ARTICLE IN PRESS

Behaviour Research and Therapy 45 (2007) 2691–2703

www.elsevier.com/locate/brat

Attention focusing versus distraction during exposure in

dental phobia
Barbara Schmid-Leuza, Karin Elsessera, Thomas Lohrmanna,
Peter Jöhrenb, Gudrun Sartorya,
a
Department of Clinical Psychology, University of Wuppertal, Max-Horkheimer-Strasse 20, 42119 Wuppertal, Germany
b
Dental Clinic, Augusta Hospital, Bergstrasse 26, 44791 Bochum, Germany

Received 10 April 2007; received in revised form 2 July 2007; accepted 17 July 2007

Abstract

A survey of the discrepant findings regarding the effects of attention focusing and distraction on exposure suggested that
subjective measures of anxiety and avoidance respond better to the latter condition, and heart rate (HR) reaction responds
to the former. To test this hypothesis, 63 dental phobics were recruited who had not visited a dentist for a mean of 6.6
(1.5–25) years. Participants received a 1-h exposure session with either attention focusing or distraction. Subjective anxiety
and HR to phobia-related pictures were assessed before and after the treatment session and again after 1 week. Avoidance
was recorded in terms of adherence to the dental treatment schedule in the following 6 months. Contrary to expectation,
state anxiety showed a greater decrease in the attention focusing than the distraction condition after 1 week. Both
treatment conditions were similarly effective with regard to HR and avoidance. HR habituated in both groups after
exposure and 73% of followed-up patients adhered to the dental treatment schedule. Comparison of the present with
previous results suggests that the differences between attentional conditions tend to be more pronounced during shorter
exposure sessions than were employed in the present study.
r 2007 Elsevier Ltd. All rights reserved.

Keywords: Specific phobia; Exposure; Distraction; Attention focusing; Dental phobia; Heart rate

Introduction

Specific phobias are strong, irrational fear responses to situations or objects, and phobics typically react
with flight or avoidance when confronted with the phobic stimulus. Prolonged exposure to the phobic stimulus
has long been found to be the most effective treatment method of this disorder, but some questions remain as
to the mode of exposure. Among the problems still to be resolved is that of the attentional focus during
exposure treatment: Need phobics attend the phobia-related material or is its presence just as efficacious in
reducing anxiety if phobics are distracted?

Corresponding author. Tel.: +49 202 4392722; fax: +49 202 4393031.
E-mail address: sartory@uni-wuppertal.de (G. Sartory).

0005-7967/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.brat.2007.07.004
 ARTICLE IN PRESS
2692 B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703

In the original model of emotional processing by Foa and Kozak (1986), attention focusing is essential for
fear reduction. According to this model, exposure activates the prototype of the fear reaction, which consists
of a neuronal network that contains all aspects of stimulus properties, and emotional, physiological and
behavioural fear responses including cognitive appraisal. This fear structure is stored in memory and recalled
on exposure to any of its elements. Exposure introduces corrective, incompatible information leading to an
uncoupling of the elements. According to this model, attention focusing, which promotes sensory encoding of
the presented phobic stimuli, is necessary for the full activation of the fear structure. In contrast, distraction
strategies prevent encoding of the relevant stimulus elements and inhibit activation of the fear structure and
thereby also emotional processing. Incomplete emotional processing is thought to result in partial relapse, also
termed return of fear (e.g. Sartory, Rachman, & Grey, 1982).
Alternatively, Bandura’s model (1983) of fear reduction stresses the importance of self-efficacy, i.e., the
individual’s conviction to be able to cope with a situation. Accordingly, it is the perceived lack of self-efficacy
that induces fear during potentially aversive situations (Bandura, 1988). Behavioural mastery of fear-inducing
situations is considered to enhance perceived self-efficacy. Coping strategies aimed at mastery have been found
to be highly effective in combination with exposure and to improve self-efficacy (Jones & Menzies, 2000).
According to this model, fear reduction results from enhancement of self-efficacy. Distraction from phobic
stimuli would be considered a coping strategy as it allows the phobic individual to master the situation with
less anxiety than would otherwise have been the case. The experience of being relatively free from anxiety in
the presence of the phobic stimulus improves perceived self-efficacy and contributes to fear reduction.
As shown in Table 1, results of the condition of attention focusing and distraction during exposure
treatment are inconsistent. In a first such study in obsessive-compulsive patients (Grayson, Foa, & Steketee,
1982), one group was asked to attend to the anxiety-inducing material and the other group was also asked to
pick it up while playing a video game with the therapist during the treatment session. Both groups showed a
similar extent of reduction in subjective discomfort at the end of the session but the attention group evidenced
a more marked reduction in heart rate (HR). The distraction group experienced return of fear at the following
session, whereas the attention group’s fear reduction remained stable across sessions. In a subsequent study,
Grayson, Foa, and Steketee (1986) failed to replicate the follow-up result but confirmed the greater HR

Table 1
Comparisons between attention focusing (A) and distraction (D) during exposure

Authors (year) Participants (st. — Design, procedure Measures Results

student sample)

