated with a 55% to 65% increase in the risk of ADHD at ages 10 and 15, with additive and multiplicative interaction between them, while maternal smoking during pregnancy and hypertensive disorders of pregnancy were not associated with ADHD, which the authors note runs contrary to previous population-based findings [55]. A review of prenatal tobacco and postnatal secondhand smoke describes the literature as consistently associating exposure with behavior problems including ADHD, while acknowledging that prenatal and postnatal roles, genetic confounding and social confounding remain unresolved [80]. Read together, these sources support a single teaching point: the risk-factor literature is largely associational, and the family history that sits in Tomas's chart is a stronger predictor than any single exposure.

Comorbidity and continuity

Nearly one third of children with ADHD have symptoms of conduct disorder, according to a clinical reference [60], and oppositional defiant disorder shares a genetic correlation of about 50% with conduct disorder, higher than its correlation with ADHD [61]. In the Dunedin cohort, 25% to 60% of adults with a disorder had a juvenile history of conduct or oppositional defiant disorder [59]; the cohort does not report ADHD-specific continuity in the abstract retrieved for this book.

Access

Two sources on access were listed but could not be resolved: a 2024 university news story on stimulant shortages amid rising diagnoses [58] and the UpToDate topic on clinical features and diagnosis [57]. For the shortage itself the record is the FDA announcement of October 2022 [297], and for diagnosis rates the CDC surveillance pages [298]; a claim about a current shortage would need current, product-specific data. Neither is cited for a figure here.

What this case cannot establish

Tomas is one child with combined-presentation symptoms, a family history and an unmedicated fortnight. Nothing in the base rates tells us which of the prenatal, genetic or environmental contributions applies to him, and the sections on function and treatment do not need them to.

Source status

#54, #55, #56 and #80 from published abstracts (#56 via the publisher page); #53 from the article text; #59 from the abstract; #60, #61 from the live StatPearls chapters (author lines differ from the references as supplied for both, and the update date for #61); #13 from the health-system news release; #79 from the open-access full text. #57 and #58 could not be resolved to a verifiable record; the FDA shortage announcement [297], the CDC pages [135, 298] and the AAP guideline [136] were added in the citation check of 2026-10-05 as their replacements.

Figures for this chapter

Presentation

Three observations, not one thing

Inattention, hyperactivity, and impulsivity are the three families of behavior the diagnosis is built from. Tomas shows all three. He looks at his teacher through a three-part instruction and does the first part. He stands up while his mother speaks. He answers before a question is finished. Each is a behavior seen from outside. What Tomas perceived, what he made of it, and what he then did are separate observations, and the referral compresses them into "ADHD" in a way that loses the material treatment will act on.

His own account keeps them apart. "I know what I'm supposed to do. Then something else is happening and I'm doing that instead" describes a goal that was represented and then lost to a competitor. "Everything is loud and my body is tired but it won't stop" describes a state, not a decision. "When she points I don't have to think about which one, so it's fast" describes what a support does to a task.

What the diagnosis requires

ADHD is diagnosed on a persistent pattern of inattentive and hyperactive-impulsive symptoms, present for at least six months, beginning before age 12, evident in two or more settings, and producing impairment; the combined presentation requires threshold symptoms in both domains [57]. Tomas meets that description on history since preschool and on parent and teacher ratings. The criteria describe a pattern. They do not require that attention fail uniformly across activities or that every failure originate in one operation, and the observations below show why that matters.

"Doesn't listen": two observations that look alike

Both parents and teacher report that Tomas doesn't listen. They are describing different events.

His mother calls him from a video game and he does not respond. The game supplies visible goals, immediate feedback, and material he values, and it holds him. The teacher gives a three-part instruction while he looks at her, and he does only the first part. He was oriented toward the speaker. Something was heard, since the first step was done. What did not happen was the holding of the second and third steps until they were needed. In the session, the same three-step request failed spoken and succeeded on a card. Not responding because attention is captured elsewhere and not completing because the instruction could not be held are different failures, produce the same word on a report, and call for different help.

"Out of seat": two more

The teacher's log records eleven departures from his seat in a morning. Asked to distinguish them, she found that most were trips to the sharpener or the window in the middle of an instruction, and two were standing at his own desk to finish a math sheet, which he then completed. The first kind interrupts the task; the second kind supports it. Motor activity that accompanies successful work is not the same observation as motor activity that replaces it, and a classroom rule that treats them alike removes one of the conditions under which Tomas finishes his work.

