Investigating the Neural Substrates of Depersonalization Symptoms in Major Depressive Disorder
Elisabeth Paul1 11082291
1Programme Group Clinical Psychology, Faculty of Social and Behavioural Sciences, University of Amsterdam, the Netherlands
Short title: Neural Substrates of Depersonalization Symptoms in MDD
Manuscript characteristics:
Number of words in abstract: 230 Number of words in text: 3999 Number of references: 61 Number of figures: 4 Number of tables: 3
Supplemental information: 0
Supervisor: Dr. Henk Cremers Second Assessor: Dr. Heleen Slagter
Daily Supervisor: Dr. Paul Hamilton
Center for Social and Affective Neuroscience, Department of Clinical and Experimental Medicine, Linköping University, Sweden
Table of Content
1. Abstract 3
2. Introduction 4
a. Altered Functional Connectivity Between the Extrastriate Body Area 4 and DMN
b. Altered Connectivity Between the Hippocampus and Primary Nodes of 5 the Default Mode Network
c. Altered Functional Connectivity Between the Medial Prefrontal Cortex 6 and Striatum
d. Altered Functional Connectivity Within the Insular Cortex 6
e. Summary 7
3. Methods and Materials 8
a. Participants 8
b. Resting-State fMRI 9
c. Correlation Analysis Between Functional Connectivity and Clinical 10 Questionnaires
4. Results 12
a. Correlations Among Clinical Questionnaires 12 b. Correlations Among Functional Connectivity and Depersonalization 12
Metrics
5. Discussion 14
a. Limitations and Future Research 16
b. Conclusion 17
6. Acknowledgements 19
7. References 20
8. Tables and Figures 24
a. Figure 1. Seed and Target Regions 24
b. Figure 2. Flow-chart of the Analysis Steps 25 c. Figure 3. Activity Map of the Correlation between Extrastriate Body 26
Area - Default Mode Network and the first principal component of the Cambridge Depersonalization Scale
d. Figure 4. Activity Map of the Correlation between Extrastriate Body 27 Area - Default Mode Network and the Cambridge Depersonalization
Scale
e. Table 1. Descriptive Statistics 28
f. Table 2. Correlations between Clinical Questionnaire Scores 29 g. Table 3. Correlation Analysis Between Cambridge Depersonalization 30
Abstract
Background: Depersonalization/derealisation disorder (DPD) is a dissociative disorder
characterized by feelings of unreality and detachment from the self and surroundings. Based on its high comorbidity with mood and anxiety disorders, with highest prevalence in inpatients, and its association with treatment resistance and longer duration of depressive episodes, it is assumed that DPD might serve as a severity index of these disorders. This study proposes and tests four hypotheses about alterations in functional connectivity in the brain that might elicit DPD symptoms in major depressive disorder (MDD), namely between 1) extrastriate body area and default mode network (DMN), 2) hippocampus and DMN, 3) medial prefrontal cortex and striatum, and 4) posterior and anterior insular.
Methods: We used resting-state functional magnetic resonance imaging to scan 28 patients
diagnosed with MDD. Functional connectivity between seed and target regions as defined in the hypotheses was computed and correlated with scores on different scales of the Cambridge Depersonalization Scale (CDS).
Results: Reduced connectivity between the extrastriate body area and DMN bilaterally was
associated with higher scores on the CDS. We did not find support for the other hypotheses.
Conclusion: Our results show that DPD-symptoms in MDD patients are related to reduced
functional connectivity in brain regions that presumably support the integration of physical body-related information (extrastriate body area) into the sense of self (DMN), leading to symptoms of depersonalization. Limitations and future directions are discussed.
Key words: Depersonalization Disorder, Major Depressive Disorder, Extrastriate Body
Investigating the Neural Substrates of Depersonalization Symptoms in Major Depressive Disorder
Depersonalization/derealisation disorder (DPD) is a dissociative disorder characterized by detachment from thoughts, feelings, bodily experiences, and from the external world. In the
Diagnostic and Statistical Manual of Mental Disorders, 5th ed. (DSM-5; 1), DPD is listed as a primary disorder. Given its frequent comorbidities with mood and anxiety disorders, it has been proposed that DPD might, rather, be an indicator of the severity of these disorders (2-4). As a comorbidity of major depressive disorder (MDD), DPD predicts heightened treatment resistance (5, 6) and a longer duration of depressive episodes (7). Further indicating a relation between DPD and MDD severity, the prevalence of DPD in MDD varies between 4% to 60% (3, 8), with the highest prevalence in inpatients. In the present investigation we tested four neural hypotheses of DPD in major depression to better understand DPD in the context of MDD. Specifically, we examined the relation between DPD-symptoms and inter-regional functional connectivity during the resting state in patients with MDD.
This study is exploratory and the analysis was conducted stepwise, the outcome of each step guided subsequent ones. We hypothesized that one or more of four patterns of connectivity could account for variation in DPD-symptoms in persons diagnosed with major depression. Furthermore, we sought to determine if these neural models accounted for DPD-symptoms specifically or if they were also related to depression severity or anxiety.
Altered Functional Connectivity Between the Extrastriate Body Area and Primary Nodes of the Default Mode Network
Symptoms of DPD can be conceptualized as impairments in the integration of
representations of the physical body into a sense of self, i.e. embodiment (9). One region that is strongly implicated in the process of embodiment is the extrastriate body area (EBA), which
responds to imagined and seen body movements and integrates multisensory body-related information (10). In addition, various nodes of the default mode network (DMN; 11) have been implicated in representing the bodily self. The temporoparietal junction, for example, responds under conditions of self-location (10), while the process of distinguishing between self and others activates the node of the posterior cingulate (12).
