Anxiety sensitivity profile: dimensional structure and relationship with temperament and character

Anxiety sensitivity profile: dimensional structure and relationship with

temperament and character

Does, A.J.W. van der; Duijsens, I.J.; Eurelings-Bontekoe, E.H.M.; Verschuur, M.J.; Spinhoven,

P.

Citation

Does, A. J. W. van der, Duijsens, I. J., Eurelings-Bontekoe, E. H. M., Verschuur, M. J., &

Spinhoven, P. (2003). Anxiety sensitivity profile: dimensional structure and relationship with

temperament and character. Psychotherapy And Psychosomatics, 72, 217-222. Retrieved from

https://hdl.handle.net/1887/15007

Version:

Not Applicable (or Unknown)

License:

Leiden University Non-exclusive license

Downloaded from:

https://hdl.handle.net/1887/15007

Departments of·Psycho logVandbPsychiatry, LeidenUniversity.Lelljen, The Netherlands

Regular Art

ic le

Wil

lem

Van de

r D

oesa

.

b

_Inge

_Du

_{ij sens >}

_Eli

_sabeth

_E

_ur

_e

_li

_{nq s-}

_B

_{o nte}

_k

_{o e}

_s

M

argot Ve

rschuu rs

Phil

ip

S

pinh ov en v>

nallv intended i

t

,

m

ight

be

a defendable soluti

on.

H

ow

-ever

,

the number of items is much to

o

h

igh

,

with m

any

subscales consist

ing

of semanti

c c

lusters. ASP

scores

we

re

found to

be

weakly rel

ated

t

o

th

e

temperam

ent

dimension h

a

rm

avoidance

, c

or robo rati ng earlier fin

d

-ings that we

re

not sta

tistically

s

ignificant be

cause o

f

small samp

le

s

izes.

Co

nclus ions

:The ASP may be

s

hort-e

ned f

r

om

6

0

t

o

24

ite

ms

wi

thout

l

oss of re

liability

o

r

co

nten t. F

ut

u

r

e s

tud ies u

sing c

hallenge

para

digm

s

and

s

tud

ie

s

with ge

neral h

ospital patients may further in

ves

-tiga

te t

he us

efu

lness

of a shortened version of th

e

ASP

.

Co9v<ogl'llC2003

s

.

Karve<

AO.BaN'

[)r \J ....\""" \kf ~

l)epalt_ ."."'" ... lndm La" .... l)

....

~"'""C'I.~

"l !n3"'~1.ado<tI(Thc"-ct~1

Td •.11'I <~ " \1"] ... ,.11~!1""~ll[.\la ,1•.u><k'<.Io.>o(.I,.. Jcodo:nun" nl

A

nxiety sens

it ivity

(A

S) r

efers

t

o

the

fear of bodrly scn

-sa

nons.

ba

sed o

n

the belief t

hat

bodi

l

y

sensation

s

have

harmfu

l co

nseque nces

[

I}.

A

S I

nclud

es.

b

ut

is no

t

sy

no

ny-m

ous

wit

h,

th

e

t

enden cy

t

o

m

isinte rp ret bodily se

n

sat

i

on s

c

atastrophically

.

F

or i

nstance

,

hi

gh A

S

indiv

idual

s

may

know perfectl

y

w

cll

that

ce

rtain

bod

il

y

se

nsat ions

(e.g

.

palpita

tion

s)

arc n

ot s

igns

of

immediat

e

dang

er

(

c.g.

im

pending

h

eart

attack). However

.

th

ey

ma

y

belie

ve

th

at

ha

ving

the

se se

nsat

i

ons cau

ses

d

ama ge i

n th

e

l

ong

r

un

.

A

ccordin g

t

o

R

eiss

and M

eNally

(

11

a

nd M

cNally (21

.

A

S

is

a

v

u

l

n

e

ra

b

ility fa

ctor

f

o r

th

e

develo

pme nt

of pani

c

Psychother Psychosom 2003;72:217-222

001

;

10

.1159I000070786

'\" 'I>1cOI'I,~ill

.. hl'JlP ...PI'> ~llS.~31J"AG.

a..xr

OOH-11900J.-o'~ol-(I!P'SI~\tl.{) 13,· ~lhlltjf,I~U I.\1J,I h'l"". l.Ircn •

... L,o'''''-

.~

...

KA

RGER

Ab

. trac t

B

ackground

:

Anxiety

sensmvrtv

(AS)

,

the

b

elief that

bodily se

nsations

have ha

rmfu

l

consequenc

es.

is a

r

e

l

i

-able

pr

ed

i

ctor of panic attacks in both c

linical

an

d n

on

-clinical populations. Recen

tly,

a new meas

u

re

o

f

AS has

been proposed. The AS pro

file

(ASP

)

was designed to be

a more comprehensive measure of AS

.

and to be more

suitable for the measurement

of different

AS

dtrnen

-s

io ns. Prel

im inary

evidence

(

co llege student sample

)

s

uggests that the ASP has

4

d

imensi ons .

In the present

s

tudy

,

the dimen

sional

structure of the ASP was furthe

r

inv

e

s

l

i

g ated

.

as

w

ell

as its relationship

w

it h tempera

-ment and character traits

.

M

eth ods

:

Exploratory

and

co

nfirm ato ry fa

ctor

analysis o

f

ASP

sco

r

es

i

n two large

s

am p

l

es of psychiat

ric

outpatients and noncl

inical

con

-tro

ls

(com

bi ned

n

=

742). Co

r

relations

a

nd

pa

rt

ial

c

or rela

-tions o

f A

SP wi

th

tempe

rament

an

d

c

haracte r.

R

esults

:

Exploratory

fact

or

analysis yie

lded

a single

A

S teeter

.

H

owever , c

o n

f

irm ato ry factor analy

sis

showed that the

6

-dimens

ional s

tructure

.

as Ta

ylor

and Cox had orig

i.

KeyWords

Anxiety sensitivity·

Personality . Temperament.

Character- Anxi

ety ·

Panic dtsorcer

.

Dep

ression

Anxiety Sensitivity

Profile:

D

imensional

Structure

and Relationship

attac

ks. n

ot a co

nseq ue nce

o

f

o

r

condi

tioned r

esponse to

panic attacks o

r

i

nte nse

bodi

ly

se

nsation

s.

T

his

impl

ies

that A

S

ca

n

also be meas

ured

in i

ndividuals

who have

ne

ver

expe

rie nced

panic a

ttacks

.

T

he most w

ide ly

use

d i

nstru men t

t

o assess

A

S is

th

e

t

e-uem

Anxiety Se

nsitivi ty Ind

ex (A

SI) (

3).

