Disruptive Mood Dysregulation Disorder

Case: Owen, 8. Five chapters follow: the case, epidemiology, presentation, functional neuroscience, and treatment.

The case: Owen, 8

Presenting concern

Owen is eight, in third grade. The school psychologist's referral reads: "Outbursts two to three times a week for over a year: screaming, throwing, has hit a classmate and an aide. Sent home four times this semester. Irritable most of the day between outbursts. Behavior chart and time-outs have not worked. Parents report the same at home, worse. A paternal uncle has bipolar disorder. Please evaluate for mood disorder versus ADHD or ODD."

Owen, seen alone, said: "I get mad really fast and then I can't stop. My body gets too hot. Everyone thinks I do it on purpose." Asked about a good day: "Saturday, when nobody tells me stuff."

Developmental and family history

Born at term; a late talker who caught up by kindergarten. An "intense" baby, hard to soothe; toddler tantrums that outlasted his peers'. By four his preschool called him "our storm cloud." First grade: "can't sit still, doesn't finish, loses everything," no evaluation. He sleeps eight to nine hours, resists bedtime, falls asleep fast. Mondays after weekends at his grandmother's, where bedtime is eleven, are the worst days.

His mother had postpartum depression after Owen. His father: "I was like him and grew out of it." The father's brother has bipolar I disorder with two hospitalizations. A sister is four. No trauma or violence in the home; the parents argue about how to handle him.

The outbursts

His mother's four-week log records fourteen outbursts at home and nine at school. Triggers: the tablet ending, homework, his sister touching his things, losing a game, "no." Typically he goes red, screams, and throws whatever is near; episodes last twenty to forty-five minutes. He has hit his mother, kicked a door off its hinge, banged his head on the floor twice, and bitten his own arm once. He cannot settle unless an adult sits nearby without talking. Afterward he is exhausted and tearful: "I'm the worst kid in the world." Between outbursts he is "grumpy," snaps at his sister, says everything is unfair, and is "happy maybe an hour a day"; school says the same. No three-month stretch has been better. His sister's tantrums last five minutes, involve stomping, and end on their own.

Two entries in the log read "meltdown." On Tuesday the tablet was taken at the end of screen time; Owen was red and screaming within seconds, hit his mother, and banged his head; forty minutes. He said later, "I didn't decide to." On Saturday, in a grocery store, he wanted candy; his volume rose over ten minutes while he watched his mother's face, and he stopped within a minute of getting it. In line at school last week a classmate bumped him, by the teacher's account accidentally, and Owen hit him: "He did it on purpose." At a birthday party last month he was loud and silly for three hours, "bouncing off the walls," and slept normally that night; his father wonders whether that is "my brother's thing."

Observations in session

Owen left his chair four times, handled everything on the desk, interrupted, and answered well. He built a Lego structure with absorption. Told there were five minutes left, he scowled, said "that's not fair," and flicked a piece across the table; given a warning and a job, he put the pieces away with the clinician. Afterward, he says, he feels "bad, like really bad." He denied feeling sad most days: "just mad." He denied wanting to die; his mother reports one "I wish I was never born" after an outburst. He warmed over the hour; speech was normal in rate. Parent and teacher ADHD ratings were elevated. Oppositional items endorsed: argues, defies, easily annoyed, angry; not spiteful or vindictive.

Collateral

His teacher: "He is a different kid one-on-one. In a group he's on edge from the moment he walks in. He's not defiant so much as he can't take a no." He does not finish work, loses homework, and "daydreams." School outbursts last fifteen to twenty minutes and end in the counselor's office; he threw a chair once, never seriously hurting anyone. An after-hours clinic saw him after the head banging; no injury. His father is wary of medication "because of my brother." His mother: "I can't do another year of this."

