Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arch Gen Psychiatry. Author manuscript; available in PMC 2012 January 9.
Published in final edited form as:
PMCID: PMC3253452

Relationship of Emotional Processing to Masked Faces in the Amygdala to Mood State and Treatment in Major Depressive Disorder

Teresa A. Victor, B.A.,1,2 Maura L. Furey, Ph.D.,1 Stephen J. Fromm, Ph.D.,1 Arne Öhman, Ph.D.,2 and Wayne C. Drevets, M.D.1



Major depressive disorder (MDD) is associated with behavioral and neurophysiological evidence for mood-congruent processing biases toward explicitly presented, emotionally-valenced stimuli. However, few studies have investigated such biases toward implicitly presented stimuli.


To investigate differential amygdala responses to sad, happy and neutral faces presented below the level of explicit conscious awareness using a backward masking task in unmedicated subjects with MDD and healthy controls.


Initial cross-sectional design followed by a longitudinal treatment trial using functional magnetic resonance imaging.


Psychiatric outpatient clinic at the National Institute of Mental Health.


Twenty-two unmedicated, currently-depressed subjects with MDD (dMDD), 16 unmedicated subjects with MDD in full remission (rMDD), and 25 healthy controls (HC).


Ten dMDD subjects underwent 8 weeks of antidepressant treatment with the selective serotonin reuptake inhibitor, sertraline.

Main Outcome Measures

Amygdala region-of-interest and whole brain analyses evaluated the hemodynamic response during exposure to masked-sad versus masked-happy faces, to masked-sad versus neutral faces, and to masked-happy versus neutral faces.


dMDD subjects showed greater amygdala responses than HC to masked-sad faces, while HC subjects showed greater amygdala responses to masked-happy faces. The bias toward sad faces also was evident in the rMDD relative to HC subjects and did not differ between the dMDD and rMDD subjects. This processing bias reversed toward the normative pattern in the dMDD subjects following sertraline treatment.


Emotional processing biases occur in the amygdala to sad faces presented below conscious awareness in currently-depressed or remitted-MDD subjects and to happy faces in controls. By influencing the salience of social stimuli, mood-congruent processing biases in the amygdala may contribute to dysfunction in conscious perceptions and social interactions in MDD. Our data suggest, however, that the negative bias resolves and a positive bias develops in MDD subjects during selective serotonin reuptake inhibitor treatment.

A mood-congruent processing bias toward negatively-valenced emotional stimuli is a consistent feature in the pathophysiology of major depressive disorder (MDD). This bias is evident in behavioral measures that evaluate memory and attention14, as well as in neurophysiological indices610. For example, neurophysiological responses to explicitly presented sad faces are exaggerated in the amygdala in depressed patients compared to healthy controls11, and this abnormality normalizes following antidepressant drug treatment (ADT)12.

The amygdala plays a pivotal role in evaluating the emotional salience of sensory stimuli through participation in two distinct types of distributed networks, one involving cortical regions that allow conscious or explicit stimulus perception, and the other involving subcortical structures that allow rapid, non-conscious assessment of stimulus features1314. Notably, healthy subjects show greater amygdala responses to happy versus sad faces when stimuli are presented below conscious awareness. This finding suggests the existence of a normal positive processing bias that is supported by subcortical networks which mediate rapid, automatic emotional evaluations15. Thus, it is conceivable that the negative processing bias that characterizes MDD also may be mediated by this rapid, non-conscious processing network involving the amygdala.

The current series of experiments used functional MRI (fMRI) to test the hypothesis that the amygdala response to emotional stimuli presented below conscious awareness would show a negative processing bias in unmedicated patients with MDD. First, this bias was evaluated between currently-depressed MDD (dMDD) and healthy control (HC) subjects. Secondly, the dependence of this emotional processing bias on current mood state was assessed by comparing amygdala responses between dMDD, HC and currently-remitted MDD subjects (rMDD). Finally, the sensitivity of the processing bias to antidepressant treatment was evaluated by comparing the amygdala responses in dMDD subjects before and after 8 weeks of sertraline treatment.



Twenty-two unmedicated adults with primary MDD in a current major depressive episode (dMDD)16, 16 adults with primary MDD currently in full remission (rMDD)16 and 25 healthy control subjects (HC) completed the fMRI protocol. Volunteers between 18–50 years were recruited through the clinical services of the NIMH or newspaper advertisements in the Washington, D.C. metropolitan area. Participants underwent a screening evaluation prior to enrollment that involved a medical and psychiatric history, laboratory testing, drug screening, physical examination and neuromorphological MRI scanning. Only right-handed individuals were selected, as assessed by the Edinburgh Handedness Inventory17. The psychiatric diagnosis was established using the Structured Clinical Interview for DSM-IV (SCID)18 and a semi-structured interview with a psychiatrist. The Family Interview for Genetic Studies19 was used to screen for family history of psychiatric disorders.

Subjects were excluded if they had: 1) serious suicidal ideation or behavior, 2) major medical or neurological disorders, 3) exposure to drugs likely to influence cerebral blood flow or neurological function within 3 weeks (8 weeks for fluoxetine), 4) a history of drug or alcohol abuse within 1 year or a lifetime history of drug or alcohol dependence16, 5) current pregnancy or breastfeeding, 6) general MRI exclusion criteria. Additional exclusions applied to the HC subjects were: 1) current or past history of any major psychiatric disorder, 2) a first-degree relative with a mood or anxiety disorder. Additional exclusions applied to the rMDD subjects were having experienced a depressive episode or having received psychotropic medications within three months prior to scanning. After receiving a complete explanation of the study procedures, all subjects provided written informed consent as approved by the NIMH-IRB. Subjects received financial compensation for their participation.

