Search

Irritability is one of the most common reasons for referral to child and adolescent mental health services and is a strong predictor of current and future functional impairment. Furthermore, irritability is one of the few psychiatric symptoms that is present in both internalising and externalising disorders. In children, irritability is part of the diagnostic criteria of every mood disorder described in the DSM-5, as well as of oppositional defiant disorder (ODD). It is also listed as an associated symptom for conduct disorder (CD), attention-deficit/hyperactivity disorder (ADHD) and autism spectrum disorders (ASD). Recently, irritability as a disorder itself has been introduced in the DSM under the term disruptive mood dysregulation disorder (DMDD). Despite the omnipresence of irritability across distinct psychiatric disorders, we still know little about its predictors, longitudinal outcomes, mechanisms, and, most importantly, effective treatment approaches. In this thesis, I examine the predictors and longitudinal correlates of irritability in young people. I also test candidate cognitive mechanisms and treatments for irritability driven by results of previous research. Specifically, I examine the potential mechanisms that might explain the specific association between irritability and depression found in longitudinal studies and test a new treatment approach motivated by this association. First, I present a systematic review and meta-analysis of the longitudinal correlates of chronic severe irritability. I searched for studies in PubMed and Web of Science that investigated DMDD or the irritability dimension of ODD as a predictor of future psychiatric disorders and symptoms. The findings from this study highlight the specific association between chronic severe irritability and future depressive and anxiety disorders. 4 Second, I examine early predictors of irritability that might explain its specific association with depression. Using a large community-based sample, I tested whether an infant's sex and variation in physiological stress reactivity (i.e., two known risk factors for depression) interact to predict different patterns of ODD symptom dimensions, including headstrong and irritability, in the pre-school age. We predicted that girls would be at higher risk for irritability whereas boys would be at higher risk for headstrong symptoms. The results show that stress reactivity predicts ODD symptoms in opposite directions in boys and girls. However, this pattern was the same for both ODD dimensions, without specific risk for girls and boys. Third, using both cross-sectional and longitudinal analyses, I test whether deficits in emotion recognition, commonly seen in youth with irritability and youth with depression, are associated with an increase in the risk of depressive symptoms in children and adolescents with DMDD. The results show that emotion recognition deficits in youth with DMDD are associated with co-occurring depressive symptoms and predict these symptoms at follow-up. Fourth, I provide the results of the first-ever randomised controlled pharmacological trial (RCT) that has used irritability as a primary outcome in youth with severe mood dysregulation (SMD), the precursor of DMDD. Given the specific association between irritability and depression/anxiety, I examine whether adding citalopram to methylphenidate improves irritability to a greater extent than adding placebo. The results of the RCT show that add-on citalopram improves irritability in youth with SMD. However, improvements of irritability are not accompanied by improvements in functional impairment, depressive symptoms or anxiety symptoms. 5 Finally, I propose a model to explain the association between irritability and future depressive disorder. To do so, I refer to both findings reported in this thesis and results from previous works. The model also includes aspects that are hypothesised to play a role in this transition but still need to be tested. The hypothesised model is based on a shared factors model, in which factors contributing to irritability are shared with those contributing to depression. I discuss genetic, environmental and early life factors, as well as aberrant responses to reward and emotional stimuli. I conclude this thesis by proposing some directions for future research.

