ARTICLE IN PRESS

Journal of Affective Disorders 1 (2001) 000000

4

www.elsevier.com / locate / jad

10

Research report

11

A general anxiety-prone cognitive style in anxiety disorders

E.H. Uhlenhuth a , *, Vladan Starcevic b , Teddy D. Warner a , William Matuzas c ,
Teresita McCarty a , Brian Roberts a , Steven Jenkusky a

12
13

14
15
16

Department of Psychiatry, University of New Mexico School of Medicine, 2400 Tucker, N.E., Albuquerque, NM 87131 -5326, USA
b
Hunter Mental Health Service, University of Newcastle, Newcastle, NSW, Australia
c
Maine General Medical Center, Waterville, ME, USA
Received 26 July 2001; accepted 25 October 2001

17

Abstract

20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35

Objective: This study compared scores on the Anxious Thoughts & Tendencies (AT&T) questionnaire, a putative measure
of a general anxiety-prone cognitive style, among patients with panic disorder without agoraphobia (PD, n 5 62), panic
disorder with agoraphobia (PDA, n 5 29), generalized anxiety disorder (GAD, n 5 43), limited social phobia (LSP, n 5 34),
generalized social phobia (GSP, n 5 33), and community residents (n 5 318). Method: Candidates for treatment studies
completed a diagnostic interview and the AT&T. AT&T scores were compared among anxious groups using analysis of
variance. Then depressed and non-depressed patients were compared. The final analysis compared anxious groups without
comorbid depressive or anxiety disorders. Results: AT&T scores were highest in PDA patients, followed by patients with
GAD or GSP, then patients with PD or LSP, with community residents lowest. Depressed patients were higher than
non-depressed patients. Patients with current or past comorbid depressive disorders did not differ. The ranking of anxious
groups on AT&T scores remained unchanged after exclusion of patients with comorbid disorders. Patients with PD or LSP
without comorbidity had the same AT&T levels as the community sample. Conclusions: The AT&T discriminates PDA and
GAD from PD per our hypothesis. The low AT&T levels among patients with PD and LSP suggest no association with a
general anxiety-prone cognitive style. LSP and GSP may be distinct disorders. The cognitive style assessed by the AT&T is
also associated with depression and may be a marker for vulnerability to depression. Definitive conclusions about a
pathogenic role for cognitions require their measurement before the onset of the disorder. 2001 Published by Elsevier
Science B.V.

36

Keywords: Anxiety disorders; Cognitive style; Comorbidity; Depressive disorders

37

5
6
7
1
2

Klein (1980) first suggested that panic disorder

39

1. Introduction

* Corresponding author. Tel.: 11-505-272-8876; fax: 11-505272-5572.

E-mail address: uhli@unm.edu (E.H. Uhlenhuth).

38

EC

TE
D

PR

18
19

with agoraphobia is a complex disorder arising out of

interaction between a biological component, manifested clinically as the unexpected panic attack, and
a cognitive component. He referred to the latter as
anticipatory anxiety and postulated that it leads to
agoraphobic avoidance. Other investigators of
agoraphobia (Marks, 1969; Wolpe and Rowan, 1988)

40
41
42
43
44
45
46

0165-0327 / 01 / $ see front matter 2001 Published by Elsevier Science B.V.

PII: S0165-0327( 01 )00457-8

2397

DTD431 29 2 NOV2 2001 10:41:02.17

v1.4.99

JAD2397

ARTICLE IN PRESS
E.H. Uhlenhuth et al. / Journal of Affective Disorders 1 (2001) 000 000
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130

2. Method

131

2.1. Subjects

132

The data for this study were drawn from candidates for pharmacotherapeutic studies, two on panic
disorder with or without agoraphobia (n560 and
n531), one on social phobia (n567), and one on
GAD (n543). All patients were diagnosed using the

133
134
135
136
137

der reflects predominantly the cognitive component,

and (c) panic disorder with agoraphobia reflects the
presence of both the biological and cognitive components (Uhlenhuth et al., 1999). Furthermore, we
suggested that different combinations of the biological and cognitive components might be pathogenic for these three disorders. Whether the same
general anxiety-prone cognitive style also characterizes other anxiety disorders is an interesting question
on which we had no hypothesis. However, our
expectation was that such a cognitive style to
different degrees is common to all anxiety disorders.
Anxiety-associated, threat-related cognitive contents have been encountered frequently in patients
with depression as well (Butler and Mathews, 1983;
Greenberg and Beck, 1989; Jolly, 1994). ASI scores
were elevated in patients with depression (Otto et al.,
1995; Taylor et al., 1996; Schmidt et al., 1997).
Patients with a major depressive episode had high
scores on the Penn State Worry Questionnaire (Starcevic, 1995). In contrast to the depression-specific
hopeless pessimism, no anxiety-related appraisal
characteristics have been found that typically discriminate anxiety from depression (Riskind, 1997).
The purpose of this study was to test the hypotheses set out above by comparing levels of a general
anxiety-prone cognitive style, as measured by the
AT&T, in three anxiety disorders: panic disorder
without agoraphobia (PD), panic disorder with
agoraphobia (PDA), and generalized anxiety disorder
(GAD). In addition, we had data available to examine AT&T scores in limited social phobia (LSP),
anxiety primarily about a specific performance, and
generalized social phobia (GSP). We also evaluated
the effects of comorbid depressive and anxiety
disorders.

