Journal of Child Psychology and Psychiatry 46:11 (2005), pp 1152–1160 doi: 10.1111/j.1469-7610.2005.00411.

A randomised controlled trial of a CBT

intervention for anxiety in children with
Asperger syndrome
Kate Sofronoff,1 Tony Attwood,2 and Sharon Hinton1
1
School of Psychology, University of Queensland, Australia; 2Asperger’s Syndrome Clinic, Queensland, Australia

Background: The aim of the study was to evaluate the effectiveness of a brief CBT intervention for
anxiety with children diagnosed with Asperger syndrome (AS). A second interest was to evaluate
whether more intensive parent involvement would increase the child’s ability to manage anxiety outside
of the clinic setting. Methods: Seventy-one children aged ten to twelve years were recruited to parti-
cipate in the anxiety programme. All children were diagnosed with AS and the presence of anxiety
symptoms was accepted on parent report via brief interview. Children were randomly assigned to one of
three conditions: intervention for child only, intervention for child and parent, wait-list con-
trol. Results: The two intervention groups demonstrated significant decreases in parent-reported
anxiety symptoms at follow-up and a significant increase in the child’s ability to generate positive
strategies in an anxiety-provoking situation. There were a number of significant differences between the
two interventions to suggest parent involvement as beneficial. Conclusions: The sample of children
with AS in this study presented with a profile of anxiety similar to a sample of clinically diagnosed
anxious children. The intervention was endorsed by parents as a useful programme for children dia-
gnosed with Asperger syndrome and exhibiting anxiety symptoms, and active parent involvement
enhanced the usefulness of the programme. Limitations of the study and future research are dis-
cussed. Keywords: Asperger syndrome, anxiety, CBT, parent involvement. Abbreviation: SCAS-P:
Spence Child Anxiety Scale – Parent.

Since the term ‘Asperger syndrome’ was first intro-

Anxiety and Asperger syndrome
duced by Lorna Wing in 1981, interest in the disorder
has generated an increasing body of research, and a The DSM-IV-TR description of Asperger syndrome
corresponding growth in articles and books on the includes reference to an association between the
subject (e.g., Attwood, 1998; Frith, 1991; Howlin, disorder and secondary mood disorders, especially
1998; Ozonoff, Dawson, & McPartland, 2002; Schop- anxiety disorders (American Psychiatric Association,
ler, Mesibov, & Kunce, 1998). However, whilst much 2000). Whilst few studies have focused solely on
has been written about the disorder, very little has anxiety, Kim, Szatmari, Bryson, Streiner, and Wilson
been produced in terms of sound evidence-based (2000) evaluated the rate of mood disorders and
interventions for the population. anxiety in a cohort of children diagnosed with High
Asperger syndrome (AS) is characterised as a dis- Functioning Autism (HFA) and AS compared with a
order primarily involving socio-emotional difficulties community sample. Using parent report, they found
(Frith, 1991). Social deficits are evident and attempts elevated rates of both mood disorders and anxiety in
made in social contexts are often clumsy and in- children with HFA and AS, with 16.9% reporting
appropriate (Wing, 1981, 1991). Whilst verbal ex- depressive symptoms at least two standard devia-
pression can be fluent, the content of conversations tions above the general population mean, 13.6%
may be narrow and reflect circumscribed interests reporting generalised anxiety at this level and 8.5%
and the child may continue a conversation with little reporting separation anxiety at this level. Further-
regard for the interest of the listener. The child may more, children with anxiety and mood problems were
be motivated to form friendships but is unlikely to also rated as more aggressive, more demanding of
possess the skills to succeed (Happé & Frith, 1996; parents and as having poorer relationships with
Myles & Simpson, 2002). The inability to read social peers and teachers.
cues and to respond to these is usually a central Another study that examined psychiatric func-
deficit. There is likely to be a preference for fixed tioning in young adolescents with AS found that
routines and structure, with significant distress at compared to a group of adolescents with conduct
even minor changes. There can be signs of motor disorder, the young people with AS showed signific-
clumsiness, problems with handwriting and hyper- antly greater levels of anxiety symptoms such as
sensitivity to specific auditory and tactile experien- worrying, hypochondriasis, nonsituational anxiety
ces. There can also be problems with organisational or panic, and specific fears (Green, Gilchrist, Burton,
and time management skills (Attwood, 1998). & Cox, 2000).

