Conduct disorder (CD) is a childhood and adolescent psychiatric diagnosis defined in the DSM-5 by a persistent pattern of behavior that violates the basic rights of others or major age-appropriate social rules. The DSM-5 organizes the diagnosis around 15 specific behaviors grouped into four categories: aggression toward people and animals, destruction of property, deceitfulness or theft, and serious rule violations. A young person needs to have shown at least three of those 15 behaviors in the past 12 months, with at least one present in the past six months, and the behavior must cause meaningful problems in social, academic, or occupational functioning. What makes the DSM-5 version of this diagnosis worth understanding in detail is that it introduced a new specifier and preserved an age-of-onset distinction that together shape how clinicians think about severity, prognosis, and treatment.
The Four Behavior Categories
The 15 criteria span a wide range of conduct. The aggression category includes bullying, threatening, or intimidating others; initiating physical fights; using a weapon; being physically cruel to people or animals; stealing while confronting a victim (mugging, purse-snatching); and forcing someone into sexual activity. Destruction of property covers deliberate fire-setting and other intentional damage. Deceitfulness or theft includes breaking into someone’s property, lying to obtain goods or favors, and stealing without confrontation (shoplifting, forgery). Serious rule violations include staying out at night against parental rules before age 13, running away from home overnight at least twice, and frequent truancy beginning before age 13.
Not every young person who gets in fights or skips school meets criteria. The DSM-5 requires that the pattern be repetitive and persistent, and clinicians are expected to rule out other explanations. If the person is 18 or older, the diagnosis is only given when antisocial personality disorder criteria are not met. That boundary matters because CD in the DSM framework is understood as a developmental diagnosis, one that describes a trajectory rather than a fixed trait.
Childhood-Onset Versus Adolescent-Onset
The DSM-5 distinguishes two subtypes based on when the first symptom appears. Childhood-onset type applies when at least one criterion behavior shows up before age 10. Adolescent-onset type applies when no criteria are present before age 10. There is also an “unspecified onset” for cases where the history is unclear. This distinction is not just administrative. Research consistently finds that the two groups differ in severity, background, and long-term outlook.
Young people with childhood-onset CD tend to commit more serious and violent offenses. In one study comparing the two groups, about 78% of those who committed serious or violent offenses had the childhood-onset type, and that group also had significantly more contact with police and juvenile courts.
The childhood-onset group also shows a stronger connection to childhood physical abuse: in that same study, roughly two-thirds of the childhood-onset group had histories of physical abuse compared with about 16% of the adolescent-onset group. Memory and learning scores were severely impaired in the childhood-onset group while falling in the average range for adolescent-onset youth.
The adolescent-onset subtype, by contrast, appears more influenced by peer context and social environment. Research testing the developmental model behind this distinction found that individual and family risk factors were more strongly tied to childhood-onset CD, while exposure to deviant peers was more strongly tied to adolescent-onset CD.
The Limited Prosocial Emotions Specifier
One of the most significant changes the DSM-5 brought to this diagnosis was the addition of the “with limited prosocial emotions” (LPE) specifier. To qualify, a young person must persistently show at least two of the following over at least 12 months, in more than one relationship or setting: lack of remorse or guilt, callous lack of empathy, lack of concern about performance at school or work, and shallow or deficient affect (emotions that seem superficial or are turned on and off for manipulation).
The LPE specifier identifies a subgroup that tends to have worse outcomes and distinct developmental roots. A large prospective study found that fearlessness, callous-unemotional traits, interpersonal difficulties, and harsh parenting assessed in early childhood were developmental antecedents of the LPE subtype.
This specifier matters clinically because it flags children who may not respond as well to standard discipline-based interventions. The emotional deficits that define LPE overlap with traits traditionally described as psychopathic in adults, though clinicians and researchers are careful to note that applying that label to children is both premature and potentially harmful. The DSM-5 frames it as a dimensional specifier for a reason: it describes a pattern of emotional functioning, not a personality type set in stone.
Genetic and Environmental Risk Factors
CD does not arise from a single cause. A large meta-analysis of twin studies estimated that about half the variance in conduct disorder is attributable to genetic influences, with shared environmental factors accounting for a further 14%.
On the genetic side, the picture is not one of a single “conduct disorder gene.” Different symptom clusters are influenced by different genetic pathways. Research on individual CD criteria found that genetic contributions were highest for rule-breaking behaviors like running away and truancy, and for overt aggression like fighting and hurting others. Covert delinquency symptoms, by contrast, showed stronger shared-environment effects, suggesting that the social context a child grows up in plays a larger role for those behaviors.
