Introduction
Etiological Models: Trauma and Dissociation
Models on Brain Structure and Function Associated With Dissociation
Cortico-Limbic Disconnection Model and Neuroimaging Research in Depersonalization Disorder
Models on Emotion Modulation and Research in the Dissociative Subtype of PTSD (D-PTSD)
Research in Dissociative Identity Disorder (DID)
Research on Structural Alterations
Interim Summary
Dissociation in Borderline Personality Disorder (BPD)
Neuroimaging Research on Dissociation in BPD
Authors, year of publication | Groups (sample size), gender | Psychotropic medication status | Comorbidities and trauma history in the patient sample | Neuroimaging technique | Measures of dissociation (trait/state and time of assessment) | Key findings concerning dissociation |
---|---|---|---|---|---|---|
Hazlett et al. (2012) | • Groups: - BPD (n = 33) - Schizotypal PD (SPD; n = 28) - HC (n = 32) • Gender: mixed (female/male) BPD, 20/13; SPD, 12/16; and HC, 20/12 | Medication-free for at least 6 weeks prior to scanning. Most patients (BPD, n = 16; SPD, n = 23) had never been medicated. | High rates of childhood abuse and neglect in BPD. Exclusion of history of schizophrenia, psychotic disorder, bipolar I, or current major depressive disorder (MDD). 21 BPD and 6 SPD had past MDD. | Event-related fMRI during processing of neutral, pleasant, and unpleasant pictures from the IAPS, each of which presented twice within their respective trial block/run. | Self-reported trait dissociation (DES) | BPD patients showed greater amygdala reactivity and prolonged amygdala activation to repeated emotional versus neutral IAPS pictures. Fewer dissociative symptoms in both patient groups were associated with greater amygdala activation to repeated unpleasant pictures. |
Hoerst et al. (2010) | • Groups: - BPD (n = 30) - HC (n = 30) • Gender: female | Free of current psychotropic medication for at least 3 months prior to scanning procedure. | Some patients met criteria for current/lifetime posttraumatic stress disorder (PTSD; 11/13), MDD (3/18), substance abuse (0/7), as well as eating disorders and (other) anxiety disorders. Lifetime schizophrenia and bipolar I were excluded. | Proton magnetic resonance spectroscopy (MRS), to measure neurometabolic concentrations (glutamate levels) in the anterior cingulate cortex (ACC) | Self-reported trait dissociation (DES). | Significantly higher levels of glutamate in the ACC in patients with BPD as compared with healthy controls. Positive correlation between glutamate concentrations and dissociation as well as between glutamate concentration and subscores of the borderline symptom |
Irle et al. (2007) | • Groups: - BPD (n = 30) - HC (n = 25) • Gender: female | 8 patients were on antidepressant medication (SSRI) and 6 were occasionally treated with sedatives (e.g., benzodiazepine). | High rates of physical and sexual abuse in childhood and adolescence. Some patients met criteria for current/lifetime depersonalization disorder (27/27), dissociative amnesia (DA) (7/7), dissociative identity disorder (DID; 4/4), and PTSD (11/11). | Structural MRI to assess volumes of the superior (precuneus and postcentral gyrus) and inferior parietal cortices | Presence of comorbid dissociative disorders (SCID-D) and dissociative symptoms such as depersonalization and derealization (DIB) | BPD patients with comorbid DA or DID had significantly increased volumes of the left postcentral gyrus compared to healthy controls (+13%) and BPD patients without these disorders (+11%). In BPD subjects, stronger depersonalization was significantly correlated to larger right precuneus volumes. |
Kluetsch et al. (2012) | • Groups: - BPD with history of self-harm (n = 25) - HC (n = 23) • Gender: female | Unmedicated sample | Some patients met criteria for current/lifetime PTSD (9/9), MDD (0/18), as well as eating disorders and (other) anxiety disorders, current MDD, substance abuse, and lifetime schizophrenia or bipolar I were excluded. | FMRI during painful heat versus neutral temperature stimulation (thermal sensory anlayzer II) | Self-reported trait dissociation (DES) and state dissociation, assessed prior to and immediately after scanning (DSS) | Higher self-reported trait dissociation was associated with an attenuated signal decrease of the default mode network in response to painful stimulation. |
