Clinical features and treatment response in social phobia: axis II ...

Psicothema
ISSN 0214 - 9915 CODEN PSOTEG
2002. Vol. 14, nº 2, pp. 426-433
Copyright © 2002 Psicothema
Clinical features and treatment response in social phobia:
axis II comorbidity and social phobia subtypes
M. A. Ruipérez, A. García-Palacios y C. Botella
Universitat Jaume I
The aim of the present study was twofold. First, we studied the differential demographic and clinical
features regarding social phobia subtype and axis II comorbidity. Second, we studied the role of social
phobia subtype and axis II comorbidity in treatment effectiveness. The sample included 28 patients
diagnosed of social phobia (DSM-IV, APA, 1994). We divided the sample attending to social phobia
subtype and axis II comorbidity. When we compared the groups, we did not find significant differen-
ces in demographic variables. However, we found differences regarding clinical variables: Patients
with generalized social phobia were more impaired than patients with circumscribed social phobia; pa-
tients with axis II comorbidity were also more impaired than patients without axis II comorbidity. All
patients were treated with a group cognitive-behavioral program, adapted from Heimberg, Juster, Ho-
pe & Mattia (1995). There were no differences in effectiveness regarding the different sub-samples.
However, some of the clinical differences found at the pre-treatment regarding some clinical variables
disappeared after the termination of the treatment.
Características clínicas y respuesta al tratamiento en fobia social: Comorbilidad en el eje II y subtipos
de fobia social. El presente estudio tiene un objetivo doble. En primer lugar, estudiamos las diferencias
demográficas y clínicas en función del subtipo de fobia social y de la comorbilidad que presentaban una
muestra de pacientes diagnosticados de fobia social. En segundo lugar, analizamos el efecto que el sub-
tipo de fobia social y la comorbilidad en eje II tenían en la mejoría terapéutica. La muestra se compuso
de 28 pacientes diagnosticados de fobia social (DSM-IV, APA, 1994). Dividimos la muestra en función
del subtipo de fobia social y la comorbilidad en eje II. Al comparar los grupos, no encontramos dife-
rencias significativas en características demográficas. Sin embargo, sí encontramos diferencias en ca-
racterísticas clínicas: Los pacientes con fobia social generalizada presentaban un mayor deterioro que
los pacientes con fobia social circunscrita; asimismo, los pacientes con comorbilidad en eje II también
presentaban un mayor grado de deterioro. Todos los pacientes recibieron un tratamiento cognitivo-com-
portamental en grupo adaptado de Heimberg, Juster, Hope & Mattia (1995). No encontramos diferen-
cias en la eficacia del tratamiento en función del subtipo de fobia social y la comorbilidad en eje II. Sin
embargo, algunas de las diferencias en características clínicas que se encontraron en el pretratamiento,
no aparecieron en la evaluación realizada tras la finalización del tratamiento.
The essential feature of Social Phobia is a persistent and inten-
This variability has opened up a discussion on the necessity of
se fear of one or more social situations in which the individual is
distinguishing among different social phobia subtypes. As a back-
exposed to the observation of others. The individual fears that
ground, we can take the Öst, Jerremalm & Johansson (1981)
he/she will behave in a way that will be humiliating. Exposure to
study. This authors, using the performance in a role-play, distin-
those stimuli provokes an immediate anxiety response that can le-
guished between «behavioral reactors» (behavioral disruption
ad to the avoidance of those situations or to the endurance of them
without cardiac acceleration) and «physiological reactors» (car-
with intense anxiety (DSM-IV, APA, 1994). The clinical features
diac acceleration without behavioral disruption). Later, they con-
of this diagnostic category include a wide number of situations,
sidered another subtype, «cognitive reactors» (predominance of
from specific fears such as eating, writing or speaking in public, to
irrational thoughts) (Jerrelmalm, Jansson & Öst, 1986). DSM-III-
more generalized fears which appear in all or almost all social si-
R (APA, 1987) specifies the generalized social phobia subtype
tuations (Heimberg, Holt, Scheneier, Spitzer & Liebowitz, 1993).
when the individual fears «most» social situations. This subtype
is maintained by DSM-IV (APA, 1994). Heimberg distinguished
three social phobia subtypes: generalized social phobia, which in-
Fecha recepción: 23-7-01 • Fecha aceptación: 2-1-02
cludes fear across almost all domains of social situations; non-ge-
Correspondencia: M. Ángeles Ruipérez
neralized social phobia, which includes individuals who fear mul-
Facultad de Psicología
tiple social situations, but who report no problems in at least one
Universitat Jaume I
social domain; and finally, discrete social phobia, which includes
12080 Castellón (Spain)
E-mail: ruiperez@psb.uji.es
individuals with fear in only one or two circumscribed social si-

