ARTICLE IN PRESS
Behaviour Research and Therapy 44 (2006) 233–247
www.elsevier.com/locate/brat
Social anxiety disorder in veterans affairs primary care clinics
Todd B. Kashdana,, B. Christopher Fruehb, Rebecca G. Knappb, Renée Hebertb,
Kathryn M. Magruderb
a
Department of Psychology, George Mason University, MS 3F5, Fairfax, VA 22030, USA
b
Medical University of South Carolina, USA
Received 27 May 2004; received in revised form 30 January 2005; accepted 16 February 2005
Abstract
To examine the prevalence and correlates of social anxiety disorder (SAD) in veterans, 733 veterans from
four VA primary care clinics were evaluated using self-report questionnaires, telephone interviews, and a
12-month retrospective review of primary care charts. We also tested the concordance between primary care
providers’ detection of anxiety problems and diagnoses of SAD from psychiatric interviews. For the multisite sample, 3.6% met criteria for SAD. A greater rate of SAD was found in veterans with than without
post-traumatic stress disorder (PTSD) (22.0% vs. 1.1%), and primary care providers detected anxiety
problems in only 58% of veterans with SAD. The elevated rate of comorbid psychiatric diagnoses and
suicidal risk associated with SAD was not attributable to PTSD symptom severity. Moreover, even after
controlling for the presence of major depressive disorder, SAD retained unique, adverse effects on PTSD
diagnoses and severity, the presence of other psychiatric conditions, and suicidal risk. These results attest to
strong relations between SAD and PTSD, the inadequate recognition of SAD in primary care settings, and
the significant distress and impairment associated with SAD in veterans.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Social anxiety disorder; Post-traumatic stress disorder; Veterans; Comorbidity; Suicidality; Specificity
Corresponding author. Tel.: +1 703 993 9486; fax: +1 703 993 1359.
E-mail address: tkashdan@gmu.edu (T.B. Kashdan).
0005-7967/$ - see front matter r 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.brat.2005.02.002
ARTICLE IN PRESS
234
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
Introduction
Social anxiety disorder (SAD) is the third most prevalent psychiatric condition in the United
States with epidemiological studies estimating a lifetime prevalence rate of 13.3%, and a 1-year
prevalence rate of 7.9% in community samples (Kessler et al., 1994). Rates apparently have
increased over the past few generations (Heimberg, Stein, Hiripi, & Kessler, 2000). Left untreated,
SAD is a persistent and disabling condition that involves the often paralyzing fear of interacting
or doing things in front of other people because of social evaluative concerns. Individuals with
SAD experience high levels of functional impairment at work and school (Schneier et al., 1994;
Wittchen, Fuetsch, Sonntag, Mueller, & Liebowitz, 2000). As for interpersonal functioning, SAD
is associated with smaller social networks, less social support and acceptance, a high probability
of being single or divorced, a low probability of being in a romantic relationship, and less sexual
satisfaction (Davidson, Hughes, George, & Blazer, 1994; Schneier et al., 1994; Wittchen et al.,
2000). Over 70% of individuals with SAD meet criteria for comorbid anxiety, mood, and alcohol
abuse disorders (e.g., Magee, Eaton, Wittchen, McGonagle, & Kessler, 1996; Schneier, Johnson,
Hornig, Liebowitz, & Weissman, 1992), and these individuals are at high risk for suicidality
(Schneier et al., 1992). The present study was interested in expanding the study of SAD to traumaexposed veterans with and without post-traumatic stress disorder (PTSD).
Several studies have shown that PTSD is associated with significant social functioning
difficulties and impaired social relations (e.g., Frueh, Turner, Beidel, & Cahill, 2001; Jordan et al.,
1992; Riggs, Byrne, Weathers, & Litz, 1998). In particular, the presence of PTSD is associated
with social skills problems, less satisfaction in intimate relationships (e.g., romantic, parent–child),
and social interactions and relationships that tend to be characterized by more conflict and
hostility, poorer communication, and less emotional expressiveness, intimacy, and positive
sharing. Despite increased attention to SAD and its role in psychological functioning, the study of
SAD in trauma-exposed veterans (or any trauma survivors) is in its infancy. For the few published
examinations of SAD in veterans, samples have ranged from 41–47 veterans (Crowson, Frueh,
Beidel, & Turner, 1998; Hofmann, Litz, & Weathers, 2003; Orsillo, Heimberg, Juster, & Garrett,
1996); the exception (n ¼ 304) narrowly focused on rates of SAD in veterans with and without
PTSD (Orsillo, Weathers, Litz, Steinberg, Huska, & Keane, 1996). For those studies using
diagnostic interviews, 15% (Hofmann et al., 2003; Orsillo et al., 1996) and 72% (Orsillo et al.,
1996) of veterans with PTSD met criteria for a diagnosis of SAD compared to 5% (Hofmann
et al., 2003), 7% (Orsillo et al., 1996), and 22% (Orsillo et al., 1996) of veterans without PTSD.
Although Orsillo et al. (1996) reported very high rates of SAD in veterans, only 41 veterans were
examined and interviewers were not blind to hypotheses. Overall, existing data support a
significant, albeit neglected, relation between these psychiatric conditions.
Only one published study has examined the correlates of SAD in veterans (Orsillo et al., 1996),
finding post-war social anxiety to be positively associated with war-related shame and adverse
homecoming experiences. Although Orsillo’s seminal work on SAD and PTSD was published
almost a decade ago, only three additional studies have been conducted on the topic with each
narrowly focusing on relations among PTSD, SAD and depressive symptoms. Moreover, the
sample sizes of these studies were small and the recruitment process tended to lack
generalizability, relying on advertisements (Hofmann et al., 2003) and outpatients from mental
health specialty clinics (Crowson et al., 1998; Orsillo et al., 1996).
