Skip to main content
NCBI home page
Social Cognitive and Affective Neuroscience logoLink to Social Cognitive and Affective Neuroscience
. 2011 Jun 11;7(5):568–577. doi: 10.1093/scan/nsr038

Amygdala responses to averted vs direct gaze fear vary as a function of presentation speed

Reginald B Adams Jr 1,✉, Robert G Franklin Jr 1, Kestutis Kveraga 2, Nalini Ambady 3, Robert E Kleck 4, Paul J Whalen 4, Nouchine Hadjikhani 2,5, Anthony J Nelson 1
PMCID: PMC3375890  PMID: 21666261

Abstract

We examined whether amygdala responses to rapidly presented fear expressions are preferentially tuned to averted vs direct gaze fear and conversely whether responses to more sustained presentations are preferentially tuned to direct vs averted gaze fear. We conducted three functional magnetic resonance imaging (fMRI) studies to test these predictions including: Study 1: a block design employing sustained presentations (1 s) of averted vs direct gaze fear expressions taken from the Pictures of Facial Affect; Study 2: a block design employing rapid presentations (300 ms) of these same stimuli and Study 3: a direct replication of these studies in the context of a single experiment using stimuli selected from the NimStim Emotional Face Stimuli. Together, these studies provide evidence consistent with an early, reflexive amygdala response tuned to clear threat and a later reflective response tuned to ambiguous threat.

Keywords: threat perception, amygdala, fMRI, eye gaze, fear expression

INTRODUCTION

When examining socio-emotional perception, particularly at the neural level, it is necessary to consider both reflexive and reflective processes (cf. Lieberman et al., 2002; Lieberman 2003; Cunningham and Zelazo, 2007). This is particularly true of threat perception, which is known to involve both early and late processing strea ms (Davis, 1992; LeDoux, 1998; Adolphs, 2002a, 2002b), which appear to be dissociable both temporally and structurally (Halgren and Marinkovic, 1995; LeDoux, 1995; Morris et al., 1998, 1999, 2001; Vuilleumier, 2002; Liddell et al., 2004; Williams et al., 2004). Herein we propose that these two processing routes are not only neurally distinct, but are functionally tuned to different aspects of the threatening stimulus. Specifically, we propose that reflexive vs reflective processes may be differentially tuned to clear vs ambiguous combinations of threat cues, respectively.

Most previous studies examining neural responses to anger and fear have tended to utilize only faces displaying direct gaze, with an underlying assumption that gaze direction should not meaningfully influence the perception of such basic threat signals as anger and fear. However, anger and fear signal different types of threat for which eye gaze direction is arguably differentially informative (Adams et al., 2010a; Adams and Nelson, 2011). Anger signals a threat from the person making the expression. When coupled with direct eye gaze, therefore, observers gain information regarding both the source of threat and that the threat is directed at them. When coupled with averted gaze, however, the anger signal is a more ambiguous message of imminent danger to the observer. Fear, on the other hand, signals that an expressor perceives an external threat in the environment. Thus, when coupled with direct gaze, there is ambiguity regarding the source of threat. In this case, averted gaze offers information regarding where that threat might be located. Notably, averted gaze and fear expressions also both signal avoidance, whereas direct gaze and anger both signal approach. Thus, direct-gaze anger and averted-gaze fear, when combined, convey congruent signals. Consistent with both of these interpretations, averted-gaze fear and direct-gaze anger have been found to be perceived as more intense and recognized more quickly and accurately than direct gaze fear and averted-gaze anger (Adams and Kleck, 2003, 2005; Adams et al., 2006; Adams and Franklin, 2009; Sander et al., 2007; Benton, 2010). A growing number of studies utilizing a variety of techniques now offer growing support for the important role of gaze direction in emotion processing, particularly in threat perception (e.g. Sato et al., 2004; Fox et al., 2007; Graham and LaBar, 2007; Hess et al., 2007; Hadjikhani et al., 2008; N’Diaye et al., 2009; Ewbank et al., 2010; Rigato et al., 2010).

When initially examining the role of eye gaze in neural responses to threat displays, we found greater amygdala activation when participants viewed ambiguous threat-gaze pairs (direct fear, averted anger) vs clear threat-gaze pairs (averted fear, direct anger; Adams et al., 2003). At the time, this finding seemed to resolve a puzzling issue in the literature: early neuroimaging papers reported robust and consistent amygdala responses to fear faces, but to a lesser extent angry faces (e.g. Whalen et al., 2001). Yet, anger—at least when coupled with direct gaze—arguably signals a clear and immediate threat to the observer, as it indicates both the source and target of imminent aggression. Our findings, therefore, corroborated a working hypothesis at the time that amygdala responses increase proportionally with the degree of ambiguity surrounding the source of threat (see Whalen, 1998).

Since this initial fMRI study, a new even more puzzling issue has emerged. Three subsequent neuroimaging studies (Sato et al., 2004; Hadjikhani et al., 2008; N’Diaye et al., 2009) have been published that—although confirming the important role of gaze in threat perception at the neural level—revealed the opposite pattern of effects as our own (i.e. greater amygdala responses to congruent vs ambiguous threat-gaze pairs). Recently, important moderators of the gaze by threat display interaction have been explored, including individual differences in trait and state anxiety levels (Fox et al., 2007; Ewbank et al., 2010), shifts in progesterone levels along the menstrual cycle (Conway et al., 2007) and the relative discriminability of gaze and expression at the stimulus level (Graham and LaBar, 2007). None of these, however, can account for the current discrepancy across these studies, leaving this issue unresolved.

One aspect of threat perception yet to be examined that may help address this new puzzle is processing speed. When scrutinizing previous studies for similarities and differences, it became evident that the presentation parameters employed (‘rapid’—300 ms presentation: Hadjikhani et al., 2008; ‘dynamic’—rapidly unfolding from neutral to expressive: N’Diaye et al., 2009; and ‘peripheral’—presented to the left or right of fixation: Sato et al., 2004), all favor visual input via the magnocellular pathway, which is implicated in initial, rapid orienting responses to threat (Vuilleumier et al., 2003). Our original study, on the other hand, used prolonged (2 s), focally presented, static threat faces, which represent presentation parameters favoring additional visual input via the parvocellular pathway. Another recent study using prolonged (1 s), focally presented, static images yielded similar amygdala effects to our own—greater amygdala responses to incongruent vs congruent pairings of gaze when coupled with two approach-oriented emotional displays, averted vs direct happy and angry expressions (Straube et al., 2009). These findings suggest the possibility that threat processing along parallel visual pathways may account for the differences reported in the literature, thereby implicating a dual process.

