Bipolar disorder (BD) is challenging to diagnose in children and adolescents, yet timely and accurate diagnosis is crucial. Delayed diagnosis can result in ineffective treatment and substantial morbidity, while mis-diagnosis risks unnecessarily exposing patients to medication side effects. Despite ongoing controversies about BD in prepubertal children, there is a body of research specific to children and adolescents. This article unpacks BD and offers recommendations for assessment.

What is the bipolar spectrum?
Many researchers use the term “bipolar spectrum disorder” (BPSD) to include cases of bipolar I (BP I), bipolar II (BP II), cyclothymia, and what DSM-IV called “bipolar not otherwise specified” (BP-NOS). Much of the research on pediatric BD used the DSM-IV, so we need to start there. Researchers in the US operationalized childhood BP-NOS to include cases in which: 

The manic episode had one fewer symptom than required for a “classic” manic episode (ie, two symptoms plus elevated, or three symptoms plus irritable mood)

The mood episode lasted at least four hours on at least four separate days, compared to “classic” episodes that require symptoms to occur on consecutive days (Birmaher B et al, Arch Gen Psychiatry 2006;63(2):175–183)

The multicenter COBY study recruited pediatric patients with “sub-syndromal” bipolar symptoms (­average age 12 years). Within five years of presentation, about 45% went on to meet full criteria for BP I or BP II, while 41% continued to meet the BP-NOS criteria. Only 14% of kids with BP-NOS showed partial or full remission within five years (Axelson DA et al, J Am Acad Child Adolesc Psychiatry 2011;50(10):1001–1016.e3). This broader definition of BD can capture many young people on their way to developing BP I or BP II; however, it is still up for debate whether this broader definition includes youth who are moody for other reasons. Well defined or not, BP-NOS in youth requires intervention because these children have problematic symptoms and impairment similar to youth with BP I or BP II, including suicide attempts (Towbin K et al, J Am Acad Child Adolesc Psychiatry 2013;52(5):466–481). 

Family history of BD is among the strongest risk factors for its development. In children with a bipolar parent (high-risk children), BPSD emerges over years, usually manifesting by the mid-20s. Kids who develop BPSD may have different symptoms at different ages, with a prodromal period preceding frank BD symptoms (Duffy A et al, Am J Psychiatry 2019;176(9):720–729). These include: 

• Sleep and anxiety symptoms at younger ages (ages 4–10) 
• Adjustment (stress sensitivity) or minor depressive disorders in early adolescence (ages 10–13)
• Major depression, psychosis, or nonresponse to SSRI (mid-late ­adolescence)
• Hypomanic or manic episodes (late adolescence and early ­adulthood) 

Irritability?
Early researchers studying pediatric BD suggested that chronic, intense irritability was a form of BD in prepubertal children. However, 20 years of research have soundly demonstrated that chronic irritability is distinct from BD, and youth with chronic impairing irritability are now diagnosed with disruptive mood dysregulation disorder (Brotman MA et al, Am J Psychiatry 2017;174(6):520–532). Irritability can be a prominent symptom in BPSD, but in BPSD the irritability must be episodic and must be accompanied by other symptoms of mania or depression. 

Prevalence in adolescents vs children
Rates of broadly defined BPSD are much higher than rates of narrowly defined BD in children and adolescents. Furthermore, BPSD is far more common in adolescents than in younger children. A meta-analysis of 19 prevalence studies, representing 56,103 subjects from 10 countries, estimated the BPSD rate among adolescents to be a whopping 8.3%, while it was only 1.7% among children under 12. Prevalence rates were similar across countries. Classic BP I in children and adolescents, however, is rare. Of 14 studies reporting on rates of BP I, four had zero cases and three found just one to two cases. The meta-analysis of all 19 studies estimated the rate of BP I to be only 0.6% among youth (Van Meter A et al, J Clin Psychiatry 2019;80(3):18r12180). Controversy persists around this topic, particularly regarding heterogeneity in the underlying epidemiological studies. For example, in the studies that found any BPSD in youth, rates of adolescent BPSD ranged from 1.2% to 14.3% (Parry P et al, Int J Bipolar Disord 2021;9(1):21). This debate goes beyond child psychiatry, as there is active discussion among adult bipolar researchers about whether current DSM definitions are too restrictive, resulting in underdiagnosis of adult BD (Parker G et al, J Affect Disord 2014;156:87–91). For more on diagnosing BD in adults, see The Carlat Psychiatry Report, Nov/Dec 2021.

