www.elsevier.

com/locate/ynimg NeuroImage 23 (2004) 483 – 499

For better or for worse: neural systems supporting the cognitive down- and up-regulation of negative emotion
Kevin N. Ochsner,a,* Rebecca D. Ray,b Jeffrey C. Cooper,b Elaine R. Robertson,b Sita Chopra,b John D.E. Gabrieli,b,c and James J. Grossb
a

Department of Psychology, Columbia University, United States Department of Psychology, Stanford University, United States c Neuroscience Program, Stanford University, United States
b

Received 11 September 2003; revised 11 June 2004; accepted 22 June 2004

Functional neuroimaging studies examining the neural bases of the cognitive control of emotion have found increased prefrontal and decreased amygdala activation for the reduction or down-regulation of negative emotion. It is unknown, however, (1) whether the same neural systems underlie the enhancement or up-regulation of emotion, and (2) whether altering the nature of the regulatory strategy alters the neural systems mediating the regulation. To address these questions using functional magnetic resonance imaging (fMRI), participants up- and down-regulated negative emotion either by focusing internally on the self-relevance of aversive scenes or by focusing externally on alternative meanings for pictured actions and their situational contexts. Results indicated (1a) that both up- and down-regulating negative emotion recruited prefrontal and anterior cingulate regions implicated in cognitive control, (1b) that amygdala activation was modulated up or down in accord with the regulatory goal, and (1c) that up-regulation uniquely recruited regions of left rostromedial PFC implicated in the retrieval of emotion knowledge, whereas down-regulation uniquely recruited regions of right lateral and orbital PFC implicated in behavioral inhibition. Results also indicated that (2) self-focused regulation recruited medial prefrontal regions implicated in internally focused processing, whereas situation-focused regulation recruited lateral prefrontal regions implicated in externally focused processing. These data suggest that both common and distinct neural systems support various forms of reappraisal and that which particular prefrontal systems modulate the amygdala in different ways depends on the regulatory goal and strategy employed. D 2004 Elsevier Inc. All rights reserved. Keywords: Emotion; Neural system; Up- and down-regulation

Introduction To cope with trying times, individuals employ a wide variety of emotion-regulatory strategies (Gross, 1998; Ochsner and Gross, 2004). One common strategy alters the trajectory of an unfolding emotional response by mentally transforming, or reappraising, the meaning of the emotion-eliciting situation. In comparison to other regulatory strategies, such as suppressing emotion expressive behavior, reappraisal appears to provide one of the most flexible and effective means of diminishing the negative impact of an aversive event (Gross, 2002; Gross and Levenson, 1993; Richards and Gross, 2000). Reappraisal is used not just to mentally better our bad situations, however. It also is used to mentally make bad situations worse. In some cases, cognitively up-regulating negative emotion may be desirable, as when athletes cultivate aggression before a big game. In others it may be understandable, as when we imagine the worst while passing a traffic accident. And in still others it may be maladaptive, as our worry mushrooms into anxiety and depression (Nolen-Hoeksema, 2000; Segerstrom et al., 2000). Whether used to make ourselves feel better or worse, reappraisal plays an important regulatory role in maintaining our mental and physical well-being (Davidson et al., 2000; Gross, 1998). Despite growing recognition of reappraisal’s importance, however, little is known about the underlying psychological and neural mechanisms. A model of the brain organization of emotional reappraisal (Ochsner et al., 2002) was proposed by analogy to well-studied forms of cognitive control such as working memory and response selection (D’Esposito et al., 2000; Miller and Cohen, 2001). On this view, the emotion-modulatory effects of reappraisal stem from interactions between cognitive control processes implemented in prefrontal and cingulate regions and emotional appraisal processes implemented in multiple emotion-related structures, including the amygdala (Ochsner and Gross, 2004; Ochsner et al., 2002). This model’s predictions are consistent with the handful of neuroimaging studies that have investigated the neural correlates of reappraisal. Three studies have examined the down-regulation of either aversive affect (Ochsner et al., 2002), sadness (Levesque et al., 2003), or sexual arousal (Beauregard et al., 2001), and all found

* Corresponding author. Department of Psychology, Columbia University, 369 Schermerhorn Hall, New York, NY 10027. Fax: +1 413 845 3695. E-mail address: ochsner@psych.columbia.edu (K.N. Ochsner). Available online on ScienceDirect (www.sciencedirect.com.) 1053-8119/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2004.06.030

and in general heightening their sense of personal or subjective experience for the sights and sounds of the event they were viewing... there also is reason to believe that the up. Kelley et al. and it is therefore not known whether up-regulation involves neural systems similar to those underlying the down-regulation of emotion. 2001) or to maintain an emotion (Schaefer et al.. Hamann et al.. arousing... there has been no systematic attempt to test this possibility..and down-regulation of emotion will depend upon distinct prefrontal control systems because the reinterpretations used to make oneself feel better or worse may differ in systematic ways. Rolls. or that she was present observing that individual and their agonized expression of pain as they lay in their hospital bed. A fourth study found that reappraisal could maintain amygdala activation to aversive photos after they disappeared (Schaefer et al. To systematically examine the effects of different types of cognitive reappraisal. We used only one sex because several studies have suggested that there are sex differences in response to negative visual materials (e. for example. whereas others provided an instruction either to suppress emotion by becoming detached (Beauregard et al. 2001. The distinction between self. By directly contrasting activations related to increasing or decreasing negative emotion via cognitive reappraisal. 2002).. Part of this variability may be attributable to the use of different reappraisal strategies by participants. Lane et al. Ochsner et al. 1998. To date. 2002). Ochsner and Gross. albeit by different means.and downregulation of emotion elicited by aversive images involve differential modulation of the amygdala. and prior work suggests that these processes depend upon prefrontal cortex (Ochsner and Gross. Whereas self-focused strategies alter the personal relevance of events. Behavioral paradigm Participants were instructed to increase and decrease their negative emotions in response to aversive images using one of two strategies (see Fig. emotion. When decreasing negative emotion.and situation-focused reappraisal strategies builds on psychological theories that demonstrate the special role for self-relevant information in memory. the specific prefrontal and emotion-processing-related activations have varied across studies. 2001. Different reappraisal strategies may recruit different cognitive control systems and thus be associated with different patterns of brain activation. making one feel more or less connected to what is going on. prior work has shown that using reappraisal to down-regulate negative emotion diminishes amygdala activity (Ochsner et al. imagining themselves or a loved one as the central figure in a photo. participants in the self-focus group were instructed to increase their sense of objective distance. Participants assigned to the self-focus group were instructed to think about the personal relevance of each image as it appeared. When generating a reappraisal that makes one feel worse. imagining themselves present as pictured actions unfolded. Whalen et al. Some studies specifically instructed participants how to actively reinterpret the meaning of events depicted in stimuli (Ochsner et al. 2001).. one might retrieve emotion knowledge and emotion labels that can be used to describe an event in increasingly negative terms. 1A). 2004. 2000. Methods Participants Twenty-four female participants (M age = 20. and unlikely to heal. a participant viewing an image of a sick person in the hospital might increase her negative emotion by imagining that she could be that individual. 1997). they were to increase their sense of subjective closeness to pictured events. By contrast.and situation-focused types of cognitive reappraisal.. 2000. viewing pictured events from a detached. 2002) without specifying exactly how participants were to do this. As of yet.g. with cognitive up-regulation associated with greater amygdala activation and cognitive downregulation associated with lesser amygdala activation. situation-focused strategies reinterpret the nature of the events themselves. terribly painful. LeDoux. On the cognitive control side of reappraisal. Gusnard et al. 2002. the present study used functional magnetic resonance imaging (1) to directly compare the effect of using reappraisal to down. 2003. 2000). To decrease negative emotion participants could . Thus. Canli et al. Morris et al. we hypothesized that cognitive control regions supporting reappraisal would modulate regions involved in appraising the affective qualities of stimuli and events.g... 2002). However. and outcomes. 2004. both within and across studies.N. Although all four of these studies found prefrontal and/or amygdala participation in reappraisal. 1998) and an emerging cognitive neuroscience literature that has identified distinct medial prefrontal systems important for self-referential processing (e.and up-regulate negative emotion. Ochsner et al. To clarify and extend our understanding of the neural bases of reappraisal. which plays an important and pervasive role in encoding evocative.. The present study sought to extend prior findings by determining whether the up. 2002) but did not report maintenance-related prefrontal activations. 2003.. On the emotional appraisal side of reappraisal. third-person perspective. and the present study sought to determine the neural correlates of self... no studies have examined the use of reappraisal to up-regulate emotion. Either type of strategy can be used to increase or decrease emotion. Thus. and especially aversive events (Anderson and Phelps... Anderson et al. the present study can help identify their common and distinct neural correlates.484 K. / NeuroImage 23 (2004) 483–499 reappraisal-related prefrontal activations in combination with modulations of regions associated with different types of emotion processing. Consistent with this hypothesis. 1998).6 years) were recruited in compliance with the human subjects regulations of Stanford University and were paid US$60 for voluntary completion of this study. Cahill et al. one might add insult to the injury depicted in an image by describing it as horrific. When increasing negative affect. and (2) to examine the effects of two different reappraisal strategies that were equally effective at achieving these regulatory goals. and motivation (Baumeister. including the amygdala. there is reason to believe that the up.. we asked participants to reappraise aversive images using strategies that either focused on the self-relevance of an event or focused on situational aspects of the event.and down-regulation of emotion will depend upon similar prefrontal control systems because both involve generating an alternative interpretation for the meaning of an event. 2002. dispositions. the down-regulation of negative emotion is not likely to involve the retrieval of emotion knowledge per se. however. 2000. and we wanted to avoid this source of variability. reevaluating others’ actions. O’Doherty et al. but rather may involve mediating interference between negative bottom-up appraisals of stimuli and neutralizing top–down reappraisals of them (Bechara et al.

