affective dysregulation and dissociative experience in female patients with borderline personality...

JOURNALOF

PSYCHIATRIC

Journal of Psychiatric Research 39 (2005) 85–92RESEARCH

www.elsevier.com/locate/jpsychires

Affective dysregulation and dissociative experience in femalepatients with borderline personality disorder: a startle response study

Ulrich W. Ebner-Priemera, Sandra Badeckb, Cornelia Beckmannb, Amy Wagnerc,Bernd Feigeb, Isabelle Weissd, Klaus Liebb, Martin Bohusa,*

a Department of Psychosomatic Medicine and Psychotherapy, Central Institute of Mental Health, Postfach 12 21 20, 68072 Mannheim, Germanyb Department of Psychiatry and Psychotherapy, University of Freiburg Medical School, Freiburg, Germany

c Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle, WA, USAd Institute of Cell Biology, Swiss Federal Institute of Technology, Zurich, Switzerland

Received 10 November 2003; received in revised form 30 March 2004; accepted 13 May 2004

Abstract

Affective dysregulation and dissociation are currently discussed as core features of borderline personality disorder (BPD). Af-

fective dysregulation is hypothesized to be correlated with increased amygdala functioning and dissociation is linked to inhibited

processing on the amygdala and dampened autonomic output, according to the corticolimbic disconnection model of dissociation

from Sierra and Berrios [Biological Psychiatry 44 (1998) 898]. We assessed startle response, which is mainly mediated by the

amygdala, to investigate the relationship between affective dysregulation and dissociation. We hypothesized that patients with BPD

would reveal enhanced responses to startling tones, but that these would be lessened by the presence of state dissociative experiences.

21 unmedicated female patients with BPD and 21 healthy female controls listened to 15 startling tones (95-dB, 500-ms, 1000-Hz)

while heart rate, skin conductance and orbicularis oculi electromyogram responses were measured. Covariance analysis showed that

the BPD group had a significantly higher startle response in the electromyogram as compared to controls. Furthermore, present-

state dissociative experiences significantly influenced the startle response. Patients with low dissociative experiences revealed en-

hanced startle responses whereas patients with high dissociative experiences showed reduced responses. Our data support affective

dysregulation in BPD as well as the corticolimbic disconnection model of dissociation, at least for EMG. Furthermore, it highlights

the importance of assessing present-state dissociation in basic research as well as psychotherapy.

� 2004 Elsevier Ltd. All rights reserved.

Keywords: Borderline personality disorder; Startle response; Affective dysregulation; Dissociation; Electromyography; Psychophysiology

1. Introduction

This study focuses on the interrelation of two DSM-

IV criteria in borderline personality disorder (BPD):

affective dysregulation or instability and dissociative

symptoms. Affective dysregulation is currently depictedin the scientific literature as the core feature of BPD

(Linehan, 1993; Sanislow et al., 2002; Skodol et al.,

2002a,b; Siever et al., 2002). Support for this view has

been found by several studies (Levine et al., 1997;

Dougherty et al., 1999; Stein, 1996). Although the neu-

* Corresponding author. Tel.: +49-621-1703426; fax: +49-621-

1703172.

E-mail address: [email protected] (M. Bohus).

0022-3956/$ - see front matter � 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jpsychires.2004.05.001

robiological basis of affective dysregulation is unknown

some research suggests that affective dysregulation

might be caused by higher activity of the amygdala

(Corrigan et al., 2000). For example Herpertz et al.

(2001a) found significantly higher activity of the amyg-

dala to unpleasant visual stimuli within BPD patientscompared to healthy controls, as assessed by fMRI.

Although the self-report of valence and arousal did not

differ between groups, the authors suggested that en-

hanced amygdala activation in BPD might reflect in-

tense and slowly subsiding emotions. In addition

alterations of the amygdala in BPD were found, using

MRI-based volumetric measurements (Driessen et al.,

2000; Tebartz van Elst et al., 2001). However, there havealso been laboratory studies using potentiated startle

mail to: [email protected]

86 U.W. Ebner-Priemer et al. / Journal of Psychiatric Research 39 (2005) 85–92

that provided no evidence for emotional hyperreactivity

or hypersensitivity in individuals diagnosed with BPD

(Herpertz et al., 1999, 2001b) neither on a psychological

nor on a physiological level.

Despite the central role of emotional reactivity inBPD, the restriction or absence of emotional experi-

encing, in the form of dissociative behavior, is also

common in BPD (Zanarini et al., 2000) and is one of the

defining DSM-IV criteria. Sierra and Berrios (1998)

published a neurobiological model for depersonalization

disorder, proposing bilateral corticolimbic disconnec-

tion during dissociation. In their model the medial pre-

frontal cortex inhibits processing on the amygdala,causing a reduced emotional experience and a damp-

ening of autonomic output. Recent studies confirm au-

tonomic blunting in dissociation. Sierra et al. (2002)

showed that subjects with depersonalization disorder

exhibited reduced magnitude and increased latency of

skin conductance response to unpleasant stimuli, but

not to non-specific stimuli suggesting a selective inhibi-

tion of emotional processing. Lanius and colleagues(2002) using a traumatic script-driven symptom provo-

cation paradigm posttraumatic stress disorder (PTSD)

could also partially support the model of Sierra and

Berrios. The dissociated PTSD subgroup (present state)

revealed reduced heart rate, increased activation in the

dorsolaterala and medial frontal cortex and did not

exhibit increased amygdala activation. Studies on peri-

traumatic dissociation and physiological response re-vealed controversial results. Griffin et al. (1997) reported

decreased heart rate and galvanic skin response in fe-

male rape victims with high dissociation compared to a

group with low dissociation during baseline and while

reporting about the trauma. In contrast, Kaufmann

et al. (2002) could not show any physiological differ-

ences to trauma-relevant stimuli between PTSD patients

with low and high peritraumatic dissociation. Ladwiget al. (2002) investigated startle response and peritrau-

matic dissociation in survivors of life-threatening car-

diac events. Using several subgroup analyses, the biggest

influence on startle response was the diagnosis of PTSD.