ARTICLE IN PRESS
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
235
There are reasons to expect the presence of SAD to amplify the difficulties of veterans with and
without PTSD. Individuals who are especially concerned about being rejected and embarrassed
are more sensitive to social threat cues, and tend to interpret neutral and ambiguous social
situations as threatening (e.g., Clark & Wells, 1995; Rapee & Heimberg, 1997). These
information-processing biases lead to intense negative emotions, and negative emotions tend to
be misinterpreted as evidence of social failure, further intensifying initial fears and social cognitive
biases. Depending on the severity and intensity of this cyclical process, ultimately, individuals
engage in some level of experiential avoidance as a means of coping. One of the by-products of
experiential avoidance is that the prerequisites to positive social interactions, such as the ability to
properly display and read social cues, as well as emotional self-disclosure, expressiveness,
responsiveness are disrupted. Thereby, social fears can lead to less positive social activity and
relationships, and greater social impairment.
The manifest behaviors of PTSD and SAD, such as distress in social interactions, behavioral
inhibition, social avoidance patterns, and impaired social relationships may be similar. However,
the etiological and maintaining factors can be expected to diverge across these conditions. With
SAD, these behaviors derive from core fears of being negatively evaluated and rejected, whereas
with PTSD, these behaviors could derive from symptoms such as feeling detached from others,
experiencing a restricted range of positive and negative emotions, and avoidance of social stimuli
associated with trauma experiences. Thus, whether SAD develops before, after, or concomitantly
with PTSD, there is reason to believe that SAD will have an incremental adverse association with
indices of distress, impairment, and well-being in veterans. Overall, there is a general absence of
empirical data on the complex relation between SAD and PTSD, and outcomes related to the
presence of SAD in veterans.
The examination of SAD in veterans is of clinical importance because: (a) positive social
activity is arguably the largest contributor to well-being and quality-of-life (Baumeister & Leary,
1995; Ryff & Singer, 2000), (b) high levels of social support offer resilience for post-trauma
recovery (King, King, Fairbank, Keane, & Adams, 1998; Solomon, Mikulincer, & Avitzur, 1988),
(c) social difficulties tend to be a primary complaint of clients treated for PTSD (Herman, 1992),
and (d) facilitating social and emotional skills has been evaluated as a useful target of PTSD
intervention (Frueh, Turner, Beidel, Mirabella, & Jones, 1996). Factors that impede social
functioning, such as the social fear and avoidance, and functional impairment associated with
SAD, represent a critical area to examine in individuals who have experienced trauma or are at
risk for adverse stress-related outcomes.
Primary care providers tend to be the initial, and sometimes only, professionals to recognize,
diagnose, and manage anxiety disorders in the large majority of medical patients (Fifer, Mathias,
Patrick, Mazaonson, Lubeck, & Buesching, 1994; Kirmayer, Robbins, Dworkind, & Yaffe, 1993;
Ormel, Koeter, van den Brink, & van de Willige, 1991). Benefits of examining SAD in veterans
recruited from Veterans Affairs (VA) primary care clinics are that findings can be generalized to
those veterans who use the VA for their healthcare (as opposed to samples recruited from mental
health or specialty clinics), and primary care providers are at the frontline of psychiatric
evaluations and treatments. Thus, using a primary care veteran sample with assessment
information from primary care providers allows for an examination of prototypical psychiatric
assessment and treatment. To our knowledge, there are no published data on the prevalence and
nature of SAD using primary care veteran samples.
ARTICLE IN PRESS
236
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
The present study was designed to extend the small body of work on SAD in veterans in several
ways. First, we used a large-scale multi-site sample of veterans from four VA primary care clinics.
To maximize generalizability, we began with an initial sample of every veteran with at least one
primary care appointment in the year under study (as opposed to randomly selecting patient
arrivals, which oversamples for heavy health service users). Second, we were interested in
correlates of SAD in veterans. We examined a broad range of socio-demographic and clinical
characteristics including comorbidity, suicidal risk, and dimensions of psychological and physical
quality-of-life. For all variables related to the presence/absence of SAD, we examined the
specificity of SAD effects by controlling for (1) PTSD severity and (2) the presence of depression.
Third, because primary care providers tend to be the first, and sometimes only, professionals that
evaluate psychological distress in veteran and non-veteran populations, we examined whether or
not primary care physicians detected the presence of SAD in their patients. We compared the
recognition rate of anxiety problems in veterans (a liberal test) compared to the presence of SAD
diagnoses as assessed by validated, semi-structured psychiatric interviews.
We hypothesized a strong relationship between SAD and PTSD in veterans. Upon examining
SAD as the index disorder, SAD was expected to increase the risk for other psychiatric conditions,
greater suicidal risk, and lower quality-of-life. Due to an absence of research on the topic,
relations between SAD and socio-demographic and clinical characteristics were exploratory.
Additionally, we expected that primary care providers would fail to detect anxiety problems in a
large percentage of their patients diagnosed with SAD by structured psychiatric interviews.
Method
Study design
We examined veterans at four VA Medical Centers (Charleston and Columbia, SC; Tuscaloosa
and Birmingham, AL) (Magruder et al., 2004). Study participants were randomly selected from a
master list of all patients during the fiscal year 1999 at each site. Consenting participants were
provided with a semi-structured clinic assessment and within 2 months, were administered a
structured telephone interview. Study measures were read aloud to all participants because many
veterans had vision problems or insufficient literacy skills. Additionally, using available medical
charts, we conducted a 12-month retrospective review of each participant’s primary care
treatment. Exclusionary criteria included dementia-related symptoms and being age 80 or older.
