trait anxiety, neuroticism, and the brain basis of...

Trim: 7in × 10in Top: 0.5in Gutter: 0.875inCUUS1847-24 CUUS1847/Armony ISBN: 978 1 107 00111 4 October 14, 2012 18:56

CHAPTER 24

Trait Anxiety, Neuroticism, and theBrain Basis of Vulnerability to Affective

Disorder

Sonia Bishop & Sophie Forster

Studies of the brain basis of “norma-tive” or “healthy” processing of emotion-ally salient stimuli have flourished overthe last two decades. An initial focus onregions implicated in the detection of emo-tionally salient stimuli (Morris et al., 1996;Whalen et al., 1998) has broadened toinclude discussion of mechanisms support-ing regulatory functions (Bishop, Duncan,Brett, & Lawrence, 2004; Davidson, 2002;Ochsner, Bunge, Gross, & Gabrieli, 2002;Kim, Somerville, Johnstone, Alexander, &Whalen, 2003; Phelps, Delgado, Nearing, &LeDoux, 2004). Running in parallel to thisliterature, psychiatric imaging studies havedescribed alterations in brain function acrossa wide range of anxiety and depressive dis-orders (for reviews and meta-analyses seeEtkin & Wager, 2007; Ressler & Mayberg,2007; Shin & Liberzon, 2010; Stein 2009). Thestudy of the brain mechanisms underlyingvulnerability to disorder has, for some rea-son, fallen outside of the primary spotlight.We argue that work of this nature is criti-cal to bridging studies in healthy volunteerand patient groups and to identifying the

pathways through which risk to affective ill-ness is conferred.

Understanding the brain basis of vulner-ability to affective disorder goes hand inhand with a focus on individual variationand, in particular, trait differences in themechanisms supporting the detection andcontrolled processing of emotional stimuli.How do we study trait differences in vul-nerability to anxiety and depressive disor-ders and try to unpack the brain mechanismsthough which these might act? A number ofapproaches have been adopted, with bothshared and unique advantages and limita-tions.

Studies of the brain basis of vulnerabil-ity to affective disorders typically rely onrecruiting nonclinical volunteer samples andthen regressing scores on self-report mea-sures of trait affect or experience of disorder-related symptomatology against indices ofbrain function or structure. It is importantto remember that this approach is correla-tional in nature, and hence no conclusionscan be drawn about the direction of causal-ity. For example, if we find that, in a student

553


554 SONIA BISHOP & SOPHIE FORSTER

population, elevated scores on a measureof neuroticism are linked to poor frontalrecruitment, this could equally plausiblyreflect individuals scoring high on neuroti-cism being less able to cope with envi-ronmental stress, resulting in diminishedfrontal function; individuals with compro-mised frontal function being more likelyto develop a neurotic personality style;both elevated neuroticism and compro-mised frontal function stemming from a pri-mary disruption to neurotransmitter func-tion; or all of these factors in combination.There are also important methodologicalissues pertaining to good practice in con-ducting correlational analyses of brain activ-ity, which are discussed briefly later in thechapter.

A shared positive feature of studies in thisarea is that constructs such as trait anxiety orneuroticism can be examined as continuousfactors, facilitating exploration of their lin-ear and nonlinear relationships with regionalbrain activity. This allows for a more com-plex and potentially accurate picture to bedrawn than one that solely uses DSM cat-egorical assessments of the binary presenceor absence of a given disease state; the lat-ter “diagnostic” approach faces limitationsarising from difficulties in applying cate-gorical cutoffs, high comorbidity betweenmany anxiety and depressive disorders, andpoor diagnostic reliability (Brown & Barlow,2009).

Studies of the brain basis of vulnerabil-ity to affective disorder can be categorizedaccording to the measure used to assessindividual differences in trait affective style.Arguably there are three main categories.The first involves measures derived fromthe clinical literature and normed for usein nonclinical populations to assess individu-als’ tendency to show disorder-related symp-tomatology, affective and cognitive styles.A prominent example is the SpielbergerState Trait Anxiety Inventory (STAI; Spiel-berger, Gorsuch, Lushene, Vagg, & Jacobs,1983), widely used in studies of the cog-nitive correlates of trait anxiety withinnonselected student samples. The second

category uses measures taken from the per-sonality literature, such as the Neuroticismscale from either the Eysenck PersonalityQuestionnaire (EPQ; Eysenck & Eysenck,1975) or the NEO Personality Inventory(Costa & MacCrae, 1992). The third cat-egory focuses on genetic markers thatdiffer among individuals (functional poly-morphisms with two or more common vari-ants) and that have been linked to differ-ences in affect-related behaviors in humansand other species. In this chapter, we focuson the first two of these categories, com-menting only briefly on the third (forfurther discussion, see Chapter 25). Themajority of studies on the brain basis of vul-nerability to affective disorder falling withinthese categories have used measures of traitanxiety (category 1) or neuroticism (cate-gory 2). We use these examples to explorethe state of the field as it stands and toaddress outstanding questions for futureresearch.

Trait Anxiety and Neuroticism:Indexing Vulnerability to AffectiveDisorders through Self-Report

The trait subscale of the Spielberger StateTrait Anxiety Inventory (Spielberger et al.,1983) is a widely used measure of traitpropensity to anxiety. It has been shownto have good concurrent validity, withpatients with anxiety disorders (ADs) scor-ing higher on the STAI trait subscale thancontrols (Bieling, Antony, & Swinson, 1998).Although fewer studies have examined pre-dictive validity, pretrauma STAI trait scoreshave been found to predict levels of post-traumatic stress disorder symptomatologyafter trauma (Weems et al., 2007). How-ever, the STAI has been criticized for havingpoor discriminative validity, with individu-als with major depressive disorder (MDD)also showing elevated scores on the traitsubscale (Mathews, Ridgway, & Williamson,1996). One possibility is that this low dis-criminant validity reflects a poor choiceof anxiety-specific items for the scale. A


TRAIT ANXIETY, NEUROTICISM, AND THE BRAIN BASIS OF VULNERABILITY TO AFFECTIVE DISORDER 555

second is that there is genuine shared vari-ance underlying vulnerability to both ADsand MDD.

Not only are STAI trait scores elevatedin patients with MDD but scores on thisand other anxiety scales, such as the Tay-lor Manifest Anxiety Scale (Taylor, 1953),have also been found to correlate highly withscores on self-report measures of depres-sive symptomatology; for example, the BeckDepression Inventory (BDI; Beck Ward,Mendelson, Mock, & Erbaugh, 1961) andpersonality indices of neuroticism (Luteijn& Bouman, 1988). Neuroticism is charac-terized by a propensity for negative affect(Watson & Clark, 1984). This trait, measuredby widely used instruments such as the NEOPersonality Inventory and the Eysenck Per-sonality Questionnaire, has emerged overthe last century as one of the most widelystudied personality traits (Costa & Mac-Crae, 1992; Eysenck & Eysenck, 1975; John,1990). The relationship with vulnerability toaffective disorder has arguably been inves-tigated more thoroughly for neuroticismthan for any other dimension of personal-ity (Brown, 2007; Brown & Rosellini, 2011,Kendler, Gardner, Gatz, & Pedersen, 2007).There is strong evidence to support not onlyshared variance but also common geneticinfluences among neuroticism, anxiety dis-orders, and depressive disorders (Hettemaet al., 2008; Kendler et al., 2007).

A likely interpretation of the high corre-lations observed among indices of anxiety,depression, and neuroticism is that they tap,at least in part, into a common underlyingtrait (Figure 24.1). According to the populartripartite model, anxiety and depression notonly have a shared component – a propen-sity to negative affect (which arguably mapson to the construct of neuroticism) – butalso unique components of anxious arousaland anhedonia, respectively (Clark & Wat-son, 1991). This conception has led to thedevelopment of the Mood and AnxietySymptoms Questionnaire (MASQ; Watson& Clark, 1991), which aims to measureboth these shared and unique components.Unfortunately the MASQ focuses on “state”

Figure 24.1. Anxiety, neuroticism anddepression: overlapping constructs? Threealternate models. (A). Self-report measures ofanxiety, depression, and neuroticism couldpotentially all be tapping the same singleunderlying trait. (B). Alternatively, anxiety anddepression might be separate components of thebroader trait of neuroticism. In keeping with thisperspective, the NEO-PI-R includes anxiety anddepression as subfactors, or “facets,” ofneuroticism (Costa & McCrae; 1995). (C). Athird theoretical stance, represented by Clarkand Watson’s (1991) tripartite model, asserts thatanxiety and depression not only have a sharedcomponent of negative affect or general distress(potentially corresponding to the construct ofneuroticism) but also unique components of“anxious arousal” (autonomic hyperactivity) and“anhedonic depression” (low positiveaffect).



or current mood levels and not on trait dif-ferences between individuals. Indeed, thescarcity of trait measures of the propensityto anxiety and depression poses a major dif-ficulty for researchers aiming to investigatethe brain basis of these tendencies. Not onlythe MASQ but also the major depressiveinventories – including both the BDI and theCenter for Epidemiologic Studies Depres-sion Scale (CES-D; Radloff 1977) – focus oncurrent levels of symptomatology. This mayexplain why several imaging studies haveused neuroticism as a proxy for vulnerabilityto depression, but doing so inevitably hin-ders attempts to disentangle the extent towhich neuroticism, trait anxiety, and traitdepression involve disruption to common orunique mechanisms at either a cognitive orsystems (regional brain structure and func-tion) level of analysis.

Studying Affective Trait-RelatedDifferences in Cognitive and BrainFunction: Lessons from the LastDecade

Over the last 10 years, much progress hasbeen made in using neuroimaging tech-niques to study individual differences inneurocognitive function; recent challengesand developments set the stage for equiva-lent progress over the next decade. In theearly 2000s, neuroimaging studies of person-ality led to a conceptual shift in the approachto the investigation of human brain func-tion. These studies argued that individualvariation need not simply be treated as asource of noise in group studies of “nor-mative” neurocognitive function but thatbetween-subject differences could be stud-ied in their own right by examining associa-tions between traits such as extraversion orneuroticism and regional brain function (fora review see Canli, 2004).

