2

Characteristics of Social Behavior of Patients with Bipolar Disorder

Measured by Proximal Spacing in Immersive Virtual Reality

Environment

Eosu Kim

Department of Medicine

The Graduate School, Yonsei University

3

Characteristics of Social Behavior of Patients with Bipolar Disorder

Measured by Proximal Spacing in Immersive Virtual Reality

Environment

Directed by Professor Hyun-Sang Cho

The Master's Thesis submitted to the Department of Medicine, the Graduate School of Yonsei University

in partial fulfillment of the requirements for the degree of Master of Medical Science

Eosu Kim

June/2008

4

This certifies that the Master's Thesis of Eosu Kim is approved.

------------------------------------ Thesis Supervisor: Hyun-Sang Cho

------------------------------------ Thesis Committee Member#1: Dong Goo Kim

------------------------------------ Thesis Committee Member#2: Jae-Jin Kim

The Graduate School Yonsei University

June/2008

5

ACKNOWLEDGEMENTS

First of all, I would like to express my deep appreciation for the warm guidance of Prof. Hyun-Sang Cho, my thesis adviser, from whom I have learned a number of important things relevant to the qualification as an independent investigator, beyond his thorough intellectual instructions regarding this paper. I cannot but mention my director's name, Prof. Byoung Hoon Oh, the chief director of the department of psychiatry, Yonsei University. His generosity has granted me an opportunity to keep going as an investigator, as well as a person. I also express my appreciation for the passionate instructions of Prof. Dong Goo Kim, a distinguished scholar in the subject of 'brain and behavior', and the chief director of the department of pharmacology, Yonsei University, College of Medicine. He showed such a passion willing to take troubles from a long distance to give valuable instructions to my work. I would like to stress upon the inspiration I derived from Prof. Jae-Jin Kim whom I always try to make my model as a neuroscientist. Especially, I wish to thank Dr. Jeonghun Ku and his colleagues in the department of biomedical engineering, Hanyang University, for their genius in construction of the virtual reality system, which enabled the experiment.

To beg my parents' pardon, I first wish to express the most of my appreciation and respect to my wife, Yehwon Lee, who has always been beside, waiting without a single word of complaints she deserves to make, for my finishing the selfish exploration of the world. For all the joy my one-year-old daughter, Dayune is giving us, do I also express my thanks to her. The names of my parents, Jin Seong Kim and Jae Sook Lee, who have not only made me grow up and stand but also actually renounced much of their rights for the sake of my development, will always stay inscribed deep in my mind, as the names of the most respectful beings always just giving.

Eosu Kim

6

TABLE OF CONTENTS

ABSTRACT ………………………………………………………..… 1

I. INTRODUCTION ……………………………...……………...…3

II. MATERIALS AND METHODS ……………………………..……6

1. Participants ………………………………………………...………6

2. Preparation of the immersive virtual environment ...………………6

3. Task procedure and measurements ……………...…..……...……..7

4. Data analyses .......………………………………………………10

III. RESULTS …………………………………………..……………11

1. Interpersonal distances .......................................................………11

2. Gaze aversions ................................................................…………13

3. Affective valences and arousals toward avatars .....………………15

IV. DISCUSSION …................................................…………………16

V. CONCLUSION …………………...………………………………20

REFERENCES ………………………………………………………21

ABSTRACT (IN KOREAN) ……………………………..……………26

7

LIST OF FIGURES

Figure 1. Virtual reality systems for measurements ·················8

Figure 2. Interpersonal distances ············································11

Figure 3. Relationship between interpersonal distances and

state/trait anxiety scores··········································13

Figure 4. Gaze behaviors during listening and speaking········14

LIST OF TABLES

Table 1. Demographic and Clinical Characteristics ·················7

1

ABSTRACT

Characteristics of Social Behavior of Patients with Bipolar Disorder Measured by Proximal Spacing in Immersive Virtual Reality

Environment

Eosu Kim

Department of Medicine The Graduate School, Yonsei University

(Directed by Professor Hyun-Sang Cho)

Objective: Bipolar manic (BM) patients have been associated with

mood-congruent positive bias in emotional processing. Given that

social/emotional behaviors can be organized along a basic approach-withdrawal

dimension, we speculated that BM patients would exhibit abnormal social

behaviors, having difficulties in regulation of approach-withdrawal patterns in

social encounters (i.e., immoderate approach behavior).

Methods: We measured interpersonal distance (IPD) and the degree of gaze

aversion of twenty BM patients and age-matched 20 normal controls (NCs)

during a self-introduction task with six avatars (happy, angry, neutral/male,

female) using immersive virtual environment technology. Trait-anxiety and

affective valence and arousal to each avatar were also measured.

Results: BM patients demanded significantly greater IPD and showed more

averted gaze than NCs, even though more manic patients approached the

avatars more closely. Trait-anxiety was inversely correlated with IPD only in

NCs. Discriminating spatial behavior was apparent towards happy avatars

2

(closest) in BM patients, but towards angry avatars (farthest) in NCs. BM

patients, but not NCs, averted their gazes further while speaking than while

listening, which seemed associated with trait-anxiety.

