An Expanded Conceptual Framework for Understanding Irritability in Childhood: The Role of Cognitive Control Processes

Irritability is not only associated with poorer functioning concurrently (Beauchaine & Cicchetti, 2019; Elvin et al., 2021), but also predicts poorer longitudinal outcomes (i.e., development of psychopathology; Copeland et al., 2015; Vidal-Ribas et al., 2016). Given that irritability is one of the most common reasons youth are referred for psychiatric care (Collishaw et al., 2010; Evans et al., 2022; Stringaris, 2011), researchers are increasingly seeking to understand the mechanisms that underlie irritability (Brotman et al., 2017a, 2017b) and elucidate key targets for intervention (Kircanski et al., 2019). First, we review current definitions and transdiagnostic perspectives of irritability. Following this, we present the proposed expanded conceptual framework and integrate the literature in relation to the key tenets in the irritability, including (A) the environmental and intraindividual vulnerability factors implicated in youth with heightened irritability; (B) the ‘setting’ factors that have a bearing on the expression and phenomenology of irritability; (C) the differential pathways of positive and negative system activation in line with the RDoC perspective; (D) the underlying cognitive control processes that are maladaptively activated under threat and reward conditions; and (E) the varying outcomes of the expression of irritability and its longitudinal course. Finally, we explore the implications of this framework for theory and intervention and provide directions for future research.

While childhood irritability is relatively common and can be associated with poor concurrent and longitudinal outcomes, only recently have researchers attempted to understand the processes underlying the development of irritability (i.e., Brotman et al., 2017a, 2017b). Alongside more recent adaptations (Kircanski et al., 2019), aberrant threat and reward processing have been specified as key psychological mechanisms within these models. Indeed, the translational model of irritability proposed by Brotman et al., (2017a, 2017b) first provided insights into the components of threat and reward processing that interact and underlie childhood irritability. Kircanski et al. (2019) then included inhibitory control in their treatment model, providing further insight into factors thought to underlie irritability. Some studies (Brænden et al., 2022; Myers et al., 2017) have also explored irritability in the context of the Research Domain Criteria (RDoC) perspective (Insel et al., 2010; NIMH, 2021). Although this prior research has allowed for progression in understanding of the mechanisms involved in irritability and potential treatment targets, gaps critical for future treatment models remain. In particular, the cognitive control processes that underlie threat and reward biases, the proximal conditions that ‘set the scene’ and trigger the activation of these processes underlying threat and reward biases, and how they may be linked to differences in the expression (i.e., tonic and/or phasic) and progression (i.e., to DMDD or co-occurring psychopathology) of irritability remains poorly understood. Further, there are key differential pathways of RDoC negative and positive valence systems, that once activated appear to play a pivotal role in the ability of youth to flexibly recruit cognitive (i.e., proactive and reactive) control. Integrating these perspectives, our review aims to explore the cognitive processes that underlie and maintain the mechanisms of threat and reward processing in childhood irritability. In particular, we build on past frameworks to examine both tonic and phasic expressions of irritability and the processes that underlie threat and reward processing through the application of existing cognitive control theories (i.e., Braver, 2012) and the RDoC domains. As cognitive control develops over the first ten years of life (Munakata et al., 2012; Troller-Renfree et al., 2020), this review aims to focus on irritability within middle childhood, however studies examining samples of youth within early childhood over a longitudinal course can also provide further understanding about how these mechanisms may arise and develop over time. Further, although irritability does present in the context of many disorders (i.e., ODD), here we focus specifically on studies with irritability as an individually measured construct or in the context of disorders with irritability as its key feature (i.e., DMDD). To this end, we aim to shed further light on childhood irritability via a proposed framework which integrates prior theoretical and empirical research and seeks to guide future research and treatment.

Presented in Fig. 1A, the model highlights vulnerability factors thought to increase a child’s susceptibility to experiencing irritability. Similar to prior studies (Brotman et al., 2017a, 2017b; Kircanski et al., 2018), we have highlighted evidence specific to vulnerabilities for irritability, however we also draw upon transdiagnostic vulnerability factors common across many childhood psychopathologies given that irritability presents across the spectrum of externalising and internalising disorders (i.e., American Psychological Association, 2000). Such factors include a combination of environmental factors (e.g., parent factors), and intrapersonal factors (e.g., genetics, protective personality characteristics) thought to predispose youth to experiencing irritability.