Grayson et al. (1982) OCD A, D; 90 min, FU SUD, HR HR: A4D FU: SUD:
A4D
Grayson et al (1986) OCD A, D; 90 min, FU SUD, HR SUD: D4A HR:
A4D
Craske, Street, and Barlow PD with agoraphobia A, D; 11 S, FU SUD, BAT No diff.
(1989)
Craske et al. (1991) Animal phobics (st.) Ex., A, D; 6 min SUD, HR, SUD: D4A
Rodriguez and Craske (1995) Animal phobics (st.) D, no D  hi/ lo intens. SUD, HR, BAT SUD: hi ex: no D4D
ex.; 15 min
Penfold and Page (1999) Blood injury phobics A, D; ex.; 10 min SUD, BAT SUD: D4A, ex.
(st.)
Kamphuis and Telch (2000) Claustrophobics (st.) A, D, AD, ex.; SUD, HR, BAT SUD: A4D
6  5 min
Mohlman and Zinbarg (2000) Animal phobics A, D; 7  3 min SUD, HR, BAT SUD: A4D
Antony, McCabe, Leeuw, Sano, Animal phobics D, ex.; 30 min SUD, HR, BAT No diff.
and Swinson (2001)
Oliver and Page (2003) Blood injury phobics A, D, ex.; 3  10 min SUD D4A, ex. FU: D4A,
(st.) weekly, FU ex.
Johnstone and Page (2004) Animal phobics (st.) A, D; 3  10 min SUD, HR, BAT SUD+BAT: D4A
weekly, FU HR: A4D

Abbreviations: OCD: obsessional compulsive patients; S: sessions; SUD: subjective units of discomfort; HR: heart rate; PD: panic disorder;
ex.: exposure alone; hi/lo intens.: high/low intensity; BAT: behavioural avoidance test.
 ARTICLE IN PRESS
B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703 2693

decrease in the attention condition. In contrast, the distraction condition led to more improvement with
regard to subjective units of discomfort (SUD). Subsequent studies showed discrepant results. Measures of
SUD appear to favour distraction (Craske, Street, Jayaraman, & Barlow, 1991; Johnstone & Page, 2004;
Penfold & Page, 1999), although there are also exceptions (Mohlman & Zinbarg, 2000; Oliver & Page, 2003).
HR, on the other hand, appears to respond better to an attention-focusing condition (Grayson et al., 1982,
1986; Johnstone & Page, 2004). There are, however, also a number of studies showing no effect on HR by
either condition (Craske et al., 1991; Kamphuis & Telch, 2000; Mohlman & Zinbarg, 2000). The elusiveness of
the effect on HR may be due to differences in the way assessments were carried out. In some studies, HR was
recorded during exposure, in others during the behavioural avoidance test. Even so, results suggest a pattern
whereby distraction appears to have a superior effect on SUDs and approach as well as on cognitive factors
such as self-efficacy and internal locus of control, whereas attention focusing appears to lead to greater
improvement in physiological reactions to phobia-related stimuli. Various components of the phobic reaction
may underlie different processing modes and therefore also respond to different treatment conditions.
The present study aimed at investigating the course of phobic response components during attention
focusing and distraction in dental phobics who were in need of dental treatment. Patients received one
exposure session with either attention focusing or distraction. HR and SUDs were assessed before and after
and at follow-up after 1 week. The follow-up assessment was included in order to monitor return of fear in any
of the measures. Patients completed questionnaires measuring phobic severity and dysfunctional cognitions
initially and at follow-up. Additionally, attendance of dental treatment was recorded in the weeks after the
experiment. SUDs, dysfunctional cognitions and avoidance behaviour were expected to benefit more from
distraction than attention focusing with the reverse being the case for HR. Additionally, the distraction
condition was expected to lead to greater return of fear than attention focusing at the follow-up measurement
after 1 week.

Method

Participants

Sixty-three patients (28 men, 35 women) with dental phobia took part in the study. They were recruited at
the Dental Anxiety Clinic (Augusta Krankenanstalten Bochum), which had received repeated media coverage
for specializing in treating dental phobia. Patients were included in the study if they met DSM-IV criteria of
specific (dental) phobia, had not visited a dentist for the last 2 years (unless it was emergency treatment under
a general anaesthetic) and took no anxiolytic medication or medication affecting the cardiovascular system.
The study was approved by the local ethics committee. All participants gave their written informed consent
before being admitted to the study and received a small remuneration to cover travel expenses.
The mean age of the participants was 35.00 (SD ¼ 9.92) years. The average duration of not having visited a
dentist was 6.62 years (SD ¼ 5.67; range: 1.5–25 years). A mean of 7.6 (SD ¼ 5.6) teeth were in need of
treatment. Thirty participants suffered from one or several comorbid disorders (16 from other specific
phobias, 11 from social phobia and the remaining from other anxiety disorders or mild depressive mood).

Experimental design (Table 2)

Following the initial diagnostic interview by a clinical psychologist (MSc level) in psychotherapy training,
participants were randomly allocated to the attention focusing or distraction condition and completed the
questionnaires. A week later, the exposure session took place with laboratory measurements before and after.
After another week, questionnaires were again completed and laboratory measurements were taken again.
Attendance of the dental appointments was recorded over the following 6 months.

