Anxiety sensitivity: multiple dimensions andhierarchic structure

Steven Taylor a,*, Brian J. Coxb

aDepartment of Psychiatry, 2255 Westbrook Mall, University of British Columbia, Vancouver, B.C.,

V6T 2Al, CanadabDepartment of Psychiatry, University of Manitoba, Winnipeg, Manitoba, Canada

Received 22 June 1997

Abstract

Anxiety sensitivity (AS) is the fear of anxiety-related sensations, based on beliefs that these sensationshave harmful consequences. AS is thought to play an important role as a diathesis for anxiety disorders,particularly panic disorder. Recent evidence suggests that AS has a hierarchical structure, consisting ofmultiple lower-order factors, which load on a single higher-order factor. If each factor corresponds to adiscrete mechanism, then the results suggest that AS arises from a hierarchic arrangement of mechan-isms. A problem with previous studies is that they were

based on the 16-item Anxiety Sensitivity Index (ASI), which may not contain enough items to revealthe type and number of lower-order factors. Accordingly, we developed the 60-item Anxiety SensitivityPro®le, which was administered to 349 university students. Factor analyses revealed four lower-orderfactors: (1) Fear of respiratory symptoms, (2) fear of cognitive dyscontrol, (3) fear of gastrointestinalsymptoms, and (4) fear of cardiac symptoms. These loaded on a single higher-order factor. The lower-order factors shared variance with the higher-order factor, but also contained unique variance. Thus, theresults suggest that AS is the product of a general factor, with independent contributions from fourspeci®c factors. # 1998 Elsevier Science Ltd. All rights reserved

1. Introduction

According to Reiss' expectancy theory (Reiss, 1991; Reiss and Havercamp, 1996; Reiss and

McNally 1985), anxiety sensitivity (AS) is an individual di�erence variable that ampli®es fears

and other anxiety reactions, and is thought to be a diathesis factor in the etiology of anxiety

disorders, especially panic disorder. AS is the fear of anxiety-related bodily sensations, based

Behaviour Research and Therapy 36 (1998) 37±51

0005-7967/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved.PII: S0005-7967(97)00071-5

BEHAVIOURRESEARCH AND

THERAPY

PERGAMON

* Author for correspondence.

on beliefs that the sensations have harmful somatic, social, or psychological consequences(Reiss, 1991; Reiss and McNally, 1985). To illustrate, palpitations are feared if the personbelieves they lead to cardiac arrest; derealization is feared if it is thought to lead to insanity;trembling is feared if the person believes it leads to ridicule or rejection. Such beliefs are at thecore of AS (Reiss, 1991; Reiss and Havercamp, 1996).A growing number of studies support Reiss' theory. AS, as assessed by the Anxiety

Sensitivity Index (ASI; Peterson and Reiss, 1987), tends to be elevated in people with anxietydisorders, compared to normal controls (Taylor et al., 1992). ASI scores predict who willrespond anxiously to panic-provocation challenges (e.g. voluntary hyperventilation: Rapee andMedoro, 1994), and who is likely to develop panic attacks and panic disorder (Maller andReiss, 1992; Schmidt et al., 1997; M.J. Telch, unpublished data, 1997; Taylor, 1995, 1996).Given the apparent importance of AS in the etiology of anxiety reactions, it is important to

determine the basic mechanisms of AS. Factor analysis is one method for addressing theseissues, because distinct factors may correspond to distinct mechanisms (Cattell, 1978). Therehave been several factor analytic studies of the ASI. Mounting evidence suggests that the ASIis multifactorial, consisting of at least three factors (see Taylor, 1996, for a review). Althoughprevious studies have di�ered somewhat in the nature of the factors extracted, it appears thatthe three factors represent: (1) fear of somatic sensations (e.g. fear of palpitations), (2) fear ofcognitive dyscontrol (e.g. fear of concentration di�culties), and (3) fear of publicly observablesymptoms (e.g. fear of blushing, trembling). In turn, these factors appear to load on a singlehigher-order factor (i.e. a general AS factor: Stewart et al., 1997; Zinbarg and Barlow, 1996).Thus, the factor analytic studies suggest that AS is composed of a combination of general andspeci®c factors, possibly representing general and speci®c mechanisms.A problem with previous factor analytic studies is that they were based on the ASI, which

contains only 16 items. It seems likely that the ASI contains too few items to reliably identifythe major AS factors. A larger scale containing a broader sampling of AS items may reveal ad-ditional factors. For example, it may be that the factor labeled `fear of somatic symptoms'actually consists of several factors, such as fear of cardiac symptoms and fear of gastrointesti-nal symptoms.The purpose of the present study was to further investigate the factorial structure of AS,

using a 60-item Anxiety Sensitivity Pro®le (ASP), which was developed for the purpose of thisstudy. The ASP was constructed on the basis of our previous research on the measurement ofAS (e.g. Cox et al., 1996; Taylor et al., 1992), and consists of six 10-item scales, each assessinga domain of AS suggested by previous studies (see Taylor, 1996, for a review). The ASPappears in the Appendix. Respondents are asked to rate the perceived dangerousness of symp-toms from six domains: (1) cardiovascular, (2) respiratory, (3) gastrointestinal, (4) publiclyobservable anxiety reactions, (5) dissociative and neurological symptoms, and (6) cognitive dys-control. Thus, the ASP assesses the cognitive aspect of AS, i.e. beliefs in the dangerousness ofvarious anxiety-related sensations.The ®rst aim of this study was to conduct an exploratory factor analysis to determine the

number of AS factors. As a check on the convergent validity of the ASP, the second aim wasto correlate the AS factor scores with scores on the ASI. It was predicted that the AS factorsshould tend to have large correlations with the ASI (i.e. r sr0.50: Cohen, 1988). The thirdaim was to conduct a hierarchical factor analysis to determine whether the ASP factors loaded

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±5138

on one or more higher-order factors. As part of this aim, we intended to compute the variancein item scores due to the higher-order factor(s), and compute the residualized variance in itemscores due to the lower-order factors. Here, residualized variance refers to the variance due tolower-order factors once the variance due to the higher-order factor(s) has been partialled out.The fourth aim was to examine the relationship (correlations) between the ASP factors

(higher- and lower-order) and trait anxiety (anxiety proneness), as assessed by the trait versionof the State±Trait Anxiety Inventory (Spielberger, 1983). Previous studies suggest that AS (asassessed by the ASI) tends to have modest correlations (r s10.30) with measures of traitanxiety, which supports the view that AS and trait anxiety are distinguishable but related con-structs. On the basis of these ®ndings it was expected that the ASP factors would have moder-ate correlations (r s10.30) with trait anxiety.

2. Method

2.1. Participants

The sample consisted of 349 university students who participated for course credit. Themean age was 19.6 yr (SD = 3.3 yr) and 61% were female. Most were white and most (94%)were single.

