research paper and annotated bibliography

keyym12

22546246.pdf

Home >English homework help >research paper and annotated bibliography

Effects of alcohol on the processing of social threat-related stimuli in socially phobic women

Alexander L. Gerlach1*, Anke Schiller2,3, Cornelia Wild1

and Fred Rist1 1

University of Münster, Germany 2

Christoph Dornier Clinic, Germany 3

Christoph Dornier Foundation, Germany

Background. Social phobics are at a higher risk of developing alcohol problems. The mechanism promoting this association is not clear. According to Sayette (1993b), alcohol attenuates anxiety responses by disrupting initial appraisal of threatening stimuli. We used the emotional Stroop test and an implicit memory test to investigate whether alcohol hinders appraisal of social threat words in patients diagnosed with social phobia.

Procedure. Thirty-two women with social phobia (DSM-IV) and 32 female controls performed an emotional Stroop test either after drinking alcohol resulting in a blood alcohol levels (BAL) of 0.6‰ or after drinking a non-alcoholic beverage. The emotional Stroop test contained social anxiety-related and neutral stimuli. Implicit memory for the words presented was tested with a word-stem completion test.

Results. Without alcohol, both controls and socially-phobic participants took longer to name the colour of socially-threatening stimuli than of neutral stimuli. Alcohol levelled response latencies to the two stimulus categories only in controls. Socially- phobic participants responded more slowly to social anxiety-related stimuli than to neutral stimuli, irrespective of their BAL. In contrast to controls, social phobics showed an implicit memory bias for social threat words. This bias was attenuated by alcohol.

Discussion. Alcohol disrupts appraisal of social anxiety-related stimuli in controls but not in social phobics; in these it hinders the consolidation of memory. This also suggests that social phobics experience similar anxiety with and without alcohol, but remember this experienced anxiety less precisely. This effect might act as a reinforcer for the use of alcohol for the purpose of self-medication in future situations.

Epidemiological studies show that socially anxious people are at a higher risk of abusing

alcohol or developing alcohol dependence (Schneier, Martin, Liebowitz, Gorman, & Fyer, 1989; Allan, 1995; Holle, Heimberg, Sweet, & Holt, 1995; Himle & Hill, 1991;

* Correspondence should be addressed to Alexander L. Gerlach, WWU Münster, Department of Clinical Psychology, Fliednerstr. 21, 48149 Münster, Germany (e-mail: [email protected]).

The British Psychological Society

279

British Journal of Clinical Psychology (2006), 45, 279–295

q 2006 The British Psychological Society

www.bpsjournals.co.uk

DOI:10.1348/014466505X49862

Page & Andrews, 1996; Nardi & Versiani, 1997; Clark & Sayette, 1993; Stockwell &

Bolderston, 1987). A prospective study found a 2.3-fold increased risk for alcohol abuse

and dependence in subclinical social phobia (Crum & Pratt, 2001). However, there are

also reports that people with a diagnosis of social phobia consume less alcohol than

controls (Holle et al., 1995). This result cannot simply be explained as a social desirability

effect, since social desirability and reported alcohol consumption werenot correlated in a different sample of patients with social phobia (Cox, Swinson, Direnfeld, & Bourdeau,

1994). Thus, different processes may be at work as different intensities of social anxiety

lead to the overlap of social phobia, alcohol abuse and dependence.

A number of models attempt to explain the connection between anxiety or stress

and alcohol use. Conger (1956) suggested that alcohol reduces tension (or anxiety) and

that people consume alcohol to this effect. In line with this suggestion, social phobia

patients consume more alcohol in a socially stressful situation (public speaking), and

they report more attenuation of anxiety by drinking alcohol than controls, especially if they expected that alcohol would have such an effect (Abrams, Kushner, Medina, &

Voight, 2001; Abrams, Kushner, Medina, & Voight, 2002). These findings support a self-

medication theory of alcohol abuse. In this model, alcohol serves as a readily-available

means to cope with anxiety, inadvertently leading to abuse and dependence. In contrast

to this model, however, direct anxiety-reducing effects of alcohol in socially anxious

people have not been found consistently. For example, a direct effect of alcohol on

anxiety was lacking when social phobics were to give a speech under the influence of

alcohol (Himle et al., 1999; Naftolowitz, Vaughn, Ranc, & Tancer, 1994). The tension-reduction theory as stated by Conger (1956) has not remained

undisputed. According to Sayette’s (1993b) appraisal-disruption model, ‘SRD

[stress-response dampening] occurs to the degree that alcohol acts pharmacologically

to interfere with a person’s appraisal of stressful information’ (p. 463). Sayette suggests

that alcohol has this effect because it reduces the propensity of relevant stimuli to

activate stressor-associated memories. In the context of social phobia, these

considerations imply that biased processing of social phobia-related stimuli is reduced

after the consumption of alcohol, since phobia-related memories are not activated. However, Sayette also proposed that in case of stressors which are most readily or

automatically apprised, alcohol is much less likely to hinder appraisal. This proposition

is based on the conviction that highly automatic cognitive processes cannot easily be

disrupted. Accordingly, alcohol may not affect processing of threat-related stimuli in

clinically-phobic subjects. Whereas this notion seems reasonable, it has not yet been

tested directly in an information-processing paradigm.

Alternatively, Josephs and Steele (1990) Steele and Josephs (1988) suggested that

alcohol will narrow the attention of an individual to immediate and salient cues, hindering the processing of more remote or less salient cues (alcohol myopia theory):

only if a cue is present that is more salient or more easily processed than the anxiety-

related cues, will alcohol alleviate anxiety by hindering the processing of the anxiety-

related cue.

The processing of social phobia-related stimuli can be assessed by various procedures.

The paradigm that is probably most often used in social phobia and anxiety disorders is

the emotional Stroop test. Colour-naming latency to social phobia-related words as

compared with neutral words was longer in a number of studies in social phobia patients (Amir, Freshman, & Foa, 2002; Amir et al., 1996; Becker, Rinck, Margraf, & Roth, 2001;

Holle, Neely, & Heimberg, 1997; Lundh & Öst, 1996; Mattia, Heimberg, & Hope, 1993;

McNeil et al., 1995; Orsillo, Lilienfeld, & Heimberg, 1994). How can this slowed

Alexander L. Gerlach et al.280

processing be accounted for? The emotional Stroop effect can at least partially be

attributed to increased activation of negative emotion-related identity nodes, which

spreads automatically to related nodes ‘in some sort of semantic recognition’ (White,

1996; p. 206). Sayette (1993b) argues that ‘alcohol disrupts initial appraisal of stressful

information by constraining the spread of activation of associated information previously

established in long-term memory’ ( p. 247). Thus, alcohol can be expected to alter

performance in the emotional Stroop test.