The afternoons

Tomas is irritable and tearful around 3:30 on school days and simply restless on weekend afternoons. The referral raises the possibility of medication rebound. The case offers a comparison but not a controlled one: school days differ from weekends in medication exposure and in six hours of accumulated demand on attention, sitting, and inhibition. Those two explanations are not exclusive, and the pattern during the two weeks without medication is informative: mornings were worse, and afternoons were "the same as always." That observation weighs toward the day's demands, or toward both, rather than toward the drug alone. It does not settle the question, and the treatment chapter returns to it.

Sleep belongs here. He falls asleep at 10:30 and is woken at 6:45 on school days, and sleeps until 9 at weekends. Insufficient sleep is a state that changes access to capacities: in sleep-restriction experiments, school-age children (aged 7 to 11) can show poorer attention and emotional regulation [104, 105]. It is not the cause of a pattern present since preschool, and it may be part of the afternoons.

Homework and "defiance"

His mother reports defiance. The arguments occur at homework and nowhere else: he is cooperative with chores, at his grandmother's, and with his coach, and the oppositional items on the rating scale were endorsed only by his mother and only for homework. Oppositional defiant disorder describes a persistent pattern of angry or irritable mood, argumentative or defiant behavior, or vindictiveness, with severity graded by the number of settings in which it appears [61]. A pattern seen in one setting can still qualify, so the distinction does not rest on setting alone. It rests on what the arguments are bound to: Tomas argues at the one task that requires sustained effort on material he finds hard, with a support that, by his own account, removes the need to decide, and he is not angry, irritable, or vindictive elsewhere. That is task avoidance under a specific load, recorded as such rather than as a second diagnosis. The question stays open: oppositional and conduct symptoms co-occur with ADHD often enough that a third of children with ADHD are described as having conduct symptoms [60], and the pattern will be asked about at each visit. At present there is no aggression, cruelty, fire-setting, or stealing, and no pattern of defiance beyond the desk.

Two other patterns are considered and set aside on present evidence. Mood is not low and there are no worries; the afternoon tearfulness is time-locked and resolves. Handwriting that is laborious and erased, and math facts that fail under time, may reflect a specific learning difference rather than inattention alone; both will be examined directly rather than assumed.

Interest improves access

Tomas drew a map for twenty minutes with sustained care. His mother's and teacher's reports describe a boy who cannot stay with anything. Both are accurate. Interest improves access to sustained activity; it does not remove every difficulty with regulating that activity. Interrupted mid-map, Tomas did not resume; he started a new drawing. Even in the activity he can hold, the return to an interrupted plan was not available. His engagement with maps is a real ability whose conditions explain why it cannot serve as an estimate of his attention in fourth-grade math.

Observation	What Tomas perceives	What he interprets	What he does	What it points to
Called from a video game	Little beyond the game	Nothing to respond to	Does not answer	Attention captured by a supported activity
Three-part spoken instruction	The teacher, the words	The first step	Does step one, stops	Steps not held; succeeds with a card
Sharpener mid-instruction	A competitor	Something else to do	Leaves the task	Goal weakly maintained
Stands at desk during math	The sheet	The same task	Finishes it	Motor activity that supports work
Homework with mother pointing	The pointed problem	"The one she points at"	Completes it fast	Decision supplied, not knowledge retrieved
3:30 on a school day	Noise, fatigue	"It won't stop"	Tearful, irritable	State: demand, medication, sleep, or all three

Where he sits in a developmental course

Tomas's symptoms began before school and have been continuous. Prospective follow-back of a birth cohort found that most adults with a psychiatric disorder had met criteria for a diagnosable disorder by adolescence, with childhood conduct and hyperactivity among the juvenile diagnoses that preceded a range of adult outcomes [59]. That is a statement about a population's history, not a forecast for Tomas. It is a reason to treat the ADHD and to watch for the oppositional pattern rather than to expect either. His father's account of himself ("the same as a kid," never assessed) is part of the family history and part of how the family explains Tomas.

What this case cannot establish

The case cannot establish whether the afternoon tearfulness is medication rebound, accumulated demand, insufficient sleep, or some combination; whether the homework arguments are confined to homework because of the task or because of the pointing; whether his handwriting and math-fact difficulties are secondary to inattention or a separate learning difference; whether his father has ADHD; or whether Tomas's maps show a preserved capacity for sustained attention or a highly supported one. Each observation was made under particular conditions, and the meaning of each behavior was obtained by asking, not by watching.

