Universal prevention of anxiety and depression in school (PDF)

ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2017 Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Social Sciences 148 Universal prevention of anxiety and depression in school children JOHAN ÅHLÉN ISSN 1652-9030 ISBN 978-91-513-0153-2 urn:nbn:se:uu:diva-333033  Dissertation presented at Uppsala University to be publicly examined in Betty Pettersson- salen (14:031), Blåsenhus, von Kraemers allé 1, Uppsala, Friday, 19 January 2018 at 13:00 for the degree of Doctor of Philosophy. The examination will be conducted in English. Faculty examiner: PhD Laura Ferrer Wreder (Department of psychology, Stockholm University). Abstract Åhlén, J. 2017. Universal prevention of anxiety and depression in school children. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Social Sciences 148. 81 pp. Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-513-0153-2. Anxiety and depression are common in children and adolescents, and involve individual suffering, risk of future psychiatric problems, and high costs to society. However, only a limited number of children experiencing debilitating anxiety and depression are identified and receive professional help. One approach that could possibly reduce the prevalence of these conditions is universal school-based prevention aimed at reducing the impact of risk factors and strengthening protective factors involved in the development of anxiety and depression. The current thesis aimed to contribute to the literature on universal prevention of anxiety and depression in children. Study I involved a meta-analysis of earlier randomized, and cluster- randomized trials of universal prevention of anxiety and depression. Overall, the meta-analysis showed small but significant effects of universal preventive interventions, meaning that lower levels of anxiety and depression were evident after intervention completion and partially evident at follow-up assessments. No variables were found to significantly enhance the effects, however, there was a tendency for larger effects to be associated with mental health professionals delivering the interventions. In Study II, a widely adopted prevention program called Friends for Life was evaluated in a large school-based cluster-randomized effectiveness trial. The results showed no evidence of an intervention effect for the whole sample. However, children with elevated depressive symptoms at baseline and children with teachers who highly participated in supervision, seemed to benefit from the intervention in the short term. Study III involved a 3-year follow-up of Study II and an examination of the effects of sample attrition. The results showed no long-term effects for the whole sample and no maintenance of the short-term subgroup effects observed in Study II. Finally, to increase our understanding of the development of anxiety in children and to assist future improvements of universal prevention, Study IV evaluated different trajectories of overall anxiety together with related patterns of disorder- specific symptoms in a school-based sample over 39 months. Evidence favored a model of three different developmental trajectories across age. One trajectory was characterized by increasing levels of overall anxiety, but fluctuating disorder-specific symptoms arguably related to the normal challenges of children’s developmental level, which warrants an increased focus on age-relevant challenges in universal prevention. The four studies provide further understanding of the overall effectiveness of universal prevention of anxiety and depression in children, the short- and long-term effects of universal prevention in a Swedish context, and ideas for further development of preventive interventions. Keywords: universal prevention; anxiety; depression; school children; cluster-randomization; long-term effects; developmental trajectories Johan Åhlén, Department of Psychology, Box 1225, Uppsala University, SE-75142 Uppsala, Sweden. © Johan Åhlén 2017 ISSN 1652-9030 ISBN 978-91-513-0153-2 urn:nbn:se:uu:diva-333033 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-333033)  List of Papers This thesis is based on the following papers, which are referred to in the text by their Roman numerals. I Ahlen, J., Lenhard, F., & Ghaderi, A. (2015). Universal Preven- tion for Anxiety and Depressive Symptoms in Children: A Meta- analysis of Randomized and Cluster-Randomized Trials. The Journal of Primary Prevention, 36, 387-403. II Ahlen, J., Hursti, T., Tanner, L., Tokay, Z., & Ghaderi, A. (2017). Prevention of Anxiety and Depression in Swedish School Children: a Cluster-Randomized Effectiveness Study. Prevention Science. Advance online publication. III Ahlen, J., Lenhard, F., & Ghaderi, A. (submitted manuscript). Long-term Outcome of a Cluster-Randomized Universal Preven- tive Intervention targeting Anxiety and Depression in School Children. IV Ahlen, J., & Ghaderi, A. (submitted manuscript). Disorder-spe- cific symptom patterns in trajectories of general anxiety: A lon- gitudinal prospective study in school aged children. Reprints were made with permission from the respective publishers.   