Patients with MDD often report a dissociation between their physical body and the self. They describe their body as an obstacle, rather than as belonging to the self (13). A reduced integration of information about the physical body and the self could lead to these symptoms and could be related to DPD in MDD. We hypothesized, therefore, that increasing DPD-symptoms in MDD will be associated with decreased EBA – DMN connectivity.
Altered Connectivity Between the Hippocampus and Primary Nodes of the Default Mode Network
Additionally, patients with DPD often report disturbances in their ability to recognize themselves and their surroundings (14), which could be due, in part, to a failure of recognition memory. To recognize an externally or internally generated stimulus as similar to one previously encountered, current and past stimuli must be compared and determined to be familiar or
unfamiliar (15, 16). The DMN is thought to be critical for such processes (15, 16). The posterior regions of the DMN, especially the hippocampus (HC), play a primary role in the recognition of similarities between past and current stimuli and contexts (17, 18) while the anterior regions of the DMN are hypothesized to evaluate stimuli from a self-relevant reference frame (17). Variability in the strength of functional connectivity between the HC and other regions of the DMN might, therefore, subserve variability in the integration of information about the present and past into a sense of self.
Reduced functional connectivity between the HC and parts of the DMN has been observed in MDD (19, 20) and could account for a partial failure to recognize oneself in a familiar situation, which might lead to DPD-symptoms in MDD. Based on the role of the DMN in recognition memory and an apparent disruption of recognition memory processes, we
hypothesized that HC – DMN connectivity is reduced with advancing severity of DPD in MDD.
Altered Functional Connectivity Between the Medial Prefrontal Cortex and Striatum
There are also impairments in DPD in the affective valuation of stimuli (21), which might lead to feelings of emotional distancing or numbness. The assessment of the value of a stimulus is thought to be subserved by processes supported by medial prefrontal cortex (mPFC) and the striatum and communication between these regions (22). More specifically, the striatum codes the reward significance of a stimulus and the mPFC codes for the personal value of a stimulus to the individual (23). Decreased connectivity between the striatum and mPFC could, therefore, be associated with decreased valuation of external stimuli. In a psychiatric context, patients with MDD consistently demonstrate decreased valuation of rewarding stimuli (22). Reductions in functional connectivity of mPFC and striatum could, further, manifest as emotional numbing and feelings of detachment in promoting symptoms of DPD in MDD. It is therefore hypothesized that as mPFC – striatum connectivity decreases DPD-symptoms in MDD will increase.
Altered Functional Connectivity Within the Insular Cortex
Some aspects of DPD can be conceptualized as impaired interoceptive processing, in particular as impairments in achieving conscious representations of internal bodily states. DPD patients, for example, have shown poor performance on a heartbeat detection task which
measures interoceptive sensitivity (24). Based on dense innervation of ascending spino-thalamic tracts conveying information from the body (25), insular cortex (IC) is assumed to be responsible
for interoceptive awareness. Interoceptive processing and representation is thought to follow along a posterior-to-anterior insular gradient (25). Basic and fundamental representations of homeostatic processes are postulated to be formed in the posterior IC (26). The mid-insular cortex is thought to complement these fundamental representations with information about the emotional salience of bodily stimuli (27). Finally, the anterior IC connects information from posterior and mid-IC with higher cognitive processes (26, 28), leading to a conscious
representation of bodily states.
Diminished interoceptive awareness has been observed in MDD (29, 30). A reduced integration of information from sensory modalities within the IC and the resulting altered sense of body awareness could result in DPD-symptoms in patients with MDD. Since interoception is thought to be the information flow along the IC, it is hypothesized that DPD-symptoms in MDD would increase with diminishing functional connectivity along the posterior-to-anterior gradient of the IC.
Summary
In the present investigation we tested four neural-functional hypotheses of DPD in MDD, namely that reduced functional connectivity between 1) EBA – DMN 2) HC – DMN 3) mPFC – striatum and 4) posterior – anterior IC are related to increased DPD. We used resting-state functional magnetic resonance imaging (rsfMRI) data to estimate levels of inter-regional functional connectivity and correlated these functional connectivity measures against clinical self-report measures of DPD to determine the relation between the putative neural-functional substrates of DPD in major depression.
Methods and Materials Participants
Participants were recruited through advertisement at medical offices. Twenty-eight female patients (mean age M = 37, SD = 11) meeting DSM-4 criteria (31) for current primary MDD participated in the study. Exclusion criteria were standard MRI scanning contraindications (e.g., implanted ferrous metal, pregnancy, and claustrophobia). Participants had to be free of
psychotropic medications in the four weeks prior to scanning. Furthermore, the presence of another Axis-I psychopathology (other than an anxiety disorder), of suicidal ideation, intent or behaviour or of psychosis to the extent that the participant was unable to provide informed consent led to exclusion. In addition, individuals who had a history of drug abuse or head trauma, medical conditions that could influence cerebral blood flow, or special education needs were excluded. The study was approved by the ethical committee of the Western Institutional Review Board (www.wirb.com), and all participants gave written consent to the scientific use of their data.
In total eight participants were excluded from the analysis: two for not finishing the study, four due to excessive movement during the scan (> 0.2 mm per acquisition for 25% or more of the functional acquisition), one because she did not complete the CDS and one for being an outlier (> 2 SD from mean) on the CDS.