Figures for this chapter

Functional neuroscience

A framework borrowed from the Intellectual Disability module (Maya)

The functional account of Maya's classroom separated representing a task, keeping the right information usable, learning, and coordinating a plan with change. Tomas has none of Maya's broad cognitive limitation, and the same operations organize his difficulties, in the same four kinds: Capacity (what he can do under good conditions), Process (an operation that runs on something), State (a condition that changes access to capacities), and Representation (what is held). The distinction that matters most for Tomas is between a capacity and access to it. He can hold three instructions; on the card he did. In the classroom, spoken, he held one.

Functional layer

Attentional selection and distractor suppression

Attentional selection, a Process, selects information from competing inputs; distractor suppression, a related Process, reduces the processing of irrelevant input. Relevance is defined by the child's present goal, and a goal that is weakly maintained cannot define it. Tomas's sharpener trips occur in the middle of an instruction: a salient event redirects behavior before the activity is complete. His own description fits this level exactly. "I know what I'm supposed to do" is a represented goal; "then something else is happening and I'm doing that instead" is selection reassigned to a competitor while the goal was still, in some sense, known.

Answering before a question is finished belongs here too, though it may involve more than selection. An action prepared on partial input has been released before the input was complete. Whether that reflects a separable inhibitory operation, as one account of ADHD proposes [106], or the same weak goal maintenance operating on output rather than input, cannot be decided from the observation. The case shows the behavior; the operation is a hypothesis.

Working memory: maintenance, updating, and the card

Working-memory maintenance, a Process, preserves information in an actively usable state over a brief interval. It is needed when the information required for the next action is no longer in front of the child. Three spoken instructions must be held while the first is carried out; by the second, the third has to survive the interference of acting. Tomas completed two and asked for the last. With the instructions on a card, he completed all three. The card did not add capacity. It moved part of the burden from internal maintenance to perception and recognition, and the operations that remained were ones he could do. External support changes the cognitive task; it does not make cognition disappear.

That result places his classroom failure. When the teacher's three-part instruction produces only the first step, the first step was represented and the rest was not held. This is not the same failure as the video game, where nothing was held because attention was captured before selection could occur. Working-memory updating, replacing current content when new information becomes relevant, is visible in the interrupted map: he did not resume, he started again. The representation of the plan in progress was displaced by the question and not restored.

The effect of demand can be abrupt, and the demands interact: an error draws attention from the instruction, correction adds language, and the sharpener is available. In an experiment with 15 children with ADHD and 14 comparison children, Kofler and colleagues [320] manipulated working-memory demands and found that observed inattention varied with executive demands and with demands exceeding storage or rehearsal capacity. That small study gives a concrete reason to examine what a child must hold and do when interpreting visible inattention.

Learning: retrieval under time and what practice buys

Memory encoding, a Process, establishes a retrievable representation; memory retrieval makes it available for present use. Tomas's math facts are inconsistent under time. The facts were encoded, since he produces them untimed. What fails is retrieval under a condition that adds its own demand, and a timed sheet also removes the option of working the answer out. Whether the retrieval itself is slow, or attention leaves the sheet while retrieval is under way, is not established by the observation. His handwriting raises the same question from the motor side: laborious, erased, and a load in its own right on every written task.

Practice reduces the number of decisions that must be made anew. The bus routes and maps are what practice on valued material produces: a coordinated structure that needs little active maintenance. The map is not evidence of intact attention in general; it is evidence of what a highly practiced, self-paced, visually present task requires of him, which is less.

Homework with his mother pointing is a learning observation as well as a support observation. "I just do the one she points at" describes a decision supplied rather than knowledge retrieved. He completes the sheet. What he has learned about choosing where to begin, or about the problem itself, is not shown by the completed sheet, and it is the question the treatment chapter puts to that arrangement. Working-memory training has been tested for whether practice transfers: Roberts and colleagues [302] randomized 452 children with low working memory to a training program or usual teaching, and an early visuospatial short-term-memory benefit was absent at two years with no broad academic benefit established. Practice at holding things does not, on that evidence, generalize to holding the teacher's instruction.

Flexible coordination: the interrupted map and the end of the day

Task switching, a Process, changes the active rule or goal set. Tomas switches readily in one direction, away from the current task toward a competitor, and poorly in the other, back to an interrupted plan. Both are switches; they differ in what drives them. The first is driven by the competitor's salience, the second requires a stored representation of where he was. "Difficulty with transitions" names the situation and does not specify the operation.

