Journal of Abnormal Psychology © 2009 American Psychological Association

2009, Vol. 118, No. 1, 28 –33 0021-843X/09/$12.00 DOI: 10.1037/a0012589

Attention Modification Program in Individuals With Generalized

Anxiety Disorder

Nader Amir Courtney Beard

San Diego State University and University of California, University of Georgia
San Diego

Michelle Burns Jessica Bomyea

University of Georgia San Diego State University and University of California,
San Diego

Research suggests that individuals with generalized anxiety disorder (GAD) show an attention bias for
threat-relevant information. However, few studies have examined the causal role of attention bias in the
maintenance of anxiety and whether modification of such biases may reduce pathological anxiety
symptoms. In the present article, the authors tested the hypothesis that an 8-session attention modification
program would (a) decrease attention bias to threat and (b) reduce symptoms of GAD. Participants
completed a probe detection task by identifying letters (E or F) replacing one member of a pair of words.
The authors trained attention by including a contingency between the location of the probe and the
nonthreat word in one group (Attention Modification Program; AMP) and not in the other (attention
control condition; ACC). Participants in the AMP showed change in attention bias and a decrease in
anxiety, as indicated by both self-report and interviewer measures. These effects were not present in the
ACC group. These results are consistent with the hypothesis that attention plays a causal role in the
maintenance of GAD and suggest that altering attention mechanisms may effectively reduce anxiety.

Keywords: attention training program for GAD, attention bias, anxiety, information processing

Researchers have used a wide range of methods borrowed from presence of the probe by pressing a button. MacLeod et al. (1986)
cognitive psychology to examine attention bias to threat in indi- found that individuals with GAD detect probes that replace threat
viduals with generalized anxiety disorder (GAD; MacLeod, words, either in the upper or in the lower portion of the screen,
Mathews, & Tata, 1986; Mathews & MacLeod, 1985, 1986; Mogg, faster than probes that replace neutral words. Thus, clinically
Millar, & Bradley, 2000). Research using these methods has con- anxious individuals with GAD consistently showed an attention
sistently produced evidence that patients with GAD preferentially bias toward threat. However, nonanxious controls tended to dem-
attend to threat-relevant stimuli over neutral stimuli when the two onstrate an attention bias away from threat in this paradigm. In a
compete for processing priority. later replication of this study, MacLeod and Mathews (1988)
In a seminal study, MacLeod et al. (1986) developed the probe calculated an attention bias score in this paradigm by subtracting
detection paradigm to measure attention bias to threat in GAD. In the mean response latency for trials in which the probe replaced
this paradigm, participants see two words, one above the other, on the threat stimuli from the mean response latency for trials in
a computer screen. One word is neutral (e.g., table) and the other which the probe replaced the neutral stimuli, such that larger
word has a threatening meaning (e.g., disease). Participants are numbers revealed greater bias for threat. Using this index, these
asked to read the upper word and ignore the lower word. On authors again found that individuals with GAD show an attention
critical trials (25%), either the upper or the lower word is replaced bias toward threat.
with a dot probe ( 䡠 ) and participants are asked to signal the Two recent reviews of attention bias in anxiety provide clear
evidence for an attention bias for threat in GAD. Mogg and
Bradley (2005) reviewed 10 studies using the probe detection task
Nader Amir and Jessica Bomyea, Department of Psychology, San Diego and concluded that individuals with GAD show an attention bias
State University; Courtney Beard and Michelle Burns, Department of for threat that is absent in nonanxious controls. A second meta-
Psychology, University of Georgia. analysis conducted by Bar-Haim, Lamy, Pergamin, Bakermans-
This study was supported by National Institutes of Health Grants R34 Kranenberg, and van IJzendoorn (2007) examined attention bias
MH073004-01 and R34 MH077129-01 awarded to Nader Amir. We thank
across 172 studies (N ⫽ 2,263 anxious, N ⫽ 1,768 nonanxious),
Heide Klumpp, Jason Elias, Allie Osborn, and Sarah Wilson for their help
concluding that this bias is a consistent and reliable finding using
in data collection.
Correspondence concerning this article should be addressed to Nader a variety of paradigms. In summary, there is reason to believe that
Amir, Joint Doctoral Program in Clinical Psychology, San Diego State individuals with GAD have an attention bias toward threat-
University/University of California, San Diego, San Diego, CA 92120- relevant information that is absent or less pronounced in nonanx-
4913. E-mail: [email protected] ious individuals.
28
 SPECIAL SECTION: ATTENTION TRAINING IN GAD 29