Epidemiology

How common depends on the rule

DMDD was written into DSM-5 in 2013 [295], and its prevalence has moved with every operationalization since. Community studies using interviews have reported rates of 3% to 8% in young children and under 1% in older children; one rating-scale study found 9.3% at a mean age of 8.7 and 5.9% in the same children at 16.4 [7]. A meta-analysis of interview-diagnosed depressive disorders in children under 13 estimated DMDD at 1.60% (95% CI 0.28–3.90), an estimate the authors note is limited by the small number of studies [8]. In the 2004 Pelotas Birth Cohort in Brazil, where 3,562 children aged 10 to 12 were assessed with a structured interview, the prevalence by strict DSM-5 criteria was 2.5% (95% CI 2.0–3.0) [9]. When the same authors applied a proposed operationalization (two of seven irritable-mood items, three of eight outburst items, combined by an OR rule [9], with an impairment-in-two-settings requirement whose exact operationalization is in the paper's tables), the figure was 3.0%, split into irritable mood only 1.12%, outbursts only 0.64%, and both 1.23% [9]. Irritable mood and outbursts explained nonoverlapping aspects of impairment [9], which is the epidemiological version of the distinction the case chapters draw between Owen's baseline and his episodes.

Owen is eight, inside the age band where the highest rates are reported and below the age-10 onset ceiling the criteria impose; the diagnosis is not made before six [7].

Stability over time

Whatever the rate at one assessment, the diagnosis is unstable across assessments. One clinical sample went from 100% to 49% to 40% meeting criteria at baseline, two and four years; in a longitudinal study of youth with elevated manic symptoms, the proportions were 52%, 29% and 19% across assessments; and in the Great Smoky Mountains Study, a lifetime rate of 3.3% for the precursor construct of severe mood dysregulation fell to 0.4% when persistence across two waves was required [7]. The base rate for "DMDD that persists" is therefore much lower than the base rate for "DMDD at one visit."

Comorbidity

Comorbidity with oppositional defiant disorder is described as extremely high, and in clinical samples the combination of ADHD and ODD co-occurs in more than 75% of children with DMDD [7]. This overlap is why the presentation chapter treats ODD and ADHD as differentials that are usually also present rather than alternatives.

What it becomes

In adult follow-up of children who met criteria for the construct, rates of anxiety and depression were significantly higher than in a psychiatric comparison group, and 36.1% had more than two disorders versus 8.8% of comparisons; bipolar disorder was not among the elevated outcomes [7]. This outcome pattern is the empirical basis for the separation of chronic irritability from pediatric bipolar disorder, a debate the bipolar chapter takes up [66].

Tantrums as a measurable phenomenon

The outburst criterion rests on distinguishing clinically significant tantrums from ordinary ones. Among 279 preschoolers aged 3 to 6 assessed by caregiver interview, children with disruptive disorders were five times more likely than healthy or depressed children to have violently aggressive or destructive tantrums, and children with both depression and a disruptive disorder nine times more likely than healthy children; mean tantrum duration was 20 minutes 30 seconds in the comorbid group versus 11 minutes 48 seconds in healthy children, and self-injurious tantrum behavior was more frequent in the depressed groups [10]. The authors identified concerning styles: aggression toward others or objects in half or more of the last 10 to 20 tantrums, any self-injury, 10 to 20 tantrums on separate days at home in 30 days or more than five a day at school, an average duration near 25 minutes, and inability to self-calm [10]. These cut points come from preschoolers; the presentation chapter uses the dimensions, not the thresholds, for an eight-year-old.

What this case cannot establish

Owen's rate of meeting criteria at one visit is the 2.5% to 3% kind of figure; whether he is in the persistent fraction is a question only follow-up answers, and the stability data say most children in his position will not be.

Source status

#7 from open-access full text; #8 from the published abstract; #9 from open-access full text; #10 from the author-archived full article; #66 from the publisher's abstract.

Figures for this chapter

Presentation

Aggression, anger, and irritability as three observations

Aggression is behavior directed toward harming or threatening another person, or, in broader clinical descriptions, damaging objects. Anger is an emotional experience. Irritability refers to a lowered threshold for anger or frustration. The three often occur together, but each requires its own description. Owen hits his mother (aggression), says his body gets "too hot" (anger), and is "grumpy" and "happy maybe an hour a day" between episodes (irritability). A child may feel intense anger without acting aggressively; another may use a threat to obtain something while feeling little anger; an irritable child may have frequent verbal outbursts and no physical aggression. Owen has all three, and they are recorded separately because they change separately.

The referral says "outbursts." The interview asks what Owen perceived, what he made of it, and what he did. When a classmate bumped him in line, he perceived contact, interpreted intent ("he did it on purpose"), and hit. The teacher's account of the bump is part of the record, not a correction of Owen's, because assessment investigates the actual events while recognizing that his action was organized around the event as he represented it.