---

CLINICAL FEATURES AND TREATMENT RESPONSE IN SOCIAL PHOBIA: AXIS II COMORBIDITY AND SOCIAL PHOBIA SUBTYPES
427
tuations (public speaking, eating in public (Heimberg, et al.,
Wolff, Spaulding & Jacob (1996) observed that there were no di-
1993).
fferences between generalized social phobics and circumscribed
In the last years, several studies have been carried out to determi-
social phobics in the response to an exposure treatment, although
ne the differences in clinical and demographic features and the diffe-
generalized social phobia showed more impairment than circums-
rential responses to treatment of different social phobia subtypes.
cribed social phobics at pre-test, and these differences in overall
In the area of psychopathological and demographic differences,
impairment also appeared at post-test. Finally, using the Heimberg
Turner, Beidel, Dancu & Keys (1986) found higher sensitivity to
subtypes, Hope, Herbert & White (1995) and Brown, Heimberg &
interpersonal relationship, depression, and worse performance in a
Juster (1995), found that patients with generalized social phobia
social skills test in generalized social phobics. Heimberg, Hope,
did not differ in their response to a group cognitive-behavioral tre-
Dodge & Becker (1990) noticed that generalized social phobics
atment. However, the same as occurred with Turner et al. (1996),
were younger, less educated, and less likely to be employed than
they noticed that generalized social phobics were more impaired
those with non-generalized social phobia. Also, generalized social
overall than non-generalized social phobics at pre-test, post-test
phobics were more impaired, and depression, anxiety and fear of
and follow-up assessments.
negative evaluation scores were higher. Turner, Beidel & Towns-
In summary, it seems that there are no relevant differences in the
ley (1992) found higher scores of distress and social anxiety in ge-
response of different social phobia subtypes to psychological treat-
neralized social phobia. Tran & Chambless (1995) compared ge-
ment, although patients with generalized social phobia present mo-
neralized social phobia with and without Avoidant Personality Di-
re impairment overall than patients with non-generalized or cir-
sorder (APD) and circumscribed social phobia. They did not find
cumscribed social phobia at pre-test assessments, and these diffe-
differences in age, socio-economic status, age at onset, and the du-
rences in impairment are maintained after completion of treatment.
ration of the disorder. However, generalized social phobics with
Another important issue in social phobia literature is the study
APD were more likely to be single or taking medication than cir-
of axis II comorbidity in social phobia. Along this research line,
cumscribed social phobics. Besides, generalized social phobics
we should mention the difficulty in distinguishing generalized so-
with APD presented a higher level of depression than the other
cial phobia and APD (Alden & Capreol, 1993). On the other hand,
two groups. Finally, in self-report measures, generalized social
regarding response to treatment, the findings are not conclusive
phobics (with or without APD) scored higher in the FNE (Fear of
(Heimberg, 1996). Many studies on this issue obtained no diffe-
Negative Evaluation Scale, Watson & Friend, 1969) and the SADS
rences in response to treatment between social phobics with APD
(Social Avoidance and Distress Scale, Watson & Friend, 1969),
and social phobics without APD (i.e., Brown et al., 1995; Van Vel-
and in measures of social impairment. Among the studies which
zen, Emmelkamp & Scholing, 1997). However, these studies also
used the Heimberg subtypes, we would like to mention the work
show a higher level of impairment in patients with APD at pre-test
by Herbert, Hope & Bellack (1992), where differences in age we-
which is maintained after treatment completion (Brown et al.,
re not found, but generalized social phobics were more likely to be
1995; Hoffman, Newman, Becker, Taylor & Roth, 1995; Hope et
male. These patients received a greater number of secondary axis
al., 1995; Turner et al., 1996). However, other studies (Turner,
II diagnosis, and clinicians judged them as being more impaired
1987; Lucas & Telch, 1993; Chambless, Tran & Glass, 1997;
overall, they showed worse social skills, and their performance in
Scholing & Emmelkamp, 1999) found that APD was associated
a social interaction behavioral test was worse. In a similar study,
with poorer outcome on several measures.
Holt, Heimberg & Hope (1992) noticed that generalized social
In this work, it is our intention to study, on one hand, the diffe-
phobics were more impaired overall, and presented extreme sco-
rences in demographic and psychopathological features, and, on
res on self-report measures of social anxiety and depression. Also,
the other hand, the differential response to a cognitive-behavioral
generalized social phobia had an earlier age of onset of the disor-
treatment in a sample of patients who meet social phobia criteria
der than non-generalized social phobia.
(DSM-IV, APA, 1994). Both objectives were studied comparing
From these studies we can conclude that generalized social
social phobia subtypes (generalized vs. circumscribed) and social
phobics, when they are compared with circumscribed or non-ge-
phobia with and without axis II comorbidity.
neralized social phobics, present more severity, more impairment
overall and higher levels of associated psychopathology, such as
Method
anxiety, depression and social skills.
With regard to the differential response to psychological treat-
Subjects
ment, several studies have been carried out. Using the Öst subty-
pes, contradictory results have been found. Öst et al., (1981) found
The sample was made up of 28 patients who attended the Uni-
that «behavioral reactors» obtained better outcomes in social skills
versity Jaume I Anxiety Disorders Clinic in response to an adver-
training, and «physiological reactors» had a better response in ap-
tisement that appeared in the local newspapers and on the radio.
plied relaxation training. However, other studies (Jerrelman et al.,
The patients were assessed using an admission interview that
1986; Mersch, Emmelkamp, Bögels & van der Sleen, 1989) did
screens the presence of anxiety disorders. Then the patients who
not find differential response to treatment when comparing these
seem to suffer from social phobia were interviewed using the An-
social phobia subtypes.
xiety Disorders Interview ADIS-R (Di Nardo & Barlow, 1988) to
The studies which compare the DSM-III-R and DSM-IV subty-
confirm the diagnosis. This instrument follows DSM-III-R (APA,
pes did not show conclusive results. On one hand, Heimberg
1987) criteria. We also evaluated each subject who took part in
(1986) found that social phobia subtype was the main predictor of
the study, following the DSM-IV (APA, 1994) criteria, and we
outcome in group cognitive-behavioral treatment. However, Holt,
found that diagnosis and severity showed no changes in any of the
Heimberg & Holt (1990), did not find that social phobia subtype
cases. Diagnoses were made by experienced clinicians. The pa-
was an outcome predictor. In this same sense, Turner, Beidel,
tients were also interviewed to determine axis II comorbidity. We