Although these studies were ground-breaking in advancing the investigationof individual differences in neurocognitivefunction, as with many first steps in a newfield, several of these studies have since been

subject to criticism (Vul, Harris, Winkiel-man, & Pashler, 2009). However, many ofthe issues raised – pertaining to whole-braincorrelational analyses, insufficient correc-tion for multiple comparisons, and biasedselection of regions of interest – can beavoided if investigations of differences inregional brain function associated with per-sonality indices or trait affective style areconstrained to ones that specifically test the-ories derived from the cognitive or socialpsychological literature. An elegant argu-ment for this approach was initially madeby Kosslyn et al. (2002). In this article, Koss-lyn and colleagues take the proposal putforward by Underwood (1975) – that nat-urally occurring individual differences canbe used to test psychological theories andto reveal the structure of psychological pro-cesses, potentially providing greater insightsthan group-based methods – and argue thatthe same logic can be applied to the useof individual differences to investigate thebiological mechanisms that underpin cogni-tive processes. The authors make the casethat there is natural variation around everycentral tendency, that individuals may differin the efficiency and recruitment of mecha-nisms (which can be studied at various lev-els including both regional brain activationand cognitive processing), and that poolinginformation across individuals may be unin-formative or misleading. They also point outthat the main dangers of unguided correla-tional studies can be avoided by theoreticallygrounding the study design and the analy-sis and interpretation of results, with alter-nate theoretical explanations of observedcorrelations being used to generate furtherhypotheses than can in turn be tested. Inthe sections that follow, we illustrate howthe approach proposed by Kosslyn and col-leagues can be applied, using the example ofstudies that have investigated the brain basisof the association between trait anxiety andattentional capture by threat. In addition,we explore if we can ascertain whether neu-roticism shows a similar, potentially com-mon, relationship to the function of thesemechanisms.



Top-downattentional control

mechanisms

Competition among stimulifor representation

Output to responseand memory systems

Bottom-up sensory-driven mechanisms

Figure 24.2. According to the biased-competitionmodel of selective attention (Desimone & Duncan,1995; Kastner & Ungerleider, 2000), top-downattentional control mechanisms, which favortask-relevant stimuli, and bottom-up sensory-drivenmechanisms, sensitive to stimulus salience, jointlydetermine which stimuli are selected for furtherprocessing. Adapted from Bishop (2008) and Kastnerand Ungerleider (2000), with permission.

(a) (b)

TUMOR

APPLE TUMOR

**

Figure 24.3. Two widely used paradigms in the attention to threat literature. (A) In the EmotionalStroop task, participants are asked to name the font color of a word (here green/gray), ignoring itsmeaning, which can be either threat-related or neutral in valence. High trait anxious subjects showRT slowing for threat-related words. (B) In the dot probe task, participants are presented with twowords, followed by a “probe” (the two asterisks presented here) in the location of one of the words.They are typically asked either to indicate when the probe appears or to specify its orientation. Onkey trials, one word is threat- related, and the other is neutral. High trait anxious individuals show RTspeeding when the location of the probe was previously occupied by a threat word (as in the examplehere), suggesting that attention was allocated to the location of the threat word.



Trait Anxiety, Neuroticism, andThreat-Related Biases in SelectiveAttention: Cognitive Models andFindings

Both AD patients and high trait anxious par-ticipants show elevated attentional captureby threat-related stimuli (Mathews & Mack-intosh, 1998). According to biased compe-tition models of selective attention, atten-tional competition is influenced both by“bottom-up” sensory mechanisms that prior-itize the processing of salient stimuli and by“top-down” attentional control mechanismsthat support the processing of task-relevantstimuli (Figure 24.2; Desimone & Duncan,1995; Kastner & Ungerleider, 2000). Stimu-lus valence – the extent to which a givenstimulus is threat or reward related – is animportant dimension of stimulus salience.A number of selective attention tasks havebeen adapted to examine how stimulusvalence, especially threat-relatedness, influ-ences attentional competition. Two notableexamples are the Emotional Stroop andprobe detection tasks (Figure 24.3). In theEmotional Stroop task, participants namethe ink color of a stimulus word while ignor-ing its semantic content. On this task, hightrait anxious individuals show slower colornaming of threat-related words than emo-tionally neutral words; in low trait anxiousindividuals this slowing is reduced or absent(Richards & Millwood, 1989). In the probedetection task, participants are presentedwith two words or pictures (e.g., faces) fol-lowed by a single dot or pair of dots in theposition previously occupied by one of thetwo stimuli. Here, high trait anxious indi-viduals are faster to detect the presence of asingle dot or to determine the orientationof a pair of dots, when the dot probe ispresented in the position previously occu-pied by a threat-related stimulus (Macleod& Mathews, 1988).

These findings have informed cognitivemodels of anxiety that extend the biasedcompetition model of selective attention tospecifically deal with attentional capture bythreat-related stimuli (Mathews & Mackin-tosh, 1998; Mogg & Bradley, 1998). These

models typically propose that anxiety actsby amplifying the signal from a bottom-up preattentive threat detection mecha-nism that biases attentional competition infavor of threat-related stimuli. When thesestimuli are distracters (non- task-relevant),this is held to interfere with the process-ing of target stimuli, as indexed by slowedreaction times (RTs) and/or elevated errorrates. These models have not, in the mostpart, argued for disrupted top-down or con-trolled processing of threat-related stimuli inanxiety.

A number of studies have used vari-ants of the probe detection task to examinewhether attentional biases are also associ-ated with elevated neuroticism (“N”) scores.These have produced rather mixed results.Reed and Derryberry (1995) found evidencefor a correlation between N scores and atten-tional bias toward negative trait adjectivespreviously rated as self-applicable. How-ever, this correlation was only observedat one of three alternate adjective-probestimulus-onset asynchronies (500 ms, not250 ms or 750 ms). In addition, Chan,Goodwin, and Harmer (2007) and Rijs-dijk et al. (2009) failed to find any rela-tionship between neuroticism and perfor-mance on probe detection tasks using socialthreat words and subliminally presentedthreat-related faces, respectively. Differ-ences among these studies in the choiceof stimuli, stimulus onset asynchrony SOA,and neuroticism scale (EPQ versus NEO)further complicate interpretation of thesefindings.

Interestingly, the limited evidence for anassociation between neuroticism and atten-tional bias toward negatively valenced stim-uli parallels similarly mixed findings withinthe subclinical depression literature. Here,two studies using the Emotional Strooptask have reported increased RT slowingfor color naming of negatively valencedwords as a function of scores on the BeckDepression Inventory (BDI). These effectswere strongest in participants with elevatedBDI scores at two time points a year apart(Williams & Nulty, 1986) and were not foundwhen a depressed mood was induced in par-



ticipants with low BDI scores (Gotlib &McCann, 1984), potentially suggesting therole of an enduring trait conferring vulnera-bility to depression, rather than simply stateaffect. Other studies have found no rela-tionship between individual differences insubclinical levels of depression and atten-tional bias toward negative or threat-relatedstimuli across both the Emotional Stroopand probe detection tasks (Bradley, Mogg,Falla, & Hamilton, 1998; Gotlib, MacLachan,& Katz, 1988; Hill & Dutton, 1989; Hill &Knowles, 1991; Macleod & Hagan, 1992).

One possibility is that a partial corre-late of depression scores, such as trait anx-iety levels, rather than depression itself,might be responsible for the intermittentlyreported attentional interference effects. Asimilar argument can be made for neuroti-cism. The NEO measure of neuroticismcomprises different subfacets, one of whichis especially related to anxiety and another todepression (Costa & MacCrae, 1992; see alsoFigure 24.1). Variability among studies inthe items that high-N participants endorsemight explain the occasional but inconsis-tent findings of attentional biases reportedby studies using this measure.

It is also possible that multiple mecha-nisms contribute to attentional capture bythreat stimuli – with trait anxiety, neu-roticism, and depression potentially sharinga common relationship with one of thesemechanisms but differing in their associa-tion with others. The most obvious can-didate mechanisms are those involved inbottom-up responsivity to threat versus top-down attentional control. Investigation ofthe brain basis of these mechanisms opensup a new door for examining how differenttrait characteristics are linked to the func-tion of these component processes. Specif-ically, it is possible to build on what isknown about the function of different brainregions to explore whether trait anxietyis associated with increased activation ofbrain mechanisms involved in processingstimulus threat value, with impoverishedrecruitment of brain mechanisms involvedin attentional control, or with altered func-tion of both processes. We can also explore

whether neuroticism is linked to a simi-lar pattern of task-specific hyperactivity andhypo-recruitment. Studies that have begunto address this question are reviewed next.

From Networks to Modules and BackAgain

The late 1930s through to the early 1950ssaw the advent of theories that proposedthat networks of brain regions includingareas such as the hippocampus, cingulategyrus, amygdala, and orbital frontal cortexwere responsible for emotional processing(MacLean 1949; Papez, 1937). Support forthe “Papez circuit” and “limbic system” wascountered by criticisms that these accountswere more descriptive than functional andlacked a clear rationale for inclusion of cer-tain regions and the exclusion of others(for more extensive discussion of these crit-icisms,see Chapter 9 in Gazzaniga Ivry, &Mangun, 2009; Early neuroimaging studiesof emotion conducted in the late 1990s andearly 2000s took a more modular approach.Much research was focused on the amygdalaand its role in the detection or evaluationof threat (Morris et al., 1996; Vuilleumier,Armony, Driver, & Dolan, 2001; Whalenet al., 1998, 2004) Indeed, a number of stud-ies from the later part of this era used scanparameters that focused data acquisition ona narrow slab of slices covering the amygdalabut omitting much of the rest of the brain.

