Hostname: page-component-745bb68f8f-l4dxg Total loading time: 0 Render date: 2025-02-11T10:20:56.386Z Has data issue: false hasContentIssue false
  • English
  • Français

Effects of electroconvulsive therapy on amygdala function in major depression – a longitudinal functional magnetic resonance imaging study

Published online by Cambridge University Press:  11 April 2017

R. Redlich ,
C. Bürger ,
K. Dohm ,
D. Grotegerd ,
N. Opel ,
D. Zaremba ,
S. Meinert ,
K. Förster ,
J. Repple  and
R. Schnelle
...Show all authors
R. Redlich*
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
C. Bürger
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
K. Dohm
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
D. Grotegerd
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
N. Opel
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
D. Zaremba
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
S. Meinert
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
K. Förster
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
J. Repple
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
R. Schnelle
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
C. Wagenknecht
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
M. Zavorotnyy
Affiliation:
Department of Psychiatry, University of Marburg, Marburg, Germany
W. Heindel
Affiliation:
Department of Clinical Radiology, University of Münster, Münster, Germany
H. Kugel
Affiliation:
Department of Clinical Radiology, University of Münster, Münster, Germany
M. Gerbaulet
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
J. Alferink
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany Cells-in-Motion Cluster of Excellence, University of Münster, Münster, Germany
V. Arolt
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
P. Zwanzger
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany Inn-Salzach Hospital, Wasserburg am Inn, Germany Ludwig-Maximilians-University Munich, Munich, Germany
U. Dannlowski
Affiliation:
Department of Psychiatry, University of Münster, Münster, Germany
*
*Address for correspondence: R. Redlich, Ph.D., Department of Psychiatry, University of Münster, Albert-Schweitzer-Campus 1, A9, 48149 Münster, Germany. (Email: r.redlich@uni-muenster.de)
Rights & Permissions [Opens in a new window]

Abstract

Background

Electroconvulsive therapy (ECT) is one of the most effective treatments for severe depression. However, little is known regarding brain functional processes mediating ECT effects.

Method

In a non-randomized prospective study, functional magnetic resonance imaging data during the automatic processing of subliminally presented emotional faces were obtained twice, about 6 weeks apart, in patients with major depressive disorder (MDD) before and after treatment with ECT (ECT, n = 24). Additionally, a control sample of MDD patients treated solely with pharmacotherapy (MED, n = 23) and a healthy control sample (HC, n = 22) were obtained.

Results

Before therapy, both patient groups equally showed elevated amygdala reactivity to sad faces compared with HC. After treatment, a decrease in amygdala activity to negative stimuli was discerned in both patient samples indicating a normalization of amygdala function, suggesting mechanisms potentially unspecific for ECT. Moreover, a decrease in amygdala activity to sad faces was associated with symptomatic improvements in the ECT sample (rspearman = −0.48, p = 0.044), and by tendency also for the MED sample (rspearman = −0.38, p = 0.098). However, we did not find any significant association between pre-treatment amygdala function to emotional stimuli and individual symptom improvement, neither for the ECT sample, nor for the MED sample.

Conclusions

In sum, the present study provides first results regarding functional changes in emotion processing due to ECT treatment using a longitudinal design, thus validating and extending our knowledge gained from previous treatment studies. A limitation was that ECT patients received concurrent medication treatment.

Keywords

Amygdaladepressionelectroconvulsive therapyemotion processingfunctional magnetic resonance imaging
Type
Original Articles
Information
Psychological Medicine , Volume 47 , Issue 12 , September 2017 , pp. 2166 - 2176
Copyright
Copyright © Cambridge University Press 2017 

Introduction

Electroconvulsive therapy (ECT) is one of the most effective treatments of major depressive disorder (MDD) (Kho et al. Reference Kho, van Vreeswijk, Simpson and Zwinderman2003). Although the clinical efficacy is well documented, little is known about the mediating neurobiological effects that accompany ECT, and how they are related to the clinical symptom improvement. Though various recent studies have investigated neuro-structural effects of ECT (Tendolkar et al. Reference Tendolkar, van Beek, van Oostrom, Mulder, Janzing, Voshaar and van Eijndhoven2013; Abbott et al. Reference Abbott, Jones, Lemke, Gallegos, McClintock, Mayer, Bustillo and Calhoun2014; Dukart et al. Reference Dukart, Regen, Kherif, Colla, Bajbouj and Heuser2014; Bouckaert et al. Reference Bouckaert, De Winter, Emsell, Dols, Rhebergen, Wampers, Sunaert, Stek, Sienaert and Vandenbulcke2016; Redlich et al. Reference Redlich, Opel, Grotegerd, Dohm, Zaremba, Bürger, Münker, Mühlmann, Wahl, Heindel, Arolt, Alferink, Zwanzger, Zavorotnyy, Kugel and Dannlowski2016), very little knowledge is available at present about functional changes in the brain due to treatment.

Recent longitudinal studies show that ECT induces massive structural plasticity in humans, particularly in the hippocampus and amygdala (Tendolkar et al. Reference Tendolkar, van Beek, van Oostrom, Mulder, Janzing, Voshaar and van Eijndhoven2013; Abbott et al. Reference Abbott, Jones, Lemke, Gallegos, McClintock, Mayer, Bustillo and Calhoun2014; Dukart et al. Reference Dukart, Regen, Kherif, Colla, Bajbouj and Heuser2014; Joshi et al. Reference Joshi, Espinoza, Pirnia, Shi, Wang, Ayers, Leaver, Woods and Narr2016). However, there is growing evidence that the extent of structural plasticity in these regions does not necessarily correlate with the extent of clinical response (Tendolkar et al. Reference Tendolkar, van Beek, van Oostrom, Mulder, Janzing, Voshaar and van Eijndhoven2013; Dukart et al. Reference Dukart, Regen, Kherif, Colla, Bajbouj and Heuser2014; Jorgensen et al. Reference Jorgensen, Magnusson, Hanson, Kirkegaard, Benveniste, Lee, Svarer, Mikkelsen, Fink-Jensen, Knudsen, Paulson, Bolwig and Jorgensen2015; Bouckaert et al. Reference Bouckaert, De Winter, Emsell, Dols, Rhebergen, Wampers, Sunaert, Stek, Sienaert and Vandenbulcke2016; Nickl-Jockschat et al. Reference Nickl-Jockschat, Palomero Gallagher, Kumar, Hoffstaedter, Brügmann, Habel, Eickhoff and Grözinger2016). This suggests a possibility that these structural alterations associated with ECT might be epiphenomena – i.e. they are not essential for the therapeutic effect (Nickl-Jockschat et al. Reference Nickl-Jockschat, Palomero Gallagher, Kumar, Hoffstaedter, Brügmann, Habel, Eickhoff and Grözinger2016) and might be a byproduct rather than the underlying mechanism of ECT (Jorgensen et al. Reference Jorgensen, Magnusson, Hanson, Kirkegaard, Benveniste, Lee, Svarer, Mikkelsen, Fink-Jensen, Knudsen, Paulson, Bolwig and Jorgensen2015).