---

428
M. ÁNGELES RUIPÉREZ, A. GARCÍA-PALACIOS AND C. BOTELLA
use an instrument designed for this aim, following the Spitzer &
Inventories and scales:
Williams SCID-II (1987) structure, but introducing the DSM-IV
criteria. We also established certain exclusionary criteria: severe
a) Weekly measures
organic disease, mood disorders, psychosis, and substance abuse
related disorders.
AFS (Avoidance and Fear Scale) (Adapted from Marks & Mat-
46.4% of the participants met the criteria for circumscribed so-
hews, 1979): The patient and the therapist established 4 behaviors
cial phobia (persistent and intense fear in one or two social situa-
or situations that the patient avoided because of social phobia and
tions) and 53.6% met criteria for generalized social phobia (per-
he/she rated the level of avoidance on a 0-10 scale where 0 was I
sistent and intense fear in most social situations). 32.1% presented
never avoid it and 10 was I always avoid it; the level of fear was
axis II comorbidity (most patients met criteria for APD, just one
rated on another 0-10 scale, where 0 was No fear and 10 was Ex-
patient met criteria for paranoid personality disorder). A binomial
treme fear.
test was carried out and results showed that in our sample, there
Therapist and patient improvement measures (Adapted from
were no differences in the proportion of participants with genera-
Guy, 1976):
lized social phobia and circumscribed social phobia (p= 0.70), and
there were no differences either between the proportion of partici-
TGI (Therapist Global Impression): The therapist answered
pants with axis II comorbidity and without axis II comorbidity (p=
the question: From your clinical experience, how would you eva-
0.24).
luate the overall severity of this patient? and evaluated from a
The participants age ranged in age from 18 to 63, with a mean
clinical point of view the global impression on the severity of the
of 28 years (SD= 11.51). 39.3% of the participants were males and
patient on a 1-6 subjective scale, where 1 was Normal, 2 was
60.7% females. 78.6% were single and 21.4% were married. Re-
Slightly perturbed, 3 was Moderately perturbed, 4 was Quite
garding the level of education, 67.9% of the patients have a degree
perturbed, 5 was Severely perturbed, and 6 was Very severely
or were studying at the university, 25% had finished high school
perturbed.
and 7.1% had finished primary school. The mean duration of the
CIT (Clinical Improvement: Therapist): The therapist evalua-
disorder was around 12 years. During the interviews, many pa-
ted patient improvement from the beginning of the treatment on a
tients reported to have suffered this problem «for ever». To clarify
0-7 scale where 1 was Much better, 2 Quite better, 3 A little bet-
the onset of the disorder we use as a criterium, the time when their
ter, 4 No changes, 5 A little worse, 6 Quite worse, and 7 Much
social fears began to cause significant impairment in their lives.
worse.
We divided the sample according the different aims of our
CIP (Clinical Improvement: Patient): The patient evaluated the
study:
level of improvement from the beginning of the treatment on a 1-
7 subjective scale, the same as CIT2.
1) Circumscribed Social Phobia (N= 13) vs Generalized Social
Phobia (N= 15). The mean age of the first group was 26 years
(SD= 12.78) and of the second group 29 (SD= 10.53). With regard
b) Pre/Post measures
to the mean duration of the disorder in the first group, this was
12.76 years (SD= 14.11) and in the second, 12.46 (SD= 9.77).
State-Trait Anxiety Inventory (STAI) (Spielberger, Gorsuch &
2) Axis II comorbidity (N= 9) vs. No Axis II comorbidity (N=
Lushene, 1970), Spanish version (TEA, 1988).
19). The mean age in the first group was 28 years (SD= 7.35) and
Beck Depression Inventory (BDI). (Beck, Ward, Mendelsohn,
in the second 28 (SD= 13.21). The mean duration of the disorder
Mock & Erbaugh, 1961), Spanish version (Conde & Franch,
in the first group was 12.11 years (SD= 7.32) and in the second
1984).
this was 12.84 (SD= 3.53).
The Social Avoidance and Distress Scale (SADS) (Watson &
Friend, 1969). This is a 28-item true-false scale that assesses an-
The data were analyzed in two assessment periods: pre-treat-
xiety and avoidance of several social situations.
ment and post-treatment.
Fear of Negative Evaluation Scale (FNE) (Watson & Friend,
1969). This is a 30-item true-false scale that assesses concern
Measures
about being evaluated negatively by others.
The Social Phobia and Anxiety Inventory (SPAI) (Turner, Bei-
Diagnostic Interviews
del, Dancu & Stanley, 1989). This instrument includes 2 scales:
Social Phobia Scale (32 items), and Agoraphobia Scale (13 items).
Admission Interview (unpublished manuscript): Through this
It assesses cognitive, behavioral, and physiological responses
interview, information on demographic and clinical variables was
across several situations.
obtained: The reason for coming to the clinic, the duration of the
Self-esteem Scale (Rosenberg, 1965). This instrument assesses
problem, severity of the problem as perceived by the patient, for-
the degree of self-satisfaction. It is a 10-item instrument that is res-
mer treatments, alcohol and substance use, and presence of organic
ponded to on a 4-point scale ranging from 0 (I completely agree)
diseases. During the interview, the patient was also asked certain
to 3 (I completely disagree).
questions to determine the presence of different anxiety disorders.
Impairment Questionnaire. (in Borda & Echeburúa, 1991):
Anxiety Disorders Interview Schedule (ADIS-R) (Di Nardo &
This questionnaire evaluated the impairment that the disorder had
Barlow, 1988): This instrument is an interview designed to assess
caused in several areas of the patient’s life: couple, family, leisu-
and diagnose Anxiety Disorders with the DSM-III-R criteria. Be-
re, social area, work, and global impairment. Each area was rated
sides the clinical diagnosis, this interview assesses the situational
on a 5-point scale with scores ranging from 0 (Not at all) to 5
and cognitive factors that have influenced in anxiety.
(Completely).