Contents Introduction.................................................................................................9 Anxiety in children...............................................................................10 Clinical features...............................................................................10 Course and onset..............................................................................11 Prevalence and consequences..........................................................13 Biological perspectives....................................................................13 Temperament...................................................................................14 Information processing....................................................................15 Emotion regulation ..........................................................................16 Parental influences...........................................................................17 Depression in children..........................................................................18 Clinical features...............................................................................18 Course and onset..............................................................................18 Prevalence and consequences..........................................................19 Biological perspectives....................................................................19 Temperament...................................................................................20 Cognitive vulnerabilities..................................................................20 Parental depression..........................................................................21 Negative life events and childhood adversities ...............................22 Overlap between anxiety and depression.............................................23 Gender differences...........................................................................24 Why prevention of anxiety and depression..........................................25 Universal prevention of anxiety and depression..............................26 Gaps in evidence and research needed ............................................28 Friends for Life................................................................................30 The empirical studies ................................................................................33 Aims .....................................................................................................33 Method .................................................................................................33 Participants ......................................................................................33 The intervention...............................................................................34 Measures..........................................................................................35 Procedure.........................................................................................38  Study I - Universal prevention for anxiety and depressive symptoms in children: A meta-analysis of randomized and cluster-randomized trials......................................................................................................42 Background......................................................................................42 Method.............................................................................................42 Results .............................................................................................43 Discussion........................................................................................43 Study II - Prevention of anxiety and depression in Swedish school children: A cluster-randomized effectiveness study ............................45 Background......................................................................................45 Method.............................................................................................45 Results .............................................................................................46 Discussion........................................................................................47 Study III - Long-term outcome of a cluster-randomized universal preventive intervention targeting anxiety and depression in school children.................................................................................................48 Background......................................................................................48 Method.............................................................................................48 Results .............................................................................................49 Discussion........................................................................................50 Study IV - Disorder-specific symptom patterns in trajectories of general anxiety: A longitudinal prospective study in school-aged children.................................................................................................51 Background......................................................................................51 Method.............................................................................................51 Results .............................................................................................52 Discussion........................................................................................53 General discussion ....................................................................................54 Is there justification for universal prevention?................................54 Friends for Life in Sweden ..............................................................55 Future directions in universal prevention ........................................58 