Exposure treatment

The exposure session lasted for 60 min. Initially, participants of both groups were shown pictures of four
dental instruments (dental probe, drill, pliers and hypodermic needle) and asked to rank them in terms of
 ARTICLE IN PRESS
2694 B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703

Table 2
Study design

Condition

Attention-focusing Distraction

Initial assessment Dental examination

DIPS
Questionnaires
Treatment session (1 week later) Pre treatment assessment:
SUDs pre1
HR pre
Appraisal of pictures
SUDs pre2
Fear rating of instruments
Exposure with attention focusing Exposure with distraction
Post treatment assessment:
Fear rating of instruments
HR post
SUDs post
Follow-up (1 week later) SUDs FU
Questionnaires
Fear ratings of instruments
HR FU
Appraisal of pictures
Avoidance (over 6 months) Adherence to dental treatment schedule

being anxiety-eliciting. Throughout the exposure session, both groups received acoustic stimuli related to
dental treatment (noises made by dental probes, by drills and during ultrasound cleaning of tartar).
Additionally, cotton pads soaked in a disinfectant, typically used during dental treatment, were placed on a
table in front of participants. The four instruments were then placed consecutively in front of the patient for
15 min each, beginning with the least and progressing to the most anxiety-provoking one. The therapist sat
next to the patients throughout the session to ensure that their attention was either focused on the phobic
stimulus or directed away.

Attention focusing
Patients were instructed that their dental fear would subside if they were to remain in the presence of the
fear-eliciting dental stimuli for a long enough time. They were asked to pick up the instrument, examine it
closely and consider its function. The therapist asked them to explore their fear reaction and to remember
experiences with the instrument during previous dental treatments.

Distraction
Patients were instructed that their dental fear would subside if they could learn to distract themselves in the
presence of the fear-eliciting stimuli. They were also asked to hold the instrument in their non-dominant hand
while playing puzzle games with the therapist. The therapist discussed the various moves and strategies for the
completion of the puzzle. All participants were asked to give fear ratings of dental instruments at the
beginning and the end of the exposure session.

Materials

Interviews and questionnaires

All questionnaires were administered in the first diagnostic session and 2 weeks later at follow-up.
Initially, patients were assessed with a structured interview (Diagnostisches Interview bei psychischen
Störungen (DIPS); Schneider and Margraf (2006), the German adaptation of the Anxiety Disorders
 ARTICLE IN PRESS
B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703 2695

Interview Schedule for DSM-IV, Brown, Di Nardo, & Barlow, 1994), which confirms DSM-IV criteria. DIPS
has a good test–retest reliability (r ¼ .64–.89) and inter-rater reliability (kappa r ¼ .80–1.00; Schneider &
Margraf, 2006).
Dental Anxiety Scale (DAS; Corah (1969), German version translated by the authors). This self-rating
questionnaire is the most frequently used instrument in dental anxiety research and consists of four items
relating to dental treatment. Patients were asked to rate how anxious they would be if they had to submit to
these situations the following day. Scores range from 4 to 20. Corah, Gale, and Illig (1978) reported a mean
score of 9.07 in 2103 non-selected participants, with women having a significantly higher score (M ¼ 9.53)
than men (M ¼ 8.56). Dental phobic patients had a mean score of 17.20 (SD ¼ 1.80). A score of 15 is
considered the cut-off for being highly anxious, and scores of 13 and 14 are considered to indicate moderate
fearfulness of dental treatment (Corah et al., 1978). An internal consistency index of Cronbach’s a ¼ .64 has
been reported (Sartory, Heinen, Pundt, & Jöhren, 2006).
Dental Cognitions Questionnaire (DCQ; De Jongh, Muris, Ter Horst and Schoenmakers (1995), German
version translated by the authors). This self-rating questionnaire consists of 38 negative cognitions (beliefs and
self-statements) related to dental treatment. Fourteen items concern negative beliefs pertaining to dentistry in
general (e.g. ‘‘Dentists don’t carey’’) and to the patients themselves (e.g. ‘‘I can’t stand pain’’); the remaining
24 items contain negative self-statements (e.g. ‘‘Everything is going wrong’’). Patients were asked to indicate
whether these negative beliefs occur to them during dental treatment and the frequency of negative cognitions
(DCQ: range ¼ 0–38) is summed. Individuals with dental phobia were found to have a significantly higher
number of negative cognitions than non-phobic controls (De Jongh et al., 1995). Data of a previous study
yielded a Cronbach’s a ¼ .90 (Sartory et al., 2006a, 2006b).
Revised Iowa Dental Control Index (IDCI; Brunsman, Logan, Patil, and Baron (2003); German version
translated by the authors). This self-rating questionnaire consists of nine items, five of which concern the
desire for control (e.g. ‘To what degree would you like control over what will happen to you in the dental
chair?’) (Cronbach’s a4.78–.79) and the other four predicted control during dental treatment (e.g. ‘How much
do you think you can control what will happen to you while in the dental chair?’) (Cronbach’s a4.75–.80).
Items are rated from 1 (none) to 5 (totally) and summed. Dental patients with a high desire for control coupled
with a low feeling of control reported high levels of dental distress compared with low scorers (Logan, Baron,
Keely, Law, & Stein, 1991).
State-Trait Anxiety Inventory (STAI X1, X2; German version by Laux, Glanzmann, Schaffner, &
Spielberger, 1981). The two scales contain 20 items each describing emotional states. Probands are asked to
indicate the degree to which a given statement applies to them at present (state) and during the last 2 weeks
(trait version). Scores range from 20 (no anxiety) to 80 (high anxiety). An internal consistency score of
Cronbach’s a4.90 was reported (Laux et al., 1981).
Beck Depression Inventory (BDI, German version by Hautzinger, Bailer, Worall, & Keller, 1994). This
21-item inventory indexes depression intensity with a score range of 0–63. A score between 11 and 17 is
considered indicative of mild depressive symptoms, and a score of 18 or above is indicative of severe
depression. Internal consistency scores have been reported between Cronbach’s a4.74 in healthy subjects and
.92 in depressed patients (Hautzinger et al., 1994).
Aachen Self-Efficacy Questionnaire (ASF; Wälte, Ebe, Brandenburg, & Kröger, 1999). This self-rating
questionnaire consists of 20 items detailing a variety of problem situations and a successful way of dealing
with them. Participants are asked to indicate from 1 to 5 whether the behaviour applies to them. A factor
analysis allocated the items to four subscales: (1) a general way of dealing with problems (e.g. ‘‘I can attain the
aims I set myself’’), (2) work (e.g. ‘‘I can adapt to new situations at work’’), (3) health (e.g. ‘‘I can deal with
health problems’’) and (4) social interaction (e.g. ‘‘I can adjust to other people during unforeseen events’’). The
total score can range from 20 to 100. A study of 650 attendants of a clinic for psychosomatic disorders found
the internal consistency to be high for the total score (Cronbach’s a ¼ .90).
SUD: Patients were asked to rate how anxious they would be if they had to attend dental treatment
at this very moment, from 0 (no anxiety) to 100 (extremely anxious). These ratings were taken at the
beginning of the treatment session (pre 1), after the first HR measurements had been taken (pre 2),
immediately after the exposure session (post) and at the beginning of the follow-up assessment a week
later (FU).
 ARTICLE IN PRESS
2696 B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703