---

CLINICAL FEATURES AND TREATMENT RESPONSE IN SOCIAL PHOBIA: AXIS II COMORBIDITY AND SOCIAL PHOBIA SUBTYPES
429
Results
Measures directly related to social phobia: We obtained a signi-
ficant Group effect, F(4,16)= 3.7, p<0.05. Univariate analysis reve-
Demographic and psychopathological features at pre-treatment
aled a Group effect in Self-esteem, F(1,19)= 7.45, p<0.05, SADS,
F(1,19)= 12.77, p<0.005, and SPAI-S, F(1,19)= 5.30, p<0.05. In all
These features were analyzed in regards to the social phobia
these measures, generalized social phobics showed a higher im-
subtype and axis II comorbidity. We carried out Chi-Square and
pairment. We also found a significant Time Effect, F(4,16)= 6.63,
MANOVAs (Wilks lambda). Dependent variables were grouped in
p<0.005), which was evident in all the measures on univariate tests:
4 categories: variables directly related to social phobia (FNE,
FNE, F(1,19)= 11.49, p<0.005, Self-esteem, F(1,19)= 23.25,
SADS, SPAI-S and Self-esteem Scale); Anxiety and depression
p<0.001, SADS, F(1,19)= 10.25, p<0.005, and SPAI-S, F(1,19)=
variables (BDI and STAI-R); clinical status: impairment and im-
21.65, p<0.001. No Group x Time effect was found.
provement measures (TGI, CIT, CIP and Impairment Scale); and,
Anxiety and Depression measures: MANOVAs revealed a
finally the Avoidance and Fear Scale. Previously we checked that
Group effect, F(2,21)= 5.01, p<0.05. In univariate tests, we found
the measures in each category were highly correlated
this effect in both measures, BDI F(1,22)= 6.44, p<0.05, and
STAI-R, F(1,22)= 10.00, p<0.005; generalized social phobics sco-
Circumscribed vs. Generalized Social Phobia
red higher in these measures than circumscribed social phobics.
Chi-Square tests did not reveal any significant differences in
sex, level of education or marital status.
Table 1
Regarding measures directly related to social phobia, MANO-
Means, Standard Deviations, F values and level of significance of the scores in
VAs revealed significant differences between the groups (F(4,20)=
the measures at pre-treatment and post-treatment in generalized social phobia
and circumscribed social phobia groups
10.77, p< 0.001). Univariate tests showed the Group effect in all
measures. We also found statistical differences in the anxiety and
Measures
Group
Pre-treatment
Post-treatment
depression measures, F(2,21)= 3.81, p<0.05). Regarding the clini-
M
SD
F
M
SD
F
cal status measures, there were also significant differences betwe-
en the groups, F(4,21)= 3.307, p<0.05). Univariate tests showed,
SPAI-S
1
089.04 20.75
08.63***
68.67
21.11
10.23**
however, differences in only one of the measures, TGI. Finally, we
2
117.48
23.64
72.64
21.83
found no significant differences in fear and avoidance. In all cases
FNE
1
020.87
03.87
23.31***
19.12
04.73
NS
where differences were encountered, the generalized social phobia
2
026.84
02.03
18.53
07.35
group scored higher (and lower in the Self-esteem Scale) than the
Circumscribed social phobia group. Table 1 shows the mean, stan-
SADS
1
012.62
05.58
24.22***
11.12
04.39
20.21**
2
023.31
05.29
15.15
06.68
dard deviations, F values and level of significance for these groups
in the measures analyzed.
Self-Esteem Scale
1
029.62
04.50
13.29***
31.62
04.56
11.39**
2
023.23
03.85
29.00
03.76
Axis II Comorbdity vs No Axis II Comorbidity
BDI
1
007.81
05.34
06.69***
04.63
03.17
06.78**
2
016.77
08.85
09.23
09.03
Chi square tests did not reveal significant differences in demo-
graphic variables: sex, level of education and marital status.
STAI-Trait
1
028.18
10.95
05.38***
19.09
07.34
07.20**
Regarding measures directly related to social phobia, the MA-
2
039.46
09.64
29.69
11.78
NOVA did not reveal any differences between the groups. There
were no differences regarding either anxiety or depression measu-
TGI (Therapist Global
1
003.83
00.39
12.18***
01.83
00.57
NS
Impression)
2
004.00
00.00
01.80
00.86
res. As for clinical status variables, multivariate analysis showed
no differences. However, univariate tests did reveal significant di-
CIT (Clinical
1
003.25
00.96
NS
01.75
00.63
NS
fferences between the groups in TGI (Therapist Global Impres-
Improvement: therapist)
2
003.40
00.82
01.66
00.82
sion). Patients with axis II comorbidity were evaluated by thera-
pists as more severe than patients without axis II comorbidity. Fi-
CIP (Clinical
1
003.25
00.86
NS
02.25
00.45
NS
Improvement: Patient)
2
004.47
00.84
02.33
00.97
nally, there were no differences between the groups in avoidance
and fear. Even though we did not find any significant differences,
Global Impairment
1
002.75
00.97
NS
01.45
00.93
NS
we would like to point out that the group with axis II comorbidity
2
003.47
01.13
02.00
01.33
scored higher than the group without axis II comorbidity in seve-
ral measures, of which we would emphasize SADS, FNE, BDI,
Avoidance
1
002.83
01.94
NS
01.83
01.40
NS
2
003.33
01.75
01.13
01.18
and Global Impairment (See table 2).
Fear
1
003.50
01.68
NS
02.66
01.07
NS
Treatment Response
2
003.66
01.40
02.07
01.28
To analyze differential treatment response, repeated measures
Group 1: Circumscribed Social Phobia Group 2: Generalized Social Phobia
Level of statistical significance: * p < .05; ** p < .005; *** p < .001; NS: Non statistically signifi-
MANOVAs (Wilks lambda) were carried out.
cant
Note: Degree of freedom (pre-treatment) for SPAI-S, FNE, SADS and Self-Esteem scale: 1, 23; for
Circumscribed vs. Generalized social Phobia
STAI-T and BDI: 1, 25 and for TGI: 1, 24.
Note: Degree of freedom (post-treatment) for SPAI-S, SADS and Self-Esteem scale: 1, 20; for STAI-
Repeated measures MANOVAs showed the following results:
T and BDI: 1, 22

