Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Jan 30.
Published in final edited form as: Psychiatry Res. 2017 Nov 8;271:82–90. doi: 10.1016/j.pscychresns.2017.11.002

Longitudinal effects of cognitive behavioral therapy for depression on the neural correlates of emotion regulation

Harry Rubin-Falcone a,b, Jochen Weber c, Ronit Kishon a, Kevin Ochsner c, Lauren Delaparte d, Bruce Doré e, Francesca Zanderigo a,b, Maria A Oquendo f, J John Mann a,b, Jeffrey M Miller a,b,*
PMCID: PMC5757508  NIHMSID: NIHMS919692  PMID: 29128142

Abstract

Cognitive behavioral therapy (CBT) is effective for a substantial minority of patients suffering from major depressive disorder (MDD), but its mechanism of action at the neural level is not known. As core techniques of CBT seek to enhance emotion regulation, we scanned 31 MDD participants prior to 14 sessions of CBT using functional magnetic resonance imaging (fMRI) and a task in which participants engaged in a voluntary emotion regulation strategy while recalling negative autobiographical memories. Eighteen healthy controls were also scanned. Twenty-three MDD participants completed post-treatment fMRI scanning, and 12 healthy volunteers completed repeat scanning without intervention. Better treatment outcome was associated with longitudinal enhancement of the emotion regulation-dependent BOLD contrast within subgenual anterior cingulate, medial prefrontal cortex, and lingual gyrus. Baseline emotion regulation-dependent BOLD contrast did not predict treatment outcome or differ between MDD and control groups. CBT response may be mediated by enhanced downregulation of neural activity during emotion regulation; brain regions identified overlap with those found using a similar task in a normative sample, and include regions related to self-referential and emotion processing. Future studies should seek to determine specificity of this downregulation to CBT, and evaluate it as a treatment target in MDD.

Keywords: Cognitive behavioral therapy, CBT, depression, MDD, functional MRI, emotion regulation

1. Introduction

Major depressive disorder (MDD) is predicted to become the leading cause of disability by 2030 (Mathers et al., 2008). While treatment with evidence-based cognitive-behavioral therapy (CBT) for depression produces remission in a minority of patients (DeRubeis et al., 2005; Elkin et al., 1989; Luty et al., 2007), with comparable efficacy to other first-line antidepressant treatments (DeRubeis et al., 1999; Dobson et al., 2008; Trivedi et al., 2006), these treatments leave many depressed patients with significant symptoms and impaired functioning despite vigorous treatment. Treatment selection for MDD involves trial-and-error, due to the lack of known clinically useful moderators of treatment outcome. Task-based functional magnetic resonance imaging (fMRI) has been used to investigate neural predictors of treatment outcome with CBT for MDD and to map longitudinal changes produced by CBT in depression (Chuang et al., 2016; Forbes et al., 2010; Franklin et al., 2016; Fu et al., 2008; Ritchey et al., 2011; Siegle et al., 2006; Thompson et al., 2015; Yoshimura et al., 2013), largely by examining the neural correlates of tasks that elicit negative affect.

Deficits in emotion regulation may contribute to depression risk (Hopfinger et al., 2016) and psychopathology. In fMRI studies of emotion regulation, differences in blood-oxygen-level dependent (BOLD) activity and connectivity have been identified in depression (Heller et al., 2009; Johnstone et al., 2007). A key goal of CBT for depression is to develop improved capacities for voluntary regulation of emotion, accomplished through techniques including cognitive restructuring, behavioral activation, and behavioral experiements (Beck, 1995). Indeed, recent evidence suggests that both in-person inpatient CBT for depression (Forkmann et al., 2014) as well a computer-based CBT intervention (Morris et al., 2015) lead to improvements in a form of emotion regulation, cognitive reappraisal, and that improvement in reappraisal correlates with reduction in depression severity (Forkmann et al., 2014). We therefore sought to examine the neural correlates of emotion regulation as a predictor of treatment outcome with CBT for depression, and to examine the relationship of longitudinal changes in such emotion regulation-related activity to treatment outcome.

Negative autobiographical memories are particularly salient stimuli for emotion regulation tasks in depression due to their personal relevance, in contrast to other standardized stimuli. Kross et al. developed a fMRI paradigm in which participants recall a series of emotionally negative autobiographical memories and are instructed to either respond naturally (“feel” condition), or regulate their emotional responses in two other conditions (“analyze” and “accept” conditions) (Kross et al., 2009). The analyze strategy was designed as a memory analog of cognitive reappraisal strategies used in previous fMRI studies (Ochsner et al., 2002; Ochsner et al., 2004; Ray et al., 2005; Wager et al., 2008). During both regulation strategies, healthy volunteers showed reduced activity in brain regions associated with self-referential and affective processing, including subgenual anterior cingulate cortex, medial prefrontal cortex, and lingual gyrus.

We used a modified version of this task to examine the neural correlates of voluntary emotion regulation in a group of individuals with current MDD, both before and after a course of CBT for depression. As the “analyze” condition maps closely onto the core CBT technique of cognitive restructuring, it was selected as the regulation strategy for this study (Beck, 1995). A comparison group of healthy volunteers performed this fMRI task at comparable time-points without a treatment intervention. Given the observed suppression of activity during emotion regulation previously observed in a normal sample, we hypothesized: 1) the magnitude of emotion regulation-dependent BOLD signal reduction (i.e. “feel” > “analyze” contrast) would increase longitudinally as a function of better treatment outcome with CBT; 2) the magnitude of emotion regulation-dependent BOLD reduction at baseline would be correlated with better treatment outcome; and 3) healthy volunteers would show greater emotion regulation-dependent BOLD reduction in relevant brain regions than MDD participants. We conducted whole-brain voxelwise analyses as well as analyses at the region-of-interest (ROI) level, examining regions associated with emotion regulation of negative autobiographical memories in previous work (subgenual cingulate, dorsomedial prefrontal cortex, and lingual gyrus) (Kross et al., 2009).

2. Methods

2.1 Sample

Participants were recruited using online and print advertisements as well as through referrals from surrounding clinics and gave written informed consent prior to research participation. MDD inclusion criteria included: 1) Age 18–60; 2) A DSM-IV diagnosis of MDD in a current major depressive episode (MDE) as assessed using the Structured Clinical Interview (SCID) for DSM-IV (First et al., 1995); 3) 17-item Hamilton Rating Scale for Depression (HRSD) score ≥ 16 (Hamilton, 1960); 4) Lack of significant benefit from any current psychiatric medications and ability to tolerate washout (if applicable); 5) Capacity to provide informed consent. MDD exclusion criteria included: 1) Unstable medical conditions; 2) alcohol or substance use disorder within the past 6 months; 3) Other current or past major psychiatric disorders including bipolar disorder; comorbid anxiety and personality disorders were allowed and are described in Table 1; 4) Pregnancy, currently lactating, planning to conceive during the course of study participation or abortion in the past two months; 5) Dementia; 6) A neurological disease or prior head trauma with evidence of cognitive impairment; 7) A first-degree family history of schizophrenia if the subject is less than 33 years old (to exclude possible prodromal phase of schizophrenia); 8) Contraindication to CBT as primary treatment for depression, including prior non-response to an adequate trial of CBT, active psychosis, or severe suicidal ideation including a plan. Healthy control inclusion criteria included: 1) Age 18–60; 2) Lack of current or past DSM-IV Axis-I diagnosis as assessed by the SCID; 3) Capacity to provide informed consent. Healthy control exclusion criteria included items 1,2,4,5 from the MDD exclusion criteria plus first-degree relative with history of major depression, schizophrenia, schizoaffective disorder, or suicide attempt, or two or more first-degree relatives with a history of substance dependence.