Th

e

ASI

has

been show

n t

o

p

rodu ce

r

elia ble scores

,

a

nd t

o

diff

erent iat e

betwee

n pan

ic d

isorde r p

at ients a

nd pati

ent s

with

o

ther

p

sychiatri c

d

isorders

(i

nclu din g a

nxiety

diso

rders)

[

4].

Furth

erm ore,

AS

I

sco

res pr

ed ict

th

e occ

ur rence

o

f pani

c

a

ttac ks in h

ealthy

individual

s

durin

g

a high

l

y s

tressful

pe

riod (

5,

61

.

The que

st

i

on

of whethe

r

AS i

s

uni-

o

r

multidimension

-a

l ha

s bee

n

c

ontroversial [

7, 8].

Thi

s

q

u

csuon

i

s

important

because i

f AS i

s

i

n

fa

ct

mu

ltid

imen sional ,

relation

sh ips

be

tween

a globa

l mea

sure of

AS

a

nd

ot

he r

v

ariables m

ay

be o

bscured . F

or i

nst ance

.

th

e re

lat ionship

be

tween

A

S

a

nd r

espon se

t

o

CO

2

ch

allenge m

ay be

l

ess s

trong than th

e

r

elat ionship be

tween a

(pos

tulated)

A

S

dime

nsio n

's

uffo-cation fear' a

nd

respo

nse

10

CO

_2.

R

esea rch fi

ndi ngs in t

his

area see

med t

o be i

nco nsiste nt.

with abou

t

an eq

ual

nu

m-ber of studies su

pporting

the uni- and mu

lt idi mensional

view.

H

owever

,

i

t h

as bee

n

s

ho wn t

hat a

hi

erarchi cal

model o

f AS

(

as m

easu red by the

ASI)

,

cons

ist ing

of 3

par

-t

ially dis

ti nct

firs

t-orde r

fac

tors

a

nd I

ge

neral

seco

nd

-orde

r f

actor .

r

esol ves

th

ese i

nco ns istencies (

9).

Th

e 3

fi

rst-o

rde r

fac

tors

a

re: ph

ysical co

nce rns: m

ental

in

capa cit

a-t

ion co

nce rns, and

soc

ial

co

ncerns

[

9]. How

ever,

th

e

r

esult s of

fac

tor

anal

ysis

depend h

ea vily o

n the

compre-hen

siven ess

of t

he

mea

sureme

nt

s.

Th

e

ASI

h

a

s

only 16

it

e

m

s,

and may

be

too short t

o

reliabl

y

assess separat

e

dim

en sions

of AS [10]. On the othe

r h

an d

,

it h

as

also be

en

ar

gued

that the AS

I

contain

s

a numb

er

of problemat

i

c

i

tems [1

1).

A rea

naly

sis o

f previou

s s

tudies showed that

an

I l-i

tem

ASI

,

with

o

nly

2 s

ubsca les, might be a m

ore

p

recise

i

nde x

of

AS

[

1

1).

As every

clin

ician

kn

ows, co

mmo n

ca

tast rophic

cogn

i-nons of

pat

ien ts w

ith

pan

ic

d

isorder

i

nclude m

ore

th

an

t

wo or

t

hr ee

t

hem es,

e.g

.

f

ear of

hea n

f

ai lure.

su

ffocat ion

fear, fear of

l

osing co

ntro l.

fea

r

of fai

nti ng,

o

r f

ear of

a

ct-i

ng ridi

culou s.

Furth

erm or e. as

n

oted abov

e.

so

me

pa-t

ients

belie

ve

tha

t th

eir sym

pto ms

cause fu

tu re

damage.

Co

nsidering th

e i

mpo rta nce

of t

he

conce

pt

of

AS f

or cog

-nitive

t

heo ry.

t

he nu

mbe r

a

nd

nature of firs

t-order AS

dime

nsions i

s

n

ot

tri

vial .

It

wo

uld

be i

nteres ti ng

,

f

or

i

nstan ce.

t

o see w

he ther d

ifferent

AS pr

ofiles e

xist f

or

pa

tients wh

o

p

resent

at

a c

ardio logy d

epartment

w

ith

a

typ ical c

hest pai

n,

pat

ient s

wh

o a

re

see

n b

y

neurol

ogists

f

or

u

nexpla inable

di

zzine ss,

and pati

ent s

w

h

o

con

sult

pul-m

on ologists

because of att

acks of

breathle

ssne ss.

Sin

ce

o

nl

y

a

sma

ll

minori

t

y

of th

ese

patient

s

have

D

SM-IV

panic disorder

{12

- 15].

these

populanons ma

y

con

stitut e

examples of sing

le-dimen

sional

AS elevation

s.

I

n an attemp

t t

o

m

easure 6 dimen

sio

ns

of AS.

Tayl

or

and Cox

(1

6) created a new measure

,

the 6Q-item Armct

y

S

ensit ivity

P

rofile (AS

P)

.

Wh

en filli

ng

out thi

s

qu

esti

on-naire

. s

ub

jec

t

s

ar

c as

ked

to imagine experiencing ea

ch of

60 bodi

ly

sensa

tions (

e.g.

p

alpit ations), and then rate on a

scale fro

m I t

o 7 the

lik

elihood t

hat t

his sensation 'will

l

ead

t

o so

me thing b

ad

h

ap pen ing

t

o yo

u'. Th

e

A

SP was

d

esigned

to measure th

e

f

ollowing 6 d

imens ions:

fe

ar

o

f

ca

rdiovascular symptom

s;

r

es

p

i

rat

o

ry

sy

mpto ms;

gas-troimcstinal

s

ympto ms: n

eurol ogical/di ssociat ive sy

mp-t

om s;

publicl

y

ob

servabl e

an

xiet y sy

mptoms,

a

nd

cogm-ri

vc

d

yscontrol.

H

owever,

a facto

r-ana lytic

s

t udy w

ith 3

49

college st

ude nts

rev

ealed 4 dim

ension s,

a

nd I

secon

d-o

rde r

gene

ral dim

en sion o

f

AS [16].

T

he 4 d

imensions

w

ere

f

ear o

f

(I

)

r

espir ato ry

sym

p

to

m

s, (2) cognitive d

ys-co

nt rol

,

(3) gas

t roin test inal

sy

mpto ms.

a

nd

(4)

cardiac

sy

mpt

oms. Note t

hat i

n co

mparison

wi

th

the A

SI

facto

rs

[9]. '

phys ical

co

nce rns

'

s

plit

into 3 dime

nsion

s.

wher

eas

th

ere is no dime

nsion

'social concerns'.

Th

e

a

pp

ropriate-ness of this

4-d

im ensional model is Q

uestionab le

.

hew-eve

r

,

since the eige

nval uc

p

lot i

ndicat ed a 1

-

or 2-dim

cn-sio

n

al s

tr uct ure( the

first 3 cigenvalues we

re

2

4.0

.

