Reappraisal is an adaptive emotion regulation strategy while the role of self-perspective in reappraisal process of depressed patients is largely unknown in terms of goals (valence/arousal) and tactics (detachment/immersion). In this study, 12 depressed individuals and 15 controls were scanned with MRI during which they either attend naturally to emotional stimuli, or adopt detachment/immersion strategy. Behaviorally, no group differences in self-reported emotion regulation effectiveness were found. In addition, we observed that (1) patients were less able to downregulate amygdala activation with recruitment of more dorsal lateral prefrontal cortex (dlPFC) when adopting detachment strategy regardless of valence, and this preserved ability to regulate emotion was inversely associated with severity of symptoms; (2) patients had deficits in upregulating amygdala activation when adopting immersion strategy, with less inferior frontal gyrus (IFG) activation and strengthening coupling of dlPFC and ventral medial prefrontal cortex (vmPFC) with amygdala; (3) comparison between groups yielded that patients showed stronger vmPFC activation under either self-detached or self-immersed condition. In conclusion, impaired modulatory effects of amygdala in depressed patients are compensated with strengthening cognitive control resources, with dissociable effects for different self-perspectives in reappraisal. These results may help clarify the role of self-perspective underlying reappraisal in major depression.
According to the fourth edition of Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), anhedonia and/or a lingering low mood are the defining characteristics of major depressive disorder (MDD). A body of evidence implies a trait-like role of maladaptive emotion regulation strategies in MDD vulnerability [
Operationally, reappraisal refers to a combination of approaches that require generation, maintenance, coordination of top-down cognitive reinterpretation, and bottom-up appraisal of affective events and monitoring them in working memory over time [
The circumplex model of affect suggests that all emotions can be distinguished in terms of varying levels of valence and arousal, with two distinct neural systems mediating the representation of affective states. It is addressed that common as well as distinct neural substrates underlie the regulation of different valences of emotion [
Another critical question remains to be answered is what is regulated. Reappraisal can be self- or situation-oriented [
The purpose of this study was to examine the neural mechanisms of self-related reappraisal in Chinese MDD outpatients. Block designs were employed to avoid naturally declining emotion processes when watching emotionally arousing pictures. Emotion control areas such as dorsal lateral prefrontal cortex (dlPFC) and ventral medial prefrontal cortex (vmPFC) were selected as regions of interest since these two areas seemed to be involved in pathogenesis of depression and influence the expression of depressive symptoms [
Twelve unmedicated major depressed outpatients and 15 normal controls were recruited. The patients were diagnosed through a structured clinical interview according to DSM-IV. For depressed patients, an inclusion criterion with current depressed episode was adopted, according to the DSM-IV. All patients were assessed with SDS, BDI, and HAMD before participating into experiment, with mild to moderate depression symptoms (HAMD ≥ 18; BDI ≥ 14; SDS ≥ 35). For healthy controls,
Demographic and clinical data of MDD and normal control groups.
Measure | MDD | Control | Sig. |
---|---|---|---|
Mean/SD |
Mean/SD | ||
Gender ratio (male/female) | 5/7 | 7/8 | 0.55 |
Age | 29.50 (8.46) | 25.80 (5.89) | 0.07 |
Years of education | 14.00 (3.77) | 14.80 (2.83) | 0.53 |
Handedness | right (12) | right (15) | — |
BDI | 26.17 (12.65) | 4.27 (4.23) | 0.00 |
SDS | 64.08 (12.60) | 36.54 (5.74) | 0.00 |
HAMD | 25.23 (4.97) | — |
Note: BDI: beck depression inventory; SDS: self-rating depression scale. Both groups were matched for age, sex ratio, and years of education.
The participants were trained on a computer during a previous session to get familiar with the reappraisal strategies they were to use during the scan. They were instructed to either attend to the visual stimuli or reappraise (self-detached versus self-immersed) their emotion reactions to each picture. Tasks were performed in three consecutive sessions after acquisition of structural images. Emotional stimuli were selected from the International Affective Pictures System (IAPS) [
For the attend condition (session 1), subjects should simply view the picture without taking efforts to alter their emotion; for the self-immersed conditions (session 2), subjects should perceive each picture as real and engage themselves in the situation depicted, by imagining themselves or a loved one in the scene; for the self-detached conditions (session 3), subjects should view the situation as fake or unreal and detach themselves from the situation. The attend condition was set as the control condition. Participants were told not to close their eyes or direct eyes away from the pictures during each trial and be able to relax during the break after each trial (Figure
Sequence of a trials in attend/reappraisal task. Instructions were given prior to each block.