This study was part of a large-scale study for which scores on a variety of questionnaires and tasks were collected. For the study at hand, however, only scores on the Cambridge
Depersonalization Scale (CDS; 32), the Beck Depression Inventory-II (BDI-II; 33), the Beck Anxiety Inventory (BAI; 34) and rsfMRI data were used.
Resting-State fMRI
Whole-head rsfMRI scans were performed on a 3.0 Tesla General Electric Signa MR scanner (Milwaukee, USA) with a 32-channel head coil. Participants were instructed to lie still during the 6.5-minute scan and to maintain attention toward a fixation cross presented in the middle of an in-scanner display. Blood oxygen level dependent (BOLD) rsfMRI was conducted using echo planar imaging (34 axial, 2.9-mm-thick interleaved slices with 3.438 mm2 in-plane resolution; time-to-repeat [TR] = 2000 ms; time-to-echo [TE] = 30 ms; flip angle = 98°; matrix size = 64 × 64, FOV = 220 mm × 220 mm; sampling bandwidth = 250 kHz, 192 volumes acquired).
Despiking and low-pass filtering algorithms available in the Analysis of Functional NeuroImages (AFNI; 35) platform were used to de-noise the rsfMRI data. In addition, the data was corrected for influences of measurement artefacts (36), respiratory motion and cardiac pulsatility (37) and changes in respiratory (38) and heart rate (39). Further, translational and rotational motion regressors, their first derivatives and regressors for unmodeled residual noise (40) were applied to the data as regressors of no interest. Lastly, a censoring/scrubbing procedure was applied for removing residual effects of participant motion greater than 0.2 mm during a given acquisition on fMRI signal. For a more detailed description, see (41).
Seed-based functional connectivity analysis was conducted to estimate the strength of connectivity between seed and target regions, depicted in Figure 1. Eight separate 5-mm radius seed regions were defined: bilateral seeds in EBA, HC, mPFC, and posterior IC. Coordinates for these seed regions were determined, respectively, by locating peak regions of reverse-inference-based meta-analytic association with the search terms ‘body,’ ‘recognition memory,’ ‘value,’ and ‘interoceptive,’ in the Neurosynth meta-analytic database (42). Target regions were defined as
the DMN, striatum, and anterior IC. The anterior IC mask and the DMN mask were taken from previously published studies (43 and, 44 respectively). The striatum mask was created using the AFNI (35) TT Daemon for caudate and putamen.
Correlation Analysis Between Functional Connectivity and Clinical Questionnaires
We constructed a stepwise analysis procedure for the study. In Figure 2 we provide a flow-chart depicting all steps of this procedure. Due to the extensiveness of the large-scale study, only frequency of DPD-symptoms was assessed, not their duration. Several exploratory factors analyses have revealed that the CDS comprises more than one dimension. This study divided the CDS total scores into the four factors determined by Fagioli et al. (45): ‘detachment from self’ (detachment), ‘anomalous bodily experiences,’ ‘numbing,’ and ‘temporal blunting’ (blunting). We excluded from further consideration the ‘anomalous bodily experiences’ factor due to its high deviation from normality and skewedness towards 0. High correlations between the CDS subscales (r between .70 and .86) suggested a central construct in addition to some independent variation particular to each of the subscales. To be able to investigate the core of all and the independent variation of each of the subscales, we computed, using MATLAB, the first principal component of the subscales, as well as the residuals for each of the subscales relative to the first principal component. For the final set of analyses, therefore, we included the first principal component (CDScore; a representation of a central CDS construct) and the residuals for each subscale (PCnumbing, PCdetachment and PCblunting). Since not all of the items of the CDS were captured by the subscales, we additionally tested our neural connectivity hypotheses of DPD against CDS total scores (CDStotal) in a non-independent analysis.
Next, we determined if any of the CDS-based variables correlated significantly with severity of depression (as assessed with the BDI-II) and/or anxiety (as assessed with the BAI) so
that we could potentially account for this via multiple regression in testing our neural
connectivity hypotheses. We computed correlations between these measures (where BAI-scores were log-transformed to correct for non-normality) using IBM SPSS Statistics v.24.0. If the two-tailed significance was p ≤ 0.10, we included BDI-II and/or BAI scores as noise covariates in the correlation analysis between CDS scores and functional connectivity.
In the final step of the analysis, we calculated the functional connectivity between our seed and target regions and correlated it with scores of the CDS-based variables. First, for each participant, and for each seed region, we conducted seed-based functional connectivity analysis relative to target regions. All seed regions were correlated with bilateral target regions except for posterior IC, which was correlated with the ipsilateral anterior IC only, based on our hypothesis. To keep familywise error at αcorrected = .05, we used AFNI’s 3dClustSim applied to our target region masks to estimate smoothness and calculate voxel-wise and cluster thresholds (voxel-wise
p = .05, kDMN = 75, kstriatum = 34 and kanterior IC = 21). Cluster sizes were determined using ‘face-to-face’ nearest neighbour clustering and bi-sided thresholding. To then render z-statistic maps reflecting the correlations between CDS-based scores and functional connectivity we used AFNI’s 3dttest++.
As an additional validation of our findings and to further account for the exploratory character of the study, we conducted sign tests. For that we specified the number of significant negative and significant total correlations for each hypothesis per hemisphere. Next, we ran a permutation test with 1,000 simulations to calculate the probability of observing the given number of negative correlations under the assumption that negative and positive correlation clusters would be equally likely under the null hypothesis. Only if both z-values and sign tests were significant we considered the null hypothesis as rejected.