Kraus et al. (2009) | • Groups: - BPD with PTSD (n = 12) - BPD without co-occurring PTSD (n = 17) • Gender: female | Free of psychotropic medication for at least 2 weeks before scanning procedure. | 12 BPD patients met criteria for current PTSD. Lifetime MDD (n = 21), eating disorders, and (other) anxiety disorders were present. Current MDD, substance abuse, and lifetime schizophrenia or bipolar I were excluded | FMRI during heat stimulation, assessed in five stimulation blocks (each for 30 s) with individually adapted temperature | Self-reported trait dissociation (DES) and state dissociation at the time of scanning (DSS) | Both groups of BPD patients did not differ significantly in pain sensitivity, while amygdala deactivation was more pronounced in BPD patients with co-occurring PTSD. Amygdala deactivation was independent of state dissociation. |
Krause-Utz et al. (2014a) | • Groups: - BPD (n = 22) - HC (n = 22) • Gender: female | Medication-free for at least 14 days (in the case of fluoxetine, 28 days) prior to scanning procedure. | All BPD patients had a history of childhood abuse/interpersonal trauma. Some patients met criteria for current/lifetime PTSD (9/11), other anxiety, and eating disorders. Current MDD, substance abuse (6 months prior to scan) and lifetime schizophrenia or bipolar I were excluded | Event-related fMRI during performance of an emotional working memory task (EWMT, adapted Sternberg item recognition task) with negative versus neutral interpersonal IAPS pictures | Self-reported trait dissociation (DES) and state dissociation immediately before and after scanning (DSS4). | In the BPD group, increase of self-reported dissociative states (DSS4 scores) over the course of the EWMT positively predicted bilateral amygdala connectivity with and left insula, left precentral gyrus, right thalamus, and right anterior cingulate during emotional distraction. |
Krause-Utz et al. (2012) | • Groups: - BPD (n = 22) - HC (n = 22) • Gender: female | Medication-free for at least 14 days (in the case of fluoxetine, 28 days) prior to scanning procedure. | All BPD patients had a history of interpersonal trauma (including severe childhood abuse/neglect). Some patients met criteria for current/lifetime PTSD (9/11). Current MDD, substance abuse, and lifetime schizophrenia or bipolar I were excluded. | Event-related fMRI during performance of an emotional working memory task (EWMT, adapted Sternberg item recognition task) with negative versus neutral interpersonal IAPS pictures. | Self-reported trait dissociation (DES) and state dissociation immediately before and after scanning (DSS4) | In the BPD group, increase of self-reported dissociative states (DSS4 scores) over the course of the EWMT negatively predicted bilateral amygdala activity during emotional distraction. |
Krause-Utz et al. (2015) | • Groups: - BPD (n = 27) - HC (n = 26) • Gender: female | Free of psychotropic medication at least 4 weeks before scanning. | Some patients met criteria for current/lifetime PTSD (14/18), other anxiety disorders. and eating disorders. Lifetime diagnosis of psychotic disorder, bipolar I disorder, and alcohol/substance abuse 6 months prior to scan were excluded. | FMRI during a differential delay aversive conditioning paradigm with an electric shock as unconditioned stimulus and two neutral pictures as conditioned stimuli (CS+ and CS−) | Self-reported trait dissociation (DES) and state dissociation before and after scan (DSS) | Amygdala habituation to CS+paired (CS+ in temporal contingency with the aversive event) during acquisition was found in HC, but not in patients. No significant correlations with dissociative symptoms. |
