Behaviour Research and Therapy 94 (2017) 9e18

Contents lists available at ScienceDirect

Behaviour Research and Therapy

journal homepage: www.elsevier.com/locate/brat

Feeling safe but appearing anxious: Differential effects of alcohol on

anxiety and social performance in individuals with social anxiety
disorder
Stephan Stevens, PhD a, *, Ruth Cooper a, Trisha Bantin b, Christiane Hermann b,
Alexander L. Gerlach a
a
University of Cologne, Cologne, Germany
b
Justus Liebig University Giessen, Giessen, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Social anxiety disorder (SAD) and alcohol use disorders (AUD) co-occur frequently and there is pre-
Received 29 August 2016 liminary evidence that alcohol might reduce social anxiety. It is, however, unclear which mechanisms
Received in revised form contribute to the anxiety reducing effect, particularly regarding key aspects of social anxiety such as
7 March 2017
deficits in social performance. We compared self-rated and physiological measures of anxiety as well as
Accepted 17 April 2017
Available online 18 April 2017
self- and observer-rated social performance in a sample of 62 individuals with SAD and 60 nonanxious
control participants during a speech task after receiving either alcohol, an alcohol-free placebo drink or
orange juice. SAD patients reported more anxiety during the speech task than did control participants.
Keywords:
Social anxiety disorder
Furthermore, SAD patients underestimated their performance in comparison to observer ratings. Alcohol
Comorbidity reduced self-report anxiety only in SAD patients, while observers rated all participants as less competent
Social performance when intoxicated. Although individuals with SAD experience a reduction in anxiety when drinking
Alcohol consumption alcohol, simultaneous decreases in social performance might contribute to negative reactions from
others and consequently increase the risk of further alcohol use to cope with these negative reactions.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction unanimously confirmed that they had previously used alcohol to

Social anxiety disorder (SAD) and alcohol use disorders (AUD)
are highly comorbid (Davidson, Hughes, George, & Blazer, 1993;
Kessler, Chiu, Demler, & Walters, 2005). Moreover, both subclini-
cal (Crum & Pratt, 2001) and clinical (Kushner, Sher, & Beitman,
1990) social anxiety often precedes the onset of pathological
drinking patterns. This heightened comorbidity is not surprising,
given that alcohol is often used as a coping mechanism for social
fears: more than half of individuals with SAD report to “often or
always” use alcohol to attenuate symptoms of SAD (Buckner &
Heimberg, 2010; Thomas, Randall, & Carrigan, 2003). Similarly,
the motive to drink alcohol to cope with anxiety in social situations
is higher in socially anxious individuals compared to non-anxious
controls (Cludius, Stevens, Bantin, Gerlach, & Hermann, 2013). In-
dividuals with SAD and comorbid alcohol use disorder
* Corresponding author. Department of Clinical Psychology, University of Co-
logne, Pohligstraße 1, 50969 Ko€ln Germany.
E-mail address: stephan.stevens@uni-koeln.de (S. Stevens).
cope with anticipatory anxiety regarding social situations (Randall,
2000). Social anxiety among students undergoing a brief alcohol
intervention has been related to poorer outcomes (Terlecki,
Buckner, Larimer, & Copeland, 2011), which highlights the need
to understand anxiety-specific consequences of present alcohol
use.
Such findings are often conceptualized within the framework of
general theories of alcohol's stress-dampening effects, for example
the “self-medication hypothesis” (Chutuape & Dewit, 1995;
Quitkin, Rifkin, Kaplan, & Klein, 1972), the “attention allocation
model” (Josephs & Steele, 1990), or the “appraisal disruption
model” (Sayette, 1993). Experimentally, the effects of alcohol on
social anxiety were previously assessed in speech tasks, yielding
mixed results regarding the anxiolytic effect of alcohol: While some
studies found an anxiolytic effect (Abrams, Kushner, Medina, &
Voight, 2001; Himle et al., 1999; Kidorf & Lang, 1999), others
failed (Himle et al., 1999; Keane & Lisman, 1980; Naftolowitz,
Vaughn, Ranc, & Tancer, 1994) or underlined the role of alcohol
expectancies and placebo effects (Abrams & Kushner, 2004).