Gene-environment interactions also play a documented role. In females, for example, the MAOA gene interacts with childhood adversity and maternal antisocial behavior to influence risk for CD.
Environmental risk factors identified across multiple meta-analyses are broad and sometimes surprising. The strongest include prenatal alcohol exposure, which roughly tripled the odds of CD, and being a child of an alcoholic parent, which more than doubled the odds. Maternal smoking during pregnancy approximately doubled the risk. Other factors with statistically meaningful associations included severe psychiatric disorders in parents, maternal depression around the time of birth, childhood overweight or obesity, and even postnatal exposure to secondhand smoke.
Childhood maltreatment is one of the strongest environmental predictors. In a nationally representative sample, childhood maltreatment was associated with CD at adjusted odds ratios ranging from about 2.4 to 4.7 depending on the type of maltreatment.
What Happens in the Brain
Brain imaging research has consistently pointed to differences in how young people with CD process emotions and make decisions. Structural studies have found reduced gray matter volume in the amygdala, a region central to threat detection and emotional learning, in both childhood-onset and adolescent-onset CD groups compared with healthy peers. The adolescent-onset group additionally showed reduced volume in the right insula, and more CD symptoms correlated with smaller insula volume within the CD group overall.
The wiring between the amygdala and the orbitofrontal cortex, a brain region involved in evaluating rewards and consequences, also appears altered. One study using brain tract imaging found abnormal connectivity in the bundle of fibers connecting these two areas in young people with CD. Functional imaging work tells a similar story: youth with CD and high levels of callous traits showed disrupted signaling in the orbitofrontal cortex and caudate during tasks involving rewards and punishments, suggesting their brains are less responsive to cues that typically guide behavior away from harmful choices.
The stress-response system also functions differently. Adolescents with CD, regardless of onset type, showed blunted cortisol and cardiovascular responses to psychosocial stress compared with controls. A dampened stress response may contribute to fearlessness and reduced sensitivity to punishment, both of which are features of the disorder.
Gender Differences
CD is more commonly diagnosed in boys, but the picture is more complicated than simple prevalence numbers suggest. Girls with CD tend to present differently. Research has found that girls with CD have higher rates of depression, anxiety disorders, PTSD, and borderline personality disorder, while boys with CD have higher rates of ADHD.
There is also what researchers call the “gender paradox”: because fewer girls meet criteria for CD, those who do tend to be more severely affected. Girls with CD showed significantly more lifetime psychiatric conditions, including alcohol use disorder, and their CD symptoms were more severe overall. Fewer girls had the childhood-onset subtype. Interestingly, girls with adolescent-onset CD showed levels of psychopathology similar to boys with childhood-onset CD, supporting the idea that girls may follow a “delayed-onset pathway” that looks milder on the surface but carries comparable risk.
The behavioral expression of CD also differs by gender. Among youth with CD, girls had a stronger link between the diagnosis and risky sexual behavior such as having multiple partners and condomless sex, while boys with CD showed a stronger link to using alcohol or drugs before sex.
Overlap with ADHD and Oppositional Defiant Disorder
CD rarely appears in isolation. The most common co-occurring conditions are ADHD and oppositional defiant disorder (ODD), and the boundaries between them are blurrier than the diagnostic manual implies. Twin research has found that the overlap among CD, ADHD, and ODD is largely explained by shared genetic influences, though each disorder also has unique genetic contributions that justify keeping them as separate diagnoses.
In practical terms, ODD appears to be the more direct stepping stone to CD. A longitudinal study tracking conduct problem trajectories found that about 4% of boys and 2% of girls followed a persistently problematic CD trajectory. Among boys, engagement in that trajectory was predicted by ODD but not by ADHD independently. Among girls, both ODD and ADHD predicted it. The study concluded that although ADHD frequently co-occurs with CD, the association between ADHD and CD is largely accounted for by accompanying ODD symptoms.
This has practical implications. A child with ADHD alone is not on an inevitable path toward CD. But a child showing both ADHD and ODD symptoms, particularly escalating defiance, hostility, and aggression, warrants closer monitoring.