Fear ratings: At the beginning and end of exposure, participants were shown the instruments individually
and asked to give a fear rating from 0 to 100. A further rating was elicited at the follow-up assessment 1 week
later.

Heart rate assessment

Identical assessment procedures were carried out before and after exposure and again at follow-up (apart
from subjective appraisal, which was only recorded before exposure and at follow-up).
Stimuli: Coloured pictures depicting either dental treatment scenes or emotionally neutral contents, i.e.,
household objects, were taken partly from the International Affective Picture System (Lang, Bradley, &
Cuthbert, 1995) and partly from other sources such as textbooks of dental treatment. Eight pictures of each
category were projected with a video beamer (NEC MultiSync MT 830+, Tokyo, Japan) on a screen,
resulting in a picture size of 100  75 cm with the participants viewing them from a distance of 1.6 m. The
pictures were presented for 4 s with a randomly varying stimulus onset asynchrony, i.e., intervals between
onsets, of 15–19 s.
Heart rate (HR): Recording and data reduction of the psychophysiological measures were carried out
with a Vitaport-III (TEMEC Instruments, B.V., Kirkrade, the Netherlands). ECG was recorded with
chest electrodes. The sampling rate was 512 Hz; R-waves were detected online and inter-beat-intervals
converted with an RR-interval delay into HR in bpm. Mean HR was calculated for 6 s epochs after onset
of stimuli and baseline-corrected taking 3 s before stimulus onset into account. HR reactions were
averaged within stimulus categories resulting in evoked responses to the phobia-relevant and neutral stimuli.
Resting HR was recorded for 1 min before presentation of any stimuli. Respiration was recorded with a
respiratory belt, which was placed around the chest. Recordings were inspected and employed for artefact
control.
Subjective appraisal: Following the removal of the physiological sensors, all 16 previously shown pictures
were presented again. This time, participants rated valence and fear-arousing property of the pictures from 1
(pleasant/not at all) to 9 (unpleasant/extremely fear arousing). Mean ratings were obtained for each stimulus
mode and type. A week later, subjective appraisal was assessed again.

Avoidance
Attendance of dental treatment was assessed based on the dental treatment records of the referral clinic over
6 months after psychological treatment. Participants were categorized according to (a) attendance and
completion of dental treatment, (b) completion of two dental cleaning sessions and subsequent avoidance of
treatment, (c) attendance of one dental cleaning session and subsequent avoidance and (d) complete
avoidance.