However, most existing research has not allowed the examina- mediating vulnerability to negative affectivity. At the practical
tion of the causal relationship between attention and GAD because level, it may be possible to use such attention training procedures
correlational designs have been used in these studies. Conclusions clinically to ameliorate anxiety symptoms.
regarding the causal role of attention bias in maintaining anxiety Although the results for the above study are consistent with the
can only be gleaned from research designs in which participants hypothesis that change in attention bias can lead to change in
are randomly assigned to conditions, and their attention is exper- anxiety vulnerability, several issues need further examination.
imentally manipulated. We now turn to studies using this design. First, because the two conditions in the above studies both actively
MacLeod, Rutherford, Campbell, Ebsworthy, and Holker (2002) trained attention (i.e., either toward threat or away from threat), it
selected 64 undergraduate students from a large participant pool is not possible to determine the effect of each training session
who scored in the middle third of the distribution of a self-report compared with a baseline condition without training contingencies.
measure of trait anxiety. These participants were then randomly Thus, in the present study, we compared the effects of an Attention
assigned to one of two probe detection tasks that were designed to Modification Program (AMP) with a baseline condition in which
train attention. Each task comprised 672 trials in which pairs of there was no contingency between the location of the probe and the
words (one threat, one neutral) were presented, one above the location of the threat or neutral information. We predicted that the
other, on a computer screen. In the attend threat condition, probes AMP would lead to decreased attention to threat and anxiety
appeared in the position of the threat word on 576 training trials. symptoms compared with the attention control condition (ACC).
The remaining 96 trials were designed to provide a measure of Second, although effective attention training procedures in non-
attention bias to threat words. In these test trials, threat word patient samples has been demonstrated in prior research, research-
position and probe position were fully crossed as in a typical probe ers have not examined the role of attention training in clinical
detection task, thus permitting measurement of a participant’s populations. Therefore, we sought to extend attention training
tendency to attend preferentially to threat-relevant or neutral procedures to populations with clinical levels of anxiety (i.e.,
words. In the attend neutral condition, probes appeared in the GAD). Application of information-processing bias modification to
position of the neutral word on 576 of the trials, with the remaining alleviating anxiety symptoms is important because a substantial
96 trials again providing a measure of attention bias. Participants portion of individuals with GAD presenting for treatment do not
were asked to indicate which type of probe (i.e., single dot or a respond to current therapies (psychotherapy: 52%, Fisher &
double dot) had appeared in each trial by pressing a corresponding Durham 1999; medication: 43%, Gorman, 2003), and for many,
button as rapidly and accurately as they could. Following the the most effective treatments are unavailable or difficult to access.
training task and a brief (4-min) rest, MacLeod et al. manipulated Although researchers have established a relationship between
the participants’ level of stress by presenting them with a series of GAD and attention bias to threatening information, this knowledge
unsolvable anagrams and telling them that video recordings of has yet to be translated into effective treatments for this disorder,
participants who performed particularly well or poorly would be thus making examination of such training procedures informative
shown to other students. Results indicated that after training, for advancing available treatment options.
participants in the attend threat condition showed faster response Finally, no study has examined attention training procedures
latencies for detecting probes that replaced threat words than with materials specific to each individual’s perception of threat.
probes that replaced neutral words. Participants in the attend Due to the varied nature of concerns for individuals with GAD, we
neutral condition showed the opposite pattern of results. Moreover, asked each participant to select the words most relevant to his or
this training extended to word pairs containing novel threat- her own concern. In summary, in the present study, we examined
relevant words and was not confined to specifically trained word the effect of a multiple-session attention training program similar
pairs. More important, participants in the attend threat condition to that described by MacLeod et al. (2002) on anxiety in individ-
reported a greater elevation of negative emotion in response to an uals with GAD.
experimental stressor.
MacLeod et al. (2002) successfully replicated in their second
Method
study the findings of their first study. During this second study,
participants’ levels of negative affect were measured prior to Participants
attention training as well as subsequent to training. This modifi-
cation provided a baseline against which the effects of attention Participants were 29 treatment-seeking individuals meeting the
training could be compared. Results again showed that participants diagnostic criteria for GAD, based on a diagnostic interview using
in the attend threat condition reported greater elevation of negative the Structured Clinical Interview for the DSM–IV (SCID-IV; First,
emotion in response to the experimental stressor than did those in Spitzer, Gibbon, & Williams, 1994). Participants were included in
the attend neutral condition. Groups did not differ in their levels of the study if they (a) had a principal Diagnostic and Statistical
negative affect before the training procedure or after training prior Manual of Mental Disorders, 4th edition (DSM–IV; American
to the stressor. Thus, the difference between the two groups Psychiatric Association [APA], 1994) Axis I diagnosis of GAD;
appears to reflect the creation of differing affective vulnerability to (b) showed no evidence of suicidal intent; (c) showed no evidence
stress that is manifested only after the presentation of the stressor. of current substance abuse; (d) had no evidence of current or past
MacLeod et al. (2002) suggested that their findings have poten- schizophrenia, bipolar disorder, or organic mental disorder; (e)
tially important theoretical and practical implications. At the the- were not currently participating in CBT; and (f) had no change in
oretical level, their results provide the strongest support to date for other psychosocial treatments or medication during the 12 weeks
the hypothesis that individual differences in the allocation of prior to study entry. Participants were asked at postassessment
attention to threat-relevant information are causally important in whether they had obtained any additional form of treatment during
30 AMIR, BEARD, BURNS, AND BOMYEA