Association for Child Psychology and Psychiatry, 2005.
Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
 14697610, 2005, 11, Downloaded from https://acamh.onlinelibrary.wiley.com/doi/10.1111/j.1469-7610.2005.00411.x by University Of Trnava, Wiley Online Library on [22/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
CBT anxiety intervention for children with Asperger syndrome 1153

In their everyday lives, children with AS experience combined group showing greater improvement on a
difficulties with social reasoning and friendships, number of measures. At 6-year follow-up, treatment
Theory of Mind abilities, empathy and the pragmatic gains were maintained, with little difference between
aspects of language. They are also more likely to the two groups (Barrett, Duffy, Dadds, & Rapee,
have a specific learning disability and heightened 2001). It is considered important to evaluate the
sensory awareness, which can result in significant outcome for both child-only and child + parent
levels of stress likely to contribute to anxiety. These involvement for the AS population using a model
deficits provide illustration of the struggle children similar to that of the Barrett et al. studies, that is,
face in navigating their day-to-day world and it may training parents as co-therapists. Given the difficulty
be, as suggested by Kim et al. (2000), that interven- for a child with AS to generalise material learned
tions aimed at reducing anxiety for children with AS from one context to another, it is expected that in-
may also reduce aggression and thereby enhance creased parental involvement will result in a better
functioning and improve relationships with family, outcome for the child, especially in use of strategies.
peers and teachers. The aim of the current intervention is to work with
children with AS who are experiencing anxiety, to
teach them effective strategies to manage feelings
Cognitive Behaviour Therapy and Asperger
and to encourage a broadening of their emotional
syndrome
and behavioural repertoire. In the present study, the
Strategies initially developed for children with aut- questions of interest are whether a brief CBT inter-
ism have been used with some success with an AS vention for anxiety is effective in reducing sympto-
population (e.g., Social Stories and Comic Strip matology in children diagnosed with AS and whether
Conversations; Gray, 1998). The intervention devel- there is a significant effect of parent involvement in
oped for this study, whilst incorporating those the intervention.
strategies, was developed specifically for children
with AS and has its basis in the sound theoretical
framework of Cognitive Behaviour Therapy (CBT). As
outlined by Attwood (2003), it is necessary to Method and procedures
accommodate the cognitive profile of the child with Participants
AS when conducting CBT and to modify the ap-
proach to achieve this. CBT for typical children with Seventy-one children, aged 10 to 12 years, were
anxiety has been refined and evaluated over several recruited to participate in the 6-week intervention.
Participants were recruited via local newspapers, radio
decades (Graham, 1998; Kendall, 2000) but has only
and the Asperger Syndrome Support Network news-
recently been applied to children and adults with AS
letter. All children included in the study had a primary
(Hare & Paine, 1997). diagnosis of AS from a paediatrician. A semi-structured
Although few empirical studies have been repor- phone interview with a parent was used to establish
ted, outcomes have been positive. A recent case- DSM-IV criteria and a further check, the Childhood
study demonstrated good results from a CBT Asperger Syndrome Test with items based on DSM-IV
approach with obsessive-compulsive disorder in a criteria (CAST; Scott, Baron-Cohen, Bolton, & Brayne,
young girl with AS (Reaven & Hepburn, 2003). Re- 2002), was included in the initial questionnaire battery.
searchers conducting a CBT intervention to facilitate The primary distinction between a diagnosis of AS and
social-emotional understanding and to increase so- High Functioning Autism is based on whether the child
cial interaction in children with HFA were also able had a significant language delay before the age of
3 years. Whilst no parents in the current study reported
to demonstrate significant improvement in positive
significant language delay, five children did not meet
social interaction, emotional understanding and
the symptom score on the CAST (>15) and, although
social problem-solving (Bauminger, 2002). they participated in the intervention, their data were
excluded from analyses. Parents were asked about
Parental involvement additional diagnoses but no children were excluded on
the basis of comorbid disorders.
There is a body of literature that focuses on the role The presence of child anxiety was established via
of parental involvement in the treatment of childhood parent report at the initial phone interview when par-
anxiety. Kendall and Choudhury (2003), in a review ents were asked a series of questions to operationalise
paper, stress the importance of parent involvement symptomatology. Parents completed consent forms and
and the need to accommodate developmental differ- children gave assent to participation. Families were
randomly assigned to Intervention 1 (child only), Inter-
ences when using CBT with children. Mendlowitz
vention 2 (child + parent) or wait-list following receipt of
et al. (1999) found that a combined child + parent
consent forms. Children were allocated to groups of
group demonstrated improvement relative to a child- three within each intervention by age and sex, with girls
only condition in increased use of adaptive coping grouped together. There were no significant differences
strategies in the child. Barrett, Dadds, and Rapee among groups at pre-intervention. See Table 1 for
(1996) also found a significant difference between demographic details of the sample. Intervention 1 and
child-only and child + family groups, with the the wait-list group ran concurrently and Intervention 2
 14697610, 2005, 11, Downloaded from https://acamh.onlinelibrary.wiley.com/doi/10.1111/j.1469-7610.2005.00411.x by University Of Trnava, Wiley Online Library on [22/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1154 Kate Sofronoff, Tony Attwood, and Sharon Hinton