2.2. Measures and procedure

Participants were administered a psychometric package, which included the ASP, ASI, andthe State-Trait Anxiety Inventory (STAI). Participants were drawn from the IntroductoryPsychology subject pool and completed the measures during large testing sessions. The ASIand STAI have been well-established as psychometrically sound measures (for reviews, seePeterson and Reiss, 1987; Spielberger, 1983).The ASP, as presented in the Appendix, was developed speci®cally for the purposes of the

present study. It is a 60-item self-report scale in which respondents rate the extent that he orshe agrees with each item on a 7-point Likert scale. As mentioned earlier, the ASP consists ofsix scales, derived from the results of previous factor analytic studies of the ASI (see Taylor,1996). The items corresponding to each scale, along with the coe�cient a obtained in the pre-sent sample, are as follows: (1) fear of cardiovascular symptoms (a = 0.92; item nos. 1, 6, 8,14, 20, 31, 34, 42, 51, 55); (2) fear of respiratory symptoms (a = 0.93; item nos. 3, 15, 19, 21,26, 30, 37, 45, 47, 59); (3) fear of gastrointestinal symptoms (a = 0.88; item nos.4, 11, 16, 27,28, 32, 40, 49, 50, 57); (4) fear of publicly observable anxiety reactions (a = 0.89; item nos. 10,12, 23, 24, 33, 35, 38, 48, 53, 58); (5) fear of dissociative and neurological symptoms (a = 0.89;item nos. 5, 9, 17, 22, 29, 39, 43, 46, 52, 60); and (6) fear of cognitive dyscontrol (a= 0.94;item nos. 2, 7, 13, 18, 25, 36, 41, 44, 54, 56). Note that all six scales had good levels of internalconsistency (Nunnally, 1978). This is important because one needs to have adequate (e.g. in-ternally consistent) measures of a given content domain (scale) to determine whether or not thedomain corresponds to a distinct factor (Comrey, 1978).

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±51 39

Table

1Anxiety

SensitivityPro®le

(ASP):patternmatrices

(loadings)

andcommunalities

(h2)forthe4-factorsolution

ASPItem

III

III

IVh2

30.Youfeel

likeyou'rechoking

0.84

*(0.84

*)ÿ0

.01

(ÿ0.10)

0.13

(0.12)ÿ0

.04

(ÿ0.06)

0.72

(0.71)

15.Youfeel

likeyou'resu�ocating

0.79

*(0.78

*)ÿ0

.18

(ÿ0.18)

0.06

(0.06)

0.10

(0.09)

0.65

(0.63)

59.Youfeel

likeyoucan'tbreatheproperly

0.79

*(0.79

*)ÿ0

.01

(ÿ0.01)

0.02

(0.01)

0.09

(0.07)

0.69

(0.68)

47.Youfeel

likeyou'renotgettingenoughair

0.76

*(0.75

*)ÿ0

.01

(ÿ0.01)

0.10

(0.09)ÿ0

.01

(ÿ0.02)

0.63

(0.61)

17.Youfeel

numballover

0.69

*(0.67

*)

0.12

(0.11)ÿ0

.04

(ÿ0.04)

0.02

(0.03)

0.54

(0.51)

06.Youhavepain

inyourchest

0.67

*(0.64

*)

0.00

(0.00)

0.10

(0.09)

0.02

(0.04)

0.52

(0.49)

60.Youfeel

likethingsare

spinningaroundyou

(vertigo)

0.67

*(0.64

*)

0.15

(0.14)

0.02

(0.03)ÿ0

.05

(ÿ0.03)

0.52

(0.48)

19.Youfeel

outofbreath

even

thoughyou

haven'tbeenexertingyourself

0.66

*(0.65

*)

0.09

(0.09)

0.03

(0.03)

0.10

(0.09)

0.59

(0.57)

52.Yourface

feelsnumb

0.64

*(0.61

*)

0.27

(0.25)ÿ0

.13

(ÿ0.11)

0.00

(0.02)

0.53

(0.49)

55.Yourchestfeelstight

0.61

*(0.60

*)ÿ0

.03

(ÿ0.02)

0.12

(0.12)

0.20

(0.16)

0.57

(0.55)

26.Youhavedi�

cultysw

allowing

0.60

*(0.59

*)ÿ0

.01

(0.00)

0.17

(0.17)

0.13

(0.10)

0.57

(0.55)

03.Youfeel

likeyoucan'ttakeadeepbreath

0.060

*(0.57

*)

0.05

(0.05)

0.08

(0.08)

0.08

(0.08)

0.48

(0.46)

45.Yourthroatfeelstight

0.50

*(0.49

*)

0.03

(0.04)

0.14

(0.14)

0.22

(0.19)

0.52

(0.50)

33.Youfeel

faintorlightheaded

0.46

*(0.45

*)

0.22

(0.21)

0.22

(0.21)

0.01

(0.01)

0.52

(0.50)

34.Yourheart

startsbeatingslower

0.44

*(0.41

*)ÿ0

.04

(ÿ0.03)

0.08

(0.08)

0.19

(0.19)

0.33

(0.31)

23.Yourface

sweats

even

thoughyou'renothot

0.42

*(0.41

*)

0.21

(0.21)

0.03

(0.04)

0.24

(0.21)

0.49

(0.48)

43.Youhavetinglingsensationsin

yourlips

0.41

*(0.38

*)

0.33

*(0.30

*)

0.00

(0.02)

0.02

(0.04)

0.39

(0.37)

21.Youfeel

likesomethingisstuck

inyourthroat0.40

*(0.40

*)

0.02

(0.04)

0.21

(0.20)

0.25

(0.20)

0.48

(0.46)

09.Youfeel

likeyou'rein

afog

0.39

*(0.36

*)

0.33

(0.30

*)

0.04

(0.05)ÿ0

.03

(0.00)

0.37

(0.34)

22.Yourbodyfeelsstrangeordi�erentin

some

way

0.38

*(0.37

*)

0.35

*(0.34

*)ÿ0

.03

(ÿ0.02)

0.16

(0.15)

0.48

(0.46)

10.Hot¯ushes

sweepover

you

0.34

*(0.33

*)ÿ0

.02

(0.00)

0.28

(0.26)

0.18

(0.16)

0.39

(0.36)

37.Youfeel

thatthere'salumpin

yourthroat

0.27

(0.27)

0.16

(0.18)

0.26

(0.25)

0.23

(0.19)

0.48

(0.47)

56.Youhavedi�

cultyconcentrating

ÿ0.14

(ÿ0.14)

0.82

*(0.82

*)

0.06

(0.05)

0.10

(0.08)

0.73

(0.71)

36.Youhavetrouble

thinkingclearly

0.06

(0.06)

0.78

*(0.79

*)

0.10

(0.09)ÿ0

.01

(ÿ0.03)

0.73

(0.71)

54.Youare

easily

distracted

ÿ0.20

(ÿ0.19)

0.78

*(0.77

*)

0.12

(0.11)

0.07

(0.05)

0.66

(0.63)

25.Youcan'tkeepyourmindonatask

ÿ0.10

(ÿ0.09)

0.77

*(0.76

*)

0.04

(0.03)