the study. None reported any additional treatment, and no partic- general anxiety (Sets A and B) and two sets of 48 matched neutral
ipant dropped from the study. words. In the present study, half the participants in each group saw
a particular word set during training (Set A) and were then tested
Materials and Tasks pretraining and posttraining using the other word set (Set B). Thus,
the tests of attention were conducted on a different set of words
Clinician-rated measures, self-report measures, and information than the one used during training. To ensure the relevance of the
processing measures were administered before the first training particular words used during training for each participant, an idio-
session (pretraining) and after the final training session (posttrain- graphic material selection procedure was used. Prior to training, each
ing). Clinician ratings were made by raters blind to treatment participant was asked to rate the emotionality (⫺3 to ⫹3) of each of
condition. The interviewer measures included the SCID Anxiety
the words from the two sets. Twenty words that were rated as most
Disorders Module (First et al., 1994), the Hamilton Rating Scale
emotionally negative by that participant from the training set were
for Anxiety (HRSA; Hamilton, 1959), and the Hamilton Rating
then used as the threat words in the training task. During testing,
Scale for Depression (HAM-D; Hamilton, 1960). Interview-based
all words from the alternate set were used.
assessments were administered to all participants by postdoctoral
and doctoral-level students with at least 2 years of cognitive- Participants were seated approximately 30 cm from the com-
behavioral training. Reliability was trained and maintained for all puter screen. Words were presented in the center of the screen,
interviews in a three-phase process. The certification procedure approximately 1.5 cm from one another in size 12 Arial font, for
required the SCID-IV trainee to first view videotaped and live 500 ms. Word pairs for each participant were presented in a
administrations of the Hamilton scales and SCID-IV by senior different random order. The computer program was written in
interviewers; trainees’ ratings of these interviews were then com- Delphi (Borland, Inc.) for this experiment.
pared with those of the senior interviewer. Next, the trainee ad- Participants in both groups completed the training procedures
ministered three HRSA, HAM-D, and SCID-IV interviews in the two times per week (on different days) for 4 weeks, for a total of
presence of the senior interviewer with the requirements that (a) eight completed sessions. Participants were informed that they
the trainee’s diagnoses matched those of the senior interviewer and would be randomly assigned to one of two attention training
(b) the Hamilton scales ratings differed by no more than 5 points groups. The protocol was described to the participants by the
from the senior interviewer’s ratings. Finally, each interviewer clinician administering the interview as an experimental procedure
maintained a video-based record of his or her interviews through- to determine the efficacy of a computer treatment for anxiety.
out the study. All assessment interviews were reviewed during Participants were informed that, depending on their random group
weekly meetings. If a trained rater no longer met the reliability assignment, the computer program they would complete could be
criterion, then he or she underwent further training using a differ- either a placebo condition that was not designed to influence their
ent set of training tapes until the criterion was reached again. The anxiety or an experimental treatment condition that was designed
interrater reliability for the primary outcome measure (HRSA) for to reduce their anxiety. This protocol was approved by the univer-
the present study was high (r ⫽ .94).
sity’s Institutional Review Board (IRB).
Self-report measures included the Spielberger State–Trait Anx-
The AMP. During each session, participants saw 240 trials that
iety Inventory (STAI; Spielberger, Gorsuch, Lushene, Vagg, &
consisted of the various combinations of probe type (E or F), probe
Jacobs, 1983), the Beck Depression Inventory—II (BDI–II; Beck,
position (top or bottom), and word type (neutral or threat). Of the
Steer, & Brown, 1996), the Worry Domains Questionnaire (WDQ;
Tallis, Eysenck, & Mathews, 1992), and the Penn State Worry 240 trials, 80 included only neutral words: 2 (probe type) ⫻ 2
Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, (probe position) ⫻ 20 (word pairs). The remaining 160 trials
1990). The information-processing measure used in the present included one neutral word and one threat word: 2 (probe type) ⫻
study was an alternate version of the probe detection task. This 2 (probe position) ⫻ 2 (threat word position) ⫻ 20 (word pairs).
task was identical in terms of probe location to the ACC procedure On trials in which participants saw one neutral word and one threat
but used an alternate set of stimuli to test for generalizability of the word (i.e., 66% of the trials), the probe always followed the neutral
training to a new set of materials. word. Thus, although there was no specific instruction to direct
attention away from threat word, on 66% of the trials, the position
Procedure of the neutral word indicated the position of the probe.
The ACC. The ACC condition was identical to the AMP proce-
Participants were randomly assigned to one of two conditions: dure except that during the presentation of the trials in which a threat
AMP (n ⫽ 14) or ACC (n ⫽ 15). The participants, independent word was present, the probe appeared with equal frequency in the
assessors, and research assistants working with the participants were position of the threat and neutral word. Thus, neither threat nor neutral
blind to participant condition. To ensure that participants remained words provided information regarding the position of the probe, and
blind with respect to their condition, each participant received an there was no contingency between the position of either threat or
envelope that contained a condition number that they entered into the neutral words and the position of the probes.
computer to start the assigned computer program. Groups did not
differ significantly on any clinician-administered or self-report
measure at pretraining ( ps ⬎ .2). Results
The stimuli used in the attention training were derived from
words used by MacLeod et al. (2002). These authors created two Groups did not differ on age, education, or gender at pretraining
sets of 48 words that were relevant to fears of individuals with ( ps ⬎ .4). Table 1 summarizes these results.
 SPECIAL SECTION: ATTENTION TRAINING IN GAD 31

