Paediatric bipolar disorder: Reality or myth?

To the Editor

‘Paediatric bipolar disorder’ (PBD) is a frequently made diagnosis in the United States. The diagnosis is based on descriptions of pre-pubertal ‘mania’ with ultra rapid mood cycles and is often comorbid with neurodevelopmental disorders such as attention-deficit hyperactivity disorder (Duffy and Malhi, 2017). However, these frequently occurring forms of childhood mood lability are unlikely to be due to the early onset of bipolar disorder. As Duffy and Malhi (2017) noted in a recent Editorial, ‘the early course of bipolar disorder charted from prospective studies of high-risk offspring is strikingly different from that derived from studies of clinical samples of children diagnosed with paediatric bipolar disorder’ (p. 761).

A recent re-analysis of epidemiological studies drew similar conclusions (Parry et al., 2017). Previously, it has been argued that PBD is common but underdiagnosed outside the United States. The average population prevalence has been estimated to be as high as 1.8% among 7–21 year olds. However, we re-analysed the child and youth epidemiological studies, and found that bipolar rates fell close to zero, when concordance of parental and youth report was required for diagnosis. It was unclear if any pre-pubertal children were diagnosed with bipolar disorder across 12 epidemiological studies using strict criteria. Bipolar rates rose when there was only a youth informant, and impairment criteria were not included.

The methodologically best study found a lifetime prevalence of bipolar-I and bipolar-II disorder of 0.1% with parent and child/youth concordance, all cases being at least 16 years old. With regard to bipolar-not otherwise specified (NOS), the parent report rate was 1.1% and the youth report rate was 1.5%; however, the correlation was no better than chance (k = 0.02), and the authors commented that bipolar-NOS appeared unrelated to bipolar-I or bipolar-II. They suggested the term ‘mood lability’ might be more appropriate (Stringaris et al., 2010: 36).

Based on the best evidence from the bipolar offspring studies and the child and youth epidemiologic surveys, it would be preferable to dispense with the term ‘PBD’ at this stage. Further research is required on illness trajectories among well-defined types of bipolar disorder before definitive models can be developed for staging the course of the illness (Duffy and Malhi, 2017). This research may lead to soundly based models of early detection and early intervention. Meanwhile, as Duffy and Malhi suggest, we should take careful aim before ‘firing’ at phenomenology that may or may not be implicated in the illness trajectories.