Conclusions: These findings on avoidant patterns of nonverbal social behaviors

suggest that the negativistic social cognition, similar to that found in patients

with depression, is implicitly internalized in BM patients. Sophisticated use of

interpersonal space and cognitive therapeutic approaches would be helpful in

developing secure therapeutic relationships with BM patients.

----------------------------------------------------------------------------------------------------------

Key words: personal space; gaze; social cognition; mania; virtual reality.

3

Characteristics of Social Behavior of Patients with Bipolar Disorder Measured by Proximal Spacing in Immersive Virtual Reality

Environment

Eosu Kim

Department of Medicine The Graduate School, Yonsei University

(Directed by Professor Hyun-Sang Cho)

I. INTRODUCTION

Bipolar disorder is a major psychiatric illness, causing considerable

difficulties in social functioning 1. In particular, bipolar mania (BM) refers to an

extreme mood state characterized by severe impairment in emotional regulation.

Thus, BM patients may disrupt social relationships owing to excessive

eagerness for interpersonal interactions or immoderate ‘prosocial’ behaviors. As

reflecting the close relationship between emotion and social cognition 2, bipolar

manic, and even euthymic patients have been found to have difficulties in social

cognitive function, such as theory of mind 3, 4 or perception of emotion of others.

Studies have reported that BM patients have impairments in the recognition of

negative (fearful) facial emotion 5, showed attenuated brain responses, as well

as behavioral responses, to sad facial emotion 6, 7.

Lembke and Ketter have proposed that such mood-congruent positive biases

could make manic patients maintain approaching behavior during social

interactions, even when withdrawal would be more adaptive 5. This assumption

may be consistent with the previous report on the positively biased thought

patterns of BM patients often showing denial of the existence of risk while

4

decision-making 8, as well as with a finding that bipolar disorder (mania) is

related with dysfunctions in the right hemisphere 9 which is thought to mediate

withdrawal-related emotions 10. As ‘emovere’, the Latin root of emotion,

literally means ‘to make move’, all behaviors relevant to emotion can be

categorized in general as approach-appetitive motions expecting favorable

inputs or withdrawal-avoidant ones trying to reduce unfavorable stimuli 11.

Taken together, it can be speculated that the regulation of these emotion-specific

approach/withdrawal patterns of nonverbal social behavior might also be

impaired in BM patients during social interactions, and that these impairments

may contribute to their social dysfunction.

Nonverbal behaviors might have particular value in understanding patients

whose verbal communication is limited or disrupted 12. Therefore, the patterns

of nonverbal social behaviors could be an objectively measurable clue for the

social cognition and emotion underlying mania rather than verbal reports

through subjective rating scales. Particularly, behaviors "which cannot be

consciously manipulated" may convey more information than verbal expression,

just as patterns may convey more information than contents, as noted by Hall,

who coined the term, proxemics, the study of human beings' spatial behavior 13.

In this vein, personal space and eye gaze have been intensively investigated

as a subject of serious inquiry on nonverbal social behaviors 12, 14. Personal

space is an invisible bubble surrounding an individual, into which an intrusion

by others would arouse discomfort, and is manifested through interpersonal

distance (IPD) in social encounters 12. In clinical or nonclinical population,

personal space has been related to various social, emotional, and cognitive

factors 12, including ego functioning, self-concept, or interpersonal attribution

style, and has been also implicated in dyadic, psychotherapeutic situations 15-17.

However, the relationship between IPD and psychopathology has been

explored mainly by studies with schizophrenic patients 16-19, including our

recent studies using a virtual reality system 20, 21. Consistent with prior findings,

5

we have found that schizophrenic patients demand abnormally large personal

space 20, 21. Such spatial behavior can be understood as a passive defensive

strategy to secure a safety zone 22, 23, or, a ‘body-buffer zone’ 16, which protects

one from external or internal threats. Therefore, optimal interpersonal space

should have therapeutic implication in the initial stage of therapy when

attempting to develop secure relationships with psychiatric patients. However,

to our knowledge, no study concerning personal space or gaze has been

conducted yet with BM patients.

We developed a protocol to measure nonverbal, interpersonal behaviors, such

as spatial behavior coupled with gaze behavior, using immersive virtual

environment technology (IVET), which provides unique, methodological

advantages for the behavioral study 24, 25. From previous findings of the positive

bias in the emotional processing of BM patients, we assumed that BM patients

would exhibit positively biased behavioral patterns during social interactions.

Thus, we initially hypothesized that BM patients would maintain abnormally

close IPD relative to normal controls. We also predicted that IPD in BM patients

would be inversely associated with the severity of manic symptoms.

6

II. MATERIALS AND METHODS

1. Participants

Inpatients with bipolar I disorder, most recent episode with manic, were

recruited at a university psychiatric hospital, and diagnosed by two psychiatrists

(E.K. and H.S.C.) according to DSM-IV-TR criteria 26. Patients with visual

problems, a history of head trauma, movement disorder, or other Axis I or II

disorders were excluded. We also excluded patients with mixed mania because

of the possible confounding effects of a depressed or anxious mood on social

behaviors. In total, twenty patients (10 male, 10 female) were enrolled, who

were clinically as stabilized as cooperation was possible. Twenty age- and

sex-matched normal control subjects were recruited from colleges and from the

local community, and were carefully interviewed by a psychiatrist to exclude

any DSM-Ⅳ-TR Axis I disorders. The Spielberger State-Trait Anxiety

Inventory 27 was administered to all subjects. The Young Mania Rating Scale

(YMRS) 28 and the Hamilton Depression Rating Scale (HAMD) 29 were used to

assess the mood symptoms of the patient group. After a complete description of

the study, written informed consent was obtained from all subjects. The study

was approved by the Institutional Review Board of the Severance Mental

Health Hospital. The demographic characteristics and clinical data of the

subjects are summarized in Table 1.