---

430
M. ÁNGELES RUIPÉREZ, A. GARCÍA-PALACIOS AND C. BOTELLA
We also found a Time effect, F(2,21)= 11.50, p<0.001 that appea-
zed measures, which revealed that our treatment was effective in
red in both measures in univariate tests, BDI, F(1,22)= 17.26,
both groups. Finally, it seemed that the treatment was equally ef-
p<0.001, and STAI-R, F(1,22)= 16.84, p<0.001. Finally, analysis
fective in both groups, given the fact that we found no interaction
revealed no Time x Group interaction effect.
effects. However, if we look at the means (see table 1), we notice
Clinical status measures: Analysis revealed no Group effect,
that in the measures directly related to social phobia, the differen-
but a Time effect, F(4,22)= 45.65, p<0.001. Univariate tests sho-
ces found between the groups before treatment tended to decrease
wed this effect in all measures: (TGI) Therapist Global Impres-
after treatment completion.
sion, F(1,25)= 63.59, p<0.001, Global Impairment, F(1,25)=
21.83, p<0.001, (CIP) Clinical Improvement: Patient, F(1,25)=
Axis II Comorbidity vs. No Axis II Comorbidity
67.19, p<0.001, and (CIT) Clinical Improvement: Therapist,
F(1,25)= 179.27, p<0.001. There was no Group x Time effect.
Variables directly related to social phobia: We did not find a
Avoidance and Fear Scale (AFS): We only found a time effect,
group effect, but there was a significant Time effect, F(4,16)=
F(2,24)= 9.95, p<0.001. Univariate tests showed this effect in both
12.44, p<0.001. Univariate tests showed this effect in all analyzed
measures, Avoidance, F(1,25)= 20.71, p<0.001, and Fear,
variables: FNE, F(1,19)= 25.98, p<0.001, Self-esteem Question-
F(1,25)= 12.45, p<0.005.
naire, F(1,19)= 33.43, p<0.001, SADS, F(1,19)= 15.65, p<0.001,
In summary, we found a Group effect in variables directly rela-
and SPAI-S, F(1,19)= 28.02, p<0.000. There were also a Time x
ted to social phobia and in anxiety and depression measures. In
Group interaction effect, (4,16)= 4.82, p<0.010). Univariate tests
these variables, generalized social phobics were more impaired th-
showed this effect in the following measures: FNE, F(1,19)=
roughout the entire process. We found a Time effect in all analy-
10.26, p<0.005, and Self-esteem Questionnaire, F(1,19)= 4.54,
p<0.05, that is to say, the treatment achieved a higher decrement
in fear to negative evaluation and a higher increment in self-este-
Table 2
Means, Standard Deviations, F values and level of significance of the scores in
em in patients with axis II comorbidity.
the measures at pre-treatment and post-treatment in axis II comorbidity and non
Anxiety and Depression measures: There was a Group effect,
axis II comorbidity groups
F(2,21)= 4.01, p<0.05. Univariate analysis showed that this effect
appears only in STAI-R, F(1,22)= 8.40, p<0.01; patients with axis
Measures
Group
Pre-treatment
Post-treatment
II comorbidity presented higher scores in this variable. We found
M
SD
F
M
SD
F
a significant Time effect, F(2,21)= 12.34, p<0.000, which appea-
SPAI-Social
1
119.21 22.43
NS
70.98 22.43
NS
red in both measures, BDI, F(1,22)= 22.37, p<0.001, and STAI-R,
2
098.90 28.39
71.21
21.19
F(1,22)= 14.53, p<0.001. No Time x Group effect was found.
Clinical status measures: We did not find a Group effect, but
FNE
1
027.75
01.38
NS
16.62
08.74
NS
we did find a Time effect, F(4,22)= 40.55, p<0.000, which appea-
2
022.61
03.94
20.08
04.21
red in all measures, TGI, F(1,25)= 54.85, p<0.001, Global impair-
SADS
1
023.62
04.13
NS
14.75
08.25
NS
ment, F(1,25)= 26.10, p<0.001, CIP, F(1,25)= 59.39, p<0.001, and
2
016.53
07.91
12.92
04.62
CIT, F(1,25)= 160.44, p<0.001. No interaction effect was found.
Avoidance and Fear Scale (AFS). Only a significant Time ef-
Self-esteem Scale
1
023.00
04.72
NS
29.50
04.72
NS
fect was found, F(2,24)= 10.32, p<0.001. Univariate tests showed
2
027.30
04.78
30.31
03.98
this effect in both measures, Avoidance, F(1,25)= 21.23, p<0.001,
BDI
1
018.12
10.29
NS
09.25
11.13
NS
and Fear, F(1,25)= 11.79, p<0.005.
2
009.93
06.35
06.06
04.34
In summary, in measures related to social phobia and anxiety
and depression variables, a group effect was evident. In these me-
STAI-Trait
1
041.12 10.03
NS
32.50 13.06
5.46*
asures, patients with axis II comorbidity scored higher. On the
2
030.87 10.98
21.00
11.18
other hand, treatment was effective, as can be concluded from the
TGI (Therapist global
1
004.66
00.86
7.48*
02.66
01.00
NS
Time effect found in all measures. Finally, we found only diffe-
Impression)
2
003.55
00.92
02.11
00.58
rential treatment effectiveness (interaction effect) in two measu-
res, the fear to negative evaluation (FNE), and self-esteem. In the-
CIT (Clinical
1
004.00
00.00
NS
01.88
00.93
NS
se measures, patients with axis II comorbidity improved more than
Improvement: Therapist)
2
003.89
00.32
01.77
00.64
patients without axis II comorbidity.
CIP (Clinical
1
003.22
00.83
NS
01.66
00.86
NS
Improvement: Patient)
2
003.40
00.91
01.72
00.67
Psychopathological Features at post-treatment
Global Impairment
1
003.00
01.51
NS
01.88
01.61
NS
Circumscribed Social Phobia vs. Generalized Social Phobia
2
002.61
01.03
01.77
01.06
Avoidance
1
003.22
01.78
NS
01.11
01.27
NS
As we have just seen, the differences found at pre-test tended
2
003.05
01.89
01.61
01.33
to decrease after treatment completion. Upon observing this ten-
dency, we planned to analyze whether or not the differences in cli-
Fear
1
003.22
01.48
NS
01.88
01.05
NS
nical measures were maintained at post-test. Results are shown in
2
003.77
01.51
02.55
01.24
table 1. Regarding measures related to social phobia, again we
Group 1: Axis II Comorbidity. Group 2: Non Axis II Comorbidity.
found significant differences, F(4,20)= 12.17, p<0.001). Univaria-
Level of statistical significance: * p < .05; NS: Non statistically significant
te tests revealed differences in SPAI-S, SADS, and Self-esteem
Note: Degree of freedom (pre-treatment) for TGI: 1, 24 and for STAI-T (post-treatment): 1, 25
Questionnaire. In anxiety and depression measures, we also found

