Neural Correlates of Emotional Stroop in Women with Abuse-Related PTSD PDF

Neural Correlates of the Classic Color and Emotional Stroop in Women with Abuse-Related Posttraumatic Stress Disorder J. Douglas Bremner, Eric Vermetten, Meena Vythilingam, Nadeem Afzal, Christian Schmahl, Bernet Elzinga, and Dennis S. Charney Background: The anterior cingulate and medial prefrontal...

Neural Correlates of the Classic Color and Emotional Stroop in Women with Abuse-Related Posttraumatic Stress Disorder J. Douglas Bremner, Eric Vermetten, Meena Vythilingam, Nadeem Afzal, Christian Schmahl, Bernet Elzinga, and Dennis S. Charney Background: The anterior cingulate and medial prefrontal cortex play an important role in the inhibition of responses, as measured by the Stroop task, as well as in emotional regulation. Dysfunction of the anterior cingulate/medial prefrontal cortex has been implicated in posttraumatic stress disorder (PTSD). The purpose of this study was to use the Stroop task as a probe of anterior cingulate function in PTSD. Methods: Women with early childhood sexual abuse-related PTSD (n ⫽ 12) and women with abuse but without PTSD (n ⫽ 9) underwent positron emission tomographic measurement of cerebral blood flow during exposure to control, color Stroop, and emotional Stroop conditions. Results: Women with abuse with PTSD (but not abused non-PTSD women) had a relative decrease in anterior cingulate blood flow during exposure to the emotional (but not color) classic Stroop task. During the color Stroop there were also relatively greater increases in blood flow in non-PTSD compared with PTSD women in right visual association cortex, cuneus, and right inferior parietal lobule. Conclusions: These findings add further evidence for dysfunction of a network of brain regions, including anterior cingulate and visual and parietal cortex, in abuse-related PTSD. Key Words: Posttraumatic stress disorder, Stroop, emotion, fear, orbitofrontal cortex, anterior cingulate (Brodmann area [BA] amygdala, hippocampus, anxiety 25—subcallosal gyrus, and BA 32), and anterior prefrontal cortex (BA 9) (Devinsky et al 1995; Vogt et al 1992). In the famous case of Phineas Gage, a physical injury resulted in damage to this part P osttraumatic stress disorder (PTSD) is an important cause of morbidity that affects 10% of women in this country at of the brain that was associated with deficits in emotional some time in their lives and is twice as common in women regulation, specifically problems inhibiting behavior, and social as in men (Kessler et al 1995). Childhood sexual abuse, which inappropriateness (Damasio et al 1994). Lesions of the medial affects 16% of women sometime before their 18th birthday prefrontal cortex in animals resulted in a failure of inhibition of (McCauley et al 1997), is the most common cause of PTSD in amygdala function (Morgan and LeDoux 1995; Morgan et al women (Kessler et al 1995). Little is known, however, about the 1993). Given the role of the amygdala in fear responding (Davis long-term biological consequences of childhood abuse. 1992; LeDoux 1993), these findings led to the hypothesis that Exposure to uncontrollable stress has been used as an animal medial prefrontal cortex mediates extinction of fear responses, model for mood and anxiety disorders (Charney et al 2000). A and by extension dysfunction in this region might underlie the variety of models of stress have been used, including exposure to pathologic emotions associated with abuse-related PTSD. electric shock, forced swim, and maternal deprivation. These Neuroimaging studies in PTSD are consistent with dysfunc- stressors have in common a physical threat to life and have tion of medial prefrontal cortex/anterior cingulate. Symptom parallels to traumatic stressors, defined by DSM-IV as a threat to provocation studies with personalized traumatic scripts, trau- self or other associated with intense fear, horror, or helplessness. matic slides and sounds, and noradrenergic challenge have Animal studies in which these types of stressors were used found decreased medial prefrontal cortical/anterior cingulate provided evidence that brain regions involved in memory are function (Bremner 1998; Bremner et al 1997, 1999a, 1999b; sensitive to stress (Bremner 2002; McEwen et al 1992; Sapolsky Lanius et al 2001; Rauch et al 1997; Shin et al 1999). Other brain 1996). These stressors result in structural changes in the brain areas implicated include parietal cortex, hippocampus (Bremner region involved in new learning and memory (Gould et al 1998; et al 1997; Bremner et al 1999a), temporal cortex (Bremner et al Sapolsky 1996), with associated deficits in memory function 1997, 1999a, 1999b), posterior cingulate (Bremner et al 1999a, (Diamond et al 1996). 1999b; Lanius et al 2001), motor cortex (Bremner et al 1999a, The medial prefrontal cortex has also been implicated in the 1999b), anteromedial prefrontal cortex (Zubieta et al 1999), stress response and emotional regulation. Medial prefrontal middle/inferior frontal gyrus, cerebellum (Bremner et al 1999a, cortex in the human consists of several related areas, including 1999b; Rauch et al 1996; Shin et al 1999, 1997), and amygdala (Rauch et al 2000). From the Departments of Psychiatry and Behavioral Sciences and Radiology The Stroop paradigm has been used as a tool in research (JDB, EV, NA, CS, BE), Emory Center for Positron Emission Tomography, related to PTSD (Stroop 1935). In the Stroop task, naming the Emory University School of Medicine, Atlanta; Atlanta Veterans Affairs color of a word with an incongruent semantic context (e.g., Medical Center (JDB, EV, NA, CS, BE), Decatur, Georgia; and the Mood and naming the color of the word “red” written in the color yellow) Anxiety Disorders Research Program (MV, DSC), National Institute of results in a delay in naming relative to naming semantically Mental Health, Bethesda, Maryland. congruent words. Several hypotheses have been developed for Address reprint requests to J. Douglas Bremner, M.D., Emory University – this effect, including inhibition of the tendency to say the name Emory West Campus, Emory Clinical Neuroscience Research Unit, 1256 rather than the color, engagement in two parallel processes, and Briarcliff Road, Atlanta, GA 30306. distraction of attention (McNally et al 1993). A variant of the Received July 26, 2002; revised August 28, 2003; accepted October 3, 2003. classic color Stroop, called the “Counting Stroop” (Whalen et al 0006-3223/04/$30.00 BIOL PSYCHIATRY 