0.12

(0.10)

0.67

(0.65)

41.Youhavetrouble

remem

beringthings

0.19

(0.19)

0.73

*(0.71

*)ÿ0

.02

(ÿ0.02)ÿ0

.05

(ÿ0.04)

0.62

(0.59)

46.Youfeel

``spacey''orspacedout

0.18

(0.18)

0.71

*(0.69

*)

0.06

(0.05)ÿ0

.06

(ÿ0.06)

0.63

(0.60)

18.Thoughts

seem

torace

throughyourmind

ÿ0.12

(ÿ0.11)

0.70

*(0.69

*)

0.07

(0.07)

0.18

(0.16)

0.63

(0.60)

44.Yourmindgoes

blank

0.10

(0.10)

0.67

*(0.64

*)

0.09

(0.08)ÿ0

.02

(ÿ0.01)

0.56

(0.52)

07.Yourthoughts

seem

jumbled

0.14

(0.14)

0.66

*(0.64

*)

0.01

(0.02)

0.09

(0.08)

0.60

(0.57)

13.Youkeepgettingdistracted

byunwanted

thoughts

ÿ0.11

(ÿ0.09)

0.64

*(0.62

*)

0.04

(0.04)

0.20

(0.18)

0.54

(0.51)

39.You'reawakebutfeel

likeyou'rein

adaze

0.21

(0.22)

0.63

*(0.62

*)

0.07

(0.07)

0.02

(0.01)

0.62

(0.60)

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±5140

02.Yourthoughts

seem

slower

thanusual

0.18

(0.17)

0.52

*(0.49*)

0.14

(0.13)

0.00

(0.02)

0.48

(0.45)

29.Familiarsurroundingsseem

strangeorunreal

toyou

0.45

*(0.42*)

0.51

*(0.47*)ÿ0

.02

(ÿ0.02)ÿ0

.13

0.09

0.51

(0.47)

24.Yourvoicequavers(tremblesorsoundsshaky)0.03

(0.04)

0.39

*(0.39*)

0.10

(0.11)

0.36

*(0.31

*)

0.52

(0.49)

05.Youhavetinglingsensationsin

yourhands

0.33

*(0.31*)

0.33

*(0.30*)ÿ0

.08

(ÿ0.06)

0.16

(0.16)

0.38

(0.35)

48.Yourface

blushes

red

ÿ0.10

(ÿ0.08)

0.33

*(0.33*)

0.28

(0.27)

0.31

*(0.26)

0.47

(0.44)

53.Themusclesin

yourface

twitch

0.02

(0.03)

0.31

*(0.30*)

0.25

(0.23)

0.18

(0.16)

0.36

(0.34)

11.Youhavediarrhea

0.15

(0.14)ÿ0

.11

(ÿ0.10)

0.76

*(0.69

*)ÿ0

.15

(ÿ0.11)

0.55

(0.45)

50.Youfeel

sick

inyourstomach

(nausea)

0.08

(0.08)ÿ0

.07

(ÿ0.06)

0.75

*(0.72

*)

0.07

(0.05)

0.62

(0.57)

40.Yourstomach

isupset

ÿ0.06

(ÿ0.05)ÿ0

.01

(0.00)

0.74

*(0.72

*)

0.14

(0.12)

0.62

(0.57)

27.Yourstomach

aches

ÿ0.06

(ÿ0.06)

0.14

(0.13)

0.73

*(0.71

*)

0.03

(0.02)

0.62

(0.59)

32.Youare

constipated

0.11

(0.10)

0.07

(0.07)

0.65

*(0.58

*)ÿ0

.06

(ÿ0.03)

0.50

(0.44)

49.Youfeel

bloated(gassy)

0.04

(0.04)

0.17

(0.17)

0.63

*(0.57

*)ÿ0

.04

(ÿ0.02)

0.52

(0.47)

04.Yourstomach

ismakingloudnoises

ÿ0.19

(ÿ0.17)

0.29

(0.28)

0.51

*(0.45

*)

0.07

(0.07)

0.43

(0.38)

57.Youhaveto

urinate

more

frequentlythan

usual

0.12

(0.13)

0.20

(0.20)

0.47

*(0.41

*)ÿ0

.04

(ÿ0.02)

0.40

(0.36)

38.Youfeel

likeyou'reaboutto

vomit

0.46

*(0.45*)ÿ0

.12

(ÿ0.10)

0.44

*(0.41

*)

0.05

(0.04)

0.53

(0.49)

28.Youhaveburningsensationsin

yourchest

(heartburn)

0.33

*(0.32*)

0.03

(0.04)

0.32

*(0.30

*)

0.16

(0.16)

0.44

(0.42)

14.Yourheart

beats

rapidly

0.03

(0.01)

0.11

(0.09)ÿ0

.02

(ÿ0.02)

0.82

*(0.82

*)

0.77

(0.76)

01.Yourheart

ispounding

ÿ0.17

(ÿ0.17)

0.08

(0.07)

0.11

(0.13)

0.79

*(0.75

*)

0.67

(0.62)

08.Yourheart

isbeatingso

loudthatyoucan

hearit

0.21

(0.21)

0.08

(0.08)ÿ0

.11

(ÿ0.10)

0.71

*(0.69

*)

0.69

(0.66)

20.Yourheart

poundsin

yourears

0.32

*(0.32*)

0.05

(0.04)ÿ0

.05

(ÿ0.05)

0.65

*(0.63

*)

0.71

(0.69)

31.Youfeel

yourheartbeatpulsingin

yourneck

0.31

*(0.30*)ÿ0

.03

(ÿ0.04)

0.04

(0.05)

0.65

*(0.63

*)

0.71

(0.69)

42.Yourheart

beats

erratically

0.43

*(0.42*)ÿ0

.05

(ÿ0.04)

0.02

(0.03)

0.50

*(0.47

*)

0.60

(0.57)

51.Yourheart

skipsabeat

0.37

*(0.36*)ÿ0

.02

(0.00)ÿ0

.02

(0.00)

0.47

*(0.41

*)

0.48

(0.44)

16.Youhaveaknotin

yourstomach

0.00

(0.01)

0.15

(0.17)

0.29

(0.29)

0.44

*(0.37

*)

0.51

(0.47)

58.Yourhandsare

trem

bling

0.16

(0.17)

0.33

(0.32*)

0.00

(0.02)

0.39

*(0.33

*)

0.49

(0.47)

12.Youare

``jumpy''oreasily

startled

ÿ0.03

(ÿ0.01)

0.34

*(0.34*)

0.20

(0.19)

0.36

*(0.31

*)

0.50

(0.47)

35.Youshiver

even

thoughyou'renotcold

0.04

(0.06)

0.25

(0.26)

0.20

(0.19)

0.35

*(0.29)

0.44

(0.41)

%variance

(rotatedfactors)

15.2

(14.4)

13.2

(12.5)

7.7

(6.7)

7.7

(6.7)

*Salient(r

0.30).Loadingspertain

toprincipalcomponents

analysis,

exceptforthose

inparentheses,whichare

those

derived

from

principalaxis

factoranalysis.