Therefore, the investigation of brain function and its changes during the course of ECT might be more promising for the discovery of a biomarker related to the improvement of symptoms. At present, however, little is known about functional correlates of ECT treatment in patients with MDD. Only a few recent neuroimaging studies (Abbott et al. Reference Abbott, Jones, Lemke, Gallegos, McClintock, Mayer, Bustillo and Calhoun2014; Liu et al. Reference Liu, Du, Li, Liu, Zhao, Liu, Zeng, Li, Fu, Qiu, Li, Qiu, Hu, Meng and Luo2015; Argyelan et al. Reference Argyelan, Lencz, Kaliora, Sarpal, Weissman, Kingsley, Malhotra and Petrides2016; Leaver et al. Reference Leaver, Espinoza, Pirnia, Joshi, Woods and Narr2016) have investigated ECT-related changes in brain function by using resting-state functional magnetic resonance imaging (fMRI). These studies show, among other, that resting-state connectivity of the hippocampus (Abbott et al. Reference Abbott, Jones, Lemke, Gallegos, McClintock, Mayer, Bustillo and Calhoun2014), the anterior cingulate gyrus (Leaver et al. Reference Leaver, Espinoza, Pirnia, Joshi, Woods and Narr2016), and the subgenual cingulate cortex (Liu et al. Reference Liu, Du, Li, Liu, Zhao, Liu, Zeng, Li, Fu, Qiu, Li, Qiu, Hu, Meng and Luo2015; Argyelan et al. Reference Argyelan, Lencz, Kaliora, Sarpal, Weissman, Kingsley, Malhotra and Petrides2016) changes due to the ECT series. Furthermore, resting-state connectivity patterns in prefrontal and cingulate cortex networks seem to be capable of predicting ECT efficacy in patients with severe unipolar depression (van Waarde et al. Reference van Waarde, Scholte, van Oudheusden, Verwey, Denys and van Wingen2015). A recent study has further shown that ECT elicits robust effects on brain chemistry – including glutamate and creatine – in the subgenual anterior cingulate cortex and hippocampus, which suggests restorative and neurotrophic processes (Njau et al. Reference Njau, Joshi, Espinoza, Leaver, Vasavada, Marquina, Woods and Narr2016).

However, longitudinal studies probing the effects of ECT on brain function during emotion processing – a domain that is consistently reported to be impaired in MDD – are completely missing. In particular, amygdala dysfunction, including abnormally increased activation to negative (mood-congruent) stimuli and decreased activation to positive (mood-incongruent) stimuli, has been repeatedly reported for MDD, not only for supraliminal (Siegle et al. Reference Siegle, Thompson, Carter, Steinhauer and Thase2007; Groenewold et al. Reference Groenewold, Opmeer, de Jonge, Aleman and Costafreda2013), but also for subliminal stimuli associated with automatic stages of emotion processing (Sheline et al. Reference Sheline, Barch, Donnelly, Ollinger, Snyder and Mintun2001; Victor et al. Reference Victor, Furey, Fromm, Ohman and Drevets2010; Stuhrmann et al. Reference Stuhrmann, Dohm, Kugel, Zwanzger, Redlich, Grotegerd, Rauch, Arolt, Heindel, Suslow, Zwitserlood and Dannlowski2013). These early, automatic processing stages in limbic circuits seem to play a particularly crucial role in acutely depressed patients (Stuhrmann et al. Reference Stuhrmann, Dohm, Kugel, Zwanzger, Redlich, Grotegerd, Rauch, Arolt, Heindel, Suslow, Zwitserlood and Dannlowski2013) and, therefore, might be related to acute depressive states more directly than brain structural alterations. It has been proposed that antidepressants exert their effect partly via altering these neural emotion processing biases, thereby explaining the time lag of antidepressant action (Harmer et al. Reference Harmer, Goodwin and Cowen2009). However, such potential mechanisms have never been investigated for ECT treatment at all.

Therefore, the present naturalistic study was designed to probe the functional neural underpinnings of emotion processing in MDD patients treated with ECT, and to investigate the following four questions:

  1. 1. Which functional changes in amygdala reactivity to rapid emotion processing accompany ECT treatment?

  2. 2. Are potential changes in amygdala reactivity specific for ECT treatment compared with pharmacotherapy alone?

  3. 3. Are these potential functional changes related to clinical response?

  4. 4. Can pre-treatment amygdala activation probed by emotion stimuli predict response to ECT?

Method

Participants and study design

The present non-randomized prospective study is comprised of 24 patients with MDD who were treated with ECT (mean age = 50.42 years, s.d. = 6.58), 23 patients with MDD who were treated solely with drugs (MED, mean age = 45.35 years, s.d. = 9.52) and 22 healthy controls (HC, mean age = 45.63 years, s.d. = 10.80). Owing to the naturalistic design of the study, the decision of treatment was only based on clinical indication and patients were not randomized into treatment groups. The MED sample was matched according to age, sex, education and depression severity (p's > 0.05).

All subjects were assessed during a subliminal affective priming paradigm at two different points of time, about 45.02 days (s.d. = 15.05) apart. Patients were recruited through the in-patient service of the Department of Psychiatry, University of Münster. HC were recruited through public notices and newspaper announcements. Diagnoses were verified employing the Structured Clinical Interview for DSM-IV (Wittchen et al. Reference Wittchen, Wunderlich, Gruschwitz and Zaudig1997). All patients suffered from a current major depressive episode and fulfilled the criteria of MDD. All patients in the ECT sample were treatment resistant – first-degree treatment resistance implies unresponsiveness to two successive adequate treatments with antidepressants – according to Kuhs (Reference Kuhs1995). In the MED sample 10 patients showed first-degree treatment resistance. Two patients showed psychotic symptoms and were diagnosed as severe depression with psychotic symptoms. The Beck Depression Inventory (BDI) (Hautzinger et al. Reference Hautzinger, Bailer, Worall and Keller1994) was used as a self-evaluation questionnaire in order to assess the severity of depressive symptoms. The Hamilton Rating Scale of Depression (HAMD) (Hamilton, Reference Hamilton1960) was additionally conducted by a clinical interviewer at both time points. In order to assess medication intake, total medication load in MDD patients was assessed as described earlier (Redlich et al. Reference Redlich, Almeida, Grotegerd, Opel, Kugel, Heindel, Arolt, Phillips and Dannlowski2014). Sociodemographic, clinical, questionnaire and behavioural data were analysed, comparing the samples (see Table 1 for final sample details). Exclusion criteria for all participants were any neurological abnormalities, organic mental disorders, dementia, brain injuries or MRI contraindications. For patients, co-morbid life-time substance-related disorders, bipolar disorders, schizophrenia and other psychotic disorders were exclusion criteria. For HC, any life-time psychiatric disorder was an exclusion criterion. The study was approved by the local institutional review board, and all participants provided written informed consent before participation.

Table 1. Sociodemographic and clinical characteristics

Data are given as mean (standard deviation) unless otherwise indicated. ECT, Electroconvulsive therapy; MED, medication; HC, healthy controls; ANOVA, analysis of variance; BDI, Beck Depression Inventory; HAMD, Hamilton Depression Rating Scale; n.a., not applicable; SSNRI, selective serotonin noradrenaline reuptake inhibitor; SSRI, selective serotonin reuptake inhibitor; SDNRI, selective dopamine noradrenaline reuptake inhibitor.