In contrast, within the last 5 to 10 years,there has been an increasing focus on theinterplay of regions involved in the evalua-tion of threat stimuli with those that enablethe regulation of emotional state and phys-iological fear responses, the (re)appraisal ofstimuli, and the control of attentional focus(Bishop, Duncan, Brett, & Lawrence, 2004a;Kim et al., 2003; Ochsner et al., 2002; Phelpset al., 2004). This focus has been accom-panied by a shift from examining brainregions in isolation and toward conceptu-alizing regions as nodes within intercon-nected networks, the activation of whichvaries with task engagement and may bemeaningful even at “rest” (Deco, Jirso, &



McIntosh, 2011). Neuroimaging investiga-tions of the association between trait anxietyand brain function and structure have simi-larly evolved across this time period. In theremaining sections of this chapter, we exam-ine what these studies can tell us about thebrain basis of the association between traitanxiety and attentional capture by threat,the shared or distinct relationship with neu-roticism, and the potential common under-pinning of function across other domains ofemotional processing.

Amygdala and Frontal Mechanismsunderlying Attentional Capture byThreat: Hyper- and Hypo-ActivityLinked to Trait Anxiety

Based on findings from the basic neuro-science literature, a relatively widely heldview (which has recently come underrenewed scrutiny; Pessoa and Adolphs,2010; see also Chapter 15) is that adirect subcortico-thalamo-amygdala path-way facilitates the preattentive processing ofthreat-related stimuli (LeDoux, 2000; Tami-etto & de Gelder, 2010). In line with thisposition, a number of neuroimaging stud-ies conducted in the early 2000s reportedthat the amygdala response to threat-relatedstimuli such as fearful faces is not modulatedby the focus of spatial attention (Anderson,Christoff, Panitz, De Rosa, & Gabrieli, 2003;Vuilleumier et al., 2001). These findings lentsupport to the proposition that the amyg-dala might provide the biological under-pinning, or instantiation, of the preatten-tive threat detection mechanism describedin cognitive models of anxiety. Accord-ing to this proposal, amygdaloid activationmight influence the competitive success ofthreat-related stimuli in winning attentionalresources through a gain function analogousto that held to underlie the facilitation ofthe processing of targets by top-down atten-tional control (see Chapter 14).

Further support for this positionappeared, initially, to be provided by find-ings that individuals with elevated anxietylevels show a stronger selective amyg-

dala response to threat-related distracters(Bishop, Duncan, & Lawrence, 2004).However, in that study, it was state ratherthan trait anxiety that showed a relationshipto the amygdala response to unattendedthreat stimuli. In addition, these results donot necessarily indicate that high anxiousindividuals show an increased preattentiveamygdala response to threat-related dis-tracters. An alternate possibility is thatattentional resources may not have beenfully occupied by the primary task, withattentional “spillover” facilitating the pro-cessing of threat-related distracters. Indeed,it has been demonstrated that when theperceptual demands or “load” of the maintask is increased, a differential amygdalaresponse to threat-related versus neutraldistracters is no longer observed (Pessoa,McKenna, Gutierrez, & Ungerleider, 2002);between-participant differences in theamygdala response to threat distracters asa function of anxiety also being eliminated(Bishop, Jenkins, & Lawrence, 2007).

An interesting model, of value in con-ceptualizing these findings, is the load the-ory put forward by Lavie (e.g., Lavie, 2005).Lavie argues that the debate between “early”and late” accounts of selective visual atten-tion (i.e., whether or not the processing ofcertain stimulus characteristics is obligatoryand unconstrained by attentional resources)may be resolved by taking into accountthe perceptual load of the task at handand allowing for two separate stages ofattentional competition. According to thismodel, there is, first, a stage of early per-ceptual competition. The processing of dis-tracters terminates at this stage when theperceptual load of the primary task is high.Second, under conditions of low percep-tual load, competition is held to occurfor further processing resources, includingthe initiation of behavioral responses, withactive recruitment of control mechanismsbeing required to inhibit the processing ofsalient distracters and support task-relatedprocessing. Lavie’s hybrid early/late modelhas primarily been used to account forfindings showing that increasing perceptualload reduces or eliminates the processing



of affectively neutral salient distracters,such as moving dot patterns, lexical stim-uli that promote competing responses tothat required by the current target, and col-orful or novel scenes (Lavie, 2005; Rees,Frith, & Lavie, 1997). It is, however, inter-esting to speculate whether the findingthat the amygdala response to threat dis-tracters is diminished under high load(Bishop et al., 2007; Pessoa et al., 2002)might be consistent with amygdaloid pro-cessing of threat-related stimuli being sub-ject to similar perceptual processing limi-tations that have been found to affect theprocessing of other classes of salient visualstimuli.

An alternate theoretical stance to thatput forward by Mathews and Mackintosh(1998) is that elevated trait anxiety is asso-ciated with impoverished recruitment ofthe frontal mechanisms required for task-focused attentional control. If high traitanxious participants show impaired recruit-ment of attentional control mechanisms,this could result in increased “capture” ofattentional resources by threat-related dis-tracters. As just outlined, Lavie argues thatthe active recruitment of attentional con-trol mechanisms to support the processingof targets and inhibit the processing of dis-tracters is particularly required under con-ditions of low perceptual load to preventsalient distracters from receiving furtherprocessing. In support of this claim, Laviecites findings that groups characterized byweakened attentional control – specificallythe elderly and children – show particularlylarge response competition effects underlow perceptual load conditions (Huang-Pollock, Carr, & Nigg, 2002; Maylor & Lavie,1998). In an interesting parallel, elevatedtrait anxiety was found to be associatedwith diminished activation of the dorsolat-eral prefrontal cortex (DLPFC), ventrolat-eral prefrontal cortex (VLPFC), and ante-rior cingulate cortex (ACC) in response tothreat distracters under conditions of lowbut not high perceptual load (Bishop etal., 2007). This finding suggests that traitanxiety might be linked to impoverishedrecruitment of frontal regions required for

attentional control under task conditionswhere these mechanisms are needed to reg-ulate trial-by-trial fluctuations in process-ing competition from emotionally salientdistracters.

This raises two further questions. First,can we say anything about the relativeattentional control functions of the lateralprefrontal and anterior cingulate corticalregions shown to be under-recruited by hightrait anxious individuals? Second, if traitanxiety is linked to difficulties in the useof these frontal regions to regulate atten-tion, will they also be apparent when dis-tracter salience is unrelated to threat value?In regard to the former question, it has beensuggested that specific subregions of the pre-frontal cortex may play differing roles intop-down attentional control, with the ACCinvolved in detecting the presence of com-petition for processing resources and the lat-eral prefrontal cortex (LPFC) responding toincreased expectation of processing compe-tition by augmenting top-down control tosupport the processing of task-relevant stim-uli (Botnivick, Cohen, & Carter, 2004; seeFigure 24.4 for illustration of the regions con-cerned). Evidence for this account has pri-marily come from studies using response-competition tasks with affectively neutralstimuli (e.g., Carter et al., 2000; Macdon-ald, Cohen, Stenger, & Carter, 2000), includ-ing ones that manipulate the frequency, andhence the expectancy, of high competitiontrials (Carter et al., 2000). Through appli-cation of an equivalent frequency manipu-lation to a task requiring attentional con-trol over threat distracters, it is possibleto investigate whether ACC and LPFCregions show parallel differential responsesto unexpected (infrequent) and expected(frequent) threat-related distracters, respec-tively. This was indeed found to be thecase (Bishop, Duncan, Brett, & Lawrence,2004). Further, the results of this study indi-cated that individuals with high levels ofanxiety showed impoverished recruitmentof both these mechanisms (state anxietyanalyses were reported, similar results wereobserved with trait anxiety, unpublisheddata).



Lateral surfaceDorsolateral

prefrontal cortex(DLPFC)

Parietalcortex

Ventrolateralprefrontal cortex

(VLPFC)

Dorsal anterior(mid) cingulatecortex (dACC)

Rostral anteriorcingulate cortex

(rACC)

Medial wall

Striatum:caudate & putamen

Cerebellum

Figure 24.4. Frontal brain regions implicated in attentional regulation of emotionally andnon-emotionally salient stimuli include the dorsolateral prefrontal cortex (DLPFC), ventrolateralprefrontal cortex, (VLPFC), rostral anterior cingulate cortex (rACC), and dorsal anterior (mid)cingulate cortex (dACC). Subgenual anterior cingulate cortex is not shown here. Adapted withpermission from Bush, 2010.

Intriguingly, a more recent study has pro-duced findings that are discrepant from boththose of Bishop and colleagues (2004) andthose from the earlier response-competitionliterature. Using a “face/word” version ofthe Stroop task, Etkin, Peraza, Kandel, andHirsch (2006) asked volunteers to indicatewhether faces showed fearful or happyexpressions while ignoring the word “happy”or “fearful” superimposed on each face.They reported that activation of the ACC,rather than the LPFC, increased when pro-cessing competition was expected (frequentface/word incongruent trials) and that LPFCactivity increased when processing compe-tition was infrequent or unexpected. Oneinteresting difference from the study byBishop, Duncan, Brett, and Lawrence (2004)is that in Etkin”s task, both the distractersand targets were emotionally valenced, withthe valence of targets and distracters beingbalanced across “high-conflict” and “low-conflict” trials; those conditions differedinstead in face/word congruence. It is diffi-cult to disentangle effects of emotional con-gruency from response congruency in thisdesign. However, this still does not easilyexplain why the conditions under whichthe LPFC and ACC were activated differedfrom those previously observed in both emo-

tional distracter and response-competitionparadigms. One hopes that these discrepantfindings will spark further research that mayhelp resolve and unite these results andadvance our theoretical understanding ofthe role of the ACC and LPFC in the detec-tion and resolution of different types of pro-cessing competition.