2004;55:612– 620 doi:10.1016/j.biopsych.2003.10.001 © 2004 Society of Biological Psychiatry J.D. Bremner et al BIOL PSYCHIATRY 2004;55:612– 620 613 1998), which involves counting the number of color and emo- the study. No subjects were taking psychotropic medication at tional words that appear on a screen, has been developed for use the time of the study; however, subjects did not discontinue with functional magnetic resonance imaging to avoid the prob- taking medication for the purposes of participating in the study. lem with head motion associated with verbalization in the classic Subjects with medical or neurologic illness, comorbid psychotic color Stroop. Different types of “emotional Stroop” tasks have disorders, retained metal, a history of head trauma, loss of also been developed for application to patients with PTSD, for consciousness, cerebral infectious disease, or dyslexia were instance asking subjects to name the color of the word “rape.” excluded. There was no difference in age between the women Delay in naming of emotional words has been hypothesized to a with (37 ⫾ 11 years [mean ⫾ SD]) and without (35 ⫾ 9 years) diversion of attention toward traumatic as opposed to neutral PTSD [t(20) ⫽ 1.3, p ⫽.76]. There was no difference in years of material in individuals who have been sensitized to specific education between women with (14 ⫾ 2) and without (15 ⫾ 3) emotional stimuli through a history of traumatization (McNally et PTSD [t(20) ⫽ 1.5, p ⫽.15]. All subjects were right handed. al 1996), in addition to other factors that apply to the delay in color naming outlined above. Psychometric Assessments Patients with PTSD have been found to color-name emotional History of childhood abuse was assessed with the Early words more slowly than other types of words, relative to control Trauma Inventory (ETI). The ETI is a 56-item, clinician-adminis- subjects. Vietnam combat veterans with PTSD, compared with tered interview that assesses physical, emotional, and sexual combat veterans without PTSD, have been found to take longer abuse, as well as general traumatic events. The ETI has been to color-name “PTSD” words than specific obsessive-compulsive demonstrated to be reliable and valid in the assessment of disorder–related words, positive words, and neutral words, and childhood trauma (Bremner et al 2000). All women had a history this delay was correlated with severity of PTSD symptomatology of childhood rape, molestation, or attempted rape as assessed by as measured by the Mississippi Scale (McNally et al 1990, 1993, the ETI. 1996). Stroop interference has also been shown in patients with Of the women with abuse and current PTSD, 10 of 12 PTSD PTSD related to the trauma of rape (Cassiday et al 1992; Foa et al subjects (83%) fulfilled criteria for a past history of major 1991) and motor vehicle accidents (Beck et al 2001; Bryant and depression and 3 of 12 (25%) for current major depression, as Harvey 1995), in mixed-trauma groups (McNeil et al 1999), and in determined by the SCID interview. One patient (8%) had current children with PTSD (Moradi et al 1999). and lifetime dysthymia. In no cases did depression precede the Increased anterior cingulate activation has been consistently onset of trauma or PTSD. One patient (8%) fulfilled criteria for shown during the Stroop interference condition (Pardo et al current and lifetime history of panic disorder without agorapho- 1990; Peterson et al 1999; Whalen et al 1998). A study of bia. Three PTSD subjects (25%) met current and lifetime criteria combat-related PTSD in which the counting Stroop was used for generalized anxiety disorder and social phobia, and one for found a failure in anterior cingulate function during counting of simple phobia (8%). Three PTSD subjects (25%) fulfilled criteria emotional (but not color) words (Shin et al 2001). In patients with for a past history of alcohol dependence, one (8%) for a past depression (George et al 1997) and schizophrenia (Carter et al history of polysubstance dependence, two (17%) for a past 1997)— disorders also hypothesized to be related to anterior history of marihuana abuse or dependence, three (25%) for a cingulate dysfunction—performance of the Stroop task is asso- past history of cocaine abuse or dependence, and one (8%) with ciated with a failure in anterior cingulate activation. No studies to a past history of opiate dependence or abuse. No PTSD subjects date have measured neural correlates of the classic color Stroop had a current history of alcohol or substance abuse or depen- task in PTSD (naming word color for semantically incongruous dence. words). The purpose of the present study was therefore to use Of nine abused women without PTSD, five (56%) fulfilled positron emission tomography (PET) in the examination of criteria for a past history of major depression, with no women neural correlates of both color and emotional Stroop tasks in having current major depression, as determined by the SCID women with a history of childhood sexual abuse with and interview. One subject (11%) had a past history of social phobia, without the diagnosis of PTSD. We hypothesized that Stroop and one subject (11%) had a past history of bulimia. Two women performance would be associated with a failure of anterior (22%) fulfilled criteria for a past history of alcohol dependence, cingulate activation in abuse-related PTSD. three (33%) for a past history of polysubstance dependence. No non-PTSD subjects had a current history of alcohol or substance Methods and Materials abuse or dependence. Subjects underwent behavioral assessment at baseline and Subjects immediately after each of the PET conditions (described below), Twenty-one physically healthy women with a history of retrospectively assessed for the period of the condition. Assess- childhood sexual abuse (rape, molestation, or attempted rape ments included the Subjective Units of Distress Scale, a widely before the age of 18 years) participated in the study. Subjects used measure of distress in the cognitive behavioral treatment of included women with (n ⫽ 12) and without (n ⫽ 9) the PTSD (Meadows and Foa 1999), the Clinician-Administered diagnosis of current PTSD. All subjects were recruited through Dissociative States Scale, a validated measure of dissociative state newspaper advertisement. The study was approved by a local symptoms (Bremner et al 1998), the PTSD Symptom Scale, a investigational review board at Yale University School of Medi- measure of PTSD symptom level, as