Prepubertal BD: US vs global findings
There is more agreement about the emergence of BPSD in adolescence but considerable debate about BPSD in prepubertal children. Five studies of genetically high-risk children (who have one parent with BD) found no evidence of BPSD in prepubertal children (Duffy A et al, Int J Bipolar Disord 2020;8(1):18). In contrast, US-based researchers consistently describe BPSD in prepubertal children. Compared to other cohorts, US kids have higher rates of comorbid attention deficit hyperactivity disorder (ADHD) and other disruptive behavior disorders, and higher rates of social adversity (Birmaher B et al, Arch Gen Psychiatry 2009;66(3):287–296). These research differences are reflected in dramatic differences in inpatient discharge diagnosis rates of bipolar disorder, with US rates for 5- to 9-year-olds hundreds of times higher than in several other countries (Clacey J et al, BJPsych Open 2015;1(2):166–171). 

Social inequity and trauma
Inequities due to race and class can also affect the diagnosis and treatment of BD. In the US, Black adults and adolescents with BD tend to be misdiagnosed with schizophrenia and undertreated with medication (Glassgow AE et al, Health Equity 2019;3(1):604–611). Economic disadvantage often results in lack of access to specialty care in the US, and this is associated with both overdiagnosis and overtreatment of BD. For example, in Kentucky, 2.4% of 6-year-olds are prescribed antipsychotics, often with a diagnosis of BD, but the diagnosis and prescriptions are usually not made by child psychiatrists (Stringaris A, Child Adolesc Ment Health 2019;24(1):106–107). 

Social disadvantage brings childhood adversity, trauma, and psychopathology. Aggressive children with histories of trauma and social disadvantage have disruptions in mood, thought, and behavior, which contribute to overdiagnosis of BD, including in prepubertal children (Havens JF et al, J Am Acad Child Adolesc Psychiatry 2022;61(3):364–365). Still, childhood abuse, neglect, exposure to parental psychiatric symptoms, and attachment difficulties may all increase the risk of early onset of BPSD. 

Tips for assessment 
Given the complexities surrounding the definition of bipolar conditions, what tools or tips can help us sort out the diagnosis? Here are a few that we find helpful: Look at risk factors, developmental trajectories of symptoms, and substance use. 

Risk factors, time course, and risk ­calculators

A key risk factor is family history of mood disorders and treatment. Children of parents with BD have an estimated eight- to 10-fold higher lifetime risk of developing BD than the general population (Duffy A et al, J Clin Psychiatry 2000;61(9):630–637). If the child has this history, BPSD is more likely. Ask about family treatment—lithium response has a genetic component.

Differentiating BPSD from ADHD, posttraumatic stress disorder, or oppositional-defiant disorder requires attention to whether symptoms occur episodically or chronically. Ask multiple informants about the time course of symptom onset and resolution. Mood rating scales or checklists can help you and your patients consistently identify symptoms, improve diagnostic decisions, and track treatment. The Parent General Behavior Inventory (www.tinyurl.com/2cj3sawr) or the Child Mania Rating Scale (www.tinyurl.com/44r84jtz) are two examples. A good social history can identify maltreatment, trauma, or extreme adversity contributing to mood dysregulation and inform your differential diagnosis. 

US researchers have developed a risk calculator that uses clinical symptoms and family risk factors to estimate the probability that a patient will meet full criteria for BD within five years; it may be used periodically to track care. This approach may motivate families to watch for symptoms and return for follow-up. The calculator is available for free at: www.cabsresearch.pitt.edu/bpriskcalculator/

Developmental course of mood ­symptoms

Parents often feel demoralized by the time they are in your office, often years after their child’s first symptoms or after trying other ineffective treatments. Explain to parents that the shifting symptoms and lack of treatment response may be part of this complicated illness and that it can take a long time for the symptoms of BD to become clearly ­visible. 

Co-occurring substance use

Substance use disorders (SUDs) occur in up to 33% of adolescent patients with BD, including alcohol, marijuana, and tobacco (Scavone A et al, J Can Acad Child Adolesc Psychiatry 2018;27(3):159–166). This combination is associated with earlier onset of BD, severe symptoms, rapid cycling, and suicidality. Assess substance use symptoms with a checklist such as CRAFFT (www.tinyurl.com/mr2wxysw). If your patient has a co-occurring SUD, treat both conditions together (see www.tinyurl.com/39mfrw4c for more information). For patients without substance use problems, talk to them and their parents about prevention strategies to avoid problem drug and alcohol use. For a good resource, see: ­www.­tinyurl.com/5fftjfya 

Carlat Verdict: BPSD is rare in prepubertal children, and BP I is vanishingly rare. Family history and history of adversity are as important as mania and depression. There is vigorous debate about narrow vs broad definitions of BD. With no age-based standards for “appropriate levels” of grandiosity, elation, or irritability, diagnosis of BPSD depends upon methodical assessment of symptoms, time course, episodic patterns, and family history. Manic symptoms often emerge from a developmental trajectory of symptoms, so reassess for bipolar symptoms as patients age and explain this to patients and their families.