Setting factors (Fig. 1B), i.e., contextual or immediate precursors that ‘set the scene’ for engagement of underlying processes have not been explicitly outlined in prior models of irritability. However, some researchers have considered conditions under which threat and reward biases thought to underlie irritability may be intensified, such as when irritable youth are acutely angry or frustrated (i.e., Deveney et al., 2013; He et al., 2013). The setting factors proposed here are interrelated circumstances which appear to exacerbate or ‘activate’ the positive and negative valence systems, leading to maladaptive employment of cognitive control and the expression of irritability. Setting factors (Fig. 1B) include internal factors (i.e., current mood) and environmental demands (i.e., task requirements) that may influence outcomes alone or in amalgamation, in activating situations (e.g., within the classroom, playing with peers) and/or resulting from changes in internal states (e.g., as a child becomes more tired or hungry).

Although this review provides an account of the research to-date of factors thought to increase an individual’s vulnerability to experiencing irritability (‘vulnerability factors’; Fig. 1A) and the more immediate contextual factors that may activate or exacerbate maladaptive responding (‘setting factors’; Fig. 1B), these factors are also thought to be interrelated. Indeed, although some studies refer to irritability as an underlying disposition or ‘trait’ (i.e., Beauchaine & Tackett, 2019), other studies have positioned irritability as a state-like concept arising from momentary experiences (Toohey & Digiuseppe, 2017). More recently, irritability has been described as a distinct factor arising from early temperamental vulnerabilities that is multifaceted and influenced across development (i.e., Silver et al., 2023a, 2023b; Sorcher et al., 2024). Indeed, the combination of specific vulnerability factors in circumstances or ‘settings’ which may lead to activation of positive and negative valence systems may then exacerbate a child’s predisposition to experiencing irritability. For example, although harsh parenting was highlighted as a vulnerability factor (Wiggins et al., 2014), children with more severe irritability appear to evoke harsher and more hostile parenting styles, indicating a bidirectional relationship (Hakulinen et al., 2013; Kiff et al., 2011). As such, although negative parenting (i.e., shouting and smacking) has been found to increase irritability in middle childhood, negative parenting also appears to be driven by greater irritability in young childhood (i.e., 4 years of age; Oliver, 2015). Together, the bidirectional influences of child temperament and parenting appear to interact to influence child outcomes and adjustment. Moreover, Lengua (2006) suggested that irritable and difficult to manage temperamental traits in children engender frustration in parents and thus lead to more inconsistent parenting approaches (and in turn promoting incompatible setting conditions; Fig. 1B), which may for example lead to changes in parental reward contingencies and foster negative mood states. Thus, although we have outlined both some key vulnerability factors and setting factors implicated in childhood irritability, it is important to note these factors are interrelated, as depicted by the bidirectional arrow in Fig. 1. Early vulnerability factors and setting factors also appear to be interrelated in the case of both threat and reward expectancies in irritable youth.

Threat expectancies. Differences in pre-existing threat and reward expectancies of irritable youth compared to non-irritable youth may also contribute to activation of negative and positive valence and cognitive control systems, such as via early vulnerability factors and activation via setting factors. For example, some evidence points to irritable youth being more likely to inaccurately perceive the intensity of emotional stimuli, suggesting that pre-existing threat expectancies influence the performance of irritable youth (i.e., Levy et al., 2022; Stoddard et al., 2016), and may also influence a child’s mood state. Indeed, prior studies have suggested that highly irritable youth may automatically identify stimuli as more threatening or hostile than non-irritable youth, reflecting a hostile attribution bias (Crick & Dodge, 1994; Leibenluft & Stoddard, 2013). Although some studies have also suggested that irritable youth demonstrate more inaccuracies in face emotion labelling in addition to requiring a greater degree of facial emotional intensity before they can accurately identify the emotion when compared to healthy controls, research in this area is sparse and has relied on irritability in the context of Severe Mood Dysregulation (SMD; characterised by severe irritability and low mood) (Guyer et al., 2007; Rich et al., 2008). Some irritable youth have been found to view neutral faces as more hostile or threatening (Crick & Dodge, 1994; Leibenluft & Stoddard, 2013; Stoddard et al., 2016), with hostile attribution biases suggested to arise from early vulnerabilities, such as parent-related factors (Griffith et al., 2021; Lee et al., 2017). In general, youth appear to preferentially process unexpected information, consistent with studies suggesting a general expectancy bias in individuals (Proulx et al., 2017). Thus, taken together, these findings reflect a vulnerability for irritable youth to preferentially attend to threat information as a result of the activation of expectancy-violating information, such as through the attribution of hostile intent. When such threat expectancies are activated (such as in specific setting conditions), the activation of negative valence systems is likely to occur.