well as analogue scales of cine. Diagnosis of PTSD was established with the Structured fear and anxiety (Bremner et al 1997). Clinical Interview for DSM-IV (SCID; First et al 1995). Subjects gave written informed consent for participation, were free of PET Imaging Methods major medical illness on the basis of history and physical All subjects underwent PET scanning in conjunction with examination, laboratory testing, and electrocardiogram, and performance of the color and emotional Stroop tasks and control were not actively abusing substances or alcohol (past 6 months). condition. The control condition consisted of naming the color of Subjects were free of all medications for at least 4 weeks before colored XXs that were presented randomly in the colors red, www.elsevier.com/locate/biopsych 614 BIOL PSYCHIATRY 2004;55:612– 620 J.D. Bremner et al 6 mm full width half maximum [FWHM]). Each subject was placed in the scanner with her head in a holder to minimize motion and positioned with the canthomeatal line parallel to an external laser light. An intravenous line was inserted for admin- istration of H20. After positioning within the camera gantry, a transmission scan of the head was obtained with an external 67 Ga/68Ge rod source, to correct emission data for attenuation due to overlying bone and soft tissue. Subjects then underwent scanning during the control task and color and emotional Stroop tasks. Subjects were randomly as- signed to a control-color-emotional-control-color-emotional or emotional-color-control-emotional-color-control order of scan- ning. Ten seconds before administration of H20, subjects received instructions for the performance of the task. Subjects were asked to name the color of the word and to proceed as quickly as they could through the list of words. Subjects then received a bolus injection of 30 mCi of H20 followed 10 sec Figure 1. Posttraumatic stress disorder (PTSD) symptoms during color and emotional Stroop task. Subjects with PTSD had increased PTSD symptoms later by a 60-sec PET scan acquisition. The onset of the PET scan during all time points relative to abused women without PTSD, with the acquisition was timed to correspond to the point of maximum highest level of symptoms during the emotional Stroop task, which in- rate of increase in uptake of tracer into the brain. Subjects were volved color naming of words like “rape” or “mutilate.” asked to stop naming the color of words after 60 sec, and the number of words correctly identified by color, representing the blue, green and yellow on a computer screen. The active color Stroop score, was recorded. condition consisted of naming the color of colored words, in the According to the logic of the study design, brain blood flow colors red, blue, green, and yellow, when the semantic context of related to the act of identifying a color would be controlled for, the word was a color name incongruous with the color (e.g., the and any differences in brain blood flow between the Stroop and word “red” spelled in the color yellow). The active emotional the control condition would be secondary to the cognitive effort condition consisted of naming the color of words that were in the to inhibit the tendency to divert attention to the linguistic content colors red, blue, green, and yellow, when the words were the rather than the color content of the presented stimulus. On the emotional words, rape, bruise, weapon, and stench. These words basis of prior imaging studies with the Stroop reviewed above, were selected based on our prior studies showing that the words and on the hypothesis of anterior cingulate dysfunction in PTSD, were associated with a high degree of emotional arousal (Brem- we hypothesized an activation of the anterior cingulate with the ner et al 2001). Although the color and emotional words were Stroop in non-PTSD (but not PTSD) subjects. A similar logic matched for syllable length, and all of the words are used applied for the emotional Stroop versus control contrast. For the frequently in the English language, it was not possible to find emotional Stroop versus color Stroop contrast, both the process emotional words that are used as frequently in the English of identifying the color and ignoring the linguistic content of the language as the very commonly used color words (Kucera and words are controlled for, and differences in blood flow are Francis 1967). specifically related to the attempt to ignore the emotional content During the study period, assessments were also made of of the words. On the basis of prior imaging studies demonstrat- psychophysiologic factors, including the parameters of heart rate ing a role for the anterior cingulate in emotion reviewed above, and blood pressure. The purpose of measuring heart rate and and on the hypothesis of anterior cingulate dysfunction in PTSD, blood pressure was to have assessments of physiologic param- we hypothesized an activation of the anterior cingulate with the eters related to the stress of the emotional and color Stroop tasks. emotional Stroop in non-PTSD (but not PTSD) subjects. Heart rate was measured continuously during the entire study period with a Polar Vantage heart rate recording device (Wood- Image Processing and Data Analysis bury, New York). The Polar Vantage heart rate recording appa- Images were realigned to the first image in the scanning ratus was moistened with water to facilitate recording and session with statistical parametric mapping (spm96; Friston et al strapped around the subject’s chest for direct measurement of 1991). The mean concentration of radioactivity in each scan was heart rate. Heart-rate data were transmitted from the recording obtained as an area-weighted sum of the concentration of each device to a Polar Vantage recording device worn as a wrist watch slice and adjusted to the nominal value of 50 mL/min/100 g. The on the subject’s wrist. After the study session, the data were data were then rescaled and transformed into a common ana- downloaded to a personal computer for analysis. Heart rate over tomic space (expressed in Talairach three-dimensional x, y, and 5-sec intervals was compared during the interval of each task z coordinates) for statistical analysis. After transformation, images (described below) and during a baseline period 2–3 min before were smoothed to 16 mm FWHM before statistical analysis. each of these sessions. Data were digitized for analysis on a Regional cerebral blood flow was compared in Stroop and personal computer. Mean