The mood between the outbursts

DMDD is defined by two things that are easy to run together: severe recurrent temper outbursts, and a persistently irritable or angry mood between them. The current criteria require outbursts grossly out of proportion to the provocation and inconsistent with developmental level, three or more times a week on average; irritable or angry mood most of the day, nearly every day, observable by others; twelve months or more without a symptom-free stretch of three months; presence in at least two of home, school, and peers, with severity in at least one; age six or older at diagnosis; onset before ten; no manic or hypomanic episode lasting more than a day; and no better explanation by another disorder, with ODD and bipolar disorder excluded by hierarchy [111].

Owen's log gives the frequency and the settings. The mood between outbursts is the observation that a log of outbursts does not contain, and it is obtained separately: "storm cloud" at four, "grumpy" now, "on edge from the moment he walks in," "happy maybe an hour a day," with school and home agreeing. In a population cohort, irritable mood and temper outbursts each explained impairment that the other did not, which supports recording them as two dimensions rather than one syndrome inferred from its louder half [9]. The "storm cloud" at four also dates onset well before ten. The uncle's history and the birthday party are the reasons the mood is characterized as chronic and nonepisodic rather than assumed to be.

The outburst as a specified event

An episode has dimensions that disappear inside a frequency count. Identify the target, the force or potential harm, whether an object was used, the opportunity to disengage, and the time needed for recovery. A reduction in the number of events accompanied by more severe injury would not be a straightforward improvement, and the same number of disagreements with earlier recovery and no physical action might be a substantial change.

Tantrums are common in healthy young children. In a study of preschoolers, the behaviors that distinguished children with depressive or disruptive disorders from healthy children were not the presence of tantrums but what happened inside them: aggression toward caregivers or objects, self-injury, tantrums that were long, tantrums that were frequent, and inability to calm down without help [10]. That sample was aged three to six, so the specific cut points do not transfer to an eight-year-old; the dimensions do. Owen's episodes are long (twenty to forty-five minutes at home), frequent (three to four a week at home), involve aggression toward his mother and objects, include self-injury (head banging, biting his arm), and end only when an adult sits with him. His sister, at four, supplies the normative comparison in the same house: five minutes, stomping, self-resolving.

"Cannot calm down" is recorded with its condition. Owen settles when an adult sits nearby without talking. He does not settle alone. Both are true and they name different things: the capacity to return to baseline exists and is reachable through another person; access to it under his own control, at the peak, is what is missing.

Reactive and instrumental outbursts

Reactive and instrumental aggression are useful descriptions of episodes. Reactive aggression follows a perceived provocation, threat, or frustration. Instrumental aggression is used in pursuit of a goal. They are not permanent kinds of children. Two log entries read "meltdown." On Tuesday the tablet was taken, and within seconds Owen was red, screaming, hitting, and banging his head; "I didn't decide to." On Saturday, in the store, his volume rose over ten minutes while he watched his mother's face and stopped within a minute of getting the candy. The first has the time course of a reaction; the second has the time course of a negotiation. Emotional activation and expected gain can both influence the same action, and their relative contributions vary between episodes. A child may lash out in anger and subsequently learn that the episode ends an unwanted task; the same child may escalate because escalation previously changed a parent's behavior. Recording which one is happening is what makes the parents' plan specific.

Differentials

Normative tantrums. The sister's tantrums are the reference: short, non-aggressive, self-terminating, and diminishing with age. Owen's differ on every marked dimension [10]. Age matters in the other direction too: DMDD is not diagnosed before six; the floor helps separate pathological irritability from earlier developmental tantrums, though at younger ages severity, frequency, and impairment still matter [134, 111].

Oppositional defiant disorder. ODD comprises angry or irritable mood, argumentative or defiant behavior, and vindictiveness, persisting for at least six months and shown with at least one person who is not a sibling [61]. Owen argues, defies, and is easily annoyed and angry; he is not spiteful, and his teacher's "he can't take a no" rather than "won't" is an observation the ODD label does not capture. Most children who meet DMDD criteria also meet ODD criteria, the two are separated more by the severity and chronicity of the irritable mood than by kind, and the diagnosis has shown instability over follow-up [7]. The manual's hierarchy assigns DMDD when both are met [111]. The label chosen matters less than the two observations the label rests on: how angry he is between episodes, and what happens inside them.