Reward expectancies. Similar to threat expectancies, differences in pre-existing reward expectancies of irritable youth may also contribute to activation of positive and negative valence systems. Although initial errors in reward (or punishment) predictions may arise from early vulnerabilities as previously described (i.e., inconsistent parenting strategies may lead to children first developing erroneous expectancies about conditions under which they may receive a reward), youth prone to irritability also demonstrate deficits in their reinforcement learning over time i.e., probabilistic and reinforcement learning; reward prediction error (Brotman et al., 2017a, 2017b; Brotman et al., 2017a, 2017b; Meyers et al., 2017). For example, Dickstein et al. (2007) examined responses on a task involving youth aged 7 to 18 years learning to adjust to changes in previously reward responses, finding that youth with BD and SMD made more errors compared to healthy controls when tasks involved reversal learning. Thus, a child’s erroneous reward expectancies are likely to lead to situations that ‘set the scene’ for the activation of negative and positive valence and cognitive control systems in response to reward stimuli (Perlman et al., 2015; Rich et al., 2011), such as pre-existing beliefs about whether or not they will receive a reward and the development of habitual responding to previously rewarded situations. Taken together, early vulnerability factors (e.g., parenting-related factors) combined with setting conditions under which a child may expect to receive or not receive a reward may lead to overactivation or underactivation of the positive valence system, or overactivation of the negative valence system. These conditions of reward expectancies may be further compounded by setting factors, such as a child’s mood state, priming irritable youth to have increased difficulty making judgements in conditions of possible reward or threat, thus also evoking activation of positive and negative valence systems.

Numerous models have been proposed to explain overlapping aspects of self-regulation of emotion, cognition, and behaviour and as a result, a more unified approach to terminology has been recommended (see Nigg, 2017 for review). Therefore, we aim to examine the underlying cognitive control processes of irritable youth, rather than complex aspects of self-regulation or control of emotions and behaviour more broadly, such as those proposed in models of emotion regulation (i.e., Ochsner et al., 2012) and the RDoC domains of arousal/regulatory systems (NIMH, 2021).

Arising from the development of top-down executive functions (i.e., working memory, response selection; Nigg, 2017), cognitive control is thought to develop across childhood by the age of 10 years (Munakata et al., 2012) and allows for active maintenance of cognitive, behavioural, and emotional goals (Miller & Cohen, 2001; Nigg, 2017). Although in the present review, the focus is on conceptualising irritability specifically within middle childhood, given nascent research in this area, we also draw upon studies specifically examining irritability and DMDD from preschool to adolescence.

Recent conceptualisations of the processes leading to irritability implicate underling system activations via evidence of both threat and reward biases (Brotman et al., 2017a, 2017b). In the model proposed by Kircanski et al. (2019), reduced inhibitory control (as a facet of cognitive control) was suggested as an underlying cognitive process associated with greater expression of frustration and behavioural outbursts, thus implicating a broader self-regulation difficulty impacting goal maintenance. Further, Kryza-Lacombe et al. (2022) described the mitigating role of executive functioning (cognitive flexibility and inhibitory control) on emotion regulation in reward biases among irritable youth. Recently, Brænden et al. (2022) highlighted key themes as related to the RDoC domains that are implicated in irritability, including underlying negative valence systems (i.e., frustrative nonreward), positive valence systems (i.e., reward anticipation and responses), social systems deficits (i.e., interpretation differences, biases toward threat faces) and cognitive systems (i.e., cognitive control, inhibition). However, highlighted within Brænden et al’s (2022) review was equivocal results given the intersection of cognitive systems with many of the RDoC domains (i.e., positive and negative valence systems, systems for social processes) thought to underlie irritability.

Thus, the proposed model extends upon the earlier models and reviews and aims to conceptualise cognitive control in terms of proactive and reactive processes that are proposed to be influenced by positive and negative valence system activation and underlie the processing of threat and reward stimuli (Miller & Cohen, 2001; Nigg, 2017) (Fig. 1). That is, common themes across the literature on both threat and reward biases in irritable youth suggest differences in initial responding, allocation of attention, sustained attention, interferences in attention, and response inhibition, arising from inflexible employment of specific top-down and bottom-up cognitive processes. Thus, in explicating the underlying cognitive control processes, we apply Braver's (2012) Dual Mechanisms of Control (DMC) framework in the context of the RDoC domains of negative (i.e., threat biases) and positive (i.e., reward biases) valence systems and cognitive control processes, as characterised within Fig. 1.