Paradigm

A frequently used paradigm (Dannlowski et al. Reference Dannlowski, Kugel, Huber, Stuhrmann, Redlich, Grotegerd, Dohm, Sehlmeyer, Konrad, Baune, Arolt, Heindel, Zwitserlood and Suslow2013; Grotegerd et al. Reference Grotegerd, Stuhrmann, Kugel, Schmidt, Redlich, Zwanzger, Rauch, Heindel, Zwitserlood, Arolt, Suslow and Dannlowski2014) has been used to investigate early, automatic stages of emotion processing. Neutral, sad and happy facial expressions (Ekman & Friesen, Reference Ekman and Friesen1976) were presented as primes for 33 ms, and subsequently masked by neutral faces. Two fixed pseudo-random sequences with 80 trials were shown: 20 each with sad, happy and neutral prime faces, as well as 20 trials with no-face stimuli as primes. Each trial lasted 9 s and proceeded according to the following pattern. A fixation cross, presented for 800 ms, preceded a prime face (for 33 ms), which was immediately followed by a picture of the same actor with a neutral expression, serving as mask and target (467 ms). A blank screen was then displayed for 7700 ms. During this period, subjects had to evaluate whether the neutral (mask) face expressed negative or positive feelings, which they indicated by pressing one of four buttons (labelled −1.5, −0.5, +0.5, and +1.5).

ECT

Brief pulse ECT was conducted three times a week using the Thymatron IV system (Somatics Inc., USA). Initially, nine to 12 sessions of ECT were given, and these were continued if they failed to achieve symptom relief (mean = 14.00, s.d. = 3.83, range: 9–24). All patients received unilateral ECT. In three patients, treatment was switched to bilateral ECT because of insufficient response to unilateral treatment. The mean stimulus intensity was 51.95% (s.d. = 18.80), the mean stimulus duration 6.98 s (s.d. = 0.50), and the mean pulse frequency 40.17 Hz (s.d. = 10.16). The mean seizure durations were 42.86 s (s.d. = 10.08; electroencephalography) and 23.98 s (s.d. = 10.64; electromyography). The mean postictal suppression index was 93.86 % (s.d. = 2.77) while the seizure generalization index was 73.27% (s.d. = 8.59). For additional details please see our previous work (Redlich et al. Reference Redlich, Opel, Grotegerd, Dohm, Zaremba, Bürger, Münker, Mühlmann, Wahl, Heindel, Arolt, Alferink, Zwanzger, Zavorotnyy, Kugel and Dannlowski2016).

Data acquisition and statistical analyses

Acquisition and preprocessing

Our functional MRI methods and statistical approach followed published protocols (Opel et al. Reference Opel, Redlich, Grotegerd, Dohm, Haupenthal, Heindel, Kugel, Arolt and Dannlowski2015; Redlich et al. Reference Redlich, Dohm, Grotegerd, Opel, Zwitserlood, Heindel, Arolt, Kugel and Dannlowski2015a , Reference Redlich, Grotegerd, Opel, Kaufmann, Zwitserlood, Kugel, Heindel, Donges, Suslow, Arolt and Dannlowski b , Reference Redlich, Stacey, Opel, Grotegerd, Dohm, Kugel, Heindel, Arolt, Baune and Dannlowski c ). Briefly, T2* functional data were acquired with a 3 Tesla scanner (Gyroscan Intera 3 T; Philips Medical Systems, the Netherlands) using a single-shot echoplanar sequence, with parameters selected to minimize distortion in the region of central interest, while retaining an adequate signal:noise ratio and T2* sensitivity. Volumes consisting of 34 slices were acquired (matrix 64 × 64, resolution 3.6 mm × 3.6 mm × 3.6 mm; repetition time = 2.1 s, echo time = 30 ms, flip angle = 90°). The slices were tilted by 25° from the anterior commissure/posterior commissure line in order to minimize drop-out artifacts in the mediotemporal and orbitofrontal regions. All stimuli were projected towards the rear end of the scanner (Sharp XG-PC10XE with additional high-frequency shielding). Data were analysed using statistical parametric mapping software (SPM8; Wellcome Department of Cognitive Neurology, UK; http://www.fil.ion.ucl.ac.uk/spm). Functional data were preprocessed, including realignment, unwarping and spatial normalization of each participant's functional images to the Montreal Neurological Institute International Consortium for Brain Mapping template. Nine patients with MDD (five patients from the ECT sample, four patients from the MED sample) and two HC had to be excluded due to excessive head movement (exclusion criterion 3 mm/3°). Images were smoothed with a Gaussian kernel of 6 mm full-width at half-maximum (FWHM).

First-level analyses

The onsets of the experimental conditions (positive, negative, neutral) were modelled in an event-related fashion by using a canonical haemodynamic response function in the context of a general linear model, and the model was corrected for serial correlations. A high-pass filter of 128 s was used to remove low-frequency noise. For each subject, two contrast images were generated in each individual first-level analysis (sad > neutral, happy > neutral) in order to investigate our research objective.

Second-level analyses

First, in order to investigate diagnosis and treatment-related functional changes, a 3 × 2 × 2 analysis of variance (ANOVA) was calculated using a full factorial model, with group (ECT v. MED v. HC) as the between-subjects factor, as well as condition (negative v. neutral; positive v. neutral) and time (T1 v. T2) as within-subjects factors. To cover for the longitudinal effects, the interaction (group × condition × time) was analysed (objectives 1 and 2). Potential interactions were followed by appropriate t contrasts, including the cross-sectional comparisons of samples at T1 and T2, as well as the comparisons of brain functional changes across time (T1–T2), again using the above-mentioned full factorial model. To address our hypotheses on differential amygdala responsiveness to negative stimuli, region-of-interest analyses of the bilateral amygdala were carried out. The mask for the bilateral amygdala was created with the aid of the WFU PickAtlas according to the Automated Anatomical Labeling (AAL) atlas definitions (Tzourio-Mazoyer et al. Reference Tzourio-Mazoyer, Landeau, Papathanassiou, Crivello, Etard, Delcroix, Mazoyer and Joliot2002). To control for multiple statistical testing, the cluster-level false-positive detection rate was kept at p < 0.05, using a voxel-level threshold of p < 0.05 with a cluster extent (k) empirically determined by Monte-Carlo simulations (n = 1000 iterations). This was performed by means of the AlphaSim (Forman et al. Reference Forman, Cohen, Fitzgerald, Eddy, Mintun and Noll1995) procedure, implemented in the REST toolbox (http://restfmri.net/forum/index.php), as reported in previous publications (Dannlowski et al. Reference Dannlowski, Grabe, Wittfeld, Klaus, Konrad, Grotegerd, Redlich, Suslow, Opel, Ohrmann, Bauer, Zwanzger, Laeger, Hohoff, Arolt, Heindel, Deppe, Domschke, Hegenscheid, Völzke, Stacey, Meyer Zu Schwabedissen, Kugel and Baune2015). The empirically determined cluster threshold was k = 33 voxels for the bilateral amygdala. For exploratory reasons, additional whole-brain analyses, with a cluster threshold of k = 100 voxels, were conducted.

Second, to determine a potential relationship between brain function changes across time (T2–T1) and the clinical response, represented by BDI percentage score changes (1 – BDI at T2/BDI at T1) in patients, the contrast values of the resulting peak voxel of the interaction analysis were extracted for both time points and further analysed by using SPSS Statistics 21 (USA). The individual functional contrast values were then correlated with the percentage of symptom relief, as represented by the changes in BDI percentage scores (objective 3).

Finally, in order to investigate pre-treatment amygdala activation of emotional stimuli as a predictor of response to ECT, pre-treatment amygdala contrast values were correlated with symptom improvement in percentage, as measured by the BDI (objective 4).