2. Preparation of the immersive virtual environment

We developed a virtual environment system where social interactions could

occur between the subjects and the avatars. A total of six avatars varied across

emotional types (happy, neutral, and angry) and gender were created and

located in the middle of the virtual room (Fig. 1). The intensity of each facial

emotional expression of avatars was determined by using the linear morphing

technique, as described in other literature 30. To augment the sense of realism,

7

each avatar was made to look at the subjects, blink its eyes naturally, and

perform gestures compatible with the avatar’s facial and verbal expression of

emotion. The height of the eyes of the avatars was set to be exactly the same as

that of each subject.

Table1: Demographic and Clinical Characteristics

Bipolar manic

patients (n=20, male=10)

Healthy normal

controls (n=20, male=10)

p value by t-test

Age (years) 30.1±5.9 27.7±4.4 0.166

Education (years) 14.5±2.0 15.5±1.5 0.081

Intelligence quotient a 102.9±9.9 109.3±3.2 0.081

State anxiety score b 41.3±5.7 40.3±8.0 0.652

Trait anxiety score b 25.0±5.2 25.9±8.6 0.691

Number of admission 3.2±2.3

Duration of illness (years) 5.2±3.6

Antipsychotic medication c (mg, n=14) 575.0±254.2

Mood stabilizers

divalproex (n=14)

dose (mg) 957.1±296.2

blood concentration (ug/mL) 76.9±15.5

lithium (n=5) 810.0±201.2

dose (mg)

blood concentration (mmol/L) 0.76±0.17

YMRS score 20.9±6.1

HAMD score 4.6±2.1

Mean±S.D.; YMRS, Young Mania Rating Scale; HAMD, Hamilton Depression Rating Scale; a Obtained using Korean WAIS (Wechsler Adult Intelligence Scale).

b Scores from Spielberger’s State-Trait Anxiety Inventory. c Chlorpromazine-equivalent doses of antipsychotic medications.

3. Task procedure and measurements

Subjects put on a head mounted display and ear phones. A receiver, used to

track head position and orientation, was placed at the vertex of subjects. A

8

magnetic-field generating transmitter, the basis of how the receiver determines

data, was placed in front of the subject (Fig. 1). To become accustomed to the

virtual environment, subjects first explored an empty virtual room and then

conducted a total of six consecutive tasks. The order of emotional types and the

gender of avatars were counterbalanced. Subjects were not told that their gaze

and distancing behavior would be measured; instead, they were informed that

this was a study about the way people feel and interact in virtual reality, and

that this information would be used for therapeutic purposes of virtual reality.

Figure1: Virtual reality systems for measurements. Note that the subjects actually see the avatars in the head mounted device (HMD).

9

In each task, subjects were asked to approach an avatar and to stop at a

distance where they felt most comfortable to have a conversation with the

avatar. From that distance, subjects were to say ‘hello’ first and wait for the

response of the avatar. The avatar was manipulated to introduce itself only after

a subject says ‘hello’, at which point the avatar would talk about topics

compatible with its facial emotion and then request that the subjects introduce

themselves. Then, subjects were to introduce themselves and to say ‘over’ to

announce that the self-introduction was finished. To make the conversation

more natural, an experimenter in a separate monitoring room controlled when

the avatars would start talking or asking questions. The starting distance

between the subject and the avatar was 280 cm.

IPD was indexed by the distance between the vertexes of a subject and an

avatar (Fig. 1). Gaze aversion was indexed by gaze angle averted from the

middle of the avatar’s forehead to any direction. We determined the final

variables of IPD between subjects and avatars, and of the gaze angle of the

subjects by averaging all data measured at every 0.03-sec interval from the

moment when the subjects said ‘hello’ to the moment they said ‘over’. Gaze

angle was further examined by dividing it into two parts: the averaged gaze

angle while listening (from the moment of saying ‘hello’ to the moment when

‘self-introduction’ was requested) and the averaged gaze angle while speaking

(from the beginning of the self-introduction to saying ‘over’). We used head

orientation as an indirect indicator of gaze angle because direct measuring

technology was not applicable in our virtual system.

After completing the six respective tasks, the subjects rated their emotional

valence and arousal when responding to each avatar, using the Self-Assessment

Manikin (SAM) scale 31, where valence ranges from -4 (most negative) to 0

(neutral) and to 4 (most positive), and arousal from 0 (least arousing) to 8 (most

arousing).