Tonic and phasic irritability. As noted, recent conceptualisations of irritability have suggested interrelated yet distinguishable expressions of irritability, namely tonic and phasic irritability (Copeland et al., 2015). Tonic irritability (Fig. 1, pink border) has been described as persistently grouchy or grumpy mood, with youth being touchy or easily annoyed. Conversely, phasic irritability (Fig. 1, blue border) is defined as developmentally inappropriate outbursts or tantrums (Copeland et al., 2015; Vidal-Ribas & Stringaris, 2021). Although tonic and phasic irritability have been identified as unique and discrete constructs, they are also interrelated in that phasic irritability has been linked to the development of tonic irritability, and tonic irritability has been linked to the development of phasic irritability (Copeland et al., 2015; Moore et al., 2019). Indeed, both tonic and phasic expressions of irritability have been associated with disruptions in functioning and poorer outcomes for youth (Copeland et al., 2015), associated with different patterns of psychopathology (as seen in Fig. 1E).

Outcomes. Depicted in Fig. 1C are the positive and negative valence systems, as related to the underlying cognitive processes of proactive and reactive control, and tonic and phasic irritability that are both observed as core components in DMDD. Although DMDD has been more closely associated with phasic presentations of irritability (Moore et al., 2019), youth who have three or more developmentally inappropriate temper outbursts per week (i.e., phasic irritability) in addition to persistent irritable mood between outbursts (i.e., tonic irritability) satisfy diagnostic criteria for DMDD (American Psychological Association, 2000). Given this interconnected nature of tonic and phasic irritability, DMDD at the age of 6 years old predicted significantly lower global functioning and increased comorbid psychopathology, including increased symptoms of ADHD, disruptive behaviour disorders, and depression (Dougherty et al., 2016). Further, persistently irritable youth (both tonic and phasic irritability) demonstrate greater risk for depression at 13 years old (Orri et al., 2019) and higher suicidality risk in adolescence (Forte et al., 2021; Orri et al., 2019).

Despite irritability being a core feature of DMDD, not all children with high irritability satisfy diagnostic criteria for DMDD. Certainly, both tonic and phasic irritability have been observed in the context of ODD, which when compared to DMDD is considered less severe with respect to irritability symptoms. As illustrated in Fig. 1E, youth with ODD may present with either phasic or tonic irritability, rather than requiring the presence of both (American Psychological Association, 2000). In addition to also displaying phasic irritability, youth with internalising disorders (i.e., depression and anxiety) more commonly exhibit tonic irritability (as presented in pink in Fig. 1), as irritable mood is a common symptom of these disorders within the DSM-5 (American Psychological Association, 2000; DeGroot et al., 2023; Stringaris, 2011). Indeed, within an adolescent sample, youth with higher tonic irritability were observed to develop later depressive disorders and anxiety (Silver et al., 2021). However, some recent evidence has suggested greater impacts on functioning for youth with selective mutism who display phasic forms of irritability (Freitag et al., 2023), indicating the need for further studies parsing these constructs in the context of youth psychopathology. On the other hand, youth with ADHD demonstrate greater phasic irritability (i.e., increased outbursts and aggression) rather than persistently irritable mood (Cardinale et al., 2021; DeGroot et al., 2023). Moreover, youth with more severe phasic expressions of irritability demonstrated associations with the development of externalising disorders (Silver et al., 2021), as presented in blue in Fig. 1. Therefore, alongside varying concurrent patterns of psychopathology with tonic and phasic irritability, youth also demonstrate associations with psychopathology and poorer outcomes in the longer-term. Although tonic and phasic irritability are depicted in Fig. 1 as associated with patterns of psychopathology and underlying cognitive processes, current research suggests that these constructs are not mutually exclusive and that instead they commonly co-occur (Copeland et al., 2015; Moore et al., 2019). Indeed, although both tonic and phasic irritability are interrelated, and together they increase the risk of poorer youth outcomes, separately these expressions of irritability may also be associated with distinct outcomes (Leibenluft & Kircanski, 2021). Further work is still required to understand the unique longitudinal outcomes associated with both tonic and phasic irritability in childhood.