4.7,

and

2.

4).

Th

ese

r

esults see

m t

o s

uggest that t

he A

SP m

ay be a

u

seful

alt

ernati ve

m

easure of

A

S, bu

t

a co

llege

studen

t

sa

mple ha

s se

rious

l

im

i

tat i

o

n

s.

I

t

i

s

a h

ighly se

lective

sam-ple. parti

cularly

a

s

regard

s

th

e

r

estri ct ed

rang

e of

a

ge

an

d

edu

cation

and the expected ran

ge o

f ASP

sco

res. Ind

eed ,

the ASP item d

istribution

s

wer

e so s

kewed that the

y

had

t

o

be

l

og-tr ansformed bef

o

r

e

fun her anal

ysis

(16

].

A

replt-c

at ion

o

f t

h

is

stud

y

in

a

m

or e

d

iverse

populat

ion see

ms

warranted

b

efo

r

e makin

g a

ny

co

ncl usions r

egardin g

t

he

numb

er o

f dime

n

sion s o

f

t

he ASP

.

Th

e

prima

ry

a

im of t

he

pr

esent s

tudy wa

s

t

o

in

vest igate

th

e

f

act or st

ruct ure

of the

A

SP

i

n tw

o l

arge

sa

mples:

(

I)

a rep

resentat ive

samp

le

of

th

e

n

ormal

populat

ion . and (2) co

nsecu tive

psyc

hiatric

o

utpat ient

a

d missions

.

R

elative ly

littl

e

i

s

kn

own abou

t

the rela

tionship

be-twee

n

A

S

a

nd pe

rsonality. with the exception of trait anx

i

-c

ry.

Co

nsens us h

as e

merged

that AS

i

s re

lated

,

but not

r

edu cible.

t

o

t

rait anxiety (

1

7

).

As

rega

rds

o

the r

perso

nali-t

y

f

actors,

AS m

easures

w

ere

m

od erat ely corre

lated

w

ith

ne

uroti cism

in a

st

udy

w

ith

9

4

o

utpatie nts

{

18J. In

a

noth-e

r

s

tudy

[

19], th

e co

rrelation

be

twee n AS

a

nd

negati

ve

e

mot iona lity wa

s

r

=

0.

40.

Nega

ti ve

e

mot ionality,

t

he t

en-den

cy

to experienc

e

n

egati ve

aff

ects,

i

s

a co

ncept re

lated

to neuroti

cism .

Tw

o s

mall

s

t udies hav

e

exam

ined

t

he

218

Van dcr Doc

sl

r

elation ship

be

tween

AS

a

nd Clonin

gers

ps

ychobiological

sys

tem

or perso

nality [

20, 2

1). A

S

wa

s mode

rately

co

rre-l

ated wi

th h

arm avoi

dance.

a

ltho ugh

t

he sampl

e s

izes

were so

s

mall

t

hat

corre

latio ns

re

il

s

hort

or s

ignificance.

H

ar m a

voida nce m

easu res se

nsitivi ty

to

d

anger, an

d i

s

re

lated t

o be

havic ral

i

nhibi t ion [

22. 23)

.

Th

e seco

nd

ai

m

or the prese

nt

s

tudy w

as

t

o

in

vestigate

th

e

re

lation ship

betwee

n

A

S. Cl

oning ers personality di

mensio ns and an

xi-c

ry

and

dep

ression

i

n bo

th cl

inical

and n

onclinical

samp

les.

M

eth od

Penta pams and Procedure

A norm

al

sam

ple

co

nsisted

of i

nd ividua ls

w

ho wer

e ra

ndo mly

se

lected

from

th

e te

lephone d

irect ory .

w

hile ca

re

was

t

ak en

1

0

select a

re

pr

esentative

proportion o

f

peo

ple

from

u

rba n a

nd ru

ral areas. A

t

ot

al

o

f 6

1

7

in

div id uals

recei

ved

a

tel

epho ne call a

nd

agreed

t

o pa

r

-nc

jpa

te.

O

f these

.

4

85 (78

.6 %)

re

turned

the ques

tio nnaires

t

ha t

were

se

nt t

o

th

em

b

y

mai

l.

A

c

linica

l

sample consis

ted

o

f 289

consec

utive

referral

s

t

o

an o

ut-panent clime

o

f a psychia

tric

hosp

ital

(n

•

106) a

nd

a

regional

am

bu-lat

ory

mental health centc

r

(n "" 183)

.

Both faci

lities

primaril

y

admit

pa

tien ts With neurQI

N::

disorders

;

ps

ydlO(lC

patient

s

are referred t

o

ot

her locat

ions o

f these

fecuuies

.

At both

o

utpa t ien t facilit

ies,

pan

e nts

were

as

ked to rill out questionnaires at the

l

im

e

of the intake

procedure

.

It

was

pointed out t

o

patient

s

that the data ...."eregathered

f

or researc

h purposes

.

and informed

co

nse nt wa

s

obtained

.

Qut'm Q/ltwi

ri'

s

I

n \

ie

/I

'

S

e

luimi

ty.

T

he

A

SP (16

)

wa

s

t

ran slated

into

Du

t

c

h. and

lh

e

oa nsteuc n

was translated b

ac k

un

o

Engl

ish

by a

n

ative

English

pro

f

ess

i

on

altranslato r.

No ch

anges in

mea

ning

be

tween th

e orig

ina l

and bad ..-

translate d

i

te ms w

ere fo

und.

~»mplonu.

Curren

t

(past

week )

leve

l

s o

f

a

nx iety

a

nd d

epr ession

were meas

ured w

ith t

he la-

item self-report

H

osp

ital

Anxiet

y

and

D

epre

ss

io n Sc

ale

(2

4]. A

n a

ut hor ized

tra

nslation

a

nd

ve

ry larg

e no

rm

groups arc availab

le fo

r

th

is m

ea sure [2

51.

P

ersonahty

T

he

T

ernpcram cm and Cha

racter

Inventory (TCll

{26

- 281.

wh

i

ch

is the s

uccesso r

o

f

the

T

rid imensional

Pe

rsonalit y

Qu

esti onna ire .

assesses

4

dimen

s

i

o

n

s

oftempcrament

(novelt

y

seek

-10

9.

h

arm av

oida

nce

.

reward dependen

ce,

and persisten

ce)

and

3

c

harac ter dimensions (self-drrec

tedn

ess .

coope

rative ness

.

a

nd

se

lf-r

ra nscc nde nce j

.

I

n one of the patienl sampl

es

(regional

m

ental hea

lth

cc

mer].

th

e orig

mal

2

40- ltem

T

CI

""'3.5

a

d m inistered .