However, we think that the disconnection model of

dissociation from Sierra and Berrios (1998) refers rather

to present state than to peritraumatic dissociation.

Consistent with the disconnection model, clinicianshave observed that dissociation is characterized by de-

creased emotional experiencing (Maldonado and Spie-

gel, 1998). Similarly, behavioral therapists (Foa and

Kozak, 1986; Wagner and Linehan, 1999) view dissoci-

ation as functioning to regulate emotional engagement

in e.g. exposure therapy.

To explore these questions related to the neurobio-

logical underpinnings of affective dysregulation anddissociation in BPD, we employed the acoustic startle

response paradigm (ASR). The neural pathway of the

EMG in the ASR involves three neuronal ‘‘relay sta-

tions’’: The cochlear root neuron, the caudal pontine

reticular nucleus (PnC), and motoneurons in the facial

motor nucleus (Davis et al., 1999). The PnC is mainly

controlled by the medial part of the central nucleus of

the amygdala (Rosen et al., 1991). Enhanced amygdalaactivation (e.g. via electrical stimulation) therefore leads

to enhanced startle response in EMG (Davis et al., 1999;

Rosen and Davis, 1988). Along these lines, studies have

shown that trauma survivors with chronic PTSD dem-

onstrate both exaggerated amygdala response (Rauch

et al., 2000) and elevated responses to startling tones

(Metzger et al., 1999; Shalev et al., 1997, 2000).

Because of the afore mentioned linkage betweenBPD, emotional dysregulation, enhanced amygdala ac-

tivation (Herpertz et al., 2001a) and enhanced startle

response in EMG (Davis et al., 1999; Rosen and Davis,

1988), the ASR is an appropriate paradigm for investi-

gating affective dysregulation in this population. The

first specific hypothesis is: (1) Patients with BPD will

show enhanced responses to startling tones. According

to the corticolimbic disconnection model of dissociationfrom Sierra and Berrios (1998) we examined whether

dissociation is linked with reduced physiological re-

sponsiveness. The second specific hypothesis is: (2)

Present-state dissociative experiences will reduce the

startle response in BPD. To ensure that reduced startle

response is related to present-state dissociative experi-

ences and not to other psychological variables, we in-

vestigated anxiety, depression, trait dissociation andcomorbid PTSD as a confounding variables.

2. Materials and methods

2.1. Subjects

Twenty one female patients with BPD (10 inpatients,11 outpatients) and a comparison group of 21 female

healthy controls (HC) participated in this study. The

BPD sample was recruited from consecutively admitted

patients to a DBT-treatment program (Bohus et al.,

2000). Patients were randomly assigned to inpatient or

outpatient treatment. All patients fulfilled DSM-IV

criteria for BPD, assessed by the appropriate segment of

the Structured Clinical Interview for DSM-IV Person-ality Disorders (SCID-II; First et al., 1996) and scored a

minimum of 8 points on the Revised Diagnostic Inter-

view for Borderlines (DIB-R; Zanarini et al., 1989). Axis

I comorbidity was assessed by the Structured Clinical

Interview for DSM-IV Axis I Disorders (SCID-I; First

et al., 1997). Patients with a lifetime history of schizo-

phrenia, bipolar I disorder or alcohol and drug addic-

tion were excluded. Trained psychologists administeredall diagnostic instruments. Eighteen patients from the

BPD group had current comorbid Axis I disorders, in-

cluding major depressive disorder ðn ¼ 6Þ, anxiety dis-

U.W. Ebner-Priemer et al. / Journal of Psychiatric Research 39 (2005) 85–92 87

orders without PTSD ðn ¼ 11Þ, PTSD ðn ¼ 9Þ, obses-sive-compulsive disorder ðn ¼ 2Þ, and eating disorders

ðn ¼ 8Þ. The healthy controls were randomly selected

individuals from the national resident register of the

City of Freiburg, Germany, contacted by telephone.Exclusion criteria for the control group included the

diagnosis of BPD (SCID-II for DSM-IV), any Axis-I

disorder (MINI-SCID; Sheehan et al., 1998), current

psychotherapy or self-report of any first-degree relative

with a mental disorder. Individuals for both groups were

required to be free of medication for a minimum of 4

weeks prior to the study, except for contraceptive.

Groups were matched for sex (all females) and age(Table 1). All subjects were paid for participating in

the study. After complete description of the study to the

subjects, written informed consent was obtained. The

study was approved by the ethical board of the Uni-

versity of Freiburg, Medical School, Germany, in ac-

cordance to the declaration of Helsinki.