Attention-deficit/hyperactivity disorder. Owen left his chair four times, handled everything, interrupted, does not finish work, loses homework, and "daydreams," with parent and teacher ratings elevated in both settings. This is a separate pattern with its own criteria [135, 136], it commonly co-occurs with chronic irritability [7], and it changes what a "no" costs him: a child who cannot hold the next step in mind experiences an interruption as a loss of the whole task. Behavior disorders have a median onset in late childhood, earlier than mood disorders [53]; Owen's inattention was noted in first grade, before anyone called his mood a problem.

Bipolar disorder. The uncle and the birthday party raise the question. Activated mood states are identified by a coordinated change from the child's own baseline over a sustained interval: elevated or irritable mood with accelerated thought and speech, increased goal-directed activity, altered judgment, and reduced need for sleep [340]. Three loud hours at a party followed by a normal night is a context-bound state that ended when the context did. The "narrow phenotype" of pediatric bipolar disorder requires episodes of elevated mood; whether chronic, nonepisodic irritability should be counted as bipolar was the central debate of the previous two decades, and the youth with chronic irritability in those studies differed from youth with episodic mania in family history and in longitudinal outcome, developing depression and anxiety rather than mania [66, 7]. DMDD was introduced to give those children a diagnosis that did not commit them to a bipolar course. Owen's mood is chronic; nothing in his history is episodic. A family history raises a prior and is recorded; it does not make a party an episode.

Depression. "I wish I was never born," said once after an outburst, is asked about directly and is not folded into the irritability. Owen denies sadness most days: "just mad." The mood is characterized separately from the outbursts because chronic irritability in childhood predicts depressive and anxiety disorders in adulthood [7], and because adult depression in a birth cohort was frequently preceded by juvenile diagnoses of a different kind, including oppositional and conduct disorders [59].

Observation	What Owen perceives	What he interprets	What he does	What it points to
Tablet taken	The activity ending	Loss, unfairness	Red within seconds; hits; head banging; forty minutes	Reactive outburst; cannot self-calm
Grocery store	Candy in view	"If I get louder, I get it"	Escalates over ten minutes, watching her face; stops on receipt	Instrumental escalation; learned
Bump in line	Contact	"He did it on purpose"	Hits	Hostile attribution; reactive aggression
Lego, five minutes left	Activity ending	"Not fair"	Flicks a piece; puts pieces away with a job	Same trigger class; recovery with structure
Birthday party	Noise, other children	Fun	Loud and silly for three hours; sleeps normally	Context-bound, not an episode
Monday after grandmother's	Ordinary demands	Everything is worse	More outbursts	Sleep as a State that lowers the threshold

What this case cannot establish

One interview, a four-week log, and a teacher's letter cannot establish whether Owen's irritable mood is one condition with the ADHD or two; whether the store escalation began as reaction and became negotiation or was negotiation from the start; whether the head banging is self-injury in the sense that predicts later harm or a feature of overwhelmed motor discharge; whether the uncle's illness signals a course that has not declared itself; or what "I wish I was never born" means to an eight-year-old on the far side of forty minutes of screaming. The parents' disagreement about handling him is a fact about the household, not yet a cause of anything. The meaning of each behavior was obtained by asking, and the answers are samples.

Figures for this chapter

Functional neuroscience

Frustration is an event with a history

When the tablet ends, the event is not simply the absence of pleasure. Owen has organized the previous half hour around an expected outcome, and the interruption violates a prediction, changes the value of what came before, and demands a new decision at the moment he is least able to make one. What an observer calls frustration is prediction, valuation, and interpersonal meaning converging in a few seconds. This chapter separates those strands and names each as a Capacity, Process, State, or Representation.

Functional layer

Expectation, omission, and frustrative nonreward

Prediction error is the discrepancy between an expected quantity and what is observed. In a deliberately simple example, let the expected outcome be five units and the received outcome zero. The signed error is minus five. The units are illustrative; they do not measure Owen's anger. The computation explains why an omitted expected reward differs from a reward that was never expected, and why the same zero can carry different information on two occasions: what changes is the prediction against which it is compared. A tablet that ends after a two-minute warning and a tablet that ends mid-game deliver the same zero against different predictions.

