Overview on irritability and the bipolar debate
The motivation for research in irritability originates in the so-called bipolar debate. Over the last 15 years, the rates of BP diagnoses in children and adolescents in the USA have increased dramatically, both in inpatient [
2] and outpatient services [
3]. The increase in rates of BP diagnoses occurred in parallel with a rise of prescription rates of antipsychotic medication [
4]. Diagnostic practice seemed to be the most likely explanation for the observed increase in BP diagnoses [
5] and irritability seems to have been at the heart of this problem. Based on the premise that mania may present differently in youths than in adults, some researchers suggested that chronic, non-episodic irritability is a core characteristic of BP in young people (e.g. [
6].). This is inconsistent with the DSM-IV criteria for BP, which specify the need for a “distinct period of abnormally and persistently elevated, expansive or irritable mood”. Moreover, irritable mood is relatively less specific to BP in the DSM-IV. It requires the presence of four additional symptoms for the diagnosis to be made, compared with three or more for elevated or expansive mood.
Longitudinal outcomes and family history of irritability
Irritability, although not formally defined in the DSM-IV, is a mood symptom present in the criteria for a large number of psychiatric disorders and refers to easy annoyance and touchiness that can manifest in anger and temper outbursts [
1]. Depending on the stringency of threshold, prevalence rates of irritability range from 3 % for severe, chronic irritability strictly defined in the Great Smoky Mountain study [
7] to 20 % in the Isle of Wight study [
8]. In response to the bipolar controversy, Leibenluft et al. [
9] conceptualised chronic irritability as severe mood dysregulation (SMD), a category generated to allow for testable comparisons with BP. SMD is characterised by chronic and severe irritability with frequent and developmentally inappropriate temper outbursts, along with negatively valenced mood in between outbursts.
Follow-up studies contradicted the notion that severe irritability is an early manifestation of mania. Children with SMD at the age of 10 years suffered from unipolar depression, but not BP at the age of 18 years [
7]. Stringaris et al. [
10] reproduced this finding using a 20-year follow-up study. However, these studies were of community samples where the rates of BP are low. Therefore, not finding an association between BP and chronic irritability may have been due to the lack of statistical power. To address this, Stringaris et al. [
11] used a referred sample to compare the course of children with SMD and those with BP over a median period of 29 months. The authors found that only one out of 84 (1.2 %) patients with SMD experienced a manic episode over this time period. By contrast, 58 out of 93 (62.4 %) patients with BP had a manic episode over the same time period. Again, such findings argue against the notion that chronic irritability is a characteristic of BP. It will be informative to follow children with SMD further as they pass through the period of maximum risk for BP.
Consistent with these findings, SMD and BP were found to differ in family history [
12], with parents of youth with narrow phenotype BP being significantly more likely to be diagnosed with BP (14/42, 33.3 %) than parents of youth with SMD (1/37, 2.7 %).
Pathophysiological findings
Recently, research has compared whether severe irritability, in the form of SMD, and BP overlap in their pathophysiological mechanisms. SMD and BP youth were both found to display deficits in face emotion labelling. Both patient groups were significantly less accurate in labelling facial emotions [
13] and less sensitive in recognising facial expressions of emotions [
14], compared to healthy controls. However, SMD and BP differ in the brain mechanisms underlying these emotion labelling deficits—a recent functional magnetic resonance imaging (fMRI) study showed that patients with SMD had lower amygdala activity compared to those with BP during a face emotion processing task [
15]. The pattern of amygdala activation in SMD resembled that of young people with depression [
16,
17].
Differences between SMD and BP youth have also been identified in studies using the affective Posner task [
18], an attention task designed to elicit frustration. Rich et al. [
19] showed that although both SMD and BP youth displayed significantly more negative affect than healthy controls in response to negative feedback, patients with SMD differed from those with BP in their event-related potentials (ERPs) [
20] and brain activation patterns [
19]. The pattern of ERPs following frustration suggested executive attention deficits in the BP, but not SMD group, who were impaired in the initial, “bottom-up” attention regardless of emotional context. These findings are consistent with neuroimaging data: in response to blocked goal attainment on the frustration task, relative to SMD youth, BP youth displayed greater activation in the superior frontal gyrus (SFG) [
19], a region implicated in the regulation of executive attention [
21]. Coupled with neuroimaging techniques, frustration tasks offer a promising way of disentangling SMD from BP, by suggesting that the seemingly similar behavioural manifestations of irritability may have different neural substrates.