The alcohol myopia theory states that alcohol influences behaviour when

conflicting cues simultaneously influence behaviour ( Josephs & Steele, 1990; Steele &

Josephs, 1988). Stroop interference is the result of a response conflict (e.g. MacLeod,

1991). Thus, according to alcohol myopia theory, ingestion of alcohol will increase the

emotional Stroop effect. Colour naming is the process that is less automatic and less

salient compared with word reading. Since alcohol reduces cognitive control, word

reading that interferes with colour naming should be processed with priority and

consequently an increased interference should occur. Indeed, Curtin and Fairchild

(2003) found an increase in error rate and a tendency for longer reaction times after

ingestion of alcohol (0.08 per mille) in incongruent trials in the colour word version of

the Stroop test.

Alcohol, in addition to potentially preventing appraisal of information, also possesses

powerful amnesic effects (Weissenborn & Duka, 2000). Generally, it is assumed that

alcohol impairs the encoding and storage of new information (Sayette, 1993b).

However, with respect to acute alcohol effects on implicit memory there is very little

and contradictory information. Hashtroudi et al. (1984) found increased performance

on an implicit memory test after consumption of alcohol. Duka et al. (2001) and Lister et

al. (1991) found that alcohol had a profound negative impact on explicit memory but

left implicit memory intact.

During the emotional Stroop test, perceptual priming takes place, resulting in

enhanced implicit memory for emotional stimuli of negative valence ( Rajaram, Srinivas,

and Travers, 2001). Although information-processing models of emotional disorders

suggest that anxious people (e.g. social phobics) have a memory bias for threat-related

stimuli (e.g. Williams & Scott, 1988), empirical support for such a memory bias is weak.

Specifically for social phobia, there is only limited evidence for either an explicit or an

implicit memory bias for social threat stimuli (Coles & Heimberg, 2002). Amir et al.

(2003) argued that methodological issues may have prevented the detection of memory

biases for social threat material in social phobics in previous studies. For example, Rapee

et al. (1994) failed to find an implicit memory bias for social threat words in social

phobics. However, participants were tested not before 15–35 minutes after first

exposure to the stimulus material. In the Lundh and Öst study (1997) that found an

implicit memory bias in non-generalized social phobics, memory performance was

tested only 5 minutes after the first exposure. Graf and Mandler (1984) argued

convincingly that the duration between acquisition and testing is critical if implicit

memory is measured by word-stem completion. In two newer studies using tests for

implicit memory other than word-stem completion, Amir and colleagues demonstrated

implicit memory biases for social threat stimuli in social phobics (Amir et al., 2003; Amir,

Foa, & Coles, 2000). Consequently, we additionally planned to explore whether implicit

memory measured by word-stem completion for social phobia-related words used in our

emotional Stroop test would be enhanced in social phobics and whether this bias would

be affected by alcohol.

Effects of alcohol on processing social threat words 281

Based on these considerations we investigated whether consumption of alcohol

attenuated the effect of social phobia-related words on colour-naming latencies and also

reduced a possible implicit memory bias for such words in social phobics. Since alcohol

has different effects on men and women (Eckardt et al., 1998), only female participants

were tested. In a number of information-processing studies, anxiety induction was

employed in order to activate social threat-relevant schemata. In his 1993 paper, Sayette argued that SRD is unlikely to occur when a stressor is sufficiently threatening to

override appraisal deficits. Consequently, we decided not to use an anxiety-inducing

procedure in order to test the appraisal disruption hypothesis explicitly under

conditions of low threat. It is still likely that social phobics will experience more

evaluation anxiety during the stress test than non-anxious controls.

Method

Recruitment Women who responded to newspaper advertisements seeking women either with or

without fear of social situations and who successfully completed a brief telephone

screening were invited to attend a diagnostic session for approximately 2-hours.

This session included completion of several questionnaires and the German Version of

the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental

Disorders (SCID; Wittchen). Interviews were conducted by clinicians with several years

of experience in treating social phobia patients and who were trained in using the SCID. Control participants received 60 or 40 German marks (DM), depending on whether they

received alcohol or not. Socially-phobic participants were offered the choice of the same

amount of money or participation in a 6-hour workshop dealing with social phobia.

Exclusion criteria for participation in the study were current or past drug or alcohol

abuse or dependence, complete abstinence from alcohol, colour-blindness, use of

psychoactive medication, liver damage, and current or past psychotic episodes.

Participants Forty-two socially phobic patients and 36 control participants took part in the

experiment. Due to equipment failure (malfunction of the throat microphone), data for

only 32 socially phobic and 32 control participants were available. We did not control

the phase of the menstrual cycle of our participants at the time of testing. Table 1 shows

significant differences in the expected direction of social phobia-related measures.

These were the German versions of the Fear of negative evaluation scale (Vormbrock & Neuser, 1983), the Social phobia scale (Stangier, Heidenreich, Berardi, Golbs, & Hoyer,

1999), the Social interaction and anxiety scale (Stangier et al., 1999), Drinking due to

social anxiety scale (‘Trinken wegen sozialer Angst’, Heidenreich, Wagner, & Stangier,

2003), and the Blushing propensity scale (Leary & Meadows, 1991), the Beck

Depression Inventory (Hautzinger, Bailer, Worall, & Keller, 1994), and an Alcohol

Expectancy Questionnaire (Demmel & Hagen, 2002a). On the Alcohol Expectancy

Questionnaire socially-phobic participants reported that they expected more tension-

reduction and regulation of negative mood and more enhanced socio-emotional functioning due to alcohol than the control group. Finally, all participants filled out

the German version of the Short Michigan Alcoholism Screening Test for father and

mother (Demmel & Hagen, 2002b). Based on the cut-off score of 6 as suggested by

Demmel and Hagen (2002b), the percentage of participants with at least one parent

with alcohol problems was calculated.

Alexander L. Gerlach et al.282

All participants gave written informed consent after learning about the experiment,

but prior to randomization. They were then randomized to either the alcohol or the

orange juice condition. Thus, all participants agreed to participate irrespective of

whether they would receive alcohol or not. The randomization procedure resulted in

18 controls and 17 social phobics receiving alcohol, and 14 controls and 15 social

phobics receiving orange juice. Neither the two resulting social phobic groups nor the

two control groups differed on any of the baseline variables. Thus, the randomization

procedure was successful in creating equivalent groups.

Procedure Once randomized, all participants were asked to eat a ‘light meal’, specified in a

hand-out, approximately 3.5 hours before the start of the experiment. They were also

asked to refrain from drinking anything containing caffeine (coffee, tea, soft drinks)

during the 4 hours prior to the experiment, and not to drink alcohol for 24 hours prior

to the experiment.