heart rate was calculated for each control conditions. Data were analyzed with spm96 (Friston et al baseline and task period, and the ⌬ of task minus baseline was 1991) with global blood flow considered as a confounding compared between groups. covariate with image data sets in which the values assigned to Each subject underwent six PET scans on a single day individual voxels correspond to t statistic. Statistical images were according to methods that have been described previously displayed with values of Z score units ⬎ 2.58 (p ⬍.005) and (Bremner et al 1999a, 1999b). Positron emission tomography clusters of more than 65 contingent significant voxels. A p value imaging was performed on a Posicam PET camera (Positron of.005 has been demonstrated to be associated with the most Corporation, Houston, Texas) (in-plane resolution after filtering, optimal minimization of type I and type II errors (Reiman et al www.elsevier.com/locate/biopsych J.D. Bremner et al BIOL PSYCHIATRY 2004;55:612– 620 615 Table 1. Behavioral and Physiological Responses to Neutral and Emotional Stroop Task in Women with Abuse with and without PTSD Time ⫻ Diagnosis Interaction Baseline Control #1 Neutral #1 Emotional #1 Control #2 Neutral #2 Emotional #2 F p Abused PTSD (n ⫽ 12) Diastolic blood pressure 75 (11) 79 (10) 77 (9) 79.8 (11) 81 (9) 79 (10) 79 (11).49.81 Dissociative states (CADSS) 4 (6) 6 (8) 6 (7) 8 (10) 6.63 (8) 7 (10) 7 (9).49.82 Anxiety (PASS) 5.90 (6)a 9.09 (8)a 8.9 (7)a 13.6 (12)a 12.7 (13)a 11.4 (11)a 10.8 (11)a 1.12.35 Fear (analogue) 3.6 (3.4) 3.6 (3.8) 3.5 (3.4) 4.1 (4.1) 2.9 (2.9) 2.9 (3.1) 3.7 (3.6)a 1.36.24 Subjective distress (SUDS).23 (.24).25 (.26).25 (.22).32 (.29).24 (.20).26 (.23).314 (.23)a.97.45 Abused Non-PTSD (n ⫽ 9) Diastolic blood pressure 72 (9) 74 (7) 73 (7) 75 (9) 73 (6) 74 (12) 75 (8) Dissociative states (CADSS) 0 (1) 0 (1) 1 (3) 2 (3) 0 (1) 1 (2) 1 (2) Anxiety (PASS) 0 (1) 1 (2) 3 (6) 3 (6) 2 (6) 3 (6) 2 (4) Fear (analogue) 1.4 (2.1) 1.0 (1.5) 1.6 (1.8) 1.8 (1.8) 2.0 (2.6) 1.5 (1.9) 1.1 (1.8) Subjective distress (SUDS).11 (.09).08 (.09).15 (.18).20 (.22).13 (.16).16 (.18).09 (.13) Data are expressed as mean (SD). PTSD, posttraumatic stress disorder; CADSS, Clinician Administered Dissociative States Scale; PASS, Panic Attack Symptom Scale; Analogue, Analogue ratings for fear; SUDS, Subjective Units of Distress Scale. a PTSD ⬎ non-PTSD by Duncan’s multiple range test. 1997). Areas of activation were identified with standard stereo- 4.28, p ⫽.04; Duncan test: control ⬎ color and emotional, p ⬍ tactic coordinates, according to the atlas of Talairach and Tour-.05). Subjects with PTSD named fewer items across conditions noux (1988). [main effect for diagnosis: F(1,5) ⫽ 3.96, p ⫽.002]. There was no Behavioral measures were analyzed with analysis of variance interaction between diagnosis and condition. Abused PTSD (ANOVA). For naming of Stroop words, the ⌬ of Stroop score compared with abused non-PTSD women showed a nonsignifi- between color Stroop and control and between emotional Stroop cant pattern of lower number of correctly named emotional versus and control was determined. The ⌬ of these “baseline” corrected color words as measured by the ⌬ between baseline adjusted scores was then determined as baseline corrected color Stroop emotional and color scores [⫺18.3 ⫾ 17.3 vs. ⫺13.7 ⫾ 25.4, minus emotional Stroop. The rationale for examining the ⌬ of the respectively; t(19) ⫽.47, p ⫽.65]. Posttraumatic stress disorder color/emotional Stroop and control condition was to examine for [F(1,19) ⫽ 6.93, p ⫽.01] (Figure 1), anxiety [F(1,19) ⫽ 7.27, p ⫽ specific Stroop effects controlling for reading speed, intelligence,.01], distress [F(1,19) ⫽ 3.08, p ⫽.10], fear [F(1,19) ⫽ 3.28, p ⫽ and other nonspecific performance-related parameters..08], and dissociative [F(1,19) ⫽ 4.44, p ⫽.04] symptoms were greater at all time points in the PTSD subjects relative to control Results subjects. There was no interaction between time and diagnosis Color naming of colored words and color naming of emo- for any behavioral measure (Table 1). Heart rate was higher at all tional words both resulted in a lower number of correctly named points in PTSD, both during the baseline periods and during the words per session than in the control condition [82 ⫾ 25 vs. 97 ⫾ tasks, and heart rate increased in both PTSD and control subjects 21 vs. 103 ⫾ 23, respectively; main effect for condition: F(1,5) ⫽ to an equal degree with all of the tasks (Figure 2). Figure 2. Heart rate response to color and emo- tional Stroop tasks. Individual symbols represent mean heart rate measured over 5-sec intervals dur- ing the period of tasks and intervening resting base- line periods. Posttraumatic stress disorder (PTSD) subjects had a higher heart rate at all time points relative to control subjects, both during the baseline periods between tasks (B) and during the control (CON), neutral color (NEU), and emotional Stroop (EMOT) conditions (see text for explanation). There was no difference in the ⌬ of task minus baseline between PTSD and control subjects for any of the conditions. www.elsevier.com/locate/biopsych 616 BIOL PSYCHIATRY 2004;55:612– 620 J.D. Bremner et al Table 2. Areas of Increased and Decreased Blood Flow with Neutral Stroop versus Control in Women with Childhood Abuse without PTSD Increased Blood Flow Decreased Blood Flow Talairach Coordinates Talairach Coordinates a a Z Score x y z Brain Region Z Score x y z Brain Region 3.13b ⫺50 ⫺42 ⫺24 Cerebellum 3.06b ⫺12 54 ⫺8 L anterior frontal cortex (10) 2.90b 10 32 2 Anterior cingulate (24,32) 2.90b 16 ⫺88 6 Cuneus (17) 3.17b 50 ⫺62 6 R middle temporal gyrus (37) 3.13b ⫺50 ⫺42 ⫺24 L fusiform gyrus (36) Numbers in parentheses represent Brodmann’s areas. PTSD, posttraumatic stress disorder; L, left; R, right. a Z score ⬎ 3.09, p ⬍.001; Z score ⬎ 2.58, p ⬍.005. b Area of greatest activation in a contiguous cluster of activated voxels that extends over several brain regions. Performance of the color Stroop compared with the control lobule (Table 3; Figure 3). There were relatively greater increases task resulted in a nonspecific increase in anterior cingulate blood in non-PTSD compared with abused PTSD women with the color flow in both PTSD and non-PTSD subjects. Direct comparison of Stroop in right visual association cortex, cuneus, and right the two groups showed no difference in activation between the inferior parietal lobule and relatively greater decreases in right two groups in anterior cingulate