http://dx.doi.org/10.1016/j.brat.2017.04.008
0005-7967/© 2017 Elsevier Ltd. All rights reserved.
10 S. Stevens et al. / Behaviour Research and Therapy 94 (2017) 9e18
More explicit social anxiety-relevant models were only recently
established and empirical support for them is still being developed
(Buckner, Heimberg, Ecker, & Vinci, 2013). These models suggest
that when explaining the complex relationship between social
anxiety and alcohol use disorders, not only the level of negative
affect due to social anxiety is relevant, but also other pivotal
components such as low positive affect, fear of scrutiny, and social
avoidance. In line with the idea that specific components of social
anxiety contribute to the risk of alcohol related problems is the
“avoidance coping” model (Bacon & Ham, 2010), which states that
the reduction of attentional biases through alcohol use increases
the risk for alcohol dependency in individuals with social anxiety.
Indeed, this is underscored by the fact that alcohol reduces atten-
tional bias toward threat as measured with attentional probe tasks
(Gerlach, Schiller, Wild, & Rist, 2006; Stevens, Rist, & Gerlach,
2009). To summarize, the effects of alcohol on state social anxiety
in individuals with SAD are much more complex than a simple
direct relation (more alcohol e less anxiety) and explicitly should
consider cognitive processes like expectancy effects, attentional
biases, drinking motives and comorbid depression.
As suggested by the above-mentioned models, there is potential
value in evaluating the components of social anxiety that may be
influenced by the use of alcohol and which may contribute to the
etiology and/or maintenance of SAD. In addition to biased cognitive
processing of social situations, self-focused attention in social sit-
uations, self-awareness and safety behaviors, cognitive models of
SAD (i.e. Clark & Wells, 1995) emphasize the role of social perfor-
mance deficits in the maintenance of social anxiety. Whereas pa-
tients usually possess adequate skills to perform in social situations,
their high level of state anxiety in feared social situations, or actual
physiological arousal, often interferes with situational demands
(anxiety inhibition) and may lead to a significant decrease in social
performance (Hofmann, Gerlach, Wender, & Roth, 1997). Support-
ing this notion, individuals with SAD rate themselves as performing
worse compared to healthy controls, expect to be rated more
negatively by others and observers rate the performance of in-
dividuals with SAD as worse compared to non-anxious controls
(Baker & Edelmann, 2002; Norton & Hope, 2001; Stevens et al.,
2010; Voncken & Bogels, 2008). Furthermore, individuals with
SAD even underestimate their actual performance when compared
to observer performance ratings in speech (Rapee & Lim, 1992) and
interaction situations (Stopa & Clark, 1993). Against this back-
ground it is not surprising that the probability of negative reactions
from others in social situations is increased in individuals with SAD
(Alden & Taylor, 2004). In a study by Keane and Lisman (1980), the
authors investigated the effects of alcohol and alcohol expectancy
on social anxiety in males. They did not find an anxiolytic effect of
alcohol or alcohol expectancy on self-reported anxiety. On the
contrary, alcohol had detrimental effects on social performance.
However, in this landmark study social skills were only assessed
regarding verbal responses (i.e. total speaking time and how often a
test subject asked a question). Self-ratings of social performance
were not measured. Against this background, the aim of the current
study is to explicitly disentangle the effects of alcohol on social
anxiety and social performance to better understand the height-
ened comorbidity between the two disorders.
In summary, it is not well-known if and how alcohol influences
social performance, and how a possible core feature of social anx-
iety, i.e. decreased social performance, is influenced by alcohol
intake in both healthy controls and socially anxious individuals.
Thus, the aims of the present study were to explore the acute effects
of alcohol during a speech task on social performance in individuals
with SAD and non-anxious control participants. We additionally
assessed self-reported anxiety and physiological arousal.
As research on a state social anxiety reducing effect of alcohol
revealed a complex pattern of results (see Battista, Stewart, & Ham,
2010), we again examined the influence of alcohol in a common
social speech task. Although the influence of alcohol on self-
awareness in individuals with SAD has been shown, evidence for
its effects on social performance is quite rare. Based on the social
anxiety literature, we predicted that sober individuals (i.e. persons
in the orange juice condition) with SAD would judge their social
performance as worse compared to controls and blind raters (e.g.,
Voncken & Bogels, 2008). If anxiety inhibits performance, de-
creases in anxiety in individuals with SAD in the alcohol group may
lead to increases in self-rated and perhaps even observer rated
social performance. While remembering that the relation between
social anxiety and alcohol is complex and additionally based on
expectancy effects, drinking motives and attentional processes
(Battista et al., 2010), we nonetheless hypothesize that social per-
formance would increase in individuals with SAD after ingesting
alcohol. Finally, we expected that participants with SAD receiving
placebo would benefit less as compared to those receiving alcohol,
as a pharmacological effect on anxiety should be more profound
than a pure expectancy based effect.
2. Method
2.1. Participants
Participants suffering from social anxiety were invited to the
experiment through newspaper ads, flyers, and emails (sent to all
students of the University of Giessen) that specifically targeted
individuals who feel insecure in social situations. Participants in the
SAD group could choose between monetary compensation (20 V
for the OJ condition, 30 V for the ALC and PLA conditions) or a 2-
hour counseling session concerning their social fears and treat-
ment options. Participants were included if they fulfilled the DSM-
IV criteria for a primary SAD diagnosis. The diagnosis of a substance
use (including alcohol use disorders), bipolar, or psychotic disorder
resulted in exclusion from the study. Further general exclusion
criteria included suffering from coronary heart diseases; undergo-
ing current psychiatric, neurological, or psychological treatment;
and the current intake of drugs affecting the central nervous system
or the cardiovascular system. Due to possible alcohol intake,
pregnant or breast-feeding female participants were excluded. In
addition, participants who were alcohol naïve (i. e., reported to
have never consumed alcohol) or reported a history of were
excluded. To recruit control participants (CONT), flyers and emails
were sent out that targeted people who do not feel anxious in social
situations. Control participants were compensated monetarily as
explained above. In addition to the general exclusion criteria, spe-
cific exclusion criteria for control subjects consisted of a diagnosis
of any current mental disorder.
Initially, we screened 99 individuals for the SAD and 78 in-
dividuals for the control group. Sixty-two SAD participants and 60
healthy control participants (CONT) participants were included. 19
individuals with SAD were excluded as they failed to fulfill DSM-IV
criteria for SAD, 10 refused to drink alcohol during the experiment
and 10 had additional diagnoses (seven substance use, two bipolar
disorder and one somatic liver disease). In the control group, 10
participants fulfilled criteria for a DSM-IV diagnosis, six refused to
drink alcohol during the experiments and 3 moved after the
completion of the initial screening. Detailed sociodemographic
information is presented in Table 1.
2.2. Measures
2.2.1. Demographics questionnaire
Participants were asked to self-report age, gender, education,
 S. Stevens et al. / Behaviour Research and Therapy 94 (2017) 9e18 11

Table 1
Sociodemographic and questionnaire data of the sample.

Individuals with SAD Controls p

Gender Male: 21 Male: 20 n.s.

Female: 41 Female: 40
Education University: 16 University: 20 n.s.
High school: 41 High school: 36
Lower education: 5 Lower education: 4
Marital status Single: 39 Single: 30 n.s.
Relationship: 13 Relationship: 18
Married: 8 Married: 9
Divorced: 2 Divorced: 3
Comorbid Diagnoses MDE: 40% n.a.
Other Anxiety: 33%
Somatoform: 25%
OCD: 1.6%
Eating Disorder: 1.6%
Total: 66%

M (SD) M (SD)

Age 30.5 (11.8) 30.5 (11.4) p ¼ 0.46

BDI 12.1 (7.0) 2.4 (2.7) p < 0.001
SPS 36.8 (11.3) 5.1 (4.8) p < 0.001
SIAS 50.3 (10.8) 10.2 (5.9) p < 0.001
LSAS-F 38.8 (12.1) 10.9 (8.0) p < 0.001
LSAS-A 32.5 (11.8) 10.0 (8.8) p < 0.001
AUDIT 5.9 (3.7) 4.5 (3.0) p < 0.05
DDSAQ 29.2 (20.1) 6.1 (7.3) p < 0.001

Note: SAD ¼ Social Anxiety Disorder; MDE: Major Depressive Episode (current and past); OCD: Obsessive-Compulsive Disorder. BDI ¼ Beck Depression Inventory; SPS ¼ Social
Phobia Scale; SIAS ¼ Social Interaction Anxiety Scale; LSAS-F ¼ Liebowitz Social Anxiety Scale e Fear Subscale; LSAS-A ¼ Liebowitz Social Anxiety Scale e Avoidance Subscale;
AUDIT ¼ Alcohol Use Disorders Identification Test; DDSAQ ¼ Drinking Due to Social Anxiety Questionnaire.