Table 1.

Demographic and clinical data

Controls (N=18) MDD (N=31) p-value (Control vs. MDD, 2-tailed t-test)
Age 33±10.3 34.2±10.2 0.7
Initial Hamilton Depression Severity (17-Item) 1±1.5 19.1±4.4 <0.001
Initial Beck Depression Inventory 0.4±1.7 28±7.7 <0.001
Final Hamilton Depression Severity (17-Item) 0.4±0.9 12.8±6.5 <0.001
Final Beck Depression Inventory 1.1±2.6 15.5±8.4 <0.001
Brown-Goodwin Aggression Score 13.3±3.1 15.5±3 0.03
Years of Education 14.9±1.9 16.5±2.6 0.03
Age at Onset n/a 16±15.4
Number of Previous Depressive Episodes n/a 6.9±25.8
Length of Current Major Depressive Episode (weeks) n/a 161.3±223.5
Categorical Variables N (%) p-value (Control vs. MDD, Fisher’s Exact)
Female 11 (61) 19 (55) 1
Scanned at Cornell 5 (28) 11 (35) 0.75
Prior Exposure to Anti-Depressants n/a 17 (62)
Suicidal Ideation Present n/a 8 (27)
First Degree Relative With Major
Depression n/a 12 (44)
Past Alchohol or Canabis Abuse n/a 4 (14)
Comorbid Anxiety Disorder n/a 12 (39)
Comorbid Personality Disorder n/a 10 (32)
Comorbid Dysthymia n/a 1 (3)
Race/Ethnicity
Asian 2 (11) 3 (10)
African American 5 (28) 3 (10)
Caucasian 8 (44) 17 (55)
Hispanic 2 (11) 3 (10)
>1 Race 0 (0) 4 (13)
Open in a new tab

Table 4.

Average in-scanner valence ratings per instruction (rated on a 4 point scale) and valence differences between groups and time points (t, p): Table 5: MDD group in-scanner valence ratings correlations with treatment outcome (r, p).

Average valence (4 point scale)
Analyze Feel Feel-Analyze
Controls Time1 1.91±0.54 2.52±0.64 0.62±0.54
Controls Time2 1.87±0.49 2.54±0.64 0.67±0.60
MDD Time1 2.28±0.51 2.93±0.51 0.65±0.48
MDD Time2 2.09±0.60 2.57±0.61 0.48±0.45
Valence contrasts (t,p)
Controls vs MDD Time1 −2.48, 0.02* −2.50, 0.02* −0.26, 0.79
Controls vs MDD Time2 −1.07, 0.29 −0.13, 0.9 1.06, 0.3
Controls Time1 vs Time2 0.00, 1 −0.15, 0.88 −0.15, 0.88
MDD Time1 vs Time2 1.70, 0.1 2.30, 0.03* 0.72, 0.47
Open in a new tab

2.2 Clinical procedures and treatment

BDI (primary, (Beck et al., 1961)) and 17-item HRDS (secondary, (Hamilton, 1960)) scores were used as measures of pre- and post-treatment depression severity (BDI scores obtained at every session, HRDS scores every fourth session). Following baseline MRI scanning, 14 sessions of CBT for depression were administered over 12 weeks according to a treatment manual (Beck, 1979). Core techniques employed included cognitive-restructuring through the use of dysfunctional thought records; behavioral activation following initial activity monitoring approaches; behavioral experiments as a means to examine negative automatic predictions; and some work to identify and modify more deeply held patterns of negative thinking about oneself, one’s life, and one’s future (“intermediate beliefs” and “core beliefs.” Forty-five-minute sessions occurred as close as possible to twice-weekly for two weeks, then weekly thereafter. Study therapists were M.D.- or Ph.D.-level therapists with extensive training and experience conducting CBT, including training at the Beck Institute for Cognitive Behavioral Therapy. Therapists met weekly for peer supervision. Sessions were audiotaped, and adherence to CBT principles was assessed on at least one session per patient by the Beck Institute using the Cognitive Therapy Rating Scale (mean CTRS = 42.7±5.9) (Young and Beck, 1980). A post-treatment MRI scan was performed at the conclusion of CBT. For healthy volunteers, the fMRI task was repeated at a comparable time-point without a treatment intervention.

Participant flow through the study is depicted in Figure 1. For the participants that did not complete 14 sessions of CBT, last observation carried forward was applied, using the last BDI measurement obtained before dropout or medication augmentation. Two participants who did not complete CBT monotherapy due to clinical worsening, for whom medication was added to augment CBT, had timepoint 2 scans performed immediately prior to medication augmentation (following sessions 9 and 11 respectively); other CBT non-completers were only included in analyses related to time 1 MRI data. Of the 31 depressed participants included in the analysis, 14 were antidepressant-naive, 15 were unmedicated at enrollment but had a history of prior antidepressant use, and 2 were on an ineffective antidepressant medication at enrollment and completed a 3-week washout prior to scanning and treatment.

Figure 1.

Figure 1

Open in a new tab

Participant flow-chart.

2.3 Image acquisition

MRI scans were acquired on two 3T SignaHDx scanners (General Electric Medical Systems, Milwaukee, WI), one at The New York State Psychiatric Institute and one at Weill Cornell Medical College, using the same 8-channel head coil. Scan site was included as a covariate in all analyses. T1-weighted MRI scans were acquired using the following parameters: TR = ~6 ms, TE = minimum 2400 ms, flip angle = 8, FOV = 25.6×25.6 cm, slice thickness = 1 mm, number of slices = 178, matrix size = 256×256 pixels. For functional scanning during the memories task, an Echo Planar Imaging (EPI) acquisition was obtained for each of four runs using the following parameters: TR = 2000 ms, TE = 26 ms, flip angle = 77, FOV = 22.4×22.4 cm, slice thickness = 3.5 mm, spacing = 3.5 mm, number of slices = 32, matrix size= 64×64 pixels, number of volumes = 115.

2.4 fMRI task design and training

This task was similar to that in Kross et al. 2009 (Kross et al., 2009). During a semi-structured interview prior to the time 1 scan, participants identified 12 upsetting memories from their lives during which they felt sad, hopeless, unlovable, incompetent, ashamed, humiliated, or neglected. Participants generated a cue phrase for each memory, which served as a prompt during the fMRI task. Six memories were used during each scan; if participants could not identify 12 memories at the baseline interview, a subsequent interview was conducted prior to the time 2 scan. Participants also identified 3 additional practice memories, of a different emotional type than used during scanning (anger, fear, or disgust), for use during task training.