1993) and were balanced for valence and arousal across instruction types. 1A). look at a stimulus and respond naturally) and strategy group (employing a reappraisal strategy that either focuses on reinterpreting aspects of a pictured situation or on reinterpreting the self-relevance of pictured events).N.K. it was emphasized that the participant should imagine events in the image getting worse. Ochsner et al. and is likely to recover quickly. Third. decrease. Participants assigned to the situation-focused group were instructed to reinterpret the emotions. participants were instructed simply to look at the image and let themselves respond naturally. On increase and decrease trials. actions. participants reappraised images as described above. participants in the situation-focused group were asked to imagine pictured events getting better. / NeuroImage 23 (2004) 483–499 485 Fig. decrease. or look) and two valences of stimuli (negative and neutral. Each trial was composed of four events (Fig. Each image was shown only once for a given participant. and tragically is unlikely ever to recover. has been suffering for quite a long time. view the sick person from the detached. or in a baseline condition. Participants then provide a rating of their current negative affect and finally have a moment to relax before the onset of the next trial. in the case of the sick person pictured in the hospital. (A) Schematic diagram showing the six different conditions in the present study defined by the crossing of two factors: type of instruction (to increase emotion. Twenty-seven images were shown for each of six types of trials defined by crossing three types of instruction (increase. clinical perspective of one not personally connected in any way to the pictured individual and the context in which she is situated. or look. or look) appeared centrally for 2 s. To increase negative emotion using this strategy. the situation-focused strategy might involve imagining that person is in great pain. and outcomes of individuals as depicted in their situational context. First. decrease. This trial type served as a baseline for comparison with the increase and decrease reappraisal trials. Fig. 1. decrease emotion.. possessed of a strong constitution. Second. and assignment of images to instruction types was counterbalanced across participants. which is followed by a photo presentation period during which participants follow this instruction. one could imagine that the sick individual is only tired. To decrease negative emotion. Participants randomly assigned to each of the two instructional groups (12 self-focused and 12 situation-focused) were cued to increase or decrease their negative emotion on a series of 162 randomly intermixed trials. a cue word (increase. An initial cue instructs participants to increase. On look trials. an aversive or neutral image appeared centrally for 10 s. Neutral images were included primarily to provide a buffer against habituation to negative images and are not a focus of the analyses presented here. (B) Timeline for events on each trial. participants performed the evaluation operations specified by the prior instructional cue. Negative and neutral images were selected from the International Affective Picture System (Lang et al. 1B). To decrease negative emotion. Thus. While the image remained on the screen. a rating scale appeared immediately after presentation of the .

The scale consisted of a horizontal rectangular bar with anchors of 0 and 7 to indicate relative strength of negative affect. A general linear model analysis was used in SPM99 to create contrast images for each participant summarizing differences between trial types. and participants were asked to rate how much effort they had exerted to perform the selfor situation-focused reappraisals. This scale allowed participants to rate the current strength of their negative affect after having looked. A high-pass filter with a cutoff period of 120 s was applied to remove drifts within sessions. This training phase ensured that participants clearly understood the specific type of strategy they were to employ inside the scanner and could effectively implement that strategy to reappraise negative images. These images were used to create SPM{t} maps for the group. The 2-s instruction period and 4-s rating period were modeled with a canonical hemodynamic response function at the onset of each period. 1-mm gap) were collected at 3T (GE Signa LX Horizon Echospeed scanner) with a T2*-sensitive gradient echo spiral-in or -out pulse sequence (30 ms TE. the word relax appeared for 4 s in the center of the screen in capital letters. 2003). In part one. participants completed a set of computer guided ratings in a separate room.. increased. 64 Â 64 data acquisition matrix). 24-cm field of view. The experimenter helped shape these reappraisals so that they fit the self. 1999). After completing the scanning session. the bar grew from left to right and participants pressed a key when the bar grew to a size that corresponded to the strength of their current negative feeling. 2001.g.001 uncorrected for multiple comparisons. Experimenters also instructed participants not to reappraise stimuli using other strategies not relevant to that participant’s group assignment.486 K. The spiral in/out sequence has been found particularly valuable in reducing susceptibility dropout in frontal and medial temporal brain regions (Glover and Law.or situation-focused strategy the participant was instructed to use. activation time courses were extracted from the peak voxel in each region. indicating that participants should relax until the next trial began. Ochsner et al. Data analysis Functional images were slice time corrected and motion corrected using SPM99 (Wellcome Department of Cognitive Neurology).g. the experimenters debriefed participants to ensure that they were able to effectively reappraise and address any questions the participant might have. An LCD projector displayed stimuli on a screen mounted on a custom head coil fitted with a bite-bar to limit head motion.N.. or changes in affect when increasing or decreasing . the participant completed a block of 21 practice trials whose length was equivalent to one of the scans the participant would later complete in the scanner. These ratings provide a measure of the extent to which participants were effortfully engaged in the task as requested. two interleaves.. Prefrontal voxels were identified from functionally defined regions of interest derived from specific contrasts of increase and decrease trials relative to the look baseline condition (described below). or decreased their negative emotion and served as a behavioral index of the success of reappraisal. the functional images were then normalized using those parameters and interpolated to 2 Â 2 Â 2 mm voxels. High order shimming was performed before functional scans using the scanner’s software (developed in the Lucas Center for GE. participants completed a separate training session in the Psychology Department at Stanford University.. This training phase had two parts. Stimulus presentation and data acquisition were controlled using Psyscope software running on a Macintosh G3 computer. / NeuroImage 23 (2004) 483–499 photo. Statistical maps for group analyses were thresholded at P b 0. Responses were made with the index finger of the right hand using one button on a four button response box. This bar provided a continuous index of participants’ subjective experience of negative emotion. 2004). MRI data acquisition Twenty-five axial slices (4-mm-thick. Glover. outlier checks were performed to identify individuals showing affect ratings. Fixed effects for each participant were modeled using a mixed design. Preston et al. Amygdala voxels were identified using a small volume correction for a structurally defined a region of interest derived from amygdala coordinates specified in the Talairach atlas and transformed into ICBM space. Anatomical images were coregistered to the mean functional image and normalized to a standard template brain. a small set of regions were hypothesized a priori to be involved in reappraisal. At the end of this practice block. Peak voxel activations were selected rather then average cluster activations because prior research has shown peak voxel activity to be more strongly correlated with electrophysiological measures of activation (Arturs and Boniface. the 10-s regulation period and 4-s relaxation period were modeled as a boxcar regressor convolved with the canonical hemodynamic response. Ochsner and Gross. 2004). 608 flip angle. To specifically characterize activation in these ROIs. 2002) and theoretical considerations (e. Results Prior to group analysis of behavioral or imaging data. During this session. Fourth. 85 ms TE). T2-weighted flowcompensated spin-echo scans were acquired for anatomical localization using the same slice prescription (2000 ms TR. Maxima are reported in ICBM152 coordinates as in SPM99. participants read a brief description of the strategy they were to employ and then viewed a series of images for which they were asked to spontaneously generate appropriate reappraisals. Functional images were smoothed with a Gaussian filter (6 mm full width-half maximum). In part two. participants received careful instruction and guidance in the performance of the reappraisal strategy they subsequently used in the scanning session. It was emphasized that participants should do their best to reappraise when asked to do so on any given trial and should accurately report the strength of their negative affect whether or not they felt reappraisal had been successful in changing the way they felt. On the basis of prior results (e. At the beginning of the 4-s rating period. Pretraining and posttesting Three to 5 days before scanning. Ochsner et al. Each image presented in the scanner was presented again. 2000 ms TR. These regions included the amygdala and prefrontal regions thought to be involved in cognitively increasing or decreasing negative emotion.

and for comparison.5 SD from the mean of the entire group.22 t (22) = 6.83. there were main effects of reappraisal instruction—participants successfully and significantly increased or decreased negative emotion relative to look baseline trials.21). . 2A.K. t (22) = 6. and success of reappraisal did not vary as a function of strategy ( F for the interaction b1. (B) Effort ratings on increase and decrease trials. Ochsner et al. decreasing was rated as more effortful.0001] than on baseline look trials.N. Overall. affect ratings for look trials with neutral photos also are shown in Fig. and decrease trials with negative photos. / NeuroImage 23 (2004) 483–499 487 emotion. and reappraisal strategy did not significantly interact with this pattern of findings. that were greater than 2. Affect reports for these trials Fig.0001] than on look trials when they let themselves respond naturally (M = 5. Overall. P b 0. participants reported significantly greater negative affect [M = 6.42) = 57. Planned comparisons demonstrated that when increasing negative emotion.62. look. participants reported significantly less negative affect [M = 4.67.94. look trials with neutral photos. (A) Self-reports of negative affect on increase. P b 0.0001] and no other significant effects (Fig. Reappraisal strategy did not significantly interact with this pattern of findings. Behavioral results Subjective reports of negative affect An ANOVA on affect ratings for negative photos with strategy group as a between-subject factor and type of instruction as a within-subject factor revealed a significant main effect of type of instruction [ F (2. 2A).62). 2. Thus. P N 0. reappraisal was successful both at increasing and decreasing negative emotion. When decreasing negative affect.39. For comparison purposes. P b 0. One participant in the situation-focused group was identified as an outlier and her data were removed from all subsequent analyses.