10

4. Data analyses

To compare variables between groups across the three emotional types of the

avatars, we performed multivariate analysis of variance. We used the averaged

data over the two avatars in different genders and the same emotion, since we

intended to examine how subjects behaved across the avatars’ emotions, not

their gender. For pairwise comparisons among the three emotions within a

group, an adjusted p-value for multiple comparisons (Least Significant

Difference) was used. The difference between the gaze angle while listening

and that while speaking within a group was analyzed by the paired t-test.

Correlation analyses were performed by Pearson’s correlation test. All

statistical analyses were conducted with SPSS 12.0 (Chicago, IL). All p values

are two-tailed, and statistical significance was set at α=0.05.

11

III. RESULTS

1. Interpersonal distances

Overall, the mean IPD in BM patients was significantly greater than that of

normal controls (NC) (BM: 172.1±44.7 [95% CI 160.6~183.7] cm vs. NC:

133.2±44.1 [95% CI 121.7~144.6] cm, F=23.1, df=1,118, P<0.001) (see Fig.

2A). Gender effects on IPD were also significant (F=8.1, df=1,36, p=0.007) in

that female subjects demanded greater IPD than male subjects. However, there

was no significant interaction between group and gender in the mean IPD

(F=1.0, df=1,36, p=0.318).

Figure 2. Interpersonal distances. A. Between-group comparison of interpersonal distances

(IPDs). Data are shown as mean±SEM. Distances from each type of emotions of avatars are all

significantly different between the groups (neutral condition, F=13.3, df=1,38, p=0.001; angry

condition, F=6.6, df=1,38, p=0.015; happy condition, F=8.4, df=1,38, p=0.006). Asterisk (*)

means a significantly differential distance within each group (p<0.05). BM, patients with

bipolar mania; NC, normal controls. B. Relationship between IPDs and total scores of Young

Mania Rating Scale (YMRS) in the patient group.

12

Within-group comparisons revealed that only IPD-H (distance from the

happy avatars) was significantly different (smallest) from IPD-N (distance from

the neutral avatars) and IPD-A (distance from the angry avatars) in the patient

group, but in the control group only IPD-A was significantly different (largest)

from distances from the other two emotional avatars (see asterisks in Fig. 2A).

The mean IPDs of BM patients were inversely correlated with total YMRS

scores (IPD-N: r=-0.458, p=0.042; IPD-A: r=-0.439, p=0.053; IPD-H: r=-0.524,

p=0.018; for overall IPD, see Fig. 2B).

In the patient group, mean IPDs were not associated with either state or trait

anxiety scores (Fig. 3A, 3B). In the control group, significant correlations were

observed between IPDs and state anxiety scores (Fig. 3C; for neutral avatars,

r=0.546, p=0.013; for angry avatars, r=0.518, p=0.019; for happy avatars,

r=0.438, p=0.053), as well as trait anxiety scores (Fig. 3D; for neutral avatars,

r=0.605, p=0.005; for angry avatars, r=0.662, p=0.001; for happy avatars,

r=0.553, p=0.011).

13

Figure 3. Relationship between interpersonal distances and state/trait anxiety scores.

Interpersonal distance (IPD) and state or trait anxiety in the patient group (A, B) and in the

control group (C, D). BM, patients with bipolar mania; NC, normal controls.

2. Gaze aversions

The overall mean of averted gaze angles of the patient group was

significantly greater than that of the control group (6.1±2.8° vs. 3.8±1.8°,

F=28.3, df=1,118, p< 0.001). The between-group differences of gaze angles

across the emotional types of avatars were all significant (all p’s < 0.007;

Fig. 4). The main effects of gender on gaze were not significant (F=2.56,

14

df=1,36, p=0.119). Group by gender interaction in overall gaze angle was not

significant either (F=0.006, df=1,36, p=0.941).

Figure 4. Gaze behaviors during listening or speaking. Gaze angles were significantly greater

in BM compared with NC across all the three emotional conditions (all p’s <0.007). Asterisk (*)

means a significant further aversion of gaze while speaking than while listening within a group

(p<0.05). BM, patients with bipolar mania; NC, normal controls.

Within-group differences of gaze angle across the emotional types of avatars

were not significant in either of the groups. However, the asterisk marks in

15

Fig. 4 indicates that gazes became significantly more averted while speaking

than while listening in the patient group, but not in the control group. In the

patient group, these differences between gaze angles while listening and those

while speaking tended to correlate with trait anxiety scores in neutral and happy

avatars (for neutral avatars, r=0.434, p=0.056; for angry avatars, r=0.219,

p=0.354; for happy avatars, r=0.410, p=0.073; for overall, r=0.440, p=0.052).

No other significant correlations between gaze angles and other psychiatric

measures (total YMRS scores, state/trait anxiety scores, and valence/arousal)

were observed in either group (data not shown).

3. Affective valences and arousals toward avatars

Affective valence toward angry avatars was significantly diminished in the

patient group relative to the control group (BM = -2.15±1.75, NC = -3.08±0.88;

F=4.47, df=1,38, p=0.041). No group differences of valence were observed in

either neutral or happy condition. Relative to the control group, the patient

group showed enhanced arousal toward neutral avatars (BM = 3.73±1.92, NC =

2.4±1.11; F=7.13, df=1,38, p=0.011) but diminished arousal toward happy

avatars (BM = 4.10±1.47, NC = 5.10±1.06, F=6.12, df=1,38, p=0.018). No

between-group difference of arousal toward angry avatars was observed (BM =

5.07±2.0, NC = 5.83±1.38; F=1.91, df=1,38, p=0.175).