Table 1 To examine change in each, we followed significant interactions

Demographic Characteristics with paired t tests. For self-report measures, these analyses revealed
that participants in the AMP group showed a significant decrease in
Variable AMP (n ⫽ 14) ACC (n ⫽ 15) their scores from pretraining to posttraining on the STAI-T, t(13) ⫽
Women, number (%) 7 (50%) 7 (47%) 4.76, p ⬍ .001, d ⫽ 1.40; STAI-S, t(13) ⫽ 4.27, p ⬍ .001, d ⫽ 1.81;
Age (SD) 26 (6.14) 25 (8.87) BDI, t(13) ⫽ 2.69, p ⬍ .02, d ⫽ 0.90; and WDQ, t(13) ⫽ 4.30, p ⬍
Education in years (SD) 15.7 (2.02) 15.4 (1.68) .001, d ⫽ 1.14. The same paired t tests in the ACC group did not
reveal significant changes on the STAI-T, t(14) ⫽ 1.89, p ⬍ .08, d ⫽
Note. AMP ⫽ Attention Modification Program; ACC ⫽ attention control
0.27; STAI-S, t(14) ⫽ 0.35, p ⬍ .74, d ⫽ 0.11; BDI, t(14) ⫽ 0.40,
condition.
p ⬍ .70, d ⫽ 0.07; or WDQ, t(14) ⫽ 1.35, p ⬍ .20, d ⫽ 0.36. Similar
analyses on the interviewer-administered measures revealed that the
Measure of Attention Bias AMP group showed a significant decrease in their scores on the
HRSA, t(13) ⫽ 4.27, p ⬍ .001, d ⫽ 1.36. In contrast, the ACC group
Our first goal was to demonstrate change in bias from pretrain- did not show a significant changed in their HRSA scores, t(14) ⫽
ing to posttraining in the AMP group using the novel words from 1.47, p ⬍ .16, d ⫽ 0.31.
the test set. These data are summarized in Table 2. We also examined the number of participants who no longer met
We calculated bias scores as per MacLeod and Mathews (1988); DSM–IV diagnosis for GAD at posttreatment. These analyses
these data are depicted in Figure 1. We submitted participants’ bias revealed that a significantly larger proportion of the participants in
scores for words in the test set to a 2 (group: AMP, ACC) ⫻ 2 the AMP group (50%) compared with the ACC group (13%) no
(time: pretraining, posttraining) analysis of variance (ANOVA), longer met diagnostic criteria for GAD after training, ␹2(1, N ⫽
with repeated measures on the second factor. The main effects of 29) ⫽ 4.55, p ⬍ .03.
group, F(1, 27) ⫽ 1.53, p ⫽ .22, and time, F(1, 27) ⫽ 0.80, p ⫽
.39, were not significant. However, there was a significant inter-
action of Group ⫻ Time, F(1, 27) ⫽ 5.40, p ⬍ .03. To follow up Mediational Analyses
this interaction, we conducted simple effects analyses. Simple