Thinking outside the box .................................................................60 Turning to targeted interventions?...................................................60 Limitation of the current thesis........................................................61 Final conclusions .............................................................................62 Acknowledgements...................................................................................63 References.................................................................................................64  Abbreviations ACME Average Causal Mediation Effect ANOVA Analysis of Variance AP Academic Performance BI Behavioral Inhibition CDI-S Children’s Depression Inventory – Short version C-RCT Cluster-Randomized Controlled Trial DSM Diagnostic and Statistical Manual of Mental Disorders EC Effortful Control FFL Friends for life ICC Intra Cluster Correlation LCGMM Latent Class Growth Mixture Model LIX Läsbarhetsindex LMM Linear Mixed Model MHP Mental Health Professional MINI-kid Mini International Neuropsychiatric Interview for Children NA Negative Affectivity NNT Numbers Needed to Treat PA Positive Affectivity PH Physiological Hyperarousal PRP Penn Resiliency Program RAP Resourceful Adolescent Program RCI Reliable Change Index RCT Randomized Controlled Trials SCAS Spence Children’s Anxiety Scale SCAS-P Spence Children’s Anxiety Scale – Parent version SCAS-12 Spence Children’s Anxiety Scale – 12 item version SDQ Strength and Difficulties Questionnaire  8  9 Introduction Mental illness is the leading cause of disability in children and adolescents globally (Erskine et al., 2015). A recent report from the Swedish National Board of Health and Welfare (2013) concluded that mental illness has in- creased in young people during the last two decades, and the latest information at hand indicates a continued increase. Young people particularly reported in- creased anxiety, which may be a precursor of future mental illness and suicide attempts. A growing fraction of mentally ill children and youths constitute a major public health problem, which is why early interventions have received specific attention in the Swedish government’s mental health strategy for the years ahead (National Board of Health and Welfare, 2017). The evidence base for prevention of mental illness in Sweden is scarce. In a review by the Swedish Council on Technology Assessment (2010), it was concluded that no prevention program targeting mental illness in children was adequately evaluated to be considered as evidence-based in a Swedish context. Consequently, an important aim for prevention research in Sweden is that pro- grams are proven to be effective in a Swedish context, as preventive programs often yield mixed results between countries (Sundell, Ferrer-Wreder, & Fra- ser, 2014; Swedish Council on Technology Assessment, 2010). One program of interest, and highlighted in the Swedish Council on Technology Assess- ment review (2010) is Friends for Life (FFL), a prevention program developed in Australia aimed at reducing anxiety and depression in children. The efficacy and effectiveness of FFL has been evaluated in several countries. A pilot study of FFL was also performed in Sweden (Ahlen, Breitholtz, Barrett, & Gallegos, 2012) where it showed preliminary evidence of reducing depressive symp- toms, and improving general mental health. The empirical studies in the current thesis all aimed to contribute to the existing literature on the effectiveness, and further development of preventive interventions of anxiety and depression in children. More specifically, these four studies embraced the overall effectiveness of universal prevention of anx- iety and depression, the short and long-term effectiveness of a universal pre- ventive intervention in a Swedish context, possible mechanisms of the effects, important methodological aspects of the design, and developmental issues that are important for the further development of preventive intervention. To in- troduce the reader to the field, a summary of recent research on anxiety and depression in children, alongside related prevention research, is provided be- fore the empirical studies are presented.  10 Anxiety in children Fear is a basic human emotion, which serves as an adaptive response when facing a threat (Gullone, 2000). The words ‘fear’ and ‘anxiety’ are frequently used interchangeably, but fear is more commonly chosen when describing a response to an objective or real threat, while anxiety concerns a subjective, or perceived threat (Huberty, 2012). Fears and anxiety are common early in life and are a normal part of a child’s development. However, for a considerably large number of children, fears and anxiety become disruptive (Hale, Raaijmakers, Muris, & Meeus, 2008; Muris, Merckelbach, Mayer, & Prins, 2000). Up-to-date models delineating the etiology of maladaptive anxiety in children suggest a complex interaction between biological, psychological, so- cial, and environmental components (Ollendick & Grills, 2016). In the follow- ing sections, clinical features of maladaptive anxiety, the onset, prevalence and consequences of anxiety disorders, alongside biological, psychological, social, and environmental risk factors associated with the development of mal- adaptive anxiety are presented. Clinical features Clinical fears, termed anxiety disorders in the literature, are in general distin- guished from normal fears based on the frequency and intensity of symptoms, persistence