2.2. Psychometric measures

To assess present-state dissociative experiences the

‘‘Dissociation-Tension-Scale’’ (DSS; Stiglmayr et al.,

2001) was used. This self-rating scale consists of 19 items

concerning somatic (e.g. perception of pain, vision and

hearing) as well as psychological dissociation (e.g. de-

realization, depersonalization, amnesia). All items were

derived from the DES (Bernstein and Putnam, 1986)and the Somatoform Dissociation Questionnaire (SDQ-

20; Nijenhuis et al., 1996). Reliability analysis (internal

consistency) of the DSS resulted in a Cronbach’s a of 0.9(Stiglmayr et al., 2001). The cut-off value for having

severe dissociative features is 2.7 in a possible range

from 0.0 to 9.0 (Stiglmayr et al., 2001). Some authors

emphasize the distinction between different aspects of

dissociation, e.g. somatoform and psychological di-mensions (Nijenhuis et al., 1996). However, Stiglmayr

and colleagues (2001) did not find any support for this

Table 1

Results of demographic, self-report and physiological resting period data of

BPD mean (SD) H

Age 28.5 (8.1) 2

Dissociation: present-state (DSS) 1.38 (1.25) 0

Dissociation: trait (DES) 22.8 (7.75) 3

Depression (BDI) 25.0 (12.2) 1

Anxiety (STAI – state) 51.7 (10.1) 3

Audiometric examination (dB) 10.9 (4.3) 1

EMG: resting perioda 1.09 (0.42) 1

SCL: resting perioda 6.51 (4.0) 6

HR: resting perioda 69.7 (9.0) 7

EMG, electromyogram; SCL, skin conductance level; HR, heart rate.aResting period¼mean of pre-experiment and post-experiment resting pe

distinction in patients with BPD. For that reason sub-

scales of dissociation were not used in this study.

Trait dissociation was assessed by the ‘‘Dissociative

Experience Scale’’ (DES; Bernstein and Putnam, 1986),

anxiety by the ‘‘State Trait Anxiety Inventory’’ (STAI;Spielberger et al., 1970), and depressive symptoms by

the ‘‘Beck Depression Inventory’’ (BDI; Beck et al.,

1961). Hearing thresholds of all participants were spec-

ified with the auditory module of the NEUROSCAN-

System.

2.3. Physiologic measures

2.3.1. Data acquisition

Dependent physiologic measures included left orbic-

ularis oculi electromyogram (EMG), skin conductance

(SC) and heart rate (HR). EMG was recorded from 4-

mm (sensor diameter) Beckmann-type Ag/AgCl surface

electrodes filled with electrolyte paste (Genuine Grass

EC2, Grass Instrument Division, Warwick, RI) and

placed according to published specifications over the leftorbicularis oculi muscle (Fridlund and Cacioppo, 1986).

All electrode impedances were kept below 5 kX. Skinconductance was recorded from 7-mm (sensor diameter)

Ag/AgCl-electrodes (Marquette Hellige GmbH, Frei-

burg, Germany) filled with an isotonic EDR jelly

TDE-246 (Steffens, Berlin, Germany; identical to Grass

EC-33) and placed on the thenar and hypothenar emi-

nences of the non-dominant hand. Conductance wasconverted into a voltage signal (10 lS/V reduced to 500

lS/V by voltage divider) using an SC-amplifier MBV

400.2 (Rimkus Medizintechnik, Riemerling, Germany)

with a time constant of 10 s and a 25 Hz low-pass. ECG

was recorded from standard limb electrocardiogram

leads (blue sensor; Medicotest; Germany). All physio-

logical measures were amplified (EMG 10–100 Hz; SC

DC-100 Hz; ECG 1–100 Hz) and recorded using a 32-channel digital EEG system (Synamps; Neuroscan, El

Paso, TX) at a rate of 500 Hz.

female patients with BPD and HC

C mean (SD) Statistics t (df) p

9.7 (8.9) )0.471(40) 0.641

.07 (0.14) 4.76 (20.5) 6 0.001

.3 (2.74) 10.89 (24.9) 6 0.001

.4 (2.7) 8.65 (22.0) 6 0.001

2.8 (6.5) 7.14 (34.5) 6 0.001

0.9 (2.3) 0.021 (40) 0.984

.67 (1.05) )2.174 (22.4) 0.041

.79 (3.1) )0.25 (37) 0.806

0.0 (9.0) )0.09 (40) 0.932

riod.


2.3.2. Off-line processing

EMG raw scores were high-pass filtered at 57.5 Hz

(extracting the important high-frequency part of the

EMG), rectified and low-pass filtered at 12.5 Hz. The

corresponding short time constant was chosen to letthe rectified signal follow EMG bursts closely, thereby

allowing a precise detection of the onset latency for

failure scores. For precise information about the influ-

ence of sampling rate and time constant on EMG scores,

see Metzger et al. (1999) or Berg and Balaban (1999).

The ECG signal was transformed into a heart rate using

the software program Bio 25 (Foerster, 1998).

2.4. Stimuli

Stimuli consisted of fifteen 95-dB (SPL), 1000 Hz,

500-ms pure tones with 60.1 ms rise and fall time,

generated by a ‘‘LabMaster’’ digital–analog converter,

controlled by the software-program Stim (NEURO-

SCAN). These stimulus parameters are similar to those

used in most of the recent startle-response studies inPTSD (Shalev et al., 1997, 2000; Metzger et al., 1999)

and were chosen for this study to enable comparability.