I

n the othe

r

panc

r u

sa

mple and in th

e

n

ormal sa

mple. an authori

zed

short v

er-sio

n

(105

ite

msj

o

f'the

T

CI was

used

,

whi

ch

co

rrelates

highl

y'

...

nh th

e

f

ull-length

\

ersK>n

(19)

.

T

o

en

ab le co

mbined data anal

ysis.

the

T

CI

dime

nsions

w

e

re

calculated

o

n the ba

SISo

f the

1

0

5

it

em s

included in

born vers

scns in

eac

h

o

f the

sa

mples (2

8].

fJaw

-1l1a{,-siJ

1

:\

p/QT

alor}

'

fa

CIQT

A

nalvsis

T

he dime

n

siona

l

structure o

f

the

·

\

S

I

'

wa

s

ev

aluated

with Sim

ultaneou

s C

ompo nent Anal

ysis

(SCA)

f-l

O

-.12

J

o

n th

e co

rrelation mal ri' of

t

h

e

60

items of the AS

P

.

Th

e

n

um ber o

f factors wa

s

d

erc rmm ed

o

n

the

basi

s o

f

the sizes of the

An~iet y

Scnsrtiv

rty Pr

ofile

e

i

gen\ alues and the scree t

est

(331, $CA dilT

ers

from

P

rin

c

ipa l C

om

-ponen

ts

A

na lysis

(

PCA)

in that Identical

co

mpo nen ts are defined

....

hkh opllma

lly

account f

or

lhe: variance

i

n several populatl

on s

slm uhaneously .When

a measure i

s

administe

red

ID

m

ore

than

o

ne

populati

o

n

(e

.g.

patient

s.

n

crmet s).

o

r repeatedl

y

in th

e

sam

e

popela

-n

on ,

separat

e

PeAs

will

not

u

suall y y

Ield

i

dentical

fact

ors,

A

con-gru

ence

coefficient

can then be ca

lculat ed .

H

o

...

ev

er .

...

hen

t

he: fac

-t

ors

are clea

rly diff

erent

i

n each populati

on .

tfus

beco

mes diffICUltt

o

i

nterpret

.

F

urt he rm o re

.

even

i

f t

he:

fa

ctors

are clearly d

ifferent

.

It

m

a

y

st

ill

be possib

le

t

hat

very

S

imila r

soI

uuo ns

cou

ld

be

found a

cross

po

pulanon

s.

b

ut

th

at tbe

y s

im

ply

did

n

ot

appear

ID

separat

e

P

CA

s

[32J

.

In SCA. co

mpo ne nt

weight

s

are d

efined

a

s

in PeA

.

Th

e a

ppro-p

riat enes s of

the

resu

lting

f

act

o

r

so

l

ution

i

s evaluated in a

S

Im ilar

w

ay as i

n

con

firm atory P

CA

:

co

mparing

the

a

mo unt of varian

ce

ex

pla ined b

y

the compon

ent s of

t

he SC

A

with the va

riance

explained

by se

pa rat e

explo ratory

PeA

s: eva

luating t

he correl

at ion

of I

tems

wi

th

th

e

correspondi

ng

compon

ent s (loadi

ngs),

a

nd examini

ng th

e

co

rre

lat

ions among co

mpon

ent s a

nd

rh

e

co

rre lations betw

een

com-pon

ent s

an

d o

the r variabl

es

fo

r each po

pulat ion

[32].

Confirmato

ry F

actor An

aly

sis

Th

e percentage of explained var

i-ance was subseq

uently

com

pared

w

ith t

he amount

o

f

vananc

c

accoun

ted f

or by

th

e

6

dimension

s

o

riginally proposed b

y

T

a

ylo

r

a

nd

Cox

(1

61

b

y

mea

ns

o

f

a

Mu

lt iple Group

M

ethod A

nal

YSIS

I

MGM

)

(

34).

i.e.

a

m

at rix wi

th

b

inary

elem

ent s o

nly.

such

t

hat each variable

has a ....elght of o

ne

on

o

ne component and zemon the o

ther

compo-nen

ts,

M

GM analysis

was

foll

o

...

ed

b

y

a varimax rotat

ion

of

weijhu,

f

ollo

...

ed by obli

que t

rans form al lOll

.

Th

i

s

rota

tion

procedure Y

i

eld s

o

pt imal resuhs in those

c

ases

wh

ere

it

I

S

unlikely thal an

o

n hogon aJ

co

mpone nt structure ...

ill

be

detected

[

32].

1be

appropriaten

ess o

f

the resulting fac

tor

solution wa

s

mdearedb

y

the followi

n

g:

(

a)

the

varia

nce

accou

nted

for b

y

the: a pr

iori

matrix in

co

mpa rison with th

e

va

ria

n

ce acco

un ted

fo

r

b

y e

xplorato ry (6-factor

solunon)

PeA

s in

both samples separatel

y.

and

(

b) the number

of

subscales

well

accoun

ted

for

,

a

s

evidenced

b

y

the

s

ubst ant ial correlation

(

>0 .40)

o

f

each item w

ith

the presupposed co

rresponding

component

in

eac

h

sample.

Th

e rela

t ionsh ips

a

mo ng

the

ASP

and

pe

rson ality d

imen

sion s

were mvesrigate

d

u

sing

P

ea rson

co

rre lations. Also. pa

rtial

c

crrela-no

ns

w

ere

c

alculated

10

contro

l

for curre

nt

levels of anx

ielY

and

d

ep ression.

R

esul t s

P

relim inary An

alyses

Pri

or

1

0

a

nalysis.

da

ta

were screene

d

for acc

ura

cy o

r

da

ta

en

try

,

m

issing

va

lues

,

n

o rmality of dis

tribution

and

t

he presence o

f

outliers (35J

.

Ni

neteen

pa

tien ts

and 13

n

ormal pa

rtici pants

w

ith m

o re than

4

mi

ss

ing A

SP it

ems

we

re

exc

luded

fro

m

analysis (

4.1

%

or a

ll

cases). Six

ty-one

cases

h

ad

I

m

issing

item

.

25 cases

h

ad 2 or 3. I case had 4

.

Th

ese cases we

re r

etained

:

t

he mi

ssing val

ues

were

re-p

laced w

ith th

e

m

ean fo

r

a

ll

ca

ses.

o

rrnalit

y

or di

smbu

-ti

ons for eac

h ASP

i

tem

was exam

ined

by calcu

lating

ske

wncss and kurt

osis. Sixtee

n

of t

he

60 items h

ad

ei

ther

s

kcwness

o

r

kurto

sis above 1 or be

low

-

I.

H

owever.

t

h

e

T

abl e

1

.

Descnpnon

O

flhc

samp

l

es

W~'"

Anxiet

y

n

..