All participants were informed whether they would receive alcohol or not before

coming to the laboratory. No attempt was made to deceive participants about the nature

Table 1. Comparison of the social phobic and healthy participants

Social phobics (N ¼ 32) Normal controls (N ¼ 32)

Age 32.5 (1.6) a 31.0 (1.6) a % Single 75.0 a 84.4.a Education (% $ high school graduation) 78.1 a 87.5.a % Smokers 59.4 a 35.5 b Beginning of social drinking (age) 17.2 (1.3) a 17.0 (1.2) a Alcohol consumption (grams) during the past week

56.9 (35.9) a 49.0 (30.2) a

Comorbid diagnosis (%) 10 (31.2) 0 (0) G-FNE 63.8 (1.5) a 28.3 (1.5) b BPS 49.1 (2.0) a 29.4 (2.0).b BDI 19.9 (1.1) a 2.5 (1.1) b ASI 33.5 (1.4) a 11.7 (1.4) b SPS 36.6 (2.3) a 3.3 (2.3) b SIAS 39.9 (1.6) a 12.1 (1.6) b TWSA 17.4 (2.4) a 4.0 (2.4) b Parental alcohol problems (%) 18.7 a 12.5 a AEQ_KO 12.6 (0.5) a 14.2 (0.5) b AEQ_SP 12.7 (0.4) a 14.0 (0.4) b

Note. Values enclosed in parentheses represent standard errors where not otherwise noted. G-FNE ¼ German version of the fear of negative evaluation scale; BPS ¼ Blushing propensity scale; BDI ¼ Beck Depression Inventory; ASI ¼ Anxiety Sensitivity Index; SPS ¼ Social phobia scale; SIAS ¼ Social avoidance and distress scale; TWSA ¼ ‘Fragebogen Trinken wegen sozialer Angst’ [Drinking due to social anxiety scale]; AEQ_KO ¼ Alcohol Expectancy Questionnaire, subscale ‘tension-reduction and regulation of negative mood’; AEQ_SP ¼ Alcohol Expectancy Questionnaire, subscale ‘enhanced socio-emotional functioning’. Parental alcohol problems were defined as a score greater than 3 on the German version of the Short Michigan Alcoholism Screening Test. Means for specific questionnaires that do not share a common superscript differ significantly in Mann–Whitney U tests p , :001. Fisher exact tests were used for comparison of percentages.

Effects of alcohol on processing social threat words 283

of the beverage they were about to get. Sayette (1993a) argues convincingly that

appraisal disruption will only take place if a person is pharmacologically sufficiently

intoxicated. Unfortunately, a full balanced placebo design cannot be achieved

successfully if behaviourally-relevant alcohol doses are used. In a thorough study by

Lyvers and Maltzman (1991), 90% of the participants that were lead to believe that they

drank a non-alcoholic beverage but received alcohol (.0.5‰) were able to sense the intoxication resulting from drinking the alcohol. In a study by Sayette et al. (1994) using

alcohol levels of 0.6‰, 94% of their subjects receiving alcohol in the placebo condition

were not deceived.

When the participants receiving alcohol arrived on the test day, a urine sample was

collected for pregnancy testing ( Hilary, Dolorgiet). None of the participants tested

positive. Subsequently, their height and weight was measured. Breath alcohol

concentrations were measured using a standard breathalyzer with an accuracy of

^0.03 mg / L ( Dräger, Alkotester 7410). Test results for all participants upon arrival were 0.00‰. For participants receiving alcohol, the necessary amount of alcohol based

on their weight and height was estimated following a version of the Widmark formula

( Widmark, 1932), modified by Fisher et al. (1987), Kapur, (1991) and Breslin et al.

(1997). We aimed for a blood alcohol concentration of 0.06% based on findings that

people are able to perform the Stroop test without significant performance deficits with

similar blood alcohol levels (BAL; Gustafson & Källmén, 1990a, 1990b).

The alcoholic beverage was one part vodka and two parts orange juice, the non-

alcoholic beverage was juice only in comparable amounts. Participants received their respective beverages in three equal doses, each to be finished within 5 minutes.

All participants were able to complete this procedure. No participant reported nausea

or other feelings of being uncomfortable. At the end of the experiment, participants

who had received alcohol either were given newspapers to read until their BAL reached

less than 0.04‰ or they were fetched by their partners.

Presentation of the stimuli ERTS software ( Beringer, 1994) was used to present the stimuli and to measure

reaction times. Words were presented in a blocked format on a computer screen: half

of the participants were first asked to name the colours of the social phobia-related words and then of the neutral words, whereas the other half of the participants first

had to name the colours of the neutral words and then the colours of the social

phobia-related words. Each word was presented individually. Within blocks, word

order was randomized. Word presentation ended as soon as the colour naming was

registered by a throat microphone. The colours were red, blue, green, and yellow,

randomly chosen for each word presentation. After naming the colour of a word there

was a 1-second interval before the next word was presented. Participants were allowed

a maximum of 3 seconds to name the colour of a word, but no participant needed that much time.

Stimuli and measures The emotional Stroop test was composed of 16 social phobia-related words and 16

neutral words. Social phobia-related words and neutral words were matched based on

word length and frequency of use in the German language. Reaction times for colour

naming were measured with a throat microphone attached with double-coated adhesive

Alexander L. Gerlach et al.284

electrode rings (Marquette Hellige). Estimates of mean reaction times for each stimulus

class were calculated after elimination of values above or below two standard deviations

(Ratcliff, 1993). Although the Stroop test is emotionally demanding, participants rarely

make errors in naming the word colour. To check the prevalence of errors, 11 socially

phobic and 10 control participants (13 under the influence of alcohol) were selected

randomly and their errors were counted using a videotape of the Stroop session. Only 31

errors were made out of the total of 2,016 trials that we checked. There was no

difference in the number of errors between social phobics and controls (Mann–Whitney

U tests: social phobia-related words Z ¼ 20:49, p ¼ :62; neutral words: Z ¼ 0, p ¼ 1:0) or between participants in the alcoholic or non-alcoholic beverage group

(Mann–Whitney U tests: social phobia-related words Z ¼ 20:07, p ¼ :94; neutral words: Z ¼ 1:01, p ¼ :20).