Startle probes were delivered within a pseudorandom,

inter-trial interval of 33–52 s and binaurally over

headphones (EAR LINK 3a). Sound was controlled

with a precision sound level meter (Br€uell and Kjaer,

Typ 2206; Darmstadt, Germany).

2.5. Procedure

The experiment took place in a sound-attenuated,

temperature-controlled room connected through wires

to the laboratory in which the experimental apparatus

was located. Participants were seated upright in a

comfortable armchair. After the participant was famil-

iarized with the laboratory conditions, the electrodeswere attached. Then psychological assessments for dis-

sociation and anxiety were administered. In a 2 min

resting-period before and after the experiment, physio-

logical parameters were recorded. The participant was

instructed as follows: ‘‘You are going to hear a series of

sounds. Please sit quietly and listen to the sounds as they

come. Keep your eyes open throughout the entire pro-

cedure, which will not last more than 15 min’’ (Englishtranslation of the German instruction).

2.6. Data analysis

Main parameters for analysis were ‘‘response’’ and

‘‘habituation’’ to the 15 startling tones.

2.6.1. Response

Response in EMG, SC and HR for each trial was

calculated by subtracting average baseline levels imme-

diately preceding the onset of the tone (EMG, SC, HR:

0–2 s before tone onset) from the maximum response in

the respective time windows after startle onset (EMG:

21–150 ms; SC: 1–4 s, HR: 1–4 s). EMG trials with re-

sponse onset 6 20 ms or P 120 ms were rejected. Re-

sponse onset was defined as time until the EMG increasereaches the response-criterion. The number of rejected

responses was below 1%. Control of SC artifacts fol-

lowed published specifications (Boucsein, 1992).

2.6.2. Habituation

Habituation was assessed by two methods: ‘‘slope’’

and ‘‘trial to non-response’’ (trial to criterion¼TTC).

Relative habituation was defined as the slope of the re-gression equation y ¼ bxþ a for trials 2–15, where y is

the square root of the response score and x is the log

trial number. The first trial was dropped as usual

(Lykken et al., 1988). The non-response criterion is de-

fined as an EMG and SC response of <0.35 lV and

<0.01 lS, respectively. TTC, as a parameter of absolute

habituation, is defined as the number of trials until the

subject reaches two consecutive non-response trials. Toreduce the variance associated with unusually large re-

sponses, square root transformations were performed on

the response scores of EMG, SC, and HR prior to the

statistical analysis. The software Bio 25 (Foerster, 1998)

was administered to control artifacts and calculate

physiological raw data.

To examine the group differences in physiological

resting period, demographics, and self-report question-naires, t test for independent samples were used. For the

analysis of group differences in the physiological re-

sponse parameters controlling for present-state disso-

ciative experiences, covariance analysis (ANCOVA) was

performed. A covariance analysis was chosen instead of

a 2-factorial analysis (with dissociation as a second

factor) because there was insufficient variation in DSS

scores in the HC group for this type of analysis. Fur-thermore, 2-factorial analysis (or post hoc tests of sub-

groups) would eliminate most of the interesting variance

in dissociation. To assess the influence of anxiety, de-

pression, and comorbid PTSD on the physiological re-

sponse, multiple linear regression analyses were used. A

p-value of 6 0.05 conferred statistical significance. All

analyses were two-tailed.

3. Results

3.1. Demographic and psychometric data

Demographic and psychometric data are presented in

Table 1. Patients with BPD reported significantly higher

values in the present-state dissociation (DSS), trait dis-sociation (DES), depression (BDI) and anxiety (STAI).

The correlation between DSS and DES was significant

(r ¼ 0:61; p6 0:001).


3.2. Audiometric examination and physiological resting

period data

Audiometric examination revealed comparable re-

sults and no hearing impairment in patients and controls(Table 1). Physiological resting period values (mean of

pre-experiment and post-experiment resting periods) are

given in Table 1. Whereas the mean resting period scores

for HR and SCL were comparable between patients and

controls, the EMG mean resting period scores were

significantly lower in the patients with BPD. Because

EMG raw scores depend strongly on the extent of skin

abrading and electrode placement, raw score compari-sons of group differences are problematic (Tassinary and

Cacioppo, 2000). However, to assess a possible impact

of low resting scores on the amplitude of the startle re-

sponse, EMG resting values were included as a control

variable in the multiple regression analysis. Anyway, it

should be noted that previous studies have not shown

such an influence (Ornitz et al., 1996).

3.3. Physiological responses

EMG, SC and HR scores and the results of the sta-

tistical analyses are shown in Table 2. Three patients

and three HC were non-responders for EMG, and two

patients and one HC were non-responders for SC; these

subjects were therefore excluded from the analyses.