Age

Educa-

Dep

res-years

_tOO"

,,,'"

Normal sample

472

57.0

4

5.5(1 5.:!)

1.9

(O.8

)

5

.7(3.8)

.1 9(. 1) )

P

at ient sa

mples

Ontpauem clime

95

62.1

3

6.0 ( 1:!.4)

1.61

0.7)

1

1.8

(4

.2)

9

.6

(.4.k)

\1

ema l health cemcr

I1S

59.4

3

7.0(9 .7)

1.5

(0

.7)

ID(4.8)

9.

9 (5.0)

R

esults

a

re mean va

lues

:

st

andard

dev

iat ions

are given

in

paren

theses. A

nxict} and

dep-ression

w

ere

m

easured w

ith

t

he H

osp ital

A

nxiet y an

d

Dep

ress i

on

Scare. educat

ion

on a

sc

aie

from

1

to 3

(

low , medium,

h

igh) .

deviati

on s

were small. Square roo

t

and

l

og trans

for matio n

w

ere bo

th examin

ed.

but thi

s

made matters worse. 11wa

s

th

erefore

decided to u

se

u

mmn

sformcd sc

o res. The KMO

i

ndex o

f

sa

mpling adequa

cy

wa

s

0

.98,

whi

ch

mea

n

s

t

hat

th

e co

rrelat ion matrix wa

s

quit

e s

uitable f

or

factor anal

y-StS.

Participants

A desc

ription

o

f

bot

h

sa

mples i

s s

hown

i

n t

able

I

.

Th

e

t

wo

pa

t ient

sa

mples d

id

n

ot

d

iffer on

an

y v

ariable and

we

re

subseq

uently treated

as o

ne

sa

mple. In

t

he

e

o mbincd

pa

tien t sa

mple, th

e

ASP mean t

otal sco

re wa

s

195

.9 (

SO

=

7

4,4),

w

hereas the n

ormal sa

mple had a mean scor

e o

f

1

53.1

(SD

=

61.

3).

T

aylor a

nd

Co

x (16

]

did not pr

ovide

m

ean sc

o res for their colleg

e s

tude nt

sa

mp

l

e.

E

xploratory

FactorAnaly

sis

T

h

ere

were 7 components w

ith

cigc

nvalue

greate

r

tha

n

I in the no

rma l

sa

mple

,

a

nd

8

s

uch compo

ne nts

in

t

he

patien

t

sample.

In

the no

rmal

sam

ple, t

he first 7 cige

n-va

lue

s

were 25.9

,

5.4

,

2

.3,

1.9

,

1.5

,

1.2 and 1.0.

I

n

t

h

e

patient sa

mple. t

hese

v

alues were al

mo

s

t

identical.

T

he

sc

ree

c

riterio n indicated a

1-

o

r

z

-

tacto r solution. The 2·

f

actor so

lut ion was in

spect ed.

Th

ese

tw

o

fac

tors

exp

lained

5

1.9%

v

ariance. compar

ed

w

ith

52

.0%

for two

se

parate

P

CAs

10

e

ac h

sa

mple. Th

e

f

actors

wer

e

highl

y

correlated

:

r

=

0

.67

in th

e

n

crmal sa

mple. and r

=

0.

68

in the patien

t

sa

mple. Furth

erm or e, eac

h

o

f th

e 6

0 it

em s

had high l

oad-in

gs o

n

bo

th

f

actors.

s

uggest ing that

t

he

so

lut ion wa

s o

ver-e

xtrac ted.

Co

nfirmatory F

actor

A

naly

s

is

Using

MGM anal

ysis.

th

e

SIX

)()·it

em sc

ales that Ta

y-lor an

d C

ox

{16] had int

ended

for the ASP were e

va

luat-e

d.

I

n term

s o

f ex

plai ned

var

ian ce.

th

ese

a

pri

ori scale

s

did rather well

:

61.7

%

(a

verage o

ve

r t

wo samp

les)

a

s

co

rn-par

ed

with 6

3.9%

for

e

-tacro

r

P

CA solutions. The corr

cla-l

io n

s

among the

s

c

a

les w

ere

hi

gh.

and varied from r

=

0.58

t

o r

=

0

.84

in the normal

s

ample. and from r

=

0

.52

t

o

r

=

0

.82

in the patient sa

mple

.

W

ith a

lmost no

e

xcept ions.

the

l

oadin gs

w

ere

h

igh

es

t

o

n th

e co

rrect

sc

ale

(

i.e.,

eac

h i

tem

had the highest

co

rrelat ion

w

it h th

e sca

le t

hat

i

t be

longs

to

a

cco

rding

t

o

Taylor an

d

C

o x

(1

6)).

H

owever. w

ithout

an~

e

xcept ion.

eac

h

i

tem al

so

h

ad h

igh

loadi

ngso n th

e ot

her

5

sc

a les

,

in

bo

th

sa

m ples

(a

ll l

oadings >0.27 and

t

ypi

c

ally

>0

,45)

.

Cro

nbach's

a

lpha

was

hi

gher than 0.88 for

e

ach

fac

to r.

Seco

nd-Order F

actorAnalysi

s

The 6 T

aylor

and

Co

x dim

ensions

wer

e subjected to a

seco

nd-order explorat

ory

fa

ctor an

alysis: SCA

was used to

d

efine

t

he

o

pt imal

so

l

ut ion for

bo

th

sa

mp

les

. PCA

o

n

the

c

omb ined samples (n

=

742) wa

s a

lso

ca

rr

ie

d

D

uI. T

he

re

s

ult

s

of bo

th

a

nal

yses

w

ere

alm

ost

identical a

nd clearly

yielded a l

-factor

solutio

n

.

I

n

P

C

A

,

t

he

fi

rst

3 eig

en values

we

re 4

,6

,

0

.6 and

0

.3. The

s

ing

le

factor ex

plai n

ed

76

% of

t

he va

riance.

Clo

ser

A

ssessment of t

he

6·

FaClorSo

lution

Loo

king

over th

e

toitem

s

th

at

mak

e

up ea

ch of

th

e 6

s

ubscales, it becom

es

appa

rent

that

so

me ar

c

alm

ost

id

en-ti

cal.

F

or

in

stance ,

the f

ollowing 3

it

em s be

long t

o the

'

f

ea

r of

r

es

pirato ry

sy

m ptoms'

sca

le:

(

I)

yo

u f

eel like

yo

u' re n

ot g

ett ing

e

no ugh

air: (

2)

yo

u f

eel li

ke

you can

'

t

breath

e

properl

y.

and

(3

)

yo

u

fee

l

o

ut

o

f

brea

th

even

th

ough y

ou haven

't bee

n

exe

rt ing

yo

urse lf. Th

e

f

ollowing

3

ar

e

part

o

f th

e

same

sca

le:

(

4)

yo

u f

eci

like

some

thing

IS

s

tuck

i

n

yo

ur throa

t: (5) you f

eci l

ike yo

u're

choking

,

and

(

6) yo

u ha

ve

diffi

cult y s

wallowing.