A compact learning rule writes the next expectation as the old expectation plus a fraction of the error: V(next) = V(now) + α × [outcome − V(now)]. The coefficient α determines how strongly a new outcome changes the represented estimate. Neither a high nor a low value is universally good; rapid learning helps in a changing environment, slower updating protects a useful expectation from one unrepresentative event. The equation is an explanatory model, not a clinical score obtained by asking how upset Owen became.

The phrase frustrative nonreward usefully identifies the event class, which in Owen's log is an expected reward withheld or removed. The explanatory work lies in showing what the child expected, what changed, and what that change did to subsequent processing. Describing Owen as "highly reward-sensitive" leaves open whether the main issue is strong anticipation, the aversiveness of omission, the interpretation of unfairness, or difficulty redirecting behavior afterward. The first two are Processes acting on a Representation of what was coming; the third is a Representation of other people; the fourth is a Process of disengagement that fails under a State. They are different targets.

The reward operations of the Depressive Disorders module (Jonah) apply here with the sign reversed. Anticipatory reward responsiveness is what the tablet recruits before it is switched on, and it is strong in Owen: "Saturday, when nobody tells me stuff" describes a child whose good hours are the ones with an uninterrupted expected reward. Effort-cost computation and effort–benefit integration run on homework under inattention, where every step costs more because the next step is not held in mind, so that a "no" to the tablet is also a "yes" to a task whose end he cannot see. Action initiation for homework and action termination for the tablet are the two transitions at which his outbursts cluster.

Affective response gain

The emotional response to an error is a further matter. The size of an unexpected omission, the importance of the outcome, the meaning of the promise, and the state of the child may all contribute. Affective response gain describes a relation between change in an eliciting input and change in a specified emotional response. It is a property of the response Process, not a trait read off one episode; characterizing it formally needs defined inputs and responses across conditions. Clinically, a pattern of strong reactions to relatively small changes across a four-week log and two settings can motivate the inquiry without establishing a measured parameter. The Lego flick and the tablet outburst were responses to the same event class, an activity ending, at very different amplitudes, and the difference tracks the setting and the warning, not the size of the loss.

Rate of affective recovery and who supplies it

An emotion has a time course: a background state, a response, a development, and a return toward some reference condition. Its peak and its duration vary independently, and describing both as "severe" loses a clinically meaningful difference. Rate of affective recovery describes how an evoked response approaches a defined reference condition after the eliciting event; describing it requires a specified response and a specified reference condition. Owen's recovery to conversational contact takes forty minutes at home, fifteen to twenty at school, and about a minute in the clinic with a warning and a job. Recovery of heart rate, return of reported anger, and renewed ability to engage need not coincide, and "back to baseline" needs care when the baseline itself is irritable: he can recover from the tablet while remaining "grumpy" at the level of the day.

The observation that he settles only when an adult sits nearby without talking specifies where the recovery comes from. The Capacity to return to baseline is present. Access to it under his own control, at the peak, is absent, and another person supplies the access. Regulation can act through attention, meaning, the situation, or another person, and these routes make different demands on working memory and language. An eight-year-old with inattention has less of both to spend at the peak than a fifteen-year-old, which is one reason co-regulation is not a failure of independence but the developmentally available route.

Attention after the disappointment

A pediatric experiment examined attention orienting following frustration in 195 young people aged eight to eighteen, including several clinical groups and young people without a disorder. The task examined processing after the disappointing event; associations between irritability, age, and neural responses were studied within a defined task, and the findings concern those conditions rather than the mechanism of an individual interaction [343]. The clinical translation is a question rather than a finding: after the tablet ends, where does Owen's attention go, and can it be moved? Attentional disengagement, the Process of releasing attention from the current information, is what the adult sitting silently seems to give time for. What ends the store escalation is different: not disengagement but receipt.

Hostile attribution as a Representation

A social event arrives with a history. When a classmate bumps Owen in line, the contact does not specify its own meaning. He must relate it to the line, the classmate's expression, what happened earlier, and what he expects from that person. Mental-state inference is the operation of estimating another person's unobserved state from behavior and context. Causal attribution concerns the explanation assigned to an event. Owen inferred that the classmate wanted to hurt him and attributed the bump to that intention; the inference influenced action even though it was, on the teacher's account, mistaken.