Results
Table 1 shows descriptive statistics of the CDS, BDI-II and BAI. For comparison, data from a demographically matched healthy control sample are shown.
Correlations Among Clinical Questionnaires
After extracting the first principal component and calculating the residuals of the CDS subscales, we correlated all CDS-based variables with each other and with BDI-II and BAI (log-transformed) to determine if the CDS-based variables were significantly associated with
depression or anxiety. As shown in Table 2, PCnumbing and BDI-II were correlated at a marginally significant level. We therefore included BDI-II as a noise covariate in the correlation analysis between functional connectivity and PCnumbing in the primary analysis.
As we were interested in exploring whether the neural hypotheses are specific to one or another subscale or to CDS-scores in total, we decided to include PCcore and the residuals of the subscales even though PCcore accounted for 98% of the variability. We also included CDStotal, despite the high correlation of CDStotal and PCcore. While the subscales did not cover all items of the CDS and we excluded anomalous bodily experiences from the principal component analysis, CDStotal captured all items of the questionnaire, potentially adding information.
Correlations Among Functional Connectivity and Depersonalization Metrics
In Table 3 we summarize the findings of the correlation analyses among CDS variables and the functional connectivity indices. We see relatively clear support in favour of the
hypothesis that a reduction in EBA – DMN functional connectivity accounts for general symptoms of depersonalization that can occur with depression.
Extrastriate body area-to-default mode network connectivity hypothesis. We found
significant negative relations between CDScore and both left (p = .03) and right (p = .02) EBA – DMN connectivity. These findings are presented in correlation maps in Figure 3.
In addition, connectivity between the left EBA and the left DMN was positively related to PCdetachment, however the sign test was not significant (p = .5).
Right EBA – DMN connectivity was, furthermore, related to PCnumbing (accounted for BDI-II scores) where connectivity between EBA and one node of the DMN was positively related, while connectivity between EBA and another node of the DMN was negatively related to the questionnaire scores. The sign test revealed this finding as non-significant (p = .5).
Additionally, reduced functional connectivity between left and right EBA and DMN nodes was associated with increased CDStotal. The sign test indicated that it is unlikely to find this result when in reality there is no negative association between EBA – DMN connectivity and CDStotal (pleft = .03; marginally significant pright = .06). In Figure 4 we present correlation maps
depicting these relations.
Hippocampus-to-default mode network connectivity hypothesis. We observed that
decreasing connectivity between the left HC and a node of the DMN was associated with increasing PCdetachment. However, this finding did not pass the sign test criterion (p = .5).
Medial prefrontal cortex-to-striatum connectivity hypothesis. We found a positive
relation between left mPFC and left striatum connectivity and PCblunting and between right mPFC and right striatum connectivity and PCblunting , however, the sign tests for both findings were non-significant (both p = .5). In addition we found a negative relation between reduced right mPFC – bilateral striatum connectivity and PCdetachment, but this finding, too, was not significant (p = .25).
Posterior-to-anterior insula connectivity hypothesis. We did not find any significant
relations between posterior – anterior IC connectivity and any of the CDS-based variables.
Discussion
In the present study, we set out to investigate the neural substrates of depersonalization symptoms in depression. We found that decreasing connectivity between bilateral EBA and DMN significantly predicted increasing depersonalization as indexed both by CDScore and CDStotal. Our other neural-functional hypotheses of depersonalization in MDD were, however, not supported by the present results.
Our results indicate that depersonalization symptoms in MDD are related to a reduction of functional connectivity between the EBA and the DMN nodes in the frontal and temporal gyrus as well as posterior cingulate cortex/precuneus. Even though strong conclusions must not be drawn based on functions of brain regions, considering these functions can aid our
understanding of how reduced functional connectivity between the EBA and DMN might constitute a neural substrate of depersonalization symptoms. The EBA responds to images of human body parts (46) but also to one’s own bodily movement – irrespective of whether the movement is actually performed or only imagined (47), suggesting that the EBA integrates information from multiple senses. The DMN is postulated to encode information in terms of an egocentric frame of reference (16, 17), generating a conscious presentation of the self (48). Reduced functional connectivity between EBA and DMN might, therefore, lead to a reduced integration of physical body-related information (EBA) into the sense of self (DMN), leading to symptoms of depersonalization.
Interestingly, even though there is much literature about the DMN in MDD, we could not find any literature implicating that EBA is involved in MDD, and also Neurosynth (42) does not
show changes in activation in the EBA using search terms related to depression. This may indicate that our finding is specific to DPD.
While we hypothesized that HC – DMN, mPFC – striatum and posterior – anterior IC connectivity could also predict DPD in MDD, the data did not bear out any of these hypotheses. This is surprising given that these regions have consistently been implicated in MDD and DPD, as we detail below.
Reduced functional connectivity between the HC and nodes of the DMN has been observed in patients with recurrent MDD (20), first-episode MDD (19) and, moreover, has been associated with depressive symptoms in temporal lobe epilepsy (49). Additionally, DPD has also been associated with alterations in HC activity: compared to healthy controls, patients with DPD showed decreased HC activity during an emotional verbal memory test (50). Moreover,
dissociative effects of ketamine, which resemble DPD-symptoms, are related to HC activity (51). Indirect support for a reduced functional connectivity between the mPFC and striatum in MDD comes from findings implicating increased mPFC activity during the resting-state (52) while striatal activity has been found to be reduced in MDD patients compared to healthy controls (53, 54). In DPD, striatal activity to facial expressions distinguishes patients and controls (2). In addition, psychopharmacological studies using positron emission tomography found a significant correlation between depersonalization symptoms triggered by psilocybin-induced psychosis and increases in ventral striatal dopamine levels (55).