It

i

s

un

su rpr ising

t

hat

thi

s s

ubscale has h

igh i

nternal

co

ns istency. Ind

eed . a fa

c

-tor

co

ns

is

t

ing

o

f th

ese

it

em s

ma

y

be

rega

rded a

s a

seman-ti

c

clu

ster

with little ps

ychological s

ignificance. Th

ere-f

ore.

a con

siderab

l

y s

hortened

ve

rsion of the ASP

w

as

a

lso

evaluated.

B

ecau se a

ll

item

s

had high intercor

relati cn

s

.

Tabla 2

. Correlations

of

AS with pc:

rsonalll y and

sympto ms

wo

rdin

gst 'u sc

ares

m

e

wh

en ....

)

, p

art icipan ts

are asked to

i

magi

ne

e

xper ienc ing

a se

nsatio n

and t

hen

rate the

proba-b

ility

that t

his 'will

l

ead

t

o some

thing b

ad happening t

o

m

e' . Second

ly, t

he

n

umber of i

te m

s

wa

s

increa

sed

dra

sti-cally

f

rom

1

6

t

o 60

.

Our result

s

show tha

t

t

he

scal

e c

an

be

reduced to 24 i

tems

,

wi

thout

d

amage

to bo

th

the

co

nte nts

and the reliabilities

o

f

th

e

s

ubsca

les

(very high

co

rrela-tio

ns

bet

ween

shortened and full-l

ength v

ersions.

a

nd

high int

ern al co

nsistencies).

Th

e s

ign

ifican t

correlat

ions

of AS with harm av

oid

-a

nce

co

rroborate

com

parable

findings with t

he

AS\.

which had

be

en

o

btained in

s

mall

samp

les and

we

r

e

there

-for

e

n

ot s

tat

ist

ically

s

ig

n

ificant

[2

0.

2

1]. Th

e

f

act

t

hat

t

h

e

p

reviou sly

r

eport ed corr

elatio ns

o

f

harm avoidanc

e

with

t

he AS

I

and t

he pr

esent cor

relations

w

ith

the AS

P

w

ere

o

f

t

he sa

me m

agn it ude

p

rovides som

e

indirect support

for

t

he

posi

tio n

t

hat

t

he

A

SP

is a

n

alt

ernative

mea

sure o

f AS

On

th

e othe

r

hand. I of the dim

en sion s o

f the AS

I

.

f

ear of

losing cognitive

c

ont

r

ol, ha

s

been f

ound

t

o

be

s

pecific ally

rela

ted

to depression {38-40]

.

wh

ich

wa

s

not th

e case in

the

p

resen t s

tudy

with t

he

AS

P

.

Of

co

urse. the (short

ened)

AS

P

and

t

he AS

I

s

hould

be compar

ed

direct

l

y

in a

s

i

ngle

s

tudy. T

he use

fulness

ofa shortene

d

ASP may further

be

eval

uated

by

in

vestigatin g (a) gene

ral

hospi

tal

pa

tient

s

w

ith

s

pec i

fic

a

nx ie

t ies.

e.g.

ca

rdiac ph

obi a. a

nd

(b) th

e

pred

icti

ve po

wer

of

AS

P

score

s

i

n

di

fferent

cha

lleng

e

p

arad

i

gms.

O

ther a

reas

of futu

re

r

esea

rch

include the r

ela-tions

hip

of AS

a

nd

l

oc us of contro

l

or

ientation

[41]

a

nd

the so-ca

lled

causal cata

s

t rophical

mis

int

erpretati on s as

assessed du

ring

c

ognitive therap

y

[42J

.

22

1

pat

ient

sample

(

n

'"

210)

-

0.35"

O.oJ

0.

16*

0.02

0.3

2

u

-

0.04

-

0.09

normal

sa

mpk

(n. 472)

-0

.07

0

.36

--0.11

0

.02

_0

.24

u

-0.

04

0.09

Temper am ent

Novelty seeking

Harm

avoida nce

R

eward depende

nce

Persistence

Character

Sdf-d

irectedn

css

Cooperati

veness

Sdf-trnncendc

nce

Symptoms Amncly

Depressjon

Two-ta iled

significa nce. "

p

<0 .05:

"

1'< 0 .001 .

ASP

total

score

Discussion

\n

\lct)

Sensmvuv

Profile

A n

umbe r

of

diff

e

r

ent c

r

i

t

eria

exis

ts

t

o

d

e

t

e

rmin

e

th

e

nu

mber o

f fa

ctor s to be

e

xtract ed

in a

fac

to

r

analy

sis,

c.g

.

t

he

n

umbe r

o

f

cige

nvalues

gre

at er

th

an

I

[3

6], th

e se

ree

te

st

[331, a

nd

inte

rp reta bi

lity

[37]

.

T

he se

ree t

est clea

rly

s

ugges

t

e

d

a

t

-fa

ct

or solut

ion f

ort

h

e AS

P

.

Sinee

t

he ex

t

rac-tion of more tha

n

I fa

cto

r

resu

lted

in

h

ighly co

rrelated

fac

tors , i

t seem

s

safe t

o

conclude t

hat

the AS

P

is

l

-d

imcn-sio

nal. Ne

vertheless .

there

m

ay bc circurn

sta

nces

in w

hi

ch

u i

s

useful to distingui

sh

the 6 dimen

sion s

as originall

y

propo

sed

b

y

Ta

ylor

and Cox (

16).

Confi

rma tory

factor

analy

sis

s

howed

that this so

lut ion

i

s

de

fendab le.

des

pite

the high

int

cr

c

cr rclat ion s a

mong

fac

tors . A

s arg

ued

ubove. certa

in

ge

neral

hos

pita l

pa

tie nts

m

ay ha

ve e

le-vate

d

sco

res

on

I

speci

fic dim

en sion of A

S,

so

th

e

ASP

m

ay he usef

ul

i

n

t

hese

po

pularions.

It

i

s also clear

.

ho

weve r,

t

hat

t

he A

S

P

can be s

horte

ned

co

ns i

derably. Co

mpared

w

i

th

the AS

!. T

aylor a

nd

Cox

[

lfi} lOt rod

u

ced tw

o c

h

anges .