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±51 41

3. Results

3.1. Lower-order factor structure

The ASP item scores tended to be skewed in their distributions. Skew was substantiallyreduced by log-transforming the responses; i.e. new variable = 1n (old variable). Accordingly,all results reported in this article are based on log-transformed ASP responses. Responses tothe other questionnaires were untransformed. For the ASP scores, we calculated the Kaiser±Meyer±Olkin index of sampling adequacy. This is a measure of the factorability of the corre-lation matrices on which the factor analyses were based. The index ranges from 0 (inadequate)to 1 (excellent). In the present study, the value was 0.96, which indicated that the correlationmatrices were suitable for factor analysis.We chose principal components analysis (PCA) as the primary method of factor analysis

instead of common factor analysis because factor scores from common factor analysis are inde-terminate (Schonemann and Wang, 1972). Indeterminacy occurs when more than one set offactor scores can be constructed from a given factor pattern. PCA does not su�er from thisproblem. However, Gorsuch (1983) recommended that hierarchical analyses be based on com-mon factor analysis rather than PCA, because PCA may tend to overestimate the percentageof variance accounted for by the higher-order factor(s). Yet, Gorsuch also criticized commonfactor analysis because it su�ers from factor indeterminacy. Thus, there are pros and cons forusing each form of factor analysis. Accordingly, all factor analyses were conducted twice; onceusing PCA, and again using common factor analysis (i.e. principal axis factor analysis: PAF).As will be seen, results from PCA and PAF were very similar to one another, suggesting thatthe results were not an artifact of the method of factor analysis. In both cases, the factors wererotated using an oblique (Oblimin) transformation.The number of factors to retain was determined by parallel analysis (Horn, 1965), which has

been found to be one of the most accurate methods for identifying the number of factors(Zwick and Velicer, 1986). Following the recommendations of Longman, Cota, Holden andFekken (1989), parallel analysis was conducted twice for each factor analysis, once using themean eigenvalues and again using the 95th percentile eigenvalues. Parallel analyses of the log-transformed ASP data indicated a 4-factor solution for both the mean and 95th percentileeigenvalues. Thus, four factors were extracted for both PCA and PAF. The ®rst 10 eigenvalueswere as follows: 24.01, 4.65, 2.36, 1.96, 1.57, 1.35, 1.21, 1.11, 1.02 and 0.94.Pattern matrices (loadings), communalities, and percentage of variance for the four rotated

factors are shown in Table 1. Here it can be seen that the results for PCA and PAF were verysimilar to one another. For both forms of factor analysis, the four-factor solution attainedgood simple structure, as indicated by the small number of items with complex loadings (i.e.salient loadings on more than one factor) and small number of hyperplane items (i.e. items fail-ing to have a salient loading on any factor). The communalities show that the factorsaccounted for a moderately large proportion of variance in most items. Table 1 also shows thatthe rotated factors accounted for 6.7±15.2% of variance in item scores.Factor I had salient loadings on all 10 respiratory items. This factor also had salient loadings

for a number of items from the dissociative and neurological scale. However, the respiratoryitems tended to have higher loadings than the dissociative/neurological items. Moreover, some

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±5142

of the dissociative/neurological items represent symptoms that could be concomitants of respir-atory distress (e.g. feeling faint or lightheaded). Thus factor I was labeled fear of respiratorysymptoms.Factor II had salient loadings on all 10 items from the cognitive dyscontrol scale, and the

loadings for these items tended to be higher than loadings of other items on that factor.Accordingly, factor II was labeled fear of cognitive dyscontrol. Factor III had salient loadingsfor 9 out of 10 gastrointestinal symptoms, and smaller loadings for items assessing other symp-toms. Accordingly, this factor was labeled fear of gastrointestinal symptoms. Finally, factor IVhad salient loadings for 7 of 10 cardiac items, and smaller loadings for items assessing othersymptoms. Thus, factor IV was labeled fear of cardiac symptoms. Items assessing dissociative/neurological symptoms and publicly observable anxiety reactions were absorbed into these fourfactors, and did not emerge as distinct factors even when a ®ve-factor (i.e. over-extracted) sol-ution was speci®ed.The stability of the four-factor solution can be estimated from the magnitude of the factor

loadings (Guadagnoli and Velicer, 1988). In a series of simulation studies, Guadagnoli andVelicer found that the factor structures are likely to be stable when components possess four ormore variables with loadings r0.60. The factors in Table 1 contain numerous loadings r0.60,thereby indicating the solution is likely to be stable.To quantify the degree of similarity between PCA and PAF solutions, congruency coe�-

cients were computed (Everett, 1983), which are correlations between factor scores for corre-sponding factors. For example, the scores for factor I obtained via PCA were correlated withthe scores for factor I obtained from PAF. The same was done for factors II, III and IV. Allfactor scores reported in this article were computed according to the regression method.Congruency coe�cients ranged from 0.80 to 0.86, which suggests that the factor solutions werevery similar to one another. Factor indeterminacy may have been responsible for some attenu-ation in the magnitude of these coe�cients.Factor scores were computed to determine the correlations among the four factors, and

between the factors and the ASI and STAI. Factors derived from PCA rather than PAF wereused because the latter are indeterminate, whereas PCA yields exact factor scores. The four fac-tors were moderately correlated with one another (r s 0.36±0.48: see Table 3). The remainingcorrelations will be discussed later in this article.

3.2. Hierarchical analyses

For the lower-order factors computed from PCA the matrix of correlations between factorswas factor analyzed (via PCA) in order to identify the higher-order factor(s). Similarly, for thelower-order factors computed from PAF, the matrix of correlations between factors was factoranalyzed (via PAF) in order to identify the higher-order factor(s). One eigenvalue was greaterthan one, and so a single higher-order factor was extracted. The four eigenvalues were 2.28,0.65, 0.59 and 0.48. Thus, the results indicated a hierarchical solution in which the four lower-order ASP factors loaded on a single higher-order factor. The higher-order factor was highlycorrelated with the lower-order factors (Table 3).To further characterize these factors, the transformation developed by Schmid and Leiman

(1957) was applied using formulae provided by Jensen and Weng (1994), to separate the var-

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±51 43

Table

2

Hierarchicalanalysis:patternsmatrices

(loadings)

foronehigher-order

factorandfourresidualizedlower-order

factors

Lower-order

factors

ASPitem

Higher-order

factor

III

III

IV

30.

0.59

*(0.50

*)

0.58

*(0.66*)

ÿ0.07

(ÿ0.08)

0.09

(0.09)

ÿ0.03

(ÿ0.04)

15.

0.56

*(0.47

*)

0.55

*(0.61*)

ÿ0.12

(ÿ0.13)

0.04

(0.04)

0.06

(0.06)

59.

0.64

*(0.55

*)

0.54

*(0.61*)

ÿ0.01

(ÿ0.01)

0.01

(0.01)

0.06

(0.05)

47.