With respect to the posterior – anterior IC connectivity hypothesis, the MDD literature points to reductions in IC volume in individuals with MDD compared to healthy controls (56, 57). Moreover, MDD patients show impairments in some functions of the IC such as disgust recognition (56) and emotional awareness (58). Evidence for alterations in IC connectivity in
DPD comes from a functional neuroimaging study which investigated subjective experience of emotion during viewing of negative and neutral pictures in patients with DPD, patients with obsessive compulsive disorder and in healthy controls (59). DPD patients rated the pictures as less emotional and showed less IC activity compared to the other groups.
The fact that we did not find any relationship between these neural-functional substrates and CDS variables could indicate that these alterations in functional connectivity are specific to MDD and are not related to DPD. Due to the high comorbidity of DPD with MDD, it is difficult to assess primary DPD independently. Most studies compare brain activation in DPD patients and healthy controls (2, 50, 59) without taking depression into account in their analysis, although the patients do show depressive symptoms (e.g. in this study (59), DPD patients had an average score of 23.5 on the BDI, indicating moderate depression). Therefore, these findings might be specific to MDD rather than to DPD. We also did not find any relationship with the CDS subscale residuals. This could be due to the first principal component accounting for 98% of the variability, leaving almost no variability unexplained. Other explanations are discussed in the limitations section.
Limitations and Future Directions
Limitations of the current study ought to be considered. We could only measure frequency, not duration of depersonalization symptoms. CDS scores including both frequency and duration could have captured DPD-symptoms in MDD more accurately. However, since the duration scale is only to be answered if the participant scores at least ‘1’ on the frequency scale, we think that including the duration scale would only have strengthened but not compromised our findings.
Statistical power of this must also be kept in mind: Due to the small sample size and the multiple comparison corrections, effect sizes had to be very high for an effect to be considered significant. Moreover, we based our hypotheses on assumed functions of brain regions.
However, when using Neurosynth (42) and our search terms, we found activation in other brain regions related to these terms as well. Future studies might want to extend seed and target regions to the areas indicated by Neurosynth (42) and thus reveal a relation between the other neural-functional hypotheses we have proposed and CDS scores. As cluster-wise analysis can inflate the false-positive rate (60) caution should be exercised. Other methods like voxel-wise independent component analysis without predefined regions of interest ought to be used to replicate our findings
Due to the exploratory character of the study, its results have to be interpreted with caution, of course. However, it lays scientific ground for future studies investigating EBA – DMN connectivity in more detail. Of special interest is to study whether our findings are specific to depersonalization symptoms in MDD only or if they can also be found in DPD without
comorbid MDD. It would also be interesting to test if our findings hold when DPD-symptoms are simulated in healthy controls, for example through ketamine administration (51, 61), or to examine the EBA – DMN connectivity in MDD without depersonalization symptoms.
More knowledge about the underlying neural substrates of DPD in MDD could help to sculpt treatment. For example, transcranial magnetic stimulation therapy or real-time
neurofeedback techniques might be used to increase EBA – DMN activity coupling.
Conclusion
Despite the frequent comorbidity of DPD and MDD, little is known about the neural-functional substrates of depersonalization symptoms in depression. This study presented several
hypotheses of how variability in functional connectivity could account for depersonalization symptoms in MDD. The results presented support the formulation that reduced connectivity between the EBA and DMN is related to DPD-symptoms in MDD. Even though future research is needed, the results of this study could be considered a first step to improve treatment for MDD patients with DPD-symptoms.
Acknowledgements
We gratefully acknowledge the Warren Foundation, the Swedish Research Council and Region Östergötland for their generous support of the research presented.
References
1. Association AP (2013): Diagnostic and statistical manual of mental disorders
(DSM-5®). American Psychiatric Pub.
2. Lemche E, Surguladze SA, Brammer MJ, Phillips ML, Sierra M, David AS, et al. (2016): Dissociable brain correlates for depression, anxiety, dissociation, and somatization in
depersonalization-derealization disorder. CNS Spectr. 21:35-42.
3. Mula M, Pini S, Cassano GB (2007): The neurobiology and clinical significance of depersonalization in mood and anxiety disorders: a critical reappraisal. J Affect Disord. 99:91-99. 4. Organization WH (1992): The ICD-10 classification of mental and behavioural
disorders: clinical descriptions and diagnostic guidelines. Geneva: World Health Organization.
5. Noyes R, Jr., Kletti R (1977): Depersonilization in response to life-threatening danger.
Compr Psychiatry. 18:375-384.
6. Sedman G, Reed GF (1963): Depersonalization phenomena in obsessional personalities and in depression. Br J Psychiatry. 109:376-379.
7. Ackner B, Grant QA, Maxwell AE (1960): The prognostic significance of
depersonalization in depressive illnesses treated with electroconvulsive therapy. J Neurol
Neurosurg Psychiatry. 23:242-246.
8. Hunter EC, Sierra M, David AS (2004): The epidemiology of depersonalisation and derealisation. A systematic review. Soc Psychiatry Psychiatr Epidemiol. 39:9-18.
9. Michal M, Reuchlein B, Adler J, Reiner I, Beutel ME, Vogele C, et al. (2014): Striking discrepancy of anomalous body experiences with normal interoceptive accuracy in
depersonalization-derealization disorder. PLoS One. 9:e89823.