16

IV. DISCUSSION

We measured IPD and gaze of BM patients and healthy normal controls to

examine the effect of manic symptoms on those nonverbal social behaviors. As

expected, BM patients with more severe manic symptoms were found to

approach avatars more closely (Fig. 2B), supporting the mood-congruent,

positively biased behaviors of BM patients. However, contrary to our initial

hypothesis, BM patients exhibited relatively avoidant-withdrawn patterns of

interpersonal behaviors when compared with normal controls, by demanding

greater personal space (Fig. 2A) and showing more gaze aversion (Fig. 4).

These findings seem to contradict each other in that it sounds as though the IPD

of BM patients become 'normalized' as their symptoms worsened.

Based on these paradoxical findings, we can rather conceptualize the

important aspect of personal space; the size of which may be determined by two

factors at different levels. One, probably developmentally- or biologically-based,

involves trait-related factors, such as the diagnostic label in this case (e.g. Fig.

2A), and the other involves state-dependent factors, such as the mood symptoms

(e.g. Fig. 2B). If the former determines the distinctive, original size of personal

space, the latter makes it flexibly vary to some extent, depending on

circumstances, as far as the pre-determined original size permits. Differentiation

among such different levels of influences might be key to interpreting and

integrating complex findings on social behaviors. Indeed, speculative attempts

to explain similar contradictory findings can be found in other literature; for

example, Hall's early delineation of the different levels of cultural influences on

IPD 32, the inverse correlation between negative symptoms and IPD 18, 21, or the

personal space expanding effect of nicotine administration 33 in schizophrenic

patients.

In this regard, the ‘between’-group difference in personal space size might be

caused by illness-associated, trait-like characteristics of bipolar disorder rather

17

than by the effect of mood symptoms, while manic symptoms per se can

diminish the size of personal space and depressive symptoms can cause the

opposite reaction ‘within’ a group. Therefore, the question about the factors

causing the between-group difference of IPD or gaze is a major concern in the

interpretation of this study.

First, the avoidant behavioral patterns of BM patients, as the similar patterns

of schizophrenic patients were reported 16, 17, 19, 34, may reflect social and

cognitive deficits (including social cognitive difficulties) in bipolar patients,

which have been reported to be comparable to those of schizophrenic patients in

not all but some previous studies 35-37. However, we did not directly measure

cognitive functions of subjects to address this issue further. But, more gaze

aversion of BM patients during speaking relative to listening (Fig. 4) may

reflect the illness-associated cognitive difficulties as well as social difficulties in

BM patients. Glenberg et al. have suggested that gaze aversion has a role in

managing environmental stimuli to lessen cognitive load rather than in avoiding

social embarrassment 38.

Second, the differential behaviors might be mediated by the effect of

medication in BM patients. However, this seems unlikely in that no relationship

was found between doses or blood concentrations of antipsychotics, valproate

or lithium, and social behaviors (all p's>0.1).

Third, as the correlation between state/trait anxiety and IPD was seen in

normal controls, the anxiety-related psychopathology might mediate the

enlarged personal space of BM patients 39. This may be supported by the report

on a significant proportion of social anxiety comorbidity in bipolar disorder 40.

However, there was neither between-group difference nor within-group

correlation with IPD of state/trait anxiety in BM patients. With respect to this

finding, we speculated that the personal space-expanding effect of anxiety seen

in the control group might have been offset in the patient group (Fig. 3A, 3B vs.

3C, 3D) by manic symptoms that caused decrease of personal space (Fig.1B).

18

However, we postulate that the between-group difference of nonverbal

behaviors might have been caused not only by abovementioned factors which

can be ‘explicitly’ measured. Rather, we suggest that these nonverbal social

behaviors (interpersonal spacing and gaze behaviors) may reflect ‘implicit’

characteristics of social cognition because these nonverbal behaviors, as ‘hidden

dimension’ 32, are supposed to be determined ‘out of conscious awareness’, that

is, unwittingly during social interactions 13. In this sense, our findings might

result from the avoidant social cognition or emotion implicitly internalized in

BM patients. Supporting this assumption, previous studies using implicit

measures have recognized that bipolar euthymic or even manic patients have

negative biased social cognition, self-esteem or attribution styles, similar to

those found in patients with depression 41-43. Furthermore, Chen et al. have

found that BM patients show abnormally enhanced brain activation to a

mood-incongruent, sad face in an implicit affect recognition task, while, as

expected, showing reduced activation in an explicit task 6. These findings

collectively support the so-called 'manic defense hypothesis' 42, which proposed

that mania is just another countenance of depression, dominated by the same

underlying complexes of depression and manifested or compensated through

extreme defense mechanisms such as denial and reaction formation.

Such negative patterns of thought process might be a state-independent (i.e.

trait) factor which is maintained across the mood states in that the underlying

self-concept in bipolar patients might be formed predominantly under the

influence of negatively biased cognitions or emotions, rather than positive ones.