Intelligence testing and mood ratings were performed using the Weschler Abbreviated Scale of Intelligence (WASI)20, Hamilton Depression Rating Scale (HAM-D)21, Automatic Thoughts Questionnaire (ATQ)22, Inventory of Depressive Symptomatology: Self-Rating (IDS-SR)23, State-Trait Anxiety Inventory (STAI)24 and Thought Control Questionnaire (TCQ)25.

Antidepressant Drug Treatment

Ten of the dMDD subjects were rescanned after 8 weeks of treatment with the selective serotonin reuptake inhibitor (SSRI), sertraline. To control for test-retest and other nonspecific order effects, 10 HC subjects were rescanned following the same interval. Following the baseline scan in the pre-treatment condition, the dMDD subjects received sertraline (50 mg daily for three days and then titrated to 100 mg as tolerated). After three weeks of follow-up the dose was increased or decreased as clinically indicated. All subjects received a stable sertraline dose for at least four weeks prior to the post-treatment scan. At the post-treatment scan the mean sertraline dose was 105±50 mg (range 50–200 mg). Additional information regarding patient selection for the treatment study and a comparison of the subjects who received treatment versus those who did not appears in the Supplemental Materials.

fMRI Data Acquisition

Images were obtained on a General Electric 3.0 Tesla scanner (GE Signa, Milwaukee, WI) with an 8-channel phased-array head coil using an echoplanar imaging (EPI) pulse sequence (39 continuous slices, TE=20ms, TR=2000ms, flip angle=90°, 64×64 matrix, field of view=22cm, and voxel dimensions=3.4 × 3.4 × 3.0 mm3). A total of 290 fMRI images were acquired in each of four 10-minute runs during the backward masking task. The first four images of each run were discarded to allow for steady-state tissue magnetization. To provide an anatomical framework for analysis of the functional images, high-resolution anatomical images were also obtained using a fast spoiled gradient echo (FSPGR) sequence (TR=780ms, TE=2.7ms, flip angle=12°, FOV=22cm, matrix=224×224, 128 axial slices, 1.2mm thick, in-plane resolution=0.98mm2).

fMRI Backward Masking Task

Subjects were scanned while performing a novel version of a backward masking task using a slow-event related design (Figure 1). Prior to each run, subjects were shown two neutral target faces. They were instructed to remember the faces and to respond as quickly as possible to indicate whether a target face appeared during the current trial. Subjects used a button box with their right hand and pressed the “1” button if the face shown was one of the two target faces, or the “2” button if the target face was not shown. Target faces displayed neutral, sad or happy expressions and subjects judged whether a target face was present based on the identity of the person pictured, irrespective of emotional expression. Subjects demonstrated their understanding of the task by performing an abbreviated version outside the scanner using flash cards. Each task-trial displayed faces in pairs of two including a 26ms “masked” face immediately followed by a 107ms “masking” face to inhibit explicit perception of the first face2627. Each face stimulus was presented in the masked position and followed by a neutral stimulus for the following pairings: sad-neutral (SN), happy-neutral (HN) or neutral-neutral (NN). In addition, a neutral face stimulus was presented in the masked position and followed by an emotional face stimulus for the following pairings: neutral-sad (NS) and neutral-happy (NH). A sad or happy face stimulus in the masked position never also presented in the unmasked position. The SN, HN, NS, and NH stimulus types were presented eight times and the NN type 16 times within each run in a pseudo-randomized, mixed-trial design. Each run employed different target faces and emotional face stimuli from distinct actors. The data from the four runs were combined so that each stimulus type was presented a total of 32 times for stimulus pairs that included an emotional face and 64 times for pairs including only neutral faces. Within a single trial, the identity of the masked face was never the same as the identity of the masking face, but the two face stimuli always depicted the same gender. The gender for all stimulus pairings was balanced across runs. A 10–13s interstimulus interval was selected to allow the hemodynamic response to return to baseline prior to the next stimulus presentation.

Figure 1
Design of the backward masking task

Stimuli were presented using E-Primesoftware (Psychological Software Tools, Pittsburgh, PA) on a Monarch Hornet PC computer with a cathode ray tube monitor at 75 Hz and a cloned projection display to subjects in the scanner gantry. Accuracy of the presentation times was verified using a photodiode and oscilloscope. Face stimuli were obtained from the NimStim Set of Facial Expressions28. For Experiment 3, the dMDD and HC subjects performed the same task before and after treatment using unique face stimuli for the separate sessions.

Assessment of Behavioral Performance During fMRI

Behavioral data were analyzed using SPSS 14.0. Accuracy of responses and reaction times for identifying whether or not each face stimulus was a “target” were recorded using E-Prime software. The efficacy of the backward masking was assessed by categorizing each subject’s response to a stimulus event according to target detection accuracy. Subjects were debriefed following the fMRI study and asked about their experience of performing the task.

fMRI Processing and Statistical Analyses

Functional imaging analyses were performed using the general linear model within SPM5 (Wellcome Trust Center for Neuroimaging, London, Whole brain fMRI volumes were realigned to the first volume, co-registered to each subject’s anatomical scan, and normalized to fit the Montreal Neurological Institute (MNI) standard brain template. The data were smoothed with a Gaussian kernel (8 mm, full-width at half-maximum), high-pass filtered with a cutoff period of 128 sec to correct low-frequency artifacts, corrected for serial correlations choosing an autoregressive model of the order 1 [AR(1)] model, and the non-specific effects of global fluctuations in blood-oxygen-level dependent (BOLD) signal were removed using global normalization. Realignment parameters were modeled in the analysis as regressors to control for motion artifacts. We excluded runs from the analyses in which the subject showed movement of more than one-half voxel (1.5 mm) translation or 1.25° rotation. Imaging data from three or four runs were included for all subjects.