Krause-Utz et al. (2014b) | • Groups: - BPD (n = 20) - HC (n = 17) • Gender: female | Medication-free for at least 14 days (in the case of fluoxetine, 28 days) prior to scanning procedure. | All BPD patients had a history of interpersonal trauma. Some patients met criteria for current PTSD (n = 9), other anxiety disorders, and eating disorders. Current MDD, substance abuse (6 months prior to scan), and lifetime schizophrenia or bipolar I were excluded. | Resting state (RS) fMRI was acquired to investigate RS functional connectivity in the medial temporal lobe network (seed: amygdala), salience network (seed: dorsal ACC), and default mode network (seed: ventral ACC). | Self-reported trait dissociation (DES) | Self-reported trait dissociation positively predicted amygdala connectivity with dorsolateral prefrontal cortex and negatively predicted amygdala connectivity with during resting state. No differences were found between BPD with or without comorbid PTSD. |
Krause-Utz et al. (submitted) | • Groups: - BPD patients exposed to a dissociation script (BPDd: n = 17) - Patients exposed to neutral script (BPDn: n = 12) - HC (n = 18) • Gender: female | Free of psychotropic medication for at least 4 weeks prior to the study. | All patients reported at least one type of severe to extreme childhood trauma. 5 patients in the BPDn group and 7 in the BPDd group met criteria for current PTSD. Comorbidity with current (other) anxiety and eating disorders was evident. Lifetime psychotic disorder, bipolar I disorder, mental retardation, and alcohol/substance abuse 6 months prior to scan were excluded. | FMRI to measure changes in BOLD signal, combining script-driven imagery (to experimentally induce dissociation) with a subsequent EWMT (to investigate working memory performance during emotional distraction) | Dissociation was induced in 17 patients, while 12 patients and 18 HC were exposed to a personalized neutral script. Self-reported trait dissociation (DES) and state dissociation at baseline, after script and after EWMT | BPDd showed overall WM impairments, significantly reduced bilateral amygdala activity (across conditions) and reduced left cuneus, lingual gyrus, and posterior cingulate activity (during negative distractors) compared to BPDn. Inferior frontal gyrus activity was higher in both BPD groups than in HC. BPDd further showed a stronger coupling of amygdala with right superior/middle temporal gyrus, right middle occipital gyrus, left inferior parietal lobule, and left claustrum than BPDn and HC. |
Lange et al. (2005) | • Groups: - BPD (n = 17) - HC (n = 9) • Gender: female | 5 BPD patients were on antidepressant medication and some were occasionally treated with benzodiazepines (n = 4) or mild neuroleptics (n = 3). | All BPD participants had experienced severe childhood sexual and physical abuse. Some patients met criteria for current PTSD (n = 6), depersonalization disorder (n = 14), DA (n = 4), and DID (n = 1). Nearly all patients met criteria for current or lifetime MDD (n = 16). | 18Fluoro-2-deoxyglucose positron emission tomography (FDG-PET) to assess glucose metabolism in temporo-parietal cortices | Presence of comorbid dissociative disorders (SCID-D) and self-reported trait dissociation (DES) | BPD patients demonstrated reduced FDG uptake in the right temporal pole/anterior fusiform gyrus and in the left precuneus and posterior cingulate cortex. Impaired memory performance among borderline subjects was significantly correlated with metabolic activity in ventromedial and lateral temporal cortices. |
Ludascher et al. (2010) | • Group: BPD patients (n = 15) • Gender: female | Medication-free for at least 14 days (in the case of fluoxetine, 28 days) prior to scanning procedure. | 10 BPD patients had comorbid PTSD following severe childhood abuse (6 patients reporting sexual abuse, 3 patients reporting physical abuse, and 1 patient reporting neglect). Some patients met criteria for current/lifetime other anxiety and eating disorders. Lifetime diagnosis of psychotic disorder, bipolar I disorder, and alcohol/substance abuse 6 months prior to scan were excluded. | FMRI to measure changes in BOLD signal during script-driven imagery: participants were exposed to a personalized dissociative-inducing script (versus a neutral script) during the fMRI scan. | Self-reported trait dissociation (DES) and state dissociation before and immediately after script inside the MRI scanner (DSS-4) | BPD patients showed a significant increase in the left inferior frontal gyrus during the dissociation script. DSS4 ratings positively predicted activity in the left superior frontal gyrus and negatively predicted activity in the right middle and inferior temporal gyrus. In BPD patients with comorbid PTSD, increased activity in the left cingulate gyrus was found during the dissociation script. In this subgroup, DSS4 positively predicted bilateral insula activity and negatively predicted activity in the right parahippocampal gyrus. |