In the ANCOVA for EMG, a highly significant groupeffect was found in response to the 15 tones, such that

the BPD group showed a higher response. Further, there

was a significant effect for the covariate, present-state

dissociative experiences. Examining habituation as

number of trials needed to reach the EMG non-response

criterion (TTC), the ANCOVA again revealed a highly

significant effect for group, such that the BPD group

Table 2

Response scores of electromyogram, skin conductance and heart rate and the

BPD HC

Mean (SD) Mean (SD)

EMG

Response 2–15 # 1.45 (0.90) 1.01 (0.47)

TTC 10.33 (5.93) 7.94 (5.42)

Slope )0.26 (0.41) )0.26 (0.36)

SC

Response 2–15 # 0.53 (0.3) 0.37 (0.2)

TTC 7.84 (4.8) 6.40 (5.0)

Slope )0.61 (0.31) )0.27 (0.47)

HR

Response 2–15 # 4.52 (2.5) 4.98 (3.3)

Slope 0.04 (1.1) )0.54 (1.2)

EMG, electromyogram; SC, skin conductance; HR, heart rate; Response

transformation; df (EMG)¼ 1.33, df (SC)¼ 1.36, df (HR)¼ 1.39, n(EMG

HC¼ 21.

needed more trials, and for the covariate, present-state

dissociation. No effect was found for relative habitua-

tion (slope). No differences were found in SC and HR

for all parameters (response, TTC and slope).

To further assess whether the results of the EMG canbe explained by other confounding variables, two mul-

tiple regression analysis were computed, predicting

mean response and TTC. Each had one fixed factor

(group: BPD vs. HC) and the following stepwise (for-

ward entry) factors: present-state dissociation (DSS),

anxiety (STAI), PTSD diagnosis, and EMG baseline

levels (resting period). For both (mean response, TTC),

only the overall model with the factors group and dis-sociation accounted significantly for the variance (mean

response: R2 (group and DSS) ¼ 0.21; F ð2:33Þ ¼ 4:32;p ¼ 0:022; TTC : R2 (group and DSS)¼ 0.34;

F ð2:33Þ ¼ 8:46; p ¼ 0:001). Anxiety, PTSD and EMG

baseline level were not significant predictors in either

equation. Furthermore, a model with group alone (BPD

vs. HC) as well as a model with present-state dissocia-

tion (DSS) alone did not account significantly for thevariance.

To show the direction of the influence of present-state

dissociation on EMG graphically, we split the group of

patients with BPD into two subgroups along the median

of the DSS. The HC group was not split because of

insufficient variation in DSS scores (see Table 1). We

designated the two BPD subgroups as low (0.53� 0.34)

and high (2.51� 1.11) present-state dissociative experi-ences. As shown in Fig. 1, the startle responses to the 15

tones in the subgroup with low present-state dissociation

(dotted line) were higher than those of the subgroup

with high present-state dissociation (dashed line) and

those of the controls (solid line). T test revealed nearly

significant difference for reactivity between the two

subgroups (T ð16Þ ¼ 2:09; p ¼ 0:054).

results of the statistical analysis of female patients with BPD and HC

Main effect: group Covariate: DSS

F p F p

8.42 0.007 4.88 0.034

12.9 0.001 14.7 0.001

0.03 0.868 0.04 0.849

2.61 0.115 0.01 0.940

1.60 0.214 0.79 0.381

2.68 0.511 0.44 0.110

0.09 0.925 0.10 0.751

0.49 0.490 1.17 0.490

1–15, response to tone 1–15; TTC, trial to non-response; #, square root

) BPD¼ 18, HC¼ 18, n(SC) BPD¼ 19, HC¼ 20, n(HR) BPD¼ 21,

0

0.5

1

1.5

2

2.5

3

3.5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

EM

G -

Rea

ctiv

ity

in µ

V (

sqr)

all BPD patientsHCBPD patients with low dissociationBPD patients with high dissociation

Fig. 1. Mean EMG response to 15 startling tones in HC ðn ¼ 18Þ, BPDwith low dissociative experiences ðn ¼ 9Þ and BPD with high disso-

ciative experiences ðn ¼ 9Þ. Please note that splitting was done only for

figures not for data analysis.


4. Discussion

The present investigation revealed an enhanced star-

tle response in individuals with BPD compared to

healthy controls, as assessed by EMG response. These

findings reflect the affective dysregulation in this popu-

lation and support the notion of greater amygdala ac-

tivity. Furthermore, there was evidence that the startle

response was influenced by the present-state of dissoci-

ation. Patients with high present-state dissociative ex-periences revealed a lower EMG response than those

with low dissociation. This supports the theory of

physiological blunting in dissociation of Sierra and

Berrios (1998), at least for EMG.

Two different interpretations of this finding are pos-

sible. First, it could be the case that present-state dis-

sociative symptoms block the (Borderline-typical)

enhancement of the startle response (state-dependentinterpretation). Second, it could be the case that there is

not an overall heightened startle response among BPD

individuals, but instead two different stable subgroups

within the BPD population, one group with dissociative

features combined with low responsivity to startling

tones and another subgroup without dissociative fea-

tures and a high responsivity (trait-dependent interpre-

tation). Although it is not possible to determinecausation from the data presented here, a comparison of

the association between present-state dissociative expe-

riences (DSS) and trait dissociative features (DES) on

ASR may shed some light on this issue. In the current

sample, for the BPD group, the correlation between

EMG-response and DES (trait) was not significant

(r ¼ �0:10; p ¼ 0:734), whereas the correlation between

EMG-response and DSS (present-state) revealed a trendfor statistical significance (r ¼ �0:44; p ¼ 0:068). The

picture is clearer with regard to absolute habituation.

The correlation, in the patient sample, between EMG-

TTC and DES (trait) revealed only a trend for statistical

significance (r ¼ �0:49; p ¼ 0:066), whereas the corre-

lation between EMG-TTC and DSS (present-state) was

highly significant (r ¼ �0:77; p6 0:001). The data

therefore favor the idea that present-state high dissoci-ation blocks the enhanced startle response in BPD.