F

ir

s

tl

y.

t

he instruc

tio ns

a

nd

w

ordi

ng

of item

s

were cha

nged

:

in

stead

of condit

i

o

na l

this version was

c

reated by s

imply

taking

t

he

first 4

i

te ms

of eac

h

subscalc. leading to a 2

4-ile m

sca

le. Ea

ch of the 6

s

horte

n

e

d

s

ubscales

h

ad very hig

h

co

rrelat ions

with i

t

s

fu

ll-length

versio

n

(ra

nge:

I'

=-

0

.92 to

I'

=

0.95)

. Th

e

int

er

-nal cons

iste nc ies

(Cro

nbac

h's alp

ha) o

f th

e

4-item

sca

les

we

re

goo

d

(ra

nge: 0

.79-0 .88).

T

h

e s

ho

rt AS

P

con

s

i

sts o

f

i

te ms 1

-19 ,21 - 24

and

27

of t

he

60

·ite

m

AS

P

[16].

R

elariunships between

A

S

.

Sym

ptom

s

an

d P

ersonality

The co

rre latio n

s

be

tween A

S a

nd

sy

mp toms

a

nd

pe

r-so

nali

t

y a

re

de

pi

cted

i

n

tab

le

2.

I

n bo

th

sa

mples

,

mode

r-atel

y

s

tro ng

cor

relat ions

ex

ist

betwee

n

A

S and

a

nx iety

,

depression and harm avoida

nce

.

A

n

egat ive correlat

ion

e

xists between AS a

nd

sclf-dircctedncs

s.

The co

rrelat ion

s

of t

he

6 AS

P

dime

nsio ns

with perso

na lity and

sy

m ptoms

w

e

re also

c

alcula

t

ed; t

he p

att ern was

n

OI d

ifferent f

rom

that of the AS

P t

otal scor

e. All

di

mensio ns

co

rrelated

h

igher

w

ith

a

nx iety

(aro

und

r

=

0

.35) than with d

epres-s

ion

(a

r

ound

I'

=

0

.25);

there w

ere

n

o s

pec

i

fically

s

tro

ng

co

rre

l

at ions

o

f an

y o

f t h

e

6 dimen

s

i

on s

wit

h

eit

her

anxi

ety

or

d

epr essio n.

A

fte r

co

nt rolling

for c

urrent l

evels of a

nx iety

a

nd

de

pression

,

t

he

pa

rt ial

corr

elations

be

twee n

A

S

and

h

arm

avoidance re

mained

signi

ficant.

bu

t dr

op ped

1

0

I'

=

0.

13

(p

<

0.05) in t

he

clinical samp

le

.

a

nd

I'

=

0.17 (p

<

0.00

I

)

m thc normal

samp

le. P

art ial correlations

wit

h

se

1'~)"ChothcrPsychosom 2003;72:211-222

R

eferenc es

R<,ISSS.McNallyRJ

l

k

expcctancy model of fear: In R~i!i!iS, Booctln Rleds): Thcorellcal

IssUl$'" BchavlOr Thmpy. N<'w Yol1c, Aca-demK: Press . 19&5,pp101-121

2 McNallyRJ:Pan icDiso rder. ACritical A

naly-~IS.

New

York .Gu llfo rd .1994

Pelerso nRA. Rei" 5:An,icty Sens u i...ityIn . dcx Manu al , ed 2,wcrthm gton,lDS,199~,

4 Ta)lor S.Koc h WJ.McNallyRJ:HQ\lodoes

anxlct~sensitl"'lt) ...'a ry acruM the anx ietyd,,...

orden'!1Aru Dtsor d1992 ~6:249-2 59.

5 Schmldt N8.

LerewD

R.Jad50n RJ-Tberok: of anxiet) scnsll1...tW tll the ~t~sof panIC:Pl'O:S(Jl'l:tl ~<'evaJuahOl\ of sponlllnc:ous panICallacb dunq acule !>ITeS5.J Abnonn PsycholI99 7J(6) 55-J6.t

6

Schrnldt

1\'

8

.

Lerew

DR

.

Jackson

RJ P

roseec-til e Claluallon of anxlCly SCl\SlIi"'ltym the pathocmcslli ofpanIC:Reph ca tion and eaten-SIOn.

J

Ab no rm1'S)'tho t 1999;108:532-H1 . 7 Lihrn feldSO. Turner SM. 1aoob RG, Anxlcty

scnsiti vu y: Anexanunano n

of

lhc:oretic.al and methodolog Ical lSS~. Adv Benav ResThcr

J99 3:15:147-1 83,

8 McNa llyRJ:Anxietysensit ivity isdisun gursh-able from traIt aOllc ty: InRapceRM (00)' Cur-rentCentre...·CTSleI 10theAnxll'IYDisorders. Newl'ort ,Guilford.1996 . PP 214- 221-9 ZmoofI RE. Mob lmanJ.1i0fll NU:

Dimen-Sion S of an;uet) SCfl5tl1"'lt); In Ta)<1or 5 (cd) An11C1)'Sl'nslllvlly.Mah...~h.ErIbaum.1999. pp&3-114.

10 Cox

BJ

,

P:lIl1cerJDA.S.. I05OORP: AnxlC1yscn-SIl1\II)<:Confirmatory C\'idmcc fora mullLdi-mrns10lllll construct. IkllavRes1bcr 1996:3-4

591~S98.

I1 SLarsMA_OUoMW. loc ker00.McNally lU. Sch mtd lNB,Fa" aM,f'oll ad MH: The .

nX1-ety senSlltVlly index Item analys is andsuggcs.

uons for refine me nt. J I'crs Assess 2001;77: 212-294.

12 1'0HackMH,Kr. d ln R,Ono MW, Worthing·

Ion

J

.

Gould R, Saba ltl\O SA.Rose n baum JF:

Prevaleflcc ofpan icto patl<'nts rc:f<'rredfor pul-mon ary fUflct1Ol1testln&at a majo r med ical

center.Am 1 Psye hlll ry1996:1H:110-1 13, 13 ....orth mllonJJ,Pollac kMH,OtmMW, Gould

RA.Sabal1noSA.GoIdnutoL Roscnba um JF

La' TH: P:lIIlICdIsorder 10 cml'l'f:CT"lCY.. -ard patwnlS "'llh C'hc:stpaJn. J Ncn M~ l Dis

1991; 18 5:274- 21f>_

14 VanPcsk i-Qo:stcrNa n AS, Spmho"<'n P.Van dcr ~ AlW. Wilk nu LNA, Sll'ft PJ, Is

t~a spa:irlCrelationship bet~ a!>Ihma

andpanICdlsor~8thavRes "fh,c,r1996;34 333-340 .

222

15 Van Pesk l-Oost crbaa n AS,Spmho>cn P,Van dcr

Does

AlW, Bruschke AVG. Rooijm ans

IIGM: Cognit"e change following cogrlltl'e

behaviou raltherapy for non-cardiacchatpaIn PsychotherPsycboeom1999 :68:214- 220.