emphasized conditioning to the context in which the

aversive stimulus, the panic attack, occurred. Cognitive theorists attempted to associate specific cognitive distortions with certain anxiety disorders
(Chambless et al., 1984; Reiss et al., 1986; Meyer et
al., 1990; Watson and Friend, 1969).
Although the evidence for an association between
these measures of cognitive distortions and the
corresponding anxiety disorders is persuasive, there
has been little research on the purported specificity of
each measured style for the corresponding disorder.
Existing research suggests that such specificity may
not exist. For example, Taylor et al. (1992) found
that the Anxiety Sensitivity Index (ASI) did not
differ significantly in patients with panic disorder
and post-traumatic stress disorder. In the same study,
the ASI also was elevated significantly in other
anxiety disorders, except specific phobia, when
compared to normal controls. Likewise, the Fear of
Negative Evaluation scale did not differentiate between patients with social phobia and other anxiety
disorders, except specific phobia (Oei et al., 1991;
Turner et al., 1987).
Blackburn and Davidson (1994), following Beck
and Emery (1985), proposed that anxious patients
have a general cognitive style characterized by
excessive perception of threat and appraisal of the
future as unpredictable. Our group suggested that
panic disorder arises from interaction between such a
general anxiety-prone cognitive style (Ganellan et
al., 1986; Uhlenhuth et al., 1997) and a biological
component, clinically identifiable as unexpected
panic attacks (Klein and Fink, 1962).
To measure this cognitive style, we developed the
Anxious Thoughts and Tendencies (AT&T) scale
(Ganellan et al., 1986; Uhlenhuth et al., 1999) based
conceptually on three types of cognitive distortion
described in anxious patients: catastrophizing potential outcomes of situations, selectively focusing
on negative aspects of situations, and intrusive
negative thoughts (Beck, 1976; Beck and Emery,
1985). These three conceptual dimensions proved to
be empirically indistinguishable (Uhlenhuth et al.,
1999). The AT&T is available in the Appendix to
Uhlenhuth et al. (1999).
We hypothesized specifically that (a) panic disorder without agoraphobia reflects predominantly the
biological component, (b) generalized anxiety disor-

47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94

PR

EC

TE
D

138

JAD2397

ARTICLE IN PRESS
E.H. Uhlenhuth et al. / Journal of Affective Disorders 1 (2001) 000 000

159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176

A factor analysis showed that the AT&T is a

homogeneous questionnaire of 15 items, each rated
on a four-point scale (rarely51; sometimes52;
often53; most of the time54). It has high internal
reliability (a 50.91) and acceptable testretest reliability (r50.75, before and after treatment) and
appears to be free of gender, age, and ethnic
(Hispanic versus non-Hispanic) effects. The development of the current version of the AT&T, the
rationale for doing so, and the questionnaire itself is
described in detail elsewhere (Uhlenhuth et al.,
1999).
Patients completed the AT&T during the screening
period of the therapeutic study for which they had
been recruited. The summed ratings for the individual items were divided by the total number of
items completed. The theoretical and actual range of
scores was from 1 to 4.

177

2.3. Data analyses

178
179
180
181
182

At their initial evaluation, panic patients rated as

none or mild on agoraphobia were grouped under the
diagnosis of PD. Those rated moderate or severe
were grouped under the diagnosis of PDA, as in our
first publication on the AT&T (Ganellan et al.,

3. Results

219

3.1. Demographic information

220

There were differences in mean age among the

diagnostic groups (P,0.001): PD, 38.55610.94
(S.D.); PDA, 34.3468.51; GAD, 43.93610.07; LSP,
46.32610.68; GSP, 38.4267.62; community sample,
48.36617.82. Women predominated in all diagnostic
groups except LSP and GSP (P,0.001): PD, 60%;