and marital status on a standard demographic questionnaire.
2.2.2. Beck Depression Inventory (BDI)
As depression frequently co-occurs with SAD, we added a
measure of depressive symptoms to our assessment battery. The
BDI (Beck, Erbaugh, Ward, Mock, & Mendelsohn, 1961; German
version:; Hautzinger, Bailer, Worall, & Keller, 1994) is a 21-item
measure of depression symptoms. The German version reliably
assesses depressive symptoms (Cronbach's a ¼ 0.92; Hautzinger
et al., 1994). Additionally, it evidences favorable convergent val-
idity with other measures of depression (0.72 - 0.89, Hautzinger
et al., 1994). In the current sample, it evidenced good internal
consistency (Cronbach's a ¼ 0.90). Retest-reliability has been
investigated only for the English version, showing satisfactory to
good stability for a retest interval of 1e2 weeks (r ¼ 0.74 - 0.96).
2.2.3. Social Phobia Scale (SPS) and Social Interaction Anxiety Scale
(SIAS)
To assess interaction as well as performance anxiety as pivotal
aspects of SAD, two frequently used self-report measures of social
anxiety were included in the study. The SPS and SIAS (Mattick &
Clarke, 1998; German version:; Stangier, Heidenreich, Berardi,
Golbs, & Hoyer, 1999) measure fear in performance situations and
fear in interactional social situations, respectively. Both scales
consist of 20 items. The German versions of the scales evidence
acceptable internal consistency (SPS: a ¼ 0.82; SIAS: a ¼ 0.77;
Stangier et al., 1999), good retest reliability among 3 weeks (SPS:
r ¼ 0.96; SIAS: r ¼ 0.93; Stangier et al., 1999) and adequate
construct validity (Eidecker, Glockner-Rist, & Gerlach, 2010). In the
current sample, Cronbach's as were 0.95 for the SPS scale and 0.96
for the SIAS scale.
2.2.4. Liebowitz Social Anxiety Scale e self-report version (LSAS-SR)
The Liebowitz Social Anxiety Scale was also specially developed
in order to assess anxiety in both performance and social interac-
tion situations. The LSAS (Liebowitz, 1987; German version:;
Stangier & Heidenreich, 2003) assesses fear (LSAS-F) and avoidance
(LSAS-A) of 11 common social performance and interaction situa-
tions. The LSAS was originally developed as a clinician-
administered scale, but Fresco et al. (2001) found high compara-
bility between the self-report and the clinician-administered
version. Stangier and Heidenreich (2003) reported adequate over-
all internal consistency for the German version (a ¼ 0.91) and good
support for convergent validity, as well as a two-factor structure
with acceptable internal consistencies for both factors (fear:
a ¼ 0.82; avoidance: a ¼ 0.80). In the current sample, Cronbach's as
were 0.96 for the fear subscale and 0.94 for the avoidance subscale.
2.2.5. Alcohol Use Disorders Identification Test (AUDIT)
The AUDIT was included as a measure to assess problematic
drinking. Given the focus of this study, the propensity for prob-
lematic alcohol use is most relevant. The AUDIT (Saunders, Aasland,
Babor, Delafuente, & Grant, 1993; German version: Rist, Glockner-
Rist, & Demmel, 2009) was developed as a method of screening
for excessive drinking behavior. It consists of 10 items and the
German version has an acceptable internal consistency (a ¼ 0.77;
Rist et al., 2009) and a good retest-reliability (r ¼ 0.95 across 31
days; Dybek et al., 2006) and convergent validity with other mea-
sures of drinking behavior and biological markers of alcoholism
(Allen, Litten, Fertig, & Babor, 1997). In the current sample, Cron-
bach's a was 0.80.
2.2.6. Drinking Due to Social Anxiety Questionnaire (DDSAQ)
As drinking motives have been previously shown to highly vary
between individuals with SAD, we decided to additionally include a
measure that directly assesses drinking motives due to social
anxiety. The DDSAQ (Wagner, Stangier, Heidenreich, & Schneider,
2004) is a 28-item measure assessing the motive to drink alcohol
in order to relieve social fears. Although originally developed based
on data from patients with alcohol use disorders, Stevens and
Gerlach (2009) reported very good internal consistency (a ¼ 0.96)
and good convergent validity with measures of alcohol
12 S. Stevens et al. / Behaviour Research and Therapy 94 (2017) 9e18
expectancies, amount, and frequency of alcohol use (r ¼ 0.42 - 0.52)
in a sample of SAD patients. In the current sample, Cronbach's a
was 0.96. Unfortunately, no indices of retest-reliability have yet
been reported.
2.2.7. Social Performance Rating Scale (SPRS)
The SPRS (German version: Fydrich & Bürgener, 1999; Fydrich,
Chambless, Perry, Buergener, & Beazley, 1998) is a standardized
rating system for social performance in social situations. Gaze di-
rection, voice quality, length of individual talking units (mono-
syllabic/overly long responses preventing discourse vs. adequate
responses encouraging discourse), discomfort (agitation and
nervousness), and conversation flow are rated on five 5-point
Likert-type scales (1 ¼ very good; 5 ¼ very poor), with higher
scores indicating poorer social performance. The total score was
calculated as the summed score of the five items. Each participant's
performance was rated by him/herself (Cronbach's a ¼ 0.86) as well
as by two observers (Cronbach's as ¼ 0.83 and 0.80) who were
blind to the participant's diagnosis and assigned condition. The
inter-rater reliability of the two observers' ratings was r ¼ 0.89.
Observers were master level students of clinical psychology, who
were trained two days to use the SPRS rating system in standard-
ized test videos to reach an interrater reliability of at least ICC ¼ 0.8
in the ratings of the test videos. They were blind to the diagnoses of
the participants.
2.2.8. Visual analogue scales (VASs)
Participants were asked to repeatedly indicate their level of
anxiety, tension, and nervousness (e.g., “How anxious do you feel
right now?”) on three 10 cm VASs ranging from 0 (“not at all”) to
100 (“extremely”).
2.3. Interview measures
2.3.1. Structured Clinical Interview for DSM-IV (SCID)
The SCID (First, Spitzer, Gibbon, & Williams, 1996; German
version:; Wittchen, Wunderlich, Gruschwitz, & Zaudig, 1997) is a
semi-structured interview based on DSM-IV criteria. Two masters-
level clinicians administered the SCID interviews and were trained
and supervised in administering the SCID by the senior author of
the study. Interrater reliability of the DSM-IV diagnosis of SAD was
r ¼ 0.8. For the number of comorbid diagnosis, please refer to
Table 1.
2.4. Physiological measures
2.4.1. Breath alcohol concentration (BAC)
In order to ensure the effectiveness of the alcohol manipulation,
all participants were repeatedly measured with a breathalyzer
(Dra€ger Alcotest 7410, Dra €ger Sicherheitstechnik GmbH, Lübeck,
Germany), which is checked and calibrated annually by the
manufacturer. Participants in the ALC condition were tested for a
targeted breath alcohol concentration of 0.07%. Participants in the
PLA condition were “assessed” with a device that looked identical
to the valid breathalyzer but falsely indicated an elevated BAC of.
0.04% during the first, of 0.5% (indicating a rising blood alcohol
level) during the second assessment. Before each breath alcohol
test, both ALC and PLA participants were additionally asked to take
a guess regarding their current BAC, which was recorded by the
experimenter (“BAC-self”).
2.4.2. Electrocardiogram (ECG)
Heart rate (HR) was continuously recorded using the Varioport
system (Becker Meditec, Karlsruhe). The ECG was recorded with a
sample rate of 512 Hz using three electrodes attached to the chest.
Two of these electrodes were active, one placed on the right
collarbone and one below the left nipple on the lowest rib. The
third electrode on the left collarbone grounded the electric circuit.
For the ECG analysis, three intervals of 3 min duration each were
marked according to the experimental phases of the task: In a
baseline period, the participants were asked to sit quietly, refrain
from moving and talking and directing their gaze to a fixation cross
on the computer screen. During the speech task, participants were
instructed to concentrate on the speech, and during recovery,
baseline instructions were repeated.
2.5. Administration of alcohol
Participants in the ALC condition received an amount of alcohol
that was meant to induce a final BAC of 0.07%. We aimed for this
BAC level based on findings that individuals with this BAC are still
able to perform experimental or behavioral tasks without signifi-
cant performance decrements (Battista, Macdonald, & Stewart,
2012). The necessary amount of alcohol was estimated following
a modified version of the Widmark formula (Widmark, 1932). This
calculation takes into account the participant's height, weight, age,
and gender. The calculated amount of 37 vol.-% vodka was equally
distributed between three drinks of vodka-orange juice (ratio of
1:3). The vodka-orange juice mixture is indistinguishable from a
drink consisting of pure orange juice (OJ), which helped to ensure
the credibility of the placebo condition. This procedure has been
successfully used in previous investigations by our group (i.e.,
Gerlach et al., 2006; Stevens, Gerlach, & Rist, 2008; Stevens et al.,
2009).
2.6. Procedure (Fig. 1)
2.6.1. First laboratory visit
During the initial laboratory visit, participants read study in-
formation and provided informed consent. Next, they underwent a
SCID diagnostic assessment that lasted approximately two hours,
and were randomly assigned to the experimental conditions (ALC,
PLA, and OJ). Participants were then scheduled for the second
laboratory visit and were given nutrition instructions intended to
optimize any administration of alcohol during that second visit.
Specifically, they were instructed to eat a light meal four hours
before the experiment and to refrain from eating or drinking
caffeine after the meal. They were also asked to abstain from
alcohol 24 h prior to the second laboratory visit. In order to ensure
the safety of participants, they were informed that they could
choose to either arrange for a designated driver if they wanted to
leave the laboratory immediately after the upcoming visit, or wait
in the laboratory until their BAC was equal to or lower than 0.02%.
2.6.2. Second laboratory visit
At the beginning of the second laboratory visit (i.e., the exper-