Dual process models have played a critical role in helping shape our understanding of a broad range of social processes and are of particular importance to consider at the neural level (Lieberman et al., 2002; Satpute and Lieberman, 2006). Specifically, these models focus on the duality of ‘reflexive’ vs ‘reflective’ responding, for which distinct neural underpinnings are known to exist. Reflexive pathways are implicated in automatic responding, whereas reflective pathways act in a more controlled, top–down manner to guide, control, respond to and fine tune information processing. Distinct neurocognitive pathways such as these exist arguably so that reflexive and reflective processes can operate in a parallel, coordinated manner for maximally efficient responding.

The dual process distinction has a direct parallel in the threat perception literature as well, the putative ‘low road’ vs ‘high road’. Low road responses are thought to involve only crude information about the environment provided predominantly by magnocellular inputs and unfold in a quick and efficient manner, allowing for survival-enhancing behaviors to be enacted before threat strikes. High road responses, on the other hand, are thought to build upon a slower and more detailed account of the environment, thereby allowing for modulation of initial low road perceptions and behaviors (LeDoux, 1998; Palermo and Rhodes, 2007). Little is known about how these systems interact, however, particularly as to whether they exhibit differential responses to combinations of threatening cues, such as eye gaze and expression. Given the existing evidence for distinct reflexive vs reflective neural responses to threat, these observations give rise to a straightforward prediction in the context of the current work: presentation conditions favoring relatively rapid vs more sustained threat processing will give rise to differentially increased amygdala responses to averted vs direct gaze fear, respectively. Addressing this hypothesis is the focus of our current series of studies.

Current study

We designed a series of studies that vary the presentation speed of fear displays, from relatively sustained to very rapid, to examine the proposed temporal dynamics of compound threat cue processing. These studies rule out other obvious potential differences in methodology that could otherwise account for such differences across studies, including features of the stimuli (e.g. total number, prototypicality) and differences in design (i.e. block vs event-related). Importantly, Hadjikhani et al. (2008) utilized 300 ms stimulus presentations (i.e. 300 ms fear expression, 1200 ms fixation). To create an identical design varying only in stimulus presentation duration, keeping all other aspects of the paradigm equivalent, we employed both 300 ms presentations with 1200 fixation for our rapid presentation condition, a direct replication of Hadjikhani et al. (2008) and a 1 s stimulus presentation with 500 ms fixation for our sustained duration comparison condition. Critically, initial reflexive attention has previously been found to be triggered as early as 50–100 ms, becoming fully engaged ∼300 ms, whereas intentional responding is thought to occur as early as 500–700 ms (e.g. Posner, 1980; Driver et al., 1999; Cooper and Langton, 2006; Jones et al., 2010), making these stimulus durations well-suited to our goal of targeting obligatory vs intentional threat responses.

We ran three fMRI studies varying the presentation speed (1 s vs 300 ms) of direct vs averted gaze fear faces. Study 1 (1 s) and Study 2 (300 ms) utilized stimuli selected from the highly standardized and widely validated Pictures of Facial Affect (Ekman and Friesen, 1976), previously used by Adams et al. (2003). Study 3 then directly replicated these findings in the context of a single experiment with stimuli selected from the NimStim Emotional Face Stimuli database (Tottenham et al., 2009), previously used by Hadjikhani et al. (2008). Thus, the current series of experiments aimed to reconcile differences across these previous studies.

Study 1: amygdala responses to 1 s presentations of direct vs averted fear faces

Methods

Participants

Fifteen participants (nine female, six male) completed this study. Fourteen of the participants were Caucasian and one male was African American. Participants were all right-handed, with normal or corrected to normal vision, between 18 and 30 years, with no reported history of neurological impairments or disorders.

Stimuli

Eight separate models (four female, four male) displaying fear expressions were selected from the Pictures of Facial Affect (Ekman and Friesen, 1976). To reduce potential habituation effects, we also included neutral displays of these same exemplar faces morphed at 20% with their corresponding happy expressions.

Design and procedure

Participants passively viewed blocks of fear faces and 20% joy faces with fixation trials interspersed to help decrease habituation. Participants completed two functional runs. Each run consisted of an initial 20 s resting fixation, followed by 24 blocks, including six direct fear (DF), six averted fear (AF), three direct neutral (DN), three averted neutral (AN) and six additional fixation blocks (+). Each block lasted 16 s, including 12 trials consisting of a face displayed for 1 s, followed by a 500 ms fixation. For each of two runs, blocks consisted of the following run order: (+, DF, DN, DF, +, AF, NA, AF, +, AF, DN, DF, +, AF, DN, DF, +, AF, DN, DN, + DF, AN, AF, +) with the reverse order used in the second run. Run orders were counterbalanced across subjects.

Data acquisition and analysis

Data was acquired using an Intera 3.0 Tesla Philips scanner with custom built 3-axis balanced-torque head gradient coil and end capped birdcage RF coils. A T1–weighted three-dimensional image was acquired for anatomical structure (TR = 8.05 ms, TE = 3.7 ms, flip angle = 8°, FOV = 256 × 150 mm3, acquisition matrix = 256 × 256 × 150 and a SENSE factor = 2). SENSE factor refers to sensitivity encoding used to reduce overall scan time using multiple image coils with parallel scan times, thereby preserving spatial resolution. It is represented as a factor of k-space samples that have been reduced. Echo planar imaging (EPI) was used for functional data acquisition (TR = 3 s, TE = 30 ms, flip angle = 90°, voxel size = 3 × 3 × 3 mm, 48 interleaved slices, 134 volumes, thickness = 3 mm, no gap and SENSE factor = 2). The first 6 volumes collected preceded the task run and were dropped from analysis due to EPI saturation effects. Preprocessing was conducted in SPM5. Images were realigned using a least squares approach and a 6-parameter rigid body spatial transformation to the first functional image, coregistered using the structural and the functional data in 3D using rigid body transformations. The coregistered anatomical images were segmented and normalized to the MNI space using these segmentation parameters, which formed the basis of spatial normalization of the functional images. To allow for inter-subject averaging, the functional images were smoothed with 8 mm FWHM isotropic Gaussian smoothing kernel.

Statistical analysis was performed using a mass-univariate GLM approach. Stimulus conditions were modeled as delayed boxcar functions convolved with a standard HRF model. We included six regressors to account for head motion (x, y and z plane and roll, pitch and yaw). These were included in the first-level analysis along with our trial conditions. Low-frequency signal components were eliminated with a standard SPM5 high-pass filter of 1/128 Hz. Subject-specific contrasts were estimated using a fixed-effects model. For group analysis, these estimates were entered in a second-level analysis treating participants as a random effect, using one-sample t-tests at each voxel. Clusters were localized based on the contrast of direct gaze minus averted gaze fear and vice versa (height: P < 0.005, uncorrected, extent: 10 voxels). Notably, Lieberman and Cunningham (2009) recently argued that this threshold is optimal to balance between Type 1 and Type 2 errors. For illustration purposes, all group contrast images were overlaid onto a representative T1 template anatomical image using MRIcron (http://www.sph.sc.edu/comd/rorden/mricro.html). Coordinates are reported in MNI space.