Table 1 Child demographics described across intervention groups

Group Age Child anxiety (SCAS) IQ CDI CAST

Intervention 1 (child only)

N ¼ 23 (+2 non AS participants)*
Boys ¼ 20; Girls ¼ 3
ADHD ¼ 10; Depression ¼ 2
Mean 10.56 29.29 107.5 11.50 21.59
Std. dev. .99 17.45 27.3 8.27 4.3
Minimum 9.00 6.00 90 1.00 15
Maximum 12.00 78.00 137 30.00 27
Wait-list group
N ¼ 23
Boys ¼ 20; Girls ¼ 3
ADHD ¼ 8; Depression ¼ 1
Mean 10.75 27.46 101.0 8.62 20.46
Std. dev. 1.04 12.19 27.2 5.80 3.3
Minimum 9.00 2.00 95 2.00 15
Maximum 12.00 49.00 125 21.00 25
Intervention 2 (child + parents)
N ¼ 25 (+3 non-AS participants)*
Boys ¼ 22; Girls ¼ 3
ADHD ¼ 12; Depression ¼ 3
Mean 10.54 32.23 105.6 10.25 20.71
Std. dev. 1.26 17.71 21.2 7.82 3.8
Minimum 9.00 6.00 90 1.00 15
Maximum 12.00 91.00 135 25.00 25

SCAS ¼ Spence Children’s Anxiety Scale (Spence, 1995).

IQ ¼ Short form WISC-III, 3 verbal + 3 performance subscales prorated.
CDI ¼ Children’s Depression Inventory (Kovacs, 1992).
*Data was not included from non-AS participants.

Table 2 Parent ratings of anxiety in children with Asperger syndrome, clinically anxious children and a non-clinical sample

AS sample (data Clinical sample (data Non-clinical sample

SCAS-P Ratings from current sample) from Nauta et al.) (data from Nauta et al.)

Total score 37.30** 31.90 14.2

Panic/Agorophobia 4.30 3.2 0.8
OCD 5.10** 3.0 1.1
Social phobia 8.03 7.7 4.2
Physical injury 5.43* 4.5 2.8
Separation anx. 7.35 6.9 2.6
GAD 7.09 6.6 2.7

*significantly higher than clinical sample, p < .05.

**significantly higher than clinical sample, p < .001.

began immediately following completion of Intervention

Procedure
1. The wait-list group completed the intervention
(child + parent) following their final data collection. The therapists conducting the child groups were post-
Fifteen families participated and their results were very graduate students enrolled in the clinical psychology
similar to Intervention 2. programme at the University of Queensland who were
Independent groups t-tests were used to compare completing their second internship. All therapists par-
scores on the SCAS-P from the children with AS, chil- ticipated in a one-day training workshop prior to com-
dren diagnosed as clinically anxious, and normally mencing the intervention, worked from a therapist’s
developing children. The clinically anxious and com- manual and received weekly supervision. Ethical
parison samples were taken from Nauta et al. (2004). approval for the study was gained from the University of
The purpose of the preliminary analysis was to provide Queensland ethics committee in accordance with
descriptive data on the type and severity of anxiety standards required by the National Health and Medical
reported by parents of the children in the current study. Research Council of Australia.
Results showed significant differences between the AS In Intervention 1 (child only), eight groups were
group and the normal sample but also significant dif- formed. Each group comprised three children and two
ferences between the AS group and the clinically anxi- therapists. Parents received no training but after each
ous children on the OCD subscale and the physical session met briefly with therapists to access feedback
injury fears subscale, with the AS group reporting sig- on the child’s participation in sessions and to be told
nificantly higher scores (see Table 2). about the weekly home-based projects.
 14697610, 2005, 11, Downloaded from https://acamh.onlinelibrary.wiley.com/doi/10.1111/j.1469-7610.2005.00411.x by University Of Trnava, Wiley Online Library on [22/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
CBT anxiety intervention for children with Asperger syndrome 1155