Niedtfeld et al. (2013) | • Groups: - BPD (n = 60) - HC (n = 60) • Gender: female | Free of psychotropic medication | 21 of the BPD patients met criteria for current PTSD. Traumatic events included severe physical and sexual childhood abuse. | Structural MRI to assess anatomical scans. Whole-brain gray matter volumes (GMV) were studied using voxel-based morphometry (VBM). | Self-reported trait dissociation (DES); scores were available in 42 BPD patients. | Trait dissociation (DES score) was positively correlated to GMV in the middle temporal gyrus in BPD. |
Paret et al. (2016) | • Group: BPD patients (n = 8) • Gender: female | All participants were on stable medication. | Real-time fMRI-based neurofeedback training comprising of four sessions, in which participants viewed aversive images and received feedback from a thermometer displaying amygdala BOLD signals. Amygdala activity and functional connectivity were studied. | Self-reported state dissociation (DSS4) at the end of each run | Task-related amygdala-ventromedial prefrontal cortex connectivity was altered across the four sessions, with an increased connectivity during instructed emotion regulation versus viewing emotional pictures without regulation. Self-reported state dissociation and “lack of emotional awareness” decreased with training. | |
Rusch et al. (2007) | • Groups: - BPD and comorbid attention deficit hyperactivity disorder (ADHD; n = 20) - HC (n = 20) • Gender: female | Medication-free (including methylphenidate) for at least 2 weeks prior to scanning procedure | 5 patients met criteria for current PTSD, 10 patients reported a history of sexual abuse in childhood, and 14 patients met criteria for past MDD. Exclusion of current major depression, substance abuse 6 months prior to study, lifetime substance dependence, schizophrenia, and bipolar I. | Diffusion tensor imaging (DTI) was used to measure mean diffusivity and fractional anisotropy in the inferior frontal white matter | Self-reported trait dissociation (DES) | Patients showed increased mean diffusivity in inferior frontal white matter, which was associated with higher levels of dissociative symptoms, dysfunctional affect regulation, anger-hostility, and general psychopathology but not associated with a history of sexual abuse. |
Sar et al. (2007) | • Groups: - DID (n = 21, 15 with BPD) - HC without history of child abuse and trauma (n = 9) • Gender: mixed (female/male) Patients, 14/7 HC, 6/3 | Medication-free for at least 1 month prior to study | All patients reported at least one type of severe childhood abuse and/or neglect, 15 patients had comorbid BPD+DID, and 6 patients had DID without BPD. | Single photon emission computed tomography (SPECT) with Tc99m-hexamethylpropylenamine (HMPAO) as a tracer was used to measure regional cerebral blood flow. | Self-reported trait dissociation (DES) | Patients showed decreased reduced cerebral blood flow in the orbitofrontal cortex (OFC) and occipital regions bilaterally. There was no significant correlation between rCBF ratios of the regions of interest and self-reported dissociation. |