Participants
A total of 1198 randomly identified veterans (known to be alive) were approached for
participation, and 885 veterans provided informed consent (74%). Overall, 746 veterans
completed the telephone interview. For those veterans with missing telephone interviews, reasons
were given for 107: 59% were not contactable (incorrect phone number, number disconnected, no
answer after multiple attempts), 23% were contacted but declined further interviews, 4% were
known to have died, and 14% listed ‘‘other’’ reasons. Because of missing follow-up data on
primary instruments, our final sample was 733 veterans. In comparing those with and without a
ARTICLE IN PRESS
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
237
telephone interview, the average PCL score was lower for those with a telephone interview than
for those without (26.4 vs. 29.4, p ¼ :02), indicating that those who did not complete the telephone
PTSD diagnostic study materials had somewhat higher levels of PTSD symptomatology.
Clinic interview measures
Demographic information: Participants were asked about socio-demographic information,
disability status, and combat exposure.
PTSD checklist military version (PCL-M; Weathers, Litz, Herman, Huska, & Keane, 1993): The
17-item PCL-M is a self-report measure of PTSD symptoms during the past month (e.g.,
Blanchard, Jones, Buckley, & Forneris, 1996). Items correspond to Diagnostic and Statistical
Manual of Mental Disorders—4th Edition (DSM-IV; American Psychiatric Association, 1994)
criteria for PTSD.
The short-form health survey (SF-36; Ware & Sherbourne, 1992): The 36-item SF-36 is a selfreport measure of several health domains including: physical functioning; social functioning; daily
role limitations because of physical health problems (role-physical); daily role limitations because
of emotional problems (role-emotional); vitality; psychological distress and well-being (mental
health); and general health perceptions (general health). For all subscales, higher scores reflect
greater health. The SF-36 has been shown to be a valid and reliable instrument for use with
veteran populations (e.g., Richardson, Engel, Hunt, McKnight, & McFall, 2002). The SF-36 raw
scores for physical functioning, vitality, mental health, and general health were transformed to a
0–100 scale (according to scoring and formulas in the SF-36 Health Survey Manual; Ware,
Kosinski, & Keller, 1997). Due to very non-normal distributions, 3 subscales were transformed
into categorical variables based on response distributions (i.e., role-physical on a 0–4 scale; social
functioning and role-emotional on 0–3 scales).
Telephone interview
Records of patients who completed clinic interviews were sent to the Charleston VA Medical
Center, where master’s level clinicians administered structured interviews via telephone.
Interviewers were all master’s level clinicians (psychologists and psychiatric nurse practitioners),
with prior clinical activities with trauma populations and/or involved in prior research projects
using identical measures to those included in this study. Interviewers underwent one training
session in the administration of clinical interviews from one of the authors (BCF). They also
received ongoing supervision. Inter-rater reliability analysis (kappa) was conducted for a random
sample of interviews (approximately 8%). Raters were 100% concordant for PTSD and SAD
diagnoses.
The trauma assessment for adults-interview version (TAA; Resnick, Kilpatrick, Dansky,
Saunders, & Best, 1993) assessed lifetime prevalence of trauma and has been widely used to screen
community and medical populations (Kilpatrick, Acierno, Saunders, Resnick, Best, & Schnurr,
2000; Resnick et al., 1993). This interview provided data to categorize patients as to whether they
met PTSD criterion A1.
The clinician-administered PTSD scale (CAPS; Blake et al., 1990) was administered to veterans
scoring positively on the TAA. The CAPS is a structured clinical interview that measures the
ARTICLE IN PRESS
238
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
intensity and frequency of DSM-IV PTSD symptoms (APA, 1994). The CAPS is considered the
‘‘gold standard’’ for assessing PTSD (Weathers, Keane, & Davidson, 2001).
The mini international neuropsychiatric interview (MINI; Sheehan et al., 1997) is a brief
structured interview that assesses the criteria for DSM-IV diagnoses and suicidality risk
(Lecrubier et al., 1997; Sheehan et al., 1997). The MINI exhibits similar sensitivity and specificity
to more time-intensive structured psychiatric interviews (e.g., Structured Clinical Interview for
DSM-IV, Composite International Diagnostic Interview) (Lecrubier et al., 1997; Sheehan et al.,
1997). We examined SAD and other anxiety, mood, and alcohol-related disorders.
Suicide risk was assessed with six differentially weighted MINI items. Participants were asked if
they engaged in any of the following in the past month (item weightings in parentheses): think
they would be better off dead or wish they were dead (1), want to harm themselves (2), think
about suicide (6), have a suicide plan (10), or attempt suicide (10). They were also asked about any
lifetime suicide attempts (4). Based on item weightings, suicidal risk was defined as low for scores
between 1 and 5, moderate for scores between 6 and 9, and high for scores X10.
Chart review
Electronic medical records for the 12 months preceding study participation were reviewed for
each consenting patient. The information recorded included medical and psychological diagnoses
and conditions. In the present study, we examined primary care providers’ recognition/
documentation of anxiety problems that require clinical attention as compared to the presence/
absence of SAD diagnoses based on MINI interviews. Anxiety problems were coded as ‘‘present’’
if any primary or specialty care progress notes indicated anxiety, synonyms of anxiety (e.g.,
nervous, jittery, stressed), or any DSM-IV anxiety disorder as a problem to be addressed. This
strategy represented a liberal effort to determine whether providers recognized, not only specific
anxiety disorders, but also any general concern about anxiety symptoms in these patients.
Assignment of PTSD diagnostic category (positive or negative)
PTSD diagnoses were based on results of structured phone interviews. Participants who scored
negative on either the TAA (failing to meet PTSD criterion A1) or the CAPS comprised the PTSD
( ) group. Corresponding with DSM-IV criteria, participants were designated as PTSD (+) if
they met criterion A1 on the TAA, criteria B, C, and D on the CAPS, and the duration of CAPS
symptoms was greater than 1 month (Blake et al., 1990).