The pattern across events, "everyone thinks I do it on purpose," "he did it on purpose," "everything is unfair," is the content of a hostile attribution model: a Representation of what other people intend toward him, held between episodes and consulted in the ambiguous instant. It is a Representation because it persists, applies across situations, and is what a benign alternative would have to compete with. The represented threat need not be physical; losing, being treated unfairly, and being blamed are what is at stake. The clinician needs to know what Owen thought was at stake without treating every interpretation as equally accurate: an ambiguous bump and an unmistakable shove call for different work, and a more benign explanation has a purpose in only one of them.

Action selection under activation, and what the outcome teaches

Action selection concerns choosing from available alternatives under particular evidence and value conditions. In the store the alternatives included asking, waiting, leaving, and escalating. Owen may know that asking is possible but expect refusal; he may know the words for a break and see no adult who will grant one. The set of practically available actions can be narrower than the set an adult lists afterward, and it narrows further with activation. Reactive and instrumental describe what the selection was organized around: a threat or loss just registered, or an outcome being pursued.

Outcome credit assignment links an outcome to its antecedent, and value updating revises the associated expectation. The store teaches: escalation was followed by candy, and the value of escalating rises by the rule above. The tablet teaches too, though the pupil is different: what Owen's mother learns is that the outburst ends when she stops talking; what Owen learns is uncertain. Being sent home is a third lesson: an outburst ends the school day. None of these lessons requires intent, only that outcomes follow actions and the rule runs.

State: sleep, hunger, and the day of the week

A State is a condition that changes access to capacities. Mondays after a weekend of eleven o'clock bedtimes are Owen's worst days. Sleep loss lowers the threshold at which the same prediction error produces an outburst and slows recovery, without changing what he expects or how he reads a bump. The chronic irritable mood itself functions as a State on the timescale of a day and as a standing condition on the timescale of years; the label does not settle which, and the log, with sleep recorded alongside, is what can.

Circuits, development, and genes

Distributed systems

The ventral striatum participates in processing reward-predicting information and its motivational significance; cortical regions contribute representations of outcomes, context, goals, and costs; dopaminergic input influences activity and learning within these circuits. Dopamine neurons in nonhuman primates respond to the difference between predicted and obtained reward [326], which is the biological root of the prediction-error account above. A signal that changes with surprise can teach an organism about opportunities; it is not a meter of anger, and a primate task does not establish what happens in Owen when the tablet ends.

Youth with chronic, nonepisodic irritability have been compared with youth with episodic bipolar disorder on family history, longitudinal outcome, and behavioral and neural responses in frustration and face-emotion paradigms, and the groups differ on several of those measures [66]. This supports treating chronic irritability as a phenotype in its own right. It does not supply a circuit for Owen; a group difference within a task is not an individual mechanism.

Development

Regulatory systems are distributed. In a reappraisal study of 112 people aged six to twenty-three, age-related changes in reported emotion accompanied changes in prefrontal recruitment and prefrontal–amygdala relationships [327]. Owen is at the young end of that range. The study concerns performance on a specified strategy, not a maturity score. What it licenses is modest: the route by which an alternative appraisal changes feeling is still developing at eight, a reason to expect co-regulation to carry more of the load now than later.

Development also changes the phenotype. The DMDD diagnosis has shown instability over follow-up, with many children no longer meeting criteria while continuing to meet criteria for something [7]. In a birth cohort followed to twenty-six, most adults with a psychiatric disorder had met criteria for a juvenile diagnosis, and adult depression and anxiety were preceded by juvenile oppositional and conduct disorders as well as by juvenile mood and anxiety disorders [59]. One reading of that continuity: a hostile attribution model, a high-gain response to omission, and slow unassisted recovery can persist while the behavior expressing them changes from throwing chairs at eight to withdrawal at eighteen.

Genes

Twin and polygenic-score studies support genetic contributions to childhood irritability, but its genetic overlap with other conditions depends on phenotype and sample [137, 138, 139]. Among children with ADHD, chronic emotional dysregulation has been linked to ADHD genetic risk [138]. In an adolescent network study, irritability shared an edge with the depression polygenic score, not the ADHD polygenic score [139]. The uncle's bipolar disorder raises a family prior and is recorded; it does not identify a process in Owen, and the families of youth with chronic irritability in the comparison studies carried less bipolar disorder than the families of youth with episodic mania [66]. Genetic liability is expressed across development in interaction with sleep, inattention, and a household that is arguing about how to respond to him.