effect of time revealed that the participants in the AMP group To test the hypothesis that the AMP exerted its influence on anxiety
showed a reduction in their attention bias from pretraining to and depression through change in attention bias to threat, we con-
posttraining, t(13) ⫽ 3.21, p ⬍ .007, d ⫽ 3.3, whereas those in the ducted mediational analyses following the procedure described by
ACC group did not, t(14) ⫽ ⫺0.84, p ⫽ .41, d ⫽ ⫺1.0. Simple MacKinnon and colleagues (MacKinnon, Fairchild, & Fritz, 2007;
effect of group revealed that groups did not differ in their bias MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). In brief,
score at pretraining, t(27) ⫽ 0.46, p ⫽ .65. However the AMP this procedure tests the product of the coefficients for the effects of (a)
group showed significantly lower bias scores than the ACC group the independent variable (group: AMP, ACC) to the mediator (atten-
at posttraining, t(27) ⫽ 2.18, p ⬍ .05. tion bias after training) [␣]) and (b) the mediator to the dependent
variable (change in scores, HRSA, STAI-T, STAI-S, BDI, WDQ,
from pre- to posttraining) when the independent variable is taken into
Measures of Anxiety and Depression
account (␤). This procedure is a variation on the Sobel (1982) test that
We submitted participants’ scores on self-report and interviewer accounts for the nonnormal distribution of the ␣␤ path through the
measures to separate 2 (group: AMP, ACC) ⫻ 2 (time: pretraining, construction of asymmetric confidence intervals (MacKinnon, Fritz,
posttraining) ANOVAs, with repeated measurement on the second Williams, & Lockwood, 2007). Our results indicated that the 95%
factor. These analyses are summarized in Table 3. With the ex- confidence interval of the indirect path (␣␤) did not overlap with zero
ception of the HAMD and the PSWQ, all interactions of Group ⫻ for change in anxiety for HRSA (lower limit ⫽ .015, upper limit ⫽
Time were significant ( ps ⬍ .05). .343). The same indirect paths for all other measures (STAI-T,

Table 2
Means and Standard Deviations of Response Latencies by Group on the Probe Detection Task

AMP ACC

Pretraining Posttraining Pretraining Posttraining

Probe position M SD M SD M SD M SD

Top
Threat word
Top 589 175 526 100 635 206 543 115
Bottom 593 179 510 89 641 214 553 135
Bottom
Threat word
Top 584 199 508 100 629 214 539 120
Bottom 581 193 514 94 633 213 533 114

Note. AMP ⫽ Attention Modification Program; ACC ⫽ attention control condition.