Procedure (Table 2)
The original assessment took place in the dental clinic where patients were referred to the psychologists. One
week later, the treatment experiment took place in the laboratory of the Clinical Psychology Deptartment at
the University of Wuppertal. Initially, participants handed in the completed questionnaires, which they had
received after agreeing to take part in the study following the diagnostic interview. SUD pre 1 rating was
taken, after which HR pre-assessment was carried out. Participants were seated in a comfortable chair in
front of a screen in a sound- and light-attenuated room. ECG electrodes and the respiratory belt were attached
after which participants were asked to sit still and relax for 3 min, the last minute of which served as resting
HR. After presentation of the pictures, fear and valence ratings were recorded. Participants were then
shown the four dental instruments and asked to give a fear rating after which they received the exposure
session. At the end, another fear rating of the dental instruments was elicited and HR post-recorded during
presentation of the pictures. Afterwards SUD post was recorded. A week later, a follow-up assessment was
conducted. Initially, participants were asked to give a SUD FU rating and completed the questionnaires
again. They were then shown the dental instruments again and gave fear ratings. There was then a final HR
FU assessment followed by another appraisal of the pictures, identical to the one carried out before the
exposure session.
 ARTICLE IN PRESS
B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703 2697

Results

Group comparisons

The attention group was significantly older than the distraction group (F(1,61) ¼ 12.79, po.01; Z2 ¼ .17;
Table 2). There was no significant group difference with regard to gender composition (w2 ¼ .21, p ¼ .12).
Questionnaires: Group means and SDs of questionnaire data are displayed in Table 3. As questionnaire data
tend to be highly inter-correlated, a principal component analysis was initially carried out to explore the factor
structure with the aim of extracting factor scores. As shown in Table 4, the principal component analysis with

Table 3
Group means and SDs of questionnaire data

Variable Attention (N ¼ 32) Distraction (N ¼ 31) Sign. effects

Age 39.03 (9.23) 30.84 (8.94) G

Sex (m/f) 17/15 11/20 n.s.
Dental anxiety (DAS)
Pre 17.53 (1.90) 17.23 (2.50) MO
FU 15.38 (3.06) 15.45 (3.17)

DCQ
Pre 21.06 (6.23) 21.53 (7.47) MO
FU 16.06 (6.34) 19.03 (7.90) G  MO*
IDCI (desired control)
Pre 21.26 (2.67) 21.03 (3.17) n.s.
FU 21.06 (2.59) 20.90 (4.20)
IDCI (predicted control)
Pre 7.50 (2.53) 7.97 (2.49) MO
FU 9.27 (3.08) 8.76 (2.55)
State anxiety (STAI)
Pre 40.00 (10.19) 37.68 (7.57) G  MO
FU 38.19 (9.54) 41.74 (8.15)
Trait anxiety (STAI)
Pre 38.81 (9.55) 40.42 (8.32) MO
FU 36.78 (8.98) 39.71 (9.99)

Depression (BDI)
Pre 6.34 (5.18) 8.87 (9.02) MO
FU 4.31 (5.54) 6.73 (6.27)
ASF (self efficacy)
Pre 3.85 (.34) 3.75 (.44) MO
FU 3.93 (.37) 3.84 (.52)
SUDs
Pre 1 83.75 (12.12) 75.48 (18.04) MO
Pre 2 80.63 (15.65) 74.84 (18.23) G  MO*
Post 60.63 (18.83) 58.06 (23.72)
FU 62.81 (22.18) 63.55 (23.88)
Resting HR (bpm)
Pre 81.80 (16.13) 81.91 (10.38) MO
Post 72.85 (13.51) 73.46 (7.83)
FU 74.96 (12.21) 74.58 (9.58)
Avoidance (% attendance of subsequent dental treatment) 66.7 (N ¼ 26) 79.3 (N ¼ 29) n.s.

Significant effects denote main effects for G (group), MO (measurement occasion) and G  MO interaction. (*indicates a marginally
significant effect; FU—follow-up assessment 1 week after exposure).
 ARTICLE IN PRESS
2698 B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703

Table 4
Principal component analysis with Varimax rotation and Kaiser normalization of questionnaire data

Variables Component

1 2

Trait anxiety (STAI) .897 .133

ASF (self efficacy) .832 .057
Depression (BDI) .805 .072
State anxiety (STAI) .777 .061
IDCI (predicted control) .093 .808
Dental anxiety (DAS) .052 .718
DCQ .379 .708
IDCI (desired control) .035 .522