blood flow with the color superior temporal gyrus and orbitofrontal cortex (Table 4). Stroop. The emotional Stroop, when compared with the color The emotional Stroop compared with the color Stroop re- Stroop, showed decreased blood flow in the PTSD (but not the sulted in only a decrease in left inferior parietal lobule in the non-PTSD) subjects in anterior cingulate. This difference was control subjects. In the PTSD subjects, the emotional Stroop significant when the two groups were directly compared. When resulted in increased blood flow in left precuneus, right cuneus, the emotional Stroop was compared with the control condition, and right lingual gyrus and in decreased blood flow in left middle the non-PTSD (but not PTSD) subjects showed increased blood frontal gyrus, left inferior frontal gyrus, right anterior frontal flow in the anterior cingulate; this difference was not significant cortex, cerebellum, insula, uncus, right hippocampal region, when the two groups were directly compared. midbrain, and left superior temporal gyrus (Table 5; Figure 4). Other regions that showed increased activation in the non- Abused non-PTSD women had greater increases in blood flow PTSD subjects during the color Stroop when compared with the with the emotional Stroop compared with abused PTSD women control condition included cerebellum, right middle temporal in left insula, pons, cerebellum, and left middle frontal gyrus and gyrus, and left fusiform gyrus, and a decrease in blood flow was relatively greater decreases in precuneus (Table 6). seen in the left anterior frontal cortex and cuneus (Table 2). In The emotional Stroop compared with the control condition the PTSD subjects, the color Stroop resulted in increased blood resulted in (in addition to increased blood flow in anterior flow in cerebellum, left inferior and middle frontal gyrus, right cingulate) decreased flow in orbitofrontal cortex and right infe- superior temporal gyrus, cerebellum, right supplementary motor rior parietal lobule in the control subjects. In the PTSD subjects, area, and right amygdala and in decreased blood flow in right the emotional Stroop resulted in increased blood flow in left visual association cortex, precuneus, and right inferior parietal inferior frontal gyrus, cerebellum, hypothalamus, and subtha- Table 3. Areas of Increased and Decreased Blood Flow with Neutral Stroop versus Control in Women with Childhood Abuse-Related PTSD Increased Blood Flow Decreased Blood Flow Talairach Talairach Coordinates Coordinates Z Scorea x y z Brain Region Z Scorea x y z Brain Region 4.28b ⫺54 ⫺42 ⫺32 Cerebellum 3.61b 34 ⫺78 28 R visual association cortex (19) 3.38 ⫺46 ⫺74 ⫺30 3.52b 16 ⫺76 6 2.94 ⫺42 ⫺88 ⫺12 Visual association cortex (19) 3.46b 16 ⫺70 40 Precuneus (7) 3.90b ⫺24 14 ⫺16 L inferior frontal gyrus (47) 3.41 18 ⫺70 32 2.88b ⫺44 22 24 L middle frontal gyrus (46) 3.27b 64 ⫺30 24 R inferior parietal lobule (40) 3.82b 12 34 8 R anterior cingulate (24,32) 3.65 8 28 30 3.45b 50 18 ⫺12 R superior temporal gyrus (38) 3.40b 26 ⫺88 ⫺48 Cerebellum 3.20 14 ⫺86 ⫺50 3.17b 14 ⫺12 50 R supplementary motor area (6) 2.76b 28 0 ⫺16 R amygdala Numbers in parentheses represent Brodmann’s areas. PTSD, posttraumatic stress disorder; L, left; R, right. a Z score ⬎ 3.09, p ⬍.001; Z score ⬎ 2.58, p ⬍.005. b Area of greatest activation in a contiguous cluster of activated voxels that extends over several brain regions (the Z scores without footnote symbols that appear below the noted Z scores). www.elsevier.com/locate/biopsych J.D. Bremner et al BIOL PSYCHIATRY 2004;55:612– 620 617 prior functional imaging studies in abuse-related PTSD (Bremner et al 1995; Shin et al 1999). The color Stroop was associated with greater decreases in function in PTSD relative to control subjects in visual association cortex (BA 19), cuneus, and right inferior parietal lobule (BA 40) and with greater increases in right superior temporal gyrus (BA 38) and orbitofrontal cortex (BA 11). In addition to anterior cingulate, greater increases in blood flow were seen with the emotional Stroop in the abused non- PTSD compared with PTSD women in left insula, pons, cerebel- lum, and left middle frontal gyrus (BA 6) and relatively greater decreases in precuneus (BA 7) (Table 5). The findings of the current study demonstrate alterations in a network of brain regions that have been implicated in previous studies of PTSD. Prior studies of PTSD used a variety of tasks, Figure 3. Statistical parametric map overlaid on a magnetic resonance including script-driven imagery or exposure to traumatic slides imaging template of areas of significant activation during performance of and sounds, to activate specific memories associated with the the color Stroop task compared with the control condition in women with traumatic event. In the current study, the stress of performing the abuse and posttraumatic stress disorder. There was increased activation of Stroop under time pressure can be seen as a generic stressor; the the anterior cingulate as well as the right amygdala. Areas displayed in emotional Stroop (involving exposure to trauma-specific words) yellow are areas of activation at the level of Z score ⬎ 3.09 (p ⬍.001). Abused women without PTSD also showed anterior cingulate activation (data not involves the stress of the task as well as that of processing shown), which suggests that anterior cingulate activation with the color negative emotional information. As in prior PTSD studies, de- Stroop is nonspecifically seen in subjects with and without PTSD. Increased creased function was seen in visual association cortex, cuneus, amygdala activation is felt to be related to the stress of performance of the and inferior parietal lobule. Visual association cortex and cuneus task. are involved in making visual associations and processing visual imagery and memory. Altered function in these regions might lamic area. There was decreased blood flow in the PTSD group represent a neural correlate of alterations in visual imagery in in precuneus, right inferior parietal lobule, uncus, right hip- PTSD. Inferior parietal cortex is involved in spatial perception pocampal region, midbrain, and superior/middle temporal gy- and processing. Alterations in this region might be related to rus. Abused non-PTSD women compared with abused PTSD changes in spatial processing related to the increased attention or women had greater increases in blood flow with the emotional fear response seen in PTSD patients when under stress. Other Stroop