64. Fig.70. 2B). Postscan effort ratings Effort ratings were placed on a 100-point scale where 100 = greatest and 1 = least effort.12. including many left-sided regions similar to those used when increasing. t (22) = 6. Planned comparisons indicated that participants exerted greater effort when decreasing (M =52.0001] and no other significant effects (Fig. Ochsner et al. (C) Regions uniquely activated when increasing as revealed by the increase–decrease contrast. Of note is bilateral activation of lateral and medial prefrontal cortex. (B) Left two panels show left and right lateral views and rightmost panel shows right medial view of regions active in the decrease–look contrast.0001. for all comparisons].42) = 28. and rightmost panel shows axial view of right lateral orbitofrontal activation. t (22) N 7.67.N. / NeuroImage 23 (2004) 483–499 were significantly smaller than for any trial type using negative photos [M = 3.47) than when increasing [M = 24. P b 0.37. (A) Left two panels show left and right lateral views and rightmost panel shows left medial view of regions active in the increase–look contrast. P b 0. P b 0. An ANOVA on postscan effort ratings with strategy group as a between-subject factor and type of instruction as a within-subject factor revealed a significant main effect of type of instruction [ F (2.0001] negative affect. Left rostral medial prefrontal cortex and posterior cingulate activation are most easily observed on the rightmost medial view of the left hemisphere. Center panel shows right lateral prefrontal activation. 3. Group contrasts showing activation when increasing or decreasing emotion. Of note is the presence of left dorsal lateral and medial prefrontal as well as anterior cingulate cortical activation. (D) Regions uniquely activated when decreasing as revealed by the decrease–increase contrast. .488 K.

88 4.17 3. The overall results of each contrast showed many similarities. Local maxima for these clusters are denoted with (L). are reported.87 3.or situation-focused reappraisal when increasing or decreasing emotion.006 uncorrected.. bilateral temporal–parietal junction. and (2) identify regions uniquely involved in self. the increase–decrease contrast was masked by the increase–look contrast.54 3. regions associated with the main effects of increasing or decreasing).40 4. For the first set of contrasts.25 4. P b 0.52 3. and cerebellum. thalamus. two-tailed P b 0. P b 0.23 3. regions associated with the interaction of strategy and direction of reappraisal). whereas decrease N look activations generally were bilateral or right-lateralized and included specific activation of orbital frontal cortex.09 corrected..K.86 3. To identify increase-related regions that were not recruited when decreasing. and second. In so doing. middle temporal gyrus.34 4. The effects of cognitively increasing and decreasing emotion Recruited by reappraisal.00 2.81 4. These dissimilarities were confirmed by directly contrasting activations on increase and decrease trials (Figs.35 3. Clusters of five or more contiguous voxels whose global maxima meet a t threshold of 3.N. decrease.05 5. Regions generally involved in increasing or decreasing emotion first were identified by greater activation in response to increase or decrease as compared to look trials (Figs.88 4. This insured that only regions more active during increasing than during the baseline look condition could be identified as increase-specific by the increase–decrease contrast.e.26 4. 3A and B and Tables 1 and 2).62 3. 3C and D. including significant activation of dorsal and ventral lateral prefrontal cortex. Coordinates are in MNI space.e.72 3.93 4. anterior cingulate cortex. / NeuroImage 23 (2004) 483–499 Table 1 Group activations for increase N look contrast Region of activation Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Medial frontal gyrus Superior frontal gyrus Cingulate gyrus Medial frontal gyrus Superior frontal gyrus Medial frontal gyrus Medial frontal gyrus Inferior frontal gyrus Inferior frontal gyrus Middle frontal gyrus Postcentral gyrus Postcentral gyrus Superior temporal gyrus Middle temporal gyrus Middle temporal gyrus Superior temporal gyrus Middle temporal gyrus Middle temporal gyrus Middle temporal gyrus Middle temporal gyrus Cuneus Precuneus Thalamus Thalamus Thalamus Putamen Putamen Globus pallidus Cerebellum Posterior cingulate Precuneus Brodmann Coordinates x L6/9 L6 L6 R6 R6 R6 L6 L6 L32 L9 L10 L9 R32 L44 L44 L9 L3 R L39 L39 L39 R38 L21 L21 R39 R19 L30 L7 L L R L L R L L30/23 L31 -50 -50 -36 60 52 56 -10 -4 -6 -6 -4 -6 18 -58 -58 -44 -22 24 -50 -54 -40 48 -66 -58 60 60 -24 -6 -2 -10 20 -22 -30 14 -10 -6 -2 y 2 0 12 -2 -6 6 2 6 18 48 66 52 10 16 12 12 -30 -30 -64 -72 -72 14 -30 -4 -72 -78 -76 -62 -16 -8 -14 6 4 -4 -64 -50 -68 z 42 50 52 50 58 54 66 62 38 40 26 26 44 16 6 30 54 62 20 22 22 -12 -10 -20 20 10 8 46 12 -2 10 0 -6 -2 -24 20 22 Z score 5.67 4.54 3.98 3. the masking procedure ensures that identified regions are functionally interpretable as increaserelated by excluding from consideration any regions that were not more active when increasing negative emotion than when simply .63 3. Ochsner et al. regardless of strategy (i. we collapsed across strategy group and compared activations on increase.63 3.24 3. dorsal medial prefrontal cortex.001 uncorrected. There were some notable dissimilarities as well: increase N look activations generally were left-lateralized and included specific activation of rostral medial and posterior cingulate cortex.74 Volume (mm3) 7344 (L) (L) 896 (L) (L) 6968 (L) (L) 9600 (L) (L) 40 1728 (L) (L) 40 80 7136 (L) (L) 88 264 14 1392 (L) 400 408 944 72 48 1472 (L) 840 6816 (L) (L) Table 1 (continued ) Region of activation Cerebellum Cerebellum Cerebellum Amygdala* Brodmann Coordinates x R R R L 38 18 32 -18 y -62 -72 -74 -10 z -34 -28 -28 -14 Z score 4. and subcortical regions including caudate.43 3.46 3.33 4. Table 4). For the second set of contrasts.60 3. and look baseline trials for all 23 participants.93 3. given that such regions had been identified.58 4.61 3. Imaging results Contrasts were performed to (1) identify regions involved in increasing or decreasing in emotion.50.98 3. * Small volume corrected. to then determine which of those regions are additionally or selectively more active when increasing than when decreasing.15 4.52 489 Volume (mm3) 15136 (L) (L) 128 Note .or situation-focused reappraisal strategies (i.72 4. we contrasted activations for each group to determine whether specific regions were more involved for self.70 3.43 4.32 3. The idea behind this masking procedure was first to identify a network of regions involved in increasing negative affect using the increase–look contrast.24 4.

The reason for this is that voxel clusters modulated by up.24 4. which was used as a mask for that increase–decrease comparison.61 3.14 5.29 3. looking at an image. differential activation of the amygdala in the increase as compared to decrease conditions cannot be revealed. For the increase N look contrast.53 3.75 3. Tables 1 and 3).N.66 3.54 4.50.80 Volume (mm3) 17768 (L) (L) 80 88 120 256 3184 (L) (L) 56 88 328 (L) 368 (L) 656 (L) 416 (L) 792 (L) 392 (L) 216 3000 (L) (L) 1496 (L) (L) 264 (L) (L) 160 312 1496 (L) (L) 2192 (L) L6 L6 R6 L6 L8/9 L8 R10 L6 L8 L9 L8 L8 R8 R6/8 R9 R8 L45 L44 L47 L47 R44 R45 R47 R47 R32 R40 R39 R39 L39 L39 L39 L21 L21 L21 R21 L20 L L L R R À10 À2 10 À34 À16 À8 22 À46 À46 À56 À36 À24 52 50 42 36 À54 À58 À44 À30 58 60 50 34 12 60 62 64 À60 À54 À48 À66 À68 À66 52 À54 À16 À20 À16 18 16 Note .02 3. Similarly.14 3.70 4.04 3.490 Table 2 Group activations for decrease N look contrast Region of activation Superior frontal gyrus Superior frontal gyrus Superior frontal gyrus Superior frontal gyrus Superior frontal gyrus Superior frontal gyrus Superior frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Middle frontal gyrus Inferior frontal gyrus Inferior frontal gyrus Inferior frontal gyrus Inferior frontal gyrus Inferior frontal gyrus Inferior frontal gyrus Inferior frontal gyrus Inferior frontal gyrus Cingulate gyrus Inferior parietal lobule Superior temporal gyrus Superior temporal gyrus Middle temporal gyrus Angular gyrus Middle temporal gyrus Middle temporal gyrus Middle temporal gyrus Middle temporal gyrus Middle temporal gyrus Inferior temporal gyrus Caudate Lateral globus pallidus Caudate Caudate Lateral globus pallidus K. Ochsner et al.42 3. whereas decrease-specific regions work exclusively right lateralized and included dorsolateral and lateral orbital prefrontal cortex.006 uncorrected). Local maxima for these clusters are denoted with (L).11 5.87 3.90 3.27 3.64 3.32 3.001 uncorrected.44 4. / NeuroImage 23 (2004) 483–499 Brodmann Coordinates x y 18 16 20 12 46 46 46 6 8 8 22 24 16 6 30 22 20 14 22 22 10 18 18 24 24 À56 À68 À62 À72 À64 À60 À32 À34 À26 À32 À8 8 À2 0 10 2 z 62 62 64 56 42 48 30 54 44 40 48 46 48 46 38 44 8 14 À8 À16 12 18 À6 À16 30 42 24 30 28 32 22 À14 À6 À8 À8 À26 14 0 16 12 À6 Z score 5.13 3.86 3. significantly activated voxels were identified in the left amygdala at a slightly liberalized two-tailed threshold ( P b 0.04 3. It should be noted that because of the masking procedure used to construct these contrasts. Small-volume corrected analyses were used to test the hypothesis that amygdala activation should be upregulated (increase N look) when increasing negative emotion and down-regulated (look N decrease) when decreasing negative emotion (Fig. 4. Thus.28 4. P b 0.73 4. which may be justified given a priori interest in the amygdala’s role in reappraisal and our . amygdala regions showing diminished activation when down-regulating (as identified in the decrease– look contrast) will not show up in the increase–decrease contrast because those down-regulated voxels are not also up-regulated in the increase–look contrast. Clusters of five or more contiguous voxels whose global maxima meet a t threshold of 3. This ensured that only regions more active during decreasing than during the baseline look condition could be identified as decrease-specific by the decrease–increase contrast.29 3.09 corrected.34 3.50 3. are reported.48 4. for example.25 3. Increase-specific regions were primarily left-lateralized and included left rostral medial prefrontal cortex (BA 9/10) and the posterior cingulate cortex. Coordinates are in MNI space.66 3. Modulation by reappraisal.53 3.30 3.50 3.00 3. decrease-related regions were identified by masking the decrease–increase contrast with the decrease–look contrast. P b 0.97 3.92 3.59 3.or down-regulation of emotion do not overlap.05 3.58 4.