During the scanning, stimuli were projected onto a screen, reflected by a mirror in front of the subjects. The task was performed in three consecutive sessions (“maintain,” “detach,” and “immerse”), in the order of the last two sequentially counterbalanced across all subjects. The instruction for each condition was given at the beginning of each block. Each trial consisted of four components: fixation, induction or regulation, rating, and rest. A fixation cross was displayed for 2 s, and then an IAPS picture appeared for 8 s during which subjects either simply viewed or reappraised the picture, followed by an affect rating screen (1 = no intensity to 4 = very intense), and a black blank screen was shown for 8 s for relaxation. A four-point scale was chosen because it forces the subjects to make emotional judgments and was proved to be reliable for emotion discrimination in a previous study [
MRI data were collected on a Siemens 3T Allegra MRI scanner. A high-resolution T1-weighted 3D image (T1WI) was acquired, with slice thickness = 4 mm, field of view (FOV) = 240 × 240 × 240 mm3, and matrix =256 × 256 × 256. Functional images were obtained from 30 gradient-echo T2*-weighted slices (slice thickness = 4 mm) per volume. A single shot gradient-recalled echo-echo planar imaging (SS-GRE-EPI) sequence was used with a time repetition of 2000 ms, a flip angle of 90°, time echo of 30 ms, FOV of 240 × 240 mm2, matrix of 64 × 64, slice thickness of 4 mm, and slice interval of 0.8 mm. For coregistration, 176 sagittal whole-brain scans were collected by 3-D magnetization-prepared rapid gradient-echo imaging (MPRAGE), with TR = 1970 ms, TE = 3.93 ms, a flip angle = 15°, slice thickness = 1.70 mm, slice interval = 0.85 mm, FOV = 250 × 250 mm2, and a matrix = 448 × 512.
The emotional state ratings during the experiment were analyzed with PASW (Version 19, SPSS Inc., Chicago, IL, USA). A two-way ANOVA was conducted to analyze the effect of the emotional picture presentation (negative, neutral, positive) on emotional state in the viewing condition. A 2 × 3 × 2 repeated-measures ANOVA including the factors group (MDD, HC), valence (negative, positive), and condition (attend, self-detachment, self-immersion) was calculated to illuminate the effects of regulation on emotional state. The neutral condition was neglected for the second analysis as there was no neutral picture in the reappraisal condition.
Data were preprocessed and statistically analyzed with SPM 8 (
We then performed region of interest (ROI) analyses upon a priori region of interest implicated in emotion reactivity and regulation (bilateral dlPFC, vmPFC, and amygdala). If vmPFC and dlPFC are critical neural substrates for pathogenesis of depression, then damage to either area should affect the expression of depressive symptoms. We used anatomical masks based on the Talairach daemon database, defined by WFU Pick atlas software (version 3.0; ANSIR Laboratory, WFU School of Medicine, Winston-Salem, NorthCarolina), and set the threshold at
A one-way ANOVA was computed to analyze the effect of picture type (positive, neutral, negative) on induced emotional reactivity during the attending task. A 2 × 3 × 2 repeated-measures ANOVA was also conducted on factors including group (MDD, normal), condition (attend, self-detachment, self-immersion), and emotion (positive, negative) to examine the effects of cognitive reappraisal on emotional induction (Figure
Means and standard deviations of intensity ratings after affect processing during the scan.
Regarding priori regions, we found activations in left inferior frontal gyrus (IFG) (109 voxels in left IFG, peak at (−48, 20, −4)
(a) Group × reappraisal interaction: regions activated in “attend,” “detach,” and “immerse” conditions in both groups. (b) Percent signal change for the following contrasts: “attend/emotional,” “detach/emotional,” and “immerse/emotional”. Note: attd = attend; dec = detach; inc = immerse.
We then performed a two-sample
In patients, a regression analysis revealed that during detachment of positive emotion, downregulation of left amygdala negatively correlated with HAMD scores (
(a) Decreased activation in the left amygdala for patients during detachment of positive emotion. (b) The signal change of amygdala between the detachment and view conditions correlated negatively with HAMD scores (
We are specifically interested in amygdala-cortical interactions during reappraisal. The PPI analysis revealed that compared to healthy controls, patients showed significantly enhanced coactivation of left amygdala with right dlPFC (MFG), right vmPFC (anterior cingulate), and right inferior parietal lobule (IPL) (Table
PPI analysis of left amygdala seed for immersion/emotional condition in both groups.