On the day of the scan, participants were reminded of the negative autobiographical memories they generated during screening: each cue was randomly presented on screen and participants were instructed to press a button as soon as they could recall the memory to which it referred. Participants were then trained in the “feel” and “analyze” strategies by a research staff member using a script-driven training including examples with a hypothetical memory. For “feel” trials, participants were instructed to focus on the specific feelings they experienced while recalling event. For “analyze” trials, participants were instructed to identify their thoughts and feelings about the event, and to re-consider the situation from a more realistic, rational, and positive perspective. These techniques were then practiced with 3 practice memories. They then practiced the analyze and feel strategy on 3 practice memories they provided for this purpose. The staff member asked the participant what alternative cognitive responses they would use to reduce their emotional reaction to each memory, and if they had difficulty identifying adaptive, effective alternative responses, the research staff member provided coaching to them to provide additional training.

The task consisted of 3 blocks of memory presentations that alternated with an active baseline condition. Each block consisted of 2 “analyze” trials and 1 “feel” trial or vice-versa, with the sequence of trials counterbalanced across blocks. Trials consisted of a 2-second fixation followed by a 10-second presentation of the cue phrase to a memory, during which participants were instructed to recall the autobiographical memory that the cue indicated, followed by 3-second fixation, after which the cue phrase re-appeared with a strategy instruction (“feel” or “analyze”) for 20 seconds. Participants then rated how emotionally negative they felt and how vivid was their recall of the memory on a 4-point scale, with question order counterbalanced within runs. Each question was shown until the participant responded, up to 3 seconds. Each memory appeared twice during the task, once during block 1 or 2, and once again during block 3 or 4, with alternating instructions with each presentation. Participants then completed an active baseline task: a 30-second spatial perception task during which they observed arrows pointing left or right and were asked to indicate the direction of the arrow. This task was designed to minimally engage the regulatory, memory, and emotional processes of interest (Stark and Squire, 2001).

2.5 Image processing

2.5.1 Pre-processing

The fMRI task data were processed using FEAT (FMRI Expert Analysis Tool) Version 5.98, which is part of FSL (FMRIB’s Software Library, www.fmrib.ox.ac.uk/fsl) (Woolrich et al., 2009). Standard pre-processing, including motion outlier removal and non-linear co-registration was performed; a detailed summary of pre-processing methods can be found in Appendix A.

2.5.2 fMRI statistical analysis

For each participant, we used a general linear model (GLM) to identify brain activity associated with the period during which the subject was processing memories using either the “feel” or “analyze” strategy. The GLM was convolved with a double gamma hemodynamic response function for 5 epochs consisting of: the “feel” and “analyze” cues, the initial memory recall cues, the valence question periods, and the vividness question periods. Rather than using periods of unmodeled fixation (which allow for unconstrained cognitive activity) as an implicit model baseline, we used periods of an active perceptual task (the arrows task) as an umodeled implicit baseline. This was done in order to minimize the occurrence of memory recall outside of the cue periods.

F-Tests were performed to identify clusters in which “feel” signal was greater than “analyze” and vice-versa. Each participant’s four runs were combined using a Fixed Effects GLM approach in order to consider the average across all runs.

To identify brain regions where the differences in response between “analyze” and “feel” differed between depressed participants and controls, an F-Test was performed between the “feel” > “analyze” images from time point 1 of the two groups. The average within each diagnostic group was also examined. In order to identify clusters where the “feel” > “analyze” contrast was associated with current depression severity, baseline BDI score was regressed onto the “feel” > “analyze” contrast at time 1. In order to identify clusters where the time 1 “feel” > “analyze” contrast predicted treatment outcome, final BDI was regressed onto the “feel” > “analyze” contrast at time 1, while covarying for baseline BDI. In order to identify clusters in which change in the “feel” > “analyze” contrast scaled with clinical improvement, F-Tests were performed between Time1 and Time2 “feel” > “analyze” parameter estimates and then final BDI score was regressed onto the result, while covarying for baseline BDI. Analyses were repeated using HDRS score as a secondary outcome measure (see Appendix B).

Higher level analysis was carried out using FLAME (FMRIB’s Local Analysis of Mixed Effects) stage 1 and stage 2 (Woolrich, 2008). Clusters were identified using a minimum z-score of 3.1 and a corrected cluster significance threshold of p < 0.05 (Worsley, 2001). The voxel-wise threshold of 3.1 was chosen in order to prevent false positives, as recent critiques of cluster-based thresholding have shown false positives to be common when less stringent thresholds are applied to null data (Eklund et al., 2016; Woo et al., 2014). Scan site was included as a nuisance regressor for each analysis.

2.6 ROI analysis

In order to examine the relationship between treatment outcome and emotion regulation-dependent BOLD suppression within the centers of emotion regulation identified in the paradigm-originating Kross et al. study (Kross et al., 2009), a region of interest (ROI) analysis was performed. ROIs considered were 3 spheres with centers corresponding to the emotion-regulation activation peaks by Kross et al. (lingual gyrus: −12, −43, −2; subgenual anterior cingulate (sgACC): 0,14,−5; and medial prefrontal cortex (mPFC): 6,59,7), with radii of 8 mm (all coordinates in talairach space). Mean parameter estimates for the emotion regulation task were extracted within each sphere for each subject at time 1, and from the time 1 vs time 2 F-test. Correlations between the values at time 1 and final BDI score were performed to look for a prediction effect, and time 1 vs. time 2 values were correlated with final BDI to look for longitudinal treatment effects. Scan site and baseline BDI were included as co-variates in these analyses.

3. Results

3.1 Clinical characteristics and treatment outcome

MDD participants were moderately depressed (mean BDI = 28.7±7.7). After treatment, mean BDI scores were 15.5±8.4 in the full MDD sample (40±35% improvement), and 12.5±6.9 in the 22 completers (51±33% improvement). Remission rate (final BDI<=10) was 32% in intent-to-treat (ITT) sample and 41% in completers; response rate (reduction in BDI > =50%) was 45% in ITT sample and 59% in completers. Clinical and demographic data are described in Table 1.

3.2 fMRI

All analyses were performed looking at both tails of our contrast of interest (emotion regulation: “feel” > “analyze”, and its inverse: “analyze” > “feel”). The inverse contrast did not result in whole-brain or ROI-level findings for any analysis; “emotion regulation-dependent BOLD reduction” below refers to the “feel” > “analyze” contrast.

3.2.1 Main effect of emotion regulation (within-group)

At time 1, MDD participants had greater BOLD signal during the “feel” condition compared to the “analyze” condition in bilateral insula and right supramarginal gyrus (three clusters, see Figure 3 and Table 2). No clusters were observed at the a priori threshold in the healthy volunteer group.

Figure 3.