The Connection to Trauma
The relationship between CD and traumatic experiences runs in both directions. Childhood maltreatment is a strong predictor of developing CD, but young people with CD are also significantly more likely to experience traumatic events afterward and to develop PTSD. In a nationally representative study, people with a history of CD were about 2.7 times more likely to report experiencing any traumatic event and about 2.2 times more likely to develop PTSD. Among those diagnosed with both CD and PTSD, conduct disorder symptoms developed first in roughly 73% of cases.
This finding challenges a common assumption that behavioral problems in traumatized youth are always a consequence of trauma. Sometimes they are. But in a substantial portion of cases, the conduct problems preceded the trauma exposure, and the aggressive and rule-violating behaviors themselves increased exposure to dangerous situations. Clinicians evaluating youth with both CD and trauma histories need to understand the sequence, because treatment that focuses solely on trauma processing may miss the underlying behavioral pattern, and vice versa.
How Often CD Progresses to Antisocial Personality Disorder
One of the most important questions about CD is whether it predicts lifelong antisocial behavior. The traditional clinical estimate has been that a minority of youth with CD go on to meet criteria for antisocial personality disorder (ASPD) in adulthood. However, at least some research challenges that reassuring narrative. In one study, approximately 75% of individuals with CD also met criteria for ASPD, leading the authors to conclude that progression was the norm rather than the exception in their sample. A separate study following adolescents after substance abuse treatment found that 61% met ASPD criteria four years later, with early onset of deviant behavior (at or before age 10), a wider variety of problem behaviors, and more extensive drug use being the strongest predictors.
These numbers are higher than what many clinicians expect, though they likely reflect the severity of the study populations involved. The broader clinical reality is that CD does resolve for many young people, particularly those with adolescent-onset, less severe presentations, and access to effective intervention. But dismissing CD as “just a phase” carries real risk, especially for youth with early onset, the LPE specifier, and limited social support.
Treatment Approaches
The strongest evidence base for treating CD in younger children belongs to parent-focused interventions. Parent management training (PMT) teaches caregivers specific strategies for reinforcing prosocial behavior and reducing coercive parent-child interactions. A meta-analysis found that PMT produced moderate-to-large reductions in parent-rated disruptive behavior compared to waitlist controls. Parent-child interaction therapy (PCIT), which adds direct coaching of parent-child exchanges in real time, showed even larger effects, particularly for younger children.
Randomized trials have confirmed these benefits in routine community settings, not just research clinics. One Swedish trial found that practitioner-led group PMT sessions reduced child conduct problems more than self-directed training, and that improvements continued at six-month follow-up. Gains in parenting skills mediated the improvements in child behavior, supporting the logic that changing how parents respond changes how children act.
For adolescents with more entrenched problems, multisystemic therapy (MST) has been widely studied. MST is an intensive home-based intervention that addresses the young person’s behavior within the context of family, school, and peer systems. Decades of research support it as an evidence-based approach for serious offending and delinquency, though a recent meta-analysis offered a more tempered view: MST had a meaningful effect on reducing time in out-of-home placement but showed no clearly important effect on new offenses, delinquency, or substance use. Functional family therapy (FFT) showed possible benefits for reducing new offenses and substance use, though evidence strength was low.
Medication is not a first-line treatment for CD. A Cochrane review found some evidence that risperidone can reduce aggression and conduct problems in the short term, but also that it causes significant weight gain. No strong evidence supports any other antipsychotic for this purpose, and the review emphasized that medication should not be used alone given the effectiveness of behavioral interventions. A broader network meta-analysis of drug treatments for disruptive behavior found that second-generation antipsychotics and stimulants both outperformed placebo, with stimulants being most relevant when ADHD co-occurs. But the overall message from the evidence is that behavioral and family-based treatments remain the backbone.
Diagnostic Bias and Cultural Context
A diagnosis this consequential carries risks beyond the clinical. Research has documented that racial and ethnic minority youth, particularly Black and Hispanic youth, are disproportionately diagnosed with disruptive behavior disorders including CD compared with white peers. Both implicit clinician biases and culturally unstandardized assessment tools contribute to this over-representation, which has downstream consequences in the juvenile justice and mental health systems.
This does not mean CD is not real or that the diagnostic criteria are inherently flawed. It means that applying them requires careful, context-aware clinical judgment. A child’s behavior does not exist in a vacuum. Poverty, neighborhood violence, discrimination, and unstable housing can all produce behaviors that superficially resemble CD criteria without reflecting the same underlying disorder. The DSM-5 includes a note that clinicians should consider whether the behavior is adaptive to the individual’s environment, but how consistently that note gets applied in practice is another matter entirely.