Results

Effects of treatment

Treatment response

Both groups responded significantly, while the ECT sample showed nominally higher, but not significant (p’s > 0.07) percentage changes in symptoms improvement compared with the MED sample, as measured by the BDI (ECT: T1 = 29.84, s.d. = 9.28; T2 = 20.31, s.d. = 10.17, p < 0.001; MED: T1 = 27.56, s.d. = 7.49; T2 = 22.50, s.d. = 8.88, p < 0.001) and the HAMD (ECT: T1 = 25.00, s.d. = 6.10; T2 = 13.79, s.d. = 8.77, p < 0.001; MED: T1 = 23.95, s.d. = 5.27; T2 = 15.58, s.d. = 9.19, p < 0.001). Improvements in symptom severity in both patient groups did not significantly correlate with age (p’s > 0.42), education (p’s > 0.43), medication load index (p’s > 0.20) or any other clinical parameter described in Table 1 (p’s > 0.16). Furthermore, no significant difference in symptom improvement was found in both patient samples when subjects with antipsychotic treatment are compared with those with non-antipsychotic treatment (p’s > 0.40). However, in the MED sample, female patients benefited significantly more than males (p < 0.02) while this effect was absent in the ECT group (p > 0.28).

Longitudinal effects on amygdala reactivity

The 3 × 2 × 2 full-factorial model yielded an interaction effect of group × condition × time within the bilateral amygdala (left: x = −30, y = 0, z = −18, F 2,220 = 6.03, Z = 2.77, k = 47, p = 0.003; right: x = 30, y = 2, z = −28, F 2,220 = 5.77, Z = 2.77, k = 37, p = 0.004), resulting from a decrease in amygdala activity in reaction to sad faces in both patient samples (ECT: x = 30, y = 2, z = −20, Z = 2.26, k = 90, p = 0.012; MED: right: x = 32, y = 4, z = −20, Z = 3.28, k = 90, p = 0.001; left: x = −24, y = −8, z = −14, Z = 3.06, k = 123, p = 0.001), and an increase in amygdala activity to positive faces in the MED sample (x = 26, y = −8, z = −12, Z = 2.09, k = 33, p = 0.018). However, the increase of amygdala responsiveness to positive stimuli in the ECT group failed to reach significance. An additional 2 × 2 × 2 ANOVA excluding the HC sample yielded no significant differences between both patient samples, indicating a rather non-specific effect of treatment. As one might expect, the amygdala activity of the HC sample did not change significantly over time for any emotion condition (see Fig. 1). Comparing patients receiving antipsychotic treatment with those receiving non-antipsychotic treatment by using the extracted peak voxel values of the interaction analysis yielded no significant differences in amygdala activity changes either in the ECT sample (p’s > 0.67) or in the MED sample (p’s > 0.42). Additionally, there was no significant association between amygdala reactivity and age – either in the ECT sample (p’s > 0.38) or the MED sample (p’s > 0.34). Furthermore, there were no significant associations between amygdala reactivity and ECT parameters [all p’s < 0.11; mean stimulus intensity (r = 0.39, p = 0.11), mean stimulus duration (r = −0.01, p = 0.97), mean pulse frequency (r = 0.28, p = 0.27), mean seizure durations by EEG (r = −0.13, p = 0.61) and EMG (r = −0.37, p = 13), mean postictal suppression index (r = −0.15, p = 0.55) and the seizure generalization index (r = −0.22, p = 0.38)].

Fig. 1. Treatment effects of electroconvulsive therapy (ECT) on amygdala reactivity during emotion processing. Left: coronal view (Montreal Neurological Institute coordinates at y = 1) depicting the interaction effect of group × condition × time within the bilateral amygdala (left: F 2,220 = 6.03, p = 0.003; right: F 2,220 = 5.77, p = 0.004), resulting from a decrease in amygdala reactivity to sad faces in both patient samples [ECT: Z = 2.26, p = 0.012; medication (MED), right: Z = 3.28, p = 0.001; left: Z = 3.06, p = 0.001], and an increase in amygdala activity to positive faces in the MED sample (Z = 2.09, p = 0.018). The colour bar shows F values. Right: bar graphs depict the estimated contrast values subdivided by group [ECT, MED, healthy control HC)] at T1 (orange) and at T2 (blue) for the negative > neutral (above) and positive > neutral (below) conditions extracted from peak values of the interaction analysis. * Significant differences (Alpha-Sim corrected, voxel threshold, p < 0.05; minimum cluster volume threshold k = 33 voxels). fMRI, Functional magnetic resonance imaging.

The whole-brain analysis indicated that no other brain area showed any significant interaction effect.

Cross-sectional effects on amygdala reactivity

Compared with HC, both patient groups showed increased amygdala activity to sad faces at T1 (ECT: x = 34, y = −2, z = −28, Z = 2.54, k = 43, p = 0.006; MED: x = 28, y = 4, z = −20, Z = 3.14, k = 42, p = 0.001; x = −24, y = 0, z = −16, Z = 2.66, k = 128, p = 0.004). In turn, there were no significant differences for the happy face condition. The post-hoc analyses of cross-sectional effects at T2 did not reveal any significant differences of amygdala activity in reaction to negative or positive faces in both patient samples compared with controls, indicating that the abnormal amygdala responsiveness pattern observed at T1 in both patient groups was entirely normalized after treatment.

Associations between brain functional changes and clinical response

The correlation analyses yielded a significant negative association between changes in amygdala activity to sad faces and symptomatic improvements in the ECT sample (r spearman = −0.48, p = 0.044), and by tendency for the MED sample (r spearman = −0.38, p = 0.098). In contrast, a positive association was found between amygdala activity changes to positive faces and symptomatic improvements for the ECT sample (r spearman = 0.52, p = 0.027). However, this association was absent for the MED sample (p = 0.806, see Fig. 2). The comparison of the differences between correlation coefficients yielded marginal significance (Z = 1.48, p = 0.07) using the Fisher r-to-z transformation.

Fig. 2. Associations between functional magnetic resonance imaging (fMRI) contrast value changes in amygdala reactivity and changes in Beck Depression Inventory scores, displayed separately for the contrast negative > neutral (above) and positive > neutral (below) subdivided by treatment condition [left: electroconvulsive therapy (ECT) sample; right: medication (MED) sample]. The graphs depict a significant negative association between amygdala activity changes to sad faces and symptom relief in the ECT sample (r spearman = −0.48, p = 0.044), and by tendency for the MED sample (r spearman = −0.38, p = 0.098). In contrast, a positive association was found between amygdala activity changes to positive faces and symptomatic improvements for the ECT sample (r spearman = 0.52, p = 0.027).

Associations between pre-treatment amygdala function and ECT response

We did not find any significant association between the pre-treatment amygdala function to emotional stimuli and individual symptom improvement in percentage as measured with the BDI and HAMD for either the ECT sample (p’s > 0.18) or the MED sample (p’s > 0.17).