Since the masked and unmasked stimuli for each stimulus pair were presented too closely in time to model the hemodynamic response to each component separately, the fMRI data were modeled as event-related correlates of the combined stimulus pairs, and the hemodynamic response to the different pair combinations were compared29. This method allowed us to evaluate the effect of varying the emotional expression of the masked face while the expression of the unmasked face remained constant (i.e. SN-HN, SN-NN and HN-NN).

Single-subject t-contrast maps were generated by computing the difference maps between emotional conditions (e.g. masked sad-neutral versus masked happy-neutral; or SN-HN). At the group level, these difference maps were compared within an amygdala region-of-interest (ROI) to evaluate significant interactions and main effects through random-effects analysis of the beta-weight values obtained from the single-subject analyses. The voxel-wise statistical analysis within the amygdala was constrained using the “small-volume correction” option within SPM5 to reduce the likelihood of Type I error. Results included differences in amygdala activation that remained significant after applying false-discovery rate error correction for multiple comparisons or consisted of clusters ≥ 10 contiguous voxels at a threshold of p<0.05 (uncorrected) within the amygdala ROI. To assess group differences in other regions an exploratory whole brain analysis was performed post-hoc. The significance threshold was set at a cluster of ≥ 10 contiguous voxels for which the voxel p<0.001. Coordinates were transformed from MNI coordinates to the stereotaxic array of Talairach and Tournoux30. Anatomical localization was performed using stereotaxic atlases3031.

Data analyses were divided into three experiments and presented in the order conducted. First we tested the hypothesis that the amygdala response to emotional stimuli presented below conscious awareness would show a negative processing bias in MDD by comparing the difference in BOLD response between dMDD and healthy control subjects. We anticipated that the difference between groups would show the greatest effect size in the direct comparison of masked-sad versus masked-happy faces. Secondly, we characterized the influence of mood and clinical state by performing the same contrasts in a separate cohort of MDD subjects scanned while remitted and unmedicated. In the third experiment, we evaluated the effects of treatment on the emotional processing biases in a longitudinal assessment of dMDD subjects scanned before and during antidepressant pharmacotherapy.

In Experiment 1 the fMRI data acquired from the dMDD (n=22) and HC subjects (n=25) were compared using two-sample t-tests to evaluate the difference between groups for masked-sad versus masked-happy faces (SN-HN). To assess the specificity of the responses to each type of masked stimulus (happy or sad), post-hoc t-tests also evaluated differences in the amygdala response to masked-sad versus masked-neutral faces (SN-NN), and to masked-happy versus masked-neutral faces (HN-NN). Finally, to assess the specificity of the results for stimuli presented below the level of conscious awareness, we compared the BOLD response to unmasked-sad versus unmasked-happy faces (NS-NH).

In Experiment 2 the fMRI data obtained in the subjects from Experiment 1 were compared to those obtained from the 16 unmedicated rMDD subjects to evaluate the mood-state dependence of the emotional processing biases found in Experiment 1. The rMDD subjects performed the same backward-masking task. A two-way repeated-measures ANOVA was used to analyze hemodynamic differences across conditions (SN, HN, NN) and groups (dMDD, rMDD, HC).

Where emotion × group interactions were significant, post-hoc analyses were performed in SPSS 14.0. Beta-weight values were extracted at the peak voxel within a cluster for each subject and compared across groups using independent t-tests to characterize specific interaction effects.

In Experiment 3, ten dMDD subjects were rescanned after sertraline treatment. To control for test-retest and other nonspecific order effects, 10 HC subjects were rescanned following the same interval. Time × Group ANOVAs were performed on the data from the contrast(s) found in Experiment 1 to provide the most specific information regarding whether the response to sad faces (i.e., SN-NN) or to happy faces (i.e., HN-NN) accounted for abnormalities identified in the SN versus HN contrast. Paired t-tests were used to characterize the significance of differences within the dMDD subject group before versus during treatment and within the HC group across the same interval.


Demographic and clinical characteristics of the study participants appear in Tables 1 and and22 and eTables 1 and 2. Groups were similar for gender composition, mean age and mean intelligence quotients. Of the MDD samples 13 dMDD and 3 rMDD subjects were naive to psychotropic drugs. For those subjects who previously had received antidepressant medications the mean drug-free period was 21±23 months and 50±54 months for the dMDD and rMDD groups, respectively. The average age-of-onset was 17±6.0 years and 18±4.3 years for the dMDD and rMDD groups, respectively.