A further point worth considering hereconcerns the role of rostral versus dor-sal subdivisions of the ACC (see Figure24.4). Early studies reported dorsal ACCactivity when processing competition arosefrom response competition, and rostralACC activity when processing competitionwas caused by the presence of emotion-ally salient but task-irrelevant distracters(Bishop, Duncan, Brett, & Lawrence, 2004;Bush, Luu, & Posner, 2000). This distinc-tion is also apparent within the psychiatricimaging literature (Bush et al., 1999; Shinet al., 2001). However, more recently thisdistinction has been challenged by a rangeof studies suggesting a role for the dorsalACC in emotional processing – not only inemotional distracter tasks but also in stud-ies investigating the anticipation and expe-rience of pain and the expression of con-ditioned fear (Bishop et al., 2007; Milad,Quirk, et al., 2007; see Shackman et al., 2011,



for a comprehensive review). This also callsfor further investigation of the precise func-tion of this region. The difference in nomen-clature conventions used across groups andthe changes in those terms across time com-plicate this investigation. In this chapter wefollow the nomenclature convention intro-duced by Mayberg and colleagues (1999)and adopted in our earlier work (Bishop,Duncan, Brett, & Lawrence, 2004 , 2007)whereby rostral ACC excludes subgenualACC (the region ventral to the corpus callo-sum). The subdivision referred to here as thedorsal ACC or dACC (see Figure 24.4) haselsewhere been renamed the dorsal anteriormid-cingulate gyrus since this term arguablyreflects more accurately the region underconsideration (Bush, 2010).

The second question raised above con-cerns the nature of the association betweentrait anxiety and the deficient recruitmentof frontal attentional control mechanisms. Isthis association specific to attentional con-trol over threat, perhaps being secondaryto hyper-responsivity of the amygdala tothreat-related stimuli? Or does it reflect amore general dysregulation of frontal atten-tional function that is independent of amyg-dala responsivity to threat? The cognitive lit-erature provides some suggestion that thelatter might be the case, with findings link-ing trait anxiety to impoverished or inef-ficient attentional control (Derryberry &Reed, 2002; Eysenck, Derakshan, Santos,& Calvo, 2007). In a test of the hypoth-esis that trait anxiety is associated withreduced recruitment of frontal attentionalcontrol mechanisms even in the absenceof threat-related stimuli, Bishop (2009)examined DLPFC recruitment while volun-teers performed a response-competition taskunder conditions of low versus high per-ceptual load. High trait anxious volunteersshowed reduced DLPFC recruitment tohigh response-competition trials under con-ditions of low but not high perceptual load,in line with the Lavie model and consis-tent with trait-anxiety-related dysregulationof frontal attentional mechanisms extendingbeyond the specific case of attentional con-trol over threat. Further support for a threat-

independent relationship between trait anx-iety and impoverished recruitment of frontalattentional control mechanisms comes fromthe ERP literature. Using an antisaccadetask (where volunteers must saccade awayfrom the position in which a cue is pre-sented), Ansari and Derakshan (2011) foundthat high trait anxious individuals showedlonger antisaccade latencies together withreduced frontocentral activity during anti-saccade preparation.

The studies discussed here provide someinitial evidence that trait anxiety is linked toimpoverished recruitment of frontal regionsimportant for attentional control both whenprocessing competition is caused by threat-related distracters and when it is caused byresponse conflict. This raises the questionof whether this deficient recruitment can beremediated by cognitive interventions suchas attentional training. A number of early tri-als provide some initial hope that this mightindeed be the case (Amir, Weber, Beard,Bomyea, & Taylor, 2008; Hakamata et al.,2010).

Although work reviewed here suggestsa relationship between trait anxiety andfrontal dysfunction in the absence of task-related differences in amygdala activity(Bishop, 2009), this leaves open the questionas to whether trait anxiety is also linked toamygdala hyper-responsivity to threat in theabsence of differential activation of frontalmechanisms. To address this question, weneed to turn to tasks that require the rel-atively passive processing of threat-relatedstimuli in the absence of demands on atten-tion or other executive processes. Surpris-ingly few such studies exist that meet theconstraints of having examined amygdalafunction and also having measured indi-vidual differences in trait anxiety. In oneearly study, Etkin et al. (2004) reported thattrait anxiety was associated with elevatedamygdala responsivity to masked threat-related faces, but not to unmasked threat-related faces. In a second study, Stein, Sim-mons, Feinstein, and Paulus (2007) reportedthat elevated STAI trait scores were linkedto greater amygdala activity during condi-tions requiring matching of facial emotions



than during conditions requiring matchingof basic shapes.

One intriguing possibility is that an asso-ciation between trait anxiety and amyg-dala responsivity to threat-related stimulimight primarily be seen when the stimuliare ambiguous or require some form of res-olution of their threat value (Whalen, 2007).This is arguably more the case for maskedthan for unmasked threat-related faces andalso potentially the case when different faceswith varying emotional expressions need tobe judged for their equivalence in expres-sion. This is clearly a tentative hypothesis,and it remains to be more firmly establishedunder precisely which conditions trait anxi-ety is linked to amygdala hyper-responsivityto threat. Furthermore, given the possibil-ity that frontal mechanisms may be impor-tant for certain forms of ambiguity resolu-tion (Kim et al., 2003; Nomura et al., 2003),it will also be important to confirm thattrait-anxiety-related differences in amygdalaresponsivity to weak or ambiguous threatstimuli are not secondary to the differentialrecruitment of frontal mechanisms.

Neuroticism and the BrainMechanisms Influencing AttentionalCapture by Threat

As reviewed in previous sections of thischapter, in recent years an increasing num-ber of studies have examined the relation-ship between trait anxiety and recruitmentof the frontal and amygdaloid mechanismsimplicated in attentional control over threatdistracters. This makes it possible to startto ask relatively specific questions aboutwhich parts of the neural circuitry influ-encing attentional capture by threat showaltered function in high trait anxious indi-viduals. Although the corresponding litera-ture on neuroticism is more limited, we canlook at the initial studies available to beginto assess whether neuroticism and trait anx-iety show similar relationships with regionalbrain function – as might be expected ifthey both represent the same single underly-ing construct – or whether neuroticism and

trait anxiety only partially covary in theirrelationship with the function of discretebrain mechanisms as might, for example, bepredicted by Clark and Watson’s tripartiatemodel (see Figure 24.1).

At the time of writing, three studieshave examined the correlates of neuroti-cism while participants perform fMRI tasksinvolving manipulation of selective atten-tion and emotional stimuli. Two of thethree – one using the Emotional Stroop taskand the other the probe detection task –reported no significant association betweenneuroticism scores and regional brain activ-ity during conditions of attentional compe-tition from emotional stimuli (Amin, Con-stable, & Canli, 2004; Canli, Amin, Haas,Omura, & Constable, 2004). However, giventhat the sample size in each case was fairlysmall for a correlational study (12 or fewersubjects), it is difficult to draw strong con-clusions from these null results.

In a subsequent larger study (n = 36),Haas, Omura, Constable, and Canli (2007)administered a Stroop-like task, similar tothat of Etkin et al. (2006), in which tri-als differed in the emotional congruenceof target words and the expressions ofbackground faces. In this study, the wordsdid not map directly onto the names ofthe facial expressions, thereby reducing theassociation between emotional incongru-ency and response conflict. Individuals withhigh neuroticism levels were found to showincreased amygdala and subgenual anteriorcingulate activity to trials with emotion-ally incongruent face/word pairs (collapsingpositive-face/negative-word and negative-face/positive-word trials). This result isintriguing, but difficult to relate to find-ings from imaging studies examining theinfluence of trait anxiety on the neuralmechanisms regulating selective attentionto threat, because the task manipulationused by Haas and colleagues (emotional con-gruency) is not one of distracter threat-relatedness, being orthogonal to distractervalence. One possibility is that the height-ened amygdala responsivity to emotion-ally incongruent stimuli in high-N individ-uals reported here might primarily reflect



sensitivity to stimuli that are ambiguous intheir emotional significance, in line with theproposals put forward by Whalen and col-leagues (Whalen, 2007).

To explore the “ambiguity sensitivity”hypothesis further, we review studies thathave examined the influence of neuroti-cism on regional brain activity to emotionalstimuli using passive viewing or cognitivelyundemanding tasks. Here, the question ofinterest is whether neuroticism is particu-larly linked to heightened amygdala respon-sivity to emotional stimuli when these stim-uli are in some form ambiguous in theirvalence or threat-relatedness. Canli and col-leagues reported that although the amygdalaresponse to happy faces and positive imageswas predicted by extraversion, there was norelationship between neuroticism and amyg-dala responsivity to either negative emo-tional faces or negative emotional images(Canli et al. 2001, 2002). Similarly, Britton,Ho, Taylor, and Liberzon (2007) found norelationship between neuroticism and amyg-dala responsivity during the passive view-ing of emotional images, facial expressions,and emotional films. Cremers et al. (2010)also found no relationship between neuroti-cism and the amygdala response to negativeemotional faces during a gender discrimi-nation task. The one exception is a studyby Chan, Norbury, Goodwin, and Harmer(2009) that examined amygdala responsiv-ity to fearful, happy, neutral, and morphedfacial expressions during a gender discrim-ination task in individuals high and lowin neuroticism. Here high N scores wereassociated with stronger amygdala activityto faces with expressions “morphed” part-way between neutral and fear. It could beargued that these morphed facial stimuli aremilder or more ambiguous in their threatvalue than the “fully” negative stimuli usedin the other studies. However, this one find-ing does not permit any definitive conclu-sions to be drawn in favor of the “ambi-guity” account without further empiricalinvestigation.

