Emotions run high when it comes to bipolar disorder (BD) in children. At conferences, I often had to speak in front of audiences that were divided into those who strongly favoured a wider use of the diagnosis of BD and those who fervently opposed it. For years, such diametrically opposed views have been held with equal vigour, despite the striking absence of much evidence either way. This is about to change as more empirical data on BD accumulate. It is research like that conducted by Boris Birmaher's group that allows us to start having a more rational discussion on youth BD, and a debate that goes beyond anecdote and prejudices (Birmaher, 2013).
There are good reasons to want to understand BD in early life. First, the presentation of bipolar among children and young people is important to elucidate in order to resolve the mystery of the origins of BD in adults. More than 50% of adult bipolar patients report onset of symptoms during childhood and adolescence, but we need more research to understand the nature of these symptoms and whether they will be useful in predicting later BD (Perlis et al., 2009). Second, in order to intervene early. BD can be a devastating illness and it seems plausible to assume that diagnosing it early and treating it actively may be beneficial.
The debate about BD started in this country a few years ago, when news had crossed the Atlantic that US psychiatrists were giving diagnoses of BD to irritable children and labeling normal childhood happiness as a form of rapid cycling mania. It was a scare story and, as such stories go, part of it, but only part of it, was true. There were, indeed, influential suggestions that chronic and severe irritability was characteristic of early-life BD. This notion was successfully countered by subsequent research conducted at NIMH by Ellen Leibenluft and her team (Leibenluft, 2011). Birmaher in his article also explains how notions of ultradian and other rapid cycling in children may have led to over-diagnosis of bipolar as well as overtreatment.
But the story rarely told is how much we have learnt about BD and mood dysregulation in youth from research that was done in the US. Studying the clinical course of BD in youth is where Boris Birmaher's research has led the way and given us crucial new insights. Here, I will only focus on the two that I think are most relevant to child psychiatrists (and CAMHS practitioners) on this side of the Atlantic.
First, we now have solid data about the devastating impairment that BD can cause in young people, from impact on academic and family life to worryingly increased rates of suicidality. All this is lucidly described by Birmaher in his Jack Tizard Lecture article. There is no excuse for ignoring BD any longer: we must be alert to this diagnosis in youth and treat it decisively.
Second, we have learnt that we should not over-rely on the arbitrary DSM criteria about the duration of manic episodes. Birmaher's research shows that children with shorter episodes of mania convert at a high rate (around 45%) to classical, full-blown bipolar with time. This crucial insight is in keeping with data from adults (Angst et al., 2003), which also suggest that shorter episode durations are on a spectrum with mania that lasts for four or or more days.
As in all good research, Birmaher's raises a number of questions that will shape future research in the field.
The first concerns the duration of the short-lived episodes. Are episodes of 2.5 or 3.5 days on the same spectrum as episodes lasting 6 or 10 hours? Do they carry the same risk of transition to classical mania, and are they equally impairing?
This leads on to the next question: where on the spectrum should we start treating young people and with what? The majority of patients with BD-not otherwise specified (NOS) do not seem to convert to BD, at least not within the time period studied so far. As we are not yet good at predicting who will convert to classical mania, how justified is it to use anti-psychotics or anti-epileptics as a prophylaxis? Could the impairment that youth with BD-NOS suffer be alleviated by any other means (e.g., family-focused treatment)? While acute classical mania requires prompt and decisive pharmacological treatment, clinicians will want to carefully balance the benefits of medication against its risks, such as weight gain and adverse metabolic changes.
The other question concerns public health. The data by Birmaher et al. come from laboriously-collected and well-characterised clinical samples. How do these findings generalise to non-clinical populations? We have found that short-lived (few hours to 3 days) episodes during which children meet all criteria for mania are 15 times more common than classical mania (1.5% vs. 0.1%) in the general population of children and adolescents (Stringaris, Santosh, Leibenluft, & Goodman, 2010). Do all these children have BP or are they at high risk for mania? It is unclear at the moment. An alternative explanation could be that some of these children experience phasic exacerbations of another underlying illness (e.g., the frequently comorbid ADHD or ODD). We will not know until we have longitudinal data from general population samples.
The good news is that we may soon be able to answer these questions if we continue to do more clinical research. Once, the debate on bipolar in children and adolescents was marred by ideological entrenchment. It now looks more like an opportunity for discovery that will be helpful to clinicians and benefit our patients.