a subsequent varimax rotation of the eight questionnaires resulted in two factors: (I) a generalized anxiety and
mood factor and (II) a specific phobia factor. Factor scores were extracted and submitted to 2  2
(group  measurement occasion, MO) ANOVAS. Factor I showed a marginally significant group  MO
interaction (F(1,56) ¼ 3.08, po.09, Z2 ¼ .05). There were no other significant group effects with regard to
either factor. Individual variables were then submitted to ANOVA with a 2  2 design comparing groups by
measurement occasion (pre-treatment, follow-up after 1 week). There was no main effect for the group with
regard to any of the questionnaires. There was, however, a measurement occasion effect with regard to most
variables showing improvement from before the treatment session to follow-up. (DAS: F(1,61) ¼ 41.03,
po.01, Z2 ¼ .40; DCQ: F(1,60) ¼ 25.47, po.01; Z2 ¼ .30; IDCI (predicted): F(1,57) ¼ 13.56, po.01; Z2 ¼ .19;
STAI trait: F(1,61) ¼ 5.16, po.03; Z2 ¼ .08; BDI: F(1,60) ¼ 17.97, po.01, Z2 ¼ .23; ASF: F(1,59) ¼ 5.41,
po.03; Z2 ¼ .08). Additionally, there was a significant group  measurement occasion effect in case of STAI
state (F(1,61) ¼ 6.34, po.02, Z2 ¼ .09). Simple comparison of means revealed a significant STAI state increase
in the distraction group from before treatment to follow-up (t(30) ¼ 2.91, po.01) with no other effects being
significant. There was also a marginally significant interaction effect with regard to the DCQ (F(1,60) ¼ 2.83,
po.10, Z2 ¼ .05). Attention engendered marginally greater improvement than distraction at follow-up.
SUDs were submitted to ANOVA with a 2  4 design. There was no significant group nor interaction effect
of group  occasion (Table 3). SUDs decreased significantly in both groups (F(3,59) ¼ 28.69, po.01,
Z2 ¼ .59). Individual comparison of means revealed a significant decrease from before to after the treatment
session whereas the other comparisons were not significant. Transformation of the repeated measures into
linear, quadratic and cubic trends confirmed the results of a significant decrease and yielded a marginally
significant group  linear trend (F ¼ 2.83, po.10, Z2 ¼ .04), with the attention group showing a greater
decrease in SUDs over measurement occasions.
Fear ratings of dental instruments are shown in Fig. 1. Data were submitted to a 2  4  3
(group  instrument  measurement occasion) ANOVA design. There were no significant group differences,
but some instruments were considered more fear-arousing than others (F(3,183) ¼ 18.02, po.01, Z2 ¼ .23).
Individual comparisons revealed the drill to be more fear-arousing than any of the other instruments and the
pliers and hypodermic needle to be more fear-arousing than the dental probe. Fear ratings decreased from
before to after treatment (F(2,122) ¼ 26.23, po.01, Z2 ¼ .30) and with a different course (instruments  MO:
F(6,366) ¼ 3.39, po.01, Z2 ¼ .05), with fear ratings to the drill, unlike to the probe, showing an increase
between the end of the treatment session and follow-up.
Appraisal of pictures: Both valence and fear ratings of pictures were submitted to an ANOVA with a
2  2  2 (group  pre-treatment/FU  stimulus type) design. There were no significant main nor interaction
effects with regard to group but main effects for measurement occasion (fear: F(1,61) ¼ 37.05, po.01,
Z2 ¼ .38; valence: F(1.61) ¼ 14.35, po01), stimulus type (fear: F(1,61) ¼ 691.65, po.01, Z2 ¼ .92; valence:
F(1,61) ¼ 483.31, po.01) and a significant interaction effect between stimulus type and measurement occasion
(fear: F(1,61) ¼ 38.12, po.01, Z2 ¼ .38; valence: F(1,61) ¼ 33.90, po.01). Valence and fear ratings decreased
more to phobia-related than to neutral pictures from before treatment to follow-up (Fig. 2).
 ARTICLE IN PRESS
B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703 2699

Fig. 1. Fear ratings of dental instruments at the beginning and end of their 15-min exposure (pre–post) and at the follow-up assessment 1
week later. The drill elicited higher fear ratings than the other instruments and lead to greater return of fear than the dental probe at
follow-up.

Fig. 2. Subjective appraisal of pictures in terms of fear and valence before the exposure session and at follow-up. Phobia-related pictures
were given higher ratings than neutral pictures and their ratings improved more at follow-up.

HR: Resting HR (Table 3) was submitted to ANOVA with a 2  3 (group  measurement occasion) design.
There was no significant group difference but a measurement occasion effect (F(2,60) ¼ 39.04, po.01,
Z2 ¼ .57). A significant quadratic trend confirmed that tonic HR decreased from before to after treatment and
did not increase again at follow-up.
 ARTICLE IN PRESS
2700 B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703

Submitting the evoked HR responses to ANOVA with a design of 2  3  2  6 (group  measurement

occasion  stimulus type  second) revealed no significant group differences. There was a significant seconds
effect and a stimulus type  seconds interaction (F(5,305) ¼ 5.07, po.01, Z2 ¼ .08) indicating that HR
increased more to phobia-related than neutral pictures. A significant measurement occasion  stimulus
type effect (F(2,122) ¼ 13.00, po.01, Z2 ¼ .18) showed that HR responses to phobia-related pictures
decreased over occasions and the significant quadratic effect of occasions  stimulus type (po.01) confirmed
that the decrease occurred from before to after the treatment session and remained stable at follow-up
(Fig. 3).
Avoidance: Attendance of dental treatment could not be assessed in 7 participants because records could not
be traced (5 of the attention and 2 of the distraction group). Forty-one (73.2%) of the remaining 56
participants underwent and completed the subsequent dental treatment, 10 attended two and 2 one dental
cleaning sessions before terminating dental treatment prematurely. Three participants stayed away altogether.

Fig. 3. HR reactions to phobia-related and neutral pictures before and after treatment and at follow-up. HR accelerated more to the
former than the latter stimuli and improved from before to after treatment.
 ARTICLE IN PRESS
B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703 2701

There was no significant group difference with regard to avoidance (Mann–Whitney U ¼ 350.00, n.s., two-
tailed).