0.61

*(0.51

*)

0.52

*(0.58*)

ÿ0.01

(0.00)

0.07

(0.07)

ÿ0.01

(ÿ0.02)

17.

0.58

*(0.49

*)

0.48

*(0.52*)

0.08

(0.08)

ÿ0.03

(ÿ0.03)

0.01

(0.02)

06.

0.57

*(0.49

*)

0.46

*(0.50*)

0.00

(0.00)

0.06

(0.07)

0.02

(0.03)

60.

0.57

*(0.49

*)

0.46

*(0.50*)

0.10

(0.10)

0.02

(0.02)

ÿ0.03

(ÿ0.02)

19.

0.65

*(0.55

*)

0.46

*(0.50*)

0.06

(0.07)

0.02

(0.02)

0.06

(0.06)

52.

0.57

*(0.49

*)

0.44

*(0.48*)

0.17

(0.18)

ÿ0.08

(ÿ0.08)

0.00

(0.01)

55.

0.65

*(0.55

*)

0.42

*(0.46*)

ÿ0.02

(ÿ0.02)

0.08

(0.09)

0.11

(0.11)

26.

0.65

*(0.56

*)

0.41

*(0.46*)

ÿ0.01

(0.00)

0.12

(0.12)

0.08

(0.07)

03.

0.59

*(0.50

*)

0.41

*(0.45*)

0.03

(0.04)

0.05

(0.06)

0.05

(0.05)

45.

0.66

*(0.57

*)

0.35

*(0.38*)

0.02

(0.03)

0.09

(0.10)

0.14

(0.13)

33.

0.68

*(0.58

*)

0.32

*(0.35*)

0.14

(0.15)

0.15

(0.15)

0.00

(0.01)

34.

0.50

*(0.42

*)

0.30

*(0.32*)

ÿ0.03

(ÿ0.02)

0.05

(0.06)

0.12

(0.13)

23.

0.67

*(0.58

*)

0.29

(0.32*)

0.13

(0.15)

0.02

(0.03)

0.15

(0.14)

43.

0.57

*(0.49

*)

0.28

(0.30*)

0.21

(0.22)

0.00

(0.01)

0.02

(0.03)

21.

0.66

*(0.56

*)

0.28

(0.31*)

0.01

(0.03)

0.14

(0.15)

0.16

(0.14)

09.

0.55

*(0.47

*)

0.27

(0.28)

0.21

(0.21)

0.03

(0.03)

ÿ0.02

(0.00)

22.

0.65

*(0.56

*)

0.26

(0.29)

0.23

(0.24)

ÿ0.02

(ÿ0.01)

0.10

(0.10)

10.

0.58

*(0.50

*)

0.23

(0.26)

ÿ0.01

(0.00)

0.19

(0.19)

0.12

(0.11)

37.

0.69

*(0.60

*)

0.18

(0.21)

0.11

(0.13)

0.17

(0.18)

0.15

(0.13)

56.

0.64

*(0.58

*)

ÿ0.10

(ÿ0.11)

0.53

*(0.59*)

0.04

(0.04)

0.06

(0.06)

36.

0.71

*(0.63

*)

0.04

(0.05)

0.50

*(0.57*

0.07

(0.06)

ÿ0.01

(ÿ0.02)

54.

0.59

*(0.53

*)

ÿ0.14

(ÿ0.15)

0.50

*(0.55*)

0.08

(0.08)

0.04

(0.04)

25.

0.64

*(0.57

*)

ÿ0.07

(ÿ0.07)

0.50

*(0.55*)

0.03

(0.03)

0.07

(0.07)

41.

0.65

*(0.57

*)

0.13

(0.15)

0.47

*(0.51*)

ÿ0.01

(ÿ0.02)

ÿ0.03

(ÿ0.03)

46.

0.67

*(0.59

*)

0.12

(0.14)

0.46

*(0.50*)

0.04

(0.04)

ÿ0.04

(ÿ0.04)

18.

0.64

*(0.57

*)

ÿ0.09

(ÿ0.09)

0.45

*(0.50*)

0.04

(0.05)

0.11

(0.11)

44.

0.63

*(0.55

*)

0.07

(0.08)

0.43

*(0.46*)

0.06

(0.06)

ÿ0.01

(ÿ0.01)

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±5144

07.

0.68

*(0.60

*)

0.09

(0.11)

0.42

*(0.46*)

0.01

(0.01)

0.05

(0.06)

13.

0.60

*(0.53

*)

ÿ0.07

(ÿ0.07)

0.41

*(0.44*)

0.02

(0.03)

0.12

(0.12)

39.

0.71

*(0.62

*)

0.15

(0.17)

0.41

*(0.44*)

0.05

(0.05)

0.01

(0.01)

02.

0.64

*(0.55

*)

0.12

(0.13)

0.34

*(0.35*)

0.10

(0.10)

0.00

(0.01)

29.

0.60

*(0.51

*)

0.31

*(0.33*)

0.33

*(0.34*)

ÿ0.01

(ÿ0.01)

ÿ0.08

(ÿ0.07)

24.

0.68

*(0.60

*)

0.02

(0.03)

0.25

(0.28)

0.07

(0.08)

0.23

(0.21)

05.

0.56

*(0.49

*)

0.23

(0.24)

0.22

(0.22)

ÿ0.06

(ÿ0.04)

0.10

(0.11)

48.

0.63

*(0.55

*)

ÿ0.07

(ÿ0.06)

0.21

(0.24)

0.19

(0.20)

0.19

(0.18)

53.

0.57

*(0.50

*)

0.01

(0.02)

0.20

(0.21)

0.16

(0.17)

0.11

(0.11)

11.

0.49

*(0.41

*)

0.11

(0.11)

ÿ0.07

(ÿ0.07)

0.50

*(0.50

*)

ÿ0.09

(ÿ0.08)

50.

0.62

*(0.54

*0.06

(0.06)

ÿ0.04

(ÿ0.04)

0.49

*(0.53

*)

0.04

(0.03)

40.

0.62

*(0.54

*)

ÿ0.04

(ÿ0.04)

ÿ0.01

(0.00)

0.49

*(0.53

*)

0.09

(0.08)

27.

0.63

*(0.55

*)

ÿ0.04

(ÿ0.04)

0.09

(0.10)

0.48

*(0.52

*)

0.02

(0.01)

32.

0.57

*(0.49

*)

0.07

(0.08)

0.04

(0.05)

0.43

*(0.43

*)

ÿ0.04

(ÿ0.02)

49.

0.60

*(0.52

*)

0.03

(0.03)

0.11

(0.12)

0.42

*(0.42

*)

ÿ0.03

(ÿ0.01)

04.

0.52

*(0.45

*)

ÿ0.13

(ÿ0.13)

0.19

(0.20)

0.33

*(0.33

*)

0.04

(0.05)

57.

0.56

*(0.48

*)

0.09

(0.10)

0.13

(0.14)

0.31

*(0.30

*)

ÿ0.03

(ÿ0.01)

38.