Table 1
Mean scores and standard deviation measurements for subject demographic characteristics and clinical symptom rating scales
Table 2
Mean scores and standard deviation measurements for MDD subject characteristics and mood assessment rating scales in the pre-versus post-treatment conditions

Clinical Ratings

One-way ANOVAs revealed a significant effect of group on mean scores for HAM-D (F2,60=274, p<0.001), ATQ (F2,60=120, p<0.001), IDS-SR (F2,60=221, p<0.001), STAI-S (F2,59=69.0, p<0.001) and STAI-T (F2,54=123, p<0.001) and three subscales of the TCQ: Distraction (TCQ-D; F2,58=61.3, p<0.005), Worry (TCQ-W; F2,58=65.6, p<0.001) and Punishment (TCQ-P; F2,58=65.1, p<0.001) (Table 1).

For the dMDD subjects who underwent treatment, ratings of illness severity significantly decreased (Table 2). Nine of the 10 MDD subjects were considered treatment responders (i.e. ≥50% improvement on HAM-D scores) and 7 of the 10 were considered to have remitted during treatment (i.e. HAM-D scores in the non-depressed range [≤7])32. Nevertheless, independent t-tests showed scores on the mood assessment ratings for dMDD subjects post-treatment remained higher than those of the unmedicated rMDD subjects on the HAM-D (t24=3.55, p<0.005), ATQ (t24=2.59, p<0.05), IDS-SR (t24=3.06, p<.01), STAI-S (t24=2.24, p<0.05), STAI-T (t7=2.10, p<0.05), TCQ-Worry (t24=2.21, p<0.05) and TCQ-Punishment (t9=2.92, p<0.01) scales.

fMRI Behavioral Performance

During debriefing no subjects reported awareness of seeing two face presentations. Participants performed at chance level in the identification of target faces when presented in the masked position (Table 3). A paired t-test comparing the correct detection rate to the incorrect detection rate (false alarm rate) showed that subjects did not differ in their response to a target face presented in the first position versus when no target face was presented (t=0.15, p>0.05). An ANOVA also revealed no difference between groups (F2,59=2.6, p>0.05). These data imply that the experimental method succeeded at “masking” the masked face stimuli so that they were not consciously perceived.

Table 3
Differential accuracy measuresa for detecting target faces when stimuli were presented in the masked (1st) versus the masking (2nd) face position

An ANOVA (5 stimulus types × 3 groups) revealed a group difference in detection of the target face in the Neutral_Neutral stimulus pair (F2,59=4.39, p<0.05). Both HC and rMDD subjects were more likely to detect the neutral target face in the masked position than dMDD subjects. No other significant between-group difference was found.

An ANOVA revealed a significant difference in reaction time to target masked-sad faces between groups (F=10.4, p<0.001). Post-hoc tests showed the rMDD subjects were faster than both dMDD subjects (p<0.05) and HC (p<0.001), while dMDD responded faster to masked-sad faces than HC (p<0.05). In HC, reaction time was faster to target masked-happy faces versus target masked-sad faces (t=3.6, p<0.001).

fMRI Results

In Experiment 1 the dMDD and HC subjects differed in the left and right amygdala response to SN-HN (t45=3.00, p<0.005 and t45=2.80, p<0.005 respectively; Figures 2A,2B). These results remained significant following false-discovery rate corrections for multiple comparisons (p<0.05, bilaterally). Post-hoc t-tests showed that the magnitude of the difference between SN-HN in left amygdala and right amygdala was greater in dMDD versus HC subjects (p<0.005, bilaterally; Figures 2C,2E). Similarly, the difference in left and right amygdala between SN-NN was greater in dMDD versus HC subjects (p<0.05, bilaterally; Figures 2C,2E). In contrast, the difference in left amygdala between HN-NN was greater in HC versus dMDD subjects (p<0.05; Figure 2D).

Figure 2
Neuroimaging results for experiments 1 and 2

Post-hoc assessments addressed the relationships between depression severity and behavioral performance in the amygdala response to masked-sad or happy faces. In dMDD subjects the HAM-D scores correlated inversely with the amygdala response to masked-happy faces (r=−0.48, p<0.05), such that the amygdala response to HN-NN decreased as depression severity increased (Figure 3). We found no significant relationship between depression severity and the amygdala response to SN-NN. Nevertheless, in dMDD subjects the reaction time to masked-sad faces was inversely correlated with the right amygdala response to SN-HN (r=−0.53, p<0.05) (eFigure 1). Additional correlational analyses of the relationship between reaction time and amygdala response are reported in the Supplemental Results.

Figure 3
Relationship between depression severity and amygdala response to HN-NN

In the exploratory whole brain analyses performed post-hoc the hemodynamic response to SN-HN was greater in left hippocampus in dMDD versus HC subjects (t45=3.28, p<0.001) and greater in left thalamus in HC versus dMDD subjects (t45=3.28, p<0.001)(eTable 3). In the post-hoc assessment of hemodynamic responses to unmasked-sad versus unmasked-happy faces (NS-NH) no significant difference was found between groups in the amygdala ROI. In the whole brain analysis of the same contrast, however, the BOLD response to NS-NH was greater in left temporopolar cortex in dMDD versus HC subjects (t45=3.28, p<0.001), and greater in HC subjects versus dMDD subjects in the superior frontal gyrus, right and left precentral gyrus, postcentral gyrus, middle temporal gyrus and parietal operculum (t45=3.28, p<0.001)(eTable 4).

In Experiment 2 the two-way repeated-measures ANOVA comparing hemodynamic differences across conditions (SN, HN, NN) and groups (dMDD, rMDD, HC) showed a condition by group interaction (F3,92=4.19, p<0.01; Figures 2F,2G). Post-hoc t-tests indicated that the magnitude of the difference in the amygdala hemodynamic response to SN versus HN was greater in both dMDD and rMDD groups than in HC (p<0.005 and p<0.05, respectively; Figure 2H).