Relations between measures

Age was only correlated with time since the last dental treatment (r ¼ .30, po.03).
As indicated by the principal component analysis, questionnaire measures and ratings were highly inter-
correlated. Avoidance was highly and positively correlated with STAI state (r ¼ .47, po.001) and trait anxiety
(r ¼ .41, po.01) as well as with IDCI (desired control) (r ¼ .39, po.01) and BDI depression (r ¼ .40, po.01)
as assessed at the first occasion. Neither resting HR nor HR responses were significantly correlated with any of
the other measures.

Discussion

Contrary to expectations, distraction failed to engender greater improvement in SUD, dysfunctional

cognitions and avoidance. Instead, STAI state anxiety showed a greater decrease in the attention focusing
than distraction group at follow-up, and dysfunctional cognitions also improved marginally more in the
former than latter group. In addition, SUDs showed marginally more improvement after treatment in the
attention focusing than distraction group. There were no significant group differences in HR reactions.
Apart from the extent to which (IDCI) control was desired over dental treatment, all variables showed
improvement from before to after the treatment session. A major difference between the present and previous
studies with discrepant results is the duration of exposure. Craske et al. (1991) exposed animal phobics
for 6 min, Rodriguez and Craske (1995) exposed their phobic participants for 15 min and Penfold and
Page (1999) exposed theirs for 10 min as did Oliver and Page (2003) and Johnstone and Page (2004), although
the latter two authors did so for three times weekly. In these studies, SUDs, and in case of Johnstone and
Page (2004) study also BAT, benefited more from distraction. In the present study, exposure lasted for
60 min, which conveyed a moderate advantage to attention focusing. The latter condition was also beneficial
after a 30-min (Kamphuis & Telch, 2000) and a 21-min (Mohlman & Zinbarg, 2000) exposure. It therefore
seems likely that the length of the present exposure session accounts for the discrepancy compared with
previous results. An exception in the pattern of results is the study by Grayson et al. (1986) in which
distraction had a better effect on SUD than attention focusing after a 90-min exposure. However,
obsessional–compulsive (OCD) patients took part rather than phobic participants as in the subsequent
studies.
One of the shortcomings of the study is the lack of a manipulation check. Although the therapist sat next to
the patients during the exposure sessions and drew their attention either toward or away from the phobic
stimulus, it is conceivable that patients may have avoided cognitively in the attention-focusing condition or
attended the dental instruments in the distraction condition. A manipulation check such as subsequent
questions as to characteristics of the instruments would have confirmed differences in the extent of stimulus
processing in the two groups. Similarly, there was no check on treatment credibility. The instructions were
intended to convince the participants in both groups that the respective treatment would help them overcome
their dental fear. In the absence of a credibility check, it, however, cannot be stated with certainty that both
groups had similar expectations with regard to the efficacy of exposure.
Both HR reactions to phobia-related pictures and behavioural avoidance showed improvement in both
groups. As far as HR is concerned, the decrease from before to after treatment may be due to habituation
to the repeated presentation of the picture material. It is one of the shortcomings of this study that there is no
control group without treatment, without which the improved HR reactions cannot be attributed to
treatment. In case of behavioural avoidance, it is likely that treatment contributed to the improvement.
Participants had not visited a dentist for 6–7 years on average but more than 60% of them did so within weeks
after the treatment session. The lengthy avoidance baseline suggests that even that brief psychological
treatment can improve dental treatment adherence.
Groups did not differ with regard to avoidance behaviour subsequent to treatment. Avoidance was
significantly correlated with IDCI desire for control over dental treatment. This is a replication of a previous
 ARTICLE IN PRESS
2702 B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703

finding (Sartory et al., 2006). In that study, the IDCI variable failed to correlate with other measures of self-
efficacy or locus of control. Rather than being a personality trait, the desire for control in the dental setting
appears to be specific to dental anxiety but independent of its severity. Instrumental control was previously
found to have an anxiety-relieving effect in phobics (Sartory & Daum, 1992). It is conceivable that patients
desiring more control experienced its anxiety-relieving effect or rather, the fear-inducing effect of its lack at
some earlier stage during the lengthy course of their phobic history. Avoidance may confer a measure of
control over phobic anxiety and therefore its immediate reduction.
Summarizing, attention focusing during a 60-min exposure session led to a better outcome than distraction
in terms of state anxiety and a marginally better one in terms of SUD and dysfunctional cognitions in dental
phobics. There were no significant group differences with regard to HR reactions to phobia-related material or
avoidance with both, together with other measures, showing improvement from before to after the treatment
session.

Acknowledgements

This study was supported by a grant from the Deutsche Forschungsgemeinschaft (SA 735/14-1). We are
grateful for the help in data collection by Tanja List, Marianne Behl, Nurten Nuyan, Elvira Sons and Julia
Taake.