imental session), female participants in the PLA and ALC conditions
were asked to conduct a routine self-administered urine pregnancy
test (HCG Ultra Strip, Ascimed Germany), but no women were
excluded because of a positive result. All participants were then
weighed in order to calculate the amount of alcohol needed for
reaching the appropriate BAC level or the amount of orange juice
needed for those in the PLA and OJ conditions. To ensure that all
participants did not drink any alcohol before the administration of
drinks, the first BAC estimation (BAC-self) and actual measurement
(BAC-actual) were conducted. The ECG electrodes were then
attached and a 2-min pre-drink physiological baseline assessment
was initiated. To assess for baseline anxiety, participants completed
VASs.
Next, participants underwent the administration of alcohol.
 S. Stevens et al. / Behaviour Research and Therapy 94 (2017) 9e18
Participants in the ALC condition were told that they would be
drinking alcohol, and were given three separate drinks containing a
vodka-orange juice mixture. They were given the three drinks
consecutively and asked to finish each one within five minutes.
Participants in the PLA condition were also told that they would be
drinking alcohol, and were also given three separate drinks. These
drinks were served in glasses that had been rubbed with a minimal
amount of vodka before the orange juice was poured in, and then
one drop of vodka was added; these measures were taken to create
the smell of alcohol and thus increase the credibility of the placebo.
They were given the three drinks consecutively and asked to finish
each one within five minutes. Participants in the OJ condition were
told that their drinks contained only orange juice, and were also
given the three drinks consecutively and asked to finish each
within five minutes.
After the drinking phase, which lasted a total of fifteen minutes,
participants sat quietly for another five minutes to allow for the
absorption of alcohol. Next, ALC and PLA participants rinsed their
mouths and the second BAC-self and BAC-actual assessment was
conducted. For the PLA group, a manipulated BAC tester was used
that first falsely indicated an elevated BAC of 0.04%. This BAC was
selected because on the one hand, it indicates a relevant BAC, but on
the other still is believable for participants in a placebo condition.
Participants were then given instructions concerning the up-
coming speech that highlighted the evaluative characteristics of the
task (i.e., “please remember that your performance will be video-
taped and judged”). The topic was on the pros and cons of the
death penalty. Directly before beginning the speech task, partici-
pants completed the VASs again and underwent the third BAC
assessment. Participants in the PLA condition were told that their
BAC was now at 0.05%, (indicating the rising arm of the breath
alcohol curve). During the 3-min speech task in front of a small
audience of two individuals, the second physiological HR assess-
ment took place, followed by the fourth BAC assessment (BAC-self
and BAC-actual). Here, neither participants in the PLA condition nor
in the ALC condition received feedback with regard to their actual
BAC. Following this, participants were asked to sit quietly, direct
their gaze towards a fixation cross on a computer screen in front of
them to assess the HR measurement for the recovery period.
Finally, participants rated their previous performance on the SPRS,
and the two audience members completed the SPRS scales.
Upon completion of the SPRS scales, participants in the PLA
group were informed of their drinks’ actual content. All participants
were debriefed concerning the purposes of the experiment.
3. Results
3.1. Data reduction
ECG raw data were visually inspected and corrected for artifacts.
Subsequently, an algorithm using MATLAB™ (Gerlach, Wilhelm,
Gruber, & Roth, 2001) calculated mean HR scores for each partici-
pant and measurement period (pre-drink, post-drink, and speech)
by counting the number of R-spikes per minute. Detected R-spikes
were checked visually to ensure the validity of the algorithm.
The three anxiety-related VASs (anxiety, tension, and nervous-
ness) were combined into a single mean anxiety rating for each of
the three measurement points (pre-drink, speech, and recovery), as
per Andor, Gerlach, & Rist, 2008). Correlations between the single
VASs ranged between r ¼ 0.77 and r ¼ 0.95 and were thus suffi-
ciently high to calculate a mean VAS score for each assessment.
Cronbach's a for VAS ratings were: a ¼ 0.95 (pre-drink), a ¼ 0.97
(speech), and a ¼ 0.95 (recovery).
Additionally, correlations between the two observers’ social
competence ratings for each participant were sufficiently high
13
(r ¼ 0.89) to be combined into a mean observer SPRS score
(Cronbach's a ¼ 0.82). Consequently, one self-rated and one
observer-derived SPRS score were entered into the analyses. Finally,
for the analyses regarding biased self-perceptions in individuals
with SAD, we calculated a performance rating difference score
(self-rating e observer-rating).
3.2. Preliminary analyses
Sample characteristics for the SAD and control groups, including
sociodemographic variables and scores on self-report measures are
reported in Table 1. We ran a series of t-tests and c2 tests to
examine whether groups differed on age, gender, education and
marital status. Results indicated that groups did not differ in any
sociodemographic variable. We also ran a one-way (Group: SAD,
control) MANOVA on questionnaire scores. SAD participants scored
significantly higher than control participants on all self-report
scales (see Table 1).
All the following analyses were also run including gender.
However, since gender was not significant in any of these analyses
and gender was not the focus of the present study, we decided to
not include the respective analyses here.
3.3. Manipulation checks
3.3.1. Efficacy of the alcohol administration
To check whether our drinking procedure actually induced the
expected BAC in the SAD and control participants who actually
drank alcohol, a repeated measures ANOVA on BAC-actual with
Group (SAD, control) and Time (baseline, pre-drink, speech, re-
covery) yielded a significant multivariate effect of time (F(3,
37) ¼ 256.8, p < 0.01, h2p ¼ 0.96). BAC levels between the two groups
were not significantly different (see Fig. 2; F(1,39) ¼ < 0.01, p ¼ 0.96,
h2p < 0.01). Breath alcohol levels increased from pre-to post-drink as
well as from post-drink to speech (mean differences ¼ 0.6 and 0.1,
respectively; ps < 0.01). Mean BAC levels were raised to 0.072%
(SD ¼ 0.02) for SAD patients and to 0.071% (SD ¼ 0.02) for CONT
participants.
3.3.2. Efficacy of the placebo manipulation
Next, we examined whether participants in the alcohol condi-
tion evidenced differences in BAC-self ratings relative to partici-
pants in the placebo condition. A 2 (Group: SAD, control)  2
(Condition: alcohol, placebo)  4 (Time: pre-drink, post-drink,
speech, recovery) mixed model ANOVA on BAC-self levels indicated
a significant interaction effect Condition  Time (F(3, 81) ¼ 25.3,
p < 0.01, h2p ¼ 0.51), such that participants in the alcohol condition
evidenced higher BAC levels during post-drink, speech and recov-
ery but not during baseline (see Fig. 2). A main effect of time
(F(1,3) ¼ 397.5, p < 0.001, h2p ¼ 0.94) also indicated that, irrespective
of group and condition, estimated BAC increased across assessment
points but did not differ between SAD and control participants.
Concerning the credibility of the placebo manipulation, all partic-
ipants in the PLA condition estimated their BAC to be higher than
zero during the speech task. In addition, 16 of 21 SAD patients (76%)
and 18 of 21 CONT participants (86%) reported a BAC of at least
0.04% in the placebo group.
3.3.3. Efficacy of the speech anxiety induction
Regarding the effects of our speech task, we wanted to ensure
that it increases self-report anxiety and HR. Regarding self-report
anxiety, a repeated measures ANOVA with assessment (pre-drink
vs. speech vs. recovery) as within-subjects factor and group (SAD,
control) as between-subjects factors on self-reported anxiety
revealed a significant main effect of group (F (1,116) ¼ 67.5,
14 S. Stevens et al. / Behaviour Research and Therapy 94 (2017) 9e18
Fig. 1. Procedure.
SCID I ¼ Structured Clinical Interview for Mental Disorders 1; VAS ¼ Visual Analogue Scales Anxiety; BACeS ¼ Breath Alcohol Level-self report; BAC-O ¼ Breath Alcohol Level-
objective; SPRS ¼ Social Performance Rating Scale.
Fig. 2. Measured and estimated BAC.
ALC ¼ alcohol condition; PLA ¼ placebo condition; BAC-Actual ¼ measured breath
alcohol concentration in ALC condition; BAC-self ¼ participants' estimated BAC before
each measurement/false feedback of BAC. Bars indicate standard errors.
p < 0.001, h2p ¼ 0.37). Individuals with SAD had higher VAS scores
than controls during the assessment (see Fig. 3). A significant main
effect of time (F (2,122) ¼ 28.4, p < 0.001, h2p ¼ 0.20) was found due
to self-reported anxiety being higher during the speech task than
during baseline (Mspeech ¼ 28.2; SDspeech ¼ 28.5; Mbaseline ¼ 18.5;
SDbaseline ¼ 19.7) and recovery across groups (Mrecovery ¼ 16.21;
SDrecovery ¼ 22.7; compare Fig. 2). For heart rate (see Fig. 4), we
found a main effect of time (F (2,121) ¼ 325.1, p < 0.01, h2p ¼ 0.77),
such that HR increased from baseline to speech (1, 122) ¼ 99.8;
p < 0.05; h2p ¼ 0.3), and decreased from speech to recovery (1,
122) ¼ 78.42; p < 0.05; h2p ¼ 0.47), irrespective of group. In addition,
a main effect of group (F (1,121) ¼ 5.1, p ¼ 0.03, h2p ¼ 0.05) indicated
that individuals with SAD had higher HR compared to controls. No
effects of condition or any interactions on HR were found.
3.4. The effects of alcohol on anxiety and social performance
All significant effects regarding our main hypotheses were fol-
lowed up hierarchically with further ANOVAs and for comparison
for two means with planned simple contrasts.
3.4.1. Effects of alcohol on self-report anxiety and HR
We examined whether alcohol reduced anxiety as measured by
self-report (see Fig. 3) or HR (see Fig. 4) in two separate repeated
measures ANOVAs. The 2 group (SAD, control)  3 condition
 S. Stevens et al. / Behaviour Research and Therapy 94 (2017) 9e18 15