It also remains to be established whetherneuroticism is associated with impover-ished recruitment of the frontal mecha-

nisms supporting attentional control, duringnonemotional task performance, in a similarmanner to that observed for trait anxiety.There is little existing work that pertains tothis issue. Using an oddball detection task,Eisenberger, Lieberman, and Satpute (2005)found that neuroticism was associated withreduced lateral prefrontal cortical and ros-tral ACC recruitment but increased dorsalACC activity. Further detailed investigationof the relationship between neuroticism andrecruitment of lateral frontal and anteriorcingulate subregions is required to form aclearer picture of the commonalities and dif-ferences in the relationship between neu-roticism and trait anxiety with regional brainfunction. In particular, as noted earlier, theprecise role of the dorsal ACC in cognitiveand emotional processing remains an issueunder active debate.

It is hoped that this chapter has provideda flavor of how, in line with the case madeby Kosslyn and colleagues, it is possibleto conduct neuroimaging studies that mayadvance both our understanding of the brainmechanisms supporting attentional captureby threat and the relationship between traitanxiety and variation in the function of thesemechanisms. Although there are fewer stud-ies pertaining to neuroticism, the availablestudies serve to generate hypotheses thatcould form the basis of future research.The work reviewed here also raises ques-tions as to the extent to which the associ-ation among trait predisposition to anxiety,amygdala hyperactivity, and frontal hypoac-tivity is dependent on task domain. Specif-ically, will the associations reported herealso be observed in the context of per-formance of other tasks? Or even in theabsence of any task? Studies pertaining tothese questions are reviewed in the next twosections.

Trait Anxiety and Hyper-Amygdalaand Hypo-Frontal Function: The Caseof Fear Conditioning

The rodent fear conditioning literature pro-vides strong evidence for the role of frontal



inhibitory influences on the amygdala inthe attenuation of physiological and behav-ioral fear responses. Findings indicate thatthe amygdala is involved in the acqui-sition and expression of cued fear, withmedial prefrontal cortical inputs inhibit-ing amygdala responsivity to conditionedfear stimuli (CSs) following extinction train-ing (Maren & Quirk, 2004; Sotres-Bayon,Bush, & LeDoux, 2004). Human studiesof conditioned fear have implicated simi-lar circuitry with ventral medial regions ofprefrontal cortex (vmPFC) facilitatingcontext-specific recall of “CS – uncondi-tioned stimulus (UCS) absent” associationsformed during extinction training (Milad,Wright, et al., 2007; Phelps et al., 2004).Disruption to this circuitry has been doc-umented in adults with posttraumatic stressdisorder (Milad et al., 2009) and has beenproposed to be of potential relevance toother anxiety disorders.

Recently, a handful of studies have begunto address whether trait vulnerability toanxiety is linked to altered functioning ofthis circuitry. Two initial studies reportedthat high trait anxious individuals showedelevated amygdala activity during extinction(Barrett & Armony, 2009; Sehlmeyer etal., 2011). Associations of extinction-relatedACC activity with trait anxiety were alsoreported, but the directionality and specificlocus of these effects did not replicate acrossthe two studies. In recent work from ourown lab, we found that elevated amygdalaactivity to a CS that predicted an aversivestimulus (the UCS) mediated the relation-ship between trait anxiety and the strengthof initial acquisition of skin conductanceresponses to the predictive CS (Indovina etal., 2011). We also found that trait anxietywas negatively related to recruitment ofventral frontal regions linked to context-appropriate down-regulation of both cuedand contextual fear prior to omission ofthe UCS. Hierarchical regression revealedthat the relationship between trait anxietyand amygdala responsivity to the predictiveCS was independent of that betweentrait anxiety and context-appropriateventral PFC recruitment. It is interesting

to note that the medial and lateral regionsof ventral prefrontal cortex reported inthis study, the activation of which wasassociated with lower cued and contextualfear responses, overlap with those reportedelsewhere to be activated during deliberateemotion regulation and affective stimulusreappraisal (see Chapter 16). Delgadoand colleagues have further reported thatactivation of similar ventral PFC regionsaccompanies a reduction in conditionedfear, regardless of whether it occurs due toextinction or emotion regulation (Delgado,Nearing, LeDoux, & Phelps, 2008).

Together with the work reviewed ear-lier, these findings raise the possibility thatindividual variation in the recruitment ofdifferent subregions of frontal cortex mayinfluence volunteers’ ability to regulate theiremotional responses to the anticipation ofaversive stimuli, as well as their locus ofattention when threat-related visual stimuliare presented. It is important to note thatcharacteristics such as trait anxiety may maponto individual differences not only in theability to regulate responses to aversive stim-uli but also in the nature of the regulationstrategy selected. Initial studies have begunto examine the brain regions activated bydifferent emotion regulation strategies (e.g.,Vrticka, Sander, & Vuilleumier, 2011), butas yet this work has not been integratedwith investigation of individual differencesin strategy selection or success in implemen-tation.

Trait Vulnerability to AffectiveDisorder: What Might We Learn fromResting State Studies?

Recently, there has been increasing inter-est in whether the functional and structuralconnectivity between different brain regionsmay be informative even in the absence oftask performance. One high-profile exam-ple is the recently launched Human Con-nectome Project, which aims to “compre-hensively map human brain circuitry . . .using cutting-edge methods of noninva-sive neuroimaging . . . yield[ing] invaluable



information about brain connectivity, itsrelationship to behavior, and the con-tributions of genetic and environmentalfactors to individual differences in braincircuitry” (http://humanconnectome.org/).This interest has been accompanied by arenewed emphasis on examining the func-tion of brain regions in the context ofthe networks in which they are embeddedas “nodes.” It has also brought increasedrecognition of the need to study individualdifferences in order to understand norma-tive brain function. It is beyond the scopeof the current chapter to provide a com-prehensive review of this literature (seeDeco et al., 2011). We limit this section toa consideration of a few findings to datethat may inform our understanding of thebrain basis of trait vulnerability to affectivedisorder.

Roy et al. (2009) provided a detailedreport of connectivity between the amyg-dala and other brain regions at rest.Positive correlations were found betweenresting state blood-oxygen-level-dependent(BOLD) activity in the amygdala and a num-ber of brain regions, including the medialfrontal gyrus, rostral ACC, dorsal ACC,insula, thalamus, and striatum. Negativecorrelations were also observed between theamygdala and areas, including the superiorfrontal gyrus, bilateral middle frontal gyrus,posterior cingulate cortex, precuneus, andparietal and occipital lobes. Further anal-yses were conducted to profile the con-nections of different amygdala subnuclei.These provided some suggestion of nega-tive connectivity between the basolateralnucleus and central nucleus of the amyg-dala, as well as opposing patterns of con-nectivity between these subnuclei with themedial frontal gyrus and anterior cingulatecortex.

These findings are intriguing, but haveseveral limitations. First, as noted bythe authors, it is extremely difficult tospecify amygdala subnuclei on echoplanarimages. Although the probabilistic approachadopted by Roy and colleagues can providean estimate, it is not clear if the level ofaccuracy achieved is sufficient for analyses

of this nature. Second, an issue that per-tains to all resting state studies is that it isnot clear how to interpret negative versuspositive patterns of BOLD connectivity atrest. Are inhibitory connections at the neu-ronal level likely to be reflected as negativeBOLD connectivity patterns? This is oftenassumed but far from established. A thirdproblem is that it is unclear as to what crite-ria should be used for including or excludingregions in different “resting state” networks,raising spectra of the criticisms applied tothe circuits of Papez (1937) and MacLean(1949). Specifically, with both seed-basedand component-based approaches, the samequestion occurs as to what threshold touse – whether in terms of significance lev-els for whole-brain seed-driven analyses orthe number of components for indepen-dent or principal-components-based anal-yses. As the field develops, the challengewill be to find ways to address theseissues.

Building on the work by Roy and col-leagues, Kim et al. (2011) examined theinfluence of individual differences in self-reported pre-scan anxiety on resting statefunctional connectivity. They reported thatanxiety levels significantly modulated con-nectivity between the amygdala and onlytwo regions. High state anxious individu-als showed negative amygdala-vmPFC con-nectivity contrasting with positive func-tional connectivity between these regionsin low state anxious individuals. In addi-tion, they also showed an absence of thenegative connectivity between the amyg-dala and dorsal medial PFC observed inlow state anxious volunteers. The effectsof anxiety in this study are particularly ofinterest because of their selectivity. Inadvancing our understanding of trait vul-nerability to affective disorder, a limita-tion is that the primary analyses presentedused state anxiety measures, with anal-yses using trait anxiety measures beingdescribed as showing similar but weakertrends. One hopes that future studies willfurther explore the extent to which anxiety-related variability in amygdala-frontal func-tional connectivity at rest reflects stable



individual differences versus effects of atemporary mood state.

Trait Vulnerability to AffectiveDisorder: From Correlation toCausation?

If, based on the studies reviewed in thischapter, we come to the conclusion thatthere are stable trait-related differences inthe function of amygdala-frontal circuitry,what might cause these differences? Thereare a number of possibilities, which areby no means mutually exclusive. Theseinclude structural or functional differencesin one or both regions reflecting eithergenetic or environmental influences, differ-ences in the integrity of tracts connect-ing these regions, and differences in neuro-chemical modulation of one or both regions,again potentially reflecting either geneticor environmental influences uniquely or incombination.