221
222
223
224
225
226

2.2. AT& T questionnaire

183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218

158

1986). Similarity of symptom and disability levels

supports collapsing these categories (Starcevic et al.,
1992).
Patients with social phobia also were divided into
those with a specific performance phobia or one or
two social phobic situations (limited social phobia,
LSP) or those with generalized social phobia (GSP).
Therefore, five diagnostic groups of anxious patients
were available for comparison: PD (n562), PDA
(n529), GAD (n543), LSP (n534), and GSP (n5
33).
For analysis, patients also were classified in
hierarchical fashion into one of the following affective disorder categories: current major depressive
disorder (MDD, n518), current dysthymia (Dysth,
n514), past major depressive disorder (Hx MDD,
n559), or no depression (No Depr, n5110).
The a priori hypotheses were tested by comparing
the mean AT&T scores for patients with PD, PDA,
and GAD. Other questions of interest were explored
by comparing mean AT&T scores among diagnostic
groups using analysis of variance (ANOVA) with one
or two factors as required, followed by post-hoc
pairwise comparisons between diagnostic groups
based on Fishers LSD, without correction for multiple comparisons (Dunn, 1961). Effect sizes were
computed using the standard deviation for the entire
sample (n5519; S.D.50.65), including the community subjects, as a best estimate, but are not
presented to save space. Since data distributions
were not always normal, all analyses were repeated
using non-parametric (ranked scores) methods. Results of these analyses are not reported, as they were
essentially identical to the results using raw scores.
All analyses were performed with the NCSS statistical package (Hintze, 1998).

full outpatient version of the Structured Clinical

Interview for DSM-III-R / DSM IV (SCID; First et
al., 1997; Spitzer et al., 1990). Patients were included in the present study if (1) their principal
diagnosis met DSM-III-R or DSM-IV criteria for the
respective therapeutic study, whether or not the other
inclusion and exclusion criteria were met, and (2)
they completed the AT&T at the time of initial
evaluation. The principal diagnosis was defined as
the one that accounted for most of the distress and
dysfunction at the time of evaluation.
A community sample (n5318) previously studied
with the AT&T (Uhlenhuth et al., 1999) also was
included in the present study to provide an additional
check on the validity of the AT&T and an estimate
of the effect of each diagnosis on the AT&T.
Psychiatric diagnosis was not performed with these
subjects.
All subjects provided written informed consent.

EC

TE
D

139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157

PR

227

JAD2397

ARTICLE IN PRESS
280

E.H. Uhlenhuth et al. / Journal of Affective Disorders 1 (2001) 000 000

266
267
268
269
270
271
272
273

PDA, 62%; GAD, 84%; LSP, 32%, GSP, 36%;

community sample, 65%. However, our previous
work with the community sample demonstrated no
age or gender effects on the AT&T (Uhlenhuth et al.,
1999). The rates of comorbidity for other anxiety
disorders in the principal diagnostic groups were
(P,0.001): PD, 50%; PDA, 62%; GAD, 47%; LSP,
15%; GSP, 33%. The rates of comorbidity for a

228
229
230
231
232

Table 1
Mean 15-item total AT&T scores by disorder
A. All patients grouped by anxiety disorder (n5201)
PD
PDA

GAD

LSP

GSP

Commun

233
234
235
236

n
Mean
S.D.
S.E.M.

43
2.42 b
0.64
0.10

34
1.90 a,c
0.73
0.13

33
2.35 b
0.61
0.11

319
1.71 c
0.50
0.03

237
238

B. All patients grouped by comorbid depression disorders (n5201)

MDD
Dysth
Hx MDD

No Depr

239
240
241
242

n
Mean
S.D.
S.E.M.

110
2.03
0.69
0.07

243
244

C. Patients with comorbid depressive disorders (n591)

PD
PDA

245
246
247
248

n
Mean
S.D.
S.E.M.

249
250

18
2.71 a
0.71
0.17

14
2.55 a
0.59
0.16

59
2.52 a
0.59
0.08

Section A of Table 1 shows the descriptive

statistics for the AT&T by anxiety disorder and the
community sample. Pairwise t-tests indicate that the

277
278
279

276

319
1.71
0.50
0.03

GSP
18
2.55 a,b
0.52
0.12

D. Patients without comorbid depressive disorders (n5110)

PD
PDA
GAD

LSP

GSP

251
252
253
254

n
Mean
S.D.
S.E.M.

27
1.74 d,c
0.55
0.11

15
2.17 a,b
0.67
0.17

319
1.71 c
0.50
0.03

255
256
257
258
259
260
261
262
263
264
265

E. Patients without comorbid depressive or anxiety disorders (n578)

n
21
7
13
Mean
1.68 a,c
2.49 b
2.02 b,c,d
S.D.
0.45
0.67
0.69
S.E.M.
0.10
0.25
0.19

25
1.72 a,d
0.57
0.11

12
2.01 b,c
0.62
0.18

319
1.71 a
0.50
0.03

10
2.50 b
0.59
0.19

39
1.93 a,d
0.66
0.11

EC

19
2.82 b
0.56
0.13

TE
D

LSP

7
2.52 a,b
1.01
0.38

23
2.41 a
0.72
0.15

GAD

3.2. Tests of a priori hypotheses

29
2.71
0.58
0.11

274
275

PR

62
2.11 a
0.71
0.09

depressive disorder were (P,0.005): PD, 37%;

PDA, 66%; GAD, 56%; LSP, 21%; GSP, 55%.