Significant modulation of the right amygdala was not observed when increasing ( P corrected b 0. was modulated when decreasing (Table 3).00.045 for a one-tailed comparison).and down-regulation of amygdala activity due to reappraisal. P b 0. P b 0. activation time courses for the left amygdala peak activated voxels were extracted for each contrast.12. as shown in the portion of the time course corresponding to the photo presentation period. and that increased or decreased amygdala activation was maintained while participants actively reappraised their emotion until the photo disappeared. As shown in Fig. downward modulation was significant for the left ( P b 0. t (22) b 1]. Activation was most significantly diminished on decrease as compared to look trials with negative stimuli at the beginning and at the end of the photo presentation period [for 7–8 and 13–14 s TRs.N. modulation due to reappraisal may first be observed 6–8 s after trial onset. For the look N decrease contrast. Left panels show amygdala activation from group contrasts. one-tailed t (22) N 2. and bottom panel shows left and right amygdala voxels active in the look N decrease contrast. as indicated in the figure.4. up. P b 0. bilateral insular cortex.18. T tests contrasting levels of percent signal change were used to confirm that reappraisal significantly up. decreases in negative affect when down-regulating emotion significantly correlated with decreases in both left . which respectively reflect up.006 uncorrected) amygdala. Top panel shows left amygdala voxels active in the increase N look contrast.s instruction cue.01 uncorrected). 4. for 9–10 and 11–12 s TRs. 4.60. specific one-tailed prediction that amygdala activation should be greater on increase trials ( P b 0. Activation time course for look trials with neutral photos is shown for comparison. To determine whether and how activation in reappraisal-related regions predicted reappraisalrelated affect change. correlational analyses were conducted that related self-reported increases (increase–look affect difference) or decreases (look–decrease affect difference) to a measure of activation (Beta values obtained from the SPM model fit for the photo presentation period on each trial type for each region of interest) in either the activated amygdala clusters listed in Tables 2 and 3 or in the increase. Affect-activation relations.087 corrected.and decrease-specific prefrontal regions listed in Table 4. all one-tailed t (22) N 2. t (22) N 1. An additional emotion-processing region.001 uncorrected) and marginally significant for the right (two-tailed: P b 0. and an initial 2. By contrast. Ochsner et al. Increases in negative affect when up-regulating emotion did not significantly correlate with activation in the amygdala or in increase-specific prefrontal or cingulate regions. which may reflect up. whereas right panels show percent signal change activation time courses for the left amygdala peak voxel from each contrast. for 15–16 s TR.to 6-s hemodynamic response lag.05. Activation was greater on increase as compared to look trials with negative stimuli for each 2-s TR during the photo presentation period when participants actively reappraised their emotional response [from 7–8 to 15–16 s.007 corrected. P b 0. With a 4. P b 0.05]. P b 0. / NeuroImage 23 (2004) 483–499 491 Fig. Time courses for peak voxels show significant modulation when increasing or decreasing negative emotion 6–8 s after trial onset.09 for a two-tailed comparison corresponds to P b 0. Time courses begin at the onset of a trial.or downregulation of amygdala activity within the first 2 s that participants attempted to reappraise a photo.K.or down-modulation of the amygdala began at the onset of the photo presentation period.or down-regulated amygdala activation. To further characterize reappraisal-related amygdala modulations.

88) as compared to increasing emotion [M = 0. The reason for a failure to observe affect-activation correlations when increasing negative Table 4 Group activations for increase N decrease and decrease N increase contrasts Region of activation Brodmann Coordinates x Increase N decrease Superior frontal gyrus Superior frontal gyrus Medial frontal gyrus Posterior cingulate Posterior cingulate Posterior cingulate Thalamus Globus pallidus Cerebellum Cerebellum Cerebellum Cerebellum Decrease N increase Superior frontal gyrus Superior frontal gyrus Superior frontal gyrus Middle frontal gyrus Inferior frontal gyrus Inferior frontal gyrus Inferior parietal lobule L9 L10 L10 L30 L23 L31 L R L L L L À10 À4 À4 À8 À6 0 À6 18 À18 À14 À8 À4 y 50 64 68 À64 À52 À62 À12 À8 À46 À54 À62 À44 z 34 32 24 14 24 24 10 À2 À16 À18 À18 2 Z score Volume (mm3) 4. P b 0.75 3.001 uncorrected. P b 0. Activation time courses for the peak voxel of representative medial and lateral right PFC regions are shown for the decrease– look contrast in Fig.41 3. t (22) = 3.34 3. 5. P b 0.26 3. P b 0.50. the self-focused group showed more activation in right medial PFC.001 uncorrected.83 3.49 3.44 3.001 uncorrected.47 3. amygdala (r = 0.560.492 Table 3 Group activations for look N decrease contrast Region of activation Insula Insula Insula Insula Insula Inferior parietal lobule Inferior parietal lobule Inferior parietal lobule Precuneus Amygdala* Amygdala* Amygdala** Brodmann Coordinates x L6/13/44 L13 L13 R13 R13 L40 R40 R40 R L L R À48 À38 À40 44 40 À58 50 38 24 À30 À28 20 y À4 À26 À16 À14 0 À24 À30 À58 À60 À2 À4 0 z K. are reported. SD = 0.61.27 3.97 4. The effects of specific types of cognitive emotion regulation strategies To identify regions more activated when employing either a self.43. Clusters of five or more contiguous voxels whose global maxima meet a t threshold of 3. Coordinates are in MNI space. Local maxima for these clusters are denoted with (L). P b 0.67 4. T tests contrasting percent signal change indicated (1) that only for the situation-focused group was left lateral frontal signal on decrease trials significantly greater than on 30 À20 À14 À24 Note . Because of a reduced range over which increases in negative affect could be observed.53 3.00 3. This hypothesis is supported by greater mean affect change and greater variability when decreasing (M = 1.58 3.005).505.91 3. Ochsner et al.78 3.24 4. * Small volume corrected.66 968 136 40 248 648 (L) 88 56 672 (L) (L) 96 R8 R6 R8 R9 R44 R47 R40 16 14 14 42 56 34 64 26 20 36 22 12 22 À56 56 64 50 42 16 À18 44 4.or situation-focused reappraisal strategy to increase or decrease emotion.23 4. whereas the situation-focused group showed greater activation in both left and right lateral PFC (Table 6). as well as increased activation in right orbitofrontal cortex (r = À0. it may have been more difficult to observe correlations between restricted-range affect change scores and measures of change in brain activation. are reported.29 3608 (L) (L) 312 384 40 1592 Note . Coordinates are in MNI space.86 4.22. P b 0.N. . no regions were more activated for the self-focused group.01) and right amygdala (r = 0. Local maxima for these clusters are denoted with (L).01) activation.49 4.22. These contrasts test for the interaction of strategy and type of reappraisal. Clusters of five or more contiguous voxels whose global maxima meet a t threshold of 3. regions of temporal and parietal cortex were more activated for the situation-focused group (Table 5).06 3. two sample t tests compared activations between groups for all of the contrasts described above (Tables 5 and 6). For the increase N look contrast that identified regions involved in increasing emotion.98 2. P b 0. ** Small volume corrected.45 3.087 corrected. For the decrease N look contrast that identified regions involved in decreasing emotion.65 3.01].50.007 corrected. two-tailed P b 0.50 Volume (mm3) 624 528 (L) 456 64 88 248 144 112 592 (L) 64 emotion may be related to an upper bound (or ceiling effect) on the amount that participants could increase their negative emotion above and beyond that already experienced during the look baseline condition. two-tailed P b 0.50 2.504.50 3.76 3.006 uncorrected. SD = 0. P b 0.34 3. / NeuroImage 23 (2004) 483–499 Z score 10 18 14 20 16 22 30 44 4.