0.61

*(0.52

*)

0.32

*(0.35*)

ÿ0.08

(ÿ0.07)

0.29

(0.30

*)

0.03

(0.03)

28.

0.63

*(0.54

*)

0.23

(0.25)

0.02

(0.03)

0.21

(0.22)

0.10

(0.11)

14.

0.72

*(0.65

*)

0.02

(0.01)

0.07

(0.07)

ÿ0.02

(ÿ0.01)

0.51

*(0.57

*)

01.

0.64

*(0.58

*)

ÿ0.12

(ÿ0.13)

0.05

(0.05)

0.08

(0.10)

0.50

*(0.52

*)

08.

0.69

*(0.62

*)

0.15

(0.16)

0.05

(0.06)

ÿ0.07

(ÿ0.07)

0.45

*(0.48

*)

20.

0.74

*(0.65

*)

0.22

(0.25)

0.03

(0.03)

ÿ0.04

(ÿ0.04)

0.41

*(0.44

*)

31.

0.74

*(0.65

*)

0.22

(0.24)

ÿ0.02

(ÿ0.03)

0.03

(0.04)

0.41

*(0.44

*)

42.

0.68

*(0.59

*)

0.29

(0.33*)

ÿ0.03

(ÿ0.03)

0.01

(0.02)

0.32

*(0.32

*)

51.

0.60

*(0.52

*)

0.26

(0.28)

ÿ0.01

(0.00)

ÿ0.01

(0.00)

0.29

(0.28)

16.

0.67

*(0.59

*)

0.00

(0.01)

0.10

(0.12)

0.19

(0.21)

0.27

(0.26)

58.

0.67

*(0.58

*)

0.11

(0.13)

0.21

(0.23)

0.00

(0.01)

0.24

(0.23)

12.

0.67

*(0.59

*)

ÿ0.02

(ÿ0.01)

0.22

(0.24)

0.13

(0.14)

0.22

(0.21)

35.

0.64

*(0.56

*)

0.03

(0.04)

0.16

(0.19)

0.13

(0.14)

0.22

(0.20)

%variance

39.5

(29.9)

7.3

(8.8)

5.5

(6.4)

3.3

(3.6)

3.3

(3.2)

*Salient(r

0.30).Loadingspertain

toprincipalcomponents

analysis,

exceptforthose

inparentheses,whichare

those

derived

from

principalaxis

factoranalysis.

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±51 45

iance in each ASP item into variance due to the higher-order factor and variance due to thelower-order factors. This procedure was also used to compute the loadings of the ASP itemson higher-order factor, and the residualized loadings of these items on the lower-order factors.Residualized loadings represent the loadings on the lower-order factors once the contributionof the higher-order factor has been partialled out. This enables one to identify the unique con-tribution of the lower-order factors after controlling for the higher-order (general) factor.The results are shown in Table 2. The pattern of loadings of the lower-order factors in

Table 2 are similar to those of Table 1, and the factor labels remain the same. The table showsthat all ASP items had salient loadings on the general AS factor, and that this factor accountedfor more variance than the lower-order factors. If factors correspond to distinct mechanisms(Cattell, 1978), then the results suggest that AS arises largely from a general AS factor, withmore modest contributions from four speci®c factors.

3.3. Correlations with the ASI and STAI-T

Table 3 shows the correlations between the ASP lower-order factors and the ASI and traitversion of the STAI. Here it can be seen that the ASP factors generally had large correlations(r sr0.50) with the ASI. These correlations were included as validity checks, and support theconvergent validity of the ASP. The table also shows hat the ASP lower-order factors tendedto have statistically signi®cant, albeit modest correlations with the trait version of the STAI,with overlapping variance ranging from 1 to 8%. These ®ndings are consistent with thosefound for the ASI (Taylor, 1996), and support the view that AS and trait anxiety are correlatedbut distinct constructs.1

4. Discussion

The ®ndings of the present study add to the growing number of studies showing that ASconsist of multiple dimensions, structured in a hierarchical manner (see Taylor, in press).Unlike previous studies, which used the 16-item ASI, our study was based on an expandedmeasured of ASÐthe 60-item ASP. The ASP, compared to the ASI, contains more items toassess each putative domain of AS. For example, the ASI contains only two items assessingfear of cardiac symptoms. In comparison, the ASP contains 10 cardiac items. With anexpanded assessment of AS we were able to identify four dimensions: (1) fear of respiratorysymptoms, (2) fear of cognitive dyscontrol, (3) fear of gastrointestinal symptoms, and (4) fearof cardiac symptoms. These loaded on a single higher-order factor. The latter, compared to thelower-order factors, accounted for more variance in item scores.

1 As a further comparison between the ASP and ASI, we could have factor analyzed the ASI items and correlated

these factors with the ASP factors. We did not perform these analyses because the ASI may contain too few itemsto yield robust (reliable) factors. Instead, we chose to simply correlate the ASP factors with the ASI total score,with the latter representing a global measure of AS.

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±5146

The magnitude of factor loadings suggest that the solution is likely to be stable (replicable)(Guadagnoli and Velicer, 1988). However, it remains to be seen whether the factor structurecan be replicated in clinical samples, such as samples of anxiety disordered patients. Studies ofthe ASI suggest that factor structures do not vary appreciably across clinical and non-clinicalpopulations (Taylor, 1996). It remains to be seen whether this holds for the ASP. This scale isenclosed in the Appendix in order to encourage further research with this instrument.The factors obtained in the present study were broadly similar to those obtained in factor

analyses of the ASI. However, here were two notable di�erences. First, evidence was found tosuggest that the fear of somatic sensations is composed of three distinct factors: fears of car-diac, respiratory, and gastrointestinal symptoms. Cox's (1996) interactional model suggests thatthese fears (and the fear of cognitive dyscontrol) are risk factors for di�erent sorts of anxietyreactions. Cox (1996) proposed that AS is multidimensional and interacts with a congruenttrigger, but not with an incongruent trigger, in producing catastrophic cognitions and accompa-nying panic. For example, fear of cardiac symptoms interacts with congruent triggers (e.g. pal-pitations) but not with incongruent triggers (e.g. derealization) to produce anxiety and panic.The second divergence from previous studies is that our factor analyses failed to reveal a fac-

tor de®ned by fear of publicly observable symptoms. We failed to ®nd such a factor eventhough more items were included to assess these fears (10 items) than were included in previousfactor analytic studies; i.e. the ASI contains only four items assessing these fears. In the presentstudy, the items assessing fear of publicly observable symptoms were absorbed into other fac-tors. Our results suggest that fear of publicly observable symptoms is not a distinct factor, butrather is an element of other factors. In other words, each of the AS factors may contain an el-ement of fear of the social consequences of one's symptoms. It will be of interest to determinewhether this result can be replicated in clinical samples.