Supporting this theory, even manic patients were found to have negativistic core

belief (schema), as much as depressed patients have 44, and social dysfunction

was strongly associated with the severity of depressive symptoms in patients

with bipolar disorder 1.

Finally, our findings might suggest dysfunction of the 'social brain' in BM

patients. Recently, it was found that defensive spatial behaviors are mediated by

19

specialized regions of the brain, such as the ventral intraparietal area and the

precentral gyrus 23. Given the recent finding that amygdala damage results in

abnormal gaze patterns including a severe reduction of eye contact 45, avoidant

gaze behavior in BM patients may also be associated with these kinds of

anomalous neural activities. In addition, the amygdala, which is known to

abnormally function in mania 7, 46, has been regarded as a critical component

which mediates between emotion and social cognition, modulating contextually

relevant social/emotional behaviors 47.

This study might be methodologically sound in its utilization of IVET to

measure behaviors. The previous methodological weakness in precisely and

consistently measuring social behaviors might have contributed to a marked

decline in research on these nonverbal behaviors after a period of considerable

research interest 48. The recent advent of the new technology has reprised

interest in these issues 24, 49, 50 because IVET can provide, in a traditional sense,

two mutually incompatible conditions that are essential to social/behavioral

studies: the ecological validity and the experimental control 24.

Nevertheless, this study has several limitations. First, we could not exclude

the confounding influences of medication and hospitalization, although no

relationship was found between medication and social behaviors as described

above. Second, sample size was not sufficient to find out detailed relationships

between nonverbal behaviors and other psychiatric symptoms, which might

further account for differential patterns of social behaviors. Third, without

recruiting bipolar depressed and euthymic patients, we could not directly

address state- versus trait-related factors influencing IPD. These should be

examined in future studies.

20

V. CONCLUSION

Our data show that BM patients exhibit avoidant-withdrawal patterns of

nonverbal social behaviors relative to normal controls, consistent with prior

findings of the negatively biased social cognition implicitly internalized in these

patients. This might imply that, regardless of apparently disruptive manic

behaviors, the sophisticated use of optimal interpersonal space and empathetic,

as well as cognitive behavioral therapy to treat avoidant social cognition veiled

behind mania could be helpful in developing secure therapeutic alliance with

BM patients. Further studies are warranted to clarify the cognitive contents and

neural correlates that affect personal space and gaze in bipolar patients across

mood states.

21

REFERENCES

1. Simon GE, Bauer MS, Ludman EJ, Operskalski BH, Unutzer J. Mood

symptoms, functional impairment, and disability in people with bipolar

disorder: specific effects of mania and depression. J Clin Psychiatry 2007

Aug;68(8):1237-45.

2. Adolphs R. Social cognition and the human brain. Trends Cogn Sci 1999

Dec;3(12):469-79.

3. Olley AL, Malhi GS, Bachelor J, Cahill CM, Mitchell PB, Berk M.

Executive functioning and theory of mind in euthymic bipolar disorder.

Bipolar Disord 2005;7 Suppl 5:43-52.

4. Kerr N, Dunbar RI, Bentall RP. Theory of mind deficits in bipolar affective

disorder. J Affect Disord 2003 Feb;73(3):253-9.

5. Lembke A, Ketter TA. Impaired recognition of facial emotion in mania.

Am J Psychiatry 2002 Feb;159(2):302-4.

6. Chen CH, Lennox B, Jacob R, Calder A, Lupson V,

Bisbrown-Chippendale R, et al. Explicit and implicit facial affect

recognition in manic and depressed States of bipolar disorder: a functional

magnetic resonance imaging study. Biol Psychiatry 2006 Jan 1;59(1):31-9.

22

7. Lennox BR, Jacob R, Calder AJ, Lupson V, Bullmore ET. Behavioural and

neurocognitive responses to sad facial affect are attenuated in patients with

mania. Psychol Med 2004 Jul;34(5):795-802.

8. Leahy RL. Decision making and mania. J Cogn Psychother

1999;13:83-105.

9. Caligiuri MP, Brown GG, Meloy MJ, Eyler LT, Kindermann SS, Eberson

S, et al. A functional magnetic resonance imaging study of cortical

asymmetry in bipolar disorder. Bipolar Disord 2004 Jun;6(3):183-96.

10. Lee GP, Meador KJ, Loring DW, Allison JD, Brown WS, Paul LK, et al.

Neural substrates of emotion as revealed by functional magnetic resonance

imaging. Cogn Behav Neurol 2004 Mar;17(1):9-17.

11. Maxwell JS, Davidson RJ. Emotion as motion: asymmetries in approach

and avoidant actions. Psychol Sci 2007 Dec;18(12):1113-9.

12. Hayduk LA. Personal space: where we now stand. Psychol Bull

1983;94(2):293-335.

13. Hall ET. Proxemics. Curr Anthropol 1968;9(2-3):83-108.

14. Kleinke CL. Gaze and eye contact: a research review. Psychol Bull 1986

Jul;100(1):78-100.

23

15. Pollack I, Kiev A. Spatial orientation and psychotherapy: an experimental

study of perception. J Nerv Ment Dis 1963 Jul;137:93-7.