Results

As predicted, responses to direct minus averted gaze fear presentations revealed greater left amygdala response—thereby replicating Adams et al. (2003); (see Figure 1). No amygdala responses were apparent when comparing averted minus direct fear, even at greatly reduced thresholds (for full list of activations, refer to Supplementary Table S1).

Fig. 1.

Fig. 1

Open in a new tab

Activation maps correspond to whole-brain analyses for Study 1 (A) and Study 2 (B). Activations for Study 3 (C and D) correspond to whole-brain analyses, but are restricted to the ROI masks determined by Studies 1 and 2. Activations in Study 3 represent points of direct overlap with Study 1 for the 1 s presentation condition (C) and with Study 2 for the 300 ms presentation condition (D) for the contrasts averted minus direct fear (displayed in the bottom panel) and direct minus averted fear (displayed in the top panel). Significant left amygdala activation (circled in green) was found for direct minus averted fear in both Studies 1 and 3 (1 s presentation condition) and significant right dorsal amygdala/SI (circled in green) was apparent to averted minus direct fear in both Studies 2 and 3 (300 ms presentation condition).

Study 2: amygdala responses to 300 ms presentations of direct and averted fear

Methods

Participants

Eighteen Caucasian participants (nine female, nine male) were recruited for financial compensation. Participants were all right-handed, in the age group of 18–30 years, with normal or corrected to normal vision and reported having no history of neurological impairments or disorders. Two participants (both male) showed significant signal drop out in the amygdala and thus were excluded from the analyses.

Stimuli

The same fear expressions used in Study 1 were used here. In order to keep this design identical to that used by Hadjikhani et al. (2008), the neutral baseline and fixation blocks were dropped.

Design and procedure

Participants completed a passive viewing task using the same procedure employed by Hadjikhani et al. (2008), which involved an ABA design with alternating blocks of averted and direct gaze fear. Participants viewed eight blocks of fear expressions. Blocks lasted 24 s and included 16 trials each. Each trial consisted of a fearful face displayed for 300 ms followed by a 1200 ms fixation.

Data acquisition and analysis

FMRI scanning employed a 3-T Siemens Tim Trio scanner with a standard 12-channel headcoil for data acquisition. Scanning consisted of two experimental functional runs and a T1-weighted anatomical scan collected for coregistraton and normalization. High-resolution T1-weighted were collected using MP-RAGE (128 sagittal slices 1.33 mm thick, 256 × 256 matrix). Functional scans used EPI involving 192 dynamic whole-brain T2*-weighted images per run collected in an oblique axial orientation (TR = 2 s, TE = 30 ms, flip angle = 90°, voxel size = 3.125 × 3.125 × 5 mm, 32 interleaved slices, 192 volumes, thickness = 5 mm, gap = 1 mm). Foam padding around the head was used to minimize head movement. Images were viewed using a back-projection system. Data were preprocessed and analyzed in the same way as reported in Study 1. Notably, these data were drawn from a larger study examining cross-cultural influences of gaze and fear processing (Adams et al., 2010b). Given evidence for cultural effects in that study, the current analysis focuses only on US Caucasian participants viewing US Caucasian fear expressions. The contrasts comparing direct and averted gaze fear reported here were not previously published.

Results

As predicted, neural responses to averted minus direct gaze fear revealed significant amygdala activation, specifically in bilateral dorsal amygdalae/substantia innominata (SI). No amygdala response was apparent in the reverse contrast, again even at greatly reduced thresholds (for full list of activations, refer to Supplementary Table S2).

Study 3: replication and extension of Studies 1 and 2 utilizing a different stimulus set

Methods

Participants

About 29 (15 female, 14 male) Caucasian participants were selected in the age ranging between 18 and 35. Participants had normal or corrected to normal vision and were free of neurological impairment. About 15 participants (7 female, 8 male) were randomly assigned to the 1 s presentation condition and 14 participants (8 female, 6 male) to the 300 ms presentation condition. Two participants (male) were subsequently dropped from analyses for excessive movement, both from the 1 s condition.

Stimuli

Stimuli used here were the same as those used by Hadjikhani et al. (2008), selected from the NimStim Emotional Face Stimuli database (Tottenham et al., 2009) and included eight models displaying fear expressions, four female and four male.

Design and procedure

Participants passively viewed fearful faces in an ABA block design alternating between averted and direct gaze, with a total of 16 blocks of fearful faces. Each block consisted of 16 stimulus presentations. For the sustained presentation condition, stimuli were presented for 1 s followed by a 500 ms fixation. For the rapid presentation condition, stimuli were presented for 300 ms followed by a 1200 ms fixation between each stimulus. Otherwise, trial length between the two conditions was identical.

Data acquisition and analysis

We collected and preprocessed fMRI data as described in Study 1. An initial whole-brain analysis was conducted using the same thresholds utilized in Studies 1 and 2, (height: P < 0.005, uncorrected, extent: 10 voxels). In order to examine a precise replication of Studies 1 and 2, we also applied inclusionary region of interest (ROI) masks derived from each of these studies for direct comparison with the current presentation conditions. Thus, an ROI mask from Study 1 was used to examine corresponding neural responses for the 1 s presentation condition and an ROI mask from Study 2 was used to examine the 300 ms presentation condition. ROI masks were thresholded at P < 0.05, uncorrected, a commonly used threshold when employing an inclusive mask in this manner (e.g. Peigneux et al., 2004; Coull et al., 2011). Significant clusters were then based on a contrast of direct gaze minus averted gaze fear and vice versa when applying these masks (height: P < 0.005, uncorrected, extent: 10 voxels).

In this way, we were able to examine activations specifically overlapping with those found from in Studies 1 and 2 to employ a more stringent assessment of the reliability and consistency of effects across studies. Finally, we ran ROI analyses pulling beta weights from an 8 mm sphere centered over the peaks of the amygdalae activation reported in Studies 1 and 2 (i.e. left amygdala in Study 1, bilateral dorsal amygdala/SI in Study 2), to directly compare activation across presentation duration for the comparison of direct vs averted gaze.