In Intervention 2 (child + parents), there were 9 Materials

groups of three children each with two therapists. Par-
ents from this intervention formed two ‘parent groups’ Therapists conducting the sessions followed the format
and a therapist trained parents to work as co-therapists outlined in the Trainer’s manual and completed
in all components of the intervention. They were asked checklists indicating components of the session that
to encourage and coach use of strategies in different had been covered. Twenty-five percent of sessions were
situations and to encourage completion of the home- videotaped and independent raters viewed these to as-
based projects. sess content fidelity. Approximately 98% of the material
was found to have been accurately covered within the
sessions.
Measures
The first measure, ‘James and The Maths Test’
Intervention
(Attwood, 2002), required each child to generate strat-
egies for ‘James’ to cope with anxiety in the situation The CBT programme was designed to be highly struc-
outlined (see Appendix A). Administration was stan- tured, informative and entertaining. Every child re-
dardized, with the scenario read aloud to each child and ceived a workbook and materials for the six two-hour
responses recorded by a therapist. sessions that included information on being happy,
The second questionnaire, a parent-report measure relaxed and anxious, with space for individual com-
of child anxiety, the Spence Child Anxiety Scale – ments and responses to questions. At the end of each
Parent (SCAS-P, Nauta et al., 2004), contains 39 session, a project was explained to participants and the
questions relating to situations in which a child could completed project was discussed within the group at
experience feelings of anxiety, and parents indicate on the start of the next session. A metaphor was created of
a scale (0–3) how applicable the feelings would be to the child as scientist or astronaut exploring a new
their child. The scale yields a total score and six sub- planet. The authors had noted that children with AS,
scales, Generalised Anxiety Disorder (GAD), Obsessive within the age range of ten to twelve years (and pre-
Compulsive Disorder (OCD), Specific Phobia (SP), Panic dominantly boys), often have a special interest in sci-
and Agoraphobia (PA), Separation Anxiety Disorder ence and science fiction.
(SAD), Social Anxiety (SA). The total scale demonstra- Session one explored two positive emotions, happi-
ted high internal validity with this population (Cron- ness and relaxation, with a range of group and indi-
bach’s Alpha .92) and internal reliability coefficients for vidual activities to measure and compare emotions in
the six subscales ranged from adequate to excellent specific situations. Session two was an exploration of
(.61 for physical injury fears to .81 for panic and ago- anxiety and recognition of the changes that occur in
raphobia). physiology, thinking, behaviour and speech. The con-
The child version of this measure was used at pre- cept of a ‘tool box’ with different types of tools to ‘fix the
intervention and these scores appear in Table 1. feeling’ was explained, with a focus on physical tools
Significant difficulty was experienced gaining anxiety- that provide a constructive release of emotional energy
related information from child informants and as other (e.g., going for a run or bouncing on the trampoline),
authors have noted similar difficulty in accessing and and relaxation tools that lower the heart rate (e.g., lis-
reporting emotions in this population (Capps, Yirmiya, tening to music or reading a book). Session three ex-
& Sigman, 1992), it was decided not to retain this plored social tools; for example, how other people can
measure at post or follow-up. help restore positive feelings through words, gestures of
The third measure was a parent-report measure reassurance and affection or how avoiding social con-
gauging the level of social worry experienced by the tact, solitude, can be an effective emotional restorative
child, the Social Worries Questionnaire–Parent for children and adults with Asperger syndrome. Also
(Spence, 1995). The questionnaire consists of 10 explored in session three were thinking tools, a category
questions relating to feelings of worry in a range of of activities or thoughts that test the reality and prob-
social situations and asks parents to rate (0–2) how ability of feared outcomes. In session four participants
true these feelings have been for their child in the discovered a range of measures of the degrees of emo-
previous four weeks. The scale yields a total score tion, with some time spent exploring the concept of a
taken across the behaviours nominated to have ‘thermometer’. Group discussion then explored how
occurred. The scale demonstrated good internal reli- each participant could share strategies or tools to suc-
ability, with a coefficient alpha of .82 for the population cessfully manage their anxiety. In session five, partici-
in this study. pants explored how Social Stories (Gray, 1998) can be
Measures were administered on three occasions, pre- used for emotion management, and the concept of cre-
intervention (Time 1), immediately post-intervention ating an ‘antidote’ to poisonous or noxious thoughts. In
(Time 2) and at six-week follow-up (Time 3). the final session, the participants worked together
designing a programme for each participant to improve
their management of anxiety.
Scoring
James and the Maths Test was scored by allocating 1
point for each positive strategy generated. Scoring was Results
checked for inter-rater reliability using independent
raters blind to the child’s intervention condition (wait- Preliminary analyses revealed that the distribution
list vs. intervention 1 or 2). The reliability check found was normal and homogeneous with no missing data.
99% agreement between raters. Five children failed to complete the intervention, one
 14697610, 2005, 11, Downloaded from https://acamh.onlinelibrary.wiley.com/doi/10.1111/j.1469-7610.2005.00411.x by University Of Trnava, Wiley Online Library on [22/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1156 Kate Sofronoff, Tony Attwood, and Sharon Hinton