32 AMIR, BEARD, BURNS, AND BOMYEA

Bias score on the probe detection task by training. Moreover, this change in bias was not specific to the
group and time words used for training but generalized to a different word set used
at pretraining and posttraining assessments. These results suggest
15 that the AMP group experienced a reduction in attention bias
AMP
10 ACC toward threat compared with the ACC group.
Although our predictions concerned the impact of attention
5 modification on anxiety, we observed similar effects on self-
Bias Score

0 reported depressive symptoms. These results are not surprising in

light of the high comorbidity between anxiety and depression,
-5 especially in individuals with GAD (Nutt, Argyropoulos, Hood, &
Potokar, 2006). However, our mediational analyses only supported
-10
the indirect effect of the attention training procedure on our inter-
-15 viewer measure of anxiety (HRSA) through reduction in attention
bias. This finding could reflect the differential sensitivity of our
-20 various measures or, alternatively, the relative efficacy of the
Pre Post training on each construct. Future research should examine the
potential mechanisms of attention training program effects on
Figure 1. Change in attention bias. AMP ⫽ Attention Modification depressive symptoms as well as on anxiety.
Program; ACC ⫽ attention control condition.
These results have theoretical and practical implications for
information-processing bias modification for anxiety. At the the-
STAI-S, BDI, WDQ) overlapped with zero, indicating an absence of oretical level, our results provide strong support for the hypothesis
a significant mediation. that the allocation of attention to threat-relevant information caus-
ally contributes to the anxiety symptoms experienced by individ-
uals with GAD. At the practical level, our findings suggest that it
Discussion
may be possible to use such attention training procedures clini-
Participants undergoing the AMP demonstrated a decrease in cally, either independent of or to augment already established
attention bias to threat and a decrease in anxiety symptoms. Fifty treatments for anxiety such as cognitive-behavioral therapy (CBT)
percent of participants in the AMP were classified as responders or medication. Despite its high prevalence, many individuals with
(no longer meeting DSM diagnosis for GAD) compared with 13% GAD who need treatment do not receive it for a variety of reasons.
of participants in the ACC. At posttreatment, independent asses- Additionally, many individuals who present for treatment do not
sors rated participants completing the AMP as significantly less respond (Fisher & Durham 1999; Gorman, 2003). This interven-
anxious from pre- to posttraining. Finally, participants’ self-report tion may provide an effective and efficient treatment accessible to
of anxiety and depressive symptoms corroborated the interviewer both individuals who are not receiving any treatment and those
ratings. These data suggest that our intervention reduced anxiety. who do not respond to traditional treatments. Moreover, this study
These results are promising, considering the short duration (i.e., 4 suggests that basic findings from experimental psychopathology
weeks) of the intervention and absence of therapist contact. The research can be translated into efficacious treatments for individ-
AMP group showed less bias for threat than the ACC group after uals with clinical levels of anxiety.

Table 3
Means and Standard Deviations for Self-Report and Interviewer Measures

AMP ACC Resultsa

Pre Post Pre Post Group Time Group ⫻ Time

Measure M SD M SD M SD M SD F p F p F p

Self-report
STAI-T 60.9 8.9 47.3 10.4 58.6 10.2 55.6 12.3 0.72 .4 26.72 .00 10.84 .003
STAI-S 59.1 10.5 41.4 8.9 51.1 9.8 49.8 13.7 0.00 .9 11.81 .002 8.87 .006
BDI-II 24.6 12.3 14.5 10.0 22.8 8.9 22.1 12.0 0.70 .4 7.17 .01 5.50 .03
WDQ 86.5 16.7 66.5 18.4 86.8 14.3 80.9 18.2 1.81 .2 16.54 .00 4.96 .04
PSWQ 70.6 4.7 63.0 11.2 67.8 7.6 63.9 10.9 0.11 .7 10.70 .003 1.15 .29
Interviewer
HRSA 20.3 6.5 11.9 5.7 19.0 7.6 16.7 7.4 0.58 .45 20.34 .001 6.46 .02
HAMD 10.4 5.4 6.4 4.4 12.0 4.7 10.3 4.3 3.29 .08 10.20 .004 1.73 .20

