Temperament and Anxiety DisordersAnxiety disorders in childhoodTemperament in childhoodLink between early temperament and childhood anxietyPotential explanatory mechanismsAnimal model for parentingParenting styleParental psychopathologyAttentional or effortful controlNeural underpinningsOutstanding issuesSummaryReferencesTemperament and Anxiety DisordersKoraly Pe´rez-Edgar, PhD*, Nathan A. Fox, PhDUniversity of Maryland, 3304 Benjamin Building, College Park, MD 20742, USAAs a relatively new contributor to the world of empirical science, the field ofpsychology has spent the last century working first to prove its scientific under-pinnings and then to generate an understanding of human behavior that truthfullyportrays the present and accurately predicts the future. This has at times provenquite difficult because the object of study is much less orderly and rule-boundthan the target of other disciplines, such as the atom or the cell. However, aconsensus has arisen in the last decade that much of this empirical frustrationmay have been self-inflicted. By subdividing the field into parsimonious andatomic subdisciplines, researchers have created units that are almost by definitionill suited to capturing the complexity and multidimensionality of human behavior.Leaders in the field are attempting to bridge these gaps by forming multi-disciplinary research groups that can bring together disparate literatures andmethodologies to create a more three-dimensional view of the phenomenonof interest.This article focuses on the atte mpt to link early appearing temperamental traitsto the later emergence of psychopa thology, particularly in the form of anxietydisorders. The discussion defines and characterizes the current understandingof temperament and anxiety as separate constructs; reviews the evidence to datelinking temperament and anxiety; and explores the environmental, cognitive, andneural mechanisms that have been suggested as potential mediators for this effect.The article also highlights the strength of bringing together converging data frommultiple sources and levels of analysis.Anxiety and anxiety disorders can have a large affect on the daily functioningof an individual, coloring interactions with both the environment and personalassessments of internal states. The affect can be particularly damaging if anxiety1056-4993/05/$ – see front matter D 2005 Elsevier Inc. All rights reserved.doi:10.1016/j.chc.2005.05.008 childpsych.theclinics.com* Corresponding author.E-mail address: [email protected] (K. Pe´rez-Edgar).Child Adolesc Psychiatric Clin N Am14 (2005) 681– 706first emerges in childhood and adolescence because this has been linked toincreases in both the severity and longevity of the disorder [1] . As such, re-searchers and clinicians have been keen on identifying factors that may predictthe emergence of anxiety. In this regard, differentiating the symptoms or char-acteristics of the disorder may help researchers delineate the multiple pathways todisease [2].A number of reviews [3–5] have noted a variety of behavioral similaritiesbetween shy or inhibited temperament groups and anxious individuals. Bothgroups are marked by social awkwardness and withdrawal, an avoidant copingstyle, and a constellation of psychophysiologic markers (Table 1). Becausetemperament is early appearing, the construct may help outline early risk factors,even before a disorder is visibly manifested.There are, however, a number of limitations to the potential bridge betweentemperament and anxiety that should be kept in mind when reviewing the dis-cussion. First, our definition of temperament must be further refined andsolidified. As Vasey and Dadds [4] have noted, many of the measures oftemperament in infan cy and early childhood have been rationally rather thanempirically deriv ed. This has led to some confusion within the temperamentliterature regarding the core characteristics of a particular temperament trait, theobserved behavioral phenotype, its developmental concomitants, and its impacton socioemotional development.Table 1Defining characteristics shared by anxiety and behavioral inhibitionBehavior CharacteristicsOverly sensitive dangerdetection systemsAnxious and behaviorally inhibited individuals show a tendency tofeel frightened by objects or situations that most individualsexperience as nonthreatening.Attentional bias to threat Individuals monitor the environment for potential threat. In addition,anxious or inhibited individuals detect and respond to ‘‘threat’’ cuesat lower thresholds. This may lead the individual to find theenvironment more subjectively threatening.Avoidant coping style Having detected an environmental threat, anxious and behaviorallyinhibited individuals often respond by withdrawing from the situationand avoiding the trigger both at that moment in time and in futureencounters.PsychophysiologicpatternsElectroencephalographic asymmetry: likely to show greater activationin the right frontal lobeStartle responses: greater potentiated startle to threat cuesHeart rate and heart rate variability: show high heart rate and lowheart rate variability at restPupil dilation: show greater dilation during cognitive tasksSalivary cortisol: tend to show higher levels of stress hormone atrest and after provocationOver-reactive amygdala The preliminary definition of behavioral inhibition was based onbehaviors linked to amygdalar activity. Recent functional MRIstudies have documented increased amygdala activation to threateningand salient stimuli for both clinical and temperament groups.pE´rez-edgar & fox682Second, even with a better-defined construct, it is unlikely that research willreveal a clear linear relationship between early emerging traits and later anxiety.Developmental changes often occur as a result of reciprocal interactions betweenan active child and his or her environmental context, making the child both theproducer and product of the environment [6]. As such, attempts to draw a linkfrom early temperament to the later emergence of psychopathology must conten dwith the fact that a difficult temperament may push a child in the direction of anynumber of developmental outcomes (multifinality), and the targeted outcome canresult from a host of predisposing pathways (equifinality). Research musttherefore account for a number of potential moderating factors that can come intoplay at various points throughout development.Third, just as temperament must be rigorously defined, our