---

CLINICAL FEATURES AND TREATMENT RESPONSE IN SOCIAL PHOBIA: AXIS II COMORBIDITY AND SOCIAL PHOBIA SUBTYPES
431
significant differences, F(2,21)= 3.610, p<0.05. Univariate analy-
The second aim was the analysis of the differential treatment
sis revealed differences in both measures. In the other measures,
response regarding social phobia subtype and axis II comorbidity.
there were no statistical differences between Circumscribed and
As for social phobia subtype, in most measures there were no di-
Generalized Social Phobia. If we compare these differences with
fferences in treatment effectiveness between generalized social
those obtained at pre-test, we see that some differences were main-
phobics and circumscribed social phobics. With regard to axis II
tained once treatment was completed. However, in some variables
comorbidity. This data is along the same line of Brown et al.
where we found differences at pre-test, we did not find any diffe-
(1995), Hope et al., (1995) and Turner et al., (1996). These last
rences at post-test, FNE, Global Therapist Impression. That is to
authors found that although the treatment showed similar effecti-
say, although at pre-test generalized social phobics presented a
veness in generalized and circumscribed social phobics, generali-
higher fear to negative evaluation and the therapist evaluated them
zed social phobics were more impaired than circumscribed social
as more impaired than circumscribed social phobics, after treat-
phobics after termination of treatment. Our findings indicate that
ment completion both groups presented similar scores in these me-
at pre-test generalized social phobics showed differences in some
asures (see table 1).
clinical variables –they were more impaired– and at post-test they
continued to show higher impairment in some of these measures.
Axis II Comorbidity vs. No Axis II Comorbidity
In two variables, however, fear to negative evaluation and thera-
pist global impression, the treatment managed to make those di-
Finally, we also analyzed clinical differences at post-test regar-
fferences disappear.
ding Axis II comorbidity. Multivariate analysis showed no signifi-
Regarding axis II comorbidity, there were no differences in tre-
cant differences at post-test. We only found differences at post-test
atment response in most measures. However, the differences bet-
in STAI-T. The group with axis II comorbidity scored higher than
ween the groups at pre-test disappeared at post-test. Besides this,
the group without axis II comorbidity. If we compare these diffe-
we also found interaction effects in FNE and Self-esteem Scale
rences with those obtained at pre-test, it can be seen that the diffe-
which revealed that patients with a personality disorder had a hig-
rences in Therapist Global Impression disappeared at post-test
her improvement in these variables than patients without persona-
(see table 2).
lity disorder. This would appear to suggest that our treatment has
been very effective regarding fear to negative evaluation and self-
Discussion
esteem in social phobics with axis II comorbidity. However we
should be cautious in regards to this finding. We will have to wait
Over the past 10 years, numerous studies have been carried out
for a follow-up assessment to see if this result is maintained. In
to demonstrate the effectiveness of cognitive-behavioral pro-
conclusion, our results about axis II comorbidity are similar to tho-
grams for the treatment of Social Phobia (Feske & Chambless,
se of Brown et al. (1995), and Hope et al. (1995), who, despite ha-
1995; Moreno, Méndez & Sánchez, 2000; Taylor, 1996; Turner,
ving found a differential tendency in social phobics with persona-
Cooley-Quille & Beidel, 1996). However, the success indices are
lity disorder, the treatment response did not differ from social
smaller than those obtained in other disorders, such as Panic Di-
phobics without personality disorder. We agree with Heimberg
sorder (Gould, Otto & Pollack, 1995). Because of this, resear-
(1996) in that it would seem reasonable to think that suffering
chers have recently been paying attention to the analysis of diffe-
from a personality disorder associated to social phobia should ma-
rential features and response to treatment regarding social phobia
ke the treatment more difficult. However, our data and that from
subtype and axis II comorbidity (i.e., Heimberg et al., 1990; Her-
other studies (see Heimberg, 1996) would support the idea that tre-
bert et al., 1992; Hoffman et al., 1995; Holt et al., 1992; Mersch
atment responses are not different when there is a personality di-
et al., 1989; Öst et al., 1981; Turner et al., 1992; Turner et al.,
sorder. These results brought us to a controversy over whether or
1996). The present work was designed to study these issues in a
not APD is qualitatively distinct from social phobia. Heimberg
Spanish population.
(1996), taking into account the results from certain studies (i.e.,
Regarding the first aim, that is, the study of demographic and
Brown et al., 1995; Holt et al., 1992; Hope et al., 1995), states that
clinical differences regarding social phobia subtype and axis II co-
the differences between these disorders are dimensional, that is to
morbidity, our results showed that there were significant differen-
say, a question of degree, in regards to the severity of the disorder
ces between social phobia subtypes in general measures as trait an-
(Heckelman & Schneider, 1995; Heimberg, 1996). Other resear-
xiety (STAI-R), and depression (BDI), and in more specific social
chers, although their results did not support the idea that these di-
phobia measures as FNE, SADS, SPAI-S, self-esteem, and also the
sorders are qualitatively different, suggest we should look for the
therapist’s clinical global impression. In all these measures, pa-
qualitative differences in measures as self-esteem (Tran & Cham-
tients with generalized social phobia were more impaired. This da-
bless, 1995) or, even, to improving the methodology of the studies
ta confirmed that generalized social phobics were more impaired
(Hoffman et al., 1995). Finally, Turner et al. (1992) are also of the
than circumscribed social phobics, Generalized social phobics had
opinion that these disorders are qualitatively different, and state
a higher fear to negative evaluation, higher social anxiety, lower
that the fact that no differences were found is due to the similarity
self-esteem, and, finally, the patients evaluated themselves as
of diagnostic criteria which tends to make these two disorders
being more impaired. These results are similar to those of other
overlap. Our findings support the idea that these disorders are only
studies (Heimberg et al., 1990 and Turner et al., 1996).
different from a quantitative point of view, given the fact that we
As for Axis II Comorbidity, results only showed differences in
did not find different treatment responses.
the therapist’s global impression. The therapist evaluated patients
Results from this study therefore support the idea that our tre-
with axis II comorbidity as being more impaired. These results are
atment is equally effective for all the patients in the sample, inde-
similar to those of Hoffman et al. (1995), Tran & Chambless
pendent of the social phobia subtype or axis II comorbidity. On the
(1996) and Turner et al. (1996).
other hand, we find encouraging the fact that, in the measure of