Table 3Pearson correlations among ASP factors, Anxiety Sensitivity Index, and measure of trait anxiety

ASP lower-order factors

ASP higher-order factor I II III IV

ASP higher-order factor Ð

ASP factor I: fear of respiratorysymptoms

0.72**

ÐASP factor II: fear of cognitive

dyscontrol

0.76** 0.36**

ÐASP factor III: fear ofgastrointestinal symptoms

0.75** 0.41** 0.45**

Ð

ASP factor IV: fear of cardiacsymptoms

0.78** 0.44** 0.48** 0.41**

ÐAnxiety Sensitivity Index 0.66** 0.41** 0.57** 0.50** 0.50**

Trait version of the State±TraitAnxiety Inventory

0.26** 0.11* 0.29** 0.17** 0.22**

* P < 0.05, **P < 0.005. ASP = Anxiety Sensitivity Pro®le. All factors were derived from principal componentsanalysis. Note that the lower-order factors (i.e. factors I±IV) were not residualized).

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±51 47

The results also indicated that the dissociative/neurological items were absorbed into otherfactors, particularly the respiratory factor. It is interesting that a similar ®nding was obtainedin a factor analysis of panic attack symptoms in a panic disorder sample (Cox et al., 1994). Inthis study a dizziness factor was identi®ed which also included parasthesias and derealization.Cox et al. (1994) speculated that many of these symptoms could be produced by hyperventila-tion.Assuming that factors correspond to distinct sets of mechanisms (Cattell, 1978), then our

results suggest that AS is composed of ®ve mechanisms; one general mechanism (correspondingto the higher-order factor) and four speci®c mechanisms (corresponding to the lower-order fac-tors). Recent evidence suggests that social learning plays a role in the development of AS(Watt et al., 1998). It seems plausible that speci®c social-learning experiences may in¯uence thespeci®c factors. For example, the extent to which one learns about the dangerousness of car-diac symptoms may determine one's fear of those sensations. Similarly, the extent to which onelearns that respiratory symptoms are dangerous may determine one's fear of respiratory sen-sations. Speci®c genetic factors may also in¯uence speci®c AS dimensions. Thus, an importantquestion for future investigation is whether the AS factors are associated with distinctive pat-terns of environmental and genetic in¯uence.

Acknowledgements

This research was supported in part by grants from the Manitoba Health Research Council,Medical Research Council of Canada, and the UBC Humanities and Social Sciences ResearchGrants program. The authors gratefully acknowledge the assistance of Sharon Borger inpreparing the data for analyses. We also thank Richard E. Zinbarg for statistical advice.

Appendix

Anxiety sensitivity pro®le.

INSTRUCTIONS. It is very important that you read these instructions carefully so that you will be able to answer the questionsthat follow. The purpose of this questionnaire is to measure your level of fear of anxiety-related sensations. There are manyanxiety-related sensations, including the following: palpitations (pounding heart or accelerated heart rate), sweating, trembling,shortness of breath, chest pain or discomfort, nausea, dizziness, feelings of unreality, chills, and hot ¯ashes. People di�er in theirfears of these sensations: some people have little or no fear, others have mild or moderately severe fears, while others have verystrong fears.

Anxiety sensations are feared if a person believes that these sensations have bad consequences. For example, people are frightenedof palpitations if they believe these sensations could lead to a heart attack. People are frightened of dizziness if they believe thatthis sensation could mean that they are going crazy. People are frightened of publicly observable anxiety reactions (e.g. blushing ortrembling) if they believe these reactions could cause others to ridicule or reject them.

We would like you to do two things for each of the items on the following pages:

1. Imagine that you are experiencing the sensation. Try to imagine this as vividly as possible.2. Using the scale provided, rate the likelihood that if YOU experienced the sensation, it would lead to something bad happeningto you, such as dying, going crazy, losing control, or being ridiculed or rejected by others. There are no right or wrong answers,and all responses will remain anonymous. Please note: We are not assessing whether or not you experience these sensations as a

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±5148

result of being anxious. We want to assess whether you believe that anxiety-related sensations would lead to something bad tohappen to you.

Practice item:

Imagine that you're experiencing the following sensation. What is the likelihood that this sensation would LEAD to somethingBAD happening to YOU? Circle the number that best indicates your choice:

Not at all likely Somewhat likely Extremely likely0. Your legs feel unsteady 1 2 3 4 5 6 7

Before you answer the following questions, please place a checkmark here if you fully understand the instructions you haveread:______.

If you don't understand the instructions, please ask for clari®cation.

What is the likelihood that this sensation would LEAD tosomething BAD happening to YOU?

Not at all likely Somewhat likely Extremely likely1. Your heart is pounding 1 2 3 4 5 6 7

2. Your thoughts seem slower than usual 1 2 3 4 5 6 73. You feel like you can't take a deep breath 1 2 3 4 5 6 74. Your stomach is making loud noises 1 2 3 4 5 6 75. You have tingling sensations in your hands 1 2 3 4 5 6 7

6. You have pain in your chest 1 2 3 4 5 6 77. Your thoughts seem jumbled 1 2 3 4 5 6 78. Your heart is beating so loud that you can hear it 1 2 3 4 5 6 7

9. You feel like you're in a fog 1 2 3 4 5 6 710. Hot ¯ushes sweep over you 1 2 3 4 5 6 711. You have diarrhea 1 2 3 4 5 6 7

12. You are ``jumpy'' or easily startled 1 2 3 4 5 6 713. You keep getting distracted by unwanted

thoughts1 2 3 4 5 6 7

14. Your heart beats rapidly 1 2 3 4 5 6 715. You feel like you're su�ocating 1 2 3 4 5 6 716. You have a knot in your stomach 1 2 3 4 5 6 717. You feel numb all over 1 2 3 4 5 6 7

18. Thoughts seem to race through your mind 1 2 3 4 5 6 719. You feel out of breath even though you haven't

been exerting yourself1 2 3 4 5 6 7

20. Your heart pounds in your ears 1 2 3 4 5 6 721. You feel like something is stuck in your throat 1 2 3 4 5 6 722. Your body feels strange or di�erent in some way 1 2 3 4 5 6 7

23. Your face sweats even though you're not hot 1 2 3 4 5 6 724. Your voice quavers (trembles or sounds shaky) 1 2 3 4 5 6 725. You can't keep your mind on a task 1 2 3 4 5 6 726. You have di�culty swallowing 1 2 3 4 5 6 7

27. Your stomach aches 1 2 3 4 5 6 728. You have burning sensations in your chest

(heartburn)1 2 3 4 5 6 7

29. Familiar surroundings seem strange or unreal toyou

1 2 3 4 5 6 7

30. You feel like you're choking 1 2 3 4 5 6 7

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±51 49

Not at all likely Somewhat likely Extremely likely31. You feel your heartbeat pulsing in your neck 1 2 3 4 5 6 7

32. You are constipated 1 2 3 4 5 6 733. You feel faint or lightheaded 1 2 3 4 5 6 734. Your heart starts beating slower 1 2 3 4 5 6 7

35. You shiver even though you're not cold 1 2 3 4 5 6 736. You have trouble thinking clearly 1 2 3 4 5 6 737. You feel that there's a lump in your throat 1 2 3 4 5 6 7