10. Arzy S, Thut G, Mohr C, Michel CM, Blanke O (2006): Neural basis of embodiment: distinct contributions of temporoparietal junction and extrastriate body area. J Neurosci. 26:8074-8081.
11. Molnar-Szakacs I, Uddin LQ (2013): Self-processing and the default mode network: interactions with the mirror neuron system. Front Hum Neurosci. 7:571.
12. van Buuren M, Gladwin TE, Zandbelt BB, Kahn RS, Vink M (2010): Reduced functional coupling in the default-mode network during self-referential processing. Hum Brain Mapp. 31:1117-1127.
13. Aho KA (2013): Depression and embodiment: phenomenological reflections on motility, affectivity, and transcendence. Med Health Care Philos. 16:751-759.
14. Guralnik O, Schmeidler J, Simeon D (2000): Feeling unreal: cognitive processes in depersonalization. Am J Psychiatry. 157:103-109.
15. Buckner RL, Andrews-Hanna JR, Schacter DL (2008): The brain's default network: anatomy, function, and relevance to disease. Ann N Y Acad Sci. 1124:1-38.
16. Spreng RN, Mar RA, Kim AS (2009): The common neural basis of autobiographical memory, prospection, navigation, theory of mind, and the default mode: a quantitative meta-analysis. Journal of cognitive neuroscience. 21:489-510.
17. Coutinho JF, Fernandesl SV, Soares JM, Maia L, Goncalves OF, Sampaio A (2016): Default mode network dissociation in depressive and anxiety states. Brain Imaging Behav. 10:147-157.
18. Maren S, Holt W (2000): The hippocampus and contextual memory retrieval in Pavlovian conditioning. Behavioural brain research. 110:97-108.
19. Geng H, Wu F, Kong L, Tang Y, Zhou Q, Chang M, et al. (2016): Disrupted Structural and Functional Connectivity in Prefrontal-Hippocampus Circuitry in First-Episode Medication-Naive Adolescent Depression. PLoS One. 11:e0148345.
20. Tahmasian M, Knight DC, Manoliu A, Schwerthoffer D, Scherr M, Meng C, et al. (2013): Aberrant intrinsic connectivity of hippocampus and amygdala overlap in the fronto-insular and dorsomedial-prefrontal cortex in major depressive disorder. Front Hum Neurosci. 7:639.
21. Phillips ML, Sierra M (2003): Depersonalization disorder: a functional neuroanatomical perspective. Stress. 6:157-165.
22. Eshel N, Roiser JP (2010): Reward and punishment processing in depression. Biol
Psychiatry. 68:118-124.
23. Phan KL, Taylor SF, Welsh RC, Ho SH, Britton JC, Liberzon I (2004): Neural correlates of individual ratings of emotional salience: a trial-related fMRI study. Neuroimage. 21:768-780. 24. Sedeno L, Couto B, Melloni M, Canales-Johnson A, Yoris A, Baez S, et al. (2014): How do you feel when you can't feel your body? Interoception, functional connectivity and emotional processing in depersonalization-derealization disorder. PLoS One. 9:e98769.
25. Craig AD (2009): How do you feel--now? The anterior insula and human awareness. Nat
Rev Neurosci. 10:59-70.
26. Deen B, Pitskel NB, Pelphrey KA (2011): Three systems of insular functional connectivity identified with cluster analysis. Cereb Cortex. 21:1498-1506.
27. Menon V, Levitin DJ (2005): The rewards of music listening: response and physiological connectivity of the mesolimbic system. Neuroimage. 28:175-184.
28. Chang LJ, Yarkoni T, Khaw MW, Sanfey AG (2013): Decoding the role of the insula in human cognition: functional parcellation and large-scale reverse inference. Cereb Cortex. 23:739-749.
29. Avery JA, Drevets WC, Moseman SE, Bodurka J, Barcalow JC, Simmons WK (2014): Major depressive disorder is associated with abnormal interoceptive activity and functional connectivity in the insula. Biol Psychiatry. 76:258-266.
30. Paulus MP, Stein MB (2010): Interoception in anxiety and depression. Brain Struct
Funct. 214:451-463.
31. Association AP Diagnostic and statistical manual of mental disorders. 4. Washington, DC: American Psychiatric Association; 2000. Text revision.589-594.
32. Sierra M, Berrios GE (2000): The Cambridge Depersonalization Scale: a new instrument for the measurement of depersonalization. Psychiatry Res. 93:153-164.
33. Beck AT, Steer RA, Brown GK (1996): Beck depression inventory-II. San Antonio. 78:490-498.
34. Beck AT, Epstein N, Brown G, Steer RA (1988): An inventory for measuring clinical anxiety: psychometric properties. Journal of consulting and clinical psychology. 56:893. 35. Cox RW (1996): AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Computers and Biomedical research. 29:162-173.
36. Jo HJ, Saad ZS, Simmons WK, Milbury LA, Cox RW (2010): Mapping sources of correlation in resting state FMRI, with artifact detection and removal. Neuroimage. 52:571-582. 37. Glover GH, Li TQ, Ress D (2000): Image-based method for retrospective correction of physiological motion effects in fMRI: RETROICOR. Magn Reson Med. 44:162-167.
38. Birn RM, Smith MA, Jones TB, Bandettini PA (2008): The respiration response function: the temporal dynamics of fMRI signal fluctuations related to changes in respiration. Neuroimage. 40:644-654.
39. Chang C, Cunningham JP, Glover GH (2009): Influence of heart rate on the BOLD signal: the cardiac response function. Neuroimage. 44:857-869.