24
2.53 a,b
0.44
0.09

19
2.29 b
0.82
0.19

319
1.71
0.50
0.03

Legend: PD, Panic disorder without agoraphobia; PDA, panic disorder with agoraphobia; GAD, generalized anxiety disorder; LSP, limited
social phobia, GSP, generalized social phobia; Commun, community sample.
Notes: AT&T pooled standard deviation50.65, n5520. Effect sizes can be computed by dividing the difference between two means by
0.65. ANOVAs comparing diagnostic groups are significant at P,0.001 in all five sections of the table. Means with the same superscript do
not differ significantly, P.0.05.

JAD2397

ARTICLE IN PRESS
E.H. Uhlenhuth et al. / Journal of Affective Disorders 1 (2001) 000 000

289
290
291
292
293
294
295
296
297
298
299
300

Mean AT&T scores differed significantly among

the five anxiety disorders in the full patient sample
and the community sample (df55 / 513; F 532.24;
P,0.001), as shown in section A of Table 1. AT&T
scores were higher in all groups of patients with
anxiety disorders, except LSP, than in the community
sample. PDA patients had the highest scores. GAD
and GSP patients had similar scores lying at an
intermediate level. PD and LSP patients had the
lowest scores. AT&T scores among LSP patients
only slightly exceeded those in the community
sample.

301
302

3.4. AT& T scores by comorbid depressive

diagnosis in the full sample

303
304
305
306
307
308
309
310
311
312
313
314
315

Mean AT&T scores differed significantly among

patients grouped by comorbid depressive disorder in
the full patient sample and the community sample
(df54 / 514; F 541.68; P,0.001), as shown in
section B of Table 1. AT&T scores were higher in
all three groups of patients with comorbid depressive
disordersand in anxious patients without comorbid
depressive disorderthan in the community sample.
Anxious patients with current or past comorbid
depressive disorders had higher AT&T scores than
anxious patients without comorbid depressive disorders. AT&T scores were similar among patients
with current and past comorbid depressive disorders.

316
317
318

3.5. AT& T scores among anxiety disorders in

patients with and without comorbid depressive
disorders

319
320
321

Mean AT&T scores differed significantly among

anxiety disorders in patients with diagnosed comorbid depression and the community sample (df55 /

3.3. AT& T scores among anxiety disorders in the

full sample

287
288

403; F 538.63; P,0.001) (Table 1, section C) and

also among anxiety disorders in patients free of
comorbid depression and the community sample
(df55 / 422; F 510.03; P,0.001) (Table 1, section
D). Two-way ANOVA showed significant main
effects for anxiety disorder (df54 / 191; F 53.42;
P,0.01) and for comorbid depressive disorder (df5
1 / 191; F 519.16; P,0.001), but no interaction
(df54 / 191; F 50.74; P.0.25). However, the power
for detecting interaction was only 0.23. AT&T scores
were higher in depressed than in non-depressed
patients with anxiety disorders. Relative AT&T
levels among anxiety disorders in depressed and
especially non-depressed patients were similar to
those in the full sample. AT&T scores in nondepressed PD and LSP patients were not substantially elevated above the community sample.

339
340
341

Finally, mean AT&T scores differed significantly

among anxiety disorders in patients free of any
comorbid anxiety or depressive diagnosis and the
community sample (df55 / 390; F 54.66; P,0.001),
as shown in section F of Table 1. PDA patients had
the highest scores. GAD and GSP patients had
similar scores lying at an intermediate level. PD and
LSP patients had the lowest scores, which were
similar to those in the community sample.

342
343
344
345
346
347
348
349
350

4. Discussion

351

The results of this investigation are partly consistent with our hypotheses concerning associations
between a general anxiety-prone cognitive style and
PD, PDA, and GAD. Low AT&T levels and panic
attacks characterized PD, high AT&T levels and
panic attacks characterized PDA, and high AT&T
levels in the absence of panic attacks characterized
GAD. However, AT&T levels were not as high in
GAD as in PDA, contrary to expectation. These
findings held when patients with comorbid depressive and anxiety disorders were excluded.

352
353
354
355
356
357
358
359
360
361
362

EC

322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338

3.6. AT& T scores among anxiety disorders in

patients free of comorbid depressive and anxiety
disorders

PDA group (t54.00; P,0.001) and the GAD group

(t52.34; P,0.05) had significantly higher scores
than the PD group, consistent with our a priori
hypotheses. The difference between the PDA and the
GAD groups was not quite significant (t51.96; P,
0.10).