Fig. 3. Anxiety Scale scores between groups, condition, and time.

SAD ¼ Social Anxiety Disorder group; CONT ¼ control group; OJ ¼ orange juice condition; PLA ¼ placebo condition; ALC ¼ alcohol condition. Bars indicate standard errors.

Fig. 4. Mean HR between groups, condition, and time.

SAD ¼ Social Anxiety Disorder group; CONT ¼ control group; OJ ¼ orange juice condition; PLA ¼ placebo condition; ALC ¼ alcohol condition. Bars indicate standard errors.

(control, placebo, alcohol)  3 assessment (baseline, speech, re-
covery) repeated measures analysis for self-report anxiety resulted
in a significant 3 way interaction (F (4, 122) ¼ 3.2; p ¼ 0.01;
h2p ¼ 0.05). In follow-up repeated measures ANOVAs separate for
groups, the analysis for controls only revealed a main effect of
assessment (F (2, 60) ¼ 20.0; p < 0.001; h2p ¼ 0.41), such that
anxiety increased from baseline to speech (F (1, 60) ¼ 234.35;
p < 0.001; h2p ¼ 0.5) and decreased from speech to recovery (F (1,
60) ¼ 199.75; p < 0.001; h2p ¼ 0.44). Thus, no effect of condition on
anxiety was found for controls (F (2, 60) ¼ 0.98; p ¼ 0.45;
h2p ¼ 0.01). For individuals with SAD, we found a significant
assessment  condition interaction (F (2, 62) ¼ 3.8; p ¼ 0.01;
16 S. Stevens et al. / Behaviour Research and Therapy 94 (2017) 9e18
h2p ¼ 0.01). No differences between conditions were revealed for
baseline (F (2, 62) ¼ 1.3; p ¼ 0.3; h2p ¼ 0.07)), while during speech
and recovery, individuals in the verum (F (1, 62) ¼ 425.35;
p < 0.001; h2p ¼ 0.7) and placebo conditions (F (1, 60) ¼ 310.55;
p < 0.001; h2p ¼ 0.54) reported less anxiety compared to those in the
control condition (see Fig. 3) and those in the verum condition
reported less anxiety compared to individuals with SAD in the
placebo condition (F (1, 60) ¼ 123.98; p < 0.001; h2p ¼ 0.3).
Next, we examined whether alcohol reduced HR in the alcohol
compared to the placebo and verum conditions initially using a 2
group (SAD, controls)  3 condition (control, verum, placebo)  3
assessment (baseline, speech, recovery) repeated measures
ANOVA. This analysis only revealed two main effects of group (F (1,
122) ¼ 5.2; p ¼ 0.025; h2p ¼ 0.1) and time (F (2, 122) ¼ 325.46;
p < 0.001; h2p ¼ 0.77). HR increased from baseline to speech (F (1,
122) ¼ 303.56; p < 0.001; h2p ¼ 0.69) and decreased from speech to
recovery (F (1, 122) ¼ 228.64; p < 0.001; h2p ¼ 0.62) irrespective of
condition and group (compare Fig. 4).
3.4.2. Effects of alcohol on social performance measures
To examine the effects of alcohol on social performance, we
conducted a 2 (Group: SAD, controls)  3 (Condition: alcohol,
placebo, OJ)  2 (source of the ratings: self vs. observer) repeated
measurement ANOVA on social performance ratings provided by
participants and observers with the SPRS as dependent variable.
This ANOVA resulted in a significant “group  rating source”
interaction (F (1,121) ¼ 25.1, p < 0.001, h2p ¼ 0.19), such that only in
individuals with SAD, self-ratings of social performance were lower
compared to observer ratings (see Fig. 5; p < 0.05). In controls,
there was no difference between self- and observer-rated social
performance measures (p > 0.05). Next, a significant rating
source  condition interaction (F (2,121) ¼ 10.23, p < 0.001,
h2p ¼ 0.16) revealed that observer-rated performance decreased
from orange juice to placebo conditions (F (1, 122) ¼ 366.42;
p < 0.001; h2p ¼ 0.7), and from placebo to alcohol conditions (1,
122) ¼ 197.82; p < 0.001; h2p ¼ 0.38), irrespective of the clinical
status of the participants. No influence of alcohol use was found for
social performance as rated by the individuals themselves.
4. Discussion
We investigated the impact of alcohol use on social performance
in individuals with SAD and healthy controls, as well as on self-
Fig. 5. Self- and observer-rated SPRS scales between groups and condition.
Higher scores indicate poorer performance ratings. SPRS ¼ Social Performance Rating Scale; Self ¼ self-rated SPRS; observer ¼ observer-rated SPRS; ALC ¼ alcohol condition;
PLA ¼ placebo condition; OJ ¼ orange juice condition; SAD ¼ social anxiety disorder group; CONT ¼ control group. Bars indicate standard errors.
report and physiological measures of anxiety. Alcohol reduced
anxiety during a speech task in individuals with SAD. Since patients
in the placebo group felt more anxious compared to those in the
alcohol group, the reduction in self-report anxiety can partly be
attributed to the pharmacological effects of alcohol. However, SAD
participants receiving placebo also reported less anxiety compared
to the control (orange juice) condition lending support to the
notion, that alcohol expectancy effects also may reduce social
anxiety.
Individuals with SAD rated their performance worse compared
to the ratings of neutral observers. Most interestingly, although
alcohol led to a decrease in self-report anxiety in individuals with
SAD, no beneficial effects of alcohol on self-rated performance were
found. In contrast, alcohol intake had detrimental effects on
observer rated social performance in both groups (compare Keane
& Lisman, 1980; for a similar effect in socially anxious males).
Interestingly, a detrimental effect of placebo on observer rated
social performance was also revealed. At this point it is unclear
what mechanism may be responsible for such an effect. Possibly,
fear of being hindered by intoxication to give the speech skillfully
may have actually hampered performance (Cludius et al., 2013).
In addition to the replication of deficits in social performance in
SAD (Baker & Edelmann, 2002; Norton & Hope, 2001; Stevens et al.,
2010; Voncken & Bogels, 2008), and anxiety reduction in in-
dividuals with SAD in a speech task (e.g., Abrams et al., 2001), the
key finding of the current study is two-fold: While participants
suffering from SAD reported less anxiety after drinking alcohol
compared to those who drank a placebo beverage or orange juice,
their performance was judged more negatively by group- and
condition-blind observers. This key finding is somewhat surprising,
as, arguably, a reduction in anxiety should lead to increases in social
performance, according to modern cognitive models of SAD (Clark
& Wells, 1995). Our results, however, suggest a rather independent
influence of alcohol on social performance and anxiety, which
might increase the risk for hazardous drinking in individuals with
SAD. Primarily, alcohol reduces social anxiety in individuals with