Our finding of a significant influences of present-

state dissociative experiences on the startle response

may possibly explain why some previous studies,

which did not consider dissociation, did not find en-

hanced startle responses during affective stimulation in

patients with BPD (Herpertz et al., 2001b). Addi-

tionally, physiologic unresponsiveness found in aportion (e.g., 30–40%) of subjects in past studies of

PTSD (see Orr and Roth, 2000) might be explained by

present-state dissociation.

Limitations of this study should be noted. First,

sample size is small and stage of menstrual cycle was not

assessed. This is a possible confound in the current

study. A clinical comparison group was not studied yet,

therefore it remains unclear whether these findings arespecific for BPD. Another remaining question is why

enhanced startle response and a detectable influence of

dissociation were only evident in the EMG. This may be

explained by relatively independent pathways of the

three physiological parameters (Dawson et al., 2000).

Since enhanced acoustic startle response is a common

finding in patients with PTSD, one might argue that the

enhanced startle response in the current study is relatedto PTSD symptomatology. However, this seems un-

likely, since the diagnosis of PTSD did not account

significantly for the variance in the multiple regression

analyses. Furthermore, post hoc analyses showed that

BPD patients with comorbid PTSD did not show dif-

ferences in EMG response compared to the group of

BPD patients without comorbid PTSD (t ¼ 0:31;df ¼ 19; p ¼ 0:759). Furthermore, no effect of stateanxiety on startle response was detected. This is sur-

prising given the vast literature on the effects of fear-

induced startle (Lang et al., 1990, 1998). We suppose

that the limited sample size may have obscured such a

potential relationship. Finally, the current study was

based on a relatively new measure of present-state dis-

sociation (DSS; Stiglmayr et al., 2001). Although initial

psychometric data on the measure are strong, furtherstudies on the reliability and validity of the measure are

needed.

We would like to suggest implications for both

psychotherapy and basic research. With regard to the

former, we would like to make a daring comparison

between ASR as a simple learning process – in terms

of learning not to react to an aversive stimuli – and

exposure therapy, as a much more complex learningprocess. Jaycox and colleagues (1998) postulates two

necessary conditions for successful exposure therapy:

high initial emotional engagement and habituation.


The results of the current study suggest that exposure

therapy in patients with BPD might be hampered by

both: patients experiencing low levels of dissociation

may reveal prolonged habituation (at least in the

TTC), whereas patients with high dissociation mayshow attenuated initial emotional activation (Fig. 1).

A treatment-outcome study of patients with panic

disorder could already confirm, that dissociative be-

havior led to poorer outcomes in exposure therapy

(Michelson et al., 1998). However, the finding of al-

tered habituation (TTC) should be interpreted care-

fully because it is inherently confounded with the level

of startle reactivity.With regard to basic research, the present findings

highlight the importance of assessing present-state dis-

sociation, in investigations of psychophysiology, neu-

ropsychology and neuroimaging in BPD as well as other

disorders with comorbid dissociative behavior.

Our findings, of course, require replication in order to

determine the strength and stability of the present

findings. Additional studies examining the extent ofemotion dysregulation in individuals with BPD to a

range of stimuli as well as the role of present-state dis-

sociation in emotional reactivity among individuals with

BPD are needed to fully understand the nature of

emotion dysregulation in this population.

Acknowledgements

This research was supported by a grant (Bo 1487/3-1)

from the German Research Society (DFG) and the

Borderline Personality Disorder Research Foundation

(BPDRF). The authors thank J. Fahrenberg, F. Foer-

ster, P. H€uttner, and V. H€oppner from the Psycho-

physiology Research Unit, University of Freiburg for all

methodological, technical and statistical support, andthe patients and control probands for their participation

in the study.

References

Beck AT, Ward CH, MendelsonM, Mock H, Erbaugh J. An inventory

for measuring depression. Archives of General Psychiatry

1961;4:561–71.

Berg WK, Balaban MT. Startle elicitation: stimulus parameters,

recording techniques, and quantification. In: Dawson ME, Schell

AM, B€ohmelt AH, editors. Startle modification: implications for

neuroscience, cognitive science, and clinical science. New York:

Cambridge University Press; 1999. p. 21–50.

Bernstein EM, Putnam FW. Development, reliability, and validity of a

dissociation scale. Journal of Nervous and Mental Disease

1986;174:727–35.

Bohus M, Haaf B, Stiglmayr CE, Pohl U, Boehme R, Linehan MM.

Evaluation of inpatient dialectical-behavioral therapy for border-

line personality disorder – a prospective study. Behaviour Research

and Therapy 2000;38:875–87.

Boucsein W. Electrodermal activity. New York: Plenum Press; 1992.

Corrigan FMD, Davidson A, Heard HL. The role of dysregulated

amygdalic emotion in borderline personality disorder. Medical

Hypotheses 2000;54:574–9.

Davis M, Walker DL, Lee Y. Neurophysiology and neuropharmacol-

ogy of startle and its affective modulation. In: Dawson ME, Schell

AM, B€ohmelt AH, editors. Startle modification: implications for

neuroscience, cognitive science, and clinical science. Cambridge:

University Press; 1999. p. 95–114.