Wingenfeld et al. (2009) | • Groups: - BPD (n = 20) - HC (n = 20) • Gender: mixed (14 females and 6 males in each group) | 12 of the BPD samples received psychotropic medication (including antidepressants and neuroleptics). | 17 BPD patients reported at least mild PTSD symptoms and 5 patients fulfilled criteria for current PTSD. Some patients met criteria for current MDD (n = 3), bulimia nervosa (n = 3), social phobia (n = 1), and somatoform disorder (n = 1). Current MDD with psychotic symptoms, schizophrenia, schizoaffective disorders, anorexia, and substance dependence 6 months prior to the study were excluded. | fMRI to measure changes in BOLD signal during performance of an individualized emotional stroop task (EST), with neutral, general negative words, and individual negative words (selected from a prior interview with each participant). | Self-reported state dissociation before and after scanning (DSS21 akut) as well as dissociation within the past 7 days (DSS21) | Overall BPD patients had slower reaction times, which however were not correlated with dissociation. Healthy controls—but not BPD patients—showed significant recruitment of the ACC for negative versus neutral and individual negative versus neutral conditions, respectively. No significant correlations between DSS scores and BOLD signal were reported. |
Winter et al. (2015) | • Groups: - BPD patients exposed to a dissociation script: BPDd (n = 18) - Patients exposed to neutral script: BPDn (n = 19) - HC (n = 19) • Gender: female | Free of psychotropic medication for at least 4 weeks prior to the study. | 7 patients in the BPDn group and 8 in the BPDd group met criteria for current PTSD. Comorbidity with current (other) anxiety and eating disorders was evident. Lifetime psychotic disorder, bipolar I disorder, mental retardation, and alcohol/substance abuse 6 months prior to scan were excluded | fMRI to measure changes in BOLD signal, combining script-driven imagery (to experimentally induce dissociation) with a subsequent emotional stroop task (EST; containing negative, neutral, and positive words—to measure cognitive control of emotional material). | Dissociation was induced in 18 patients, while 19 patients and 19 HC were exposed to a personalized neutral script. Self-reported trait dissociation (DES) and state dissociation at baseline, after script (twice before EST, twice within EST, and once after EST) | BPD patients after dissociation induction (BPDd) showed overall slower and less accurate responses as well as longer reaction times for negative versus neutral words than BPDn. Moreover, BPDd showed increased activity in the inferior frontal gyrus and dorsolateral prefrontal cortex (dlPFC) during negative than neutral words. BPDn patients showed increased activity in the right superior temporal gyrus for emotional (positive and negative) versus neutral words compared to HC. |
Wolf et al. (2011) | • Groups: - BPD (n = 17) - HC (n = 17) • Gender: female | All BPD patients were on stable medication for at least 2 weeks before scanning (psychotropic medication included antidepressants, mood stabilizers, and antipsychotics). | Some patients met criteria for current/lifetime MDD (9/5), past substance abuse (n = 6), eating disorders, and anxiety disorders other than PTSD. Current PTSD, lifetime schizophrenia, bipolar disorder, ADHD, and substance abuse 6 months prior to study were excluded. | RS-fMRI was acquired to investigate RS functional connectivity in large-scale brain networks. | Self-reported dissociation (DSS) | Self-reported state dissociation and tension (DSS) positively predicted RS functional connectivity of the insula and precuneus in the BPD group. |
Wolf et al. (2012) | • Groups: - BPD (n = 16) - HC (n = 16) • Gender: female | All BPD patients were on stable medication for at least 2 weeks before scanning (psychotropic medication included antidepressants, mood stabilizers, and antipsychotics). | Some patients met criteria for current/lifetime MDD (8/5), past substance abuse (n = 6), eating disorders, and anxiety disorders other than PTSD. Current PTSD, lifetime schizophrenia, bipolar disorder, ADHD, and substance abuse within 6 months prior to study were excluded. | Continuous arterial spin labeling magnetic resonance imaging | Self-reported dissociation (DSS) | Compared to controls, BPD patients exhibited decreased blood flow in the medial OFC, whereas increased blood flow was found in the left and right lateral OFC. Correlation analyses revealed a positive relationship between medial and lateral orbitofrontal blood flow and impulsivity Barrett impulsiveness scale (BIS), but not with dissociation (DSS). |