By 3:30 on a school day he has spent six hours selecting, holding, and returning under conditions that supply little external support. "Everything is loud and my body is tired but it won't stop" is a report of a State. Whether it is produced by medication exposure ending, by the accumulated demand, by short sleep, or by all three cannot be read from the report, and the two unmedicated weeks, when afternoons were "the same as always," weigh against exposure alone. A State changes access to every capacity above; it is not a fourth symptom cluster.

Construct	Kind	What it names in Tomas
Attentional selection / distractor suppression	Process	The sharpener mid-instruction; the video game he cannot be called from
Working-memory maintenance	Process	Two of three spoken steps; three with a card
Working-memory updating	Process	The map not resumed after the question
The goal in progress	Representation	"I know what I'm supposed to do"
Memory retrieval	Process	Math facts untimed but not timed
Task switching	Process	Easily away, poorly back
Sustained attention on practiced, self-paced material	Capacity	Twenty minutes on the map
Afternoon fatigue; short sleep; medication exposure	State	"It won't stop" at 3:30

Circuits, development, and genes

Keeping a representation active

Keeping the instruction usable recruits systems for maintenance and control; acting on it involves motor and subcortical loops. Funahashi and colleagues [301] recorded prefrontal neurons in monkeys whose location-sensitive activity persisted while a remembered location was held across a delay: a biological example of information remaining available to guide an action after the stimulus has gone. Recurrent excitation lets a network sustain its own activity, and neuromodulators change how strongly that activity influences subsequent processing. Wang and colleagues [306] showed in prefrontal models that alpha-2A adrenergic stimulation reduced cAMP-related opening of HCN channels and strengthened task-related network activity. A molecular event can change the effectiveness of communication within an already learned network, on a timescale far shorter than learning; modulating access to a representation differs from acquiring it. Gamo and colleagues [307] found that methylphenidate and atomoxetine improved working-memory performance in monkeys within a dose range and through adrenergic and dopamine receptor actions, with starting state, dose, and task all influencing the benefit. Whether a related action mediates Tomas's morning improvement is a further question, and his teacher's log is not an answer to it.

Development

Klingberg and colleagues [304] related visuospatial working-memory performance and age to frontal and parietal activity across childhood and adolescence, connecting improving performance with the systems engaged. Shaw and colleagues [305] analyzed repeated scans from children with ADHD and comparison participants and found later attainment of modeled peak cortical thickness across much of the cerebrum in the ADHD group: an aspect of ADHD-associated structural variation located in developmental timing. It is a group comparison. It does not say where Tomas's cortex is, and it does not license "his brain is two years behind" as a formulation.

Genes and prenatal exposures

ADHD is highly heritable, with an estimated 10 to 40 percent of variance attributed to environmental factors [56]. Demontis and colleagues [310] identified 27 common-variant loci, and their rare-variant study [311] implicated MAP1A, ANO8, and ANK2, connecting risk to neuronal biology beyond a simple dopamine-deficiency story. These associations identify liability at the population level. His father's account of himself is consistent with familial liability and does not establish it; he was never assessed.

Tomas's mother smoked for the first eight weeks of pregnancy. Maternal smoking in pregnancy has repeatedly been associated with ADHD, but a review emphasizing designs that account for genetic and familial confounding concluded that none of the proposed prenatal risk factors can be confirmed as causal, and that the stronger the study design, the less likely it is to support an association [56]. In an extremely-low-gestational-age cohort, elevated pre-pregnancy maternal BMI and diabetes were each associated with increased ADHD risk, while maternal smoking during pregnancy was not, contrary to earlier population studies [55]; that cohort is not Tomas's, who was born at term. The exposure is recorded in his history. It is not offered to his mother as the cause, because the evidence does not support that inference, and because the inference would do her harm without helping him.

Moving between levels of evidence

A task experiment tests specified operations under controlled conditions. A developmental group comparison describes a distribution across children. A genetic or prenatal association identifies liability at the population level, and its strength depends on how well the design handles confounding. A treatment trial estimates benefits and harms on its measured outcomes. Connecting any of these to Tomas requires an additional inference. Neither an imaging trajectory nor a risk association by itself identifies his functional difficulty or his preferred medication.