Construct	Kind	What it names in Owen
Prediction error on omission (frustrative nonreward)	Process, on a Representation of the expected reward	The tablet ending
Anticipatory reward responsiveness	Process	"Saturday, when nobody tells me stuff"
Effort-cost computation; effort–benefit integration	Process, under inattention	Homework as a task with no visible end
Affective response gain	Property of the response Process	Lego flick versus tablet outburst
Rate of affective recovery	Property of the response Process; Capacity reachable through another person	Forty minutes, fifteen, one
Attentional disengagement	Process	What silence gives time for
Hostile attribution model	Representation	"He did it on purpose"
Action selection; credit assignment; value updating	Process	The store; being sent home
Sleep loss; chronic irritable mood	State	Mondays after grandmother's; "grumpy" all day

What this case cannot establish

The account names the operations; it does not measure them. The case cannot establish Owen's response gain, his learning rate, or his neural response to omission; whether his hostile attribution model preceded the peer rejection or grew from it; whether the ADHD and the irritability are one liability or two; or which operation will move first under treatment. The studies cited are of other children, in groups, under task conditions. They license the vocabulary and the questions, not a mechanism read off the log.

Figures for this chapter

Treatment

What Owen and his family want

Owen wants people to stop thinking he does it on purpose. His mother wants a year she can do. His father wants to avoid his brother's path, which for him means avoiding medication. The school wants him in the classroom. These are different goals and they can be counted separately: minutes to calm, outbursts with aggression, days sent home, and the hour a day he is happy, which the plan aims to lengthen as a target in its own right rather than as a byproduct of fewer outbursts.

What the evidence covers and what it does not

Treatment evidence specific to DMDD is thin. The diagnosis is recent, most trial evidence comes from children diagnosed with ODD, ADHD, or irritability within another condition, and reviews of the disorder recommend treating the comorbid conditions and the target behaviors with the treatments that have evidence for them [7]. The functional account cannot fill that gap. It can say where each treatment is arranged to act, so that what Owen encounters can count, and it can say what to measure.

Assessment before treatment

The four-week log becomes the instrument. Each entry records the trigger, the warning given, the duration, aggression toward people or objects, self-injury, what ended the episode, whether an adult sat with him, and the time to conversational contact, because those are the dimensions that separate clinical from typical tantrums [10] and the dimensions the differentials turn on. Sleep is logged alongside, since Mondays after grandmother's are the worst days and a State that lowers the threshold is cheaper to change than a Representation. Head banging is asked about at each visit: force, surface, and whether it occurs outside outbursts. His sister's safety during episodes is planned for explicitly.

Parent management training

Parent training programs have the strongest evidence for oppositional and disruptive behavior in children of Owen's age, with reductions in oppositional and aggressive behavior [61]. They are the first treatment here, and they are not a judgment on the parents. The program teaches the parents to notice and attend to what they want more of, to give clear and few commands, to warn before transitions, to hold a consequence without a lecture, and to stop responding to escalation with the thing being escalated for.

The functional account explains what each piece is arranged around. A two-minute warning before the tablet ends changes the prediction, so the omission arrives against a smaller expectation and the prediction error is smaller; this is the difference between the Lego flick and the tablet outburst, arranged deliberately. Not buying the candy removes the outcome that has been teaching escalation; the first weeks are expected to be worse in the store, because an expectation being unlearned produces larger errors before it produces smaller ones, and the parents are told so in advance. Sitting nearby without talking is kept, because it is the route to a capacity Owen cannot reach alone; it is named as co-regulation rather than "giving in," which is what the father has called it. The parents' disagreement is treated as a treatment target because two adults running two rules supply two predictions.

Work with Owen

Cognitive behavioral approaches for childhood irritability, including versions built around graded exposure to frustration with practiced responses, have early support from a randomized multiple-baseline study, not a conventional parallel-group trial [140]. With an eight-year-old the work is concrete. "Too hot" becomes a signal he can name before the peak. The bump in line is revisited with the question "what else could it have been," without insisting that everyone is always benign, because sometimes children do push. The store is revisited with what he wanted and what else he could have done. The work stays at the level of specific events, since a benign interpretation that Owen can produce in the clinic is a Capacity; using it in line is access under a State, and access is built by practice in conditions close to the real ones.