over time, and to what extent it affects the child’s life and function (Gullone, 2000). There are several anxiety disorders defined in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM–5; American Psy- chiatric Association, 2013). The most common are separation anxiety disor- der, specific phobia, social anxiety disorder, panic disorder, agoraphobia, and generalized anxiety disorder. In the earlier version of the Diagnostic and Sta- tistical Manual of Mental Disorders (4th ed., text rev.; DSM–IV–TR; Ameri- can Psychiatric Association, 2000), obsessive-compulsive disorder was also classified as an anxiety disorder. However, in the DSM-5 it has been relocated to the new chapter called Obsessive-Compulsive and Related Disorders. Separation anxiety disorder is characterized by an excessive and develop- mentally deviant level of anxiety in separations or anticipated separations from an attachment figure. Separation anxiety disorder is most common in children, but also rather prevalent in adults (Bögels, Knappe, & Clark, 2013). Specific phobia is defined as an evident and exaggerated fear when con- fronting or anticipating a specific object or situation. Specific phobias in chil- dren mainly surround fear of animals (e.g., spiders, dogs) and fear of natural environments (e.g., storms, darkness), but also blood-injection-injury fear (Es- sau, Conradt, & Petermann, 2000; LeBeau et al., 2010) Social anxiety disorder is characterized by disproportional fear and anxiety in social situations where the individual feels noticed, observed or scrutinized. For children, these features have to be observed in situations with peers and  11 not solely in interactions with adults. The recent update in DSM-5 more care- fully stresses that social anxiety disorder typically includes multiple social fears and consequences rather than just a specific fear of speaking or perform- ing in front of others (Heimberg et al., 2014). Panic disorder is characterized by unexpected and repeated panic attacks together with a persistent worry of future panic attacks and their conse- quences. Agoraphobia is defined by an exaggerated fear and engagement in avoidant behavior when confronting or expecting public situations like crowds and public transportation. Contrary to the former version of DSM, panic dis- order and agoraphobia are now separated as two specific disorders (Asmund- son, Taylor, & Smits, 2014). Generalized anxiety disorder includes excessive and persistent worry re- garding several day-to day situations or activities, together with difficulties controlling worry. School, family, significant others’ health, and things going on in the world have been found to be the most common domains of worry in children with generalized anxiety (Jarrett, Black, Rapport, Grills-Taquechel, & Ollendick, 2015). Obsessive-compulsive disorder is characterized by time-consuming obses- sions (i.e., annoying and intrusive thoughts or impulses), compulsions (i.e., repetitive compulsive behaviors or mental rituals) or both. A majority of ex- perts around the world agreed that obsessive-compulsive disorder should be moved from the construct of anxiety to a separate construct in DSM-5, as in- trusive thoughts and repetitive behaviors rather than anxiety are the primary features of the disorder (Mataix-Cols, Pertusa, & Leckman, 2007). Course and onset The overall course of anxiety symptoms across childhood and adolescent years has been described in several studies (e.g., Hale et al., 2008; Olatunji & Cole, 2009). These studies have typically reported a general decrease in anxi- ety symptoms from childhood through adolescence, except for a relatively sta- ble course of social anxiety symptoms (Hale et al., 2008). A comprehensive model of the continuity and change of anxiety symptoms has been presented by Weems (2008). The model suggests that there are core features of anxiety such as worry, avoidance, and somatic symptoms, and, that there are second- ary features (e.g., fear of separation from parents, fears of bodily symptoms) which discriminate between the anxiety disorders as defined in the DSM-5 (American Psychiatric Association, 2013). Weems (2008) suggested that the core features are rather stable over time, whereas the secondary features might vary across age. In part, the variations in secondary features are thought to be due to normative challenges, meaning typical challenges in childhood devel- opment tied to certain ages, for example, separation from caregivers in early school years, or interpersonal interactions in adolescent years. Further, the  12 model suggests that subgroups of children primarily follow four different tra- jectories, characterized by (1) low and stable, (2) high and stable, (3) low and increasing, and (4), high and decreasing anxiety levels across age. Numerous recent studies have examined the evidence of different trajectories of anxiety in children (e.g., Allan et al., 2014; Duchesne, Larose, Vitaro, & Tremblay, 2010; Feng, Shaw, and Silk, 2008; Weeks et al., 2014). A couple of these studies have found support for the proposed trajectories as presented by Weems (2008) (e.g., Duchesne et al., 2010; Feng et al., 2008). However, other studies have found another pattern, characterized by rather homogeneous lev- els of anxiety at the early ages and diverging trajectories over time (e.g., Allan et al., 2014; Crocetti, Klimstra, Keijsers, Hale, & Meeus, 2009; Letcher, San- son, Smart, & Toumbourou, 2012). Evidence that supports the association be- tween anxiety and