---

432
M. ÁNGELES RUIPÉREZ, A. GARCÍA-PALACIOS AND C. BOTELLA
fear to negative evaluation (FNE) and the Therapist Global Im-
neity that clinicians find in their daily practice with social phobia.
pression, those patients who showed a more severe impairment
These results will hopefully encourage researchers to continue in
(axis II comorbidity) achieved the same level of impairment as pa-
delimiting the differential features and achieve treatments that are
tients who were less impaired. Follow-up studies, however, are ne-
equally effective for all sufferers of social phobia, including or de-
eded to be able to confirm this finding.
leting therapeutic differential components that can provide an ans-
This study agrees with writings on the high effectiveness of
wer for these differences.
cognitive-behavioral programs in social phobia, which have beco-
me the first lines of therapeutic action for this disorder. We feel,
Agradecimientos
however, that research must design treatment alternatives that can
help patients with higher levels of severity.
This study was partially supported by Conselleria d’Educació i
In summary, our findings support the idea that social phobia
Ciència, Generalitat Valenciana (project GV-2421/94) and funda-
subtype and axis II comorbidity can help to explain the heteroge-
ció Caixa-Castelló (project P1B97-02).
References
Alden, L.E. & Capreol, M.J. (1993). Avoidant personality disorders: Inter-
Heimberg, R.G., Hope, D.A., Dodge, C.S. & Becker, R.E. (1990). DSM-
personal problems as predictors of treatment response. Behavior The-
III-R subtypes of phobia social: Comparison of generalized social pho-
rapy, 24, 357-376.
bics and public speaking phobics. Journal of Nervous and Mental Dise-
American Psychiatric Association (APA) (1987). Diagnostic and statisti-
ase, 178, 172-179
cal manual of mental disorders (3ª ed. Rev.) (DSM-III-R). Washington,
Heimberg, R.G., Holt, C.S., Schneider, F.R., Spitzer, R.L. & Liebowitz,
DC: APA (Spanish translation: Masson, Barcelona, 1988).
M.L. (1993). The issue of subtypes in the diagnosis of social phobia.
American Psychiatric Association (APA) (1994). Diagnostic and statisti-
Journal of Anxiety Disorders, 7, 249-270.
cal manual of mental disorders (DSM-IV). Washington, DC: APA
Heimberg, R.G., Juster, H.R., Hope, D.A. & Mattia, J.I. (1995). Cognitive
(Spanish translation: Masson, Barcelona, 1995).
behavioral group treatment for social phobia: description, case presen-
Beck, A.T., Ward, C.H., Mendelson,M., Mock, J. & Erbaugh, J. (1961). An
tation, and empirical support. In M. Stein (Ed.), Social Phobia: Clini-
inventory for measuring depression. Archives of General Psychiatry, 4,
cal and Research Perspectives (pp. 293-321). American Psychiatric
561-571.
Press: Washington, DC.
Borda, M. & Echeburúa, E. (1991). La autoexposición como tratamiento
Herbert, J.D., Hope, D.A. & Bellack, A.S. (1992). Validity of the dis-
psicológico en un caso de agorafobia. Análisis y Modificación de Con-
tinction between generalized social phobia and avoidant personality di-
ducta, 18, 101-103.
sorder. Journal of Abnormal Psychology, 101, 332.339.
Brooks, R.B., Baltazar, P.L. & Munjack, D.J. (1989). Co-occurrence of
Hoffman, S.G., Newman, M.G., Becker, E., Taylor C.B. & Roth, W.T.
personality disorders with panic disorders, social phobia, and generali-
(1995). Social phobia with and without avoidant personality disorder:
zed anxiety disorders: A review of the literature. Journal of Anxiety Di-
Preliminary behavior therapy outcome findings. Journal of Anxiety Di-
sorders, 3, 259-285.
sorders, 9, 427-438.
Brown, E.J., Heimberg R.G. & Juster, H.R. (1995). Social phobia subtype
Hoffman, S.G., Newman, M.G., Ehlers, A. & Roth, W.T. (1995). Psy-
and avoidant personality disorder: Effect on severity of social phobia,
chophysiological differences between subgroups of social phobia.
impairment, and outcome of cognitive behavioral treatment. Behavior
Journal of Abnormal Psychology, 104, 224-231.
Therapy, 26, 467-486.
Holt, C.S., Heimberg, R.G. & Hope, D.A. (1992). Avoidant personality di-
Chambless, D.L., Tran, G.Q. & Glass, C.R. (1997). Predictors of response
sorder and generalized subtype of social phobia. Journal of Abnormal
to cognitive-behavioral treatment for social phobia. Journal of Anxiety
Psychology, 101, 318-325.
Disorders, 11, 221-240.
Holt, C.S., Heimberg, R.G. & Hope, D.A. (1990). Success from the outset:
Conde, V. & Franch, J.I. (1984). Escalas de evaluación comportamental
Predictors of cognitive-behavioral therapy outcome among social pho-
para la cuantificación de la sintomatología psicopatológica en los
bics. Paper presented at the annual meeting of the Association for the
trastornos angustiosos y depresivos. Upjohn.
Advancement of Behavior Therapy, San Francisco.
DiNardo, P.A. & Barlow, D.H. (1988). The Anxiety Disorders Interview
Hope, D.A., Herbert, J.D. & White, C. (1995). Diagnosis subtype, avoidant
Schedule-Revised (ADIS-R). Graywind Publications: Albany, Nueva
personality disorder, and efficacy of cognitive behavioral group the-
York.
rapy for social phobia. Cognitive Therapy and Research, 19, 399-417.
Feske, V. & Chambless, D.L. (1995). Cognitive-behavioral versus exposu-
Jerrelmalm, A., Jansson, L. & Öst, L.G. (1986). Cognitive and physiologi-
re treatment for social phobia: A meta-analysis. Behavior Therapy, 26,
cal reactivity and the effects of different behavioral methods in the tre-
695-720.
atment of social phobia. Behavior Research and Therapy, 24, 171-180.
Gould, R.A., Otto, M.W. & Pollack, M.H. (1995). A meta-analysis of tre-
Juster, H.R., Heimberg, R.G. & Engelberg, B. (1995). Self selection and
atment outcome for panic disorder. Clinical Psychology Review, 15,
sample selection in a treatment study of social phobia. Behavior Rese-
819-844.
arch and Therapy, 33, 321-324.
Guy, W. (1976). ECDEU. Assessment Manual for Psychopharmacology
Marks, I. & Mathews, A. (1979). Brief standards self-rating for phobic pa-
Revised. NIMH Publ. DHEW Publ. No (Adm.), 76-338.
tients. Behavior Research and Therapy, 17, 263-267.
Heckelman, L.R. & Scheneider, F.R. (1995). En R.G. Heimberg, M.R. Lie-
Mersch, P.P.A., Emmelkamp, P.M.G., Bögels, S.M. & van der Sleen, J.
bowitz, D.A. Hope & F.R. Schneider (Eds.), Social phobia: Diagnosis,
(1989). Social phobia: Individual response patterns and the effects of
assessment and treatment. New York: Guilford Press.
behavioral and cognitive interventions. Behavior Research and The-
Heimberg, R.G. (1986). Predicting the outcome of cognitive-behavioral
rapy, 27, 421-434.
treatment of social phobia. Paper presented at the annual meeting of the
Moreno, P.J., Méndez, X. & Sánchez, J. (2000). Effectiveness of cogniti-
Society for Psychotherapy Research, Wellwsley, MA.
ve-behavioural treatment in social phobia: a meta-analytic review. Psi-
Heimberg, R.G. (1996). Social Phobia, Avoidant Personality Disorder and
cothema, 12, 246-352.
the Multiaxial Conceptualization of Interpersonal Anxiety. In P.M. Sal-
Öst, L.G., Jerremalm, A. & Johansson, J. (1981). Individual response pat-
kovskis (Ed.), Trends in Cognitive and Behavioral Therapies. New
terns and the effects of different behavioral methods in the treatment of
York: Wiley.
social phobia. Behavior Research and Therapy, 19, 1-16.