SAD, qualifying negative reinforcement as one possible mechanism
that increases the risk of further alcohol intake (Battista et al.,
2010). This anxiety reducing effect does not, however, directly
lead to improvements in social performance, as the physiological
consequences of alcohol intake tend to include typical behaviors
associated with reduced social performance (e.g., having problems
concentrating, slurring words, see also Keane & Lisman, 1980).
 S. Stevens et al. / Behaviour Research and Therapy 94 (2017) 9e18
Indeed, reduction of anxiety in individuals with SAD and decreases
in social performance were particularly high in the alcohol
compared to the placebo and orange juice groups. These results
suggest that individuals with SAD are explicitly susceptible to the
effects of alcohol with regards to its anxiety-reducing properties,
and might be particularly vulnerable to observable decreases in
social performance: Social rejection in performance situations due
to worse social performance might further increase the vicious
circle of anxiety, alcohol use, decreases in social performance, and
negative reactions from others (Alden & Taylor, 2004).
Regarding our placebo design, we found more pronounced ef-
fects of alcohol on anxiety and social performance in the alcohol
compared to the placebo and the orange juice conditions. As sug-
gested by the recent literature on alcohol effects in SAD (Battista
et al., 2010), a relatively large amount of alcohol is necessary to
investigate the effect of alcohol on anxiety in the laboratory (tar-
geting at least a blood alcohol level of 0.06%). However, as the
placebo group in individuals with SAD consistently differed from
the orange juice group in the current study, considerations of the
psychological effects of alcohol use should not be neglected.
Recently, alcohol expectancies have been regularly linked to the
anxiety reducing effects of alcohol (Abrams & Kushner, 2004). In
particular, expectations regarding tension reduction and improve-
ment in social performance after drinking alcohol have been found
to influence the anxiety reduction properties of alcohol. In addition
to alcohol expectancies, further psychological aspects such as the
motive to drink alcohol to reduce social anxiety (Cludius et al.,
2013), and the adequacy of alcohol use depending on the social
situation (i.e., speech vs. interaction), should be considered when
attempting to explain placebo effects in social anxiety research.
In a clinical sense, our results point to the pivotal role of in-
terventions that address problematic alcohol use in social anxiety.
On the one hand, alcohol leads to immediate anxiety reduction and
is comparable to other safety behaviors as it prevents learning in
social situations, i.e. that social situations are not dangerous
without the use of alcohol. On the other hand, while alcohol use
might help to alleviate anxiety symptoms, it decreases social per-
formance, and thus might contribute to increased social isolation of
individuals with SAD.
Naturally, our study has some limitations, which have to be
carefully considered. Individuals in the placebo condition believed
that they consumed less alcohol compared to those in the alcohol
condition. Ideally, both the alcohol group and the placebo group
should not differ in this regard. However, as has been pointed out
by Martin and Sayette (1993), this cannot be accomplished if higher
BAC levels (i. e., 0.07%) are used, because of the strong pharmaco-
logical effect of this level of alcohol. Nonetheless, our placebo group
significantly differed in perceived BAC from the control group,
allowing to at least partially disentangle the pharmacological vs.
expectancy effects of alcohol intake. Furthermore, due to ethical
consideration, we excluded participants who currently or previ-
ously suffered from alcohol related disorders, as these groups may
be particularly susceptible to the effects of alcohol in social situa-
tions. In addition, as our experimental conditions might have
differed regarding the stress they caused our participants, i.e.
drinking alcohol or not, we were asked by the ethical review board
to compensate differentially depending on the experimental group.
Unfortunately, this might be a meaningful threat to the internal
validity of our study. Similarly problematic, only female partici-
pants in the ALC and PLA condition were given a urine pregnancy
test. Also, we used a speech task as social stress task. Individuals
with SAD may, however, use alcohol more often in social interac-
tion situations because in social interactions impairment of per-
formance may be of less concern and positive effects of alcohol may
be more salient (e.g. higher social assertiveness; Cludius et al.,
17
2013). Nonetheless, in this study a speech task was chosen since
this task is more easily controlled experimentally, has been studied
most extensively, and most reliably induces social anxiety in both
socially anxious and healthy individuals. Also, arguably, at least
pharmacological effects of alcohol on anxiety should not be
differentially affected by the situation in which anxiety occurs.
However, future research should also focus on other common social
situations like small talk or eating, where alcohol use may better fit
into the social situation and ambulatory assessment might be used
to measure real life effects of alcohol on social anxiety. Finally, the
speech did not induce much anxiety in the control group, which
might explain why we did not find an anxiety reducing effect of
alcohol in the control group.
In summary, our study reveals a complex interaction of social
anxiety, social performance, and alcohol use. While alcohol atten-
uates social anxiety on a self-report level, social performance de-
creases in intoxicated individuals. Negative reactions from others
based on these performance deficits might, in turn, increase social
anxiety, which would then contribute to the maintenance of the
vicious circle of anxiety and alcoholism.
Conflict of interest
The authors declare that there are no conflicts of interest.
References
Abrams, K., & Kushner, M. G. (2004). The moderating effects of tension-reduction
alcohol outcome expectancies on placebo responding in individuals with so-
cial phobia. Addictive Behaviors, 29(6), 1221e1224. http://dx.doi.org/10.1016/
j.addbeh.2004.03.020.
Abrams, K., Kushner, M., Medina, K. L., & Voight, A. (2001). The pharmacologic and
expectancy effects of alcohol on social anxiety in individuals with social phobia.