TE
D

281
282
283
284
285
286

PR

363

JAD2397

ARTICLE IN PRESS
E.H. Uhlenhuth et al. / Journal of Affective Disorders 1 (2001) 000 000

These results replicate the finding by Ganellan et

al. (1986) that PDA patients had significantly higher
AT&T scores than PD patients. The finding of high
AT&T levels in patients with PDA adds to the large
body of evidence that certain cognitive styles are
more highly associated with PDA than with PD
(Craske et al., 1988; Telch et al., 1989; Barlow et al.,
1994; Fleming and Faulk, 1989; Noyes et al., 1987).
Our results are consistent with the view that panic
patients with a marked general tendency to catastrophize may be more likely to develop
agoraphobia (Clum and Knowles, 1991) in the
context of unexpected panic attacks. The finding of
high AT&T levels in patients with GAD also is
consistent with recent formulations of GAD as a
condition characterized primarily by abnormal cognitive processes (Borkovec and Inz, 1990; Woody and
Rachman, 1994; Freeston et al., 1994; Davey, 1994;
Dugas et al., 1998a,b).

385
386

4.2. General anxiety-prone cognitive style in PD

and social phobia

387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408

We were surprised to find that patients with PD

and LSP had AT&T levels equivalent to levels in the
community sample when patients with comorbid
conditions were excluded. These results suggest that
comorbid conditions, especially depression, account
for cognitive distortions in PD and LSP, i.e., that a
general anxiety-prone cognitive style is not intrinsic
to these disorders. Alternatively, there may be cognitive styles specific to PD and LSP and independent
of the general style measured by the AT&T. This
possibility seems unlikely in view of the existing
evidence cited in the introduction against the specificity of cognitive measures conceptually related to
particular disorders.
PD and LSP may arise primarily from biological
predispositions. Barlow et al. (1994) noted in their
review that catastrophic cognitions are not essential
to conceptualize panic attacks. Goddard and Charney
(1997) proposed autonomic hyper-arousal or instability as the essential (biological) basis for panic
attacks.
Kagan et al. (1988) and Rosenbaum et al. (1994)

366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384

identified an inborn trait of behavioral inhibition to

the unfamiliar as a frequent precursor of social
phobia. Kendler et al. (1992) estimated heritability in
social phobia (not separated into LSP and GSP) at
30%. Moreover, there appear to be several distinct
types of genetic and temperamental vulnerabilities to
social phobia (Hirshfeld-Becker et al., 1999).
These considerations invite a formulation for
social phobia parallel to the one proposed earlier for
panic and agoraphobia; i.e., LSP reflects only a
biological component, whereas GSP reflects both a
biological component and a cognitive component.
Moreover, we suggest that unique biological components rather than different cognitive styles may
distinguish panic disorder with agoraphobia from
generalized social phobia.

4.3. General anxiety-prone cognitive style in

depressive disorders and relationship between
anxiety and depression

EC

409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424

425
426
427

4.1. General anxiety-prone cognitive style in PDA

and GAD

364
365

The findings on comorbid depressive disorders

indicate that the cognitive style measured by the
AT&T is not specific to anxiety disorders, but is also
a prominent feature of depressive disorders. Correlations between the AT&T and the SCL-90 factor
scores (Lipman et al., 1979) for anxiety (n5318;
r50.53) (Uhlenhuth et al., 1999) and depression
(n5318; r50.57) (Uhlenhuth, unpublished data) in
the community sample further support this inference.
It is appealing to try to understand the contrast
between anxiety and depression through anxious
cognitions. However, they more likely represent a
non-specific distress factor called negative affectivity by Watson and Clark (1984).
An interesting sidelight is the similarity of AT&T
levels among patients with current MDD, current
dysthymia, and past MDD. This finding is consistent
with the possibility that a general anxiety-prone
cognitive style at least in part expresses a trait
predisposing to the development of depressive symptoms, rather than a state that resolves with successful
treatment. Although the correlations of the AT&T
with anxiety and depressive symptoms (states) are
substantial, they are not so high as to rule out a trait
component in the AT&T. However, in the majority
of related studies, as reviewed by Segal and Ingram
(1994), individuals who recovered from MDD re-

PR

TE
D

455

428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454

JAD2397

ARTICLE IN PRESS
E.H. Uhlenhuth et al. / Journal of Affective Disorders 1 (2001) 000 000