Although the mechanisms of threat and reward biases have been previously been included in pathophysiological models of irritability in youth (Brænden et al., 2022; Brotman et al., 2017a, 2017b; Kircanski et al., 2019), the differential pathways of positive and negative valence system activations that may drive maladaptive activation of cognitive control (i.e., proactive and reactive control) processing differences among irritable youth is not clearly understood or explained. As such, we present a novel and expanded framework of irritability in which we have drawn together a summary of evidence regarding vulnerability factors for irritability (Fig. 1A) and how this likely interacts with setting conditions (Fig. 1B), thus leading to differential positive and negative system activation (Fig. 1C) and accentuating differential recruitment of cognitive control processes (Fig. 1D) and contributing to the expression and differential progression of irritability (Fig. 1E).

Findings point to clear processing differences for irritable compared to non-irritable youth under threat and reward conditions. However, these conditions also appear to interrelate, with potential rewards increasing an individual’s motivation to recruit more cognitively demanding methods of control (i.e., proactive control), such as via positive valence system activation or overactivation. On the other hand, the presence of threat and reward appears to reduce the capacity of irritable youth to employ proactive and reactive control effectively and flexibly (i.e., negative valence system overactivation; positive valence system underactivation). Indeed, although the availability of rewards may enhance the ability of irritable youth to engage in tasks, once contingencies change or rewards are removed, activation of the negative valence system (i.e., via frustrative nonreward) leads to more difficulty for youth to engage flexibly in both proactive and reactive control and thus is likely to lead to phasic outbursts. Moreover, highly irritable youth also demonstrate an overall requirement for increased neural resources in order to employ proactive control, likely leading to increased tonic irritability. These cognitive control differences therefore explain the differences in outcomes and expression observed among irritable youth compared to non-irritable youth. In particular, when interacting within a social context, irritable youth likely trigger hostile interactions with their peers (Brotman et al., 2017a, 2017b; Forte et al., 2021), likely arising from their sensitivity to threats and rewards and their reduced capacity for flexible cognitive control processing in the context of threats or rewards. Such differences evoking poor interactions with peers may involve a higher sensitivity to minor threats (negative valence system overactivation; e.g., name calling), greater competition with peers when a potential reward is involved (positive valence system overactivation; e.g., difficulties with turn taking or impulsivity), increased frustration when not receiving a desired reward (frustrative nonreward; e.g., phasic expressions of irritability when not winning a game), or reduced motivation when there are no potential rewards (e.g. positive valence system underactivation; reduced effort in challenging tasks) and longer-term tonic irritability. As such, these pathways of system activation impacting on effective and flexible recruitment of proactive and reactive control are likely to exacerbate early vulnerabilities (such as parent–child relationships and hostile attribution biases) and lead to ingrained and maladaptive responses to proximal setting factors (i.e., fatigue, mood state). Over time, this perpetuation of activation of underlying systems and processes is likely to lead to greater functional challenges for irritable youth and longer-term psychopathology. Illustratively, Sorcher et al. (2022) found that early irritability increases the risk for social difficulties in the longer-term. Indeed, Barclay et al. (2022) found that low social skills mediated the interaction between childhood irritability and adolescent internalising problems. In a sample of 739 8- to 11-year-olds, physical victimisation and peer rejection were found to exacerbate symptoms of ADHD and irritability, both concurrently and in the longer-term (Bellaert et al., 2023). Moreover, Barker and Salekin (2012) suggested that youth with high irritability and peer victimisation at age 10 was associated with increased callus-unemotional traits and internalising difficulties at age 13. Such interactions with peers over time combined with reduced capacities to self-regulate emotional and behavioural control, likely serve to increase risk for psychopathology in the long term (Orri et al., 2019).