38. You feel like you're about to vomit 1 2 3 4 5 6 739. You're awake but feel like you're in a daze 1 2 3 4 5 6 740. Your stomach is upset 1 2 3 4 5 6 7

41. You have trouble remembering things 1 2 3 4 5 6 742. Your heart beats erratically 1 2 3 4 5 6 743. You have tingling sensations in your lips 1 2 3 4 5 6 744. Your mind goes blank 1 2 3 4 5 6 7

45. Your throat feels tight 1 2 3 4 5 6 746. You feel ``spacey'' or spaced out 1 2 3 4 5 6 747. You feel like you're not getting enough air 1 2 3 4 5 6 7

48. Your face blushes red 1 2 3 4 5 6 749. You feel bloated (gassy) 1 2 3 4 5 6 750. You feel sick in your stomach (nausea) 1 2 3 4 5 6 7

51. Your heart skips a beat 1 2 3 4 5 6 752. Your face feels numb 1 2 3 4 5 6 753. The muscles in your face twitch 1 2 3 4 5 6 7

54. You are easily distracted 1 2 3 4 5 6 755. Your chest feels tight 1 2 3 4 5 6 756. You have di�culty concentrating 1 2 3 4 5 6 757. You have to urinate more frequently than usual 1 2 3 4 5 6 7

58. Your hands are trembling 1 2 3 4 5 6 759. You feel like you can't breathe properly 1 2 3 4 5 6 760. You feel like things are spinning around you

(vertigo)

1 2 3 4 5 6 7

References

Cattell, R. B. (1978). The scienti®c use of factor analysis in the behavioral and life sciences. New York: Plenum.

Clark, D. M. (1986). A cognitive approach to panic. Behaviour Research and Therapy, 24, 461±470.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences. (2nd edn.). Hillsdale, NJ: Erlbaum.

Comrey, A. L. (1978). Common methodological problems in factor analytic studies. Journal of Consulting and Clinical Psychology, 46,

648±659.

Cox, B. J. (1996).The nature and assessment of catastrophic thoughts in panic disorder. Behaviour Research and Therapy, 34, 363±374.

Cox, B. J., Swinson, R. P., Endler, N. S., & Norton, G. R. (1994).The symptom structure of panic attacks. Comprehensive Psychiatry,

35, 349±353.

Cox, B. J., Taylor, S., Borger, S., Fuentes, K. & Ross, L. (November, 1996). Development of an expanded Anxiety Sensitivity Index:

Multiple dimensions and their correlates. In Taylor S. (Chair) New studies on the psychopathology of anxiety sensitivity.

Symposium presented at the 30th annual meeting of the Association for Advancement of Behavior Therapy. New York, NY.

Everett, J. E. (1983). Factor comparability as a means of determining the number of factors and their rotation.Multivariate Behavioral

Research, 18, 197±218.

Gorsuch, R. L. (1983). Factor analysis. (2nd edn.). Hillsdale, NJ: Erlbaum.

Guadagnoli, E., & Velicer, W. F. (1988). Relation of sample size to the stability of component patterns. Psychological Bulletin, 103,

265±275.

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±5150

Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179±185.

Jensen, A. R., & Weng, L-J. (1994). What is a good g?. Intelligence, 18, 231±258.

Longman, R. S., Cota, A. A., Holden, R. R., & Fekken, G. C. (1989). A regression equation for the parallel analysis criterion in

principal components analysis: Mean and 95th percentile eigenvalues. Multivariate Behavioral Research, 24, 59±69.

Maller, R. G., & Reiss, S. (1992). Anxiety sensitivity in 1984 and panic attacks in 1987. Journal of Anxiety Disorders, 6, 241±247.

Nunnally, J. C. (1978). Psychometric theory. (2nd edn.). New York: McGraw-Hill.

Peterson, R. A. & Reiss, S. (1987). Anxiety Sensitivity Index Manual. Palos Heights, IL: International Diagnostic Systems.

Rapee, R. M., & Medoro, L. (1994). Fear of physical sensations and trait anxiety as mediators of the response to hyperventilation in

nonclinical subjects. Journal of Abnormal Psychology, 103, 693±699.

Reiss, S. (1991). Expectancy theory of fear, anxiety, and panic. Clinical Psychology Review, 11, 141±153.

Reiss, S., & Havercamp, S. (1996). The sensitivity theory of motivation: Implications for psychopathology. Behaviour Research and

Therapy, 34, 621±632.

Reiss, S. & McNally, R. J. (1985). The expectancy model of fear. In S. Reiss & R. Bootzin (Eds.), Theoretical issues in behavior therapy

(pp. 107±121). New York: Academic Press.

Schmid, J., & Leiman, J. M. (1957). The development of hierarchical factor solutions. Psychometrika, 22, 53±61.

Schmidt, N. B., Lerew, D. R., & Jackson, R. J. (1997). The role of anxiety sensitivity in the pathogenesis of panic: Prospective evalu-

ation of spontaneous panic attacks during acute stress. Journal of Abnormal Psychology, 106, 355±364.

Schonemann, P. H., & Wang, M. M. (1972). Some new results on factor indeterminacy. Psychometrika, 37, 61±91.

Spielberger, C. D. (1983). Manual for the State±Trait Anxiety Inventory (Form Y). Palo Alto, CA: Consulting Psychologists Press.

Stewart, S. H., Taylor, S., & Baker, J. M. (1997).Gender di�erences in dimensions of anxiety sensitivity. Journal of Anxiety Disorders,

11, 179±200.

Taylor, S. (1995). Anxiety sensitivity: Theoretical perspectives and recent ®ndings. Behaviour Research and Therapy, 33, 243±258.

Taylor, S. (1996). Nature and assessment of anxiety sensitivity: Reply to Lilienfeld, Turner and Jacob (1996). Journal of Anxiety

Disorders, 10, 425±451.

Taylor, S. (Ed.) (in press). Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety. Mahwah, NJ: Erlbaum.

Taylor, S., Koch, W. J., & McNally, R. J. (1992). How does anxiety sensitivity vary across the anxiety disorders?. Journal of Anxiety

Disorders, 7, 249±259.

Taylor, S., Koch, W. J., Rachman, S. J. & McLean, P. D. (November 1992). Fear of cognitive and somatic phenomena. Presented at the

26th meeting of the Association for Advancement of Behavior Therapy, Boston, MA.

Watt, M. C., Stewart, S. H. & Cox, B. J. (1998). A retrospective study of the learning history origins of anxiety sensitivity. Behaviour

Research and Therapy, (in press).

Zinbarg, R. E. & Barlow, D. H. (November 1996). The hierarchical structure of anxiety sensitivity. Paper presented at the 30th annual

meeting of the Association for Advancement of Behavior Therapy, New York, NY.

Zwick, W. R., & Velicer, W. F. (1986). Comparison of ®ve rules for determining the number of components to retain. Psychological

Bulletin, 99, 432±442.

S. Taylor and B. J. Cox / Behaviour Research and Therapy 36 (1998) 37±51 51