In additional post-hoc t-tests, the dMDD group showed greater amygdala activity to SN versus NN (p<0.05), while HC subjects showed no such effect (p=0.51), and the difference between groups was significant (p<0.05; Figure 2I). In contrast, the controls showed higher amygdala activity to HN versus NN (p<0.05), while the dMDD and rMDD subjects showed no such difference (p=0.87 and p=0.83, respectively), although the difference across groups was not significant (Figure 2J). The dMDD and rMDD groups did not differ significantly in their amygdala response to any task condition.

In Experiment 3, following treatment the dMDD subjects showed reduced activity to SN-NN in right amygdala (t9=3.26, p<0.01; Figures 4A,4B), and elevated activity to HN-NN in left amygdala (t9=2.59, p<0.05; Figures 4A,4C).

Figure 4
Neuroimaging results for experiment 3

A time × group ANOVA on responses to SN-NN revealed a significant interaction in the right amygdala (t18=2.21, p<0.05; Figures 3D,3E). Individual comparisons performed post-hoc showed a reduction in the amygdala response to SN-NN in dMDD in the pre-versus post-treatment conditions (p<0.05) with no significant change across time in the controls (p=0.17; these post-hoc comparisons excluded a single healthy control whose contrast beta-weight value exceeded three standard deviations beyond the mean.


These data demonstrate that negative emotional processing biases occur automatically, below conscious awareness, in unmedicated-depressed MDD subjects. Both unmedicated-depressed and unmedicated-remitted MDD subjects showed greater amygdala activity than controls when processing masked-sad versus masked-happy faces. Depressed MDD subjects also responded faster than HC subjects to masked-sad faces, despite being unaware of the masked face (Table 3). In contrast, healthy subjects showed larger responses in the amygdala and faster behavioral responses to masked-happy faces relative to masked-sad or neutral faces, consistent with other evidence that healthy humans show a processing bias toward positively valenced stimuli15,3334.

This non-conscious processing of emotional stimuli is consistent with evidence that the amygdala contains cells that are tuned selectively to specific stimulus characteristics, facilitating early detection of biologically salient information35. The coordinates for the emotional processing biases found herein (Figure 1) appear to implicate specifically the lateral nucleus of the amygdala31, which receives monosynaptic projections both from the sensory cortices that allow conscious or explicit stimulus perception and from the subcortical structures that support rapid, non-conscious assessment of stimulus features1314. The rapid response system facilitates detection of and behavioral adaptation to stimuli that are novel, threatening, rewarding or socially significant3642. The exploratory whole brain analysis (eTable 3) implicated hippocampus and thalamus in the extended anatomical network that, together with the amygdala, responds to non-conscious stimuli. Projections from the hippocampus to amygdala provide input during emotional processing about the environmental context43, while the thalamus plays a role in gating the transmission of sensory information to other brain regions based upon the anticipated salience of this information within the behavioral context44. Non-conscious emotional mood-congruent processing biases in the amygdala of MDD subjects may negatively influence their conscious perceptions of experiential stimuli and impact social interactions.

The comparison of hemodynamic responses to unmasked-sad versus unmasked-happy faces showed the specificity of our amygdala results to masked stimuli, as we found no difference across groups in the amygdala response to explicitly presented emotional faces. This suggests using backward masking confers an advantage in identifying emotional processing biases involving the amygdala in MDD.

The emotional processing abnormalities found in unmedicated-depressed MDD subjects extended to unmedicated-remitted MDD subjects, suggesting that MDD is associated with a trait-like bias toward processing negative stimuli independently of current mood state. Nevertheless, while rMDD cases met criteria for full remission, they showed an elevation of trait anxiety ratings and negative thought patterns (Table 1). These symptom clusters appear endophenotypic in individuals who develop MDD45 and conceivably relate to the persistent emotional processing bias observed herein. This finding is suggestive of an “illness-congruent” processing bias in remitted subjects that may serve as a biomarker for the vulnerability to depressive relapse and recurrence within MDD.

This pattern of amygdala activity reversed during treatment, however, as the response bias to masked-sad faces disappeared (Figure 2B), while a bias to masked-happy faces developed in MDD individuals receiving treatment (Figure 2C). Previous studies reported that the amygdala response to unmasked-sad faces12 or masked-fearful faces4647 attenuated during treatment, whereas our study was the first to identify a reciprocal increase in amygdala activity in response to masked-happy faces together with a concomitant decrease in amygdala activity in response to masked-sad faces associated with treatment. These findings thus provide the first evidence of a non-conscious negative processing bias toward sad faces in unmedicated patients with MDD that resolves, while a positive processing bias emerges, during treatment.

Our results appear compatible with the hypothesis48 that antidepressant drugs exert their primary therapeutic mechanism by normalizing the negative bias on information processing. This hypothesis was based partly on evidence that in healthy humans short-term administration of citalopram enhanced the amygdala response to happy faces49 and in depressed subjects acute administration of reboxetine enhanced the behavioral responses to positively valenced stimuli.50 Longitudinal studies are needed to assess whether the ability of antidepressant drugs to reduce relapse vulnerability and improve clinical outcomes relates directly to the attenuation of the automatic amygdala response to negative stimuli.