Discussion

The present prospective study is the first to investigate functional changes in rapid emotion processing in the course of ECT in a psychiatric sample by using fMRI. The principal findings suggest pre-treatment elevated amygdala activation to subliminally presented sad faces, replicating previous reports of a mood-congruent neuronal processing bias in the amygdala (Victor et al. Reference Victor, Furey, Fromm, Ohman and Drevets2010; Stuhrmann et al. Reference Stuhrmann, Dohm, Kugel, Zwanzger, Redlich, Grotegerd, Rauch, Arolt, Heindel, Suslow, Zwitserlood and Dannlowski2013; Suslow et al. Reference Suslow, Kugel, Ohrmann, Stuhrmann, Grotegerd, Redlich, Bauer and Dannlowski2013). More importantly, the present data suggest a normalization of amygdala function after a series of ECT (objective 1). However, again consistent with previous studies, this normalization of amygdala activity was also found in the MED sample, indicating a non-specific effect of treatment rather than an ECT-specific mechanism of action (objective 2). Furthermore, the findings of the regression analyses indicated that the degree of changes in amygdala reactivity to negative emotional stimuli went along with clinical symptom improvement (objective 3). However, we were not able to predict individual response to treatment through pre-treatment amygdala function (objective 4).

The present results are in line with previous studies consistently reporting functional abnormalities in implicit emotion-regulation circuitry in acutely depressed MDD patients. Abnormally elevated activity in the amygdala in response to negative, mood-congruent stimuli (Sheline et al. Reference Sheline, Barch, Donnelly, Ollinger, Snyder and Mintun2001; Victor et al. Reference Victor, Furey, Fromm, Ohman and Drevets2010; Hamilton et al. Reference Hamilton, Etkin, Furman, Lemus, Johnson and Gotlib2012) and, to a smaller extent, reduced activity in response to positive, mood-incongruent emotional stimuli (Suslow et al. Reference Suslow, Konrad, Kugel, Rumstadt, Zwitserlood, Schöning, Ohrmann, Bauer, Pyka, Kersting, Arolt, Heindel and Dannlowski2010; Victor et al. Reference Victor, Furey, Fromm, Ohman and Drevets2010; Stuhrmann et al. Reference Stuhrmann, Dohm, Kugel, Zwanzger, Redlich, Grotegerd, Rauch, Arolt, Heindel, Suslow, Zwitserlood and Dannlowski2013), for both subliminal (Suslow et al. Reference Suslow, Konrad, Kugel, Rumstadt, Zwitserlood, Schöning, Ohrmann, Bauer, Pyka, Kersting, Arolt, Heindel and Dannlowski2010; Stuhrmann et al. Reference Stuhrmann, Dohm, Kugel, Zwanzger, Redlich, Grotegerd, Rauch, Arolt, Heindel, Suslow, Zwitserlood and Dannlowski2013) as well as supraliminal emotion processing (Siegle et al. Reference Siegle, Thompson, Carter, Steinhauer and Thase2007; Groenewold et al. Reference Groenewold, Opmeer, de Jonge, Aleman and Costafreda2013) were repeatedly observed. Furthermore, consistent with longitudinal studies investigating antidepressant treatment (Sheline et al. Reference Sheline, Barch, Donnelly, Ollinger, Snyder and Mintun2001; Anand et al. Reference Anand, Li, Wang, Gardner and Lowe2007), both patient samples showed a decrease of amygdala reactivity to sad faces across time points. Importantly, the degree of functional changes in amygdala reactivity to ‘normal’ was associated with symptom improvements in both the ECT sample and the MED sample. Moreover, our results also revealed increased amygdala responses to happy facial expressions after antidepressant treatment in the MED sample as recently shown in studies investigating antidepressant treatment (Norbury et al. Reference Norbury, Taylor, Selvaraj, Murphy, Harmer and Cowen2009; Victor et al. Reference Victor, Furey, Fromm, Ohman and Drevets2010; Williams et al. Reference Williams, Korgaonkar, Song, Paton, Eagles, Goldstein-Piekarski, Grieve, Harris, Usherwood and Etkin2015). While these effects did not reach significance in the ECT sample, there was a similar tendency at an uncorrected significance level. To summarize, the present data suggest that a change in amygdala function in response to emotional stimuli seems to constitute a key factor and a general mechanism of action of therapy in depression, probably independent of the type of treatment. While recent longitudinal, structural ECT studies showed heterogeneous findings regarding the associations of structural (e.g. hippocampal) changes and clinical symptom improvements, functional changes seem to reflect a more valid state marker for symptom improvement.

However, to date, it is not finally clarified whether amygdala hyper-reactivity to negative stimuli is present before the onset of depression, and therefore represents a predisposition or trait factor for MDD, or whether these observed findings are the consequence of (acute) depression rather representing a state factor. At present, evidence exists for both interpretations. On the one hand, several neuroimaging studies are in line with our results, which provide a body of evidence proving that this pattern constitutes a state marker by reporting a normalization of abnormally elevated activity in limbic circuits in response to pharmacotherapy (Sheline et al. Reference Sheline, Barch, Donnelly, Ollinger, Snyder and Mintun2001; Fu et al. Reference Fu, Williams, Cleare, Brammer, Walsh, Kim, Andrew, Pich, Williams, Reed, Mitterschiffthaler, Suckling and Bullmore2004; Anand et al. Reference Anand, Li, Wang, Gardner and Lowe2007). On the other hand, these abnormal amygdala activation patterns are also found in healthy subjects at genetic (Dannlowski et al. Reference Dannlowski, Konrad, Kugel, Zwitserlood, Domschke, Schöning, Ohrmann, Bauer, Pyka, Hohoff, Zhang, Baune, Heindel, Arolt and Suslow2010; Joormann et al. Reference Joormann, Cooney, Henry and Gotlib2011) or environmental risk for MDD (Dannlowski et al. Reference Dannlowski, Kugel, Huber, Stuhrmann, Redlich, Grotegerd, Dohm, Sehlmeyer, Konrad, Baune, Arolt, Heindel, Zwitserlood and Suslow2013; van Harmelen et al. Reference van Harmelen, van Tol, Demenescu, van der Wee, Veltman, Aleman, van Buchem, Spinhoven, Penninx and Elzinga2013), thus indicating a trait marker for the onset of depression. However, the two conclusions are not mutually exclusive. Since hyper-reactivity of the amygdala might exist before the onset of depression, constituting a trait marker, these abnormal patterns might be more pronounced in depressive states, and furthermore, might be reversed or normalized due to pharmacotherapy or ECT.

Regarding our fourth objective – to verify whether pre-treatment amygdala function might have predictive value to indicate ECT response – we were unable to detect an association between pre-treatment amygdala activity and clinical response. While some fMRI studies of treatment response prediction in depression report that a higher degree of pre-treatment amygdala activity to emotional stimuli predicts better outcome of cognitive–behavioural therapy (Siegle et al. Reference Siegle, Carter and Thase2006) and antidepressant treatment (Canli et al. Reference Canli, Cooney, Goldin, Shah, Sivers, Thomason, Whitfield-Gabrieli, Gabrieli and Gotlib2005), some other studies report the opposite effects (Salvadore et al. Reference Salvadore, Cornwell, Colon-Rosario, Coppola, Grillon, Zarate and Manji2009; Williams et al. Reference Williams, Korgaonkar, Song, Paton, Eagles, Goldstein-Piekarski, Grieve, Harris, Usherwood and Etkin2015). Since these studies vary dramatically with respect to the form and duration of the therapy, medication types and functional paradigms, no meaningful conclusions can be drawn at present. Here, future studies with standardized paradigms and randomized clinical trials, preferably with a sham control group, are needed to prove the predictive power of the amygdala function as a pre-treatment biomarker.