What this case cannot establish

The case cannot establish whether Tomas's answering before the question is a separate inhibitory failure or the same weak goal maintenance seen at the sharpener; whether his maintenance limit is a capacity limit or an access limit under classroom conditions, since the card test was done once, in a quiet room, on medication; whether his timed math failure is slow retrieval or attention leaving the sheet; whether the afternoon state is exposure, demand, or sleep; or whether any imaging or genetic finding above applies to him. No operation was measured. Each was inferred from what Tomas did under described conditions, and a behavior that fits a construct does not demonstrate the mechanism it is named after.

Figures for this chapter

Treatment

What Tomas and his family want

Tomas wants the afternoons to stop being "the worst part of the day." His mother wants homework without a fight and asks whether she is doing it wrong. His teacher wants him in his seat, and on questioning wants him to finish his work, which is not the same thing. These goals can be counted: tearful afternoons per week, homework sheets completed without pointing, instructions followed to the third step, math sheets finished. Symptom ratings are kept; they are not the goal.

Assessment before changing anything

The referral asks about a different medication. Before that question is answered, the conditions under which the present one is being judged are examined. Weight and height are plotted against his prior curve; a kilogram in three months is a change to act on, and appetite at lunch is asked about directly. Blood pressure and pulse are recorded. Sleep is measured rather than estimated, because time in bed can overestimate actual sleep. Even as scheduled, 10:30 to 6:45 gives 8 hours 15 minutes. That is below the 9 to 12 hours recommended for ages 6 to 12, and short sleep is a State that, in experiments, impairs attention and emotional regulation in school-age children [104, 105, 107]. Handwriting and timed math are screened for a specific learning difference rather than assumed to be inattention. Rating scales from both settings are repeated, and the teacher's log is changed so that it separates departures that replace the task from standing that accompanies it.

The two unmedicated weeks are treated as data with known limits. Mornings were worse without medication; afternoons were unchanged. That pattern was not planned, not blinded, and not counted with a consistent instrument. It still weighs against the afternoon being rebound alone, and it is the kind of observation a planned comparison could sharpen.

Does the support make knowledge available or supply the decision?

The distinction drawn in the Intellectual Disability module (Maya) applies directly. When Tomas's mother points to the next problem, she supplies the decision. The sheet is finished quickly and he has practiced nothing about choosing where to start or returning after losing the task; "I just do the one she points at." A numbered sequence on the sheet, or a card that says "find the next unfinished one," makes the same information available while leaving the decision to him. It will be slower at first. The change is made by agreement, with his mother in the room, doing less. When a cue makes knowledge available, it supports participation while he practices the routine; when an adult supplies the answer, later opportunities to choose show whether the routine has been learned. Both are useful; only the first builds toward homework without her.

The card in the session is the model for the classroom. Instructions are given in writing or one step at a time, because the failure was in holding steps, not in hearing them. The two kinds of out-of-seat are treated differently: standing at his own desk to finish a sheet is allowed, because it accompanied completed work, and trips mid-instruction are met with a redirection that names the step he is on. The interrupted map suggests a further support: a visible marker of where he was when a task is interrupted, so that returning does not require reconstructing the plan.

Parent and classroom behavioral approaches

Behavioral parent training and classroom behavior management are established components of ADHD treatment, with trial evidence for improvement in parent-rated behavior and classroom functioning and smaller effects on core symptom ratings than medication [108, 109]. In the source chapter, in 128 children with autism and ADHD, structured parent training improved ADHD symptoms relative to placebo, as did atomoxetine, in a design that compared neither with a stimulant [316]. The population differs from Tomas's; what transfers is that a parent-mediated intervention changed measured ADHD symptoms under controlled conditions. For his mother, the training is specific: what to do at homework instead of pointing, how to give an instruction he can hold, and how to respond to the argument that follows the harder arrangement without it becoming the pattern the ODD items describe.

Stimulant medication

Methylphenidate inhibits dopamine and norepinephrine reuptake. Its pharmacology offers a way to alter signaling in systems supporting goal-directed activity; it does not establish that Tomas has too little dopamine, and its exact therapeutic mechanism in ADHD remains incompletely established [317]. Short-term stimulant efficacy for ADHD symptoms in children and adolescents is supported by a large trial literature, summarized in a network meta-analysis [110]. In the source chapter, Simonoff and colleagues [313] found methylphenidate superior to placebo in children with severe ADHD and intellectual disability, a population that differs from Tomas's. Gamo and colleagues [307] found that methylphenidate improved monkeys' working-memory performance within a dose range through specified receptor actions; that does not show that Tomas's morning improvement was produced by better working memory, and the teacher's log cannot show it either.

