Paediatric bipolar disorder: What are the dangers of treating a hypothetical disorder as a real disease?

To the Editor

Amerio et al. (2016) have recently reviewed the treatment of children and adolescents (aged 4–17 years) diagnosed with both paediatric bipolar disorder (PBD) and obsessive–compulsive disorder (OCD). They report Osler’s view, ‘medicine should be treatment of diseases, not of symptoms’ (p. 594), and, on this basis, recommend treating PBD as the underlying disease, using adult mood stabilisers for children and adolescents.

Osler’s view is highly relevant to PBD: Is PBD really a ‘disease’ or just a loose collection of common ‘symptoms’ such as irritability and mood lability that arise from many causes in childhood?

PBD began as an interesting research question, investigated by child psychiatrists working in US universities during the 1990s. They wondered whether adult bipolar disorder (BD) could be detected early in life and treatment instituted from childhood. Subsequent US studies focused on symptom profiles among young children including offspring of adults diagnosed with BD. However, after many studies over two decades, it was found to be difficult to predict a low prevalence condition like adult BD from high prevalence childhood symptoms such as irritability and mood lability measured decades earlier (Malhi, 2016). By definition, it is challenging to predict relatively rare outcomes from variable collections of frequently occurring antecedents.

While PBD remained an interesting research hypothesis in Ivy League universities, it presented few dangers. However, the concept was prematurely translated into clinical practice, resulting in the widespread off-label prescribing of a broad range of adult psychotropic medications for children and adolescents, especially in the United States (Malhi, 2016). For instance, Amerio et al. (2016) located seven clinical studies of PBD–OCD, which included the use of clozapine (for a 13-year-old boy), lithium (including a 4-year-old boy), lamotragine, divalproex sodium, olanzapine, risperidone, quetiapine, aripiprazole, clonazepam, clomipramine and escitalopram among others (Table 1: p. 595). The risks and benefits of these 11 medications have not been fully investigated with randomised controlled trials in children, but the original clinical studies reported a variety of side effects such as increased appetite, sedation, slurred speech, gait disturbance, low blood pressure, neurological symptoms, agitation, manic switch and suicidal intent.

In Australia and New Zealand, child psychiatrists have been careful observers of the US PBD phenomenon with the vast majority choosing not to adopt the hypothesis into clinical practice (Parry et al., 2009). This caution was well founded, as it has proved impossible to diagnose PBD accurately, and adult medications present iatrogenic dangers for children.