Results

For sustained presentations (1 s), left amygdala responses to direct minus averted gaze fear directly replicated those found in Study 1. Likewise, for rapid presentations (300 ms), neural responses to averted minus direct gaze fear displays replicated those found in the right dorsal amygdala/SI found in Study 2, but not for left dorsal amygdala/SI (Figure 1). Using activations pulled based on our three ROIs derived from Studies 1 and 2, we next computed a 2 (presentation duration: 300 ms, 1 s) × 3 (ROI: left amygdala—Study 1, left and right dorsal amygdala/SI—from Study 2) mixed factorial (between-within) ANOVA. This analysis revealed a main effect of presentation duration, F(1, 25) = 9.47, P < 0.005, partial η2 = 0.275, such that amygdala responses to fast presentations were greater for averted than for direct gaze fear, whereas amygdala responses to slow presentations were greater for direct than for averted gaze fear (Figure 2). Although there was not an interaction between type of ROI and presentation duration, for the sake of completeness, we also ran single sample t-tests for each ROI for each presentation duration separately. The only regions reaching significance based on these analyses were left amygdala in the sustained presentation condition, t(12) = 2.52, P < 0 .05, such that activation was greater for direct minus averted gaze (β = 0.087, s.e. = 0.035) and right dorsal amygdala/SI in the rapid presentation condition, t(13) = −2.23, P < 0.05, such that activation was greater for averted minus direct gaze (β = −0.051, s.e. = 0.025; all other P’s > 0.2), thereby directly replicating the pattern found in Studies 1 and 2 for these ROIs. Overlapping activations when employing the whole-brain inclusionary masks derived from Studies 1 and 2 were also evident in several additional regions of interest (Table 1).

Fig. 2.

Fig. 2

Open in a new tab

Bars represent the mean beta weights of 8 mm clusters centered over the peaks of the three amygdala ROIs derived from Studies 1 and 2 (i.e. left amygdala in Study 1, bilateral dorsal amygdala/SI in Study 2).

Table 1.

ROI masks derived from Studies 1 and 2 to show overlapping activation in Study 3 (height: p < .005, uncorrected; extent: 10 voxels)

Anatomical location MNI coordinates
x y z t-value Extent
1 s presentation
Direct minus averted
    L. cuneus −18 −66 20 3.8 27
    L. cerebellum −8 −46 −8 3.87 19
    R. cerebellum 12 −38 −8 4.39 119
    R. posterior cingulate 10 −28 44 4.12 41
    L. posterior cingulate −18 −26 36 7.67 220
    L. superior temporal sulcus −52 −26 0 5.38 173
    R. superior temporal sulcus 58 −26 10 3.84 12
    R. midbrain 2 −16 −8 4.01 11
    R. claustrum 34 −12 −6 4.25 29
    L. amygdala −28 8 −26 3.95 7a
    R. posterior orbitofrontal cortex 18 14 −16 3.57 16
Averted minus direct
    N/A
300 ms Presentation
Direct minus averted
    N/A
Averted minus direct
    L. cerebellum −20 −66 −44 6.42 283
    R. middle temporal gyrus 42 −64 14 3.59 20
    L. temporo-parietal junction −46 −40 26 4.57 60
    L. postcentral gyrus −16 −38 58 4.13 36
    R. cingulate 20 −32 60 3.49 11
    Midbrain 0 −30 −18 3.97 28
    L. thalamus −12 −26 −4 6.01 226
    R. temporo-parietal junction 44 −26 20 4.22 40
    R. thalamus 6 −20 −2 3.4 15
    L. superior temporal gyrus −60 −14 −2 3.7 25
    L. premotor cortex −30 −12 40 4.26 83
    L. cingulate −8 −10 52 4.25 55
    R. supplementary motor area 30 −4 46 3.8 44
    R. dorsal amygdala/SI 30 0 −8 4.03 81
    R. caudate 6 2 8 3.7 21
Open in a new tab

Study 1 was used to derive a mask to examine the 1 s condition and Study 2 for the 300 ms condition. Regions are reported posterior to anterior.

aMonte Carlo simulations have previously determined that a threshold of P < 0.01, extent = 5 voxels corresponds to P < 0.05, small-volume correction for bilateral amygdalae (see Kim et al., 2003). When following this approach, P < 0.01, the cluster extent here = 18 voxels.

DISCUSSION

Amygdala responses have been documented in early, presumably reflexive, detection of highly salient threat cues (LeDoux, 1998; Morris et al., 1998; Adolphs and Tranel, 2000). Research has also implicated the amygdala in later, top–down modulated, processing of ambiguity, such as in high-level decision making (Hsu et al., 2005), responses to complex visual stimuli (Hamann et al., 2002) and ambiguity surrounding threat (Adams et al., 2003). The current results are consistent with these findings, suggesting that amygdala responses to averted vs direct gaze fear expressions vary depending on whether presented rapidly and thus presumably engaging reflexive responses, vs more slowly, presumably allowing for more reflective responses to the threat displays.

Specifically, Study 1 revealed greater left amygdala activation to direct vs averted fear using 1 s presentation durations, replicating Adams et al.’s (2003) original pattern of findings. Using these same stimuli, we also replicated the reverse pattern of activation in Study 2 using 300 ms stimulus durations (see Hadjikhani et al., 2008). Specifically, we found greater response to averted vs direct fear in bilateral dorsal amygdalae/SI, which has been previously implicated in threat vigilance and action preparedness (see Whalen et al., 2001; Kim et al., 2003). Study 3, employing identical trial timing parameters as Studies 1 and 2 with a unique stimulus set, resulted in highly reliable and consistent replications of these effects in both the left amygdala response during sustained presentations and the right dorsal amygdala/SI response during rapid presentations. Critically, direct comparisons of these ROIs showed a significant flip in amygdala sensitivity to direct vs averted gaze fear, displays as a function of presentation duration. These findings, therefore, offer both a direct replication of Studies 1 and 2, while at the same time providing direct evidence consistent with an early response system tuned to clear threat and a later response system tuned to threat-related ambiguity.

A pattern of amygdala activation that emerged across Studies 1 and 2 that was replicated in Study 3, might suggest evidence for a laterality effect. Specifically, reflective threat perception was found to be more left lateralized (Studies 1 and 3–1 s condition) whereas reflexive processing was found to be more right lateralized (Studies 2 and 3–300 ms condition). This finding is consistent with work showing more right-lateralized amygdala activation to subliminal threat vs more left-lateralized amygdala activation to supraliminal threat (see Morris et al., 1998). These findings are also consistent with the notion of a dual process threat response where threat cues are processed along dissociable pathways. However, we urge strong caution when interpreting the pattern of laterality in the current studies. Laterality effects for amygdala responses to rapidly presented threat remain mixed (e.g. Hadjikhani et al., 2008; Carlson et al., 2009). Further, a direct comparison of activations across our three ROIs in Study 3 yielded no statistical evidence for laterality, even though when tested independently, the activations reaching threshold across the timing conditions did seem consistent with a laterality account. In general, patterns of lateralized amygdala responses are often based on activations that do or do not reach threshold. Future work examining direct statistical comparisons of left vs right amygdala activation will be necessary to clarify the role that laterality may play in reflexive vs reflective threat responses.