versus control condition in right hippocampus and right brain areas implicated in the current study include the insula, inferior temporal gyrus and relatively greater decreases in pre- which plays a role in the fear response, and orbitofrontal cortex, cuneus and posterior cingulate. which is involved in emotional and social processing. We found increased right amygdala activation with the color Stroop in Discussion PTSD, although the differences did not reach statistical signifi- Performance of the color Stroop task (color naming of seman- cance when a direct comparison was made with non-PTSD tically incongruent words) resulted in a nonspecific activation of subjects. The amygdala plays a critical role in the fear response the anterior cingulate (BA 24 and 32) in abused women both with and emotional processing, and increased amygdala activation and without PTSD. Comparison of the emotional Stroop with the might be related to the stress performance of the color Stroop. color Stroop showed relatively lower blood flow response in the The absence of activation with the emotional– color Stroop anterior cingulate in the abused PTSD compared with the abused contrast in PTSD suggests that increased amygdala activation is non-PTSD women. These findings suggest that dysfunction of not associated with processing of emotional words in PTSD. This the anterior cingulate in PTSD is specific to the neural circuitry of is consistent with prior studies that showed that external threat processing of emotional stimuli and is not a nonspecific finding (but not internally generated emotion) is more consistently that can be elicited with a generic “anterior cingulate probe,” associated with amygdala activation in PTSD (Rauch et al 2000). such as the color Stroop task. The findings of the current study are also congruent with prior Other brain areas implicated in this study are consistent with studies of PTSD in showing decreased function in anterior Table 4. Areas of Greater Increased and Decreased Blood Flow with Neutral Stroop versus Control in Abused Women without PTSD compared with Women with PTSD Increased Blood Flow Decreased Blood Flow Talairach Talairach Coordinates Coordinates Z Scorea x y z Brain Region Z Scorea x y z Brain Region 3.81b 50 ⫺64 6 R visual association cortex (19) 2.98b 48 10 ⫺12 R superior temporal gyrus (38) 2.90b 16 ⫺74 32 Cuneus (19) 2.90b ⫺16 54 ⫺12 Orbitofrontal (11) 2.89b 60 ⫺36 32 R inferior parietal lobule (40) Numbers in parentheses represent Brodmann’s areas. PTSD, posttraumatic stress disorder; L, left; R, right. a Z score ⬎ 3.09, p ⬍.001; Z score ⬎ 2.58, p ⬍.005. b Area of greatest activation in a contiguous cluster of activated voxels that extends over several brain regions. www.elsevier.com/locate/biopsych 618 BIOL PSYCHIATRY 2004;55:612– 620 J.D. Bremner et al Table 5. Areas of Increased and Decreased Blood Flow with Emotional Stroop versus Neutral Stroop In Women with Childhood Abuse-Related PTSD Increased Blood Flow Decreased Blood Flow Talairach Talairach Coordinates Coordinates Z Scorea x y z Brain Region Z Scorea x y z Brain Region 3.11b 12 ⫺54 42 L precuneus (7) 4.27b ⫺46 22 24 L middle frontal gyrus (46) 2.72b 18 ⫺80 8 R cuneus (17) 3.67 ⫺24 14 ⫺2 L putamen 2.66 16 ⫺64 ⫺6 R lingual Gyrus (17,18) 3.64 ⫺42 24 4 L inferior frontal gyrus (45) 3.52b ⫺32 4 56 L middle frontal gyrus (6) 3.88b ⫺52 ⫺40 ⫺28 Cerebellum 3.86b ⫺38 ⫺12 2 Insula 3.73b 26 ⫺2 ⫺28 Uncus 3.12 42 ⫺16 ⫺18 R hippocampal region 3.39b 4 ⫺24 ⫺22 Midbrain 3.23b 24 ⫺74 ⫺30 Cerebellum 2.62 22 ⫺82 ⫺46 3.14b 10 48 20 R anterior frontal cortex (9) 3.07 6 30 32 R anterior cingulate (32) 3.04 16 54 10 R anterior frontal cortex (10) 3.78b ⫺64 ⫺48 14 L superior temporal gyrus (22) Numbers in parentheses represent Brodmann’s areas. PTSD, posttraumatic stress disorder; L, left; R, right. a Z score ⬎ 3.09, p ⬍.001; Z score ⬎ 2.58, p ⬍.005. b Area of greatest activation in a contiguous cluster of activated voxels that extends over several brain regions (the Z scores without footnote symbols that appear below the noted Z scores). cingulate (Bremner et al 1999a, 1999b; Shin et al 2001). The differences in the neural circuitry mediating the behavior, relative findings suggest that an interrelated network of brain regions, to the “classic” Stroop, which involves saying the color out loud which are involved in memory function, are dysfunctional in while inhibiting the tendency to speak the word. PTSD. Unlike prior studies, this study did not find a difference in The current findings are consistent with a prior study by Shin response time to emotional versus color Stroop in PTSD versus et al (2001), who showed a relative decrease in blood flow in control subjects. There was a pattern of difference in this anterior cingulate activation in combat-related PTSD. Similar to direction, however, that might have been significant with a larger the current study, the Shin et al study showed decreased blood sample size. The findings suggest that differences in brain flow only for the emotional (combat) and not the color Stroop activation between the groups were not solely secondary to condition. Those authors used a “counting Stroop,” which uti- differences in performance of the tasks. lized the method of silently counting the number of times a word These findings, interpreted in conjunction with prior imaging is seen in a particular color. This task might have important studies in PTSD, suggest that deficits in anterior cingulate func- tion are associated with PTSD. The neural circuitry of PTSD has been hypothesized to involve a variety of interconnected brain regions, including hippocampus, amygdala, thalamus, visual and parietal cortex, insula, and dorsolateral and medial prefrontal cortex (Bremner 2002; Pitman 2001). The medial prefrontal cortex has specifically been hypothesized to underlie the failure of emotional regulation and inability to turn off the fear response (through impairment of the ability to inhibit the amygdala) seen in PTSD subjects. The current study is consistent with a role for the anterior cingulate/medial prefrontal cortex area in the neural circuitry of PTSD. Deficits in anterior cingulate function are not specific to PTSD. Prior studies in stress-related disorders such as depression in which the Stroop task was used have found deficits in anterior cingulate function (George et al 1997). The George et al (1997) study also showed anterior cingulate deficits with use of a “sad” type of emotional Stroop. Studies have also found deficits