Dawson ME, Schell AM, Filion DL. The electrodermal system. In:

Cacioppo JT, Tassinary LG, Berntson GG, editors. Handbook of

psychophysiology. Cambridge: University Press; 2000. p. 200–23.

Dougherty DM, Bjork JM, Huckabee HCG, Moeller FG, Swann AC.

Laboratory measures of aggression and impulsivity in women with

borderline personality disorder. Psychiatry Research 1999;85:315–

26.

Driessen M, Herrmann J, Stahl K, Zwaan M, Meier S, Hill A,

Osterheider M, Petersen D. Magnetic resonance imaging volumes

of the hippocampus and the amygdala in women with borderline

personality disorder and early traumatization. Archives of General

Psychiatry 2000;57:1115–22.

First MB, Spitzer RL, Gibbon M, Williams JBW. User’s guide for the

Structured Clinical Interview for DSM-IV personality disorders

(SCID-II). Washington, DC: American Psychiatric Press; 1996.

First MB, Spitzer RL, Gibbon M, Williams JBW, Benjamin LS. User’s

guide for the Structured Clinical Interview for DSM-IV axis I

disorders (SCID-I) – clinical version. Washington, DC: American

Psychiatric Press; 1997.

Foa EB, Kozak MJ. Emotional processing of fear: exposure to

corrective information. Psychological Bulletin 1986;99:20–35.

Foerster F. Programm-Paket BIO (Computersoftware). Germany:

Psychophysiology Research Unit, University of Freiburg; 1998.

Fridlund AJ, Cacioppo JT. Guidelines for human electromyographic

research. Psychophysiology 1986;23:567–89.

Griffin MG, Resick PA, Mechanic MB. Objective assessment of

peritraumatic dissociation: psychophysiological indicators. Amer-

ican Journal of Psychiatry 1997;154:1081–8.

Herpertz SC, Dietrich TM, Wenning B, Krings T, Erberich SG,

Willmes K, Thron A, Sass H. Evidence of abnormal amygdala

functioning in borderline personality disorder: a functional MRI

study. Biological Psychiatry 2001;50:292–8.

Herpertz SC, Kunert HJ, Schwenger UB, Sass H. Affective

responsiveness in borderline personality disorder: a psycho-

physiological approach. American Journal of Psychiatry

1999;156:1550–6.

Herpertz SC, Werth U, Lukas G, Qunaibi M, Schuerkens A, Kunert

HJ, Mueller-Isberner R, Osterheider M, Sass H. Emotion in

criminal offenders with psychopathy and borderline personality

disorder. Archives of General Psychiatry 2001;58:737–45.

Jaycox LH, Foa EB, Morral AR. Influence of emotional engagement

and habituation on exposure therapy for PTSD. Journal of

Consulting and Clinical Psychology 1998;66:185–92.

Kaufman ML, Kimble MO, Kaloupek DG, McTeague LM, Bachrach

P, Forti AM, Keune TM. Peritraumatic Dissociation and Physi-

ological Response to Trauma Relevant Stimuli in Vietnam Combat

Veterans with Posttraumatic Stress Disorder. Journal of Nervous

and Mental Disease 2002;190:167–74.

Ladwig K-H, Marten-Mittag B, Deisenhofer I, Hofmann B, Schap-

perer J, Weyerbrock S, Erazo N, Schmiff C. Psychophysiological

Correlates of Peritraumatic Dissociative Responses in Survivors

of Life-Threatening Cardiac Events. Psychopathology 2002;35:

241–8.

Lang PJ, Bradley MM, Cuthbert BN. Emotion, attention, and the

startle reflex. Psychological Review 1990;97:377–95.

Lang PJ, Bradley MM, Cuthbert BN. Emotion, motivation, and

anxiety: brain mechanisms and psychophysiology. Biological

Psychiatry 1998;44:1248–63.


Lanius RA, Williamson PC, Boksman K, Densmore M, Gupta M,

Neufeld RWJ, Gati JS, Menon RS. Brain activation during script-

driven imagery induced dissociative response in PTSD: a functional

magnetic resonance imaging investigation. Biological Psychiatry

2002;52:305–11.

Levine D, Marziali E, Hood J. Emotion processing in borderline

personality disorders. Journal of Nervous and Mental Disease

1997;185:240–6.

Linehan MM. Cognitive-behavioral treatment of borderline personal-

ity disorder. New York: The Guildford Press; 1993.

Lykken DT, Iacono WG, Haroian K, McGue M, Bouchard TJ.

Habituation of the skin conductance response to strong stimuli: a

twin study. Psychophysiology 1988;25:4–15.

Maldonado JR, Spiegel D. Trauma, dissociation, and hypnotizability.

In: Bremner JD, Marmar CR, editors. Trauma, memory and

dissociation. Washington, DC: American Psychiatric Association;

1998. p. 57–106.

Metzger LJ, Orr SP, Berry NJ, Ahern CE, Lasko NB, Pitman RK.

Physiologic reactivity to startling tones in women with posttrau-

matic stress disorder. Journal of Abnormal Psychology

1999;108:347–52.

Michelson L, Vives A, Testa S, Marchione N, June K. The role of

trauma and dissociation in cognitive-behavioral psychotherapy

outcome and maintenance for panic disorder with agoraphobia.

Behaviour Research and Therapy 1998;36:1011–50.