Note. AMP ⫽ Attention Modification Program; ACC ⫽ attention control condition; STAI ⫽ State–Trait Anxiety Inventory; BDI–II ⫽ Beck Depression
Inventory—II; WDQ ⫽ Worry Domains Questionnaire; PSWQ ⫽ Penn State Worry Questionnaire; HRSA ⫽ Hamilton Rating Scale for Anxiety;
HAM-D ⫽ Hamilton Rating Scale for Depression.
a
dfs ⫽ 1, 27.
 SPECIAL SECTION: ATTENTION TRAINING IN GAD 33

A number of meta-analyses have compared effective treatments Gould, R. A., Safren, S. A., Washington, D. O., & Otto, M. W. (2004). A
for GAD (i.e., CBT and pharmacological treatment). Three meta- meta-analytic review of cognitive-behavioral treatments. In R. Heimberg, C.
analyses of treatment outcome for GAD (Borkovec & Ruscio, Turk, & D. Mennin (Eds.), Generalized anxiety disorder: Advances in
2001; Gould, Saffren, Washington, & Otto, 2004; Lydiard & research and practice (pp. 248 –264). New York: Guilford Press.
Hamilton, M. (1959). The assessment of anxiety states by rating. British
Monnier, 2004) have suggested some variability in the estimates of
Journal of Medical Psychology, 32, 50 –55.
effect sizes. However, psychosocial treatments have resulted in
Hamilton, M. (1960). A rating scale for depression. Journal of Neurology,
between-group average effect sizes ranging from 0.71 to 0.90 and Neurosurgery, and Psychiatry, 23, 55– 62.
pharmacological treatments from .42 to .90 (Gelenberg et al., Lydiard, R. B., & Monnier, J. (2004). Pharmacological treatment. In R. Heimberg,
2000; Möller, Volz, Reimann, & Stoll, 2001; Rickels, Pollack, C. Turk, & D. Mennin (Eds.), Generalized anxiety disorder: Advances in
Sheehan, & Haskins, 2000). The effect sizes for the present study research and practice (pp. 351–379). New York: Guilford Press.
(0.72– 0.88) place this treatment in the range of present treatments MacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007). Mediation
for GAD. analysis. Annual Review of Psychology, 58, 593– 614.
Our study has limitations. First, our sample size was relatively MacKinnon, D. P., Fritz, M. S., Williams, J., & Lockwood, C. M. (2007).
small, limiting the generaliziaility of our results. Second, we only Distribution of the product confidence limits for the indirect effect:
Program PRODCLIN. Behavior Research Methods, 39, 384 –389.
obtained a significant interaction of Group ⫻ Time on one of our
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., &
two self-report measures of worry (i.e., on WDQ but not on
Sheets, V. (2002). A comparison of methods to test the significance of
PSWQ). Furthermore, examination of effect sizes indicated that the mediated effect. Psychological Methods, 7, 83–104.
the training was more effective for general anxiety symptoms than MacLeod, C., & Mathews, M. (1988). Anxiety and the allocation of
for worry symptoms. Therefore, this study does not provide a clear attention to threat. The Quarterly Journal of Experimental Psychology
discrimination of the symptoms of GAD that may be most ame- Section A: Human Experimental Psychology, 40, 653– 670.
nable to change using attention training. Third, we did not include MacLeod, C., Mathews, A., & Tata, P. (1986). Attentional bias in emo-
an independent measure of attention bias in our study. Future work tional disorders. Journal of Abnormal Psychology, 95, 15–20.
should include different measures of attention bias (e.g., Posner MacLeod, C., Rutherford, E., Campbell, L., Ebsworth, G., & Holker, L.
task, Stroop task) to examine the generalizability of attention (2002). Selective attention and emotional vulnerability: Assessing the
causal basis of their association through the experimental manipulation
training to other measures of attention.
of attentional bias. Journal of Abnormal Psychology, 111, 107–123.
In summary, our results suggest that the translation of basic
Mathews, A., & MacLeod, C. (1985). Selective processing of threat cues in