Region of coactivation | Side | BA | Talairach coordinates |
|
||
---|---|---|---|---|---|---|
|
|
| ||||
Control > MDD | ||||||
Middle temporal gyrus | L | 39 | −52 | −70 | 30 | 2.18 |
Middle temporal gyrus | R | 21 | 60 | −18 | −24 | 3.26 |
Superior temporal gyrus | R | 38 | 44 | 18 | −34 | 3.05 |
Medial frontal gyrus | L | −12 | 56 | 8 | 2.89 | |
Control < MDD | ||||||
Insula | L | −40 | −10 | 6 | 5.01 | |
Middle frontal gyrus | R | 9 | 42 | 42 | 32 | 2.75 |
Superior frontal gyrus | R | 26 | 54 | 32 | 2.46 | |
Inferior parietal lobule | R | 40 | 42 | −32 | 62 | 2.34 |
Precuneus | R | 7 | 18 | −36 | 58 | 2.19 |
Anterior cingulate | R | 32 | 10 | 36 | −10 | 2.16 |
Precuneus | L | −6 | −44 | 50 | 1.97 |
The current study extends previous findings about neural correlates of reappraisal in MDD along a self-relatedness dimension and confirms the hypotheses that self-relatedness may differentially modulate neural circuits underlying reappraisal for MDD versus normal group, demonstrating inflexible amygdala reactivity and strengthening frontolimbic connection in MDD. Since these neural circuits are involved in the pathology of depression [
Behaviorally, IAPS stimuli significantly induced emotion in both groups (
Interestingly, Dillon’s study using valence ratings as emotion responses revealed comparable regulation outcome between groups in increasing emotion, which differs from current study. Moreover, post hoc contrasts of repeated-measures ANOVA demonstrate that the left amygdala of depressed patients is less activated when immersion strategy is adopted regardless of valence (see in Figure
In contrast to healthy controls, depressed individuals exhibit diminished activation in left IFG when detachment strategy is adopted and enhanced activation in left IFG when immersion technique is adopted (see in Figure
Our study observes that in depressed patients, left vmPFC is more strongly recruited for self-detachment from positive affect, while right vmPFC is more strongly activated for negative affect. Observations are similar concerning self-immersion strategy. Medial prefrontal cortex (MPFC) is central to neural models of depression [
We further assume that comparisons within each valence might suggest a possibility for hemispheric asymmetry of prefrontal cortex in MDD: left vmPFC is more involved in positive affect reappraisal and right vmPFC more in negative affect reappraisal. However, more conclusive results may rely on ANOVA including “hemisphere” as one of the factors, which is not considered in this paper because SPM 8 does not allow four-way ANOVA. It is in line with the consensus that positive/negative emotion is parallel to approach/avoidance motivation system [
The PPI analysis exhibited a strengthened coupling between left amygdala and right prefrontal cortex (including dlPFC and vmPFC) and right IPL in MDD when increasing emotion. In comparison, two-sample
A wealth of data suggests that amygdala, insula, and anterior temporal pole are responsible for separately mediating the cognitive, physiological, and experiential aspects of emotional responses, respectively [
In conclusion, depressed individuals tend to rely more on cognitive control brain networks and enhanced functional coupling between left amygdala and right prefrontal cortex when using reappraisal strategy accompanied by unrestrained self-related affective processing, which applies for both valence of emotion.
There may be several clinical implications of our findings. First, group differences in the ability to regulate emotion may represent a sign of vulnerability to depressed mood and depressive disorders under stress [
Until now, the effect of cross-culture variability on reappraisal strategies remains largely unknown. Our study may provide preliminary insights into relevant research. The linguistic nature of emotion regulation strategies may vary among different cultures. As previous findings suggest, emotion-regulation strategies may contribute to differences in emotional experience across western and east Asian cultures [
The present research has several limitations. First, stimuli of personal relevance may need to be adopted (such as autobiographical experiences) in further study. Reappraisal involves momentary relevance and meaning of current stimuli which may vary among individuals, and subject-specific stimuli according to personal relevance may maintain stimulus consistency within individuals [
Dorsal lateral prefrontal cortex
Inferior frontal gyrus
Ventral medial prefrontal cortex
Fourth edition of Diagnostic and Statistical Manual of Mental Disorders
Major depressive disorder
Cognitive behavioral therapy
Beck depression inventory
Self-rating depression scale
Hamilton rating scale for depression
International affective pictures system
Field of view
Time echo
Time repetition
Region of interests
Volume of interests
Psychophysiological interaction.
The authors declare that there is no conflict of interests regarding the publication of this paper.
This research was financially supported by National Natural Science Foundation of China (NSFC30970898). The authors gratefully acknowledge the valuable assistance of Professor Huaqing Meng (the First Affiliated Hospital of Chongqing Medical University, China) and Dr. Chenggang Jiang (Institute of Surgery Research of Daping Hospital, China) for patient recruitment, Dr. Jiuquan Zhang for professional advice of imaging data processing, and Chengju Liao, Jiawen Li, Yun Liu, and Liying Gan for helpful comments.