---

CLINICAL FEATURES AND TREATMENT RESPONSE IN SOCIAL PHOBIA: AXIS II COMORBIDITY AND SOCIAL PHOBIA SUBTYPES
433
Poulton, R.G. & Andrews, G. (1996). Change in danger congnitions in
personality disorders. Journal of Abnormal Psychology, 101, 326-
agoraphobia an social phobia during treatment. Behavior Research and
331.
Therapy, 34, 413-421.
Turner, S.M., Beidel, D.C., Dancu, C.V. & Stanley, M.A. (1989). An em-
Rosenberg, B. (1965). Society and the adolescent self-image. New Jersey:
pirically derived inventory to measure social fears and anxiety: The so-
Princenton.
cial Phobia Anxiety Inventory (SPAI). Psychological Assessment, 1,
Salaberría, K. & Echeburúa, E. (1995). Tratamiento psicológico de la fo-
35-40
bia social: Un estudio experimental. Análisis y modificación de con-
Turner, S.M., Beidel, D.C., Dancu, C.V. & Keys, D.J. (1986). Psychopat-
ducta, 21, 151-179.
hology of social phobia and comparison to avoidant personality disor-
Scholing, A. & Emmelkamp, P.M.G. (1999). Prediction of treatment out-
ders. Journal of Abnormal Psychology, 95, 389-394
come in social phobia: a cross-validation. Behaviour Research and
Turner, S.M., Beidel, D.C., Wolff, P.L. & Spaulding, S. (1996). Clinical fe-
Therapy, 37, 659-670.
atures affecting treatment outcome in social phobia. Behaviour Rese-
Spielberger, C.D., Gorsuch, R.L. & Lushene, R.E. (1970). The State-Trait
arch and Therapy, 34, 795-804.
Anxiety Inventory. Palo Alto, California: Consulting Psychologist
Turner, S.M., Cooley-Quille, M.R. & Beidel, D.C. (1996). Behavioral and
Press.
pharmacological treatment of social phobia: Long term outcome. In
Spitzer, R.L. & Williams, J.B.W. (1987). Structured Clinical Interview for
Mavissakalian, M. & Prien, R. (Eds.), Anxiety disorders: Psychologi-
DSM-III-R Personality Disorders (SCID-II). New York: Biometrics
cal and pharmacological treatments. Washington D.C.: American Psy-
Research Department, New York State Psychiatric Institute.
chiatric Press.
Taylor, S. (1996). Meta-analysis of cognitive-behavioral treatments for so-
van Velzen, C.J.M. & Emmelkamp, P.M.G. (1996). The assessment of per-
cial phobia. Journal of Behavior Therapy and Experimental Psy-
sonality disorders: Implications for cognitive and behavior therapy. Be-
chiatry, 27, 1-9.
havior Research and Therapy, 34, 655-668.
Tran, G.Q. & Chambless, D.L. (1995). Psychopatology of social phobia:
van Velzen, C.J.M., Emmelkamp, P.M.G. & Scholing, A. (1997). The im-
Effects of subtypes and avoidant personality disorders. Journal of An-
pact of personality disorders on behavioral treatment for social phobia.
xiety Disorders, 9, 489-501
Behaviour Research and Therapy, 35, 889-900.
Turner, S.M., Beidel, D.C. & Townsley, R.M. (1992). Social phobia:
Watson, D. & Friend, R. (1969). Measurement of social evaluative anxiety.
A comparation of specific and generalized subtypes and avoidant
Journal of Consulting and Clinical Psychology, 33, 448-457.

---