Discussion Cognitive reappraisal was effective during the picture viewing paradigm such that participants significantly decreased negative Table 6 Regions showing greater activation for the situation or self-strategy group when decreasing emotion Region of activation Brodmann Coordinates x Situation (decrease–look) N self (decrease–look) Middle frontal gyrus L10 Inferior frontal gyrus L45 Middle frontal gyrus R6 Middle frontal gyrus R6 Middle frontal gyrus R46 Inferior frontal gyrus L46 Inferior frontal gyrus L44 Inferior frontal gyrus R45 Precentral gyrus L4 Precentral gyrus L6 Superior temporal gyrus L22 Superior temporal gyrus L21/22 Inferior temporal gyrus L37 Inferior temporal gyrus L19/37 Supramarginal gyrus L40 Cuneus L Middle occipital L19 Cerebellum L Cerebellum Cerebellum L Cerebellum L Cerebellum L Cerebellum L Cerebellum L Cerebellum R Cerebellum R Self (decrease–look) N situation (decrease–look) Cingulate gyrus R32 Inferior parietal gyrus L40 À36 À48 42 38 56 À50 À52 52 À28 À38 À60 À66 À46 À42 À58 À24 À26 À14 0 À10 À30 À22 À16 À12 18 10 y 38 36 0 À8 30 30 6 36 À26 À4 À34 À12 À64 À74 À50 À76 À80 À70 À70 À66 À64 À54 À52 À40 À44 À60 z 16 10 48 62 30 18 22 6 56 46 6 0 À6 À2 28 28 18 À40 À26 À26 À40 À36 À28 À26 À30 À28 3. P b 0.13 Z score 493 Volume (mm3) 64 64 416 (L) (L) Note . / NeuroImage 23 (2004) 483–499 Table 5 Regions showing greater activation for the situation strategy group when increasing emotion Region of activation Brodmann Coordinates x Situation (increase–look) N self (increase–look) Precentral gyrus Middle temporal gyrus Middle temporal gyrus Middle temporal gyrus Superior occipital gyrus R3/4 L21 R19 R19 R19 14 À64 48 54 40 y À38 À62 À86 À76 À86 z 72 2 16 16 22 4.05.26 3.31 3. 15–16 s TR t (22) = 1.N.36 3.46 3.33 3.90 3. 11–12.83 3.75 3.36 4.47 3.50.34 80 64 Note .40 3.78 3.62 3. one-tailed t (22) N 2.97 4.06]. Ochsner et al.76 3. P b 0. P b 0. and 13–14 s TRs. Coordinates are in MNI space.15.54 3. P b 0. look trials during the photo presentation period in which participants down-regulated negative emotion [for 9–10 and 11– 12 s TRs. one-tailed t (22) N 2. and (2) that only for the self-focused group was right medial frontal signal significantly greater on decrease than on look trials [for 9–10.02 3. Clusters of five or more contiguous voxels whose global maxima meet a t threshold of 3.98 3. are reported.60. P b 0.37 3. are reported. .48 3.35 3.57 3. Local maxima for these clusters are denoted with (L).32 Z score Volume (mm3) 216 (L) 128 80 168 360 168 328 264 200 344 80 424 136 120 472 (L) 208 864 (L) 88 112 296 (L) 128 128 6 À48 30 À66 À12 50 3.32 3.45 3. neither group showed significantly greater activation relative to the other.05]. Coordinates are in MNI space.50.58 3. For the direct contrasts between increase and decrease trials that identified regions selectively involved in each type of reappraisal.44 3.55 3.001 uncorrected. Clusters of five or more contiguous voxels whose global maxima meet a t threshold of 3.69 3.42 3.01 3. Local maxima for these clusters are denoted with (L).K.93.001 uncorrected.

and self-focused reappraisal strategies suggests that these two strategies differ in their reliance on externally as compared to internally focused processes. Up. Smith and Jonides. 2000). By contrast. Self. Although the precise functional significance of lateralized amygdala activations (in general) is not clear. 1997). Ochsner et al.N. and the induction of sad moods through verbal scripts (Schneider et al. (2) activated regions of the dorsal anterior cingulate implicated in the on-line monitoring of performance (Botvinick et al. 2001). the acquisition of classically . Inverted time courses for this medial prefrontal region are commonly observed (Gusnard and Raichle. Miller and Cohen.. these data suggest that common modulation of the left amygdala by the up. 1999) that may support the generation and maintenance of reappraisal strategies. This is the first imaging study to examine both the cognitive and up.. (B) Regions more active for the interaction contrast self(decrease–look) – situation(decrease–look). Strange et al. Time courses show greater recruitment of this left lateral prefrontal region when decreasing negative emotion. 2001.versus situation-focused reappraisal. Ochsner and Gross. rather than enhancing. Ochsner and Feldmann Barrett.and down-regulation of emotion may reflect the use of verbal strategies that can influence affective encoding in the amygdala. Relative recruitment of lateral as compared to medial prefrontal regions by the situation. down-regulation decreased activation bilaterally. Down-regulation was felt to be more difficult. which allowed for analysis of the neural systems involved specifically in self. (3) activated regions of dorsal medial prefrontal cortex implicated in self-monitoring and self-evaluation of emotion (Gusnard et al. modulation of the right amygdala during down-regulation may reflect the modulation of processes related to encoding the affective or arousing properties of nonverbal stimuli. but only for participants in the situation-focused group. Time courses show greater recruitment of this right medial prefrontal region when decreasing negative emotion. 2002.. Left panels show regions more engaged for the situation or self group when decreasing emotion. Both up. Simpson et al. 2002).. selectively decreasing or increasing its activation. the lateralization of these activations differed in the two conditions: whereas up-regulation increased activation in the left amygdala. albeit more significantly of the left side. 2001).and down-regulation of negative emotion involved both neural systems that were similarly activated and other systems that were selectively activated for one or the other kind of regulation. 2001). 2001. which probably reflects the greater challenge in reversing. which allows for consideration of the common and distinct neural systems involved in each kind of emotion regulation. / NeuroImage 23 (2004) 483–499 Fig. as evidenced by right amygdala activation to film clips (Beauregard et al. the initial emotional response to an aversive scene. Taken together. 2004. Hamann and Mao. Regions more active when using either a situation. 1997. Lane et al. Although amygdala activation was modulated for both up. there are indications that the left amygdala may be associated with processing of and memory for affectively charged verbal stimuli (Buchanan et al. 1999. (A) Regions more active for the interaction contrast situation(decrease–look) – self(decrease–look). 5.and down-regulating emotion (1) activated regions of left lateral prefrontal cortex implicated in working memory and cognitive control (Knight et al.and situation-focused reappraisal strategies were equally effective. This is also the first imaging study to systematically vary reappraisal strategy. 2001... respectively.. the mental representation of an event that is eliciting fear (Phelps et al. and (4) modulated activation of the left amygdala..or self-focused strategy to decrease emotion.and down-regulation. but only for participants in the self-focused group. 2001). Center and right panels show percent signal change activation time courses for representative peak voxels from regions highlighted by the red circle in left panels. emotion when intentionally down-regulating emotional responses and significantly increased negative emotion when intentionally up-regulating emotional responses. 2001.versus down-regulation of emotion Up..and down-regulation of emotion.494 K..