Drug Alcohol Depend, 64(2), 219e231. dx.doi.org/10.1016/S0376-8716(01)00125-
9.
Alden, L. E., & Taylor, C. T. (2004). Interpersonal processes in social phobia. Clinical
Psychology Review, 24(7), 857e882. http://dx.doi.org/10.1016/j.cpr.2004.07.006.
Allen, J. P., Litten, R. Z., Fertig, J. B., & Babor, T. (1997). A review of research on the
alcohol use disorders identification test (AUDIT). Alcoholism-clinical and
Experimental Research, 21(4), 613e619. http://dx.doi.org/10.1097/00000374-
199706000-00008.
Andor, T., Gerlach, A. L., & Rist, F. (2008). Superior perception of phasic physiological
arousal and the detrimental consequences of the conviction to be aroused on
worrying and metacognitions in GAD. Journal of Abnormal Psychology, 117(1),
193e205. http://dx.doi.org/10.1037/0021-843x.117.1.193.
Bacon, A. K., & Ham, L. S. (2010). Attention to social threat as a vulnerability to the
development of comorbid social anxiety disorder and alcohol use disorders: An
avoidance-coping cognitive model. Addictive Behaviors, 35(11), 925e939. http://
dx.doi.org/10.1016/j.addbeh.2010.06.002.
Baker, S. R., & Edelmann, R. J. (2002). Is social phobia related to lack of social skills?
Duration of skill-related behaviours and ratings of behavioural adequacy. British
Journal of Clinical Psychology, 41, 243e257. http://dx.doi.org/10.1348/
014466502760379118.
Battista, S. R., Macdonald, D., & Stewart, S. H. (2012). The effects of alcohol on safety
behaviors in socially anxious individuals. Journal of Social and Clinical Psychol-
ogy, 31(10), 1074e1094. http://dx.doi.org/10.1521/jscp.2012.31.10.1074.
Battista, S. R., Stewart, S. H., & Ham, L. S. (2010). A critical review of laboratory-based
studies examining the relationships of social anxiety and alcohol intake. Current
Drug Abuse Reviews, 3(1), 3e22. http://dx.doi.org/10.2174/
1874473711003010003.
Beck, A. T., Erbaugh, J., Ward, C. H., Mock, J., & Mendelsohn, M. (1961). An inventory
for measuring depression. Archives of General Psychiatry, 4(6). http://dx.doi.org/
10.1001/archpsyc.1961.01710120031004, 561-&.
Buckner, J. D., & Heimberg, R. G. (2010). Drinking behaviors in social situations
account for alcohol-related problems among socially anxious individuals. Psy-
chology of Addictive Behaviors, 24(4), 640e648. http://dx.doi.org/10.1037/
a0020968.
Buckner, J. D., Heimberg, R. G., Ecker, A. H., & Vinci, C. (2013). A biopsychosocial
model of social anxiety and substance use. Depression and Anxiety, 30(3),
276e284. http://dx.doi.org/10.1002/da.22032.
Chutuape, M. A. D., & Dewit, H. (1995). Preferences for ethanol and diazepam in
anxious individuals - an evaluation of the self-medication hypothesis. Psycho-
pharmacology, 121(1), 91e103. http://dx.doi.org/10.1007/Bf02245595.
Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. In R. G. Heimberg
(Ed.), Social phobia - diagnosis, assessment, and treatment (pp. 69e93). New
York: The Guilford Press.
18 S. Stevens et al. / Behaviour Research and Therapy 94 (2017) 9e18
Cludius, B., Stevens, S., Bantin, T., Gerlach, A. L., & Hermann, C. (2013). The motive to
drink due to social anxiety and its relation to hazardous alcohol use. Psychology
of Addictive Behaviors, 27(3), 806e813. http://dx.doi.org/10.1037/a0032295.
Crum, R. M., & Pratt, L. A. (2001). Risk of heavy drinking and alcohol use disorders in
social phobia: A prospective analysis. American Journal of Psychiatry, 158(10),
1693e1700. http://dx.doi.org/10.1176/appi.ajp.158.10.1693.
Davidson, J. R. T., Hughes, D. L., George, L. K., & Blazer, D. G. (1993). The epidemi-
ology of social phobia - findings from the duke epidemiologic catchment-area
study. Psychological Medicine, 23(3), 709e718. http://dx.doi.org/10.1017/
S0033291700025484.
Dybek, I., Bischof, G., Grothues, J., Reinhardt, S., Meyer, C., Hapke, U., et al. (2006).
The reliability and validity of the alcohol use disorders identification test
(AUDIT) in a German general practice population sample. Journal of Studies on
Alcohol, 67(3), 473e481.
Eidecker, J., Glockner-Rist, A., & Gerlach, A. L. (2010). Dimensional structure of the
Social Interaction Anxiety Scale according to the analysis of data obtained with
a German version. Journal of Anxiety Disorders, 24(6), 596e605. http://
dx.doi.org/10.1016/j.janxdis.2010.03.020.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1996). Structured clinical
interview for DSM-IV Axis I disorders, clinician version (SCID-CV). Washington, DC:
American Psychiatric Press, Inc.
Fresco, D. M., Coles, M. E., Heimberg, R. G., Liebowitz, M. R., Hami, S., Stein, M. B.,
et al. (2001). The Liebowitz social anxiety scale: A comparison of the
psychometric properties of self-report and clinician-administered formats.
Psychological Medicine, 31(6), 1025e1035. http://dx.doi.org/10.1017/
S0033291701004056.
Fydrich, T., & Bürgener, F. (1999). Ratingskalen für soziale Kompetenz. In J. Margraf,
& K. Rudolf (Eds.), Vol. 2. Soziale kompetenz - soziale Phobie: Anwendungsfelder,
entwicklungslinien, erfolgsaussichten (pp. 81e98). Baltmannsweiler: Schneider-
Verlag.
Fydrich, T., Chambless, D. L., Perry, K. J., Buergener, F., & Beazley, M. B. (1998).
Behavioral assessment of social performance: A rating system for social phobia.
Behaviour Research and Therapy, 36(10), 995e1010. http://dx.doi.org/10.1016/
S0005-7967(98)00069-2.
Gerlach, A. L., Schiller, A., Wild, C., & Rist, F. (2006). Effects of alcohol on the pro-
cessing of social threat-related stimuli in socially phobic women. British Journal
of Clinical Psychology, 45, 279e295. http://dx.doi.org/10.1348/
014466505x49862.
Gerlach, A. L., Wilhelm, F. H., Gruber, K., & Roth, W. T. (2001). Blushing and physi-
ological arousability in social phobia. Journal of Abnormal Psychology, 110(2),
247e258. http://dx.doi.org/10.1037/0021-843x.110.2.247.
Hautzinger, M., Bailer, M., Worall, H., & Keller, F. (1994). Beck-depressions-inventar
(BDI). Bern: Huber.
Himle, J. A., Abelson, J. L., Haghightgou, H., Hill, E. M., Nesse, R. M., & Curtis, G. C.
(1999). Effect of alcohol on social phobic anxiety. American Journal of Psychiatry,
156(8), 1237e1243.
Hofmann, S. G., Gerlach, A. L., Wender, A., & Roth, W. T. (1997). Speech disturbances
and gaze behavior during public speaking in subtypes of social phobia. Journal
of Anxiety Disorders, 11(6), 573e585. http://dx.doi.org/10.1016/S0887-6185(97)
00040-6.
Josephs, R. A., & Steele, C. M. (1990). The 2 faces of alcohol myopia - attentional
mediation of psychological stress. Journal of Abnormal Psychology, 99(2),