Notably, Suslow et al (2009) reported that MDD patients who both were antidepressant-medicated and persistently depressed showed hemodynamic responses in right amygdala that were exaggerated to masked-sad faces and blunted to masked-happy faces. Although this study did not include an unmedicated sample for comparison, when their data are considered together with ours, the combined results suggest that the decrement in right amygdala responses to masked-sad faces we found (Figure 4) during pharmacotherapy may depend upon treatment effectiveness. Nine of the 10 cases we studied post-treatment showed good clinical responses, so we could not compare neurophysiological effects between responders and non-responders.

The importance of laterality effects also is raised by these data, although neither study addressed laterality effects specifically. Suslow et al. (2009) observed evidence of emotional processing biases in the right--but not in the left--amygdala in medicated-depressed MDD subjects. Our study, the first to examine the responses to masked-happy or sad stimuli in the amygdala of unmedicated-depressed MDD subjects and also the first to examine this phenomenon in unmedicated-remitted MDD subjects, additionally found these emotional processing biases exist in both subject groups in the left amygdala. Moreover in our longitudinal study the “normal” positive processing bias that emerged post-treatment was significant only in the left amygdala, while the attenuation of the negative processing bias was significant on the right (Figure 4).

Some researchers have suggested emotional processing biases are limited to late or controlled information processing in MDD5153. In contrast, our data suggest these biases are also evident at an automatic or early processing level. Studies that have suggested processing biases in depression are limited to late or controlled processing employed behavioral assessments of attention and memory for anxiety-related or socially threatening stimuli, or sad words. Depressed patients exhibit a specific bias toward sad stimuli, but have not consistently shown processing biases to socially or physically threatening stimuli54. In this study, the bias was for sad words shown for 500–1000ms. However, an implicit emotional processing bias was not found toward briefly presented sad words shown for 14ms53. Verbal stimuli may require longer processing times to detect their emotional salience. Given the biological salience of faces, face stimuli can be processed rapidly, within the time frame needed for backward masking techniques14. Also, the effects of antidepressant medication were not controlled in these studies. Antidepressant treatment has been shown decrease negative emotional information processing in depressed patients so it is plausible that previous studies were unable to detect a processing bias earlier during information processing due to confounding medication effects.

Several limitations of our study merit comment. First, we did not address the generalizability of these findings to other mood disorders or to other antidepressant drug classes. Second, the rMDD and dMDD samples were not the same subjects studied in distinct illness phases, and the rMDD sample had fewer subjects with comorbid anxiety disorders than the dMDD sample. Longitudinal component of this study did not include a placebo arm, so causal evidence for a pharmacotherapeutic effect could not be established by the results.

In summary, our findings provide behavioral and neurophysiological support for an emotional processing bias in depression to negatively-valenced stimuli presented below conscious awareness that persists independently of the current mood-state. Developmental studies are needed to explore whether this processing bias constitutes a potential endophenotype in MDD and to characterize its relationship to the emergence of depressive episodes.

Supplementary Material

Supplementary Figure 1

Text and Tables


Grant support for the work was provided by the National Institutes of Health, National Institute of Mental Health, grant number Z01-MH002792.

Grant support for the work was provided by the National Institutes of Health, National Institute of Mental Health (NIMH), via grant number Z01-MH002792. The NIMH DIRP arranged peer review of the study design, provided IRB oversight of the data collection, management and analysis, and approved submission of the manuscript for publication. However, this sponsor did not influence directly the interpretation of the results or preparation of the manuscript.

We thank Dr. Harvey Iwamoto for programming the backward masking task, Joan Williams, Michele Drevets, and Dr. Paul Carlson for recruitment assistance and clinical support, Drs. Allison Nugent and Sean Marrett for technical support and scientific direction with fMRI scanning and data analysis, and Jeanette Black and Renee Hill for MRI technologist support. We also thank the developers of the NimStim Set of Facial Expressions for use of their stimuli in the fMRI task. Development of the MacBrain Face Stimulus Set was overseen by Nim Tottenham and supported by the John D. and Catherine T. MacArthur Foundation Research Network on Early Experience and Brain Development. Please contact Nim Tottenham at ude.nmu.ct@6000ttot for more information concerning the stimulus set.


The paper was presented in part at the 36th and 38th annual meetings of the Society for Neuroscience, Atlanta, GA, Oct. 14–18, 2006, and Washington, DC, Nov. 15–19, 2008, the 46th annual meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, Dec. 9–13, 2007, and the 12th and 13th annual meetings of the Organization for Human Brain Mapping, Florence, Italy, June 11–15, 2006, and Chicago, IL, June 10–14, 2007.

The experiments described herein were performed at the National Institute of Mental Health, Division of Intramural Research Programs, in the Section on Neuroimaging in Mood and Anxiety Disorders at the National Institutes of Health, Bethesda, Maryland.

All of the authors on this manuscript report no competing interests. The antidepressant drug, sertraline, used in the treatment portion of these experiments, was provided by the National Institutes of Health (NIH), Division of Intramural Research Program (DIRP), NIH Clinical Center.

The corresponding author, Dr. Wayne C. Drevets, had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.