Implicit memory was assessed with a word-stem completion test. Participants were

asked to complete 32 three-letter word stems to the first word that came to their mind,

without a time restriction. Of these 32 word stems, 16 word stems could only be

completed to previously-presented social phobia-related words and 16 word stems

could only be completed to previously-presented neutral words. With the exception of

one neutral word (‘dringend’, English: urgent) and one social phobia-related word

(‘peinlich’, English: embarrassing), all word stems could be completed to more frequent

words other than the primed words. This was ensured using the Corpus Search,

Management, and Analysis System (COSMAS; Belica, Herberger, & al-Wadi, 1992). For the

results presented here, only words that were identical to the ones presented during the

Stroop test were counted. However, more liberal scoring including words that shared

the same root as the words from the emotional Stroop test did not affect the pattern of

findings. Participants could achieve a maximum implicit memory score of 16 for anxiety-

related and 16 for neutral stimuli.

In addition, we also assessed explicit memory for the words employed in the Stroop

task. While the instructions for the Stroop task make it unlikely that the participants will

remember many words explicitly, assessing explicit recollection of the words by free-

recall will allow for the control of effects of explicit recollection on the implicit memory

test. Presentation of the implicit and the explicit memory test was counterbalanced.

We also measured anxiety state using 10-centimetre visual analogue scales.

Participants were asked to rate how anxious they felt ‘right now’ at baseline and after

completing the emotional Stroop test.

Timeline After arrival, all participants were weighed. Participants in the alcohol condition were

also tested for pregnancy. Then everybody was seated in the experimenter room and the

first breath alcohol measurement (BAC 1) was taken. Then participants were asked to

drink their respective beverages within 15 minutes (three cups presented at 0, 5, and 10

minutes). After 5 additional minutes allowed for absorption, the breath alcohol was

measured again (BAC 2). Then all participants performed the Stroop test (5–6 minutes)

and the third breath alcohol measurement was taken (BAC 3) and participants were

asked to report their amount of anxiety (SUDS 1). The two memory tests were presented

in balanced order. Half of the participants were first asked to perform the explicit

(5 minutes) and then the implicit memory test (5 minutes), half of the participants were

asked to first perform the implicit and then the explicit memory test. Finally, breath

Effects of alcohol on processing social threat words 285

alcohol was measured again (BAC 4) and the participants were asked to report their

respective intensity of anxiety (SUDS 2).

Data analysis Colour-naming reaction times were analyzed with repeated measures ANOVAs/MANOVAs

with group (social phobics and controls) and condition (alcohol vs. orange juice) as

between-subjects factors and category (anxiety stimuli vs. neutral stimuli) as the repeated

measurement factor. Planned comparisons were calculated between social phobics and

controls. An alpha level of 0.05 was used for all statistical tests. F-statistics are misleading

when the means are correlated with variances across cells of the design (Winer, Brown, &

Michels, 1991). Therefore, for the comparison of self-report measures of social phobics and

controls, Mann–Whitney U tests were employed.

Results

Blood alcohol concentration and anxiety state Figure 1 shows that the social phobic and the control participants who had received

alcohol reached comparable levels of BAL. There was no difference in blood alcohol

concentration between the two groups (F ¼ 1:0, p ¼ :4). Socially-phobic participants reported significantly more anxiety than controls, Fð1; 60Þ ¼ 17:64, p , :001 (social phobics with alcohol after Stroop test: M ¼ 2:5 (SE: 0.4); after the two memory tests: M ¼ 3:1 (SE: 0.4), social phobics without alcohol after Stroop test: M ¼ 3:4 (SE: 0.4); after the two memory tests: M ¼ 3:0 (SE: 0.4), controls with alcohol after the Stroop test M ¼ 1:3 (SE: 0.4), after the two memory tests: M ¼ 1:1 (SE: 0.4); controls without alcohol after the Stroop test M ¼ 1:6 (SE: 0.4), after the two memory tests:

Figure 1. Blood alcohol concentration in participants that received alcohol.

Note. BAC 1 ¼ blood alcohol concentration at baseline, BAC 2 ¼ blood alcohol concentration

5 minutes after last drink, BAC 3 ¼ blood alcohol concentration after Stroop test, BAC 4 ¼ blood

alcohol concentration after memory test at the end of the session.

Alexander L. Gerlach et al.286

M ¼ 1:3 (SE: 0.5). However, neither consuming alcohol nor performing the emotional Stroop test influenced self-report anxiety level. There was neither an effect of time

(post-Stroop test vs. post-memory tests), condition (alcohol vs. no alcohol), nor an

interaction effect of Group £ Time £ Condition, Fð1; 60Þ # 0:8.

Emotional Stroop test The analysis of reaction times with group (social phobics vs. controls) and condition

(alcohol vs. no alcohol) as a between-subjects factor and category (anxiety stimuli vs.

neutral stimuli) as a within-subjects factor resulted in a significant interaction

Group £ Condition £ Category, Fð1; 60Þ ¼ 4:02, p ¼ :049. Also, a significant main effect of category Fð1; 60Þ ¼ 8:57, p ¼ :005 was found, but no main effect of group or condition or any two-way interaction effects were (see Fig. 2 for an illustration

of the significant effects). The three-way interaction was due to an interaction of

Condition £ Category in the control group, Fð1; 60Þ ¼ 4:92, p ¼ :030, but not in the social phobia group Fð1; 60Þ ¼ 0:37, p ¼ :54). Accordingly, differential effects of alcohol on colour-naming latencies were confined to the control group without alcohol. Controls

responded more slowly to anxiety stimuli than to neutral stimuli, Fð1; 60Þ ¼ 5:08, p ¼ :028, as did the social phobics. With alcohol, colour-naming latencies to anxiety- related words in controls were reduced to the level of colour-naming latencies for neutral

words, Fð1; 60Þ ¼ 0:63, p ¼ :42. In contrast, the social phobia patients took longer to name the colour of the anxiety stimuli compared with neutral stimuli, irrespective of the

beverage consumed, Fð1; 60Þ ¼ 8:89, p ¼ :004). Thus, social phobics show an attentional bias, but not more so than controls. Alcohol reduces this bias in controls,

but not in patients.

An index for the observed interference effect was formed by subtracting the colour-

naming latency for neutral words from the colour-naming latency for anxiety-related

words. This index was significantly correlated (Spearman rank correlation, p , :05)

with the SPS (r ¼ :26), the SIAS (r ¼ :27), and the TWSA (r ¼ :27). The correlation with the German FNE was not significant (r ¼ :17). Neither during baseline nor following the Stroop test did interference correlate with anxiety state (r ¼ 2:06 and r ¼ :01). Finally, alcohol expectancy as measured with the AEQ-KO and AEQ-SP was also not

Figure 2. Emotional Stroop interference depending on stimuli and diagnostic group.

Effects of alcohol on processing social threat words 287

correlated with Stroop interference, neither in the social phobic group (r ¼ :08 AEQ- KO, r ¼ 2:1 AEQ-SP), nor in the control group (AEQ-KO: r , 2:01, AEQSP: r ¼ 2:1).