with Figure 4. Statistical parametric map overlaid on a magnetic resonance the Stroop in schizophrenia (Carter et al 1997). We are not aware imaging template of areas of decreased blood flow during emotional Stroop of any studies in non-PTSD anxiety disorders in which either task in women with posttraumatic stress disorder (PTSD). Women with the color Stroop or a variant of an emotional Stroop was used, abuse-related PTSD showed decreased blood flow in anterior cingulate and thus we cannot comment on specificity within the anxiety right hippocampus. Areas displayed in blue are areas of decreased blood flow at the level of Z score ⬎ 3.09 (p ⬍.001). There were greater decreases disorders. The current study and the study by Shin et al (2001) seen in the anterior cingulate in abused women with versus without PTSD of men with combat-related PTSD, however, only found deficits during the emotional Stroop (data not shown). with the emotional Stroop. This suggests that PTSD might not www.elsevier.com/locate/biopsych J.D. Bremner et al BIOL PSYCHIATRY 2004;55:612– 620 619 Table 6. Areas of Greater Increased and Decreased Blood Flow with Emotional Stroop versus Neutral In Abused Women without PTSD Compared with Abused Women with PTSD Increased Blood Flow Decreased Blood Flow Talairach Talairach Coordinates Coordinates Z Scorea x y z Brain Region Z Scorea x y z Brain Region 3.18b ⫺38 ⫺20 4 L insula 3.28b 14 ⫺48 50 Precuneus (7) 2.92 ⫺34 ⫺8 ⫺2 2.93b ⫺52 ⫺14 ⫺20 2.69b 8 40 20 R anterior cingulate (32) 3.33b 0 ⫺24 ⫺30 Pons 2.93b 32 ⫺80 ⫺20 Cerebellum 2.84b ⫺32 4 56 L middle frontal gyrus (6) Numbers in parentheses represent Brodmann’s areas. PTSD, posttraumatic stress disorder; L, left; R, right. a Z score ⬎ 3.09, p ⬍.001; Z score ⬎ 2.58, p ⬍.005. b Area of greatest activation in a contiguous cluster of activated voxels that extends over several brain regions (the Z scores without footnote symbols that appear below the noted Z scores). be associated with nonspecific deficits in anterior cingulate women with and without posttraumatic stress disorder. Am J Psychiatry function but rather that tasks that involve emotional circuits 156:1787–1795. might be required to demonstrate this deficit in anterior cingulate Bremner JD, Randall PR, Capelli S, Scott TM, McCarthy G, Charney DS (1995): Deficits in short-term memory in adult survivors of childhood abuse. function. Psychiatry Res 59:97–107. Some limitations of this study are worthy of mention. Mea- Bremner JD, Soufer R, McCarthy G, Delaney R, Staib LH, Duncan JS, Charney surement of Stroop performance used speed of response of DS (2001): Gender differences in cognitive and neural correlates of re- correctly identified words and did not measure accuracy and membrance of emotional words. Psychopharmacol Bull 35:55–87. speed separately. The current study used only a group of abused Bremner JD, Staib L, Kaloupek D, Southwick SM, Soufer R, Charney DS non-PTSD controls and did not include a group of nonabused (1999b): Neural correlates of exposure to traumatic pictures and sound in Vietnam combat veterans with and without posttraumatic stress dis- non-PTSD women. This allowed us to examine only the effects of order: a positron emission tomography study. Biol Psychiatry 45:806 – PTSD and not effects that might be related to childhood abuse in 816. the absence of a psychiatric disorder. Additionally, the results are Bremner JD, Vermetten E, Mazure CM (2000): Development and prelim- specific to childhood abuse–related PTSD in women and cannot inary psychometric properties of an instrument for the measurement be generalized to men with PTSD or to PTSD related to other of childhood trauma: The Early Trauma Inventory. Depress Anxiety 12:1– traumas, such as combat. Future studies should measure neural 12. Bryant RA, Harvey AG (1995): Processing threatening information in post- correlates of the color and emotional Stroop in these other traumatic stress disorder. J Abnorm Psychol 104:537–541. gender and population groups. Carter CS, Mintun M, Nichols T, Cohen JD (1997): Anterior cingulate gyrus dysfunction and selective attention deficits in schizophrenia: [15O]H20 This study was supported by National Institute of Mental PET study during single-trial Stroop task performance. Am J Psychiatry 154:1670 –1675. Health Grant R01 #MH56120 to JDB, a National Center for Cassiday KL, McNally RJ, Zeitlin SB (1992): Cognitive processing of trauma Posttraumatic Stress Disorder grant, and a Veterans Administra- cues in rape victims with posttraumatic stress disorder. Cognit Ther Res tion Career Development Award grant (JDB). 16:283–295. We thank Helen Sayward, M.S., for image processing and data Charney DS, Nagy LM, Bremner JD, Goddard AW, Yehuda R, Southwick SM analysis; and Tammi Rowe-Padh, M.S., Sara Dubuque, M.S., (2000): Neurobiological mechanisms of human anxiety. In: Fogel BS, Nadeem Afzal, M.D., and Sajid Siddiq, M.D. for assistance in Schiffler RB, Rao SM, eds. Synopsis of Neuropsychiatry. Baltimore, MD: Lippincott Williams & Wilkins, 273–288. data collection. Damasio H, Grabowski T, Frank R, Galaburda AM, Damasio AR (1994): The return of Phineas Gage: Clues about the brain from the skull of a famous Beck JG, Freeman JB, Shipherd JC, Hamblen JL, Lackner JM (2001): Specificity patient. Science 264:1102–1105. of Stroop interference in patients with pain and PTSD. J Abnorm Psychol Davis M (1992): The role of the amygdala in fear and anxiety. Annu Rev 110:536 –543. Neurosci 15:353–375. Bremner JD (1998): Neuroimaging in posttraumatic stress disorder. Psychiatr Devinsky O, Morrell MJ, Vogt BA (1995): Contributions of anterior cingulate Ann 28:445–450. to behavior. Brain 118:279 –306. Bremner JD (2002): Does Stress Damage the Brain? Understanding Trauma- Diamond DM, Fleshner M, Ingersoll N, Rose GM (1996): Psychological stress Related Disorders from a Mind-Body Perspective. New York: W. W. Norton. impairs spatial working memory: Relevance to electrophysiological Bremner JD, Innis RB, Ng CK, Staib LH, Salomon RM, Bronen RA, et al (1997): studies of hippocampal function. Behav Neurosci 110:661–672. Positron emission tomography measurement of cerebral metabolic cor- First MB, Spitzer RL, Williams JBW, Gibbon M (1995): Structured Clinical Inter- relates of yohimbine administration in posttraumatic stress disorder. view for DSMIV-Patient Edition (SCID-P). Washington, DC: American Psy- Arch Gen Psychiatry 54:246 –256. chiatric Press. Bremner JD, Krystal JH, Putnam FW, Southwick SM, Marmar C, Charney DS, Foa EB, Feske U, Murdock TB, Kozak MJ, McCarthy PR (1991): Processing of Mazure CM (1998): Measurement of dissociative states with the Clinician threat related information in rape victims. J Abnorm Psychol 100:156 – Administered Dissociative States Scale (CADSS). J Trauma Stress 11:125– 162. 