Finally, despite the fact that ECT has proven efficacy for severe depression, nearly a third of patients do not benefit from ECT. For these patients, deep brain stimulation or transcranial magnetic stimulation might be alternative treatments for treatment-resistant depression (De Raedt et al. Reference De Raedt, Leyman, Baeken, Van Schuerbeek, Luypaert, Vanderhasselt and Dannlowski2010; Taghva et al. Reference Taghva, Malone and Rezai2013; Bogod et al. Reference Bogod, Sinden, Woo, Defreitas, Torres, Howard, Ilcewicz-Klimek, Honey, Yatham and Lam2014). However, research on the working mechanisms of these neurostimulation techniques is necessary to develop more efficient treatment protocols.

Some limitations of this study must be acknowledged. First and foremost, patients were not randomized into treatment groups due to the naturalistic design of the study. Second, due to the naturalistic design, all ECT patients additionally received medication that might have influenced the reported effects. Therefore, a contribution of additional pharmacologically induced functional changes cannot be ruled out. Such a contribution, moreover, is very probable, particularly since similar neurobiological effects were found in the MED sample. To shed light on the specific mechanisms of actions, future studies should aim to investigate medication-naïve patients treated with ECT. Third, it should be noted that re-testing may cause habituation problems. In particular, the amygdala is known to show strong habituation effects. However, all subjects – including the HC – were measured twice and decreases in amygdala reactivity were found only for patients, but not for HC. Furthermore, it cannot be ruled out that some specific mechanisms have been missed due to power issues. While our sample size was sufficient to detect differences between the patient samples and controls, potential group differences between treatment groups might have been too weak to be detect with the current sample size. Finally, the study only allows statements about the early, automatic stages of automatic emotion processing. Future studies should focus on supraliminal emotion-processing paradigms to detect whether the observed effects are specific for automatic emotion processing or not.

In summary, despite the above-mentioned limitations, the present study provides the first results with regard to functional changes in emotion processing due to ECT treatment by using a longitudinal design, thus validating and extending our knowledge about neurobiological effects of treatment in patients with MDD. While recent longitudinal and structural ECT studies show heterogeneous results regarding the associations of often-reported structural (e.g. hippocampal) changes and clinical symptom improvements, functional changes seem be more valid state markers for acute depression and, thus, might constitute a key factor and mechanism of action in the therapy of depression.

Acknowledgements

The study was supported by grants from the Innovative Medizinische Forschung (RE111604 to R.R.) and the German Research Foundation (DFG; grant FOR 2107; DA1151/5-1 to U.D.).

Declaration of Interest

V.A. is a member of the advisory board of, or has given presentations on behalf of, the following companies: Astra-Zeneca, Janssen-Organon, Lilly, Lundbeck, Servier, Pfizer, Otsuka and Trommsdorff. H.K. has received consultation fees from MR:comp GmbH, Testing Services for MR Safety. These affiliations are of no relevance to the work described in the paper. All other authors state that they have no actual or potential conflict of interest to declare, including any financial, personal or other relationships with other people or organizations within 3 years of beginning the present work that could influence or bias their work.