40. Marx M, Pauly KB, Chang C (2013): A novel approach for global noise reduction in resting-state fMRI: APPLECOR. Neuroimage. 64:19-31.
41. Frost Bellgowan J, Molfese P, Marx M, Thomason M, Glen D, Santiago J, et al. (2015): A neural substrate for behavioral inhibition in the risk for major depressive disorder. J Am Acad
Child Adolesc Psychiatry. 54:841-848.
42. Yarkoni T, Poldrack RA, Nichols TE, Van Essen DC, Wager TD (2011): Large-scale automated synthesis of human functional neuroimaging data. Nat Methods. 8:665-670.
43. Hamilton JP, Furman DJ, Chang C, Thomason ME, Dennis E, Gotlib IH (2011): Default-mode and task-positive network activity in major depressive disorder: implications for adaptive and maladaptive rumination. Biol Psychiatry. 70:327-333.
44. Yeo BT, Krienen FM, Sepulcre J, Sabuncu MR, Lashkari D, Hollinshead M, et al. (2011): The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J Neurophysiol. 106:1125-1165.
45. Fagioli F, Telesforo L, Dell'Erba A, Consolazione M, Migliorini V, Patane M, et al. (2015): Depersonalization: An exploratory factor analysis of the Italian version of the Cambridge Depersonalization Scale. Compr Psychiatry. 60:161-167.
46. Downing PE, Jiang Y, Shuman M, Kanwisher N (2001): A cortical area selective for visual processing of the human body. Science. 293:2470-2473.
47. Astafiev SV, Stanley CM, Shulman GL, Corbetta M (2004): Extrastriate body area in human occipital cortex responds to the performance of motor actions. Nat Neurosci. 7:542-548. 48. Davey CG, Pujol J, Harrison BJ (2016): Mapping the self in the brain's default mode network. Neuroimage. 132:390-397.
49. Kemmotsu N, Kucukboyaci NE, Cheng CE, Girard HM, Tecoma ES, Iragui VJ, et al. (2013): Alterations in functional connectivity between the hippocampus and prefrontal cortex as a correlate of depressive symptoms in temporal lobe epilepsy. Epilepsy Behav. 29:552-559. 50. Medford N, Brierley B, Brammer M, Bullmore ET, David AS, Phillips ML (2006): Emotional memory in depersonalization disorder: a functional MRI study. Psychiatry Res. 148:93-102.
51. Lahti AC, Holcomb HH, Medoff DR, Tamminga CA (1995): Ketamine activates psychosis and alters limbic blood flow in schizophrenia. Neuroreport. 6:869-872.
52. Lemogne C, Delaveau P, Freton M, Guionnet S, Fossati P (2012): Medial prefrontal cortex and the self in major depression. J Affect Disord. 136:e1-e11.
53. Epstein J, Pan H, Kocsis JH, Yang Y, Butler T, Chusid J, et al. (2006): Lack of ventral striatal response to positive stimuli in depressed versus normal subjects. Am J Psychiatry. 163:1784-1790.
54. Forbes EE, Hariri AR, Martin SL, Silk JS, Moyles DL, Fisher PM, et al. (2009): Altered striatal activation predicting real-world positive affect in adolescent major depressive disorder.
Am J Psychiatry. 166:64-73.
55. Vollenweider FX, Vontobel P, Hell D, Leenders KL (1999): 5-HT modulation of dopamine release in basal ganglia in psilocybin-induced psychosis in man--a PET study with [11C]raclopride. Neuropsychopharmacology. 20:424-433.
56. Sprengelmeyer R, Steele JD, Mwangi B, Kumar P, Christmas D, Milders M, et al. (2011): The insular cortex and the neuroanatomy of major depression. Journal of affective
disorders. 133:120-127.
57. Stratmann M, Konrad C, Kugel H, Krug A, Schöning S, Ohrmann P, et al. (2014): Insular and hippocampal gray matter volume reductions in patients with major depressive disorder.
PLoS One. 9:e102692.
58. Kranzler A, Young JF, Hankin BL, Abela JR, Elias MJ, Selby EA (2016): Emotional Awareness: A Transdiagnostic Predictor of Depression and Anxiety for Children and
Adolescents. J Clin Child Adolesc Psychol. 45:262-269.
59. Phillips ML, Medford N, Senior C, Bullmore ET, Suckling J, Brammer MJ, et al. (2001): Depersonalization disorder: thinking without feeling. Psychiatry Research: Neuroimaging. 108:145-160.
60. Eklund A, Nichols TE, Knutsson H (2016): Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates. Proc Natl Acad Sci U S A. 113:7900-7905. 61. Pikwer A (2011): Depersonalization disorder may be related to glutamate receptor activation imbalance. Med Hypotheses. 77:593-594.
Tables and Figures
Figure 1. Seed (yellow) and target (green) regions of the functional connectivity analysis.
Regions were (seeds are mentioned first) a) extrastriate body area and default mode network b) hippocampus and default mode network c) medial prefrontal cortex and striatum and d) posterior and anterior insular cortex.
Figure 2. Flow-chart depicting all analysis steps.
Note: CDStotal = total score on the Cambridge Depersonalization Scale; CDScore = principal component of CDS subscales projected back on the subscales; PCnumbing = residuals of the principal component for CDS subscale numbing; PCdetachment = residuals of the principal component for CDS subscale detachment; PCblunting = residuals of the principal component for CDS subscale blunting, BDI-II = Beck Depression Inventory II, BAI = Beck Anxiety Inventory (log-transformed).