For Tomas, the morning benefit is real and counted. The three problems are appetite and weight, the afternoon, and supply. A formulation and schedule should cover the activities that matter while allowing assessment of appetite, sleep, cardiovascular measures, mood, and growth [317]. Breakfast before the dose and a substantial meal after school are arranged first; if weight continues to fall, dose and formulation are reconsidered before the drug is abandoned.

The afternoon

An apparent afternoon decline needs to be considered against exposure timing and the day's accumulated demands. Three accounts are open: the drug wearing off, six hours of demand, and short sleep. The unmedicated weeks weigh against the first alone. The plan tests rather than assumes. Sleep is extended first, by moving bedtime earlier in steps, because it is the cheapest change and bears on all three accounts. The after-school hour is restructured so that homework does not begin at 3:30. If tearfulness persists with adequate sleep and a supported afternoon, formulation and timing are adjusted and the log records whether the afternoon changes.

Guanfacine and atomoxetine

Guanfacine provides a different approach through alpha-2A receptor agonism. Its preclinical rationale is the strengthening of prefrontal network activity shown by Wang and colleagues [306]. In an eight-week trial in children with autism, 15 of 30 on extended-release guanfacine versus 3 of 32 on placebo were rated much or very much improved, while a brief working-memory and motor-planning battery showed no group difference [315]. The trial's population is not Tomas's. Its lesson is general: behavioral improvement was demonstrated without benefit on the cognitive measures, so efficacy cannot be treated as confirmation that improved working memory mediated the response, and the null result leaves open effects those measures did not detect. If guanfacine were added for the afternoons or substituted, the outcome to watch is participation, not a presumed mechanism, and sedation can reduce visible activity without improving it. Blood pressure and pulse are assessed before treatment, after dose increases, and periodically; the extended-release preparation is tapered rather than stopped, in decrements no greater than 1 mg every three to seven days [318].

Atomoxetine, a norepinephrine-transporter inhibitor, is a further option if a stimulant is not tolerated. The Handen trial found it effective against placebo and did not compare it with a stimulant [316]. Exposure depends on CYP2D6 activity and on strong CYP2D6 inhibitors; labeling requires attention to suicidal ideation, cardiovascular effects, and symptoms suggesting liver injury [319]. A liver enzyme is not a psychiatric formulation. The choice among options considers previous response, eating and sleeping patterns, baseline alertness, cardiovascular history, formulation practicality, and the hours of needed support; neuroscience explains why these factors can matter, and comparative clinical evidence establishes how much confidence to place in each option.

Access

The two-week gap was not a decision; the pharmacy could not fill the prescription. Shortages of stimulant medication have affected availability in the United States as diagnoses have risen [58]. The plan includes a refill schedule that anticipates the gap, a named alternative formulation the prescriber will accept, and an agreement with the school on what changes when he is unmedicated, so that an interruption in supply does not become an interruption in support.

Watching for what may follow

The homework arguments are recorded and revisited. A third of children with ADHD are described as having conduct symptoms [60], and oppositional defiant disorder is diagnosed on a persistent pattern that can begin in one setting [61]. Prospective follow-back has shown that most adults with a psychiatric disorder had a diagnosable disorder by adolescence, with hyperactivity and conduct problems among the juvenile diagnoses preceding a range of outcomes [59]. None of this forecasts Tomas. It is the reason the homework plan is designed to end the argument rather than win it, and the reason his cooperation at his grandmother's and with his coach is written down as a baseline.

Measuring what changes

Tearful afternoons, sheets completed without pointing, third steps followed, and math sheets finished are counted separately, because they will change at different rates and for different reasons. Weight, appetite, sleep duration, blood pressure, and pulse are recorded at each visit; the teacher's log and both rating scales are reviewed. Tomas's own criterion, whether the afternoons are still the worst part of the day, is asked each time.

What this case cannot establish

The case cannot establish whether the afternoon will respond to sleep, to the school day, or to the prescription; whether removing the pointing will produce independent homework or a longer argument; whether a change of medication would help him, since the trials that describe the alternatives were run in populations that are not his; or whether his morning improvement reflects the mechanism the drug was chosen for. A response to treatment does not demonstrate the mechanism, and an absence of response does not show the functional account was wrong. The account offers a way to arrange the next three months so that what changes can be seen and counted.

Figures for this chapter

---