136. Friston KJ, Frith C, Liddle P, Frackowiak R (1991): Comparing functional (PET) Bremner JD, Narayan M, Staib LH, Southwick SM, McGlashan T, Charney DS images: The assessment of significant change. J Cereb Blood Flow Metab (1999a): Neural correlates of memories of childhood sexual abuse in 11:690 –699. www.elsevier.com/locate/biopsych 620 BIOL PSYCHIATRY 2004;55:612– 620 J.D. Bremner et al George MS, Ketter TA, Parekh PI, Rosinsky N, Ring HA, Pazzaglia PJ, et al Pardo JV, Pardo PJ, Janer KW, Raichle ME (1990): The anterior cingulate (1997): Blunted left cingulate activation in mood disorder subjects dur- cortex mediates processing selection in the Stroop attentional conflict ing a response interference task (the Stroop). J Neuropsychiatry Clin paradigm. Proc Natl Acad Sci U S A 87:256 –259. Neurosci 9:55–63. Peterson BS, Skudlarski P, Gatenby JC, Zhang H, Anderson AW, Gore JC Gould E, Tanapat P, McEwen BS, Flugge G, Fuchs E (1998): Proliferation of (1999): An fMRI study of Stroop word-color interference: Evidence for granule cell precursors in the dentate gyrus of adult monkeys is dimin- cingulate subregions subserving multiple distributed attentional sys- ished by stress. Proc Natl Acad Sci U S A 95:3168 –3171. tems. Biol Psychiatry 45:1237–1258. Kessler RC, Sonnega A, Bromet E, Hughes M, Nelson CB (1995): Posttraumatic Pitman RK (2001): Investigating the pathogenesis of posttraumatic stress stress disorder in the national comorbidity survey. Arch Gen Psychiatry disorder with neuroimaging. J Clin Psychiatry 62:47–54. 52:1048 –1060. Rauch SL, Savage CR, Alpert NM, Fischman AJ, Jenike MA (1997): The func- Kucera H, Francis WN (1967): Computational Analysis of Present-Day Ameri- tional neuroanatomy of anxiety: A study of three disorders using can English. Providence, RI: Brown University Press. positron emission tomography and symptom provocation. Biol Psychia- Lanius RA, Williamson PC, Densmore M, Boksman K, Gupta MA, Neufeld RW, try 42:446 –452. et al (2001): Neural correlates of traumatic memories in posttraumatic Rauch SL, van der Kolk BA, Fisler RE, Alpert NM, Orr SP, Savage CR, et al (1996): stress disorder: A functional MRI investigation. Am J Psychiatry A symptom provocation study of posttraumatic stress disorder using 158:1920 –1922. positron emission tomography and script driven imagery. Arch Gen Psy- LeDoux JE (1993): Emotional memory: In search of systems and synapses. chiatry 53:380 –387. Ann N Y Acad Sci 702:149 –157. Rauch SL, Whalen PJ, Shin LM, McInerney SC, Macklin ML, Lasko NB, et al (2000): McCauley J, Kern DE, Kolodner K, Dill L, Schroeder AF, DeChant HK, et al Exaggerated amygdala response to masked facial stimuli in posttraumatic (1997): Clinical characteristics of women with a history of childhood stress disorder: A functional MRI study. Biol Psychiatry 47:769 –776. abuse: Unhealed wounds. JAMA 277:1362–1368. Reiman EM, Lane RD, Ahern GL, Schwartz GE, Davidson RJ, Friston KJ, et al (1997): Neuroanatomical correlates of externally and internally gener- McEwen BS, Angulo J, Cameron H, Chao HM, Daniels D, Gannon MN, et al ated human emotion. Am J Psychiatry 154:918 –925. (1992): Paradoxical effects of adrenal steroids on the brain: Protection Sapolsky RM (1996): Why stress is bad for your brain. Science 273:749 –750. versus degeneration. Biol Psychiatry 31:177–199. Shin LM, Kosslyn SM, McNally RJ, Alpert NM, Thompson WL, Rauch SL, et al McNally RJ, Amir N, Lipke HJ (1996): Subliminal processing of threat cues in (1997): Visual imagery and perception in posttraumatic stress disorder: A posttraumatic stress disorder? J Anxiety Disord 10:115–128. positron emission tomographic investigation. Arch Gen Psychiatry McNally RJ, English GE, Lipke HJ (1993): Assessment of intrusive cognition in 54:233–241. PTSD: Use of the modified Stroop paradigm. J Trauma Stress 6:33–41. Shin LM, McNally RJ, Kosslyn SM, Thompson WL, Rauch SL, Alpert NM, et al McNally RJ, Kaspi RJ, Riemann BC, Zeitlin SB (1990): Selective processing of (1999): Regional cerebral blood flow during script-driven imagery in threat cues in posttraumatic stress disorder. J Abnorm Psychol 99:398 – childhood sexual abuse-related PTSD: A PET investigation. Am J Psychia- 402. try 156:575–584. McNeil DW, Tucker P, Miranda R, Lewin M, Nordgren JC (1999): Response to Shin LM, Whalen PJ, Pitman RK, Bush G, Macklin ML, Lasko NB, et al (2001): An depression and anxiety Stroop stimuli in posttraumatic stress disorder, fMRI study of anterior cingulate function in posttraumatic stress disor- obsessive-compulsive disorder and major depressive disorder. J Nerv der. Biol Psychiatry 50:932–942. Ment Dis 187:512–516. Stroop JR (1935): Studies of interference in serial verbal reactions. J Exp Meadows EA, Foa EB (1999): Cognitive-behavioral treatment of traumatized Psychol 18:643–661. adults. In: Saigh PA, Bremner JD, eds. Posttraumatic Stress Disorder: A Talairach J, Tournoux J (1988): Co-Planar Atlas of the Human Brain. New York: Comprehensive Text. Needham Heights, MA: Allyn & Bacon, 376 –390. Thieme Medical Publishers. Moradi AR, Doost HT, Taghavi MR, Yule W, Dalgleish T (1999): Everyday Vogt BA, Finch DM, Olson CR (1992): Functional heterogeneity in cingulate memory deficits in children and adolescents with PTSD: Performance on cortex: The anterior executive and posterior evaluative regions. Cereb the Rivermead Behavioural Memory Test. J Child Psychol Psychiatry Cortex 2:435–443. 40:357–361. Whalen PJ, Bush G, McNally RJ, Wilhelm S, McInerney SC, Jenike MA, Rauch Morgan CA, LeDoux JE (1995): Differential contribution of dorsal and ventral SL (1998): The emotional counting Stroop paradigm: A functional mag- medial prefrontal cortex to the acquisition and extinction of conditioned netic resonance imaging probe of the anterior cingulate affective divi- fear in rats. Behav Neurosci 109:681–688. sion. Biol Psychiatry 44:1219 –1228. Morgan CA, Romanski LM, LeDoux JE (1993): Extinction of emotional learn- Zubieta J-K, Chinitz JA, Lombardi U, Fig LM, Cameron OG, Liberzon I (1999): ing: Contribution of medial prefrontal cortex. Neurosci Lett 163:109 – Medial frontal cortex involvement in PTSD symptoms: A SPECT study. 113. J Psychiatry Res 33:259 –264. www.elsevier.com/locate/biopsych