The evidence pertaining to these alter-natives is relatively limited. Diffusion ten-sor imaging findings suggest that reducedintegrity of white matter tracts (as indexedby fractional anisotrophy) that link theamygdala to vmPFC in humans may indeedbe associated with trait vulnerability to anx-iety (Kim & Whalen, 2009). Positron emis-sion topography studies meanwhile point toa relationship between neuroticism and rest-ing state frontal hypo-perfusion (Deckers-bach et al., 2006). Arguably the most intrigu-ing findings are those emerging from therodent and human literature on the effectsof stress and gene-environment interactionson amygdala-frontal circuitry (see Arnsten,2009, and Chapters 22 and 25). Periods ofacute stress are associated with neurochem-ical changes, including elevated levels ofnoradrenaline and dopamine release (Gold-stein, Rasmusson, Bunney, & Roth, 1996).High levels of these catecholamines enhanceamygdala function (Debiec & LeDoux,2006), but undermine prefrontal corti-cal function (Arnsten, Mathew, Ubriani,Taylor, & Li, 1999). Chronic stress leadsto long-term alterations in the frontal-

amygdala network, with contrasting pat-terns of dendritic change being reported inthe frontal cortex and amygdala. Whereasamygdaloid dendrites increase (Vyas, Mitra,Shankaranarayano Rao, & Chattarji, 2002),neurons in the PFC show a reduction indendritic branches, which appears linked todeficient performance on measures of exec-utive control (see Holmes & Wellman, 2009,for a review). Genetic differences, includingcommon genetic polymorphisms influenc-ing catecholamine metabolism in the frontalcortex (e.g., the COMT val 158 met poly-morphism), as well as ones affecting sero-tonergic modulation of amygdala activity(e.g., polymorphisms in the serotonin trans-porter gene), may well interact with suchenvironmental influences (Hyde, Bogdan, &Hariri, 2011) and also with the effects ofearly life stress on gene expression (Francis,Champagne, Liu, & Meaney, 1999). Hence,a combination of genetic and environmen-tal influences potentially leads to changesin both frontal and amygdala integrity andfunction. Increased integration of the stress,epigenetics, and functional genomics litera-tures with that reviewed in the earlier sec-tions of this chapter may/might enable usto move beyond description of the cognitiveand neural correlates of trait vulnerability toaffective disorders to begin to outline causaltrajectories underlying observed individ-ual differences in affective style, disorder-related symptomatology, processing of emo-tionally salient stimuli, and associated brainfunction.

Conclusions

Studying trait vulnerability to affective dis-order may both inform our understand-ing of healthy brain function and pro-vide an important bridge to studies ofpsychiatric disorder. It may also enableus to establish markers of elevated riskfor psychiatric illness that could be usedto identify individuals who might benefitfrom preventive interventions (e.g., cogni-tive training) before a deepening spiral intoclinically significant illness occurs. To date,


TRAIT ANXIETY, NEUROTICISM, AND THE BRAIN BASIS OF VULNERABILITY TO AFFECTIVE DISORDER569

cross-group studies of normative brain func-tion and between-group studies of specificaffective disorders far outnumber studiesusing continuous trait measures to study thebrain basis of vulnerability to affective dis-order. These latter studies face a number ofchallenges. In addition to the need for well-validated trait measures of affective style andvulnerability, rigor in neuroimaging designand analysis is likely to be an importantdeterminant of progress in this field. Excit-ing advances are being made in the meth-ods and techniques available, both withinneuroimaging and converging approaches.By incorporating these advances and bydrawing on models derived from both thehuman cognitive and basic neuroscience lit-eratures, it will be possible to test increas-ingly sophisticated hypotheses regarding thebrain basis of vulnerability to affectivedisorder.

Outstanding Questions and FutureDirections

� How do different dimensions of personal-ity relate to vulnerability to affective dis-order?

� Are there multiple “pathways” by whichdysregulation of amygdala/frontal cir-cuitry confers vulnerability to affectivedisorder?

� To what extent does this dysregulationreflect genetic influences, the effects ofearly life stress on gene expression, or thedirect effect of chronic or acute stress onthis circuitry?

� On a methodological front, how can webest balance hypothesis-driven researchwith exploratory investigations? Withinthe area of neuroimaging, what are therespective limitations of different meth-ods of data acquisition and styles of anal-ysis (e.g. region-of-interest approachesversus whole-brain analyses)?.

� If we seek to understand genetic influ-ences on brain mechanisms implicated invulnerability to affective disorder, howcan we deal with the multiple com-parisons problem due to both brain

voxel and genetic polymorphism arraysize?

References

Amin, Z., Constable, R. T., & Canli, T. (2004).Attentional bias for valenced stimuli as afunction of personality in the dot-probe task.Journal of Research in Personality, 38, 15–23.

Amir, N., Weber, G., Beard, C., Bomyea, J., &Taylor, C. T. (2008). The effect of a single-session attention modification program onresponse to a public-speaking challenge insocially anxious individuals. Journal of Abnor-mal Psychology, 117(4), 860–68.

Anderson, A. K., Christoff, K., Panitz, D., DeRosa, E., & Gabrieli, J.D. (2003). Neural cor-relates of the automatic processing of threatfacial signals. Journal of Neuroscience, 23(13),5627–33.

Ansari, T. L., & Derakshan, N. (2011). The neu-ral correlates of impaired inhibitory control inanxiety. Neuropsychologia,49(5),1146–53.

Arnsten, A. F. T. (2009). Stress signaling path-ways that impair prefrontal cortex structureand function. Nature Reviews Neuroscience, 10,410–22.

Arnsten, A. F. T., Mathew, R., Ubriani, R., Tay-lor, J. R., & Li, B.-M. (1999). α-1 noradrenergicreceptor stimulation impairs prefrontal cor-tical cognitive function. Biological Psychiatry,45, 26–31.

Barrett, J., & Armony, J. L. (2009). Influence oftrait anxiety on brain activity during the acqui-sition and extinction of aversive conditioning.Psychological Medicine, 39(2), 255–65.

Beck, A. X, Ward, C. H., Mendelson, M., Mock,J., & Erbaugh, J. (1961). An inventory for mea-suring depression. Archives of General Psychi-atry, 4, 561–71.

Bieling, P. J., Antony, M. M., & Swinson, R. P.(1998). The state–trait anxiety inventory, traitversion: Structure and content re-examined.Behaviour Research and Therapy, 36,777–88.

Bishop, S. J. (2008). Neural mechanisms underly-ing selective attention to threat. Annals of theNew York Academy of Sciences, 1129, 141–52.

Bishop, S. J. (2009) Trait anxiety and impover-ished prefrontal control of attention. NatureNeuroscience, 12, 92–98.

Bishop, S. J., Duncan, J., Brett, M., &Lawrence, A. D. (2004). Prefrontal cortical



function and anxiety: Controlling attention tothreat-related stimuli. Nature Neuroscience,7(2),184–88.

Bishop, S. J., Duncan, J., & Lawrence, A. (2004b).State anxiety modulation of the amygdalaresponse to unattended threat-related stimuli.Journal of Neuroscience, 24, 10364–68.

Bishop, S. J., Jenkins, R., & Lawrence, A. (2007).The neural processing of task-irrelevant fear-ful faces: Effects of perceptual load and indi-vidual differences in trait and state anxiety.Cerebral Cortex, 17, 1595–1603.

Botvinick, M. M., Cohen, D. D., & Carter, C. S.(2004). Conflict monitoring and anterior cin-gulated cortex: An update. Trends in CognitiveSciences, 12, 539–46.

Bradley, B. P., Mogg, K., Falla, S. J., & Hamilton,L. R. (1998). Attentional bias for threateningfacial expressions in anxiety: Manipulation ofstimulus duration. Cognition Emotion, 12, 737–53.

Britton, J. C., Ho, S. H., Taylor, S. F., & Liberzon,I. (2007). Neuroticism associated with neuralactivation patterns to positive stimuli. Psychi-atric Research – Neuroimaging, 156(3), 263–67.

Brown, T. A. (2007). Temporal course and struc-tural relationships among dimensions of tem-perament and DSM-IV anxiety and mooddisorder constructs. Journal of Abnormal Psy-chology, 116, 313–28.

Brown, T. A., & Barlow, D. H. (2009). A proposalfor a dimensional classification system basedon the shared features of the DSM-IV anxietyand mood disorders: Implications for assess-ment and treatment. Psychological Assessment,21(3), 256–71.

Brown, T. A., & Rosellini, A. J. (2011). The directand interactive effects of neuroticism and lifestress on the severity and longitudinal courseof depressive symptoms. Journal of AbnormalPsychology, 120(4), 844–56

Bush, G. (2010). Attention-deficit/hyperactivitydisorder and attention networks. Neuropsy-chopharmacology, 35, 278–300.

Bush, G., Frazier, J. A., Rauch, S. L., Seidman,L .J., Whalen, P. J., Jenike, M. A., & Bie-derman, J. (1999). Anterior cingulate cortexdysfunction in attention-deficit/hyperactivitydisorder revealed by fMRI and the CountingStroop. Biological Psychiatry, 45(12), 1542–52.

Bush, G., Luu, P., & Posner, M. I. (2000). Cogni-tive and emotional influences in anterior cin-gulated cortex. Trends in Cognitive Sciences,4(6), 215–22.

Canli, T. (2004). Functional brain mapping ofextraversion and neuroticism: Learning fromindividual differences in emotion processing.Journal of Personality, 72, 1105–32.

Canli, T., Amin, Z., Haas, B., Omura, K., &Constable, R. T. (2004). A double dissociationbetween mood states and personality traitsin the anterior cingulate. Behavioral Neuro-science, 18, 897–904.

Canli, T., Sivers, H., Whitfield, S. L., Gotlib,I. H., & Gabrieli, J. D. (2002). Amygdalaresponse to happy faces as a function ofextraversion. Science, 296, 2191.

Canli, T., Zhao, Z., Desmond, J. E., Kang, E.,Gross, J., & Gabrieli, J. D. E. (2001). An fMRIstudy of personality influences on brain reac-tivity to emotional stimuli. Behavioral Neu-roscence, 115, 33–42.

Carter, C. S., Macdonald, A. M., Botvinick,M., Ross, L. L., Stenger, V. A., Noll, D.,& Cohen, J. D. (2000). Parsing executiveprocesses: Strategic vs. evaluative functionsof the anterior cingulated cortex. Proceed-ings of the National Academy of Sciences, 97,1944–48.

Chan, S. W., Goodwin, G. M., & Harmer, C.J. (2007). Highly neurotic never-depressedstudents have negative biases in informa-tion processing. Psychological Medicine, 37,1281–91.