Given that the current set of studies examined differential responses only to fear expressions, future research will also be necessary to determine the extent to which these findings generalize to other clear vs ambiguous threat-gaze pairs, such as those involving anger and disgust expressions. The current findings also should not be taken to imply that the same timing parameters will apply to all expressions, paradigms, or individuals. Overall intensity of expression and salience of gaze as visual cues, as well as the relative discriminability of these cues, can influence how readily threat is detected and even whether gaze is integrated in the threat response at all (see Graham and LaBar, 2007). Differences in amygdala responses have also been found when viewing fear faces from different cultural groups (Chiao et al., 2008; Adams et al., 2010c). Amygdala responses have further been found to be sensitive to a variety of individual differences including temperament (Schwartz et al., 2003), state and trait anxiety (Etkin et al., 2004), social phobia (Stein et al., 2002), sex differences (McClure et al., 2004; Phan et al., 2006) and genetic variation (e.g. Bertolino et al., 2005; Canli et al., 2005). Of particular relevance to the current findings, the tendency toward attentional biases to fearful expressions are positively correlated with both state and trait anxiety (Mathews and MacLeod, 1985; MacLeod et al., 1986; Mogg et al., 2000). Individuals high in trait anxiety (Mathews et al., 2003; Putman et al., 2006; Fox et al., 2007) and trait fearfulness (Tipples, 2006) also show enhanced attentional shifts in the direction cued by fearful gaze. Even more specifically, recent work has demonstrated that gaze by emotion (anger and fear) interactions in amygdala responses are modulated by state anxiety (Ewbank et al., 2010). Given that high anxious individuals have particular problems reflectively disengaging from threatening stimuli (Bishop et al., 2007), future work of this kind may offer critical insight into the reflexive/reflective nature of such responses.

Research into the interplay of reflexive and reflective threat responses remains nascent. One possibility is that reflexive and reflective processing streams have distinct neural underpinnings, allowing them to operate in parallel (e.g. Satpute and Lieberman, 2006). In the current context, such processing streams could work in tandem to produce the most efficient and adaptive threat response, allowing for rapid threat detection followed by later confirmation or disconfirmation of the threat before a behavioral response is fully engaged. Another possible mechanism has been proposed in the ‘iterative reprocessing’ model (Cunningham and Zelazo, 2007; see also Pessoa and Adolphs, 2010). This model posits that, rather than employing separate processing streams per se, the brain proceeds from reflexive to reflective processing via successive processing cycles, engaging higher-level cortical regions as the processing proceeds from reflexive to reflective. The initial few cycles may engage the amygdala and early-response cortical threat detection regions, such as the orbitofrontal cortex (OFC) (Thorpe et al., 1983; Tomita et al., 1999). As the stimulus is being evaluated through successive iterations, each cycle may incorporate more detailed information about the stimulus, engaging additional prefrontal regions (see Cunningham et al., 2004). Both the ‘dual process’ and ‘iterative reprocessing’ models predict greater activation in the lateral prefrontal cortex (LPFC), rostral anterior cingulate cortex (rACC) and medial PFC (mPFC) during more reflective processing. Our work does not address which of these approaches can best explain our findings. Closer examination of spatiotemporal dynamics and effective connectivity among these regions while performing tasks similar to ours will be very helpful in clarifying a general understanding of the mechanisms underlying reflexive and reflective responses to compound threat cues.

Another possibility is that congruent threat cues not only capture attention more readily, but due to their aversive nature, may lead to earlier attentional disengagement as compared to ambiguous threat cues (see also Mogg et al., 2007). Indirect support for this explanation is exemplified in a study conducted by Cooper and Langton (2006), using a dot probe attention task, in which they found that when presented for 100 ms, angry faces captured attention more readily than happy faces, whereas by 500 ms happy faces appeared to hold attention more than anger. The regulation explanation for these findings assumes a temporal stage model of emotion processing, with an early orienting and a later attentional maintenance response (e.g. Halgren and Marinkovic, 1995; see also Serences et al., 2005). Supporting such a temporal model, an ERP study compared subliminal vs supraliminal responses to fear faces, finding temporally and cortically dissociable activations associated with each that were argued to directly correspond to dual process models of threat detection (Liddell et al., 2004).

Methodologically, the current work holds implications for future fMRI studies of this kind by underscoring the importance of considering temporal dynamics in neural responses to threat and perhaps other socio-emotional cues. As previous research has documented that threat responses are both temporally and spatially distributed, it is critical to consider how brain responses vary across the temporal aspect of the perceptual stream. A single ‘snapshot’ of activation at a particular point along that stream will miss such variation and can thus result in seemingly discrepant, even opposing findings across studies, as was the case here. The current work offers one example of how ostensibly minor differences in presentation parameters can give rise to major discrepancies in responses acquired. Focal vs peripheral, large vs small and dynamic vs static presentations may also yield differential neural responses to threat given that these can differentially favor visual input via magnocellular and parvocellular pathways. This observation further underscores the importance of considering more broadly how presentation methods systematically influence neural responses in neuroimaging paradigms of this type.

Conceptually, this work has implications for our growing understanding of threat perception more generally. To date the interplay of reflexive and reflective processing streams in threat perception remains not well understood. What is known is that they appear to have distinct neural underpinnings, which may allow them to operate in an interactive parallel manner (see also Townsend and Wenger, 2004). Clearly, closer examination of spatiotemporal dynamics will be essential in clarifying our understanding of the threat response. Attention to similar issues within other research domains is also likely to prove fruitful, given the prevalence of dual modes of processing across a wide variety of socio-emotional, cognitive and perceptual phenomena (e.g. attention, memory, appraisal, categorization).

SUPPLEMENTARY DATA

Supplementary data are available at SCAN online.

Conflict of Interest

None declared.

Supplementary Material

Supplementary Data
supp_7_5_568__index.html (922B, html)

Acknowledgments

This work was partially supported by Social Science Research Institute research grant, the Pennsylvania State University, National Science Foundation research grant (0544533 to R.B.A. Jr.), National Institute of Mental Health grant (K01MH84011 to K.K.).