Related findings come from a series of studies using response reversal paradigms, where participants have to modify their responses to match the environmental changes. These are usually changes as to which stimulus is rewarded and which one is not. Poor performance on these tasks was suggested to lead to frustration [
22]. Although response reversal deficits were more often found in young people with BP (e.g. [
23]), a recent study found that both SMD and BP youth showed deficits on response reversal with compound stimuli [
24]. Emerging data suggest that SMD and BP youth may differ in the underlying pathophysiology of response reversal deficits, with SMD but not BP youth showing reduced inferior frontal gyrus (IFG) activation in response to errors on the task relative to healthy controls [
25]. Of note, the same study found both SMD and BP youth to have reduced task-related caudate activation compared to healthy controls.
Taken together, these findings suggest that pathophysiological mechanisms differ between SMD and BP youth; however, they also show that there are shared pathways between the two conditions. The next important step will be to compare severe irritability with other childhood psychopathology, such as attention deficit hyperactivity disorder (ADHD), and establish their distinguishing neurobiological features. It will also be important to take a dimensional view of irritability. New measures, such as the Affective Reactivity Index (ARI) [
26], could be useful in studying the brain correlates of irritability intensity. This will further help to understand the mechanisms of irritability and may be useful for diagnostic purposes.
An important clinical problem is irritability in young people with autism spectrum disorders (ASD). For good reasons, a diagnosis of ASD was an excluding factor in studies on mood dysregulation mentioned above, and therefore irritability in ASD has been understudied. Emerging evidence points on important differences between how mood problems present in ASD compared to typically developing youth. For example, in contrast to studies of typically developing young people with SMD [
13,
24], similarly defined severe mood problems in ASD youth were not associated with deficits in executive function or emotion recognition (except for identifying the facial expression of surprise) [
27]. Future research should determine which aspects of ASD affect the presentation of mood problems in this clinical group. This may inform the diagnosis and treatment of irritability not only in the ASD population, but potentially also in typically developing youth who display some autistic-like traits, such as rigidity or social interaction difficulties.
Implications for assessment and clinical guidelines
The assessment of irritability has until recently lagged behind other research in the field. In most studies, irritability was measured using items from the oppositional defiant disorder (ODD) section of interviews [
28]. More recently, a concise instrument to measure irritability, the ARI has been developed and validated in clinical samples [
26]. SMD can be assessed using a specially developed module [
11] of the Kiddie Schedule for Affective Disorders and Schizophrenia for School Age Children (K-SADS) [
29].
Treatment implications
In contrast to classical BP [
30], chronic non-episodic irritability lacks appropriate evidence-based treatments, largely due to its unclear diagnostic and aetiological status. Extrapolating treatment options from classical BP to SMD, relying on the premise that severe, chronic irritability constitutes a developmental manifestation of BP, may be harmful. For example, it is unclear whether atypical antipsychotics, such as risperidone, used in the treatment of irritability in autism [
31], would be a viable treatment option for irritability in typically developing youth with SMD. Lithium was already found to be ineffective in SMD [
32]. A trial of a serotonin reuptake inhibitor in children with SMD is ongoing (Leibenluft, personal communication). Clinicians will of course want to treat other comorbid disorders, particularly ADHD, in children presenting with severe irritability. Treatment with methylphenidate was found to reduce symptoms of oppositionality in children [
33], and in our practice and according to anecdotal evidence, it also seems to reduce irritability in those with ADHD. In our experience, cognitive-behavioural therapy (CBT) appears to be a well-tolerated and often effective treatment for children with irritability [
34]; however, no trial data are yet available about the effects of psychological therapy on children’s irritability.