Overview of analytic strategies
All tests were two-tailed. The primary data strategy involved examining relationships between
study variables and the presence/absence of SAD. We examined the prevalence rate of SAD in the
full sample and then conducted a chi-square analysis to examine the comorbidity of SAD and
PTSD. To examine correlates of SAD, t-tests were used for continuous outcomes (e.g., severity of
PTSD symptom clusters) and chi-square analyses for categorical outcomes (e.g., presence of
comorbid psychiatric conditions). To examine the unique effects of SAD on study outcomes, we
conducted linear and logistic regression models with SAD and PTSD severity (i.e., PCL-M total
ARTICLE IN PRESS
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
239
score) entered simultaneously. To examine whether SAD effects were a function of shared
variance with depression, we conducted secondary analyses statistically controlling for depression
(with a series of linear and logistic regression models).
Our final analyses focused on primary care providers’ recognition of anxiety problems. Using
chi-square analyses, we tested the agreement between SAD diagnoses and primary care providers’
recognition of (a) anxiety problems other than PTSD and (b) any anxiety problems (including
PTSD).
Results
Preliminary analyses
Comparisons were made between the final sample (n ¼ 733) and excluded participants
(n ¼ 152) on PTSD symptoms and clinical and socio-demographic variables. The only significant
difference was a higher rate of living with someone in the final sample (70.5–57.2%), w2 (1,
883) ¼ 57.38, po.001.
To examine the inter-rater reliability of PTSD and SAD diagnoses, interviews were conducted
by speakerphone for a random 8% (n ¼ 61) of the sample. Raters were 100% concordant on
PTSD and SAD diagnoses.
Sample characteristics
The average age of the final sample was 61.2 (711.9), with nearly 43% age 65 years or older
(ranging from 25.5 to 81.1 years); 93.3% was male; 62.7% was Caucasian; 22.1% had less than a
high school education; 48.5% reported serving in a war zone; and 32.9% were employed.
Prevalence of SAD and relations with PTSD
For the entire sample, 3.6% of veterans met criteria for SAD diagnoses. We found a strong
relationship between the likelihood of meeting criteria for PTSD and SAD diagnoses; w2 (1,
733) ¼ 98.13, po.001. Specifically, 73.1% of veterans with SAD had comorbid PTSD whereas
26.9% of veterans without SAD met criteria for PTSD. Alternatively, 22.1% of veterans with
PTSD had comorbid SAD whereas 1.1% of veterans without PTSD met criteria for SAD.
Relationships between the presence of SAD and PTSD severity are shown in Table 1. Using
interviews (i.e., CAPS) and self-reports (i.e., PCL-M), veterans with SAD exhibited higher scores
for each PTSD symptom cluster and total symptom severity (all p-valueso.001).
Relation between SAD and demographic and clinical characteristics
We examined clinical characteristics associated with SAD. There were no significant differences
between veterans with and without SAD in gender, ethnicity, marital status, education, or current
disability status. Veterans with SAD were significantly younger (X ¼ 55:4; SD ¼ 10:5) than
patients without SAD [SAD (+), X ¼ 55:4; SD ¼ 10:5; SAD ( ), X ¼ 61:4; SD ¼ 11:9];
ARTICLE IN PRESS
240
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
Table 1
Presence of social anxiety disorder and PTSD symptom severity
PTSD symptom severity
CAPS cluster B
CAPS cluster C—avoidance
CAPS cluster C—numbing
CAPS cluster D
CAPS total
PCL cluster B
PCL cluster C—avoidance
PCL cluster C—numbing
PCL cluster D
PCL cluster total
Group
t (728)
SAD (+) (mean7SD)
SAD ( ) (mean7SD)
19.00710.48
16.7876.60
13.7376.84
24.1976.68
75.88727.31
11.8576.44
10.4674.52
7.5073.48
14.4675.81
44.27717.99
8.1979.33
5.4376.82
4.5676.28
9.39710.45
25.85731.79
6.9874.14
5.9273.35
4.6372.81
8.5774.71
25.78713.35
5.58***
8.34***
7.29***
7.16***
7.92***
5.75***
6.69***
5.07***
6.21***
6.84***
Notes: * po.05, ** po.01, *** po.001. To assess PTSD severity on the CAPS, intensity and frequency symptom scores
were aggregated.
t (726) ¼ 2.55, p ¼ :01: A greater percentage of SAD (+) patients were unemployed (84.6%)
compared to SAD ( ) patients (61.5%); w2 (1, 730) ¼ 5.70, p ¼ :01: The majority of SAD (+)
patients were unemployed due to disability (81.8%) whereas the majority of SAD ( ) patients
were unemployed due to retirement (62.6%). Compared to SAD ( ) patients, SAD (+) patients
were significantly more likely to have had combat experience (80.8% to 49.4%, respectively; w2 (1,
724) ¼ 9.85, p ¼ :002) and witnessed someone seriously injured or violently killed (61.5% to
35.8%, respectively; w2 (1, 722) ¼ 7.16, p ¼ :01). Moreover, SAD (+) patients were significantly
more likely to pursue disability compensation compared to SAD ( ) patients (57.7% to 24.3%,
respectively; w2 (1, 730) ¼ 14.81, p ¼ :001).
Relation between SAD and psychological and physical health
Comorbid psychiatric conditions and suicidality risk: We examined the risk of psychiatric
disorder and suicidal risk associated with the presence of SAD. As shown in Table 2, SAD was
highly comorbid with other mental disorders. On average, SAD (+) patients had 4.04
(SD ¼ 1:78) comorbid mental disorders compared to .52 (SD ¼ 1:08) for SAD ( ) patients; t
(728) ¼ 15.89, po.001. As can be seen in Table 2, the three most common psychiatric conditions
found in SAD (+) patients were Major Depressive Disorder (MDD), Dysthymic Disorder, and
Generalized Anxiety Disorder (rates ranging from 50% to 69.2%); all p-values o.001.