Figure 3

Open in a new tab

Regions demonstrating significant reduction of emotion regulation-dependent BOLD signal (“Feel” > “Analyze” contrast) in the depressed group at time 1 (voxel-wise z > 3.1, cluster-wise FWE-corrected p < 0.05). Significant clusters identified in bilateral insula and right supramarginal gyrus.

Table 2.

Clusters demonstrating significant emotion regulation-dependent BOLD reduction (“feel” > “analyze”) within MDD group. All coordinates in MNI space, mm.

Region Cluster Size (voxels) Corrected p-value Maximum z-stat Coordinates of Maximum Center of Gravity
Right supramarginal gyrus 736 0.00518 4.88 (62, −36, 26) (58.4, −40.1, 19.5)
Left insula 578 0.0132 4.24 (−36, 12, 4) (−36, 7.2, 5.32)
Right insula 494 0.0224 4.65 (60, 0, 0) (39.1, 7.61, 1.27)
Open in a new tab

3.2.2 Effect of diagnosis (MDD vs. Controls)

Emotion regulation-dependent reduction of BOLD signal did not differ between MDD and control groups at the a-priori threshold.

3.2.3 Effect of depression severity within MDD

Emotion regulation-dependent reduction of BOLD signal at time 1 was not correlated with current depression severity (BDI-score) at time 1 at the a priori threshold.

3.2.4 Baseline emotion regulation signal and treatment outcome

Whole brain voxelwise analysis examining the relationship of pre-treatment emotion regulation-dependent reduction of BOLD signal to subsequent treatment outcome with CBT in the MDD group yielded no significant clusters at our a priori thresholds (voxel-level z >3.1, corrected cluster level p < 0.05). There were no significant correlations between time 1 emotion regulation-dependent reduction of BOLD signal and treatment outcome at the ROI level (p > 0.75, all ROIs).

3.2.5 Longitudinal change in emotion regulation contrast as a function of treatment outcome

Change in emotion regulation-dependent reduction of BOLD signal from time 1 to time 2 for MDD participants was correlated with treatment outcome (defined as final BDI score while controlling for time 1 BDI), in a whole-brain analysis at a priori statistical threshhold (voxel-wise z > 3.1, cluster-wise FWE-corrected p < 0.05) in medial frontal pole, left subgenual cingulate, lingual gyrus/cerebellum, left precentral/supramarginal gyrus, and left putamen (4 clusters, Figure 4 and Table 3). Greater emotion regulation-dependent reduction of BOLD signal in these clusters at time 2 than at time 1 was associated with better treatment outcome. Examination of this contrast using a secondary measure of treatment outcome, the HDRS, revealed activation in one of these regions, the lingual gyrus (details in supplement including figure B1 and table B1). At the ROI level, treatment outcome was correlated with change in emotion regulation-dependent reduction of BOLD signal in the same direction as voxelwise findings in mPFC ROI (r = −0.4397, p = 0.0344), and at a trend in the lingual gyrus ROI (r = −0.374, p = 0.077). Correlation was in the same direction but was not significant in sgACC ROI (r = −0.317, p = 0.139); this sgACC ROI was medial to the cluster identified in the whole-brain analysis above.

Figure 4.

Figure 4

Open in a new tab

Regions where greater emotion regulation-dependent BOLD reduction at time 2 compared to time 1 was associated with better treatment outcome (lower final BDI score while covarying for baseline BDI score; voxel-wise z > 3.1, cluster-wise FWE-corrected p < 0.05). Clusters included lingual gyrus, left precentral gyrus/putamen, left medial frontal pole, left subgenual cingulate and left supramarginal gyrus.

Table 3.

Clusters where greater emotion regulation-dependent BOLD reduction at time 2 compared to time 1 are associated with better treatment outcome (lower final BDI score while covarying for baseline BDI score; voxel-wise z > 3.1, cluster-wise FWE-corrected p < 0.05). All coordinates in MNI space, mm.

Regions Cluster Size (voxels) Corrected p-value Maximum z-stat Coordinates of Maximum Center of Gravity
Lingual gyrus/cerebellum 1482 0.00027 4.5 (20, −60, −24) (7.97, −67.9, −15.2)
Left precentral gyrus/putamen 1405 0.000372 4.59 (−56, −2, 10) (−39.2 −18, 13.3)
Left medial frontal pole/subgenual cingulate 1008 0.00217 4.8 (2, 54, −12) (−10.3, 42, −14.1)
Left supramarginal gyrus 524 0.0265 4.73 (−60, −24, 10) (−59.8, −27.1, 10.4)
Open in a new tab

We did not observe a main effect of time on emotion regulation-dependent reduction of BOLD signal (time 1 vs. time 2) in either the healthy volunteers, who received no intervention, or in the MDD group; longitudinal effects in the MDD group were only observed when treatment outcome was considered.

3.3 Behavioral data

During baseline scanning, participants in both groups rated memories during the “feel” instruction as more emotionally negative than those during the “analyze” instruction (MDD: t = −7.986, p <0.001; controls: t = −4.95, p <0.001); MDD participants also rated emotions recalled during the “feel” prompt as more negative at time 1 compared to time 2 (t = 2.30, p = 0.03) (see Figure 5). Detailed behavioral data are presented in Table 4. We calculated a behavioral measure of emotion regulation success (BERS) for each participant, measured as the difference between mean emotional valence ratings during all “feel” condition trials vs. all “analyze” condition trials within scan. BERS did not differ significantly between MDD participants and controls at time 1, nor did it change significantly from time 1 to time 2 in either group. In the MDD group, neither baseline BERS nor change in BERS were associated with treatment outcome (all analyses p ≥ 0.47). Data are summarized in Tables 4 and 5.

Figure 5.

Figure 5

Open in a new tab

Within-scanner valence ratings during “feel” and “analyze” conditions at time 1 and time 2 for depressed participants; “feel” at time 2 was rated significantly less upsetting than at time 1 (p = 0.027), and “analyze” was rated as less upsetting than “feel” at both time 1 and time 2 (p < 0.001 for both).

Table 5.

MDD group in-scanner valence ratings correlations with treatment outcome (r, p).

Time1 valence correlations with: Time2 valence correlations with: Time1-Time2 Valence with:
Contrast Time1 BDI Change in BDI (Time1 – Time2) Time2 BDI Change in BDI (Time1 – Time2) Change in BDI (Time1 – Time2)
Analyze −0.016, 0.93 0.098, 0.60 −0.60, .003* −0.47, 0.023* 0.40, 0.06
Feel 0.13, 0.50 0.053, 0.78 −0.65, .001* −0.43, 0.039* 0.33, 0.13
Feel-Analyze 0.072, 0.70 −0.047, 0.80 −0.076, .73 0.046, 0.83 −0.033, 0.88
Open in a new tab

4. Discussion

4.1 Primary findings

The primary finding of this study is that emotion regulation-dependent reduction in BOLD signal is enhanced longitudinally as a function of clinical improvement with CBT for depression. This occurs in brain regions relevant to the cognitive control of emotion. While replication and extension in a placebo-controlled study would be necessary to confirm the specificity of these findings to CBT, these data support the possibility that enhancement of the neural correlates of emotion regulation may mediate the effects of CBT on depression severity. This finding appears to be associated with clinical improvement specifically, and not necessarily with general treatment effects, as we observed no mean change in the emotion regulation contrast from pre- to post-treatment at the group level in the MDD group irrespective of treatment outcome.