115e126. http://dx.doi.org/10.1037/0021-843x.99.2.115.
Keane, T. M., & Lisman, S. A. (1980). Alcohol and social anxiety in males - behavioral,
cognitive, and physiological-effects. Journal of Abnormal Psychology, 89(2),
213e223.
Kessler, R. C., Chiu, W. T., Demler, O., & Walters, E. E. (2005). Prevalence, severity,
and comorbidity of 12-month DSM-IV disorders in the national comorbidity
survey replication. Archives of General Psychiatry, 62(6), 617e627. http://
dx.doi.org/10.1001/archpsyc.62.6.617.
Kidorf, M., & Lang, A. R. (1999). Effects of social anxiety and alcohol expectancies on
stress-induced drinking. Psychology of Addictive Behaviors, 13(2), 134e142.
http://dx.doi.org/10.1037/0893-164X.13.2.134.
Kushner, M. G., Sher, K. J., & Beitman, B. D. (1990). The Relation between alcohol-
problems and the anxiety disorders. American Journal of Psychiatry, 147(6),
685e695. http://dx.doi.org/10.1176/ajp.147.6.685.
Liebowitz, M. R. (1987). Social phobia. In T. A. Ban, T. Pichot, & W. Poldinger (Eds.),
Modern problems of pharmacopsychiatry (pp. 141e173). New York: Karger.
Martin, C. S., & Sayette, M. A. (1993). Experimental-design in alcohol administration
research - limitations and alternatives in the manipulation of dosage-set.
Journal of Studies on Alcohol, 54(6), 750e761. http://dx.doi.org/10.15288/
jsa.1993.54.750.
Mattick, R. P., & Clarke, J. C. (1998). Development and validation of measures of
social phobia scrutiny fear and social interaction anxiety. Behaviour Research
and Therapy, 36(4), 455e470. http://dx.doi.org/10.1016/S0005-7967(97)10031-
6.
Naftolowitz, D. F., Vaughn, B. V., Ranc, J., & Tancer, M. E. (1994). Response to alcohol
in social phobia. Anxiety, 1(2), 96e99.
Norton, P. J., & Hope, D. A. (2001). Kernels of truth or distorted perceptions: Self and
observer ratings of social anxiety and performance. Behavior Therapy, 32(4),
765e786. http://dx.doi.org/10.1016/S0005-7894(01)80020-4.
Quitkin, F. M., Rifkin, A., Kaplan, J., & Klein, D. F. (1972). Phobic anxiety syndrome
complicated by drug dependence and addiction - treatable form of drug-abuse.
Archives of General Psychiatry, 27(2). http://dx.doi.org/10.1001/arch-
psyc.1972.01750260013002, 159-&.
Randall, C. L. (2000). Alcoholism and social anxiety (Denver, CO). In Paper presented
at the annual meeting of the research society on alcoholism.
Rapee, R. M., & Lim, L. (1992). Discrepancy between self and observer ratings of
performance in social phobics. Journal of Abnormal Psychology, 101(4), 728e731.
http://dx.doi.org/10.1037/0021-843X.101.4.728.
Rist, F., Glockner-Rist, A., & Demmel, R. (2009). The Alcohol Use Disorders Identi-
fication Test revisited: Establishing its structure using nonlinear factor analysis
and identifying subgroups of respondents using latent class factor analysis.
Drug and Alcohol Dependence, 100(1e2), 71e82. http://dx.doi.org/10.1016/
j.drugalcdep.2008.09.008.
Saunders, J. B., Aasland, O. G., Babor, T. F., Delafuente, J. R., & Grant, M. (1993).
Development of the alcohol-use disorders identification test (audit) - who
collaborative project on early detection of persons with harmful alcohol-
consumption .2. Addiction, 88(6), 791e804. http://dx.doi.org/10.1111/j.1360-
0443.1993.tb02093.x.
Sayette, M. A. (1993). An appraisal-disruption model of alcohols effects on stress
responses in social drinkers. Psychological Bulletin, 114(3), 459e476. http://
dx.doi.org/10.1037/0033-2909.114.3.459.
Stangier, U., & Heidenreich, T. (2003). Die Liebowitz soziale angst-skala (LSAS). In
C. I. P. Scalarum (Ed.), Internationale Skalen für Psychiatrie. Go € ttingen: Hogrefe.
Stangier, U., Heidenreich, T., Berardi, A., Golbs, U., & Hoyer, J. (1999). Assessment of
social phobia by the social interaction anxiety scale and the social phobia scale.
Zeitschrift Fur Klinische Psychologie-Forschung Und Praxis, 28(1), 28e36. http://
dx.doi.org/10.1026//0084-5345.28.1.28.
Stevens, S., & Gerlach, A. L. (2009). Self-medication and social phobia: Psychometric
properties of the questionnaire “drinking due to social anxiety”. Zeitschrift Fur
Klinische Psychologie Und Psychotherapie, 38(2), 118e125. http://dx.doi.org/
10.1026/1616-3443.38.2.118.
Stevens, S., Gerlach, A. L., & Rist, F. (2008). Effects of alcohol on ratings of emotional
facial expressions in social phobics. Journal of Anxiety Disorders, 22(6), 940e948.
http://dx.doi.org/10.1016/j.janxdis.2007.09.007.
Stevens, S., Hofmann, M., Kiko, S., Mall, A. K., Steil, R., Bohus, M., et al. (2010). What
determines observer-rated social performance in individuals with social anxi-
ety disorder? Journal of Anxiety Disorders, 24(8), 830e836. http://dx.doi.org/
10.1016/j.janxdis.2010.06.005.
Stevens, S., Rist, F., & Gerlach, A. L. (2009). Influence of alcohol on the processing of
emotional facial expressions in individuals with social phobia. British Journal of
Clinical Psychology, 48, 125e140. http://dx.doi.org/10.1348/014466508x368856.
Stopa, L., & Clark, D. M. (1993). Cognitive-processes in social phobia. Behaviour
Research and Therapy, 31(3), 255e267. http://dx.doi.org/10.1016/0005-7967(93)
90024-O.
Terlecki, M. A., Buckner, J. D., Larimer, M. E., & Copeland, A. L. (2011). The role of
social anxiety in a brief alcohol Intervention for heavy-drinking college stu-
dents. Journal of Cognitive Psychotherapy, 25(1), 7e21. http://dx.doi.org/10.1891/
0889-8391.25.1.7.
Thomas, S. E., Randall, C. L., & Carrigan, M. H. (2003). Drinking to cope in socially
anxious individuals: A controlled study. Alcoholism:Clinical and Experimental
Research, 27(12), 1937e1943. http://dx.doi.org/10.1097/
01.ALC.0000100942.30743.8C.
Voncken, M. J., & Bogels, S. M. (2008). Social performance deficits in social anxiety
disorder: Reality during conversation and biased perception during speech.
Journal of Anxiety Disorders, 22(8), 1384e1392. http://dx.doi.org/10.1016/
j.janxdis.2008.02.001.
Wagner, C., Stangier, U., Heidenreich, T., & Schneider, R. (2004). “Drinking due to
social anxiety” and “social anxiety due to Drinking”: Development of two scales
for alcohol-dependency. Zeitschrift Fur Klinische Psychologie Und Psychotherapie,
33(4), 308e315. http://dx.doi.org/10.1026/1616-3443.33.4.308.
Widmark, E. M. P. (1932). Die theoretischen Grundlagen und die praktische Ver-
wendbarkeit der gerichtlich-medizinischen Alkoholbestimmung. Berlin: Urban &
Schwarzenberg.
Wittchen, H. U., Wunderlich, U., Gruschwitz, S., & Zaudig, M. (1997). SKID-I -
strukturiertes Klinisches Interview für DSM-IV - achse I: Psychische Sto €rungen.
Go €ttingen: Hogrefe-Verlag.