Future studies employing a shorter TR and/ or combined electrophysiological recording may serve to address this issue. Ochsner et al. These findings replicate and extend our initial findings concerning the use of reappraisal to down-regulate negative emotion (Ochsner et al. To the extent that up. Ochsner et al. which reinterpreted the actions and outcomes for a given image. and presentation of emotional facial expressions (Anderson et al.. Taken together. and common activation of many of the prefrontal and amygdala systems described above.versus situation-focused reappraisal Self-focused reappraisals. The limited temporal resolution of fMRI also prevents a precise evaluation of the temporal dynamics of amygdala modulation.and situation-focused reappraisal strategies depend upon neural systems generally involved in internally focused as compared to externally focused processing.. as revealed by greater activation in an attend baseline condition than in a reappraisal condition in which participants decreased negative emotion. 2001. when the height threshold for the decrease– attend contrast from our prior study was lowered to 0. Consistent with this argument. which is consistent with the idea that reappraisal can cognitively reverse the aversive connotations of a stimulus. 2000). both of which are required when attempting to supplant negative stimulus appraisals with neutralizing top–down reappraisals. 1999).. 2004). 2004). In that study.. Taken together. 2003). or shocking. The reliance of down-regulation on right lateral PFC could reflect the broad role of this region in behavioral inhibition and interference resolution (Bunge et al. The present study may have had substantially greater power to detect reappraisal-related activations and hence may have detected bilateral and orbital PFC involvement in the cognitive downregulation of emotion. 2000. self. Rolls.. On the other hand. in press). and (3) more rigorous and extensive prescan training in reappraisal.g.. participants may have continuously monitored the self-relevance of aversive scenes to ensure that they were remaining distant from them. This medial PFC region.and down-regulation also selectively recruited distinct regions of left rostral medial prefrontal and posterior cingulate cortex and right lateral or orbital prefrontal cortex.. Altering the affective value of a stimulus may depend critically upon representation of an event in orbitofrontal cortex and amygdala (Rolls. This difference may be due to the fact that in the present study. but it is possible that faster-acting (and perhaps automatic) amygdala processes not detectable with this study’s 2-s temporal resolution remained unaffected by the reappraisal process. In this way.. Preston et al. and decreases in negative affect when down-regulating emotion.. Sources of greater power may include (1) the use of 23 as compared to 15 participants. which participants did when thinking about aversive scenes as even more affecting. sad. In the initial study but not in the present one.. 2002). 2001). When adopting a detached and distanced perspective..and down-regulation recruit processes specifically associated with elaborating the affective properties of the stimulus as compared to mediating interference between competing appraisals and reappraisals. which may have led to greater medial orbital frontal activation relative to the present study. which modified the personal relevance for a given image.and down-regulation was proportionally greater than that uniquely associated with each type of regulatory goal may reflect the degree of overlap in underlying processes—that is.001 threshold. we observed only left PFC involvement and no lateral orbitofrontal activation when decreasing negative emotion and modulating amygdala activation. Reversal of stimulus-reward mappings activates lateral orbitofrontal cortex (O’Doherty et al. Kelley et al. Self-focused reappraisal differentially recruited medial PFC when decreasing emotion. respectively. horrific... we asked participants to look at images and respond naturally... This study revealed amygdala modulations starting as early as the first 2-s epoch of reappraisal. which . These two strategies did differ in their relative reliance on medial as compared to lateral prefrontal cortex.K. The fact that prefrontal activation reflecting overlap of processes related to up. but only when down-regulating emotion. cognitively reframing an affective event may draw most heavily upon strategic verbal and visuospatial processes used to construct reappraisal narratives. A similar pattern of differential medial versus lateral prefrontal recruitment has been observed when participants either judged the valence of their own emotional response to a photo or judged the valence of the emotion expressed by the central figure depicted in those photos (Ochsner et al. 1997). Self. situation-focused reappraisal differentially recruited regions of lateral PFC generally implicated in the maintenance and manipulation of information about stimuli in the external world (D’Esposito et al. regardless of the reappraisal goal. whereas in the initial study. Rolls. Selective attention can enhance processing (Culham and Kanwisher.. The reliance of down-regulation on right orbitofrontal cortex could reflect the role of this region in altering and updating the contextsensitive motivational relevance of stimuli (Bechara et al. The reliance of up-regulation on left rostral medial and posterior cingulate cortices could reflect the role of these regions in generating words that describe emotional events (Crosson et al. (2) the use of a spiral in–out as compared to spiral out pulse sequence that is more sensitive to detecting activation in prefrontal cortex and other regions (Glover and Law.05. such as the amygdala. 2000). corresponding to BA 32. 1997.. Up. 2001. as suggested by significant correlations between increased activation of right orbitofrontal cortex. has been associated with self-referential judgments (e. a bilateral pattern of activation is observed that very closely resembles the pattern observed in the present decrease–look contrast at a 0. The absence of differential medial or lateral PFC activation during up-regulation may be the result of task instructions. decreased bilateral amygdala activation. 2000). 2000). 2003) and is impaired by lateral OFC lesions (Dias et al. One other difference between the results of the two studies may have been attributable to a change in instructions in the baseline condition. respectively (Christoff et al. however. Jonides et al. 2001. emotion knowledge may be retrieved to add cognitive insult to the physical injuries depicted in photos.N. and more generally with processes thought to reflect a default self-monitoring state of brain activation (Gusnard et al. 2001). these findings support the hypothesis that beyond their common reliance on a core network of reappraisalrelated systems. past and present findings are consistent with the hypothesis that reappraisal involves interactions between prefrontal systems that implement cognitive control processes and systems that appraise the affective properties of stimuli. / NeuroImage 23 (2004) 483–499 495 conditioned responses (Furmark et al. we specifically instructed participants to attend to their feelings. had two notable similarities: similar success in modulating emotion. additional prefrontal control systems are recruited. 2002). and situation-focused reappraisal. 1998). we observed modulation of medial orbitofrontal cortex by reappraisal.

and cerebellar regions activated when increasing or decreasing emotion. but not close-other. Canli et al. Mayberg et al. Some studies show amygdala modulation (Hariri et al. 2003). The present results challenge these views.g. This may have diminished reliance on self-reflective processing. A second question concerns the active role that prefrontal control processes play in modulating amygdala activation. This explanation is unlikely. and temporal cortices may reflect attentional selection of cognitively reorganized perceptual inputs held in working memory during reappraisal (Cabeza and Nyberg... Modulation of thalamus and various portions of the basal ganglia may reflect the recruitment of cognitive control circuitry that links prefrontal regions to subcortical structures in functional loops (Alexander et al. but simply because one has engaged in effortful cognitive processing that either draws resources away from. Liberzon et al. because none can account for the fact that the effortful cognitive processing engaged when one is attempting to upregulate negative emotion not only makes one feel worse. it will be important to relate the mechanisms underlying reappraisal to the mechanisms underlying other ways in which cognition can regulate emotion. Critchley et al. Culham and Kanwisher.. This explanation also appears unlikely.. however. Indeed.. 2001. Participants in the self-focused reappraisal group were informed that they could increase emotion by imagining either themselves or a loved one involved taking part in.. 2000). 2000. 2000. This is not to say that various types and applications of attention. 2003). 1999. PFC (e. 2002). and greater right PFC activation was observed when emotion decreases.and down-regulation of emotion.. whereas others suggest that the relationship is more specific.and down-regulating emotions reflect the recruitment of cognitive processes specifically related to each regulatory goal: Up-regulating negative emotion may recruit left PFC systems used to self-generate affective descriptors that intensify emotion. Davidson. however. Levesque et al. 1998. Lane et al.. as suggested by prior studies showing medial PFC activation only for self. 1999)... clarifying these relationships currently is difficult because results have been mixed from studies examining modulation of amygdala activation to fearful or threatening faces or images under conditions of full as compared to divided attention.. 2000. A third question concerns the importance to reappraisal of other cortical. pictured events. 2002). the most likely explanation for the present results is that the active mental transformation of the meaning of an event is responsible for up. One possibility is that this difference is due to down-regulation being more effortful and therefore recruiting bilateral as compared to unilateral prefrontal regions. the downregulation of emotion via reappraisal occurs not because one has actively transformed the meaning of a stimulus. if not necessarily transform. Given that the specific .. Pessoa et al. 1998. Although the precise reasons for these discrepancies are not clear. 2000.g. cognitive load. which is exactly the opposite of what would be expected based on prior work associating positive emotion with left. 2001. in the long term. which should not have been observed if effort alone was driving observed prefrontal activations A second possibility is that the observed findings reflect differences in the amount of negative affect experienced when increasing as compared to decreasing emotion. One question is why down-regulation recruited both left and right PFC. Implications and future directions The present findings have important implications for the design of neuroimaging studies of emotion.. and elaboration of. 2002. 1996. for example) very much like the processes recruited by situationfocused reappraisal. Modulation of inferior parietal. is that differential PFC activation when up. because multiple regions were more active when increasing than when decreasing. or evaluative judgments might not modulate emotion processing (cf. Winston et al. 2003). Nature of reappraisal-related dynamics Although the present study provides some initial answers to questions concerning the nature of neural mechanisms underlying the up. one possibility is that instructions in some cases may induce participants to reappraise the meaning of stimuli. Ochsner et al.. Furthermore.. because greater left PFC activation was observed as negative emotion increases. 2001. 2003. Modulation of cerebellum may reflect recruitment of circuitry important for the accurate selection and timing of linguistic and cognitive operations supporting the narrative reframes used when reappraising (Fiez. subcortical. 2000) and is consistent with numerous studies showing cerebellum activation during emotion (e. it is possible that imagining increased self-involvement with complex events recruits lateral prefrontal systems mediating attention to. it raises a number of questions as well.. or associated with. Reiman et al. A similar view is that reappraisal is simply a form of distraction and disrupts emotion processing in much the same way that any secondary task disrupts a primary task (McRae et al.. / NeuroImage 23 (2004) 483–499 unexpectedly may have diminished differences between the selfand situation-focused strategies. Vuilleumier et al.. however. occipital. Various researchers have proposed a reciprocal relationship between cognitive and emotional processes. Paradiso et al. 2003). This finding should not have been observed if emotion processing is inhibited simply by engaging in any type of cognitive processing or is disrupted only when labeling emotional states and stimuli.. whereas up-regulation recruited primarily left-lateralized systems. a conclusion that is supported by the finding of prefrontal– amygdala interactions underlying other recent studies examining reappraisal (Beauregard et al. related judgments (Wyland et al. Prabhakaran et al. Schaefer et al. and one that we advanced above.or down-regulating amygdala activity. Phan et al. Some have suggested that deactivation of emotion systems is an indirect byproduct of most any type of cognitive processing (Drevets and Raichle. 1997.. In our view. or is wired up to disrupt (Lieberman. 1997). A third possibility. 1999). occurring only when linguistic processes are used to label affective stimuli (Lieberman. 2003) and others show the amygdala response is invariant with respect to these manipulations (Anderson et al. 2002). 2003) or the use of cognitive control to maintain. On these views. however. emotion processing.. aspects of those events (as when imagining what it would be like to be a sick person lying in a hospital bed. and negative emotion with right. but also increases amygdala activity.. which our participants did when upregulating emotion.496 K.N. Smith and Jonides. 1986).. an emotional response (Schaefer et al. Unfortunately. whereas down-regulating negative emotion may recruit right PFC systems used to mediate interference between competing prepotent affective responses and cognitively controlled reappraisals. or when directly as compared to indirectly evaluating the affective properties of these stimuli.