The longitudinal changes in emotion regulation-dependent BOLD contrast that correlate with treatment outcome may represent a change in self-referential processing of negative autobiographical memories, or in processing of the negative meaning of memories. A recent meta-analysis found self referential processing to be associated with brain activation in cortical midline structures, including medial orbitofrontal cortex and anterior cingulate (Nejad et al., 2013). We found that greater emotion regulation-dependent reduction of BOLD signal in these regions longitudinally to be associated with better treatment outcome. This appeared to be driven by reductions in BOLD signal during the analyze condition, rather than by increases during the feel condition (see Appendix C). Rostral mPFC is also involved in elaborating the negative meaning of stimuli to increase negative responses (Ochsner et al., 2009), and with attention to and awareness of emotion (Satpute et al., 2013). Moreover, its connectivity with amygdala and insula increase in the context of labeling a moderately aversive state as “bad” as opposed to “neutral” (Satpute et al., 2016). An alternative interpretation of this finding is that effective CBT allows individuals to better modulate these mPFC-dependent emotional elaboration/labeling processes.

Our findings are partially consistent with previous work examining longitudinal changes in BOLD fMRI responses following CBT using other task designs (Fu et al., 2008). A 2015 review looking at resting state fMRI data pre- and post- CBT reported that changes were most commonly observed in some of the same brain regions observed in our study, including anterior cingulate cortex (ACC) and ventromedial prefrontal cortex/orbitofrontal cortex (VMPFC/OFC) (Franklin et al., 2016). One study using a reward task showed that a reduction in clinical symptoms following CBT was accompanied by decreases in activation within bilateral sgACC (Straub et al., 2015). An fMRI study examining BOLD responses to visual stimuli of faces with parametrically-modulated levels of sadness identified longitudinal reductions in lingual gyrus activity to negative valence following CBT, in the same direction as our primary finding (Fu et al., 2008). This suggests a convergence of changes in BOLD-fMRI signal following CBT across a range of tasks.

Kross and colleagues’ original work using this autobiographical memory emotion regulation task in healthy volunteers found differences in activation between the “feel” and “analyze” conditions in lingual gyrus, subgenual anterior cingulate, and anterior medial prefrontal cortex (Kross et al., 2009). We found that changes in treatment outcome were associated with changes in emotion regulation activation in overlapping parts of lingual gyrus and mPFC and adjacent (more lateral) left subgenual cingulate in whole-brain voxelwise analysis, and in mPFC, with a trend in lingual gyrus, in ROI analyses. This implies that therapeutic effects of CBT for depression might be mediated by an enhancement of the neural correlates of emotion regulation observed in healthy individuals. Differences in MRI acquisition and analysis, as well as uncertainty associated with reported activations, may have also contributed to localization differences.

We found that changes in treatment outcome as assessed by HDRS-17 scores were also associated with changes in emotion regulation activation in lingual gyrus (see Appendix B), although not in other regions. The lingual gyrus is therefore strongly implicated in this analysis, as it was correlated with both treatment outcome measures despite their differences (patient vs therapist assessed, differential weights of mood vs. physical symptoms).

4.2 Main effects of emotion regulation

At baseline, MDD participants exhibited decreased activity in the “analyze” as compared to the “feel” condition in bilateral anterior insula. This finding is consistent with work identifying decreased insula activity during reappraisal in healthy volunteers (Goldin et al., 2008; Holland and Kensinger, 2013). In contrast, a meta-analysis examining fMRI studies of emotional reappraisal found increased activity during reappraisal in left anterior insula (Buhle et al., 2014). However, the majority of the task designs from the studies included in that meta-analysis used reappraisal to images rather than memories, and were performed in healthy volunteers. Regulation of anterior insula activity during reappraisal is not unexpected given its role in emotion processing (Gasquoine, 2014), of adjacent regions in cognitive control (Wager and Barrett, 2017), and of pre-stimulus activity in insula predicting subsequent emotion regulation success (Denny et al., 2014). Pre-treatment resting glucose metabolism in anterior insula was identified as a predictor of differential outcome to CBT vs. escitalopram in a recent randomized clinical trial (McGrath et al., 2013), although we did not observe prediction effects related to insula activity in the current study.

4.3 Other findings

We did not observe differences in emotion regulation-dependent BOLD contrast between the depressed and control groups, nor did we observe a main effect of emotion regulation within the control group. This may be partly explained by limitations in statistical power due to sample size of the control group (n = 18), and to the less aversive memories elicited by the control group. While some studies have identified emotion regulation deficits in depression at the behavioral level (Gotlib and Joormann, 2010; Kircanski et al., 2012), a recent fMRI study has suggested intact emotion regulation-related activity in depressed individuals (Dillon and Pizzagalli, 2013). Similarly, pre-treatment emotion regulation-related activity did not predict treatment outcome with CBT in the MDD group. Although previous studies have been able to predict CBT outcome with fMRI (Costafreda et al., 2009; Siegle et al., 2006; Thompson et al., 2015), they generally used tasks related to emotional reactivity, and not regulation. These findings suggest that while emotion regulation-related activity may be intact in MDD, enhancement of this activity via CBT may contribute to clinical improvement.

4.4 Study limitations

The primary limitations of the study were the lack of randomization to different treatment conditions, including a placebo condition, and the modest sample size, particularly in the healthy volunteer group. In addition, our sample included patients with comorbid anxiety disorders (N = 12) and personality disorders (N = 10, five of whom also had anxiety disorders), so we cannot be sure of our finding’s specificity to MDD. Replication in a larger sample and with randomization to different treatment conditions will be crucial in future studies to evaluate the specificity of our findings to CBT.

Although the relationship between treatment outcome and change in emotion regulation-related BOLD suppression may represent a change in self-referential processing of autobiographical memories, this could not be examined more closely in our study because ratings of self-referential thinking were not acquired. Future work could acquire longitudinal measures of self-referential processing (using a scale such as the Ruminative Response Scale (Treynor et al., 2003)) to relate to fMRI data and clinical outcome.

The stimuli for the task used here have high levels of face validity, as they are highly relevant to the participant and potentially to their depression. However, they are not standardized across participants, and use of novel stimuli may also have a role in testing cognitive reappraisal.

It is well known that MDD patients tend to over-generalize negative memories (Williams et al., 2007). It is possible that the recall of memories could have differed between MDD and controls groups for this reason. Of note, depressed participants and healthy volunteers were similarly able to associate each memory with a specific date. The primary findings of this study are related to treatment effects within the depressed group alone; however, an overgeneralization effect in the depressed sample could be a possible confound in the analyses comparing diagnostic groups.