child from Intervention 1, three from the wait-list difference between the two intervention groups at
group, and one from Intervention 2. ANOVA showed Time 3 (p < .025), with the combined group
no significant differences between completers and (child + parent) showing greater improvement.
noncompleters on any measure. A Bonferroni The separation anxiety subscale showed a main
adjustment was used with all analyses to take into effect for Time, F(2,78) ¼ 11.34, p < .0001 and a
account the number of analyses conducted. main effect for Group, F(2,79) ¼ 5.24, p < .01. There
was also a significant Time · Group interaction,
F(4,156) ¼ 11.21, p < .0001. Post-hoc tests showed
Parent reports of anxiety – SCAS-P
that there was a significant difference for both In-
A series of repeated measures analyses of variance tervention 1 (p < .01) and Intervention 2 (p < .0001)
was conducted to compare parent reports of child between Time 1 and Time 3. There were no differ-
anxiety across time (T1 pre-intervention, T2 post- ences across time for the wait-list group. Whilst
intervention and T3 follow-up) and between groups there was no difference between groups at Time 1,
(Intervention 1 (child only), Intervention 2 there was a significant difference between the wait-
(child + parent), and wait-list control group). Analy- list group and both Intervention 1 (p < . 02) and In-
ses were conducted for the total scale and for the six tervention 2 (p < .001) at Time 3 and also between
subscales. the two intervention groups (p < .02), with the
Results from the total score of the SCAS-P showed combined group showing greater improvement.
a significant main effect for Time, F(2,78) ¼ 38.95, Results for the subscales measuring obsessive
p < .0001. There was also a significant Time · Group compulsive tendencies, socially phobic tendencies,
interaction, F(4,158) ¼ 9.16, p < .0001. Post-hoc and generalised anxiety tendencies also demon-
comparisons of the simple effects showed that scores strated main effects for Time and Time by Group
on the SCAS-P for the two intervention groups were interactions such that scores were lower in the
significantly different from Time 1 to Time 3 at intervention groups by follow-up but not in the wait-
p < .0001, such that parents reported fewer anxious list group. The means and standard deviations are
symptoms at Time 3. There was also a significant presented in Table 3.

Table 3 Means and standard deviations for parent reported anxiety scores across time

Time 1 Time 2 Time 3

SCAS-P total Pre Post Follow-up Significance T1 to T3

Wait-list 36.64 (16.67) 35.61 (13.34) 36.32 (13.3)