Explicit memory test As expected, all participants did poorly on the free-recall test. The average number of

correctly-remembered social threat words was 1.8 (SE ¼ 1:19) and 0.19 (SE ¼ 0:06) for neutral words. That is, 27 participants per group (social phobics and controls) did not

remember one single neutral word, four participants in each group remembered one

word and one participant in each group remembered two neutral words. Seven social

phobics and six controls remembered no social threat-related word, 11 social phobics

and 10 controls remembered one social threat-related word, five social phobics and six

controls remembered two social threat-related words, and nine social phobics and 10

controls remembered three or more social threat words. Based on these low rates no

statistical analyses were computed.

Implicit memory test The ANOVA of the implicit memory scores with category (anxiety stimuli vs. neutral

stimuli) as a within-subjects factor, group (social phobics vs. controls) and condition

(alcohol vs. no alcohol) as between-subjects factors, resulted in a significant effect

of category, Fð1; 60Þ ¼ 5:93, p , :05, and a trend for group, Fð1; 60Þ ¼ 3:72, p ¼ :06, which was moderated by a Group £ Category effect, Fð1; 60Þ ¼ 5:43, p , :05. (see Fig. 3 for an illustration of the significant effects). The socially-phobic participants remembered more socially negative words than neutral words, Fð1; 60Þ ¼ 11:42, p , :05. Furthermore, social phobics remembered more social threat words than controls, Fð1; 60Þ ¼ 7:01, p , :05. There were no differences between social phobics and control participants on implicit memory for neutral words, Fð1; 60Þ ¼ 0:54, ns.

In addition, we found a tendency for an interaction of Group £ Condition £

Category, Fð1; 60Þ ¼ 3:04, p # :09. In order to evaluate the tendency for a three-fold interaction, we analyzed the Condition £ Category interaction in the social phobia

group only and found a significant effect, Fð1; 60Þ ¼ 4:54, p , :05. Hence, social

Figure 3. Implicit memory values depending on stimuli and diagnostic group.

Alexander L. Gerlach et al.288

phobics showed a significantly lesser implicit memory bias for social threat words if they

were drinking alcohol. The implicit memory bias for social threat (memory scores for

anxiety-related words minus memory score for neutral words) was significantly

correlated (Spearman–Brown) with the SIAS, r ¼ :31, tð62Þ ¼ 2:5, p , :01, and the SPS, r ¼ :29, tð62Þ ¼ 2:1. There was also a trend for the TWSA, r ¼ :27, tð62Þ ¼ 1:8, p , :08, and the German FNE, r ¼ :21, to be correlated with implicit memory bias for social threat words. Thus, a higher self-rating on measures of social anxiety tended to be

related to better retention of social anxiety-related words.

Discussion

We used an emotional Stroop procedure with neutral and social anxiety-related word

stimuli and an implicit memory test for these words to assess the effect of alcohol on cognitive processing in social phobics as compared with non-anxious controls.

We hypothesized an information-processing bias in social phobics, evident in longer

colour-naming latencies and a better implicit memory for social threat words. Based on

functional analysis, considerations of the link between social phobia and alcohol

consumption and on the core proposition of Sayette’s appraisal disruption theory, we

further expected that alcohol would have a stronger effect on social phobics in the sense

of reducing both the interference produced by anxiety stimuli and their enhanced

memory effect. Alternative expectations were derived from the alcohol myopia theory, namely that alcohol should increase Stroop interference. In line with these

considerations, social phobia patients did take longer to name the colour of anxiety

words than of neutral words, and they remembered more anxiety words than the

controls. But contrary to our expectation, alcohol attenuated the effect of anxiety-

related word stimuli on colour-naming latencies only in non-anxious controls, but not in

social phobics, and it did not affect memory performance. Hence, we disconfirmed the

alcohol myopia theory and at least partially failed to confirm the appraisal disruption

theory as applied to social phobia patients. Anxiety stimuli provoked an interference effect in colour-naming responses, in line

with the literature on emotional Stroop effects. But why did this interference effect

occur in controls also, and why was the effect not greater in social phobics than in

controls? Interference effects for social threat words in non-anxious controls have been

repeatedly reported (Amir et al., 1996; Mattia et al., 1993). Emotionally-salient words

generally cause more interference than neutral words (MacLeod, 1991), and it seems

reasonable to assume that social threat words also qualify as emotionally-salient words

for non-anxious controls. Also, only the anxiety-related words were semantically related, whereas the neutral words were not. In a blocked presentation, this could result in

increased interference (Waters, Sayette, & Wertz, 2003).

These considerations may explain the interference in controls in general, but Mattia

et al. (1993), in contrast to our findings and the findings of Amir et al. (1996), found that

the interference effect was nevertheless stronger in social phobics than in controls.

We have no ready explanation for the lack of a similar group difference in our colour-

naming latencies, but we will consider several possibilities drawn from the literature on

emotional Stroop effects. One possible explanation might be taken from Amir et al. (2002), who suggested that

social phobics can inhibit emotional Stroop interference if they have the opportunity to

strategically influence their reactions. When ‘opportunity’ was operationalized as a low

or a high ratio of threat words, social phobics showed more emotional interference with

Effects of alcohol on processing social threat words 289

rare than with frequent social threat words (Amir et al., 2002). On the other hand,

blocked presentation of threat words increases interference (Holle et al., 1997),

although one could argue with Amir et al., that in a blocked format, patients have the

perfect opportunity to strategically influence their reactions. We attempted to examine

these intriguing possibilities by calculating interference effects separately for the first,

second, third, and fourth quarters of the block of threat words. However, no interaction of time (first, second, third, and fourth quarters) with category (social threat vs. neutral

words) and/or with condition (alcohol vs. no alcohol) was observed in the social phobic

groups. In other words, interference was stable throughout the duration of the blocked

presentation, which supports neither the hypothesis of Amir et al., nor the notion that

interference will increase with increasing confrontation (blocked format).

We should also consider the possibility that particular aspects of our social anxiety

challenge condition might have hindered the expected group difference to appear.

Performance anxiety may attenuate the emotional Stroop interference in social phobics

(Amir et al., 1996) and our socially-phobic participants did report more anxiety than controls. But the levels of anxiety were generally low (3 on a scale of 0–10). Also, level of

acute anxiety was not correlated with emotional Stroop interference. Thus, emotional

override cannot explain the lack of a higher interference effect in social phobics.

We found that alcohol did attenuate the emotional Stroop interference in controls

but not in social phobics. According to the core assumption of the appraisal disruption

theory, alcohol should constrain the spread of activation of associated information and

thus attenuate the emotional Stroop interference produced by presenting words related

to social anxiety. Sayette, Martin, Perrott, Wertz, and Hufford (2001) have shown this

effect of alcohol in a non-anxious control group. Specifically, they demonstrated that alcohol, if ingested before conducting the emotional Stroop test, reduces the

interference for social stress words. This finding has been replicated in our study.