Given our strong prior interest in the three regions used for ROI analyses, we believe that a finding (in mPFC) that is not corrected for multiple comparisons across the ROIs (using a Bonferroni factor of 3) constitutes a significant finding. Replication of this finding in an independent sample is required.

4.5 Future directions

We found that longitudinal increases in emotion regulation BOLD contrast in several anatomically distant regions was associated with treatment outcome with CBT. Further work examining functional connectivity may identify functional networks whose synchronous activity may relate to treatment outcome in CBT.

Our study examined a form of emotion regulation seeking to decrease emotionally negative responses to aversive memories, but given fMRI findings in depression related to upregulation of positive affect (Johnstone et al., 2007), it would also be interesting to examine this form of emotion regulation and its relationship to CBT outcome.

Combination of task-based fMRI with other biomarkers (Fujino et al., 2015; Harkness et al., 2012; Lueken et al., 2015; Nemeroff et al., 2003; Stiles-Shields et al., 2015) may enhance power to predict treatment outcome and understand longitudinal effects of treatment.

Finally, if reduced engagement of medial prefrontal cortex, anterior cingulate, and lingual gyrus during conscious regulation of emotion is validated as a mediator of outcome with CBT for MDD, novel treatment approaches may be considered to enhance neuroplasticity in these regions during CBT, using pharmacological or brain stimulation interventions (Bajbouj and Padberg, 2014; Herrmann et al., 2017).

4.6. Conclusions

Clinical improvement following CBT for depression is associated with longitudinal changes in emotion regulation-dependent BOLD contrast, in regions including anterior cingulate, medial orbitofrontal cortex, and lingual gyrus. This suggests a possible neural mediator of CBT effects in MDD, related to conscious regulation of affective and self-relevance processing of aversive memories.

Supplementary Material

supplement
NIHMS919692-supplement.docx (598.4KB, docx)

Figure 2.

Figure 2

Open in a new tab

Schematic of emotion regulation fMRI task design.

Highlights.

  • We used an emotion regulation fMRI task to study cognitive behavioral therapy

  • Participants with major depression scanned pre- and post-treatment

  • Treatment outcome scales with changes in emotion regulation-dependent BOLD signal

  • Regions identified have been associated with emotion regulation in normative sample

  • The findings suggest a possible neural target for future treatment interventions

Acknowledgments

Funding

NIMH (K08MH085061)

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Financial disclosures:

Dr. Miller’s family previously owned stock in Johnson & Johnson, unrelated to the current manuscript.

Dr. Oquendo receives royalties for the commercial use of the Columbia Suicide Severity Rating Scale. Her family owns stock in Bristol Myers Squibb.

Dr. Ochsner has no conflict of interest to declare.

Dr. Zanderigo has no conflict of interest to declare.

Dr. Kishon has no conflicts of interest to declare.

Ms. Delaparte has no conflicts of interest to declare.

Mr. Rubin-Falcone has no conflict of interest to declare.

Mr. Weber has no conflict of interest to declare.

Dr. Doré has no conflict of interest to declare.

Dr. Mann receives royalties from the Research Foundation for Mental Hygiene for commercial use of the C-SSRS.

