We’ve all noticed the trend: children are increasingly being diagnosed with bipolar disorder. According to the Wall Street Journal, the number of children diagnosed with bipolar increased by 26% from 2002 to 2004 (May 25, 2005).

This issue is controversial, because psychiatrists are already accused of overmedicating children. Over the last several years, we have been criticized for prescribing children antidepressants that may cause suicidal ideation, for overusing stimulants, which may exacerbate pediatric cardiac problems, and most recently for a fivefold rise in the use of antipsychotics in children from 1993 to 2002 (Arch Gen Psychiatry 2006;63:679-685).

Now, with a surge of research interest in establishing bipolar disorder as a bona fide diagnosis in children, we may be setting ourselves up for another public relations fiasco.

The key questions are:

1. Does bipolar disorder exist in children?


2. If it does exist, is it the same creature as bipolar disorder in adults?


3. Do medications help pediatric bipolar disorder?

Recently, two intriguing articles were published to help us answer these questions. The first was a 10-year update on pediatric bipolar disorder published in the “orange journal” (J Am Acad Child Adolesc Psychiatry 2005;44(9):872-887), and the second was an editorial by child psychiatrist Jennifer Harris in Psychiatric Services (2005;56:529-531).

The review article in JAACAP is thorough and readable, and takes a refreshingly objective stance in a field rife with debate. The authors begin by acknowledging wide disagreement about what might constitute the diagnostic phenotype of pediatric bipolar disorder. The two most influential research groups in the U.S. are Barbara Geller’s group at Washington University in St. Louis, and Joseph Biederman’s group at Massachusetts General Hospital in Boston. Geller’s people believe that grandiosity and elevated mood should be requirements of the diagnosis, whereas Biederman’s people believe that irritability is central to pediatric bipolar disorder.

Geller’s criteria correspond to what is termed a “narrow” phenotype and closely mirror the standard DSM-IV criteria used for adults. Like their adult counterparts, these children have alternating episodes of depression and mania. However, unlike adults, children frequently exhibit what Geller and colleagues refer to as “complex cycling.” According to this theory, the DSM-IV’s requirements that mania last 7 days and hypomania 4 days are not applicable to children. Instead, Geller has coined the terms “ultrarapid cycling” (5-364 cycles per year) and “ultradian cycling” (>365 cycles per year, with a minimum cycle length of 4 hours). With cycles this short, some experts dispense with the whole notion of cycles and talk instead about “affective lability” or “mood dysregulation.”

However, Harris, the child psychiatrist/editorialist, notes that affective lability is hardly specific for bipolar disorder, and that a careful diagnostic assessment of such children often leads to more reliable diagnoses, such as pervasive developmental disorder, posttraumatic stress disorder (PTSD), and reactive attachment disorder. Since these assessments are difficult and time-consuming, the “easy out” of a bipolar diagnosis may be sorely tempting: “The enormity of the problems many children face makes the simplicity of a biological explanation tremendously appealing,” she writes.

What about the Biederman diagnosis? For Biederman and his colleague Janet Wozniak, irritability is king, even in the absence of many symptoms that we commonly associate with bipolar disorder, like depression, elevated mood, grandiosity, or even clear periodicity. This relatively non-specific notion obviously casts a wide diagnostic net, because most children have episodes of irritability, which beleaguered parents through the ages have termed “tantrums.” However, according to the MGH team, when these tantrums are particularly explosive, long-lasting, and lack reasonable triggers, they may be symptoms of bipolar disorder (J Clin Psychiatry 2001;62(suppl 14):10-15.)

The Biederman/Wozniak criteria are part of the “broad phenotype” concept, and is generally coded as “bipolar disorder, not otherwise specified (NOS)” for reimbursement purposes. This broad category may lead to high rates of bipolar diagnosis, at the expense of diagnostic validity. And according to Harris, in Psychiatric Services, there are other, non-scientific, factors encouraging psychiatrists to make the bipolar diagnosis in children. These include positive feedback from parents, who are often relieved to finally get an “explanation” for their child’s behavior; and encouragement from funding sources, like state agencies and private insurers, who are more likely to fund care for a child with the bipolar label.

Harris’s arguments are compelling: children with mood lability and severe, violent behavioral problems are not necessarily bipolar, and giving them that diagnosis feels sloppy and imprecise. Nevertheless, many drugs approved for bipolar disorder do help kids with these symptoms. These drugs include familiar standbys like Depakote (valproic acid), lithium, Tegretol (carbamazepine), and the atypical antipsychotics. Both clinical experience and clinical trials endorse their effectiveness for kids with severe affect dysregulation (the JAACAP article mentioned above thoroughly reviews this research).

Hopefully, the architects of DSM-V will figure out a more valid diagnostic term to guide our treatment, because “bipolar disorder, NOS” is starting to wear a bit thin.

TCPR Verdict:
Pediatric bipolar disorder: The diagnostic battle continues.