16 Taylor S.CO"DJ:At\~ietysensitivity. Multiplc

dimension s and hlerarehic structu re. Bchav

ResTh<'r199 8;36:37- 51.

11 Lrlicnfeld

SO

:

AnXiety scflsitiv ity and the

structure ofpenonal lty:mTaylo rS{ed )An~i.

<'Iy Senslllvlty. \b hwah. Erlba um. 1999,pp

149- 180 ,

18 Arrindcll WA:Thefear

of

fear COflCqlI::

Evi-ceece

in favourofmuhldimrnsiooality. lkhli '

Reo;'fbIor 1993:31:S07-518.

19 !.Ilicnfdd ~Thcrelation of anxiety 5alS, tIV>-I) 10hlgbcrandiowaordl'r pcnonahty dim

S1Ol1s..JAbnorm Ps)~hoIl991;106:539-S44

20 SavlOlti FM.Grandt5,Sa,lUlIG.Erme nnrnR,

8artolucc i

G

.

Cont i

S

,

Fa...a GA: Charac ter.

oIogia.l tf1lllsof

recovered

pati<'nt!i""lIh

pa

mc

disorde r and agora phobi a. J Affect Drscrd

199I :HllJ _I17,

21 Fava GA, Grandi S,Belluaedc P. Savron G. Ram AR,COnli5,Savioui FM:

Bcnzodraze-pinesandan.,iety5l'ns itivi ty in paniediSQrder.

Pros Neuropsycbopharmacol Bioi P!)'Cht<llry 1')94: 18;1163-1168

n

Gra)' JA;The Neur0p6~holoc of Anxll't). I\.cwYon . Oxford Um"'cnl ty Press.1982. 23 Carver

cs.

Wh,teTL:8ehavioral in hlbltion.

bcha ...iontIactl~.hon.andalTcetio-rn:sponses tolmprndiq~llrdandpu.nlshmrn l:theBISI

BAS !ClIks. J Pen Sec PS)'t hol 1994 :6 1:319 -333.

24 Lr,mond AS,Sna llh RP,TbeHospit alAn~lC1y

and lkpress ion Scale. Acta Ps)'thl3 tr Sand

19 83:6 1:361-310 .

25 Sptnhov cn P,C>rtM1J.S1Ol'kersI'P'\' Kcmpcn

GUM,Speckeos AEM ,Van Ifemert AM: A validation sludyof thc Hos p ilal Anxi,1Yand Depressio nScale (I[ADS) indiffere nt groups

of Dutch subjects,I'sychol Med 1997;21:363 -370.

26 CIon ingcr CR. Pn ybc:ck TR. Svrak ic OM.

WeUd RD,TheTemperam ent andCharacter 100'rnt o<y[Tell: A GUIde10lISIXvd opm en t andU~.St

Louis.Center for tl'll'

Psychob

iolo-C)ofPe~ahty.I994

27 DclaRICS.DuIJ!iol'fl$IJ.

C1oniflla"

eR:

T ern-pc<amrnt.ctl3DCter and penonallty duor dcts.

JP<'BDisord 199 8:12:362- 312.

28 DuIJ5C1l~LJ.Spmoo'rn P.Ooekooe JG,

.1

'

mon A.Eu~lmgs-8ontek~EHM ."fh,c,Dutch

Tem JXram<,n t and Character In v<,ntol)' ffCl)

Drmc

nssonal

struc ture. reliability and \alidlt) ina nor ma landpsychiatricoutpatient sample

Pers

Ind

Diff

er

2000;28:487- 499.

29 Duijsen~lJ.Spinhu\'en

P

,

Vcrschu ur

M

. Pur

e-lings-Bontekoc EII M, De Nederlandse ver -koereTem peramen t en Karaktersc haal(VTCll NedTijd sch rPs~hol l 999;54:116-283 . JOMill$apRE. Mercdlth

w

,

Component anaJY~1

In CroM-!iocuonaJ and Iongltudmal data. Ps)-ChornC10b 19 88 ;53 123--134.

31 Kilm ItAL. Ten

Ikrtl:

JMF Alt<'lmtilll kllSl

squares algpnthms for simu lu ll«lW C'Ompl.)-llC111!ianalysJSWIthequalcomponent ..-rrght rn;ttrica Intwoor man::popu1at~ ~ho>

mctnla.1989;5 4461-.47).

J2 K1CTSIlA L:SCAUser's Manual. A Progam for SlmullallCOUSComponents AnalysIS of

Vanabk:s Mea su red in Two or More"o pu la

non s.Gronmgen.I'mG am ma . l m . 33 Ca n cll RB:The screetestfo rthenumbc r c f fac

-tors.MuluvarIkha"Res1966:J:245-216 34 Nun nally 1:Psychom et ricTheo ry. New Yorl

McGraw·H, ll,1918

35 Tabac h n N:k

00.

FidellLS: USlO&MuUI" an at<, Stansncs.

W

2. Ne'/> York. Harpe r Colues,

1989.

36 Kaee rH:A llQ(e 011GUUllWl'S\oIo-nbound for the number ofoommon factOR..Multi.ar Bc:h3...R<'SI96 L 1:249-216

31 Gorusch R. Fattor AniIIyslS. HL1IsdaJc. Erl -baum,19&3,

3!

Ouo MW.PoIlackMH,Fa'a M. Llcttllo R

gosen ba u m 1F: E1e'llted 8n1ic1)sensin ...u\ scores In peueetsWIth majord~il:,""Cor -rclata and ehanp w,th anlldeprcss;lnt treat-mcnt.

J

An1 Drsord 199 5:9: 117-123.

39 Taylc r S,KochWJ,Wood yS.McLcan P:AnxI

-ety scnsil iv ity and depressiOl1:How arc the) related'! J AbnormPsyeholl996 ;10 5:474-479

40 Schmidt NB,

Ltrei\-

'

OR,Join er YE:An XIl'!) SCIIsit ivltyand thepatholl'enesis of anxiety and

depression:E...Id<,1lCC:for symptom speciflcity

BehavRes

Thcr

1998;36:16 S- 171.

41

Batt

«

A. Spmoo'rn P. Vandn

Doo::5

Aj....

Van BaB.om AlLM.Van

Dyet

R.

I...ocus

of control onrntatlOll In panic dIsorder and the d,ffermtw effects of tl'C3tlDalt. Ps)'Chothtf PsydJo5om 2OO2:7l;8}-89 .

42 Ball er A.Spmholen p. Van Balkom AlL\{ VanDyd R Rd e...anaofas!ieSSmmt

of

C'OglI

-tions dutint panIC"anacb iDthetrea tme n tof

panIC dlloOrdcr Ps)<:hother Psychoso m 2001 , 11:15&-J6 1.

Van dcr DocsIDUtJsenslE urelings-Bonlek oelV erschuurlSpinho"en