Additionally, 53.8% of SAD (+) patients were at moderate to high suicidal risk; po.001.
Compared to SAD ( ) patients, SAD (+) patients were at greater risk for all psychiatric
conditions (OR’s ranged from 2.31 to 74.01; all p-values o.01) (the exception being alcohol
abuse).
Relation between SAD and health-related quality of life: We examined differences between SAD
(+) and SAD ( ) groups on quality-of-life dimensions (SF-36 subscales). Table 3 shows that the
ARTICLE IN PRESS
241
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
Table 2
Psychiatric comorbidity and suicidal risk associated with social anxiety disorder
Psychiatric conditiona
Diagnosis from MINI
SAD (+) n ¼ 26
w2 (1, 730)
OR (95% CI)b
7.94 (3.35, 18.84)
9.28 (3.88. 22.18)
74.01 (22.42, 244.35)
8.39 (3.75, 18.76)
SAD ( ) n ¼ 707
Anxiety disorders
Panic disorder
Agoraphobia
OCD
GAD
9
9
9
13
(34.6%)
(34.6%)
(34.6%)
(50.0%)
44
38
5
75
(6.3%)
(5.4%)
(.7%)
(10.7%)
29.96***
35.53***
153.24***
36.62***
Major depressive disorder
Dysthymic disorder
Alcohol abuse
Alcohol dependence
18
16
1
4
(69.2%)
(61.5%)
(3.8%)
(15.4%)
103
80
12
12
(14.6%)
(11.4%)
(1.7%)
(1.7%)
54.06***
55.27***
.66
17.38**
13.13 (5.56, 30.98)
12.48 (5.48, 28.44)
2.31 (.29, 18.44)
8.35 (2.56, 27.22)
Suicidal risk
14 (53.8%)
59.77***
13.24 (5.85, 29.98)
57 (8.1%)
Notes: * po.05, ** po.01, *** po.001.
a
Presence of psychiatric conditions is reported in descriptive statistics.
b
Odds of having given co-morbid condition for SAD (+) group compared to SAD ( ) group.
presence of SAD was associated with poor scores on all dimensions of psychological health; all pvalueso.01. In contrast, no significant relations were found with physical health indices.
Specificity of SAD effects
Statistically controlling for PTSD severity: Focusing on study outcomes with significant
relations to SAD, we examined the unique variance attributable to SAD after statistically
controlling for PTSD severity. PTSD severity was a significant predictor in all logistic regression
models; all p-valueso.01. After statistically controlling for PTSD severity, veterans with SAD
were still at significantly higher risk for comorbidity and higher suicidal risk compared to veterans
without SAD (ORs ranged from 2.92 for panic disorder to 34.58 for obsessive–compulsive
disorder; all p-valueso.05). As for clinical characteristics, veterans with SAD retained a greater
likelihood of combat experience (OR ¼ 2.75, 95% CI ¼ .99–7.63, po.05). No other clinical
characteristics or quality-of-life dimensions retained significant relations with SAD.
Statistically controlling for depression: Focusing on study outcomes with significant relations to
SAD, we examined the unique variance attributable to SAD after statistically controlling for the
presence of MDD. Using linear and logistic regression models, the presence/absence of SAD and
the presence/absence of MDD were entered simultaneously as predictors. Except for alcohol
abuse and dependence, MDD was a significant predictor in all linear and logistic regression
models; all p-valueso.01. After statistically controlling for the presence of MDD, veterans with
SAD still exhibited significantly higher rates of PTSD (OR ¼ 12.75, 95% CI ¼ 4.26–38.15,
po.001) and greater PTSD severity (i.e., each symptom cluster and total severity) compared to
veterans without SAD; all p-valueso.01 (the only exception was emotional numbing on the PCLM). After statistically controlling for the presence of MDD, veterans with SAD were still at
ARTICLE IN PRESS
242
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
Table 3
Relations between social anxiety disorder and quality-of-life dimensions
SF-36 subscales
Role-physicala
0
1
2
3
4
Role-emotionala
0
1
2
3
Social functioninga
0
1
2
3
Mental healthb
Vitalityb
General healthb
Physical functioningb
SAD (+) n ¼ 24
SAD ( ) n ¼ 670
Test of group differences
p ¼ .15
13
3
3
2
3
(54.2%)
(12.5%)
(12.5%)
(8.3%)
(12.5%)
219
88
62
74
227
(32.7%)
(13.1%)
(9.3%)
(11.0%)
(33.9%)
12
4
4
4
(50.0%)
(16.7%)
(16.7%)
(16.7%)
105
57
47
461
(15.6%)
(8.5%)
(7.0%)
(68.8%)
8
7
3
6
(33.3%)
(29.2%)
(12.5%)
(25.0%)
72
122
104
372
(10.8%)
(18.2%)
(15.5%)
(55.5%)
54.33
36.25
59.63
52.50
(21.2)
(24.6)
(12.2)
(29.4)
76.70
49.94
58.84
60.75
(20.8)
(25.0)
(10.9)
(29.2)
w2 (3, 694) ¼ 30.29,
po.001
w2 (3, 694) ¼ 15.85,
p ¼ :001
t (692) ¼ 5.18, po.001
t (692) ¼ 2.64, p ¼ .008
p ¼ :73
p ¼ :18
Notes: Due to missing data on the SF-36, the sample size for all analyses was 694. For all subscales, higher scores reflect
greater health and functioning.
a
Role-physical (assessing role limitations due to physical health) was on a 0–4 scale, role-emotional (assessing role
limitations due to emotional problems) and social functioning were on 0–3 scales, and the raw count and percentage of
veterans reporting each score are shown.
b
Continuous raw scores were transformed to a 0–100 scale and means and standard deviations are shown.
significantly higher risk for comorbidity and higher suicidal risk compared to veterans without
SAD (ORs ranged from 2.27 for panic disorder to 22.25 for obsessive–compulsive disorder; all pvalueso.05). As for clinical characteristics and quality-of-life dimensions, none retained
significant relationships with SAD after controlling for MDD. Evidence suggests that the effects
of SAD on psychiatric comorbidity and suicidal risk were not attributable to the presence of
MDD.