REFERENCES

  1. Adams RB, Jr, Ambady N, Macrae CN, Kleck RE. Emotional expressions forecast approach-avoidance behavior. Motivation and Emotion. 2006;30(2):177–86. [Google Scholar]
  2. Adams RB, Jr, Franklin RG., Jr Influence of emotional expression on the processing of gaze direction. Motivation and Emotion. 2009;33:106–12. [Google Scholar]
  3. Adams RB, Jr, Franklin RG, Jr, Nelson AJ, Stevenson MT. Compound social cues in face processing. In: Adams RB Jr, Ambady N, Nakayama K, Shimojo S, editors. The Science of Social Vision. New York: Oxford University Press; 2010a. pp. 90–107. [Google Scholar]
  4. Adams RB, Jr, Franklin RG, Rule NO, et al. Culture, gaze, and the neural processing of fear expressions: an fMRI investigation. Social Cognitive and Affective Neuroscience. 2010b;5:340–8. doi: 10.1093/scan/nsp047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Adams RB, Jr, Gordon HL, Baird AA, Ambady N, Kleck RE. Effects of gaze on amygdala sensitivity to anger and fear faces. Science. 2003;300:1536. doi: 10.1126/science.1082244. [DOI] [PubMed] [Google Scholar]
  6. Adams RB, Jr, Kleck RE. Perceived gaze direction and the processing of facial displays of emotion. Psychological Science. 2003;14:644–7. doi: 10.1046/j.0956-7976.2003.psci_1479.x. [DOI] [PubMed] [Google Scholar]
  7. Adams RB, Jr, Kleck RE. Effects of direct and averted gaze on the perception of facially communicated emotion. Emotion. 2005;5:3–11. doi: 10.1037/1528-3542.5.1.3. [DOI] [PubMed] [Google Scholar]
  8. Adams RB, Jr, Nelson AJ. Intersecting identities and expressions: the compound nature of social perception. In: Decety J, Cacioppo J, editors. The Handbook of Social Neuroscience. New York: Oxford University Press; 2011. [Google Scholar]
  9. Adams RB, Jr, Rule N, Franklin RG, Jr, et al. Cross-cultural reading the mind in the eyes: An fMRI investigation. Journal of Cognitive Neuroscience. 2010c;22:97–108. doi: 10.1162/jocn.2009.21187. [DOI] [PubMed] [Google Scholar]
  10. Adolphs R. Neural systems for recognizing emotion. Current Opinion in Neurobiology. 2002a;12:169–77. doi: 10.1016/s0959-4388(02)00301-x. [DOI] [PubMed] [Google Scholar]
  11. Adolphs R. Recognizing emotion from facial expressions: psychological and neurological mechanisms. Behavioral and Cognitive Neuroscience Reviews. 2002b;1:21–62. doi: 10.1177/1534582302001001003. [DOI] [PubMed] [Google Scholar]
  12. Adolphs R, Tranel D. Emotion recognition and the human amygdala. In: Aggleton JP, editor. The Amygdala: A Functional Analysis. USA: Oxford University Press; 2000. pp. 587–630. [Google Scholar]
  13. Benton CP. Rapid reactions to direct and averted facial expressions of fear and anger. Visual Cognition. 2010;18:1298–319. [Google Scholar]
  14. Bertolino A, Arciero G, Rubino V, et al. Variation of human amygdala response during threatening stimuli as a function of 5′HTTLPR genotype and personality style. Biological Psychiatry. 2005;57:1517–25. doi: 10.1016/j.biopsych.2005.02.031. [DOI] [PubMed] [Google Scholar]
  15. Bishop SJ, Jenkins R, Lawrence AD. Neural processing of fearful faces: effects of anxiety are gated by perceptual capacity limitations. Cerebral Cortex. 2007;17:1595–603. doi: 10.1093/cercor/bhl070. [DOI] [PubMed] [Google Scholar]
  16. Canli T, Congdon E, Gutknecht L, Constable RT, Lesch KP. Amygdala responsiveness is modulated by tryptophan hydroxylase-2 gene variation. Journal of Neural Transmission. 2005;112:1479–85. doi: 10.1007/s00702-005-0391-4. [DOI] [PubMed] [Google Scholar]
  17. Carlson JM, Reinke KS, Habib R. A left amygdala mediated network for rapid orienting to masked fearful faces. Neuropsychologia. 2009;47:1386–9. doi: 10.1016/j.neuropsychologia.2009.01.026. [DOI] [PubMed] [Google Scholar]
  18. Chiao JY, Iidaka T, Gordon HL, et al. Cultural specificity in amygdala response to fear faces. Journal of Cognitive Neuroscience. 2008;20:2167–74. doi: 10.1162/jocn.2008.20151. [DOI] [PubMed] [Google Scholar]
  19. Conway CA, Jones BC, DeBruine LM, et al. Salience of emotional displays of danger and contagion in faces is enhanced when progesterone levels are raised. Hormones and Behavior. 2007;51:202–6. doi: 10.1016/j.yhbeh.2006.10.002. [DOI] [PubMed] [Google Scholar]
  20. Cooper RM, Langton SR. Attentional bias to angry faces using the dot-probe task? It depends when you look for it. Behaviour Research and Therapy. 2006;44:1321–9. doi: 10.1016/j.brat.2005.10.004. [DOI] [PubMed] [Google Scholar]
  21. Coull JT, Nazarian B, Vidal F. Timing, storage, and comparison of stimulus duration engage discrete anatomical components of a perceptual timing network. Journal of Cognitive Neuroscience. 2011;20:2185–97. doi: 10.1162/jocn.2008.20153. [DOI] [PubMed] [Google Scholar]
  22. Cunningham WA, Raye CL, Johnson MK. Implicit and explicit evaluation: fMRI Correlates of valence, emotional intensity, and control in the processing of attitudes. Journal of Cognitive Neuroscience. 2004;16:1717–29. doi: 10.1162/0898929042947919. [DOI] [PubMed] [Google Scholar]
  23. Cunningham WA, Zelazo PD. Attitudes and evaluations: a social cognitive neuroscience perspective. Trends in Cognitive Sciences. 2007;11:97–104. doi: 10.1016/j.tics.2006.12.005. [DOI] [PubMed] [Google Scholar]
  24. Davis M. The role of the amygdala in fear and anxiety. Annual Review of Neuroscience. 1992;15:353–75. doi: 10.1146/annurev.ne.15.030192.002033. [DOI] [PubMed] [Google Scholar]
  25. Driver J, Davis G, Ricciardelli P, Kidd P, Maxwell E, Baron-Cohen S. Gaze perception triggers reflexive visuospatial orienting. Visual Cognition. 1999;6(5):509–40. [Google Scholar]
  26. Ekman P, Friesen WV. Pictures of Facial Affect. Palo Alto, CA: Consulting Psychologists Press; 1976. [Google Scholar]
  27. Etkin A, Klemenhagen KC, Dudman JT, et al. Individual differences in trait anxiety predict the response of the basolateral amygdala to unconsciously processed fearful faces. Neuron. 2004;44:1043–55. doi: 10.1016/j.neuron.2004.12.006. [DOI] [PubMed] [Google Scholar]