16. Horowitz MJ, Duff DF, Stratton LO. Body-buffer zone; exploration of

personal space. Arch Gen Psychiatry 1964 Dec;11:651-6.

17. Horowitz MJ. Spatial behavior and psychopathology. J Nerv Ment Dis

1968 Jan;146(1):24-35.

18. Nechamkin Y, Salganik I, Modai I, Ponizovsky AM. Interpersonal distance

in schizophrenic patients: relationship to negative syndrome. Int J Soc

Psychiatry 2003 Sep;49(3):166-74.

19. Srivastava P, Mandal MK. Proximal spacing to facial affect expressions in

schizophrenia. Compr Psychiatry 1990 Mar-Apr;31(2):119-24.

20. Ku J, Jang HJ, Kim KU, Park SH, Kim JJ, Kim CH, et al. Pilot study for

assessing the behaviors of patients with schizophrenia towards a virtual

avatar. Cyberpsychol Behav 2006 Oct;9(5):531-9.

21. Park S-H, Ku J, Kim J-J, Jang HJ, Kim SY, Kim SH, et al. Increased

personal space of patients with schizophrenia in the virtual social

environment. In submission.

22. Dosey MA, Meisels M. Personal space and self-protection. J Pers Soc

Psychol 1969 Feb;11(2):93-7.

24

23. Graziano MS, Cooke DF. Parieto-frontal interactions, personal space, and

defensive behavior. Neuropsychologia 2006;44(13):2621-35.

24. Loomis JM, Blascovich JJ, Beall AC. Immersive virtual environment

technology as a basic research tool in psychology. Behav Res Methods

Instrum Comput 1999 Nov;31(4):557-64.

25. Tarr MJ, Warren WH. Virtual reality in behavioral neuroscience and

beyond. Nat Neurosci 2002 Nov;5 Suppl:1089-92.

26. American Psychiatric Association. Diagnostic and statistical manual of

mental disorders : DSM-IV-TR. 4th ed. Washington, DC: American

Psychiatric Association; 2000.

27. Hahn D-W, Lee C-H, Chon K-K. Korean adaptation of Spielberger's STAI

(K-STAI). Korean J Health Psychol 1996;1(1):1-14.

28. Young RC, Biggs JT, Ziegler VE, Meyer DA. A rating scale for mania:

reliability, validity and sensitivity. Br J Psychiatry 1978 Nov;133:429-35.

29. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry

1960 Feb;23:56-62.

30. Ku J, Jang HJ, Kim KU, Kim JH, Park SH, Lee JH, et al. Experimental

results of affective valence and arousal to avatar's facial expressions.

Cyberpsychol Behav 2005 Oct;8(5):493-503.

25

31. Bradley MM, Lang PJ. Measuring emotion: the Self-Assessment Manikin

and the Semantic Differential. J Behav Ther Exp Psychiatry 1994

Mar;25(1):49-59.

32. Hall ET. The hidden dimension. Garden City, N.Y.: Anchor Books; 1969.

33. Kim E, Jang HJ, Ku J, Kim IY, Kim C-H, Kim J-J, et al. Effects of nicotine

on interpersonal distance of patients with schizophrenia in immersive

virtual environments: a preliminary study. Korean J Psychopharmacol

2006;17(6):517-27.

34. Brune M, Sonntag C, Abdel-Hamid M, Lehmkamper C, Juckel G, Troisi A.

Nonverbal behavior during standardized interviews in patients with

schizophrenia spectrum disorders. J Nerv Ment Dis 2008

Apr;196(4):282-8.

35. Shean G, Murphy A, Meyer J. Social cognition and symptom dimensions. J

Nerv Ment Dis 2005 Nov;193(11):751-5.

36. Addington J, Addington D. Facial affect recognition and information

processing in schizophrenia and bipolar disorder. Schizophr Res 1998 Aug

17;32(3):171-81.

37. Dickerson FB, Sommerville J, Origoni AE, Ringel NB, Parente F.

Outpatients with schizophrenia and bipolar I disorder: Do they differ in

26

their cognitive and social functioning? Psychiatry Res 2001 May

10;102(1):21-7.

38. Glenberg AM, Schroeder JL, Robertson DA. Averting the gaze disengages

the environment and facilitates remembering. Mem Cognit 1998;26:651-8.

39. Brady AT, Walker MB. Interpersonal distance as a function of situationally

induced anxiety. Br J Soc Clin Psychol 1978 Jun;17(2):127-33.

40. Pini S, Maser JD, Dell'Osso L, Abelli M, Muti M, Gesi C, et al. Social

anxiety disorder comorbidity in patients with bipolar disorder: a clinical

replication. J Anxiety Disord 2006;20(8):1148-57.

41. Winters KC, Neale JM. Mania and low self-esteem. J Abnorm Psychol

1985 Aug;94(3):282-90.

42. Bentall RP, Thompson M. Emotional Stroop performance and the manic

defence. Br J Clin Psychol 1990 May;29 ( Pt 2):235-7.

43. Lyon HM, Startup M, Bentall RP. Social cognition and the manic defense:

attributions, selective attention, and self-schema in bipolar affective

disorder. J Abnorm Psychol 1999 May;108(2):273-82.