Concordance between SAD diagnoses and assessments by primary care providers
We examined primary care providers’ recognition of anxiety difficulties in SAD (+) and SAD
( ) patients. As shown in Table 4, for SAD (+) patients, 42.3% (11/26) were recognized as
having anxiety problems above and beyond PTSD, and 57.7% (15/26) were recognized as having
any type of anxiety difficulty (including PTSD); p-valueso.001. Thus, primary care providers
exhibited less than adequate recognition of anxiety problems.
ARTICLE IN PRESS
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
243
Table 4
Concordance between social anxiety disorder diagnoses and anxiety problems assessed by primary care providers
Based on primary care clinical charts
Social anxiety disorder diagnosis from MINI
Test of group differences
SAD (+) (n ¼ 26)
SAD ( ) (n ¼ 704)
Diagnosis of Anxietya
Yes
No
11 (42.3%)
15 (57.7%)
67 (9.5%)
637 (90.5%)
w2 (1, 730) ¼ 28.25,
po:001
Diagnosis of Anxiety or PTSD
Yes
No
15 (57.7%)
11 (42.3%)
94 (13.4%)
610 (86.6%)
w2 (1, 730) ¼ 38.81,
po:0001
Notes: aPrimary care physician ratings of clinically significant anxiety refers to concerns over and above PTSD-related
problems.
Discussion
Prior studies of SAD in veterans have relied on small samples and examined a limited number
of variables. Using a large multi-site sample of veterans in primary care clinics, 3.6% of the
veterans met criteria for SAD and veterans with PTSD had a greater rate of SAD compared to
veterans without PTSD (22.0% vs. 1.1%). These prevalence rates are comparable to those
reported in other studies (Hofmann et al., 2003; Orsillo et al., 1996). The presence of SAD also
was associated with greater PTSD severity for each of the DSM-IV symptom clusters. Evidence
suggested that the adverse effects of SAD on psychiatric comorbidity and suicidal risk were not
attributable to PTSD severity or the presence of MDD. In other findings, primary care providers
exhibited a poor rate of detecting anxiety problems in veterans diagnosed with SAD.
The elevated psychiatric comorbidity and suicidal risk associated with SAD were not
attributable to PTSD severity. Consistent with prior work (e.g., Brown, Chorpita, & Barlow,
1998; Kessler et al., 1994), we found MDD to be the most highly comorbid psychiatric condition
associated with SAD in veterans. Also similar to non-veteran samples (Magee et al., 1996;
Schneier et al., 1992), SAD was associated with alcohol dependence, but not alcohol abuse. One
interpretation is that veterans with SAD fall at two extremes, frequently consuming large
quantities of alcohol that inevitably lead to dependence, or altogether avoiding excessive alcohol
use. The temporal sequence among SAD, avoidance behaviors, and substance abuse requires
further investigation in veterans.
Both SAD and MDD share core features such as high negative affectivity and low positive
affectivity (Brown et al., 1998; Kashdan, 2002, 2004) and interpersonal sensitivity is a primary
symptom of atypical depression (APA, 1994). These data suggest the need for specificity tests in
examinations of either condition. After controlling for the presence of MDD, SAD was uniquely
associated with an elevated rate of PTSD diagnoses and other psychiatric conditions, greater
PTSD severity, and elevated suicidal risk. Recent data suggest that after controlling for depressive
symptoms, differences in social anxiety among veterans with and without PTSD, and healthy
controls disappears (Hofmann et al., 2003). Using a larger sample than Hofmann et al., and
diagnoses as opposed to dimensional ratings, we found evidence for unique SAD effects. Over
ARTICLE IN PRESS
244
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
two-thirds of veterans in our sample with SAD also met criteria for MDD. Thus, our specificity
tests should be considered conservative.
In an examination of over 1000 adult outpatients, the majority of clients with a current
principal diagnosis of SAD had an additional Axis I diagnosis (72%) (Brown, Campbell, Lehman,
Grisham, & Mancill, 2001). In our study, 73.1% of veterans with SAD met criteria for an
additional Axis I diagnosis. Thus, our data fit with prior work suggesting that only a small
percentage of adults meet criteria for SAD without the presence of other psychiatric conditions.
Other data suggest that SAD has an earlier average age of onset than PTSD (15.7 vs. 21.5) and
there is a greater frequency of cases with SAD preceding the onset of comorbid anxiety and mood
disorders (68% and 85%, respectively) (Brown et al., 2001). More research is needed on the
temporal sequence of these psychiatric conditions and whether the presence of SAD, MDD, and
PTSD in veterans reflects multiple conditions or manifestations of a single condition. Taxometric
analyses support the dimensional nature of social anxiety and depression (e.g., Ruscio & Ruscio,
2002). The use of dimensional measures may be a more accurate reflection of reality than the
potentially artificial, diagnostic boundaries inherent in categorical diagnoses.
We found an alarmingly low rate of anxiety problem detection by primary care physicians.
Primary care provider ratings were liberal as anxiety problems were not specific to SAD and could
have reflected difficulties stemming from any anxiety disorder or anxiety-related condition.
Although physicians’ tended not to diagnose SAD given the absence of SAD (i.e., good
specificity), there is reason to believe that physicians missed other anxiety disorders in this sample.
Prior data find that approximately 46% of veterans with PTSD and 27% of veterans without
PTSD meet criteria for at least one current anxiety disorder (Orsillo et al., 1996). In our study, for
those veterans without SAD, physicians diagnosed anxiety problems in only 9.5% of the sample.