psychopathology research to address a clinical condition may anxiety states. Behaviour Research & Therapy, 23, 563–569.
prove useful in developing new interventions. Moreover, these Mathews, A., & MacLeod, C. (1986). Discrimination of threat cues without
interventions may help identify the mechanisms that are involved awareness in anxiety states. Journal of Abnormal Psychology, 95, 131–138.
in the pathogenesis of psychiatric conditions. Future studies should Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. D. (1990).
examine the additive and/or interactive effects of attention training Development and validation of the Penn State Worry Questionnaire.
and traditional interventions (i.e., medication and CBT) as well as Behaviour Research & Therapy, 28, 487– 495.
the combination of other types of information-processing training Mogg, K., & Bradley, B. P. (2005). Attentional bias in generalized anxiety
(e.g., interpretation modification). disorder versus depressive disorder. Cognitive Therapy and Research,
29, 29 – 45.
Mogg, K., Millar, N., & Bradley, B. P. (2000). Biases in eye movements
References to threatening facial expressions in generalized anxiety disorder and
depressive disorder. Journal of Abnormal Psychology, 109, 695–704.
American Psychiatric Association. (1994). Diagnostic and statistical man-
Möller, H. J., Volz, H. P., Reimann, I. W., & Stoll, K.-D. (2001). Opi-
ual of mental disorders (4th ed.). Washington, DC: Author.
pramol for the treatment of generalized anxiety disorder: A placebo-
Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M. J., &
controlled trial including an alprazolam-treated group. Journal of Clin-
van IJzendoorn, M. H. (2007). Threat-related attentional bias in anxious
ical Psychopharmacology, 21, 59 – 65.
and nonanxious individuals: A meta-analytic study. Psychological Bul-
Nutt, D., Argyropoulos, S., Hood, S., & Potokar, J. (2006). Generalized
letin, 133, 1–12.
anxiety disorder: A comorbid disease. European Neuropsychopharma-
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for Beck
cology, 16, S109 –S118.
Depression Inventory-II. San Antonio, TX: Psychological Corporation.
Rickels, K., Pollack, M. H., Sheehan, D., & Haskins, J. (2000). Efficacy of
Borkovec, T. D., & Ruscio, A. (2001). Psychotherapy for generalized
extended-release venlafaxine in nondepressed outpatients with general-
anxiety disorder. Journal of Clinical Psychiatry, 62, 37– 45.
ized anxiety disorder. American Journal of Psychiatry, 157, 968 –974.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. (1994).
Sobel, M. E. (1982) Asymptotic confidence intervals for indirect effects in
Structured clinical interview for DSM-IV-patient edition (SCID-I/P,
structural equation models. In S. Leinhardt (Ed.), Sociological methodology
version 2.0). New York: Biometrics Research Department.
(pp. 290 –312). Washington, DC: American Sociological Association.
Fisher, P. L., & Durham, R. C. (1999). Recovery rates in generalized
Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs,
anxiety disorder following psychological therapy: An analysis of clini-
G. A. (1983). Manual for the State-Trait Anxiety Inventory. Palo Alto,
cally significant change in the STAI-T across outcome studies since
CA: Consulting Psychologist Press.
1990. Psychological Medicine, 29, 1425–1434.
Tallis, F., Eysenck, M. W., & Mathews, A. (1992). A questionnaire for the
Gelenberg, A. J., Lydiard, R. B., Rudolph, R. L., Aguiar, L., Haskins, J. T.,
measurement of nonpathological worry. Personality and Individual Dif-
& Salinas, E. (2000). Efficacy of venlafaxine extended-release capsules
ferences, 13, 161–168.
in nondepressed outpatients with generalized anxiety disorder: A
6-month randomized controlled trial. JAMA: Journal of the American
Medical Association, 283, 3082–3088. Received July 12, 2007
Gorman, J. M. (2003). Treating generalized anxiety disorder. Journal of Revision received April 4, 2008
Clinical Psychiatry, 64, 24 –29. Accepted April 9, 2008 䡲