Moreover, differences in underlying cognitive control processes may help to explain multifinality associated with early irritability symptoms. For example, Kessel et al. (2016) examined longitudinal associations of parent-reported irritability at age 3, with performance on a go/no-go task with continuous EEG monitoring at age 6. Higher irritability at age 3 and enhanced negative change in error-related negativity (ERN) at age 6 predicted more internalising symptoms at age 9. Conversely, blunted change in ERN at age 6 predicted more externalising symptoms among irritable youth. Such findings may arise due to underlying pathways of system activation (i.e., positive valence, negative valence, and cognitive systems) in which irritable youth may inflexibly engage cognitive control modes in an attempt to reduce potential interference from threats or inhibit phasic irritability, and as a result likely experience more tonic forms of irritability (pink border, Fig. 1). Intensive threat or behavioural monitoring over time is likely associated with increases in internalising symptoms in the longer term (i.e., depression). Alternatively, some youth who are prone to irritability may experience difficulties monitoring their performance and/or recognising the impact of their behaviours on others (perhaps resulting from difficulties recognising the emotions of others, particularly at low intensities) (Guyer et al., 2007; Rich et al., 2008), and may thus display more phasic irritability (i.e., temper tantrums or outbursts; blue border, Fig. 1) and externalising difficulties in the long term. These challenges with downregulation intersect with the wider RDoC domains of arousal/regulatory systems (i.e., arousal) and systems for social processes (i.e., perception and understanding of others) (Insel et al., 2010). Thus, differences in irritable youth with respect to their preferential recruitment of cognitive control processes during goal-directed tasks may be associated with variability in the expression of irritability and its associated psychopathology, and poorer functional outcomes and longer-term psychopathology.

Building on prior proposed models and conceptualisations of irritability (Brænden et al., 2022; Brotman et al., 2017a, 2017b; Kircanski et al., 2019) this expanded theoretical framework highlights reward and threat processing as putative treatment targets. At the same time, it also recognises the implications of various setting factors, some of which may be intractable, and their role in stoking how irritable youth respond to environmental stimuli and demand triggers. This understanding in turn may lead to purposeful and specific investigations into irritability and how it contributes to psychopathology in youth across varied environments, elucidating targets for the treatment of irritability in children. Indeed, given gaps in our current understanding of aspects of childhood irritability (e.g., the role of protective personality characteristics, the role of expectancies in threat processing), this framework further highlights avenues for needed investigation. At the same time, this theoretical and conceptual review has limitations that warrant further discussion and need for future research. Although we have examined irritability in a variety of contexts, such as in clinical samples of DMDD, community samples, and in the context of co-occurring psychopathology (i.e., anxiety), many childhood disorders which have irritability as a symptom have not been included here. Such disorders (i.e., PTSD, childhood mania) require subsequent investigation regarding the applicability of this theoretical conceptualisation. Moreover, given that this is indeed a theoretical review, tasks designed to directly investigate underlying cognitive processes such as proactive and reactive control (e.g., the AX-CPT; Braver, 2012; Gonthier et al., 2016) are needed in order to provide empirical data to support or falsify the proposed framework. As a result, future research should consider tasks that directly map onto these underlying cognitive processes to better assess this theoretical framework and understand their recruitment in the context of various setting factors. Specifically, youth with heightened irritability demonstrate a general reliance on reactive control, leading to increased distraction from task goals, particularly under conditions of threat (i.e., overactivation of the negative valence system) or when setting factors are negatively impacting (e.g., task demands increase, increased frustration). When irritable youth are presented with rewards, these youth demonstrate more proactive control methods likely due to overactivation of the positive valence system, however this can be inflexible or require greater recruitment of neural resources to employ (and thus limit their capacity for continued recruitment of proactive control), particularly under conditions which activate frustrative nonreward. At the same time, it remains unclear the varying degrees to which setting factors and potential threats and rewards may interact to influence youth with heightened irritability—that is, how combinations of positive and negative setting factors together may influence cognitive control capacity, and how this may be expressed with regard to tonic and/or phasic irritability and its progression over the longer term. Consequently, further understanding of these interactions between setting conditions leading to activation of these systems, the recruitment of cognitive control, and associations with tonic and phasic presentations of irritability in both community and clinical samples, may prove advantageous for our understanding of interventions that may reduce irritability in youth. Additionally, further research into the longitudinal associations between these underlying cognitive processes and the development of psychopathology in youth is required to understand how these cognitive processes may map onto youth with less severe trajectories of irritability (Orri et al., 2019; Pagliaccio et al., 2018; Wiggins et al., 2014). Such future research may also consider wider application of underlying cognitive control processes to youth irritability, such as the systems for social processes (i.e., RDoC domains mediating interpersonal settings; Insel et al., 2010). Indeed, understanding these pathways in relation to underlying cognitive control processes will help highlight critical opportunities for future treatment research.