normative challenges comes from studies of disorder onset, and studies showing that symptoms of social anxiety typically increase over time in school-aged samples, whereas symptoms of separation anxiety typi- cally decrease (Weems & Costa, 2005; Westenberg, Gullone, Bokhorst, Heyne, & King, 2007). Anxiety disorders often have their onset early in life. In a large American nationally representative adult sample, retrospective reports yielded a median age of 11 years regarding the overall onset of anxiety disorders (Kessler et al., 2005). Kessler and colleagues (2005), however, found substantial differences between anxiety disorders, where specific phobia and separation anxiety dis- order had their median onset at age seven, social phobia at age 13, obsessive- compulsive disorder at age 19, panic disorder at age 24, agoraphobia at age 20, and generalized anxiety disorder at age 31. In a longitudinal study, includ- ing a large sample of children and adolescents up to the age of 21, a mean onset before age ten was found for separation anxiety disorder, specific pho- bia, generalized anxiety disorder, and social phobia. However, a later onset was found for agoraphobia and panic disorder (Costello, Egger, Copeland, Er- kanli, & Angold, 2011). The continuity and change of anxiety disorders (rather than anxiety symp- toms) have also been examined in several studies (Costello et al., 2011). In a longitudinal study, Last, Perrin, Hersen, & Kazdin (1996) found that children were typically free from their baseline anxiety disorder after three years, but commonly met criteria for another anxiety disorder (so called heterotypic con- tinuity). Further evidence of heterotypic continuity was found in a recent study by Lieb et al. (2016), who found that specific phobia in childhood largely pre- dicted panic disorder, generalized anxiety disorder, and obsessive-compulsive disorder in young adults. Finally, some support for homotypic continuity have also been found especially regarding separation anxiety disorder, and social anxiety disorder (Bittner et al., 2007).  13 Prevalence and consequences Anxiety disorders are the most prevalent psychiatric disorders in children worldwide (Polanczyk, Salum, Sugaya, Caye, & Rohde, 2015). Prevalence rates differ substantially between studies, with point-prevalence studies using strict criteria (including structured questions of distress and impact of symp- toms) showing lower prevalence rates (e.g., 3.7%; Ford, Goodman, & Meltzer, 2003). Lifetime-prevalence and retrospective cross-sectional studies typically provide significantly higher prevalence rates (e.g., 31.9%; Merikangas et al., 2010a). A good estimation based on a longitudinal prevalence study following three large cohorts of children between the ages of 9 and 16 showed a cumu- lative prevalence of 9.9% before the age of 16 (Costello, Mustillo, Erkanli, Keeler, & Angold, 2003). The prevalence seems to differ between age-groups. Copeland, Angold, Shanahan, and Costello (2014) found a relatively high prevalence of anxiety disorders in children aged 9-10 years, but a much lower prevalence in children aged 11-12 years, but thereafter, an increasing preva- lence through adolescence. The high prevalence at younger ages was due to a large number of separation anxiety disorders, and the increased prevalence in adolescence was due to an increased number of generalized anxiety, panic, and agoraphobia disorders. In recent years, anxiety disorders have been identified as the sixth leading cause of disability worldwide (Baxter, Vos, Scott, Ferrari, & Whiteford, 2014), and have been found to be as disabling as depression and oppositional defiant disorder in children and adolescents (Ezpeleta, Keeler, Erkanli, Cos- tello, & Angold, 2001) including adverse effects on several life domains like school functioning (i.e., Weeks, Coplan, & Kingsbury, 2009) peer relation- ships (i.e., Danzig et al., 2013) and family life (i.e., Ezpeleta et al., 2001). Furthermore, anxiety disorders in children have been found to predict future anxiety disorders and depression (i.e., Bittner et al., 2007), alcohol and drug abuse and suicidal behavior in adolescence (Marmorstein, White, Loeber, & Stouthamer-Loeber, 2010; Woodward, & Fergusson, 2001) and health, finan- cial, and interpersonal problems in young adulthood (Copeland et al., 2014). Finally, anxiety disorder in children also involves significant costs to society (Bodden, Dirksen, & Bögels, 2008; Snell et al., 2013). Biological perspectives Not surprisingly, research indicates that childhood anxiety is influenced by both genes and the environment (Gregory & Eley, 2011). Twin studies have typically found evidence of a genetic influence on anxiety, but the effect dif- fers between sources of reporting (e.g., parent ratings vs. self-reports), and the child’s gender and age (Ask, Torgersen, Seglem, & Waaktaar, 2014; Gregory & Eley, 2007). For example, studies using parent reports of the child’s anxiety  14 have found that the genetic influence is more central than environmental fac- tors, while studies using children’s self-reports have suggested the opposite (Ask et al., 2014; Thapar & McGuffin, 1995). A general suggestion is that about one-third to one-half of the variation in etiology is influenced by genes (Fisak & Grills-Taquechel, 2007; Spatola et al., 2007). The genetic influence on anxiety is complex to study, because, there are probably plentiful genes involved which interact with each other, and the environment (Smoller, Block, & Young, 2009). Although