But why was this not true for social phobics as well, and what does this imply for the

functional analysis of the relation between social phobia and a tendency to alcohol

abuse and dependence?

It can be argued that the spread of social threat-related activation is much better

organized in social phobics than in non-anxious controls. Sayette (1993b) argued that

alcohol is less likely to hinder spread of activation if a stressor is ‘relatively easy to

process’ or ‘sufficiently threatening to override appraisal deficits’ (p. 469). It seems conceivable that, for social phobia patients, not only indicators of acute stressors, but

also social threat words, may either be easier to process or are associated strongly

enough with impending danger so that alcohol may no longer effectively reduce spread

of activation. However, in this context our second, exploratory finding is of interest.

We found a significantly better implicit memory for social threat words in social phobics

compared with controls. Furthermore, this memory bias was reduced by alcohol in the

expected direction.

Generally, it is assumed that two different processes are responsible for implicit

memory biases. For the word-stem completion task it is assumed that more data-driven (or perceptually driven) processes are responsible. In this study, however, it is clear that

conceptually-driven processes must be involved in order to produce a valence-

associated (or threat-associated) bias. Indeed, at least four prior studies have

demonstrated that conceptual/associative processing on perceptual implicit memory

tests can be independent of explicit memory processes (e.g. Hirshman, Passannante, &

Arndt, 2001). Note that participants had almost no traces of explicit memory for either

neutral or social threat words. Thus, it is highly unlikely that explicit recollection of

Alexander L. Gerlach et al.290

social threat words may have contaminated performance on the stem completion task.

Furthermore, due to our balanced design only half of the participants were first asked to

consciously recollect the Stroop words. It is generally acknowledged that implicit

memory processes play an important role in ‘human affairs’ (Tulving & Schacter, 1990).

These authors proposed that ‘conceptually driven priming reflects a process of semantic

learning: the modification of, or adding of new information to, semantic memory’ (p. 304). Consequently, our results suggest that, in clinically-anxious social phobics, not

only is appraisal hindered by alcohol, but elaboration after initial processing also is also

affected. If that notion is correct, a threatening situation entered by a social phobic after

drinking alcohol may be experienced with similar anxiety with and without alcohol, but

this anxiety is remembered less intensively, thus still encouraging the use of alcohol as

self-medication in future situations. This may partly explain why studies focusing on

direct effects of alcohol on anxiety in clinically socially anxious people have repeatedly

failed to find significant anxiety-reducing effects of alcohol (Abrams et al., 2001; Himle

et al., 1999; Naftolowitz et al., 1994). It has been found that alcohol has a different effect on anxiety in women compared

with men. Especially alcohol expectancy – the belief that one has received alcohol and

that it will have specific effects – has been studied in this respect. For example, de Boer

et al. (1993) showed that alcohol expectancy reduced social anxiety in women but not

in men. However, paradoxical effects of alcohol expectancy have been demonstrated in

women as well: for example, Abrams and Wilson (1979) found that women were more

anxious during a social interaction test if they believed that they had received alcohol.

Hence, findings regarding the effects of alcohol expectancy on women are inconclusive

(Schippers, de Boer, van der Staak, & Cox, 1997). However, in the only experimental study testing the direct effects of alcohol on anxiety involving socially-phobic

participants, alcohol expectancy reduced reported social anxiety while giving a public

speech (Abrams et al., 2001). We did not directly test whether the pharmacological

effects of alcohol were solely responsible for our findings or if alcohol expectancy or

other psychological factors may have additionally led to the reduction of the emotional

Stroop interference in controls. Our participants did know whether they received

alcohol or orange juice. The socially-phobic women in our sample reported that they

expected more tension-reduction and regulation of negative mood and also more

enhanced socio-emotional functioning after drinking alcohol (as measured with the AEQ) than the control participants. Possibly this may have modulated the effects of

alcohol in some way or another. We did not find any association between our measure of

alcohol expectancy and Stroop interference. However, it may be that alcohol

expectancies are a transient- or an affect-dependent form of cognition and consequently

have to be measured in vivo rather than at baseline.

There are a number of other limitations of our study. Based on ethical considerations

we recruited social drinkers without a history of abuse or dependence. It is possible that

this may have led to recruitment of a group of participants who are especially unlikely to

experience anxiety reduction by ingestion of alcohol. Also, we have only indirect information on the alcohol status of the relatives of our participants. Paternal alcoholism

has been associated with an increased risk to develop alcohol problems and is

also known to affect the propensity to use alcohol to dampen stress responses

(Finn & Pihl, 1987). We cannot exclude the possibility that our recruitment procedure

implicitly excluded a subgroup that could potentially react differently to alcohol. Also,

we exclusively recruited women. Thus, our results are limited to female social phobics.

Finally, whereas a full balanced placebo design may not be suitable if BAL above 0.6‰

Effects of alcohol on processing social threat words 291

(Sayette et al., 1994; Lyvers & Maltzman, 1991), our design would clearly benefit from at

least a third cell with participants receiving placebo but believing that they receive

alcohol. Future studies consequently should include such a group in order to further

explore the contribution of alcohol expectancies.

In summary, we were able to confirm the notion that alcohol disrupts appraisal

of threatening information in non-anxious women. This effect was not observable

in socially-phobic women, which supports the notion that highly effective processing

of threatening information lessens such an effect of alcohol. In addition, we found an

implicit memory bias for socially-threatening words in social phobics as compared with

controls. This implicit memory bias was attenuated by alcohol. The overlap of excessive

alcohol use and social phobia may not be the result of appraisal disruption exclusively,

but also the result of a curtailed memory for the anxiety experienced during social

situations.

Acknowledgements

Alexander L. Gerlach and Anke Schiller have equally contributed to this study and are listed in

alphabetical order. We thank Ralf Demmel and Michael Sayette for comments and

recommendations in the planning phase of this study. This research was supported by the

Christoph Dornier Foundation and the Christoph Dornier Clinic. Portions of this study were

presented at the XXXIII Annual Congress of the EABCT, Prague, The Czech Republic, September

2003.

References

Abrams, D. B., & Wilson, G. T. (1979). Effects of alcohol on social anxiety in women: Cognitive

versus physiological processes. Journal of Abnormal Psychology, 88, 161–173.

Abrams, K., Kushner, M. G., Medina, K. L., & Voight, A. (2001). The pharmacologic and

expectancy effects of alcohol on social anxiety in individuals with social phobia. Drug and

Alcohol Dependence, 64, 219–231.