Chan, S. W. Y., Norbury, R., Goodwin, G. M., &Harmer C. J. (2009). Risk for depression andneural responses to fearful facial expressionsof emotion. British Journal of Psychiatry, 194,139–45.

Clark, L. A., & Watson D. (1991). Tripartitemodel of anxiety and depression: Psychomet-ric evidence and taxonomic implications. Jour-nal of Abnormal Psychology, 100(3), 16–36.

Costa, P. T, Jr., & McCrae, R. R. (1995). Domainsand facets: Hierarchical personality assess-ment using the Revised NEO PersonalityInventory. Journal of Personality Assessment,64, 21–50.

Costa, P., & McCrae, R. R. (1992). Normalpersonality assessment in clinical practice:The NEO Personality Inventory. PsychologicalAssessment, 4, 5–13.

Cremers, H. R., Demenescu, L. R., Aleman, A.,Renken, R., van Tol, M.J., van der Wee, N. J.,& Roelofs, K. (2010). Neuroticism modulatesamygdala-prefrontal connectivity in responseto negative emotional facial expressions. Neu-roimage, 49, 963–70.



Davidson, R .J. (2002). Anxiety and affectivestyle: Role of prefrontal cortex and amygdala.Biological Psychiatry, 51(1), 68–80.

Debiec, J., & LeDoux, J. E. (2006). Noradrenergicsignaling in the amygdala contributes to thereconsolidation of fear memory: Treatmentimplications for PTSD. Annals of the New YorkAcademy of Sciences, 1071, 521–24.

Deckersbach, T., Miller, K. K., Klibanski, A.,Fischman, A., Dougherty, D. D., Blais, M.A., & Rauch, S. L. (2006). Regional cere-bral brain metabolism correlates of neuroti-cism and extraversion. Depression and Anxi-ety, 23(3), 133–38.

Deco, G., Jirsa, V. K., & McIntosh, A. R. (2011).Emerging concepts for the dynamical orga-nization of resting-state activity in the brain.Nature Reviews Neuroscience, 12(1), 43–56.

Delgado, M. R., Nearing, K. I., Ledoux, J. E., &Phelps, E. A. (2008). Neural circuitry underly-ing the regulation of conditioned fear and itsrelation to extinction. Neuron., 59(5), 829–38.

Derryberry, D., & Reed, M .A. (2002). Anxiety-related attentional biases and their regulationby attentional control. Journal of AbnormalPsychology, 111, 225–36.

Desimone, R., & Duncan, J. (1995). Neural mech-anisms of selective attention. Annual Reviewof Neuroscience, 18, 193–222.

Eisenberger, N. I., Lieberman, M. D., & Satpute,A. B. (2005). Personality from a controlledprocessing perspective: An fMRI study of neu-roticism, extraversion, and self-consciousness.Cognitive Affective & Behavioral Neuroscience,5, 169–81.

Etkin, A., Egner, T., Peraza, D. M., Kandel, E. R.,& Hirsch, J. (2006). Resolving emotional con-flict: A role for the rostral anterior cingulatecortex in modulating activity in the amygdala.Neuron, 51, 871–82.

Etkin, A., Klemenhagen, K. C., Dudman, J. T.,Rogan, M. T., Hen, R., Kandel, E. R., &Hirsch, J. (2004). Individual differences in traitanxiety predict the response of the basolateralamygdala to unconsciously processed fearfulfaces. Neuron, 44(6), 1043–55.

Etkin, A., & Wager, T. D. (2007) Functional neu-roimaging of anxiety: A meta-analysis of emo-tional processing in PTSD, social anxiety dis-order, and specific phobia. American Journalof Psychiatry, 164(10), 1476–88.

Eysenck, H. J., & Eysenck, S. B. G. (1975). EysenckPersonality Questionnaire manual. San Diego:Educational and Industrial Testing Service.

Eysenck, M. W., Derakshan, N., Santos, R.,& Calvo, M. G. (2007). Anxiety and cogni-tive performance: Attentional control theory.Emotion, 7, 336–53.

Francis, D. D., Champagne, F. A., Liu, D., &Meaney, M .J. (1999). Maternal care, geneexpression, and the development of individ-ual differences in stress reactivity. Annalsof the New York Academy of Sciences, 896,66–84.

Gazzaniga, M. S. Ivry, R. B., & Mangun, G.R. (2009). Emotion. In M.S. Gazzaniga, R.B.Ivry, & G.R. Mangun (Eds.), Cognitive neuro-science – the biology of the mind (3rd ed., pp.364–281). New York: Norton.

Goldstein, L. E., Rasmusson, A. M., Bunney, S.B., & Roth, R. H. (1996). Role of the amygdalain the coordination of behavioral, neuroen-docrine and prefrontal cortical monoamineresponses to psychological stress in the rat.Journal of Neuroscience, 16, 4787–98.

Gotlib, I. H., MacLachlan, A., & Katz, A. (1988).Biases in visual attention in depressed andnondepressed individuals. Cognition Emotion,2, 185–200.

Gotlib, I. H., & McCann, C.D. (1984). Constructaccessibility and depression: An examinationof cognitive and affective factors, Journal ofPersonality and Social Psychology, 47, 427–39.

Haas, B. W., Omura, K., Constable, R. T., &Canli, T. (2007). Emotional conflict and neu-roticism: Personality-dependent activation inthe amygdala and subgenual anterior cingu-late. Behavioral Neuroscience, 121, 249–56.

Hakamata, Y., Lissek, S., Bar-Haim, Y., Britton,J. C., Fox, N. A., Leibenluft, E., & Pine, D. S.(2010). Attention bias modification treatment:A meta-analysis toward the establishment ofnovel treatment for anxiety. Biological Psychi-atry, 68(11), 982–90.

Hettema, J. M., An, S. S., Bukszar, J., vanden Oord, E. J., Neale, M. C., Kendler,K. S., & Chen, X. (2008). Catechol-O-methyltransferase contributes to genetic sus-ceptibility shared among anxiety spectrumphenotypes. Biological Psychiatry, 64(4), 302–10.

Hill, A. B., & Dutton, F. (1989). Depression andselective attention to self-esteem threateningwords. Personality and Individual Differences,10, 915–17.

Hill, A. B., & Knowles, T. H. (1991). Depressionand the emotional Stroop effect. Personalityand Individual Differences, 12, 481–85.



Holmes, A., & Wellman, C. L (2009). Stress-induced prefrontal reorganization and execu-tive dysfunction in rodents. Neuroscience andBiobehavioral Reviews, 33, 773–83.

Huang-Pollock, C. L., Carr, T. H., & Nigg, J.T. (2002). Development of selective attention:Perceptual load influences early versus lateattentional selection in children and adults.Developmental Psychology, 38, 363–75.

Hyde, L. W., Bogdan, R., & Hariri, A. R.(2011). Understanding risk for psychopathol-ogy through imaging gene-environment inter-actions. Trends in Cognitive Sciences, 15(9), 417–27.

Indovina, I., Robbins, T. W., Nunez-Elizalde, A.O., Dunn, B. D., Bishop, S. J. (2011). Fear-conditioning mechanisms associated with traitvulnerability to anxiety in humans. Neuron,69(3), 563–71.

John, O. P. (1990). The “Big Five” factor taxon-omy: Dimensions of personality in the natu-ral language and in questionnaires. In L. A.Pervin (Ed.), Handbook of personality theoryand research (pp. 66–100). New York: Guil-ford Press.

Kastner, S., & Ungerleider, L.G. (2000). Mecha-nisms of visual attention in the human cortex.Annual Review of Neuroscience, 23, 315–41.

Kendler, K. S., Gardner, C. O., Gatz, M., &Pedersen, N. L. (2007). The sources of co-morbidity between major depression and gen-eralized anxiety disorder in a Swedish nationaltwin sample. Psychological Medicine, 37(3),453–62.

Kim, M. J., Gee, D. G., Loucks, R. A., Davis,F. C., & Whalen, P. J. (2011). Anxiety dissoci-ates dorsal and ventral medial prefrontal cor-tex functional connectivity with the amygdalaat rest. Cerebral Cortex, 21(7), 1667–73.

Kim, H., Somerville, L. H., Johnstone, T.,Alexander, A. L., & Whalen, P. J. (2003).Inverse amygdala and medial prefrontal cor-tex responses to surprised faces. Neuroreport,14(18), 2317–22.

Kim, M. J., & Whalen, P. J. (2009). The structuralintegrity of an amygdala-prefrontal pathwaypredicts trait anxiety. Journal of Neuroscience,29(37), 11614–18.

Kosslyn, S. M., Cacioppo, J. T., Davidson, R.J.,Hugdahl, K., Lovallo, W. R., Spiegel, D., &Rose, R.(2002). Bridging psychology and biol-ogy. American Psychologist, 57, 341–51.

Lavie, N. (2005). Distracted and confused?: Selec-tive attention under load. Trends in CognitiveSciences, 9, 75–82.

LeDoux, J. E. (2000). Emotion circuits in thebrain. Annual Review of Neuroscience, 23, 155–84.

Luteijn, F., & Bouman, T. K. (1988). The con-cepts of depression, anxiety, and neuroticismin questionnaires. European Journal of Person-ality, 2, 113–20.

MacDonald, A. W., Cohen, J. D., Stenger, V. A.,& Carter, C. S. (2000). Dissociating the roleof dorsolateral prefrontal cortex and anteriorcingulate cortex in cognitive control. Science,288, 1835–38.

MacLean, P. D. (1949). Psychosomatic diseaseand the “visceral brain”: Recent developmentsbearing on the Papez theory of emotion. Psy-chosomatic Medicine, 11, 338–53

MacLeod, C., & Hagan, R. (1992). Individual dif-ferences in the selective processing of threat-ening information, and emotional responses toa stressful life event. Behaviour Research andTherapy , 30, 151–61.