School

The school plan changes what an outburst produces. Being sent home has been teaching that an outburst ends the school day, which is credit assignment running on the wrong outcome. The plan replaces it with a place to calm within the building, an adult who will sit rather than talk, and a return to the room without a debrief at the peak. Transition warnings are built into the day. The ADHD is accommodated in its own right: shorter tasks, a visible next step, and a way to ask for a break before "no" becomes the only exit. The log is shared, so home and school count the same dimensions.

Medication

The evidence for medication runs through the comorbid ADHD. Owen's inattention and hyperactivity are present in both settings. In a secondary analysis of the MTA trial, existing irritability in children with ADHD improved with treatment [141]. A stimulant meta-analysis examined irritability as an adverse event and found different risks with methylphenidate and amphetamine; it does not establish a uniform class effect on irritability [142]. A stimulant is therefore the first medication considered, aimed at the ADHD, with irritability tracked as a secondary outcome rather than promised as a result. Monitoring covers appetite, sleep, growth, pulse and blood pressure, and any emergence of the very irritability the family hopes to reduce, which can occur as the dose wears off [136]. The father's worry is answered directly: a stimulant is not a treatment for bipolar disorder, is not chosen because of the uncle, and can be stopped.

If irritability remains after the ADHD is treated, adding a selective serotonin reuptake inhibitor to the stimulant has been examined in a small randomized trial of children with severe irritability [143]. In that trial, citalopram improved irritability response over placebo, but functional impairment did not differ at endpoint. The mechanism is not a serotonin deficiency account of Owen's temper; the evidence is for outcomes under that combination.

Among second-generation antipsychotics, evidence for reducing aggression in children with disruptive behavior disorders is concentrated in risperidone and is not equally established for every agent [144]. Evidence for irritability associated with autism is separate: risperidone and aripiprazole carry that labeled indication [125, 126]. These agents carry weight gain, metabolic change, sedation, and movement effects, which require scheduled metabolic and neurological monitoring [145]. They are reserved for aggression that has not responded to the above and that endangers Owen or others, and they are used with an explicit plan to stop.

Mood stabilizers and antimanic agents are not indicated by a family history or by chronic irritability. The children with chronic, nonepisodic irritability in the studies that shaped this diagnosis were not shown to follow a bipolar course [66, 7]; a manic episode, characterized as in the presentation chapter, would change the plan, and nothing yet has met that description. If a stimulant or an SSRI is started, sleep need, speech, and activity are watched with that question in mind.

The long view

Children who meet DMDD criteria often stop meeting them while continuing to meet criteria for something else, and the adult outcomes of chronic childhood irritability are depression and anxiety more than bipolar disorder [7]. Juvenile oppositional and conduct diagnoses preceded adult depression and anxiety in a birth cohort followed to adulthood [59]. This changes what "better" means. Fewer outbursts is one measure. The mood between them, which explains impairment the outbursts do not [9], is a second, and it is the one that predicts the longer course. A boy who throws fewer chairs but is still happy an hour a day has had one treatment target met and not the other.

Measuring what changes

Five things are counted from the shared log: outbursts per week, duration, episodes with aggression or self-injury, time to conversational contact with and without an adult present, and Owen's happy hours per day by his mother's and teacher's estimate. Days sent home are counted by the school. A parent and teacher irritability rating is repeated at intervals. Recovery is expected to shorten before frequency falls, and frequency to fall before the mood between episodes lifts, so that an early log showing the same number of shorter outbursts is read as progress rather than failure.

What this case cannot establish

The case cannot establish whether Owen's irritability will respond to treatment of his ADHD, whether parent training will be enough, or whether medication beyond a stimulant will be needed. If he improves, it cannot show whether the change reflects the warning before transitions, the candy no longer arriving, the stimulant, the school plan, sleep, or a year of growth. A molecular event can change the effectiveness of communication within an already learned network, yet this does not explain the individual outcome. A response does not confirm the mechanism the treatment was arranged around, and a non-response does not show that the functional account was wrong. The account offers a way to arrange care so that what Owen encounters can teach him something other than that escalation works.

Figures for this chapter

---