Figure 3. Relation between left (A) and right (B) extrastriate body area and default mode
network connectivity and the first principal component of the Cambridge Depersonalization subscales numbing, temporal blunting and detachment from bodily experiences (CDScore). The red mask depicts the default mode network mask as used in this experiment, yellow displays the nodes that were significantly related to the principal component. The orange circle indicates the location of the extrastriate body area mask.
Figure 4. Relation between left (A) and right (B) extrastriate body area and default mode
network connectivity and the total score on the Cambridge Depersonalization Scale (CDStotal). The red mask depicts the default mode network mask as used in this experiment, yellow displays the nodes that were significantly related to the questionnaire score. The orange circle indicates the location of the extrastriate body area mask.
Table 1
Descriptive Statistics of Cambridge Depersonalization Scale, Beck Depression Inventory II and Beck Anxiety Inventory in MDD and a Healthy Control Group.
MDD (N = 20) Healthy Control Group (N = 24)
Questionnaire M (SD) min max M (SD) min max
CDStotal 19.35 (13.66) 1 42 2.33 (4.44) 0 20
BDI-II 23.6 (11.37) 6 46 .67 (1.63) 0 6
BAI 14.75 (10.58) 1 43 2.08 (2.5) 0 9
Note. CDStotal = Cambridge Depersonalization Scale; BDI-II = Beck Depression Inventory II, BAI = Beck Anxiety Inventory
Table 2
Correlations Between Scores on Clinical Questionnaires.
CDStotal CDScore PCnumbing PCdetachment PCblunting BDI-II BAI CDStotal .99* .00 -.06 .07 .34 .26 CDScore .99* .00 .00 .00 .32 .27 PCnumbing .00 .00 -.56* -.06 .39* .29 PCdetachment -.06 .00 -.56* -.79* -.19 -.21 PCblunting .07 .00 -.06 -.79* -.06 .04 BDI-II .34 .32 .39* -.19 -.06 .30 BAI .26 .27 .29 -.21 .04 .30
Note. Asterisk (*) indicates a significant correlation at p = .10. CDStotal = Cambridge Depersonalization Scale total score; CDScore = principal component of CDS subscales projected back on the subscales; PCnumbing = residuals of the principal component for CDS subscale numbing; PCdetachment = residuals of the principal component for CDS subscale detachment; PCblunting = residuals of the principal component for CDS subscale blunting, BDI-II = Beck Depression Inventory BDI-II, BAI = Beck Anxiety Inventory (log-transformed)
Table 3
Correlation Analysis between Cambridge Depersonalization Scale Variables and Neural Substrates.
CDS variable Anatomical Location Cluster Size Centre of Mass Location z-score Sign Test p-value X Y Z
Left Extrastriate Body Area – Default Mode Network
PCcore Right Posterior Cingulate Cortex 146 4 -54 21 -2.39 .03 Left Inferior Frontal Gyrus 131 -43 31 -3 -2.39
Left Middle Frontal Gyrus 103 -27 16 52 -2.41 Left Posterior Cingulate Cortex 102 -6 -52 20 -2.47 Right Middle Temporal Gyrus 86 45 -64 25 -2.62
PCdetachment Left Anterior Cingulate Cortex 95 -5 44 5 2.31 .5 CDStotal Right Posterior Cingulate Cortex 173 4 -54 21 -2.41 .03
Left Inferior Frontal Gyrus 113 -42 31 -3 -2.35 Left Posterior Cingulate Cortex 106 -6 -52 20 -2.44 Left Middle Frontal Gyrus 91 -27 16 52 -2.35 Right Middle Temporal Gyrus 82 45 -64 25 -2.57
Right Extrastriate Body Area – Default Mode Network
PCcore Right Middle Temporal Gyrus 129 55 -10 -10 -2.49 .02 Left Middle Temporal Gyrus 126 -51 -62 20 -2.31
Right Precuneus 93 3 -61 25 -2.39
Right Medial Frontal Gyrus 90 8 52 14 -2.48 Left Posterior Cingulate Cortex 90 -10 -56 16 -2.37 Right Posterior Cingulate Cortex 87 7 -52 19 -2.39
PCnumbing Left Superior Temporal Gyrus 83 -57 -59 19 -2.24 .5
Left Angular Gyrus 81 -54 -56 36 2.33
CDStotal Right Middle Temporal Gyrus 95 54 -12 -10 -2.55 .06
Right Precuneus 90 3 -61 24 -2.37
Right Medial Frontal Gyrus 85 8 52 14 -2.48 Left Posterior Cingulate Cortex 85 -10 -56 16 -2.31
Left Hippocampus – Default Mode Network
Left medial Prefrontal Cortex – Striatum
PCblunting Left Striatum 38 -29 -15 1 2.54 .5
Left Striatum 36 -25 -6 4 2.52
Right medial Prefrontal Cortex - Striatum
PCblunting Right Striatum 62 12 12 8 2.38 .5
PCdetachment Right Striatum 43 12 12 11 -2.38 .25
Left Striatum 41 -19 8 5 -2.52
Right Striatum 35 24 -4 8 -2.40
Note. The table shows cluster-sizes, centres of mass in Talairach space, z-scores of the t-test and
p-values of the sign test.
CDScore = principal component of CDS subscales projected back on the subscales; PCdetachment = residuals of the principal component for CDS subscale detachment;
PCnumbing = residuals of the principal component for CDS subscale numbing with Beck Depression Inventory score as noise covariates; CDStotal = Cambridge Depersonalization Scale total score. Cluster-thresholds were kDMN = 75, kstriatum = 34 and kanterior IC = 21.