References

Abbott, CC, Jones, T, Lemke, NT, Gallegos, P, McClintock, SM, Mayer, AR, Bustillo, J, Calhoun, VD (2014). Hippocampal structural and functional changes associated with electroconvulsive therapy response. Translational Psychiatry 4, e483.CrossRefGoogle ScholarPubMed
Anand, A, Li, Y, Wang, Y, Gardner, K, Lowe, MJ (2007). Reciprocal effects of antidepressant treatment on activity and connectivity of the mood regulating circuit: an fMRI study. Journal of Neuropsychiatry and Clinical Neurosciences 19, 274–282.CrossRefGoogle ScholarPubMed
Argyelan, M, Lencz, T, Kaliora, S, Sarpal, DK, Weissman, N, Kingsley, PB, Malhotra, AK, Petrides, G (2016). Subgenual cingulate cortical activity predicts the efficacy of electroconvulsive therapy. Translational Psychiatry 6, e789.CrossRefGoogle ScholarPubMed
Bogod, NM, Sinden, M, Woo, C, Defreitas, VG, Torres, IJ, Howard, AK, Ilcewicz-Klimek, MI, Honey, CR, Yatham, LN, Lam, RW (2014). Long-term neuropsychological safety of subgenual cingulate gyrus deep brain stimulation for treatment-resistant depression. Journal of Neuropsychiatry and Clinical Neurosciences 26, 126–133.CrossRefGoogle ScholarPubMed
Bouckaert, F, De Winter, FL, Emsell, L, Dols, A, Rhebergen, D, Wampers, M, Sunaert, S, Stek, ML, Sienaert, P, Vandenbulcke, M (2016). Grey matter volume increase following electroconvulsive therapy in patients with late life depression: a longitudinal MRI study. Journal of Psychiatry and Neuroscience 41, 105–114.CrossRefGoogle ScholarPubMed
Canli, T, Cooney, RE, Goldin, P, Shah, M, Sivers, H, Thomason, ME, Whitfield-Gabrieli, S, Gabrieli, JDE, Gotlib, IH (2005). Amygdala reactivity to emotional faces predicts improvement in major depression. Neuroreport 16, 1267–1270.CrossRefGoogle ScholarPubMed
Dannlowski, U, Grabe, HJ, Wittfeld, K, Klaus, J, Konrad, C, Grotegerd, D, Redlich, R, Suslow, T, Opel, N, Ohrmann, P, Bauer, J, Zwanzger, P, Laeger, I, Hohoff, C, Arolt, V, Heindel, W, Deppe, M, Domschke, K, Hegenscheid, K, Völzke, H, Stacey, D, Meyer Zu Schwabedissen, H, Kugel, H, Baune, BT (2015). Multimodal imaging of a tescalcin (TESC)-regulating polymorphism (rs7294919)-specific effects on hippocampal gray matter structure. Molecular Psychiatry 20, 398–404.CrossRefGoogle ScholarPubMed
Dannlowski, U, Konrad, C, Kugel, H, Zwitserlood, P, Domschke, K, Schöning, S, Ohrmann, P, Bauer, J, Pyka, M, Hohoff, C, Zhang, W, Baune, BT, Heindel, W, Arolt, V, Suslow, T (2010). Emotion specific modulation of automatic amygdala responses by 5-HTTLPR genotype. NeuroImage 53, 893–898.CrossRefGoogle ScholarPubMed
Dannlowski, U, Kugel, H, Huber, F, Stuhrmann, A, Redlich, R, Grotegerd, D, Dohm, K, Sehlmeyer, C, Konrad, C, Baune, BT, Arolt, V, Heindel, W, Zwitserlood, P, Suslow, T (2013). Childhood maltreatment is associated with an automatic negative emotion processing bias in the amygdala. Human Brain Mapping 34, 2899–2909.CrossRefGoogle ScholarPubMed
De Raedt, R, Leyman, L, Baeken, C, Van Schuerbeek, P, Luypaert, R, Vanderhasselt, M-A, Dannlowski, U (2010). Neurocognitive effects of HF-rTMS over the dorsolateral prefrontal cortex on the attentional processing of emotional information in healthy women: an event-related fMRI study. Biological Psychology 85, 487–495.CrossRefGoogle ScholarPubMed
Dukart, J, Regen, F, Kherif, F, Colla, M, Bajbouj, M, Heuser, I (2014). Electroconvulsive therapy-induced brain plasticity determines therapeutic outcome in mood disorders. PNAS 111, 1156–1161.CrossRefGoogle ScholarPubMed
Ekman, P, Friesen, WV (1976). Pictures of Facial Affect. Consulting Psychologists Press: Palo Alto, CA.Google Scholar
Forman, SD, Cohen, JD, Fitzgerald, M, Eddy, WF, Mintun, MA, Noll, DC (1995). Improved assessment of significant activation in functional magnetic resonance imaging (fMRI): use of a cluster-size threshold. Magnetic Resonance Medicine 33, 636–647.CrossRefGoogle ScholarPubMed
Fu, CHY, Williams, SCR, Cleare, AJ, Brammer, MJ, Walsh, ND, Kim, J, Andrew, C, Pich, EM, Williams, PM, Reed, LJ, Mitterschiffthaler, MT, Suckling, J, Bullmore, ET (2004). Attenuation of the neural response to sad faces in major depression by antidepressant treatment: a prospective, event-related functional magnetic resonance imaging study. Archives of General Psychiatry 61, 877–889.CrossRefGoogle ScholarPubMed
Groenewold, NA, Opmeer, EM, de Jonge, P, Aleman, A, Costafreda, SG (2013). Emotional valence modulates brain functional abnormalities in depression: evidence from a meta-analysis of fMRI studies. Neuroscience and Biobehavioral Reviews 37, 152–163.CrossRefGoogle ScholarPubMed
Grotegerd, D, Stuhrmann, A, Kugel, H, Schmidt, S, Redlich, R, Zwanzger, P, Rauch, AV, Heindel, W, Zwitserlood, P, Arolt, V, Suslow, T, Dannlowski, U (2014). Amygdala excitability to subliminally presented emotional faces distinguishes unipolar and bipolar depression – an fMRI and pattern classification study. Human Brain Mapping 35, 2995–3007.CrossRefGoogle ScholarPubMed
Hamilton, JP, Etkin, A, Furman, DJ, Lemus, MG, Johnson, RF, Gotlib, IH (2012). Functional neuroimaging of major depressive disorder: a meta-analysis and new integration of base line activation and neural response data. American Journal of Psychiatry 169, 693–703.CrossRefGoogle ScholarPubMed
Hamilton, M (1960). A rating scale for depression. Journal of Neurology, Neurosurgery, and Psychiatry 23, 56–63.CrossRefGoogle ScholarPubMed
Harmer, CJ, Goodwin, GM, Cowen, PJ (2009). Why do antidepressants take so long to work? A cognitive neuropsychological model of antidepressant drug action. British Journal of Psychiatry: the Journal of Mental Science 195, 102–108.CrossRefGoogle Scholar
Hautzinger, M, Bailer, M, Worall, H, Keller, F (1994). Beck Depressions-Inventar (BDI). Testhandbuch. Hans Huber: Bern.Google Scholar
Joormann, J, Cooney, RE, Henry, ML, Gotlib, IH (2011). Neural correlates of automatic mood regulation in girls at high risk for depression. Journal of Abnormal Psychology 121, 61–72.CrossRefGoogle ScholarPubMed
Jorgensen, A, Magnusson, P, Hanson, LG, Kirkegaard, T, Benveniste, H, Lee, H, Svarer, C, Mikkelsen, JD, Fink-Jensen, A, Knudsen, GM, Paulson, OB, Bolwig, TG, Jorgensen, MB (2015). Regional brain volumes, diffusivity, and metabolite changes after electroconvulsive therapy for severe depression. Acta Psychiatrica Scandinavica 133, 154–164.CrossRefGoogle ScholarPubMed
Joshi, SH, Espinoza, RT, Pirnia, T, Shi, J, Wang, Y, Ayers, B, Leaver, A, Woods, RP, Narr, KL (2016). Structural plasticity of the hippocampus and amygdala induced by electroconvulsive therapy in major depression. Biological Psychiatry 79, 282–292.CrossRefGoogle ScholarPubMed
Kho, KH, van Vreeswijk, MF, Simpson, S, Zwinderman, AH (2003). A meta-analysis of electroconvulsive therapy efficacy in depression. Journal of ECT 19, 139–147.CrossRefGoogle ScholarPubMed
Kuhs, H (1995). [Stages of treatment resistance in depressive disorders, defined after somatotherapeutic methods]. Der Nervenarzt 66, 561–567.Google ScholarPubMed
Leaver, AM, Espinoza, R, Pirnia, T, Joshi, SH, Woods, RP, Narr, KL (2016). Modulation of intrinsic brain activity by electroconvulsive therapy in major depression. Biological Psychiatry 1, 77–86.Google ScholarPubMed
Liu, Y, Du, L, Li, Y, Liu, H, Zhao, W, Liu, D, Zeng, J, Li, X, Fu, Y, Qiu, H, Li, X, Qiu, T, Hu, H, Meng, H, Luo, Q (2015). Antidepressant effects of electroconvulsive therapy correlate with subgenual anterior cingulate activity and connectivity in depression. Medicine 94, e2033.CrossRefGoogle ScholarPubMed
Nickl-Jockschat, T, Palomero Gallagher, N, Kumar, V, Hoffstaedter, F, Brügmann, E, Habel, U, Eickhoff, SB, Grözinger, M (2016). Are morphological changes necessary to mediate the therapeutic effects of electroconvulsive therapy? European Archives of Psychiatry and Clinical Neuroscience 266, 261–267.CrossRefGoogle ScholarPubMed
Njau, S, Joshi, SH, Espinoza, R, Leaver, AM, Vasavada, M, Marquina, A, Woods, RP, Narr, KL (2016). Neurochemical correlates of rapid treatment response to electroconvulsive therapy in patients with major depression. Journal of Psychiatry and Neuroscience 41, 150–177.Google Scholar
Norbury, R, Taylor, MJ, Selvaraj, S, Murphy, SE, Harmer, CJ, Cowen, PJ (2009). Short-term antidepressant treatment modulates amygdala response to happy faces. Psychopharmacology 206, 197–204.CrossRefGoogle ScholarPubMed
Opel, N, Redlich, R, Grotegerd, D, Dohm, K, Haupenthal, C, Heindel, W, Kugel, H, Arolt, V, Dannlowski, U (2015). Enhanced neural responsiveness to reward associated with obesity in the absence of food-related stimuli. Human Brain Mapping 36, 2330–2337.CrossRefGoogle ScholarPubMed
Redlich, R, Almeida, JRC, Grotegerd, D, Opel, N, Kugel, H, Heindel, W, Arolt, V, Phillips, ML, Dannlowski, U (2014). Brain morphometric biomarkers distinguishing unipolar and bipolar depression: a voxel-based morphometry-pattern classification approach. JAMA Psychiatry 71, 1222–1230.CrossRefGoogle ScholarPubMed
Redlich, R, Dohm, K, Grotegerd, D, Opel, N, Zwitserlood, P, Heindel, W, Arolt, V, Kugel, H, Dannlowski, U (2015 a). Reward processing in unipolar and bipolar depression: a functional MRI study. Neuropsychopharmacology 40, 2623–2631.CrossRefGoogle ScholarPubMed
Redlich, R, Grotegerd, D, Opel, N, Kaufmann, C, Zwitserlood, P, Kugel, H, Heindel, W, Donges, US, Suslow, T, Arolt, V, Dannlowski, U (2015 b). Are you gonna leave me? Separation anxiety is associated with increased amygdala responsiveness and volume. Social Cognitive and Affective Neuroscience 10, 278–284.CrossRefGoogle ScholarPubMed
Redlich, R, Opel, N, Grotegerd, D, Dohm, K, Zaremba, D, Bürger, C, Münker, S, Mühlmann, L, Wahl, P, Heindel, W, Arolt, V, Alferink, J, Zwanzger, P, Zavorotnyy, M, Kugel, H, Dannlowski, U (2016). Prediction of individual response to electroconvulsive therapy via machine learning on structural magnetic resonance imaging data. JAMA Psychiatry 73, 557–564.CrossRefGoogle ScholarPubMed
Redlich, R, Stacey, D, Opel, N, Grotegerd, D, Dohm, K, Kugel, H, Heindel, W, Arolt, V, Baune, BT, Dannlowski, U (2015 c). Evidence of an IFN-γ by early life stress interaction in the regulation of amygdala reactivity to emotional stimuli. Psychoneuroendocrinology 62, 166–173.CrossRefGoogle ScholarPubMed
Salvadore, G, Cornwell, BR, Colon-Rosario, V, Coppola, R, Grillon, C, Zarate, CA, Manji, HK (2009). Increased anterior cingulate cortical activity in response to fearful faces: a neurophysiological biomarker that predicts rapid antidepressant response to ketamine. Biological Psychiatry 65, 289–295.CrossRefGoogle Scholar
Sheline, YI, Barch, DM, Donnelly, JM, Ollinger, JM, Snyder, AZ, Mintun, MA (2001). Increased amygdala response to masked emotional faces in depressed subjects resolves with antidepressant treatment: an fMRI study. Biological Psychiatry 50, 651–658.CrossRefGoogle ScholarPubMed
Siegle, GJ, Carter, CS, Thase, ME (2006). Use of fMRI to predict recovery from unipolar depression with cognitive behavior therapy. American Journal of Psychiatry 163, 735–738.CrossRefGoogle ScholarPubMed
Siegle, GJ, Thompson, WK, Carter, CS, Steinhauer, SR, Thase, ME (2007). Increased amygdala and decreased dorsolateral prefrontal BOLD responses in unipolar depression: related and independent features. Biological Psychiatry 61, 198–209.CrossRefGoogle ScholarPubMed
Stuhrmann, A, Dohm, K, Kugel, H, Zwanzger, P, Redlich, R, Grotegerd, D, Rauch, AV, Arolt, V, Heindel, W, Suslow, T, Zwitserlood, P, Dannlowski, U (2013). Mood-congruent amygdala responses to subliminally presented facial expressions in major depression: associations with anhedonia. Journal of Psychiatry and Neuroscience 37, 249–258.CrossRefGoogle Scholar
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 (2010). Automatic mood-congruent amygdala responses to masked facial expressions in major depression. Biological Psychiatry 67, 155–160.CrossRefGoogle ScholarPubMed
Suslow, T, Kugel, H, Ohrmann, P, Stuhrmann, A, Grotegerd, D, Redlich, R, Bauer, J, Dannlowski, U (2013). Neural correlates of affective priming effects based on masked facial emotion: an fMRI study. Psychiatry Research 211, 239–245.CrossRefGoogle ScholarPubMed
Taghva, AS, Malone, DA, Rezai, AR (2013). Deep brain stimulation for treatment-resistant depression. World Neurosurgery 80, S27.e17–S27.e24.CrossRefGoogle ScholarPubMed
Tendolkar, I, van Beek, M, van Oostrom, I, Mulder, M, Janzing, J, Voshaar, RO, van Eijndhoven, P (2013). Electroconvulsive therapy increases hippocampal and amygdala volume in therapy refractory depression: a longitudinal pilot study. Psychiatry Research 214, 197–203.CrossRefGoogle ScholarPubMed
Tzourio-Mazoyer, N, Landeau, B, Papathanassiou, D, Crivello, F, Etard, O, Delcroix, N, Mazoyer, B, Joliot, M (2002). Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. NeuroImage 15, 273–289.CrossRefGoogle ScholarPubMed
van Harmelen, A-L, van Tol, M-J, Demenescu, LR, van der Wee, NJA, Veltman, DJ, Aleman, A, van Buchem, MA, Spinhoven, P, Penninx, B, Elzinga, BM (2013). Enhanced amygdala reactivity to emotional faces in adults reporting childhood emotional maltreatment. Social Cognitive and Affective Neuroscience 8, 362–369.CrossRefGoogle ScholarPubMed
van Waarde, JA, Scholte, HS, van Oudheusden, LJB, Verwey, B, Denys, D, van Wingen, GA (2015). A functional MRI marker may predict the outcome of electroconvulsive therapy in severe and treatment-resistant depression. Molecular Psychiatry 20, 609–614.CrossRefGoogle ScholarPubMed
Victor, TA, Furey, ML, Fromm, SJ, Ohman, A, Drevets, WC (2010). Relationship between amygdala responses to masked faces and mood state and treatment in major depressive disorder. Archives of General Psychiatry 67, 1128–1138.CrossRefGoogle ScholarPubMed
Williams, LM, Korgaonkar, MS, Song, YC, Paton, R, Eagles, S, Goldstein-Piekarski, A, Grieve, SM, Harris, AW, Usherwood, T, Etkin, A (2015). Amygdala reactivity to emotional faces in the prediction of general and medication-specific responses to antidepressant treatment in the randomized iSPOT-D trial. Neuropsychopharmacology 40, 2398–2408.CrossRefGoogle ScholarPubMed
Wittchen, H-U, Wunderlich, U, Gruschwitz, S, Zaudig, M (1997). Strukturiertes Klinisches Interview für DSM-IV. Hogrefe: Goettingen.Google Scholar
Figure 0