References

  1. Bajbouj M, Padberg F. A perfect match: noninvasive brain stimulation and psychotherapy. European archives of psychiatry and clinical neuroscience. 2014;264(Suppl 1):S27–33. doi: 10.1007/s00406-014-0540-6. [DOI] [PubMed] [Google Scholar]
  2. Beck AT. Cognitive therapy of depression. Guilford Press; New York: 1979. [Google Scholar]
  3. Beck AT, Ward CH, Mendelson M, Mock J, Erbauh J. An inventory for measuring depression. Arch Gen Psychiatry. 1961;4:53–63. doi: 10.1001/archpsyc.1961.01710120031004. [DOI] [PubMed] [Google Scholar]
  4. Beck JS. Cognitive therapy: basics and beyond. Guilford Press; New York: 1995. [Google Scholar]
  5. Buhle JT, Silvers JA, Wager TD, Lopez R, Onyemekwu C, Kober H, et al. Cognitive reappraisal of emotion: a meta-analysis of human neuroimaging studies. Cerebral cortex. 2014;24:2981–2990. doi: 10.1093/cercor/bht154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chuang JY, K JW, Murray GK, Elliott R, Hagan CC, Graham JM, et al. Aberrant brain responses to emotionally valent words is normalised after cognitive behavioural therapy in female depressed adolescents. Journal of affective disorders. 2016;189:54–61. doi: 10.1016/j.jad.2015.09.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Costafreda SG, Khanna A, Mourao-Miranda J, Fu CH. Neural correlates of sad faces predict clinical remission to cognitive behavioural therapy in depression. Neuroreport. 2009;20:637–641. doi: 10.1097/WNR.0b013e3283294159. [DOI] [PubMed] [Google Scholar]
  8. Denny BT, Ochsner KN, Weber J, Wager TD. Anticipatory brain activity predicts the success or failure of subsequent emotion regulation. Social cognitive and affective neuroscience. 2014;9:403–411. doi: 10.1093/scan/nss148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. DeRubeis RJ, Gelfand LA, Tang TZ, Simons AD. Medications Versus Cognitive Behavior Therapy for Severely Depressed Outpatients: Mega-Analysis of Four Randomized Comparisons. Am J Psychiatry. 1999;156:1007–1013. doi: 10.1176/ajp.156.7.1007. [DOI] [PubMed] [Google Scholar]
  10. DeRubeis RJ, Hollon SD, Amsterdam JD, Shelton RC, Young PR, Salomon RM, O’Reardon JP, Lovett ML, Gladis MM, Brown LL, Gallop R. Cognitive Therapy vs Medications in the Treatment of Moderate to Severe Depression. Arch Gen Psychiatry. 2005;62:409–416. doi: 10.1001/archpsyc.62.4.409. [DOI] [PubMed] [Google Scholar]
  11. Dillon DG, Pizzagalli DA. Evidence of successful modulation of brain activation and subjective experience during reappraisal of negative emotion in unmedicated depression. Psychiatry research. 2013;212:99–107. doi: 10.1016/j.pscychresns.2013.01.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dobson KS, Hollon SD, Dimidjian S, Schmaling KB, Kohlenberg RJ, Gallop RJ, et al. Randomized trial of behavioral activation, cognitive therapy, and antidepressant medication in the prevention of relapse and recurrence in major depression. Journal of consulting and clinical psychology. 2008;76:468–477. doi: 10.1037/0022-006X.76.3.468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Eklund A, Nichols TE, Knutsson H. Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates. Proceedings of the National Academy of Sciences of the United States of America. 2016;113:7900–7905. doi: 10.1073/pnas.1602413113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Elkin I, Shea MT, Watkins JT, Imber SD, Sotsky SM, Collins JF, Glass DR, Pilkonis PA, Leber WR, Docherty JP, et al. National Institute of Mental Health Treatment of Depression Collaborative Research Program. General effectiveness of treatments. Arch Gen Psychiatry. 1989;46:971–982. doi: 10.1001/archpsyc.1989.01810110013002. discussion 983. [DOI] [PubMed] [Google Scholar]
  15. First M, Spitzer R, Gibbon M, Williams J. Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I/P, Version 2.0) Biometrics Research Dept., New York State Psychiatric Institute; New York: 1995. [Google Scholar]
  16. Forbes EE, Olino TM, Ryan ND, Birmaher B, Axelson D, Moyles DL, et al. Reward-related brain function as a predictor of treatment response in adolescents with major depressive disorder. Cognitive, affective & behavioral neuroscience. 2010;10:107–118. doi: 10.3758/CABN.10.1.107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Forkmann T, Scherer A, Pawelzik M, Mainz V, Drueke B, Boecker M, Gauggel S. Does cognitive behavior therapy alter emotion regulation in inpatients with a depressive disorder? Psychology Research and Behavior Management. 2014;7:147–153. doi: 10.2147/PRBM.S59421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Franklin G, Carson AJ, Welch KA. Cognitive behavioural therapy for depression: systematic review of imaging studies. Acta neuropsychiatrica. 2016;28:61–74. doi: 10.1017/neu.2015.41. [DOI] [PubMed] [Google Scholar]
  19. Fu CH, Williams SC, Cleare AJ, Scott J, Mitterschiffthaler MT, Walsh ND, Donaldson C, Suckling J, Andrew C, Steiner H, Murray RM. Neural responses to sad facial expressions in major depression following cognitive behavioral therapy. Biological Psychiatry. 2008;64:505–512. doi: 10.1016/j.biopsych.2008.04.033. [DOI] [PubMed] [Google Scholar]
  20. Fujino J, Yamasaki N, Miyata J, Sasaki H, Matsukawa N, Takemura A, et al. Anterior cingulate volume predicts response to cognitive behavioral therapy in major depressive disorder. Journal of affective disorders. 2015;174:397–399. doi: 10.1016/j.jad.2014.12.009. [DOI] [PubMed] [Google Scholar]
  21. Gasquoine PG. Contributions of the insula to cognition and emotion. Neuropsychology review. 2014;24:77–87. doi: 10.1007/s11065-014-9246-9. [DOI] [PubMed] [Google Scholar]
  22. Goldin PR, McRae K, Ramel W, Gross JJ. The neural bases of emotion regulation: reappraisal and suppression of negative emotion. Biol Psychiatry. 2008;63:577–586. doi: 10.1016/j.biopsych.2007.05.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gotlib IH, Joormann J. Cognition and depression: current status and future directions. Annual review of clinical psychology. 2010;6:285–312. doi: 10.1146/annurev.clinpsy.121208.131305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psych. 1960;23:56–62. doi: 10.1136/jnnp.23.1.56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Harkness KL, Bagby RM, Kennedy SH. Childhood maltreatment and differential treatment response and recurrence in adult major depressive disorder. J Consult Clin Psychol. 2012;80:342–353. doi: 10.1037/a0027665. [DOI] [PubMed] [Google Scholar]
  26. Heller AS, Johnstone T, Shackman AJ, Light SN, Peterson MJ, Kolden GG, Kalin NH, Davidson RJ. Reduced capacity to sustain positive emotion in major depression reflects diminished maintenance of fronto-striatal brain activation. Proceedings of the National Academy of Sciences of the United States of America. 2009;106:22445–22450. doi: 10.1073/pnas.0910651106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Herrmann MJ, Katzorke A, Busch Y, Gromer D, Polak T, Pauli P, et al. Medial prefrontal cortex stimulation accelerates therapy response of exposure therapy in acrophobia. Brain stimulation. 2017;10:291–297. doi: 10.1016/j.brs.2016.11.007. [DOI] [PubMed] [Google Scholar]
  28. Holland AC, Kensinger EA. The neural correlates of cognitive reappraisal during emotional autobiographical memory recall. Journal of cognitive neuroscience. 2013;25:87–108. doi: 10.1162/jocn_a_00289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Hopfinger L, Berking M, Bockting CL, Ebert DD. Emotion regulation mediates the effect of childhood trauma on depression. Journal of affective disorders. 2016;198:189–197. doi: 10.1016/j.jad.2016.03.050. [DOI] [PubMed] [Google Scholar]
  30. Johnstone T, van Reekum CM, Urry HL, Kalin NH, Davidson RJ. Failure to regulate: counterproductive recruitment of top-down prefrontal-subcortical circuitry in major depression. The Journal of neuroscience: the official journal of the Society for Neuroscience. 2007;27:8877–8884. doi: 10.1523/JNEUROSCI.2063-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kircanski K, Joormann J, Gotlib IH. Cognitive Aspects of Depression. Wiley interdisciplinary reviews. Cognitive science. 2012;3:301–313. doi: 10.1002/wcs.1177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kross E, Davidson M, Weber J, Ochsner K. Coping with emotions past: the neural bases of regulating affect associated with negative autobiographical memories. Biol Psychiatry. 2009;65:361–366. doi: 10.1016/j.biopsych.2008.10.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lueken U, Straube B, Wittchen HU, Konrad C, Strohle A, Wittmann A, et al. Therapygenetics: anterior cingulate cortex-amygdala coupling is associated with 5-HTTLPR and treatment response in panic disorder with agoraphobia. Journal of neural transmission. 2015;122:135–144. doi: 10.1007/s00702-014-1311-2. [DOI] [PubMed] [Google Scholar]