475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501

Problems with the results reported above include

the fact that the potential effects of sub-syndromal
depression cannot be excluded. The primary studies
from which the data were drawn did not include
sub-syndromal diagnoses. Neither are data from a
dimensional measure of depression available in most
of the primary studies at this time. Therefore, one
cannot confidently assert that the elevated AT&T
levels among PDA, GAD, and GSP patients after
exclusion of patients with diagnosed comorbid depressive disorders are intrinsic to these anxiety
disorders, i.e., that they are not related to comorbid
sub-syndromal depression. In addition, evaluating the
relationship of the AT&T to depression is limited by
the absence of data on AT&T levels in patients with
depressive disorders uncomplicated by comorbid
anxiety.
Throughout this paper, we have implied a pathogenic role for a general anxiety-prone cognitive style
in some but not other anxiety disorders. While our
findings are consistent with those hypotheses, they
demonstrate only associations between high AT&T
scores and PDA, GAD, and GSP. Cross-sectional
studies like the present one inherently are limited to
generating causal hypotheses and perhaps excluding
them. Definitive support for a pathogenic role for
cognitive distortions in anxiety or depression re-

Acknowledgements

544

The studies from which the data for these analyses

were drawn were supported in part by Hoffmann-La

545
546

4.4. Limitations of the study and prospects for the

future

502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543

473
474

quires measurement of those cognitions before the

onset of the disorder. Unfortunately, research on
cognitive styles up to the present has not produced
studies that satisfy this requirement. The same
reservation holds for many investigations of potential
biological factors in pathogenesis. Follow-up of
high-risk subjects, such as children of parents with
PDA (Biederman et al., 1991), is a feasible means to
obtain the relevant data. It is critical to address these
issues more effectively in future research.
Meanwhile an alternative interpretation of the
association between the AT&T and certain anxiety
disorders cannot be ruled out. These distorted cognitions may be conceptualized as cognitive symptoms
of the anxiety and depressive disorders in which they
are found. Additional data on the response of the
AT&T to effective treatment may help to clarify this
possibility. From the perspective of cognitions as
symptoms, it is also noteworthy that the anxiety
disorders with high AT&T levels are more severe by
definition than the disorders with low AT&T levels.
Comorbidity, which is associated with elevations of
AT&T scores, similarly implies more severe disorder. In addition, AT&T scores in the community
sample correlated at moderate levels with symptom
dimensions of somatization (r50.36), anxiety (r5
0.53), and phobia (r50.44) (Uhlenhuth et al., 1999),
as well as depression (r50.57) (Uhlenhuth, unpublished data) on the SCL-90.
From these data, one could argue that the AT&T
simply measures the severity of anxious and depressive conditions, rather than defining qualitatively
distinct types of disorder. This line of reasoning
opens the larger and controversial issue of typological versus dimensional conceptualization of psychiatric disorder, which is beyond the scope of this paper.
For present purposes we underscore the fact that the
patient groups included in this study conform to
currently accepted criteria for their respective principal diagnoses. Finally, the data presented here,
irrespective of conceptual interpretation, support the
validity of the AT&T.

sembled non-depressed comparison subjects with

respect to depression-related cognitions, suggesting
that such cognitions are not stable traits that might
indicate vulnerability to depression. Some investigators have hypothesized latent cognitive structures in vulnerable persons, which can be activated
under certain conditions to contribute to a depressive
episode (Segal and Ingram, 1994). If so, the cognitive abnormalities measured by the AT&T may be a
marker of such a trait-like vulnerability to depression. A recent monograph devoted to generalized
anxiety (Akiskal et al., 1998) and a paper more
specifically dealing with the cognitive traits of
generalized anxiety from an emotional perspective
(Akiskal, 1998) dealt in greater detail with the
relationship between such a putative trait and depression.

EC

TE
D

456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472

PR

547

JAD2397

ARTICLE IN PRESS
E.H. Uhlenhuth et al. / Journal of Affective Disorders 1 (2001) 000 000