  28. Ewbank MP, Fox E, Calder AJ. The interaction between gaze and facial expression in the amygdala and extended amygdala is modulated by anxiety. Frontiers in Human Neuroscience. 2010;4:56. doi: 10.3389/fnhum.2010.00056. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Fox E, Mathews A, Calder A, Yiend J. Anxiety and sensitivity to gaze direction in emotionally expressive faces. Emotion. 2007;7:478–86. doi: 10.1037/1528-3542.7.3.478. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Graham R, LaBar KS. Garner interference reveals dependencies between emotional expression and gaze in face perception. Emotion. 2007;7:296–313. doi: 10.1037/1528-3542.7.2.296. [DOI] [PubMed] [Google Scholar]
  31. Hadjikhani N, Hoge R, Snyder J, de Gelder B. Pointing with the eyes: the role of gaze in communicating danger. Brain and Cognition. 2008;68:1–8. doi: 10.1016/j.bandc.2008.01.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Halgren E, Marinkovic K. Neurophysiological networks integrating human emotions. In: Gazzaniga M, editor. The Cognitive Neurosciences. Cambridge, MA: MIT Press; 1995. pp. 1137–51. [Google Scholar]
  33. Hamann SB, Ely TD, Hoffman JM, Kilts CD. Ecstasy and agony: activation of the human amygdala in positive and negative emotion. Psychological Science. 2002;13:135–41. doi: 10.1111/1467-9280.00425. [DOI] [PubMed] [Google Scholar]
  34. Hess U, Adams RB, Jr, Kleck RE. Looking at you or looking elsewhere: the influence of head orientation on the signal value of emotional facial expressions. Motivation and Emotion. 2007;31:137–44. [Google Scholar]
  35. Hsu M, Bhatt M, Adolphs R, Tranel D, Camerer CF. Neural systems responding to degrees of uncertainty in human decision-making. Science. 2005;310:1680–3. doi: 10.1126/science.1115327. [DOI] [PubMed] [Google Scholar]
  36. Jones BC, DeBruine LM, Main JC, et al. Facial cues of dominance modulate the short-term gaze-cuing effect in human observers. Proceedings of the Royal Society B: Biological Sciences. 2010;277:617–24. doi: 10.1098/rspb.2009.1575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Kim H, Somerville LH, Johnstone T, Alexander AL, Whalen PJ. Inverse amygdala and medial prefrontal cortex responses to surprised faces. NeuroReport. 2003;14:2317–22. doi: 10.1097/00001756-200312190-00006. [DOI] [PubMed] [Google Scholar]
  38. LeDoux JE. Emotion: clues from the Brain. Annual Reviews in Psychology. 1995;46:209–35. doi: 10.1146/annurev.ps.46.020195.001233. [DOI] [PubMed] [Google Scholar]
  39. LeDoux JE. The Emotional Brain. New York: Touchstone; 1998. [Google Scholar]
  40. Liddell BJ, Williams LM, Rathjen J, Shevrin H, Gordon E. A temporal dissociation of subliminal versus supraliminal fear perception: An event-related potential study. Journal of Cognitive Neuroscience. 2004;16:479–86. doi: 10.1162/089892904322926809. [DOI] [PubMed] [Google Scholar]
  41. Lieberman MD. Reflective and reflexive judgment processes: A social cognitive neuroscience approach. In: Forgas JP, Williams KR, von Hippel W, editors. Social judgments: Implicit and Explicit Processes. New York: Cambridge University Press; 2003. pp. 44–67. [Google Scholar]
  42. Lieberman MD, Cunningham W. Type I and Type II error concerns in fMRI research: re-balancing the scale. Social Cognitive and Affective Neuroscience. 2009;4:423–8. doi: 10.1093/scan/nsp052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Lieberman MD, Gaunt R, Gilbert DT, Trope Y. Reflexion and reflection: a social cognitive neuroscience approach to attributional inference. Advances in Experimental Social Psychology. 2002;34:199–249. [Google Scholar]
  44. MacLeod C, Mathews A, Tata P. Attentional bias in emotional disorders. Journal of Abnormal Psychology. 1986;95:15–20. doi: 10.1037//0021-843x.95.1.15. [DOI] [PubMed] [Google Scholar]
  45. Mathews A, Fox E, Yiend J, Calder A. The face of fear: effects of eye gaze and emotion on visual attention. Visual Cognition. 2003;10:823–35. doi: 10.1080/13506280344000095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Mathews A, MacLeod C. Selective processing of threat cues in anxiety states. Behaviour Research and Therapy. 1985;23:563–9. doi: 10.1016/0005-7967(85)90104-4. [DOI] [PubMed] [Google Scholar]
  47. McClure EB, Monk CS, Nelson EE, et al. A developmental examination of gender differences in brain engagement during evaluation of threat. Biological Psychiatry. 2004;55:1047–55. doi: 10.1016/j.biopsych.2004.02.013. [DOI] [PubMed] [Google Scholar]
  48. Mogg K, Garner M, Bradley BP. Anxiety and orienting of gaze to angry and fearful faces. Biological Psychology. 2007;76:163–9. doi: 10.1016/j.biopsycho.2007.07.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Mogg K, Millar N, Bradley BP. Biases in eye movements to threatening facial expressions in generalized anxiety disorder and depressive disorder. Journal of Abnormal Psychology. 2000;109:695–704. doi: 10.1037//0021-843x.109.4.695. [DOI] [PubMed] [Google Scholar]
  50. Morris JS, DeGelder B, Weiskrantz L, Dolan RJ. Differential extrageniculostriate and amygdala responses to presentation of emotional faces in a cortically blind field. Brain. 2001;124:1241–52. doi: 10.1093/brain/124.6.1241. [DOI] [PubMed] [Google Scholar]
  51. Morris JS, Öhman A, Dolan RJ. Conscious and unconscious emotional learning in the human amygdala. Nature. 1998;393:467–70. doi: 10.1038/30976. [DOI] [PubMed] [Google Scholar]
  52. Morris JS, Öhman A, Dolan RJ. A subcortical pathway to the right amygdala mediating “unseen” fear. Proceedings of the National Academy of Sciences. 1999;96:1680–5. doi: 10.1073/pnas.96.4.1680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. N'Diaye K, Sander D, Vuilleumier P. Self-relevance processing in the human amygdala: Gaze direction, facial expression, and emotion intensity. Emotion. 2009;9:798–806. doi: 10.1037/a0017845. [DOI] [PubMed] [Google Scholar]