  34. Luty SE, Carter JD, McKenzie JM, Rae AM, Frampton CM, Mulder RT, Joyce PR. Randomised controlled trial of interpersonal psychotherapy and cognitive-behavioural therapy for depression. Br J Psychiatry. 2007;190:496–502. doi: 10.1192/bjp.bp.106.024729. [DOI] [PubMed] [Google Scholar]
  35. Mathers C, Fat DM, Boerma JT, Organization WH. The Global Burden of Disease: 2004 Update. World Health Organization; 2008. [Google Scholar]
  36. McGrath CL, Kelley ME, Holtzheimer PE, Dunlop BW, Craighead WE, Franco AR, Craddock RC, Mayberg HS. Toward a neuroimaging treatment selection biomarker for major depressive disorder. JAMA Psychiatry. 2013;70:821–829. doi: 10.1001/jamapsychiatry.2013.143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Morris RR, Schueller SM, Picard RW. Efficacy of a Web-based, crowdsourced peer-to-peer cognitive reappraisal platform for depression: randomized controlled trial. Journal of medical Internet research. 2015;17:e72. doi: 10.2196/jmir.4167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Nejad AB, Fossati P, Lemogne C. Self-referential processing, rumination, and cortical midline structures in major depression. Frontiers in human neuroscience. 2013;7:666. doi: 10.3389/fnhum.2013.00666. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Nemeroff CB, Heim CM, Thase ME, Klein DN, Rush AJ, Schatzberg AF, Ninan PT, McCullough JP, Jr, Weiss PM, Dunner DL, Rothbaum BO, Kornstein S, Keitner G, Keller MB. Differential responses to psychotherapy versus pharmacotherapy in patients with chronic forms of major depression and childhood trauma. Proceedings of the National Academy of Sciences of the United States of America. 2003;100:14293–14296. doi: 10.1073/pnas.2336126100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Ochsner KN, Bunge SA, Gross JJ, Gabrieli JD. Rethinking feelings: an FMRI study of the cognitive regulation of emotion. J Cogn Neurosci. 2002;14:1215–1229. doi: 10.1162/089892902760807212. [DOI] [PubMed] [Google Scholar]
  41. Ochsner KN, Ray RD, Cooper JC, Robertson ER, Chopra S, Gabrieli JD, et al. For better or for worse: neural systems supporting the cognitive down- and up-regulation of negative emotion. NeuroImage. 2004;23:483–499. doi: 10.1016/j.neuroimage.2004.06.030. [DOI] [PubMed] [Google Scholar]
  42. Ochsner KN, Ray RR, Hughes B, McRae K, Cooper JC, Weber J, et al. Bottom-up and top-down processes in emotion generation: common and distinct neural mechanisms. Psychological science. 2009;20:1322–1331. doi: 10.1111/j.1467-9280.2009.02459.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Ray RD, Ochsner KN, Cooper JC, Robertson ER, Gabrieli JD, Gross JJ. Individual differences in trait rumination and the neural systems supporting cognitive reappraisal. Cognitive, affective & behavioral neuroscience. 2005;5:156–168. doi: 10.3758/cabn.5.2.156. [DOI] [PubMed] [Google Scholar]
  44. Ritchey M, Dolcos F, Eddington KM, Strauman TJ, Cabeza R. Neural correlates of emotional processing in depression: changes with cognitive behavioral therapy and predictors of treatment response. J Psychiatr Res. 2011;45:577–587. doi: 10.1016/j.jpsychires.2010.09.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Satpute AB, Nook EC, Narayanan S, Shu J, Weber J, Ochsner KN. Emotions in “Black and White” or Shades of Gray? How We Think About Emotion Shapes Our Perception and Neural Representation of Emotion. Psychological science. 2016;27:1428–1442. doi: 10.1177/0956797616661555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Satpute AB, Shu J, Weber J, Roy M, Ochsner KN. The functional neural architecture of self-reports of affective experience. Biol Psychiatry. 2013;73:631–638. doi: 10.1016/j.biopsych.2012.10.001. [DOI] [PubMed] [Google Scholar]
  47. Siegle GJ, Carter CS, Thase ME. Use of FMRI to predict recovery from unipolar depression with cognitive behavior therapy. Am J Psychiatry. 2006;163:735–738. doi: 10.1176/ajp.2006.163.4.735. [DOI] [PubMed] [Google Scholar]
  48. Stark CE, Squire LR. When zero is not zero: the problem of ambiguous baseline conditions in fMRI. Proceedings of the National Academy of Sciences of the United States of America. 2001;98:12760–12766. doi: 10.1073/pnas.221462998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Stiles-Shields C, Corden ME, Kwasny MJ, Schueller SM, Mohr DC. Predictors of outcome for telephone and face-to-face administered cognitive behavioral therapy for depression. Psychological medicine. 2015;45:3205–3215. doi: 10.1017/S0033291715001208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Straub J, Plener PL, Sproeber N, Sprenger L, Koelch MG, Groen G, et al. Neural correlates of successful psychotherapy of depression in adolescents. Journal of affective disorders. 2015;183:239–246. doi: 10.1016/j.jad.2015.05.020. [DOI] [PubMed] [Google Scholar]
  51. Thompson DG, Kesler SR, Sudheimer K, Mehta KM, Thompson LW, Marquett RM, et al. FMRI activation during executive function predicts response to cognitive behavioral therapy in older, depressed adults. The American journal of geriatric psychiatry: official journal of the American Association for Geriatric Psychiatry. 2015;23:13–22. doi: 10.1016/j.jagp.2014.02.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Treynor W, Gonzalez R, Nolen-Hoeksema S. Rumination Reconsidered: A Psychometric Analysis. Cognitive Therapy and Research. 2003:27. [Google Scholar]
  53. Trivedi MH, Rush AJ, Wisniewski SR, Nierenberg AA, Warden D, Ritz L, Norquist G, Howland RH, Lebowitz B, McGrath PJ, Shores-Wilson K, Biggs MM, Balasubramani GK, Fava M. Evaluation of outcomes with citalopram for depression using measurement-based care in STAR*D: implications for clinical practice. Am J Psychiatry. 2006;163:28–40. doi: 10.1176/appi.ajp.163.1.28. [DOI] [PubMed] [Google Scholar]
  54. Wager TD, Barrett LF. From affect to control: Functional specialization of the insula in motivation and regulation. bioRxiv 2017 [Google Scholar]
  55. Wager TD, Davidson ML, Hughes BL, Lindquist MA, Ochsner KN. Prefrontal-Subcortical Pathways Mediating Successful Emotion Regulation. Neuron. 2008;59:1037–1050. doi: 10.1016/j.neuron.2008.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Williams JM, Barnhofer T, Crane C, Herman D, Raes F, Watkins E, Dalgleish T. Autobiographical memory specificity and emotional disorder. Psychological bulletin. 2007;133:122–148. doi: 10.1037/0033-2909.133.1.122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Woo CW, Krishnan A, Wager TD. Cluster-extent based thresholding in fMRI analyses: pitfalls and recommendations. NeuroImage. 2014;91:412–419. doi: 10.1016/j.neuroimage.2013.12.058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Woolrich M. Robust group analysis using outlier inference. NeuroImage. 2008;41:286–301. doi: 10.1016/j.neuroimage.2008.02.042. [DOI] [PubMed] [Google Scholar]
  59. Woolrich MW, Jbabdi S, Patenaude B, Chappell M, Makni S, Behrens T, Beckmann C, Jenkinson M, Smith SM. Bayesian analysis of neuroimaging data in FSL. NeuroImage. 2009;45:S173–186. doi: 10.1016/j.neuroimage.2008.10.055. [DOI] [PubMed] [Google Scholar]
  60. Worsley KJ. Ch 14, in Functional MRI: An Introduction to Methods, Statistical analysis of activation images. OUP; 2001. [Google Scholar]
  61. Yoshimura S, Okamoto Y, Onoda K, Matsunaga M, Okada G, Kunisato Y, Yoshino A, Ueda K, Suzuki SI, Yamawaki S. Cognitive behavioral therapy for depression changes medial prefrontal and ventral anterior cingulate cortex activity associated with self-referential processing. Social cognitive and affective neuroscience. 2013 doi: 10.1093/scan/nst009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Young JE, Beck AT. Cognitive Therapy Scale 1980 [Google Scholar]

Associated Data

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

Supplementary Materials

supplement
NIHMS919692-supplement.docx (598.4KB, docx)

RESOURCES