References

Antony, M. M., McCabe, R., Leeuw, I., Sano, N., & Swinson, R. P. (2001). Effect of distraction and coping style on in vivo exposure for
specific phobia of spiders. Behaviour Research and Therapy, 39, 1137–1150.
Bandura, A. (1983). Self-efficacy determinants of anticipated fears and calamities. Journal of Personality and Social Psychology, 45,
464–469.
Bandura, A. (1988). Self-efficacy conception of anxiety. Anxiety Research, 1, 77–98.
Brown, T. A., Di Nardo, P. A., & Barlow, D. H. (1994). Anxiety Disorders Interview Schedule for DSM-IV (ADIS-IV). San Antonio, TX:
Psychological Corporation/Graywind Publications Inc.
Brunsman, B. A., Logan, H. L., Patil, R. R., & Baron, R. S. (2003). The development and validation of the revised Iowa Dental Control
Index (IDCI). Personality and Individual Differences, 34, 1113–1128.
Corah, N. L. (1969). Development of a dental anxiety scale. Journal of Dental Research, 48, 596–602.
Corah, N., Gale, E. N., & Illig, S. J. (1978). Assessment of a dental anxiety scale. Journal of the American Dental Association, 97, 816–819.
Craske, G. M., Street, L., & Barlow, D. H. (1989). Instructions to focus upon or distract from internal cues during exposure treatment of
agoraphobic avoidance. Behaviour Research and Therapy, 6, 663–672.
Craske, G. M., Street, L. L., Jayaraman, J., & Barlow, D. H. (1991). Attention versus distraction during in vivo exposure: Snake and
spider phobias. Journal of Anxiety Disorders, 5, 199–211.
De Jongh, A., Muris, P., Ter Horst, G., & Schoenmakers, N. (1995). Negative cognitions of dental phobics: Reliability and validity of
Dental Cognitions Questionnaire. Behaviour Research and Therapy, 5, 507–515.
Foa, E. B., & Kozak, M. J. (1986). Emotional processing of fear: Exposure to corrective information. Psychological Bulletin, 1, 20–35.
Grayson, J. B., Foa, E. B., & Steketee, G. S. (1982). Habituation during exposure treatment: Distraction versus attention focusing.
Behaviour Research and Therapy, 20, 323–328.
Grayson, J. B., Foa, E. B., & Steketee, G. S. (1986). Exposure in vivo of obsessive-compulsives under distracting and attention-focusing
conditions: Replication and extension. Behaviour Research and Therapy, 24, 475–479.
Hautzinger, M., Bailer, M., Worall, H., & Keller, F. (1994). Beck-Depressions-Inventar. Bern: Huber.
Johnstone, K. A., & Page, A. C. (2004). Attention to phobic stimuli during exposure: The effect of distraction on anxiety reduction, self-
efficacy and perceived control. Behaviour Research and Therapy, 42, 249–275.
Jones, M. K., & Menzies, R. G. (2000). Danger expectancies, self-efficacy and insight in spider phobia. Behaviour Research and Therapy,
38, 585–600.
Kamphuis, J. H., & Telch, M. J. (2000). Effects of distraction and guided threat reappraisal on fear reduction during exposure-based
treatments for specific fears. Behaviour Research and Therapy, 38, 1163–1181.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1995). International affective picture system: Technical manual and affective ratings.
Gainesville: University of Florida.
Laux, L., Glanzmann, P., Schaffner, P., & Spielberger, C. D. (1981). Das State-Trait-Angst-Inventar. Weinheim: Beltz.
Logan, H. L., Baron, R. S., Keely, K., Law, A., & Stein, S. (1991). Desired control and felt control as mediators of stress in a dental
setting. Health Psychology, 10, 352–359.
Mohlman, J., & Zinbarg, R. E. (2000). What kind of attention is necessary for fear reduction? An empirical test of the emotional
processing model. Behavior Therapy, 31, 113–133.
 ARTICLE IN PRESS
B. Schmid-Leuz et al. / Behaviour Research and Therapy 45 (2007) 2691–2703 2703

Oliver, N. S., & Page, A. C. (2003). Fear reduction during in vivo exposure to blood-injection stimuli: Distraction vs. attentional focus.
British Journal of Clinical Psychology,, 42, 13–25.
Penfold, K., & Page, A. C. (1999). The effect of distraction on within-session anxiety reduction during brief in vivo exposure for mild
blood-injection fears. Behavior Therapy, 30, 607–621.
Rodriguez, B. I., & Craske, M. G. (1995). Does distraction interfere with fear reduction during exposure? A test among animal-fearful
subjects. Behavior Therapy, 26, 337–349.
Sartory, G., & Daum, I. (1992). Effects of instrumental controllability on the phasic cardiac response and subjective fear of phobic
stimulation. Journal of Psychophysiology, 6, 131–139.
Sartory, G., Heinen, R., Pundt, I., & Jöhren, P. (2006). Predictors of behavioral avoidance in dental phobia: The role of gender,
dysfunctional cognitions and the need for control. Anxiety, Stress, and Coping, 19, 279–291.
Sartory, G., Rachman, S., & Grey, S. J. (1982). Return of fear: The role of ‘rehearsal’. Behaviour Research and Therapy, 20, 123–133.
Schneider, S., & Margraf, J. (2006). Diagnostisches Interview bei psychischen Störungen: DIPS. 3. Auflage. Berlin: Springer.
Wälte, D., Ebe, H., Brandenburg, U., & Kröger, F. (1999). Kognitive Selbstregulation bei somatoformen Störungen. In F. Kröger, &
E. Petzold (Eds.), Selbstorganisation und Ordnungswandel in der Psychosomatik. Stuttgart: VAS.