  54. Palermo R, Rhodes G. Are you always on my mind? A review of how face perception and attention interact. Neuropsychologia. 2007;45:75–92. doi: 10.1016/j.neuropsychologia.2006.04.025. [DOI] [PubMed] [Google Scholar]
  55. Peigneux P, Van der Linden M, Garraux G, et al. Imaging a cognitive model of apraxia: the neural substrate of gesture–specific cognitive processes. Human Brain Mapping. 2004;21:119–42. doi: 10.1002/hbm.10161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Pessoa L, Adolphs R. Emotion processing and the amygdala: from a “low road” to “many roads” of evaluating biological significance. Nature Reviews Neuroscience. 2010;11:773–83. doi: 10.1038/nrn2920. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Phan KL, Fitzgerald DA, Nathan PJ, Tancer ME. Association between amygdala hyperactivity to harsh faces and severity of social anxiety in generalized social phobia. Biological Psychiatry. 2006;59:424–9. doi: 10.1016/j.biopsych.2005.08.012. [DOI] [PubMed] [Google Scholar]
  58. Posner MI. Orienting of attention. The Quarterly Journal of Experimental Psychology. 1980;32:3–25. doi: 10.1080/00335558008248231. [DOI] [PubMed] [Google Scholar]
  59. Putman P, Hermans E, van Honk J. Anxiety meets fear in perception of dynamic expressive gaze. Emotion. 2006;6:94–102. doi: 10.1037/1528-3542.6.1.94. [DOI] [PubMed] [Google Scholar]
  60. Rigato S, Farroni T, Johnson MH. The shared signal hypothesis and neural responses to expressions and gaze in infants and adults. Social Cognitive and Affective Neuroscience. 2010;5:88–97. doi: 10.1093/scan/nsp037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Sander D, Grandjean D, Kaiser S, Wehrle T, Scherer KR. Interaction effects of perceived gaze direction and dynamic facial expression: evidence for appraisal theories of emotion. European Journal of Cognitive Psychology. 2007;19:470–80. [Google Scholar]
  62. Sato W, Yoshikawa S, Kochiyama T, Matsumura M. The amygdala processes the emotional significance of facial expressions: an fMRI investigation using the interaction between expression and face direction. NeuroImage. 2004;22:1006–13. doi: 10.1016/j.neuroimage.2004.02.030. [DOI] [PubMed] [Google Scholar]
  63. Satpute AB, Lieberman MD. Integrating automatic and controlled processes into neurocognitive models of social cognition. Brain Research. 2006;1079:86–97. doi: 10.1016/j.brainres.2006.01.005. [DOI] [PubMed] [Google Scholar]
  64. Schwartz CE, Wright CI, Shin LM, Kagan J, Rauch SL. Inhibited and uninhibited infants “grown up”: adult amygdalar response to novelty. Science. 2003;300:1952–3. doi: 10.1126/science.1083703. [DOI] [PubMed] [Google Scholar]
  65. Serences JT, Shomstein S, Leber AB, Golay X, Egeth HE, Yantis S. Coordination of voluntary and stimulus-driven attentional control in human cortex. Psychological Science. 2005;16:114–22. doi: 10.1111/j.0956-7976.2005.00791.x. [DOI] [PubMed] [Google Scholar]
  66. Stein MB, Goldin PR, Sareen J, Zorrilla LTE, Brown GG. Increased amygdala activation to angry and contemptuous faces in generalized social phobia. Archives of General Psychiatry. 2002;59:1027–34. doi: 10.1001/archpsyc.59.11.1027. [DOI] [PubMed] [Google Scholar]
  67. Straube T, Langohr B, Schmidt S, Mentzel H, Miltner WHR. Increased amygdala activation to averted versus direct gaze in humans is independent of valence of facial expression. Neuroimage. 2009;49:2680–6. doi: 10.1016/j.neuroimage.2009.10.074. [DOI] [PubMed] [Google Scholar]
  68. Thorpe SJ, Rolls ET, Maddison S. The orbitofrontal cortex: Neuronal activity in the behaving monkey. Experimental Brain Research. 1983;49:93–115. doi: 10.1007/BF00235545. [DOI] [PubMed] [Google Scholar]
  69. Tipples J. Fear and fearfulness potentiate automatic orienting to eye gaze. Cognition and Emotion. 2006;20:309–20. [Google Scholar]
  70. Tomita H, Ohbayashi M, Nakahara K, Hasegawa I, Miyashita Y. Top-down signal from prefrontal cortex in executive control of memory retrieval. Nature. 1999;401:699–703. doi: 10.1038/44372. [DOI] [PubMed] [Google Scholar]
  71. Tottenham N, Tanaka J, Leon AC, et al. The NimStim set of facial expressions: judgments from untrained research participants. Psychiatry Research. 2009;168:242–9. doi: 10.1016/j.psychres.2008.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Townsend JT, Wenger MJ. A theory of interactive parallel processing: new capacity measures and predictions for a response time inequality series. Psychological Review. 2004;111:1003–35. doi: 10.1037/0033-295X.111.4.1003. [DOI] [PubMed] [Google Scholar]
  73. Vuilleumier P. Facial expression and selective attention. Current Opinion in Psychiatry. 2002;15:291–300. [Google Scholar]
  74. Vuilleumier P, Armony JL, Driver J, Dolan RJ. Distinct spatial frequency sensitivities for processing faces and emotional expressions. Nature Neuroscience. 2003;6:624–31. doi: 10.1038/nn1057. [DOI] [PubMed] [Google Scholar]
  75. Whalen PJ. Fear, vigilance, and ambiguity: initial neuroimaging studies of the human amygdala. Current Directions in Psychological Science. 1998;7:177–88. [Google Scholar]
  76. Whalen PJ, Shin LM, McInerney SC, Fischer H, Wright CI, Rauch SL. A functional MRI study of human amygdala responses to facial expressions of fear versus anger. Emotion. 2001;1:70–83. doi: 10.1037/1528-3542.1.1.70. [DOI] [PubMed] [Google Scholar]
  77. Williams LM, Liddell BJ, Rathjen J, et al. Mapping the time course of nonconscious and conscious perception of fear: An integration of central and peripheral measures. Human Brain Mapping. 2004;21:64–74. doi: 10.1002/hbm.10154. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_7_5_568__index.html (922B, html)
supp_nsr038_Supp_materials.pdf (73.9KB, pdf)

Articles from Social Cognitive and Affective Neuroscience are provided here courtesy of Oxford University Press

RESOURCES