Advancing our understanding of the processes underlying irritability provides pathways for novel treatment research (i.e., targeting underlying attention biases; Brotman et al., 2017a, 2017b). As such, attention training may provide a fruitful avenue for the treatment of irritability in youth by modifying attention biases for threat (such as in the treatment of anxiety disorders; Waters et al., 2013, 2015). Further, exposure to frustration-inducing conditions may provide opportunities to modify reward contingencies (i.e., expectancies) and condition more effective methods of regulating emotional and behavioural outbursts (Kircanski et al., 2019). Indeed, recent studies (Grossman & Ehrenreich-May, 2020; Hawks et al., 2020) have found preliminary evidence for modular cognitive behavioural therapy (CBT) as a transdiagnostic intervention that may assist in the treatment of irritability. Naim et al. (2023) found further preliminary evidence for exposure-based CBT with parent-management training for severe irritability in children, with results suggesting stronger effects in targeting temper outbursts compared to irritable mood. Although these results highlight the utility of exposure-based approaches in the treatment of irritability, setting factors, differential activation of positive and negative valence systems, and inflexible and ineffective cognitive control recruitment are key factors that appear to interact to negatively influence poorer youth outcomes and expressions of irritability and thus further treatment avenues are also implicated. Specifically, in training youth to attend more flexibly to positive environmental cues (i.e., via positive search training), youth may be less susceptible to negative setting factors that may hinder their performance and thus improve more effective engagement of cognitive control whilst engaging in a goal-directed activity (i.e., reduced negative valence system overactivation). Further, some recent studies have highlighted mindfulness training as an effective approach to enhancing flexible cognitive control engagement. For example, in a sample of 94 undergraduate students, a mindfulness intervention group (compared to non-intervention and a passive relaxation group), demonstrated more flexible engagement in proactive and reactive control modes on the AX-CPT at post-intervention (Chang et al., 2018). Although these findings were within a sample of undergraduate students, mindfulness training shows some promising treatment indicators for irritable children. Illustratively, children with DMDD have demonstrated reduced irritability following Dialectical Behaviour Therapy which incorporates mindfulness training (Perepletchikova et al., 2017). Thus, findings point to interventions which include mindfulness components as one possible avenue for bolstering treatment for irritability in youth. Future research is required to fully investigate the efficacy of mindfulness training in order to characterise changes in flexible recruitment of proactive and reactive control in irritable children. Although prior research has highlighted the value in psychopharmacology for irritability, such as in ADHD (Coghill et al., 2014), direct targeting of cognitive control processes (i.e., recruitment of proactive and reactive control) in irritability through medication management requires further investigation. In particular, some studies (Posner et al., 2011; Schulz et al., 2018) have found psychostimulants to disrupt neural connections that may underlie emotional processing within adolescent and adult ADHD samples, thus calling for the need to examine the impact of these treatment approaches for irritable youth. Taken together, our framework highlights key opportunities for further investigation into the treatment of irritability in youth.

The expanded conceptualisation of irritability proposed here provides a needed framework for understanding how underlying cognitive control processes of proactive and reactive control linked with setting and vulnerability factors through the activation of positive and negative valence systems contributes to outcomes for irritable youth. Based on research spanning clinical and community populations of irritability in childhood, findings point to differential recruitment of proactive and reactive control in the context of negative and positive system activation.

In particular, variations in employing proactive and reactive control appear to arise resulting from interactions between early vulnerability factors and setting factors, including internal mood state (Deveney et al., 2013; He et al., 2013), environmental demands (Adleman et al., 2011; Liuzzi et al., 2020; Toohey & DiGiuseppe, 2017), and expectancies (Dickstein et al., 2007; Proulx et al., 2017) that can lead to system activations (positive or negative). Given the influence of these setting conditions, irritable youth differentially recruit proactive and reactive control when working towards a potential reward, in conditions of nonreward, or in responding to potential threats. Further, increased cognitive load in monitoring environmental distractions and threats and reducing socially unacceptable behavioural responses in irritable youth likely increases more tonic irritability and reduced capacity to effectively employ proactive control in order to meet task goals. At the same time, reduced flexible recruitment of proactive and reactive control as required likely leads to outbursts of frustration (phasic irritability) when presented with negative feedback or the loss of potential or earned rewards due to poorer performance on tasks. Developmentally, irritability in combination with continued inflexible recruitment of cognitive control processes that may be incongruent with the environmental demands, likely contributes to poorer outcomes in youth, such as psychopathology (Kessel et al., 2016). All told, this conceptualisation offers a guiding framework for the investigation of irritability in children and highlights key avenues for treatment research.