Abrams, K., Kushner, M. G., Medina, K. L., & Voight, A. (2002). Self-administration of alcohol

before and after a public speaking challenge by individuals with social phobia. Psychology of

Addictive Behaviors, 16, 121–128.

Allan, C. A. (1995). Alcohol problems and anxiety disorders – a critical review. Alcohol and

Alcoholism, 30, 145–151.

Amir, N., Bower, E., Briks, J., & Freshman, M. (2003). Implicit memory for negative and positive

social information in individuals with and without social anxiety. Cognition and Emotion, 17,

567–583.

Amir, N., Foa, E. B., & Coles, M. E. (2000). Implicit memory bias for threat-relevant information in

individuals with generalized social phobia. Journal of Abnormal Psychology, 109, 713–720.

Amir, N., Freshman, M., & Foa, E. (2002). Enhanced Stroop interference for threat in social phobia.

Journal of Anxiety Disorders, 16, 1–9.

Amir, N., McNally, R. J., Riemann, B. C., Burns, J., Lorenz, M., & Mullen, J. T. (1996). Suppression of

the emotional Stroop effect by increased anxiety in patients with social phobia. Behaviour

Research and Therapy, 34, 945–948.

Becker, E. S., Rinck, M., Margraf, J., & Roth, W. T. (2001). The emotional Stroop effect in anxiety

disorders: General emotional or disorder specificity? Journal of Anxiety Disorders, 15,

147–159.

Belica, C., Herberger, A., & al-Wadi, D. (1992). COSMAS. Linguistische Datenverarbeitung.

Mannheim: Institut für deutsche Sprache.

Alexander L. Gerlach et al.292

Beringer, J. (1994). ERTS: A flexible software tool for developing and running psychological

reaction time experiments on IBM PCs. Behavior Research Methods, Instruments, and

Computers, 26, 368–369.

Breslin, F. C., Kapur, B. M., Sobell, M. B., & Cappell, H. (1997). Gender and alcohol dosing:

A procedure for producing comparable breath alcohol curves for men and women.

Alcoholism: Clinical and Experimental Research, 21, 928–930.

Clark, D. B., & Sayette, M. A. (1993). Anxiety and the development of alcoholism. American

Journal on Addictions, 59–76.

Coles, M. E., & Heimberg, R. G. (2002). Memory biases in the anxiety disorders: Current status.

Clinical Psychology Review, 22, 587–627.

Conger, J. J. (1956). Alcoholism theory, problem and challenge: II. Reinforcement theory and the

dynamics of alcoholism. Quarterly Journal of Studies on Alcohol, 13, 296–305.

Cox, B. J., Swinson, R. P., Direnfeld, D. M., & Bourdeau, D. (1994). Social desirability and self-

reports of alcohol abuse in anxiety disorder patients. Behaviour Research and Therapy, 32,

175–178.

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, 1693–1700.

Curtin, J. J., & Fairchild, B. A. (2003). Alcohol and cognitive control: Implications for regulation of

behavior during response conflict. Journal of Abnormal Psychology, 112, 424–436.

de Boer, M. C., Schippers, G. M., & Van Der Staak, C. P. (1993). Alcohol and social anxiety in

women and men: Pharmacological and expectancy effects. Addictive Behaviours, 18,

117–126.

Demmel, R., & Hagen, J. (2002a). Faktorenstruktur und psychometrische Eigenschaften einer

gekürzten deutschsprachigen Version des Alcohol Expectancy Questionnaire (Brief AEQ-G).

Zeitschrift für Differentielle und Diagnostische Psychologie, 23, 205–216.

Demmel, R., & Hagen, J. (2002b). Screening for paternal alcoholism: Reliability of a German

version of the F-SMAST. European Addiction Research, 8, 128–132.

Duka, T., Weissenborn, R., & Dienes, Z. (2001). State-dependent effects of alcohol on recollective

experience, familiarity and awareness of memory. Psychopharmacology, 153, 295–306.

Eckardt, M. J., File, S. E., Gessa, G. L., Grant, K. A., Guerri, C., Hoffman, P. L., et al. (1998). Effects of

moderate alcohol consumption on the central nervous system. Alcoholism: Clinical and

Experimental Research, 22, 998–1040.

Finn, P. R., & Pihl, R. O. (1987). Men at high risk for alcoholism: The effect of alcohol on

cardiovascular response to unavoidable shock. Journal of Abnormal Psychology, 96,

230–236.

Fisher, H. R., Simpson, R. I., & Kapur, B. M. (1987). Calculation of blood alcohol concentration

(BAC) by sex, weight, number of drinks and time. Canadian Journal of Public Health, 78,

300–304.

Graf, P., & Mandler, G. (1984). Activation makes words more accessible, but not necessarily more

retrievable. Journal of Verbal Learning and Verbal Behavior, 23, 553–568.

Gustafson, R., & Källmén, H. (1990a). Effects of alcohol on cognitive performance measured with

stroop’s color word test. Perceptual and Motor Skills, 71, 99–105.

Gustafson, R., & Källmén, H. (1990b). Effects of alcohol on prolonged cognitive performance

measured with stroop’s color word test. Psychological Reports, 67, 643–650.

Hashtroudi, S., Parker, E. S., Delisi, L. E., Wyatt, R. J., & Mutter, S. A. (1984). Intact retention in

acute alcohol amnesia. Journal of Experimental Psychology-Learning Memory and

Cognition, 10, 156–163.

Hautzinger, M., Bailer, M., Worall, H., & Keller, F. (1994). Beck-depressions-inventar (BDI).

Zeitschrift für Klinische Psychologie, 26, 162–163.

Heidenreich, T., Wagner, C., & Stangier, U. (2003). Skala Trinken wegen sozialer Angst und Skala

Soziale Angst wegen Trinken. In J. Hoyer & J. Margraf (Eds.), Angstdiagnostik Grundlagen

und Testverfahren ( pp. 284–287). Berlin: Springer.

Effects of alcohol on processing social threat words 293

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, 1237–1243.

Himle, J. A., & Hill, E. M. (1991). Alcohol abuse and the anxiety disorders: Evidence from the

epidemiological catchment area survey. Journal of Anxiety Disorders, 5, 237–245.

Hirshman, E., Passannante, A., & Arndt, J. (2001). Midazolam amnesia and conceptual processing

in implicit memory. Journal of Experimental Psychology: General, 130, 453–465.

Holle, C., Heimberg, R. G., Sweet, R. A., & Holt, C. S. (1995). Alcohol and caffeine use by social

phobics: An initial inquiry into drinking patterns and behavior. Behaviour Research and

Therapy, 33, 561–566.