1. Murphy FC, Sahakian BJ, Rubinsztein JS, Michael A, Rogers RD, Robbins TW, Paykel ES. Emotional bias and inhibitory control processes in mania and depression. Psychol Med. 1999;29(6):1307–1321. [PubMed]
2. Murray LA, Whitehouse WG, Alloy LB. Mood congruence and depressive deficits in memory: a forced-recall analysis. Memory. 1999;7(2):175–196. [PubMed]
3. Gotlib IH, Krasnoperova E, Yue DN, Joormann J. Attentional biases for negative interpersonal stimuli in clinical depression. J Abnorm Psychol. 2004;113(1):121–135. [PubMed]
4. Bradley BP, Mogg K, Williams R. Implicit and explicit memory for emotion-congruent information in clinical depression and anxiety. Behav Res Ther. 1995;33(7):755–770. [PubMed]
5. Suslow T, Konrad C, Kugel H, Rumstadt D, Zwitserlood P, Schöning S, Ohrmann P, Bauer J, Pyka M, Kersting A, Arolt V, Heindel W, Dannlowski U. Automatic mood-congruent amygdala responses to masked facial expressions in major depression. Biol Psychiatry. 2009 Epub ahead of print. [PubMed]
6. Drevets WC, Price JL, Furey ML. Brain structural and functional abnormalities in mood disorders: implications for neurocircuitry models of depression. Brain Struct Funct. 2008;213(1–2):93–118. [PMC free article] [PubMed]
7. Taylor Tavares JV, Clark L, Furey ML, Williams GB, Sahakian BJ, Drevets WC. Neural basis of abnormal response to negative feedback in unmedicated mood disorders. Neuroimage. 2008;42(3):1118–1126. [PMC free article] [PubMed]
8. Leppanen JM. Emotional information processing in mood disorders: a review of behavioral and neuroimaging findings. Curr Opin Psychiatry. 2006;19(1):34–39. [PubMed]
9. Murphy FC, Michael A, Robbins TW, Sahakian BJ. Neuropsychological impairment in patients with major depressive disorder: the effects of feedback on task performance. Psychol Med. 2003;33(3):455–467. [PubMed]
10. Siegle GJ, Steinhauer SR, Thase ME, Stenger VA, Carter CS. Can’t shake that feeling: event-related fMRI assessment of sustained amygdala activity in response to emotional information in depressed individuals. Biol Psychiatry. 2002;51(9):693–707. [PubMed]
11. Surguladze S, Brammer MJ, Keedwell P, Giampietro V, Young AW, Travis MJ, Williams SC, Phillips ML. A differential pattern of neural response toward sad versus happy facial expressions in major depressive disorder. Biol Psychiatry. 2005;57(3):201–209. [PubMed]
12. Fu CH, Williams SC, Cleare AJ, Brammer MJ, Walsh ND, Kim J, Andrew CM, Pich EM, Williams PM, Reed LJ, Mitterschiffthaler MT, Suckling J, Bullmore ET. Attenuation of the neural response to sad faces in major depression by antidepressant treatment: a prospective, event-related functional magnetic resonance imaging study. Arch Gen Psychiatry. 2004;61(9):877–889. [PubMed]
13. LeDoux JE. The Emotional Brain: The Mysterious Underpinnings of Emotional Life. New York, New York: Simon and Schuster; 1996.
14. Morris JS, Öhman A, Dolan RJ. A subcortical pathway to the right amygdala mediating “unseen” fear. Proc Natl Acad Sci U S A. 1999;96(4):1680–1685. [PubMed]
15. Killgore WD, Yurgelun-Todd DA. Activation of the amygdala and anterior cingulate during nonconscious processing of sad versus happy faces. Neuroimage. 2004;21(4):1215–1223. [PubMed]
16. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4. Washington, DC: American Psychiatric Association; 2000. Text Revision.
17. Oldfield RC. The assessment and analysis of handedness: the Edinburgh Inventory. Neuropsychologia. 1971;9(1):97–113. [PubMed]
18. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition (SCID-I/P) New York, NY: New York State Psychiatric Institute, Biometrics Research; 2002.
19. Maxwell E. Family Interview for Genetic Studies (FIGS): A Manual for FIGS. Bethesda, MD: Clinical Neurogenetics Branch, Intramural Research Program, National Institute of Mental Health; 1992.
20. Wechsler D. Wechsler Abbreviated Scale of Intelligence (WASI) San Antonio, TX: Harcourt Assessment; 1999.
21. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960;23:56–62. [PMC free article] [PubMed]
22. Hollon SD, Kendall PC. Cognitive self-statements in depression: development of an automatic thoughts questionnaire. Cog Ther Res. 1980;4:383–395.
23. Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH. The Inventory of Depressive Symptomatology (IDS): psychometric properties. Psychol Med. 1996;26(3):477–486. [PubMed]
24. Spielberger CD, Gorsuch RL, Lushene RE. Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press; 1970.
25. Wells A, Davies MJ. The thought control questionnaire: a measure of individual differences in the control of unwanted thoughts. Behav Res Ther. 1994;32(8):871–878. [PubMed]
26. Wiens S, Fransson P, Dietrich T, Lohmann P, Ingvar M, Öhman A. Keeping it short: a comparison of methods for brief picture presentation. Psychol Sci. 2004;15(4):282–285. [PubMed]