44. Goldberg JF, Gerstein RK, Wenze SJ, Welker TM, Beck AT.

Dysfunctional attitudes and cognitive schemas in bipolar manic and

unipolar depressed outpatients: implications for cognitively based

psychotherapeutics. J Nerv Ment Dis 2008 Mar;196(3):207-10.

27

45. Spezio ML, Huang PY, Castelli F, Adolphs R. Amygdala damage impairs

eye contact during conversations with real people. J Neurosci 2007 Apr

11;27(15):3994-7.

46. Altshuler L, Bookheimer S, Proenza MA, Townsend J, Sabb F, Firestine A,

et al. Increased amygdala activation during mania: a functional magnetic

resonance imaging study. Am J Psychiatry 2005 Jun;162(6):1211-3.

47. Adolphs R, Spezio M. Role of the amygdala in processing visual social

stimuli. Prog Brain Res 2006;156:363-78.

48. Uzzell D, Horne N. The influence of biological sex, sexuality and gender

role on interpersonal distance. Br J Soc Psychol 2006 Sep;45(Pt 3):579-97.

49. Bailenson JN, Blascovich J, Beall AC, Loomis JM. Equilibrium theory

revisited: mutual gaze and personal space in virtual environments. Presence

2001;10(6):583-98.

50. Bailenson JN, Blascovich J, Beall AC, Loomis JM. Interpersonal distance

in immersive virtual environments. Pers Soc Psychol Bull 2003

Jul;29(7):819-33.

28

ABSTRACT (IN KOREAN)

가상현실공간에서가상현실공간에서가상현실공간에서가상현실공간에서 대인대인대인대인 접근접근접근접근 행동으로행동으로행동으로행동으로 측정한측정한측정한측정한

양극성양극성양극성양극성 조증조증조증조증 환자의환자의환자의환자의 사회사회사회사회 행동행동행동행동 특징특징특징특징

<지도교수 조 현 상>

연세대학교 대학원 의학과

김 어 수

연구연구연구연구목적목적목적목적:::: 사회적/감정적 행동은 접근-회피의 차원에서 이해될 수 있

다. 조증 환자에서 정서처리의 긍정 편향이 보고되어 온 바대로, 조

증 환자는 부적절하게 가까운 대인거리를 유지하는 등, 사회적 행동

에 있어 결손을 보일 가능성이 있으나, 이에 대한 연구는 부족하였다.

따라서 본 연구에서는 가상현실기술을 이용하여 조증 환자들의 비언

어적 사회 행동을 측정하였다.

연구방법연구방법연구방법연구방법:::: 20명의 조증 환자와 나이와 성별을 맞춘 20명의 정상대조

군이 참여하여,‘자기소개하기 과제’를 수행하였다. 이들은 가상공

간내에서 성별과 감정가가 다른 여섯 명의 아바타 (남,녀/행복,중립,

분노)를 순차적으로 만났으며, 자기를 소개하기에 가장 가까운 거리

까지 다가간 후 간단한 인사를 먼저 건내고, 아바타의 소개를 듣고,

이후 자신의 소개를 하였다. 이 기간 동안 피험자와 아바타간의 대인

거리와 피험자의 시선각도를 측정하였다. 피험자의 특성불안, 각 아

바타에 대한 정서가와 각성도도 측정하였다.

연구결과연구결과연구결과연구결과:::: 애초의 예상과 달리, 조증 환자들은 정상군에 비하여 더

긴 대인거리를 유지하였고, 시선도 더 회피하는 것으로 나타났다. 그

러나 예상대로, 환자군 내에서는 조증 점수가 높을수록 대인거리는

줄어드는 것이 관찰되었다. 정상대조군에서는 특성불안이 높을수록

더 먼 거리를 유지하는 것이 드러났으나, 환자군에서는 이러한 현상

이 관찰되지 않았다. 아바타의 얼굴 정서에 따른 차별적 거리 두기

행동은 조증 환자에서는 행복한 아바타에서 (가장 근거리), 정상군에

29

서는 화난 아바타에서 (가장 원거리) 통계적으로 유의한 차이가 나타

났다. 특이하게, 조증 환자의 시선은 아바타의 소개를 들을 때보다

자신의 소개를 할 때 더욱 회피되는 것으로 나타났고, 이러한 차이는

특성불안과 관련된 것으로 보이나, 정상군에서는 두 기간에서의 시선

차이가 없었다.

결론결론결론결론:::: 본 연구에서 얻은 회피적 사회 행동이라는 결과는 다음의 시사

점을 갖는다고 할 수 있다. 외현적인 조증 증상에도 불구하고, 조증

환자들이 암묵적으로는, 우울증과 비슷한, 부정적인 사회인지구조를

내재화하고 있을 가능성이 있다. 따라서, 종종 호전적인 증상으로 인

한 어려움 가운데에서도, 치료자의 세련된 대인거리 조절 및 이용과,

부정적 인지구조에 대한 인지적 치료 접근이 조증 환자들과의 안전한

치료적 관계형성 및 치료과정에 있어 중요한 역할을 할 수 있을 것이

다.

----------------------------------------------------------------------------------------------------------

핵심되는 말: 개인공간, 시선, 사회행동, 조증, 가상현실