References

558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597

Akiskal, H.S., 1998. Toward a definition of generalized anxiety

disorder as an anxious temperament type. Acta Psychiatr.
Scand. 98 (Suppl. 393), 6673.
Akiskal, H.S., Hantouche, E., Judd, L. (Eds.), 1998. Generalized
Anxiety Disorder: From Science to Art. Acta Psychiatrica
Scandinavica, Vol. 97, Suppl. 343.
Barlow, D.H., Brown, T.A., Craske, M.G., 1994. Definitions of
panic attacks and panic disorder in the DSM-IV: implications
for research. J. Abnorm. Psychol. 103, 553564.
Beck, A.T., 1976. Cognitive Therapy and the Emotional Disorders. International Universities Press, New York.
Beck, A.T., Emery, G., 1985. Anxiety Disorders and Phobias: A
Cognitive Perspective. Basic Books, New York.
Biederman, J., Rosenbaum, J.F., Bolduc, E.A., Faraone, S.V.,
Hirshfeld, D.R., 1991. A high-risk study of young children of
parents with panic disorder and agoraphobia with and without
comorbid major depression. Psychiatry Res. 37, 333334.
Blackburn, I.-M., Davidson, K.M., 1994. In: Cognitive Therapy
for Depression and Anxiety: A Practitioners Guide. Blackwell
Science, Oxford, p. 24.
Borkovec, T.D., Inz, J., 1990. The nature of worry in generalized
anxiety disorder: a predominance of thought activity. Behav.
Res. Ther. 28, 153158.
Butler, G., Mathews, A., 1983. Cognitive processes in anxiety.
Adv. Behav. Res. Ther. 5, 5162.
Chambless, D.L., Caputo, G.C., Bright, P., Gallagher, R., 1984.
Assessment of fear in phobics: the body sensations questionnaire and the agoraphobic cognitions questionnaire. J.
Consult. Clin. Psychol. 52, 10901097.
Clum, G.A., Knowles, S.L., 1991. Why do some people with panic
disorders become avoidant? A review. Clin. Psychol. Rev. 11,
295313.
Craske, M.G., Rapee, R.M., Barlow, D.H., 1988. The significance
of panic-expectancy for individual patterns of avoidance.
Behav. Ther. 19, 577592.
Davey, G.C.L., 1994. Worrying, social problem-solving abilities,
and problem-solving confidence. Behav. Res. Ther. 32, 327
330.
Dugas, M.J., Freeston, M.H., Ladouceur, R., Rheaume, J., Provencher, M., Boisvert, J.-M., 1998a. Worry themes in primary

557

GAD, secondary GAD, and other anxiety disorders. J. Anxiety

Disord. 12, 253261.
Dugas, M.J., Gagnon, F., Ladouceur, R., Freeston, M.H., 1998b.
Generalized anxiety disorder: a preliminary test of a conceptual
model. Behav. Res. Ther. 36, 215226.
Dunn, O.J., 1961. Multiple comparisons among means. J. Am.
Stat. Assoc. 56, 5264.
First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 1997.
Structured Clinical Interview for DSM-IV Axis I Disorders
(SCID-I)Clinician Version. American Psychiatric Press,
Washington, DC.
Fleming, B., Faulk, A., 1989. Discriminating factors in panic
disorder with and without agoraphobia. J. Anxiety Disord. 3,
209219.
Freeston, M.H., Rheaume, J., Letarte, H., Dugas, M.J., Ladouceur,
R., 1994. Why do people worry? Personal. Individ. Diff. 17,
791802.
Ganellan, R.J., Matuzas, W., Uhlenhuth, E.H., Glass, R., Easton,
C.R., 1986. Panic disorder, agoraphobia, and anxiety-relevant
cognitive style. J. Affect. Disord. 11, 219225.
Goddard, A.W., Charney, D.S., 1997. Toward an integrated
neurobiology of panic disorder. J. Clin. Psychiatry 58 (Suppl.
2), 411.
Greenberg, M.S., Beck, A.T., 1989. Depression versus anxiety: a
test of the content-specificity hypothesis. J. Abnorm. Psychol.
98, 913.
Hintze, J.L., 1998. NCSS 2000: Statistical System for Windows.
Number Cruncher Statistical Systems, Kaysville, UT.
Hirshfeld-Becker, D.R., Fredman, S.J., Robin, J.A., Rosenbaum,
J.F., 1999. The aetiology of social anxiety disorder. In:
Westenberg, H.G.M., den Boer, J.A. (Eds.), Social Anxiety
Disorder. Syn-Thesis Publishers, Amsterdam.
Jolly, J.B., 1994. A multi-method test of the cognitive contentspecificity hypothesis in young adolescents. J. Anxiety Disord.
7, 223233.
Kagan, J., Reznick, J.S., Snidman, N., 1988. Biological bases of
childhood shyness. Science 240, 167171.
Kendler, K., Neale, M., Kessler, R., Heath, A., Eaves, L., 1992.
The genetic epidemiology of phobias in women: the interrelationship of agoraphobia, social phobia, situational phobia and
simple phobia. Arch. Gen. Psychiatry 49, 273281.
Klein, D.F., 1980. Anxiety reconceptualized. Comp. Psychiatry
21, 411427.
Klein, D.F., Fink, M., 1962. Psychiatric reaction patterns to
imipramine. Am. J. Psychiatry 119, 432438.
Lipman, R.S., Covi, L., Shapiro, A.K., 1979. The Hopkins
Symptom Checklist (HSCL): factors derived from the HSCL90. J. Affect. Disord. 1, 924.
Marks, I.M., 1969. Fears and Phobias. William Heinemann
Medical Books, London.
Meyer, T.J., Miller, M.L., Metzger, R.L., Borkovec, T.D., 1990.
Development and validation of the Penn State Worry Questionnaire. Behav. Res. Ther. 28, 487495.
Noyes, R., Clancy, J., Garvey, M.J., Anderson, D.J., 1987. Is
agoraphobia a variant of panic disorder or a separate illness? J.
Anxiety Disord. 1, 313.
Oei, T.P.S., Kenna, D., Evans, L., 1991. The reliability, validity

Roche, Sandoz Pharmaceutical Corporation (now

Novartis Pharmaceutical Corporation), The Upjohn
Company (now Pharmacia & Upjohn), The University of Chicago, and BRSG S07 RR-05583-22,
awarded to the University of New Mexico by the
Biomedical Research Support Grant Program, Division of Research Resources, National Institutes of
Health. Data analyses for this report were supported
by the University of New Mexico.