T. and the ability to control it.J. J. films.. Botvinick. Fiske. 1998. Pers.. Development. Spec. Barch. 21 (18). Neurosci. E. (Eds.. 2001. DeLong.M.. Fallon. M.. 133 (1). Drevets.C. Neurophysiol. Reson. 680 – 740. T. Hum. L.. there is no way to know whether prefrontal and amygdala activations vary because participants spontaneously appraised or reappraised stimuli in different ways (Ochsner and Feldmann Barrett.. J. Damasio.J. Behav. Cato.J.S. . A. A. G.K. E. Gabrieli.R. Magn. A... Soc.. 2000. Neural bases of learning and memory: functional neuroimaging evidence. Explicit and implicit neural mechanisms for processing of social information from facial expressions: a functional magnetic resonance imaging study.. 1196 – 1214. R. Culham.. T.T. Glover. aging. 46 (3). Neuroimaging of cognitive functions in human parietal cortex. 353 – 385. J. C.. Brain 124 (10). M. S. Strick.. T..L. 64 (6). In: Gilbert. studies should systematically control the way in which stimuli are appraised and reappraised in order to draw inferences about the relationship of prefrontal and amygdala activation to emotion. K.L. J. Cereb. The emerging field of emotion regulation: an integrative review. 3 – 11. U. 13 – 15.. K. 11 (2). 9... D.. Science 289 (5479). 2001. 624 – 652.K. C.E. D. Rev. Levesque. J. Christoff...D. P. NY. Gokcay.. Boniface... 2000.G. Baumeister. Psychol.. 2001. 10789 – 10794.E. Sadek. Curr. K.D.. 1 – 9..W. Putnam. For studies that allow participants to freely experience photos. J. B.. Desmond. J. M... Sex differences in the neural basis of emotional memories. Neurobiol. H. and resilience: brain mechanisms and plasticity. R.D. Dias. Radonovich. Cortex 10 (3). K. Davidson. Sex-related difference in amygdala activity during emotionally influenced memory storage. Bechara. Briggs.D. De Rosa. New York. David. Law. Rypma. Glover...R.H. L.. Desmond. Canli. Gabrieli.F.M. R. 1993. 1997.D. Neural correlates of the automatic processing of threat facial signals. Beauregard. D. vol. et al.K. Gen. D. Gross.N... 357 – 381.H. 2001. Daly.G. / NeuroImage 23 (2004) 483–499 497 prefrontal systems engaged by reappraisal. Proc.W. Cerebellar contributions to cognition. What aspect of the MRI BOLD signal best represents the underlying electric physiology in humans somatosensory cortex? Clin. Auerbach.. A. Left-hemisphere processing of emotional connotation during word generation. Van Amelsvoort. Arturs. Anderson.. 6 (2).. Prefrontal regions involved in keeping information in and out of mind.S...S. 23 (13). Robertson. Murphy.J.. Gabrieli. R. Affect. and social consequences.N. and Karen Milch for assistance with data collection and analysis... The amygdala and individual differences in human fear conditioning. 2000.. Curr. Ochsner et al. 8 (6).. Psychol. Neuron 16. Haier. 2001. A. 305 – 309. Psychol. Emotional suppression: physiology. S. Magn. Bourgouin. and the presence of depression and anxiety thus may influence the way in which emotion systems generate responses and the efficacy with which reappraisal systems modulate them (Davidson. Christoff.T. Sci. Dissociated neural representations of intensity and valence in human olfaction. and the nature of amygdala modulations reappraisal causes. Learn. National Science Foundation grant BCS-93679.. 2003. Levenson. C. J. Robbins. Exp. 1998.. Gabrieli.. in press. Potkin. Bauer.S. 5627 – 5633. Acad. Anderson. Rev. Neurosci. J. 114. Buchanan. D’Esposito. 1203 – 1209. 42. 12 (3)..J.. Dysfunction in the neural circuitry of emotion regulation—A possible prelude to violence... 1998. Emotion regulation: affective. Z. self-report. A.E. 415 – 421.. 295 – 307. 99 (6).A.M. Mem. R.. Ream. Spiral-in/out BOLD fMRI for increased SNR and reduced susceptibility artifacts. Handb. C. The present findings also have implications for the study of interindividual variability in emotion and emotion regulatory capacity. 515 – 522. Lawrence. Neurol. Neurosci. Phelps. L. NeuroReport 10 (12).. T. Med. G. J. Damasio.R. Z. Fischler. E. 271 – 299.. decision making and the orbitofrontal cortex. Soc. Mem.W.E..J. Evaluating self-generated information: anterior prefrontal contributions to human cognition. NeuroReport 9 (14)... S. B. E. J. leading either to deficits in the ability to make one’s self feel better by down-regulating negative emotion or to improved ability to make one’s self feel worse by upregulating negative emotion (for a discussion. Reson. Hemispheric asymmetry for emotional stimuli detected with fMRI. 1999. Learn. L. 157 – 163. Cohen. Affective style. et al. M. Phillips.L. Alkire.. Davidson. a NARSAD Young Investigator grant to the first author. Neurobiol. Bullmore. Denburg. Opin. Gross.. Ochsner. Brain Res. 3D Z-shim method for reduction of susceptibility effects in BOLD fMRI. J... 93 – 105.J.D.. 2. 1986.S. Psychol. Furmark... that systematically influence the specific reappraisalrelated systems identified here. 290 – 299. 2002). Geddes. 2001. Glover.. 2004).....M. J. J. Parallel organization of functionally segregated circuits linking basal ganglia and cortex. White.. J. S. A. 2001. T. Cabeza. M.. 2000. Acknowledgments The authors thank Janet Eckart. Raichle. 2074 – 2086. G... M. J. Gabrieli... N. J. G. J. Nat..T. 55 (11).. R...K. Browd..J.C. McGraw-Hill. M. 2004). Zhao. Anxiety Disord. Psychol. psychopathology. The self. Cogn. 2.. Nyberg. 326 – 335.M. 2003.. G. Neurosci. 2000. 108 (3). 196 – 202. 9285 – 9297. 2449 – 2455. Kanwisher. Annu.R. Issue: Neuropsychol. 3957 – 3960.A.W. 1999. O. Kirkpatrick.). H. 2000. Emotion..A. D. Critchley.... Williams. Future studies may identify individual differences in the tendency to experience negative emotion. et al. 1998.A. A. Neural correlates of conscious self-regulation of emotion. Rev. 591 – 594. Natl. 2003. S. see Ochsner and Gross...E. Cahill. T. N. Braver. S. R. and expressive behavior. 2000. 17 (23). R.. Emot. K. 1997. R. Desmond. Fischer. Gross. 281 – 291. D. Anderson.R. Panitz. 9 (2). Nature 411 (6835).. Conflict monitoring and cognitive control. J. K. Opin..P. and National Institute of Health grants MH58147 and MH066957 for support of this research. Reciprocal suppression of regional cerebral blood flow during emotional versus higher cognitive processes: implications for interactions between emotion and cognition. J.. J. Fiez... both may vary systematically as a function of goal and strategy. cognitive. Lesions of the human amygdala impair enhanced perception of emotionally salient events.C. J. Neurosci. M. Glover. 1996. M. Med. Verbal and nonverbal emotional memory following unilateral amygdala damage. Zhao. 2001. Am. NeuroReport 8 (18).J. Brammer. S. L. 75. or memories. 2002. Neurosci. Larson. 3233 – 3239. References Alexander. I. Maron.. pp. 2002. E. Prefrontal cortical contributions to working memory: evidence from event-related fMRI studies. Carter... 13 (4).. Dissociable forms of inhibitory control within prefrontal cortex with an analog of the Wisconsin Card Sort Test: restriction to novel situations and independence from on-line processing..E. Ambili Ramachandran. P... N.. Roberts. Christoff. Perspect. Canli. H. Psychophysiology 39 (3).T.. B. W. 970 – 986. Postle. Bunge.. Mather et al. Crosson. RC165. Brain Mapp..