recent research has found a strong association be- tween anxiety and specific gene regions (e.g., Otowa et al., 2016), replicating studies, and especially those involving children, have found very mixed results (Gregory & Eley, 2011). The neural foundation of childhood anxiety is not comprehensively ex- plored, and the understanding of the fear circuitry in humans is largely based on functional neuroimaging studies in adults (Blackford & Pine, 2012). In summary, adult studies suggest that anxiety disorder may be characterized by a hyperactive amygdala, and a hypoactive prefrontal cortex (Diekhof, Geier, Falkai, & Gruber, 2011; Etkin & Wager, 2007). Neuroimaging studies in chil- dren have shown similar locations of abnormalities in the brain compared to studies on adults (Mana, Martinot, & Martinot, 2010). However, contrary to adults, an increased (rather than decreased) activity has been found in the pre- frontal cortex in children with anxiety disorders (Blackford & Pine, 2012). Temperament Temperament is defined as a heritable, biologically based alteration that af- fects the character and behavior of an individual (Lonigan, Phillips, Wilson, & Allan, 2011). Although the distinction between temperament, personality, and psychopathology is not completely clear (Lonigan et al., 2011; Muris & Ollendick, 2005), temperament is assumed to (partially) underlie and predis- pose anxiety disorders. Two models of temperament relevant to the develop- ment of anxiety disorder are briefly described in this section; (1) affective re- activity and effortful control and (2), behavioral inhibition. According to Rothbart and Rueda (2005), temperament refers to affective reactivity and self-regulation processes which alter this reactivity. In short, affective reactiv- ity is divided into two higher-order dimensions labeled negative affectivity (NA) and positive affectivity (PA). NA includes feelings of sadness, anger and fear, and these feelings activate, for example, avoidant behaviors. PA includes feelings of happiness and activeness, which activates approaching behaviors (Rothbart & Rueda, 2005). Additionally, a third higher-order dimension la- beled Effortful Control (EC) includes attention shifting and inhibitory control, which alter negative emotions and activate coping strategies to reach long- term goals (Lonigan et al., 2011; Rothbart & Rueda, 2005). In a longitudinal study, Lonigan, Phillips and Hooe (2003) found evidence supporting the no- tion that NA predicted change in anxiety symptoms in children. Further,  15 Meesters, Muris, and van Rooijen (2007) showed that NA was positively as- sociated, and EC negatively associated with anxiety symptoms, and, addition- ally, that NA and EC had an interactive effect on anxiety symptoms in chil- dren. More recent research has showed that the strength of the association with NA varies between different anxiety disorders and that the association be- tween NA and EC is valid only for generalized anxiety disorder, separation anxiety disorder, panic disorder, and agoraphobia (Lonigan et al., 2011). Re- cent research has also found that in children displaying high NA, EC only work as a regulator at low, but not at high, stress levels (Gulley, Hankin, & Young, 2016). Behavioral inhibition is characterized by high levels of physiological arousal and behavioral avoidance to novel or unfamiliar situations, persons, or objects (Degnan, Almas, & Fox, 2010; Fox & Pine, 2012). The expression of BI is considered to vary between age groups, from motor reactivity and negative emotions in babies, crying and clinginess in toddlers, quietness and shyness in preschoolers, to avoidant behavior in social contexts in childhood and adolescents, and to cautiousness and restraint in conversations in adult- hood (Hirshfeld Becker et al., 2008). About 15% of all children are thought to display BI, and these children have been found to be at increased risk of anx- iety disorders, especially social anxiety disorder (Muris, van Brakel, Arntz, & Schouten, 2011). Referring to studies of the neurobiology of anxiety, Pérez- Edgar and Fox (2005) and Fox and Pine (2012) have suggested that the tem- perament of BI involves reacting immediately to threats (i.e., a hyperactive amygdala). Information processing A considerable amount of research has been conducted on the relationship between anxiety and information processes (Field, Hadwin, & Lester, 2011). Two forms of information processing biases commonly addressed in the anx- iety literature are: (1) attentional bias and (2) interpretation bias. Attention bias is characterized by a hyper-attention towards threatening or fearful stim- uli in the environment (Muris & Fields, 2008). A recent meta-analysis showed that an attention bias towards threat is found in all children, but significantly greater in anxious children compared to non-anxious children (Dudeney, Sharpe, & Hunt, 2015). Field and Lester (2010) have proposed a model delin- eating the development of attention bias, which suggests that children gener- ally display attention bias to threats early in life, but over time, children de- velop in different ways. This model has been supported by Dudeney and col- leagues (2015) who found that the difference in attention bias between anxious and non-anxious children increased with age. Interpretation bias concerns the tendency to imagine overly threatening in- terpretations of ambiguous stimuli (Field et al., 2011). Several studies have found that children with an anxiety disorder interpret ambiguous stories more