548
549
550
551
552
553
554
555
556

PR

EC

TE
D

655

598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654

JAD2397

ARTICLE IN PRESS
E.H. Uhlenhuth et al. / Journal of Affective Disorders 1 (2001) 000 000

Taylor, S., Koch, W.J., McNally, R.J., 1992. How does anxiety
sensitivity vary across the anxiety disorders? J. Anxiety Disord.
7, 249259.
Taylor, S., Koch, W.J., Woody, S., McLean, P., 1996. Anxiety
sensitivity and depression: how are they related? J. Abnorm.
Psychol. 105, 474479.
Telch, M.J., Broulliard, M., Telch, C.F., Agras, W.S., Taylor, C.B.,
1989. Role of cognitive appraisal in panic-related avoidance.
Behav. Res. Ther. 27, 373383.
Turner, S.M., McCanna, M., Beidel, D.C., 1987. Validity of the
Social Avoidance and Distress and Fear of Negative Evaluation
scales. Behav. Res. Ther. 25, 113115.
Uhlenhuth, E.H., Matuzas, W., Warner, T.D., Thompson, P.M.,
1997. Growing placebo response rate: the problem in recent
therapeutic trials? Psychopharmacol. Bull. 33, 3139.
Uhlenhuth, E.H., McCarty, T., Paine, S., Warner, T., 1999. The
revised Anxious Thoughts and Tendencies (AT&T) scale: a
general measure of anxiety-prone cognitive style. J. Affect.
Disord. 52, 5158.
Watson, D., Clark, L.A., 1984. Negative affectivity: the disposition to experience aversive emotional states. Psychol. Bull. 96,
465490.
Watson, D., Friend, R., 1969. Measurement of social evaluative
anxiety. J. Consult. Clin. Psychol. 33, 448457.
Wolpe, J., Rowan, V.C., 1988. Panic disorder: a product of
classical conditioning. Behav. Res. Ther. 26, 441450.
Woody, S., Rachman, S., 1994. Generalized anxiety disorder
(GAD) as an unsuccessful search for safety. Clin. Psychol.
Rev. 14, 743753.

and utility of the SAD and the FNE scales for anxiety disorder
patients. Personal. Individ. Diff. 12, 111116.
Otto, M.W., Pollack, M.H., Fava, M., Uccello, R., Rosenbaum,
J.F., 1995. Elevated Anxiety Sensitivity Index scores in
patients with major depression: correlates and changes with
antidepressant treatment. J. Anxiety Disord. 9, 117123.
Reiss, S., Peterson, R.A., Gursky, D.M., McNally, R.J., 1986.
Anxiety sensitivity, anxiety frequency and the prediction of
fearfulness. Behav. Res. Ther. 24, 18.
Riskind, J.H., 1997. Looming vulnerability to threat: a cognitive
paradigm for anxiety. Behav. Res. Ther. 35, 685702.
Rosenbaum, J.F., Biederman, J., Pollock, R.A., Hirshfeld, D.R.,
1994. The etiology of social phobia. J. Clin. Psychiatry 55
(Suppl. 6), 1016.
Schmidt, N.B., Lerew, D.R., Jackson, R.J., 1997. The role of
anxiety sensitivity in the pathogenesis of panic: prospective
evaluation of spontaneous panic attacks during acute stress. J.
Abnorm. Psychol. 106, 355364.
Segal, Z.V., Ingram, R.E., 1994. Mood priming and construct
activation in tests of cognitive vulnerability to unipolar depression. Clin. Psychol. Rev. 14, 663695.
Spitzer, R.L., Williams, J.B.W., Gibbon, M., First, M.B., 1990.
Structured Clinical Interview for DSM-III-R. American Psychiatric Press, Washington, DC.
Starcevic, V., 1995. Pathological worry in major depression: a
preliminary report. Behav. Res. Ther. 33, 5556.
Starcevic, V., Uhlenhuth, E.H., Kellner, R., Pathak, D., 1992.
Patterns of comorbidity in panic disorder and agoraphobia.
Psychiatry Res. 42, 171183.

9
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713

EC

TE
D

656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684

PR

714

JAD2397