Table 1. Sociodemographic and clinical characteristics

Figure 1

Fig. 1. Treatment effects of electroconvulsive therapy (ECT) on amygdala reactivity during emotion processing. Left: coronal view (Montreal Neurological Institute coordinates at y = 1) depicting the interaction effect of group × condition × time within the bilateral amygdala (left: F2,220 = 6.03, p = 0.003; right: F2,220 = 5.77, p = 0.004), resulting from a decrease in amygdala reactivity to sad faces in both patient samples [ECT: Z = 2.26, p = 0.012; medication (MED), right: Z = 3.28, p = 0.001; left: Z = 3.06, p = 0.001], and an increase in amygdala activity to positive faces in the MED sample (Z = 2.09, p = 0.018). The colour bar shows F values. Right: bar graphs depict the estimated contrast values subdivided by group [ECT, MED, healthy control HC)] at T1 (orange) and at T2 (blue) for the negative > neutral (above) and positive > neutral (below) conditions extracted from peak values of the interaction analysis. * Significant differences (Alpha-Sim corrected, voxel threshold, p < 0.05; minimum cluster volume threshold k = 33 voxels). fMRI, Functional magnetic resonance imaging.

Figure 2

Fig. 2. Associations between functional magnetic resonance imaging (fMRI) contrast value changes in amygdala reactivity and changes in Beck Depression Inventory scores, displayed separately for the contrast negative > neutral (above) and positive > neutral (below) subdivided by treatment condition [left: electroconvulsive therapy (ECT) sample; right: medication (MED) sample]. The graphs depict a significant negative association between amygdala activity changes to sad faces and symptom relief in the ECT sample (rspearman = −0.48, p = 0.044), and by tendency for the MED sample (rspearman = −0.38, p = 0.098). In contrast, a positive association was found between amygdala activity changes to positive faces and symptomatic improvements for the ECT sample (rspearman = 0.52, p = 0.027).