Major Depressive Disorder (MDD) is a common, recurrent and disabling mental illness which is poorly treated by currently prescribed drug therapies. The discovery of treatment tools that target putative mechanisms of illness in depression is thus a clinical priority. Depression is characterised by hyporeactivity to reward and hyperactivity to aversive stimuli, which putatively reflects altered function in fronto-striatal-limbic brain regions innervated by monoamines. Yet, very few studies have used dopaminergic drugs to probe the association between neural reward- and especially penalty- signalling and depression. Such intervention designs are important for overcoming the limitations of correlational studies through randomization and experimental manipulation. Preliminary findings raise the intriguing possibility that dopamine antagonists with antidepressant properties may exert their effects via reward and/or penalty signal normalisation, however further studies are warranted. This thesis aims to address this knowledge gap by exploiting the advantages of a placebo-controlled design to examine how a novel D2 antagonist with antidepressant properties, lurasidone, modifies activity in the brain’s reward/penalty network in depression. We recruited 43 medication-naïve participants across the range of depression severity (Beck’s Depression Inventory –II score range: 0-43), including healthy volunteers, as well as people meeting full-criteria for MDD. In a double-blind placebo-controlled cross-over design, all subjects received either placebo or lurasidone (20mg) across two visits separated by one week. Functional magnetic resonance imaging (fMRI) with the Monetary Incentive Delay (MID) task assessed reward functions via neural responses during anticipation and receipt of gains and losses. The analyses focused on these two phases of reward processing as well as medication and depression effects on Prediction Error (PE), the brain’s key dopaminergic learning signal encoding the difference between reward or loss outcome and their anticipation. We hypothesised that subjects scoring high on depression would show a baseline difference in fronto-striatal activity which would be reverted by acute-dose lurasidone. Moreover, we sought to address a key concern in pharmacoimaging studies, namely that shifts in global or regional CBF could underlie changes observed in BOLD fMRI signal. We therefore also used arterial spin labelling (ASL), an imaging modality that allows the quantification of cerebral blood flow at rest, to disentangle global and regional CBF changes from BOLD fMRI signal. As such, this was the first investigation examining the acute effects of lurasidone in the human brain (across a spectrum of depression severity), on a well validated neuroimaging reward task, together with a concerted attempt to control for known potential confounds. Our findings showed that lurasidone altered fronto-striatal activity during anticipation and outcome phases of the MID task without modification of behaviour. There was a significant three-way Medication-by-Depression severity-by-Outcome interaction in the anterior cingulate cortex (ACC) after correction for multiple comparisons. Follow up analyses revealed significantly higher ACC activation to Penalty Outcomes in high- versus low depression participants in the placebo condition, with a normalisation by lurasidone. We found an opposite pattern of signal normalisation for Reward Outcomes in the ACC and Nucelus Accumbens (NAcc). Lurasidone enhanced ACC and NAcc signalling to positive feedback in depressed individuals, however, this pattern did not remain significant after stringent correction for multiple ROI comparisons. Instead, we found that lurasidone significantly increased NAcc activation in individuals with higher symptoms of anhedonia. For the PE analyses, we found support for a normalisation in reward-related PE encoding in the amygdala and penalty-related PE encoding in the ACC in one of the three PE models tested. Finally, sensitivity analyses demonstrated that comorbid anxiety symptoms, self-reported changes in sedation and state anxiety and increased striatal CBF under lurasidone did not confound lurasidone’s effects on reward and penalty processing in depression. Taken together, an acute dose of lurasidone normalises (reduces) neural ACC responses to negative outcomes and PE encoding, without modification of behaviour in individuals with elevated depressive symptoms. Lurasidone also normalises (increases) striatal (NAcc) responses to positive feedback as a function of anhedonia severity. Potential mechanisms at the receptor level are discussed with reference to activity at Dopamine D2 and Serotonin 5-HT1A, 5-HT2A, and 5-HT7 receptors. The results provide evidence for abnormalities in neural reward-penalty systems in depression and highlight the potential of targeted pharmacological treatments to normalise penalty and reward-related processing in depression. The thesis brings increased knowledge and precision to our understanding of the effects of lurasidone, during different phases of reward and penalty processing across the continuum of depression and anhedonia severity.

Previous research indicates that psychosocial risks and contextual risks such as poverty can have detrimental effects on children’s development. In this thesis I use data from the Avon Longitudinal study of Parents and Children (ALSPAC) to examine prospective relationships between risk factors present early in life and later child outcomes related to cognitive functioning, psychopathology, and brain structure. In Chapter 1 and Chapter 2 I describe theories and methods behind the research presented in this thesis. Chapter 3 examines how maternal depression, contextual risks, and interpersonal stress early in life relate prospectively to children’s cognitive outcomes at the age of 8 years. Chapter 4 introduces a metaTanalysis of brain imaging studies I conducted to identify brain regions showing reduced metabolism in depression. Chapter 5 examines how early adverse experiences within the first six years of life relate to children’s internalising symptoms during childhood, and how early adversity and internalising symptoms relate to variation in grey matter structure in early adulthood in the regions of interest identified in Chapter 4. In Chapter 6 I examine how preT and postnatal psychosocial adversity (prenatal maternal stress, childhood adversity, and peer victimisation) relate to psychosisTlike experiences in adolescence, and how psychosocial adversity and psychosisTlike experiences relate to altered white matter microstructure in adolescence. Finally, Chapter 7 presents a summarising discussion of the general themes emerging from my research, as well as limitations and future directions for this field of research. The results presented in this thesis indicate that different types of risks during prenatal and early postnatal development relate to poorer child outcomes, including poorer cognitive functioning, increased levels of psychopathological symptoms, and altered grey and white matter. These findings have implications for intervention since they highlight the importance of intervening early in life.

Beyond core impairments in social interaction skills and restricted, repetitive behaviours, young people with autism spectrum disorder (ASD) often also suffer from debilitating mood and anxiety problems. The reasons for this overlap are poorly understood and often complicated by difficulties in symptom reporting in youth with ASD. This thesis used a multi-method approach to investigate two possible mechanisms underlying mood and anxiety problems in ASD youth, and examined the usefulness of neuroimaging methods (arterial spin labelling) to aid the detection of mood states without relying on clinical report. We first investigated symptom reporting of irritability and physiological stress responses in boys with high-functioning ASD and typically-developing controls. Boys with ASD reported reliably on their irritability, and showed reduced cortisol and heart rate responsiveness to stress compared to controls. Cortisol hypo-responsiveness in boys with ASD was associated with irritability, while lower heart rate was associated predominantly with anxiety. Second, we examined the interplay between ASD traits and anxiety during a neuroimaging reward paradigm in 1472 adolescents from the community. Shared and unique neural correlates of anxiety were found in those with high vs. low ASD traits; while specific brain activations during reward anticipation predicted future, new-onset anxiety in participants high on ASD traits. Symptoms of depression and irritability had minimal impact on the results. Third, arterial spin labelling was sensitive to experimentally-induced mood states in healthy youth. It showed some usefulness for detecting specific mood states, whereby sad and happy moods were distinguished from neutral, but not from each other, based on brain activation patterns alone. Our results suggest that some pathophysiological mechanisms of anxiety may be different in ASD youth compared to controls, bringing important clinical implications. The effects of anxiety on cortisol stress responsiveness, but not heart rate responsiveness or reward processing, may be better explained by co-occurring irritability.