MacLeod, C., & Mathews, A. (1988). Anxiety andthe allocation of attention to threat. QuarterlyJournal of Experimental Psychology, 40, 653–70.

Maren, S., & Quirk, G. J. (2004). Neuronal sig-naling of fear memory. Nature Reviews Neu-roscience, 5, 844–52.

Mathews, A., & Mackintosh, B. (1998). A cogni-tive model of selective processing in anxiety.Cognitive Therapy Research, 22, 539–60.

Mathews, A., Ridgeway, V., & Williamson, D.A., (1996). Evidence for attention to threaten-ing stimuli in depression. Behaviour Researchand Therapy, 34, 695–705.

Mayberg, H. S., Liotti, M., Brannan, S. K.,McGinnis, S., Mahurin, R. K., Jerabek, P.A., & Fox, P. T. (1999). Reciprocal limbic-cortical function and negative mood: Con-verging PET findings in depression and normalsadness. American Journal of Psychiatry, 156,675–82.

Maylor, E., & Lavie, N. (1998). The influence ofperceptual load on age differences in selectiveattention. Psychology and Aging, 13, 563–73.

Milad, M. R., Pitman, R. K., Ellis, C. B., Gold,A. L., Shin, L. M., Lasko, N. B., & Rauch,S. L. (2009). Neurobiological basis of failureto recall extinction memory in posttraumaticstress disorder. Biological Psychiatry, 66(12),1075–82.

Milad, M. R., Quirk, G. J., Pitman, R. K., Orr,S. P., Fischl, B., & Rauch, S.L. (2007). A rolefor the human dorsal anterior cingulate cortexin fear expression. Biological Psychiatry, 62(10),1191–94.



Milad, M. R, Wright, C. I., Orr, S. P., Pitman, R.K., Quirk, G. J., & Rauch, S. L. (2007). Recallof fear extinction in humans activates the ven-tromedial prefrontal cortex and hippocam-pus in concert. Biological Psychiatry, 62(5),446–54.

Mogg, K., & Bradley, B. P. (1998). A cognitive-motivational analysis of anxiety. BehaviourResearch and Therapy, 36, 809–48.

Morris, J. S., Frith, C. D., Perrett, D. I., Rowland,D., Young, A. W., Calder, A. J., & Dolan, R.J. (1996). A differential neural response in thehuman amygdala to fearful and happy facialexpressions. Nature, 383(6603), 812–15.

Nomura, M., Iidaka, T., Kakehi, K., Tsukiura,T., Hasegawa, T., Maeda, Y., & Matsue, Y.(2003). Frontal lobe networks for effective pro-cessing of ambiguously expressed emotions inhumans. Neuroscience Letters, 348(2), 113–16.

Ochsner, K. N., Bunge, S. A., Gross, J. J., &Gabrieli, J. D. (2002). Rethinking feelings:An FMRI study of the cognitive regulationof emotion. Journal of Cognitive Neuroscience,14(8), 1215–29.

Papez, J. W. (1937). A proposed mechanism ofemotion. Archives of Neurology and Psychiatry,38, 725–74.

Pessoa, L., & Adolphs, R. (2010). Emotion pro-cessing and the amygdala: From a “low road”to “many roads” of evaluating biological sig-nificance. Nature Reviews Neuroscience, 11(11),773–83.

Pessoa, L., McKenna, M., Gutierrez, E., & Unger-leider, L.G. (2002). Neural processing of emo-tional faces requires attention. Proceedingsof the National Academy of Sciences, 99(17),11458–63.

Phelps, E. A., Delgado, M.R., Nearing, K. I., &LeDoux, J. E. (2004). Extinction learning inhumans: Role of the amygdala and vmPFC.Neuron, 43(6), 897–905.

Radloff, L. S. (1977) .The CES-D Scale: A self-report depression scale for research in thegeneral population. Applied Psychological Mea-surement, 1, 385–401.

Reed, M. A., & Derryberry, D. (1995). Tempera-ment and attention to positive and negativetrait information. Personality and IndividualDifferences, 18, 135–47.

Rees, G, Frith, C. D., & Lavie, N. (1997). Modu-lating irrelevant motion perception by varyingattentional load in an unrelated task. Science,278(5343), 1616–19.

Ressler, K. J., & Mayberg, H.S. (2007). Targetingabnormal neural circuits in mood and anxiety

disorders: From the laboratory to the clinic.Nature Neuroscience, 10, 1116–24.

Richards, A., & Millwood, B. (1989). Colour-identification of differentially valenced wordsin anxiety. Cognition Emotion, 3, 171–76.

Rijsdijk, F. V., Riese, H., Tops, M., Snieder, H.,Brouwer, W. H, Smid, H. G. O. M., & Ormel,J. (2009). Neuroticism, recall bias and atten-tion bias for valenced probes: A twin study.Psychological Medicine, 39(1), 45–54.

Roy, A .K., Shehzad, Z., Margulies, D. S., Kelly,A. M., Uddin, L. Q., Gotimer, K., Biswal, B.B., & Milham, M. P. (2009) Functional con-nectivity of the human amygdala using restingstate fMRI. Neuroimage, 45(2), 614–26.

Sehlmeyer, C., Dannlowski, U., SchAning, S.,Kugel, H., Pyka, M., Pfleiderer, B., Zwitser-lood, P., & Konrad, C. (2011). Neural correlatesof trait anxiety in fear extinction. PsychologicalMedicine, 41(4), 789–98.

Shackman, A. J., Salomons, T. V., Slagter, H.A., Fox, A. S., Winter, J. J., & Davidson, R.J. (2011). The integration of negative affect,pain and cognitive control in the cingulatecortex. Nature Reviews Neuroscience, 12(3),154–67.

Shin, L .M., & Liberzon, I. (2010). The neuro-circuitry of fear, stress, and anxiety disorders.Neuropsychopharmacology, 35(1), 169–91.

Shin, L. M., Whalen, P. J., Pitman, R. K., Bush,G., Macklin, M. L., Lasko, N. B., Orr, S.P., & Rauch, S. L. (2001). An fMRI study ofanterior cingulate function in posttraumaticstress disorder. Biological Psychiatry, 50(12),932–42.

Sotres-Boyen, F., Bush, D. E. A., & LeDoux, J. E.(2004). Emotional perseveration: An updateon prefrontal–amygdala interactions in fearextinction. Learning and Memory, 11, 525–35.

Spielberger C. D., Gorsuch R. L., Lushene P. R.,Vagg P. R., & Jacobs A. G. (1983). Manual forthe State-Trait Anxiety Inventory (Form Y). PaloAlto, CA: Consulting Psychologists Press.

Stein, M. B. (2009). Neurobiology of generalizedanxiety disorder. Journal of Clinical Psychiatry,70(Suppl. 2), 15–19.

Stein, M. B., Simmons, A. N., Feinstein, J. S.,& Paulus, M. P. (2007). Increased amygdalaand insula activation during emotion process-ing in anxiety-prone subjects. American Jour-nal of Psychiatry, 164(2), 318–27.

Tamietto, M., & de Gelder, B. (2010). Neuralbases of the non-conscious perception of emo-tional signals. Nature Reviews Neurosciences,11(10), 697–709.



Taylor, J. A. (1953). A personality scale of man-ifest anxiety. Journal of Abnormal and SocialPsychology, 48, 285–90.

Underwood, B. J. (1975). Individual differencesas a crucible in theory construction. AmericanPsychologist, 30, 128–34.

Vrticka, P., Sander, D., & Vuilleumier, P. (2011).Effects of emotion regulation strategy on brainresponses to the valence and social content ofvisual scenes. Neuropsychologia, 49(5), 1067–82.

Vuilleumier, P., Armony, J. L, Driver, J., &Dolan, R. J. (2001). Effects of attention andemotion on face processing in the humanbrain: An event-related fMRI study. Neuron,30, 829–41.

Vul, E., Harris, C., Winkielman, P., & Pashler, H.(2009). Puzzlingly high correlations in fMRIstudies of emotion, personality, and social cog-nition. Perspectives on Psychological Science, 4,274–290.

Vyas, A., Mitra, R., Shankaranarayana Rao,B. S., & Chattarji, S. (2002). Chronicstress induces contrasting patterns of den-dritic remodeling in hippocampal and amyg-daloid neurons. Journal of Neuroscience, 22,6810–18.

Watson, D., & Clark, L. A. (1984). Nega-tive affectivity: The disposition to experience

aversive emotional states. Psychological Bul-letin, 96, 465–90.

Watson, D., & Clark, L. A. (1991). The Mood andAnxiety Symptom Questionnaire. Iowa City:University of Iowa.

Weems, C. F., Pina, A. A., Costa, N. M., Watts, S.E., Taylor, L. K., & Cannon, M. F. (2007). Pre-disaster trait anxiety and negative affect pre-dict posttraumatic stress in youth after Hurri-cane Katrina. Journal of Consulting and Clini-cal Psychology, 75, 154–59.

Whalen, P. J. (2007). The uncertainty of it all.Trends in Cognitive Sciences, 11(12), 499–500.

Whalen, P. J., Kagan, J, Cook, R. G., Davis, F. C.,Kim, H., Polis, S., McLaren, D. G., & John-stone, T. (2004). Human amygdala respon-sivity to masked fearful eye whites. Science,306(5704), 2061.

Whalen, P. J., Rauch, S. L., Etcoff, N. L., McIn-erney, S. C., Lee, M. B., & Jenike, M. A(1998). Masked presentations of emotionalfacial expressions modulate amygdala activitywithout explicit knowledge. Journal of Neuro-science, 18(1), 411–18.

Williams, J. M. G., & Nulty, D.D. (1986). Con-struct accessibility, depression and the emo-tional Stroop task: Transient mood or stablestructure? Personality and Individual Differ-ences, 7, 485–491.

trait anxiety, neuroticism, and the brain basis of...

Documents