. Feldmann Barrett. van der Linden. A. V.E. H. McKenna. M... 1618 – 1629. K.. D.. Lane. Lane. J.. Nat... Proceedings of the 31st Annual meeting of the Society for Neuroscience. R. Acta Psychol. Comparison of Spiral-In/Out and Spiral-Out BOLD fMRI at 1. S. Silva.. Ludlow... Am. J. Schneider.M. Neuropsychologia 39 (3).O..T. Cogn. J.M. Neurosci. 2000. J. Cosgrove.. J.. 2.. Neural correlates of hot and cold emotional processing: a multilevel approach to the functional anatomy of emotion. 2004. S. 14 (6). 24. Cooper. In: Forgas.. M. S. G..B.. Psychol.G. J. Annu. A. P.L. Rev. Lieberman. 3 (1).... 2002. Neuroanatomical correlates of externally and internally generated human emotion.. Cogn. G. Degueldre.. Searching for a baseline: functional imaging and the resting human brain..R. LeDoux. Chao. Chen. Williams. Prefrontal cortex regulates inhibition and excitation in distributed neural networks. G.. Ochsner. K. Reiman. Bunge.. J..J..L. W. R.. C... W..N. K. Conscious and unconscious emotional learning in the human amygdala (see comments).C. Bourgouin. G. Luxen. 98 (7). pp. Gutierrez. J. Neurosci. Caglar. Ahern. Pessoa. P.....J. Interaction with cognitive task. Lancaster. T.).. T.M. W. Decker.. L. Psychiatry 154 (7)... Paquette. Modulation of amygdalar activity by the conscious regulation of negative emotion.E. Martin. Ohman. Psychiatry 53 (6). Emotion regulation and memory: the cognitive costs of keeping one’s cool. R.. S.. 15. 7931 – 7939. A multiprocess perspective on the neuroscience of emotion. Mahurin.C.L.D.. Sci.. O’Connor. 2001.A. D..R. Liotti.. In: Mayne. Dolan. D.A. S.S. Psychol.R. Narayanan. H. 291–301. NeuroReport 8 (18)... 2001. Reuter-Lorenz.. Hippel.A. The orbitofrontal cortex and reward. 167 – 202.. Salmon... H..K. 11458 – 11463. 3969 – 3972. Segerstrom.N.C. Gross. Nat.E. Hanelin. E. NeuroReport 13 (1).J. Sci. 211 – 215. Cortex 10 (3).L. Schwartz. Thinking makes it so: a social cognitive neuroscience approach to emotion regulation..L. Bonanno. Modulating emotional responses: effects of a neocortical network on the limbic system. Orlando. Monarch. Biol. M. Neurosci. Knierim.)...... D.. K. Laureys. Phelps. Hichwa. Kimberg. et al.... Gabrieli. 159 – 178. Friston. Staines. S. 2000. Richards. 8410 – 8413..D. E.C. An integrative theory of prefrontal cortex function.S. Tsao.. Mather. L. NeuroReport 11 (1).. NeuroImage 21. Ochsner. Phan.498 K.. R. Limbic activation and psychophysiologic responses to aversive visual stimuli.. Reciprocal limbic-cortical function and negative mood: converging PET findings in depression and normal sadness..E. Cereb. E. S. A. Hamm.. Ramachandran. Social Judgments: Explicit and Implicit Processes.C. A.. K. and neutral visual stimuli in a PET study of normal subjects.D. K.. 82 – 92. Macrae. Inhibition in verbal working memory revealed by brain activation.K.A. K.... Craske. NY. J.. E.. Davidson.T. K. J. O’Leary. Proc. Price.K. Ponto. (Eds. pp.L. W.. Liberzon. Kosslyn.. 284 – 294. 2001. J. K. Fink. Deficits in visual cognition and attention following bilateral anterior cingulotomy. E.. Goldstein. Bradley.v. Nierenberg.. Gabrieli. Am. In: Vohs. 155 – 184. Jerabek. 62–83. J. R. 918 – 925..C. Nature 393 (6684). J.H. Memory enhancement for emotional stimuli is impaired in early Alzheimer’s disease. Kelley. L. Koeppe. Acad. Natl. 14 (5). Psychiatry 156 (5).. 785 – 794.. 2001. Beaudoin. English. Proceedings of the 31st Annual Meeting of the Society for Neuroscience. Tekell. 1999. Rethinking feelings: an FMRI study of the cognitive regulation of emotion. B..J. Mazziotta. J.. J.. 1999. 259–263. NJ. Gross. A.R.. Morris.. 2003. G. Davidson. E. Nichols... Gabrieli. R.K.C. Psychiatry 156 (10).A.K. Natl.N.G. Neurosci. 508 – 516. Philippot. 2000. Raichle.E. pp. 1993. T.. Wyland. Neural circuitry underlying voluntary suppression of sadness.. 44 – 67. 671 – 688. K. Taylor. M. C. 261 – 273.L. R.. Amygdala responses to emotionally valenced stimuli in older and younger adults. S.. J.. Joanette. Minoshima.S. Ungerleider. Proc. G. Ochsner.. 938 – 949. Eugene. Miller..T. Positive and negative emotional verbal stimuli elicit activity in the left amygdala. Psychol. 2000. Annu. 2003. Andreasen. L. Knight. Rev. R.. Common and distinct neural systems underlying the perception of emotion in self and other.A. K.C.. G. 2000. 24 (6). Neurosci. Integration of diverse information in working memory within the frontal lobe. S. 913 – 921. facial.D.N.. D. Zhao. O’Doherty.. R. S. / NeuroImage 23 (2004) 483–499 Ochsner. (Eds.. S. C.L. D. New York. Ochsner. 2003.C. The Handbook of Self-Regulation.. Schaefer.. 2002. K... Grodd. P. and behavioral reactions. Cogn.R. P. A..C. G. Emotion circuits in the brain.N. Paradiso.E. Rauch.L. Neural activation during selective attention to subjective emotional responses. Leroux.Z.S.. Gross. J.... J. 1997. Erlbaum. Jonides. Neuropsychology 14 (1). Ochsner. L. Biol. 502 – 510. 1997.S. K.A. C. 38 – 81. Watkins. Heatherton.. Worry and rumination: repetitive thought as a concomitant and predictor of negative mood. 2000.A.D. J. J.D. Welsh. L. J. 85 – 90. A. J. 2003. G. Mackey. R.. Neurosci. 14 (8). S.. 467 – 470. Levesque. K.. Psychophysiology 30 (3).H. R. Acad.R. Nat. Akbudak. 504 – 511. B.. Reflective and reflexive judgment processes: a social cognitive neuroscience approach.. Noll. Cartensen.N. Shulman.C.N. S.. Raichle. Ochsner. . J. 4 (4). Neurosci. R. J.. 99 (17). Ochsner..C.F. C. Swick. Decker.. Prabhakaran. Medial prefrontal cortex and self-referential mental activity: relation to a default mode of brain function.I.J. Koeppe.H. U. Cohen. Cogn. A. Gusnard. Proc. Acad.. Gusnard. Emotion-induced changes in human medial prefrontal Gusnard.. Neurosci. Thompson-Schill. McRae. Raichle. Inati. J. 437 – 441.. 95 (14). J.Y. 15 – 19... Cambridge Univ. P. Cerebral blood flow changes associated with attribution of emotional valence to pleasant. 219 – 230.. Sci. Psychiatry 53 (3).E. visceral.F. Ochsner et al. Sci. The Guilford Press.M. R. Fig.J.. M.N. Hariri. Neurosci. M. Finding the self? An event-related fMRI study.. M. Chau. Mensour. The role of rumination in depressive disorders and mixed anxiety/depressive symptoms. Liberzon.. Canli.. Drevets. F. Greenwald.. Abnorm. G.. 2002.E. Neuropsychopharmacology 23 (5).M...). P. F.. J... Whitfield. 2003. 675 – 682. S.E... Bookheimer. Marshuetz. S. Dissociating valence of outcome from behavioral control in human orbital and ventral prefrontal cortices. L. P..M. S.. Fox... 76 (2–3). L.J. E. J. 1998.) 101 (2-3). J.R. 2001...N. E.. Dolan. Glover. L. 2002. Wais. J. E. 2002. 1215 – 1229.. A.. F. Activation of the medial prefrontal cortex and extended amygdala by individual ratings of emotional arousal: a fMRI study. Gross.. 79 (3). F.. M. Schaefer. E. Thomason. Smith. Alden. Pers. G.A. Am. T. I. D.. L. Rev.. M. K. Res. W..5T and 3T. S. M. S. Hamann. A. Natl. 410 – 424. S...D.. Yun. Aguirre. (Amst. M.R. M.. 43 – 48.R. 2000. Emotions: Currrent Issues and Future Directions. Activation of the left amygdala to a cognitive representation of fear. Johnson. J. N. Neural processing of emotional faces requires attention.J. U. J. U. 2002. 23. Deichmann.. S. unpleasant. J.L.L...J. Functional MRI reveals left amygdala activation during emotion. 2002.. 2000. I. Maquet. Britton.A.. P. K. Lang.. A. 1999. S. Critchley. S.. Looking at pictures: affective. 2004. Exploring the neural bases of emotion regulation: comparing the effects of cognitive reappraisal and working memory load.. 1997. Taylor.... Jackson. L. K. Psychiatry Res. Soc. Delfiore. 1998. et al. Ther. Rolls. Mayberg.F. Simpson Jr. D. D.E.J.P.A. T.L... M. Nolen-Hoeksema. et al. D.D. NeuroImage 18 (4). 2004. 23. 4259 – 4264. T. Mao. Baumeister.. Brannan... Snyder. R. McGinnis. Y.M. 2004.. 685 – 694. V.M. S. 109 (3).. Preston. J. 2000. P.. J.M.. Gabrieli.A. (Eds.. 75 – 82. Cassem.. New York. Collette..J. Beauregard. Hamann. Press.J.. 2001.. FL. Z.. S.Y.

T. Proc. R.S. Vuilleumier... Driver. semantic. S..B.. R.J. 1998. Science 283 (5408). E.A. Is your best friend more like you or the President? Proceedings of the 14th Annual Meeting of the American Psychological Society Atlanta. Neurosci.N. Nat. Automatic and intentional brain responses during evaluation of trustworthiness of faces.N. Henson. 425 – 433. Brain mechanisms for detecting perceptual.. S.C.. McInerney...K. Etcoff.. J.A.. O’Doherty. Armony. N.. 18 (1)... B. Heatherton. J. N. 2003. and emotional deviance. Winston. R. Jonides. Lee.J. 1999. M. . Kelley. 2001. J.. M. U. Wyland.. Storage and executive processes in the frontal lobes. 2002. W. Effects of attention and emotion on face processing in the human brain: an eventrelated fMRI study. 5 (3). / NeuroImage 23 (2004) 483–499 cortex: II.. McRae. 1657 – 1661. Masked presentations of emotional facial expressions modulate amygdala activity without explicit knowledge.. et al. B. J. P. Neuron 30 (3). Georgia. 277 – 283. J. During anticipatory anxiety. 499 Whalen. 411 – 418.E. P. 829 – 841. Strange. Dolan.A. Rauch.L.L. Acad.. J.. Ochsner et al. C. Dolan. A..L... Sci. Strange. S. 688 – 693.. NeuroImage 12 (4). 2000.J. Jenike. Natl. Smith. Neurosci. 98 (2).

Sign up to vote on this title
UsefulNot useful