27. Esteves F, Öhman A. Masking the face: recognition of emotional facial expressions as a function of the parameters of backward masking. Scand J Psychol. 1993;34(1):1–18. [PubMed]
28. Tottenham N, Tanaka JW, Leon AC, McCarry T, Nurse M, Hare TA, Marcus DJ, Westerlund A, Casey B, Nelson C. The NimStim set of facial expressions: judgments from untrained research participants. Psychiatry Res. 2009;168(3):242–249. [PMC free article] [PubMed]
29. Menon RS, Goodyear BG. Spatial and temporal resolution in fMRI. In: Jezzard P, Matthews PM, Smith SM, editors. Functional MRI: An Introduction to Methods. New York, NY: Oxford University Press; 2001. pp. 145–158.
30. Talairach J, Tournoux P. Coplanar Stereotaxic Atlas of the Human Brain. New York, NY: Thieme Medical Publishers; 1998.
31. Mai JK, Assheuer J, Paxinos G. Atlas of the Human Brain. 2. San Diego, CA: Academic Press; 2003.
32. Nierenberg AA, DeCecco LM. Definitions of antidepressant treatment response, remission, nonresponse, partial response and other relevant outcomes: a focus on treatment-resistant depression. J Clin Psychiatry. 2001;62(suppl 16):5–9. [PubMed]
33. Erickson K, Drevets WC, Clark L, Cannon DM, Bain EE, Zarate CA, Jr, Charney DS, Sahakian BJ. Mood-congruent bias in affective go/no-go performance of unmedicated patients with major depressive disorder. Am J Psychiatry. 2005;162(11):2171–2173. [PubMed]
34. Sharot T, Riccardi AM, Raio CM, Phelps EA. Neural mechanisms mediating optimism bias. Nature. 2007;450(7166):102–105. [PubMed]
35. Bordi F, LeDoux J. Sensory tuning beyond the sensory system: an initial analysis of auditory response properties of neurons in the lateral amygdaloid nucleus and overlying areas of the striatum. J Neurosci. 1992;12(7):2493–2503. [PubMed]
36. Leonard CM, Rolls ET, Wilson FAW, Baylis GC. Neurons in the amygdala of the monkey with responses selective for faces. Behav Brain Res. 1985;15(2):159–176. [PubMed]
37. Morris JS, Frith CD, Perrett DI, Rowland D, Young AW, Calder AJ, Dolan RJ. A differential neural response in the human amygdala to fear and happy facial expressions. Nature. 1996;383(6603):812–815. [PubMed]
38. Whalen PJ, Rauch SL, Etcoff NL, McInerney SC, Lee MB, Jenik MA. Masked presentations of emotional facial expressions modulate amygdala activity without explicit knowledge. J Neurosci. 1998;18(1):411–418. [PubMed]
39. Calder AJ, Lawrence AD, Young AW. Neuropsychology of fear and loathing. Nat Rev Neurosci. 2001;2(5):352–363. [PubMed]
40. Adolphs R. Neural systems for recognizing emotion. Curr Opin Neurobiol. 2002;12(2):169–177. [PubMed]
41. Phillips ML, Drevets WC, Rauch SL, Lane R. The neurobiology of emotion perception I: the neural basis of normal emotion perception. Biol Psychiatry. 2003;54(5):504–514. [PubMed]
42. Murray EA. The amygdala, reward and emotion. Trends Cogn Sci. 2007;11(11):489–497. [PubMed]
43. Phillips RG, LeDoux JE. Differential contribution of amygdala and hippocampus to cued and contextual fear conditioning. Behav Neurosci. 1992;106:274–285. [PubMed]
44. Posner MI, Petersen SE. The attention system of the human brain. Annu Rev Neurosci. 1990;13:25–42. [PubMed]
45. Hasler G, Drevets WC, Manji HK, Charney DS. Discovering endophenotypes for major depression. Neuropsychopharmacology. 2004;29(10):1765–1781. [PubMed]
46. Sheline YI, Barch DM, Donnelly JM, Ollinger JM, Snyder AZ, Mintun MA. Increased amygdala response to masked emotional faces in depressed subjects resolves with antidepressant treatment: an fMRI study. Biol Psychiatry. 2001;50(9):651–658. [PubMed]
47. Harmer CJ, Mackay CE, Reid CB, Cowen PJ, Goodwin GM. Antidepressant drug treatment modifies the neural processing of nonconscious threat cues. Biol Psychiatry. 2006;59(9):816–820. [PubMed]
48. Harmer CJ. Serotonin and emotional processing: Does it help explain antidepressant drug action? Neuropharmacology. 2008;55:1023–1028. [PubMed]
49. Norbury R, Taylor MJ, Selvaraj S, Murphy SE, Harmer CJ, Cowen PJ. Short-term antidepressant treatment modulates amygdala response to happy faces. Psychopharmacology. 2009;206:197–204. [PubMed]
50. Harmer CJ, O’Sullivan O, Favaron E, Massey-Chase R, Ayres R, Reinecke A, Goodwin GM, Cowen PJ. Effect of acute antidepressant administration on negative affective bias in depressed patients. Am J Psychiatry. 2009;166:1178–1184. [PubMed]
51. Mathews A, MacLeod C. Cognitive Vulnerability to Emotional Disorders. Ann Rev Clin Psychol. 2005;1:167–95. [PubMed]
52. Mathews A, Ridgeway V, Williamson DA. Evidence for attention to threatening stimuli in depression. Behav Res Ther. 1996;34:695–705. [PubMed]
53. Mogg K, Bradley B, Williams R. Attentional bias in anxiety and depression: The role of awareness. Br J Clin Psychol. 1995;34:17–36. [PubMed]
54. Gotlib IH, Kasch KL, Traill S, Jorrmann J, Arnow BA, Johnson SL. Coherence and specificity of information-processing biases in depression and social phobia. J Abnorm Psychol. 2004;113:386–398. [PubMed]