Holle, C., Neely, J. H., & Heimberg, R. G. (1997). The effects of blocked versus random

presentation and semantic relatedness of stimulus words on response to a modified Stroop

Task among social phobics. Cognitive Therapy and Research, 21, 681–697.

Josephs, R. A., & Steele, C. M. (1990). The 2 faces of alcohol myopia – Attentional mediation of

psychological stress. Journal of Abnormal Psychology, 99, 115–126.

Kapur, B. M. (1991). CBAC: Computerized blood alcohol concentration – A computer model as a

clinical and an educational tool. Annales de biochimie clinique du Québec, 30, 36–39.

Leary, M. R., & Meadows, S. (1991). Predictors, elicitors, and concomitants of social blushing.

Journal of Personality and Social Psychology, 60, 254–262.

Lister, R. G., Gorenstein, C., Risherflowers, D., Weingartner, H. J., & Eckardt, M. J. (1991).

Dissociation of the acute effects of alcohol on implicit and explicit memory processes.

Neuropsychologia, 29, 1205–1212.

Lundh, L. G., & Öst, L. G. (1996). Stroop Interference, self-focus and perfectionism in social

phobia. Personality and Individual Differences, 20, 725–731.

Lundh, L. G., & Öst, L. G. (1997). Explicit and implicit memory bias in social phobia. The role of

subdiagnostic type. Behaviour Research and Therapy, 35, 305–317.

Lyvers, M. F., & Maltzman, I. (1991). The balanced placebo design – Effects of alcohol and beverage

instructions cannot be independently assessed. International Journal of the Addictions, 26,

963–972.

MacLeod, C. M. (1991). Half a century of research on the stroop effect: An integrative review.

Psychological Bulletin, 109, 163–203.

Mattia, J. I., Heimberg, R. G., & Hope, D. A. (1993). The revised stroop color-naming task in social

phobics. Behaviour Research and Therapy, 31, 305–313.

McNeil, D. W., Ries, B. J., Taylor, L. J., Boone, M. L., Carter, L. E., Turk, C. L., et al. (1995).

Comparison of social phobia subtypes using stroop tests. Journal of Anxiety Disorders, 9,

47–57.

Naftolowitz, D. F., Vaughn, B. V., Ranc, J., & Tancer, M. E. (1994). Response to alcohol in social

phobia. Anxiety, 1, 96–99.

Nardi, A. E., & Versiani, M. (1997). Social phobia and alcohol abuse: Comorbidity predicting poor

therapeutic response. Biological Psychiatry, 42, 238S.

Orsillo, S. M., Lilienfeld, S. O., & Heimberg, R. G. (1994). Social phobia and response to challenge

procedures: Examining the interaction between anxiety sensitivity and trait anxiety. Journal

of Anxiety Disorders, 8, 247–258.

Page, A., & Andrews, G. (1996). Do specific anxiety disorders show specific drug problems?

Australian and New Zealand Journal of Psychiatry, 30, 410–414.

Rajaram, S., Srinivas, K., & Travers, S. (2001). The effects of attention on perceptual implicit

memory. Memory and Cognition, 29, 920–930.

Rapee, R. M., McCallum, S. L., Melville, L. F., Ravenscroft, H., & Rodney, J. M. (1994). Memory bias

in social phobia. Behaviour Research and Therapy, 32, 89–99.

Ratcliff, R. (1993). Methods for dealing with reaction time outliers. Psychological Bulletin, 114,

510–532.

Sayette, M. A. (1993a). An appraisal-disruption model of alcohol’s effects on stress responses in

social drinkers. Psychological Bulletin, 114, 459–476.

Alexander L. Gerlach et al.294

Sayette, M. A. (1993b). An appraisal-disruption model of alcohol’s effects on stress responses in

social drinkers. Psychological Bulletin, 114, 459–476.

Sayette, M. A., Breslin, F. C., Wilson, G. T., & Rosenblum, G. D. (1994). An evaluation of the

balanced placebo design in alcohol administration research. Addictive Behaviors, 19,

333–342.

Sayette, M. A., Martin, C. S., Perrott, M. A., Wertz, J. M., & Hufford, M. R. (2001). A test of the

appraisal-disruption model of alcohol and stress. Journal of Studies on Alcohol, 62, 247–256.

Schippers, G. M., de Boer, M. C., van der Staak, C. P. F., & Cox, W. M. (1997). Effects of alcohol and

expectancy on self-disclosure and anxiety in male and female social drinkers. Addictive

Behaviors, 22, 305–314.

Schneier, F. R., Martin, L. Y., Liebowitz, M. R., Gorman, J. M., & Fyer, A. J. (1989). Alcohol abuse in

social phobia. Journal of Anxiety Disorders, 3, 15–23.

Stangier, U., Heidenreich, T., Berardi, A., Golbs, U., & Hoyer, J. (1999). Die Erfassung sozialer

Phobie durch die Social Interaction Anxiety Scale (SIAS) und die Social Phobia Scale (SPS).

Zeitschrift für Klinische Psychologie, 28, 28–36.

Steele, C. M., & Josephs, R. A. (1988). Drinking your troubles away: 2. An attention-allocation

model of alcohols effect on psychological stress. Journal of Abnormal Psychology, 97,

196–205.

Stockwell, T., & Bolderston, H. (1987). Alcohol and phobias. British Journal of Addiction, 82,

971–979.

Tulving, E., & Schacter, D. L. (1990). Priming and human-memory systems. Science, 247, 301–306.

Vormbrock, F., & Neuser, J. (1983). Konstruktion zweier spezifischer Trait-Fragebogen zur

Erfassung von Angst in Sozialen Situationen (SANB und SVSS). Diagnostica, 29, 165–182.

Waters, A. J., Sayette, M. A., & Wertz, J. M. (2003). Carry-over effects can modulate emotional

Stroop effects. Cognition and Emotion, 17, 501–509.

Weissenborn, R., & Duka, T. (2000). State-dependent effects of alcohol on explicit memory:

The role of semantic associations. Psychopharmacology, 149, 98–106.

White, M. (1996). Automatic affective appraisal of words. Cognition and Emotion, 10, 199–211.

Widmark, E. M. P. (1932). Die theoretischen Grundlagen und die praktische Verwendbarkeit der

gerichtlich-medizinischen Alkoholbestimmung. Berlin: Urban and Schwarzenberg.

Williams, J. M., & Scott, J. (1988). Autobiographical memory in depression. Psychological

Medicine, 18, 689–695.

Winer, B. J., Brown, D. R., & Michels, K. M. (1991). Statistical principles in experimental design

(3rd ed.). New York: McGraw-Hill.

Received 20 July 2004; revised version received 02 March 2005

Effects of alcohol on processing social threat words 295