Overall, primary care providers appear to be underidentifying and thus, undertreating anxiety
disorders in veterans.
All of the limitations associated with cross-sectional examinations are relevant to the present
study (e.g., retrospective reporting biases, inability to address temporality and causality). Our
multi-site, large-scale sampling precluded the ability to conduct the extensive anxiety disorder
assessments used in specialty clinics. As the next step, there is great merit in conducting multimethod assessments of social anxiety before and after military deployment to gauge the causal
impact of social anxiety on psychological functioning. In generalizing from the present data, the
average age of our sample (455 years old) and the use of patients who utilize the VA health care
system should be taken into consideration. Nonetheless, the majority of our results fit with prior
work on SAD and PTSD using other veteran, community, and college student samples.
Acknowledgements
This work was partially supported by Grant VCR-99-010-2 from the Veterans Affairs Health
Services Research and Development program to Dr. Magruder, and fellowships from the
National Institute of Mental Health (F31-MH63565) and the Anxiety Disorder Association of
America to Todd B. Kashdan. This work also was supported by the Office of Research and
Development, Medical Research Service, Department of Veterans Affairs. We thank Ken
Ruggerio for his feedback on earlier versions.
ARTICLE IN PRESS
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
245
References
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington,
DC.
Baumeister, R., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental
human motivation. Psychological Bulletin, 117, 497–529.
Blake, D. D., Weathers, F. W., Nagy, L. N., Kaloupek, D. G., Klauminzer, G., Charney, D. S., & Keane, T. M. (1990).
A clinician rating scale for assessing current and lifetime PTSD: The CAPS-1. The Behavior Therapist, 18,
187–188.
Blanchard, E. B., Jones, A. J., Buckley, T. C., & Forneris, C. A. (1996). Psychometric properties of the PTSD Checklist
(PCL). Behaviour Research Therapy, 34, 669–673.
Brown, T. A., Campbell, L. A., Lehman, C. L., Grisham, J. R., & Mancill, R. B. (2001). Current and lifetime
comorbidity of the DSM-IV anxiety and mood disorders in a large clinical sample. Journal of Abnormal Psychology,
110, 585–599.
Brown, T. A., Chorpita, B. F., & Barlow, D. H. (1998). Structural relationships among dimensions of the DSM-IV
anxiety and mood disorders and dimensions of negative affect, positive affect, and autonomic arousal. Journal of
Abnormal Psychology, 107, 179–192.
Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. In R. G. Heimberg, M. R. Liebowitz, D. A.
Hope, & F. R. Schneier (Eds.), Social phobia: Diagnosis, assessment, and treatment (pp. 69–93). New York: Guilford
Press.
Crowson, J. J., Frueh, B. C., Beidel, D. C., & Turner, S. M. (1998). Self-reported symptoms of social
anxiety in a sample of combat veterans with posttraumatic stress disorder. Journal of Anxiety Disorders, 12,
605–612.
Davidson, R. T., Hughes, D. C., George, L. K., & Blazer, D. G. (1994). The boundary of social phobia: Exploring the
threshold. Archives of General Psychiatry, 51, 975–983.
Fifer, S. K., Mathias, S. D., Patrick, D. L., Mazaonson, P. D., Lubeck, D. P., & Buesching, D. P. (1994). Untreated
anxiety among adult primary care patients in a health maintenance organization. Archives of General Psychiatry, 51,
740–750.
Frueh, B. C., Turner, S. M., Beidel, D. C., & Cahill, S. P. (2001). Assessment of social functioning in combat veterans
with PTSD. Aggression and Violent Behavior, 6, 79–90.
Frueh, B. C., Turner, S. M., Beidel, D. C., Mirabella, R. F., & Jones, W. J. (1996). Trauma Management Therapy: A
preliminary evaluation of a multicomponent behavioral treatment for chronic combat-related PTSD. Behaviour
Research and Therapy, 6, 533–543.
Heimberg, R. G., Stein, M. B., Hiripi, E., & Kessler, R. C. (2000). Trends in the prevalence of social phobia in the
United States: A synthetic cohort analysis of changes over four decades. European Psychiatry, 15, 29–37.
Herman, J. L. (1992). Trauma and recovery. Boston: Basic Books.
Hofmann, S. G., Litz, B. T., & Weathers, F. W. (2003). Social anxiety, depression, and PTSD in Vietnam veterans.
Journal of Anxiety Disorders, 17, 573–582.
Jordan, B. K., Marmar, C. R., Fairbank, J. A., Schlenger, W. E., Kulka, R. A., Hough, R. L., & Weiss, D. S. (1992).
Problems in families of male Vietnam veterans with posttraumatic stress disorder. Journal of Consulting and Clinical
Psychology, 60, 916–926.
Kashdan, T. B. (2002). Social anxiety dimensions, neuroticism, and the contours of positive psychological functioning.
Cognitive Therapy and Research, 26, 789–810.
Kashdan, T. B. (2004). The neglected relationship between social interaction anxiety and hedonic deficits:
Differentiation from depressive symptoms. Journal of Anxiety Disorders, 18, 719–730.
Kessler, R. C., McGonagle, K. A., Zhao, S., Nelson, C. B., Hughes, M., Eshelman, S., Wittchen, H., & Kendler, K. S.
(1994). Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Archives of
General Psychiatry, 51, 8–19.
Kilpatrick, D. G., Acierno, R., Saunders, B., Resnick, H. S., Best, C. L., & Schnurr, P. P. (2000). Risk factors for
adolescent substance abuse and dependence: Data from a national sample. Journal of Consulting and Clinical
Psychology, 68, 19–30.