Nijenhuis ERS, Spinhoven P, Van Dyck R, van der Hart O. The

development and psychometric characteristics of the Somatoform

Dissociation Questionnaire (SDQ-20). Journal of Nervous and

Mental Disease 1996;184:688–94.

Ornitz EM, Russell AT, Yuan H, Liu M. Autonomic, electroenceph-

alographic, and myogenic activity accompanying startle and its

habituation during mid-childhood. Psychophysiology 1996;33:507–

13.

Orr SP, Roth WT. Psychophysiological assessment: clinical applica-

tions for PTSD. Journal of Affective Disorders 2000;61:225–40.

Rauch SL, Whalen PJ, Shin LM, McInerney SC, Macklin ML, Lasko

NB, Orr SP, Pitman RK. Exaggerated amygdala response to

masked facial stimuli in posttraumatic stress disorder: a functional

MRI study. Biological Psychiatry 2000;47:769–76.

Rosen JB, Davis M. Enhancement of acoustic startle by electrical

stimulation of the amygdala. Behavioral Neuroscience

1988;102:195–202.

Rosen JB, Hitchcock JM, Sananes CB, Miserendino MJ. A direct

projection from the central nucleus of the amygdala to the acoustic

startle pathway: anterograde and retrograde tracing studies.

Behavioral Neuroscience 1991;105:817–25.

Sanislow CA, Grilo CM, Morey LC, Bender DS, Skodol AE,

Gunderson JG, Shea MT, Stout RL, Zanarini MC, McGlashan

TH. Confirmatory factor analysis of DSM-IV criteria for border-

line personality disorder: findings from the collaborative longitu-

dinal personality disorders study. American Journal of Psychiatry

2002;159:284–90.

Shalev AY, Peri T, Brandes D, Freedman S, Orr SP, Pitman RK.

Auditory startle response in trauma survivors with posttraumatic

stress disorder: a prospective study. American Journal of Psychi-

atry 2000;157:255–61.

Shalev AY, Peri T, Orr SP, Bonne OB, Pitman RK. Auditory startle

responses in help-seeking trauma survivors. Psychiatry Research

1997;69:1–7.

Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller

E, Hergueta T, Baker R, Dunbar GC. The Mini International

Neuropsychiatric Interview (M.I.N.I): the development and vali-

dation of a structured diagnostic psychiatric interview for DSM-IV

and ICD-10. Journal of Clinical Psychiatry 1998;59:22–33.

Sierra M, Berrios GE. Depersonalization: neurobiological perspec-

tives. Biological Psychiatry 1998;44:898–908.

Sierra M, Senior C, Dalton J, McDonough M, Bond A, Phillips ML,

O‘Dwyer AM, David AS. Autonomic response in depersonaliza-

tion disorder. Archives of General Psychiatry 2002;59:833–8.

Siever LJ, Torgersen S, Gunderson JG, Livesley WJ, Kendler KS. The

borderline diagnosis III: identifying endophenotypes for genetic

studies. Biological Psychiatry 2002;51:964–8.

Skodol AE, Gunderson JG, Pfohl B, Widiger TA, Livesley WJ, Siever

LJ. The borderline diagnosis I: psychopathology comorbidity, and

personaltity structure. Biological Psychiatry 2002;51:936–50.

Skodol AE, Siever LJ, Livesley WJ, Gunderson JG, Pfohl B, Widiger

TA. The borderline diagnosis II: biology, genetics, and clinical

course. Biological Psychiatry 2002;51:951–63.

Spielberger CD, Gorsuch RL, Lushene RE. STAI manual. Palo Alto:

Consulting Psychologists Press; 1970.

Stein KF. Affect instability in adults with a borderline personality

disorder. Archives of Psychiatric Nursing 1996;10:32–40.

Stiglmayr CE, Shapiro DA, Stieglitz RD, Limberger MF, Bohus M.

Experience of aversive tension and dissociation in female patients

with borderline personality disorder – a controlled study. Journal

of Psychiatric Research 2001;35:111–8.

Tassinary LG, Cacioppo JT. The skeletomotor system: surface

electromyogramm. In: Cacioppo JT, Tassinary LG, Berntson

GG, editors. Handbook of psychophysiology. 2nd ed. Cambridge:

University Press; 2000. p. 163–99.

Tebartz van Elst L, Thiel T, Hesslinger B, Henke M, Lieb K,

Bohus M, Henning J, Ebert D. Evidence of subtle prefrontal

neuropathology in patients with borderline personality disorder

as assessed by Short Echo 1H – magnetic resonance spectros-

copy study. Journal of Neuropsychiatry and Clinical Neuro-

science 2001;13:511–4.

Wagner AW, Linehan MM. Dissociation. In: Follette VM, Ruzek

JJ, Abueg FR, editors. Trauma in context: a cognitive-

behavioral approach. New York: Guildford Press; 1999. p.

191–225.

Zanarini MC, Gunderson JG, Frankenburg FR, Chauncey DL. The

revised diagnostic interview for borderlines: discriminating BPD

from other axis II disorders. Journal of Personality Disorders

1989;3:10–8.

Zanarini MC, Ruser TF, Frankenburg FR, Hennen J, Gunderson JG.

Risk factors associated with the dissociative experiences of

borderline patients. Journal of Nervous and Mental Disease

2000;188:26–30.

affective dysregulation and dissociative experience in female patients with borderline personality...

Documents