ARTICLE IN PRESS
246
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
King, L. A., King, D. W., Fairbank, J. A., Keane, T. M., & Adams, G. A. (1998). Resilience-recovery factors in posttraumatic stress disorder among female and male Vietnam veterans: Hardiness, postwar social support, and
additional stressful life events. Journal of Personality and Social Psychology, 74, 420–434.
Kirmayer, L. J., Robbins, J. M., Dworkind, M., & Yaffe, M. J. (1993). Somatization and the recognition of depression
and anxiety in primary care. American Journal of Psychiatry, 150, 734–741.
Lecrubier, Y., Sheehan, D. V., Weiller, E., Amorim, P., Bonora, I., Sheehan, K. H., Janavs, J., & Dunbar, G. C. (1997).
The Mini International Neuropsychiatric Interview (MINI). A short diagnostic structured interview: Reliability and
validity according to the CIDI. European Psychiatry, 12, 224–231.
Magee, W. J., Eaton, W. W., Wittchen, H., McGonagle, K. A., & Kessler, R. C. (1996). Agoraphobia,
simple phobia, and social phobia in the national comorbidity study. Archives of General Psychiatry, 53,
159–168.
Magruder, K., Frueh, B. C., Knapp, R. G., Johnson, M. R., Vaughan, J. A., Carson, T. C., Powell, D. A., & Hebert, R.
(2004). PTSD symptoms, demographic characteristics, and functional status among veterans treated in VA primary
care clinics. Journal of Traumatic Stress, 17, 293–301.
Ormel, J., Koeter, M. W., van den Brink, W., & van de Willige, G. (1991). Recognition, management, and course of
anxiety and depression in general practice. Archives of General Psychiatry, 48, 700–706.
Orsillo, S. M., Heimberg, R. G., Juster, H. R., & Garrett, J. (1996). Social phobia and PTSD in Vietnam Veterans.
Journal of Traumatic Stress, 9, 235–252.
Orsillo, S. M., Weathers, F. W., Litz, B. T., Steinberg, H. R., Huska, J. A., & Keane, T. M. (1996). Current and lifetime
psychiatric disorders among veterans with war-zone-related posttraumatic stress disorder. Journal of Nervous and
Mental Disease, 184, 307–313.
Rapee, R. M., & Heimberg, R. G. (1997). A cognitive–behavioral model of anxiety in social phobia. Behaviour Research
and Therapy, 35, 741–756.
Resnick, H. S., Kilpatrick, D. G., Dansky, B. S., Saunders, B. E., & Best, C. L. (1993). Prevalence of civilian trauma
and posttraumatic stress disorder in a representative national sample of women. Journal of Consulting and Clinical
Psychology, 61, 984–991.
Richardson, R. D., Engel, C. C., Jr., Hunt, S. C., McKnight, K., & McFall, M. (2002). Are veterans seeking veterans
affairs’ primary care as healthy as those seeking department of defense primary care? A look at gulf war veterans’
symptoms and functional status. Psychosomatic Medicine, 64, 676–683.
Riggs, D. S., Byrne, C. A., Weathers, F. W., & Litz, B. T. (1998). The quality of the intimate relationships of male
vietnam war veterans: Problems associated with posttraumatic stress disorder. Journal of Traumatic Stress, 11,
87–101.
Ruscio, A. M., & Ruscio, J. (2002). The latent structure of analogue depression: Should the BDI be used to classify
groups? Psychological Assessment, 14(2), 135–145.
Ryff, C. D., & Singer, B. (2000). Interpersonal flourishing: A positive health agenda for the new millennium. Personality
and Social Psychology Review, 4, 30–44.
Schneier, F. R., Heckelman, L. R., Garfinkel, R., Campeas, R., Fallon, B. A., Gitow, A., Street, L., Bene, D. D.,
& Liebowitz, M. R. (1994). Functional impairment in social phobia. Journal of Clinical Psychiatry, 55,
322–331.
Schneier, F. R., Johnson, J., Hornig, C. D., Liebowitz, M. R., & Weissman, M. (1992). Social phobia: Comorbidity and
morbidity in an epidemiologic sample. Archives of General Psychiatry, 49, 282–288.
Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Janavs, J., Weiller, E., Keskiner, A., Schinka, J., Knapp, E., Sheehan,
M. F., & Dunbar, G. C. (1997). The validity of the Mini International Neuropsychiatric Interview (MINI) according
to the SCID-P and its reliability. European Psychiatry, 12, 232–241.
Solomon, Z., Mikulincer, M., & Avitzur, E. (1988). Coping, locus of control, social support, and combat-related
posttraumatic stress disorder: A prospective study. Journal of Personality and Social Psychology, 55,
279–285.
Ware, J. E., Kosinski, M., & Keller, S. D. (1997). SF-36 health survey manual and interpretation guide. Boston: The
Health Institute, New England Medical Center.
Ware, J. E., & Sherbourne, C. D. (1992). The MOS 36-item Short Form Health Survey (SF-36). I. Conceptual
framework and item selection. Medical Care, 30, 473–483.
ARTICLE IN PRESS
T.B. Kashdan et al. / Behaviour Research and Therapy 44 (2006) 233–247
247
Weathers, F. W., Keane, T. M., & Davidson, J. R. T. (2001). Clinician-Administered PTSD Scale: A review of the first
ten years of research. Depression and Anxiety, 13, 132–156.
Weathers, F. W., Litz, B. T., Herman, D. S., Huska, J. A., Keane, T. M. (1993). The PTSD Checklist (PCL):
Reliability, validity, and diagnostic utility. Paper presented at the 9th annual meeting of the ISTSS, San Antonio,
TX.
Wittchen, H. U., Fuetsch, M., Sonntag, H., Mueller, N., & Liebowitz, M. (2000). Disability and quality of life in pure
and comorbid social phobia: Findings from a controlled study. European Psychiatry, 15, 46–58.