Intervention1 40.23 (20.42) 36.17 (16.21) 29.42** (15.3) p < .01
Intervention 2 35.25 (16.44) 32.25 (14.56) 21.11a** (10.1) p < .01
Separation Anxiety
Wait-list 8.18 (4.45) 7.46 (4.52) 8.89 (4.5)
Intervention 1 7.01 (4.46) 6.59 (3.83) 5.38** (3.7) p < .01
Intervention 2 6.79 (4.71) 5.03* (3.55) 3.03b*** (2.2) p < .0001
OCD
Wait-list 4.96 (4.08) 5.11 (3.86) 5.71 (4.4)
Intervention 1 5.11 (4.65) 4.09 (3.67) 2.88*** (2.9) p < .0001
Intervention 2 5.21 (3.67) 3.86* (2.58) 2.21*** (1.9) p < .0001
Social phobia
Wait-list 7.21 (4.75) 8.18 (4.74) 6.61 (4.49)
Intervention 1 9.03 (4.45) 8.04 (3.78) 7.38* (4.43) p < .05
Intervention 2 7.93 (5.03) 8.50 (4.95) 5.96** (4.32) p < .01
Panic
Wait-list 4.18 (4.28) 3.25 (3.21) 3.35 (2.7)
Intervention 1 4.88 (4.67) 3.77 (4.32) 3.31** (3.5) p < .01
Intervention 2 3.89 (2.84) 3.86 (3.35) 2.68* (2.4) p < .05
Personal Injury
Wait-list 5.54 (3.28) 5.18 (3.44) 5.17 (3.8)
Intervention 1 5.69 (3.37) 6.17 (3.98) 5.00 (3.9)
Intervention 2 5.07 (2.97) 5.71 (3.54) 3.32*** (2.2) p < .0001
GAD
Wait-list 6.57 (3.65) 6.42 (3.19) 6.57 (3.56)
Intervention 1 8.42 (3.41) 7.50 (3.19) 5.46*** (2.67) p < .0001
Intervention 2 6.36 (3.30) 5.28* (2.53) 3.89*** (2.36) p < .0001

a ¼ significantly different from Intervention 1, p < .05.

b ¼ significantly different from Intervention 1, p < .02.
*Time 2 significantly different from Time 1, p < .01.
**Significantly different from Time 1 at p < .01.
***Significantly different from Time 1 at p < .0001.
 14697610, 2005, 11, Downloaded from https://acamh.onlinelibrary.wiley.com/doi/10.1111/j.1469-7610.2005.00411.x by University Of Trnava, Wiley Online Library on [22/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
CBT anxiety intervention for children with Asperger syndrome 1157

14 8
Intervention
12 Wait-List
7
Number of social worries

Int. + Parents
10
6

Number of strategies
8
5
6

4
4 Wait-list
child only
2 3
child+parents

0 2
Time 1 Time 2 Time 3
Time
1
Figure 1 Parent ratings of social worries in their chil-
dren across time
0
Time 1 Time 2 Time 3
Time
Parent reports of social worries – SWQ (Spence,
1995) Figure 2 Number of strategies generated by children
pre-intervention, post-intervention and at follow-up
A repeated measures analysis of variance was con-
ducted on the Social Worries Questionnaire. The
There was also a significant difference between In-
results showed a significant main effect for Time,
tervention 1 and Intervention 2 at Time 2, p < .01
F(2,78) ¼ 11.86, p < .0001 and a significant Time by
and at Time 3, p < .0001, with the combined group
Group interaction, F(4,156) ¼ 11.06, p < .0001.
generating significantly more strategies. These re-
Post-hoc tests showed a significant effect between
sults appear in Figure 2.
Time 1 and Time 3 for Intervention 1 (child only),
p < .001 and for Intervention 2 (child + parents),
p < .0001, with parents reporting a decreased num-
Discussion
ber of social worries in their children. There was no
difference across time for the wait-list group and One question posed by this study was whether a
there was a significant difference between the two brief CBT intervention for anxiety would effectively
intervention groups and the wait-list group at Time 3 reduce symptomatology in children diagnosed with
– Intervention 1, p < .01 and Intervention 2, AS. A second question was whether there would be a
p < .0001. The difference between the two interven- positive effect of parent involvement on children’s
tion groups approached significance at Time 3, p ¼ use of the strategies and therefore an increase in
.06. These results are presented in Figure 1. effectiveness.
Whilst it is recognised that anxiety is a significant
feature in the presentation of many children, there is
James and the Maths Test – strategies generated by
little written about the nature and extent of anxiety
children
in children diagnosed with AS. The initial compar-
This analysis evaluated differences over time and ison between the sample of children with AS and a
between groups for the number of strategies gener- clinically anxious sample showed that the current
ated in the hypothetical scenario. There was a sig- sample more closely resembled the clinical sample
nificant main effect for Time, F(2,78) ¼ 104.76, than the non-clinical sample (see Table 2). These
p < .0001 and a significant main effect for Group, results support the findings outlined in Kim et al.
F(2, 79) ¼ 38.19, p < .0001. There was a significant (2000). Furthermore, the children with AS were rated
Time · Group interaction, F(4,158) ¼ 28.31, p < significantly higher by parents than the clinical
.0001. Significantly more strategies were generated sample on two of the anxiety subscales, obsessive
at Time 2, p < .0001 and at Time 3, p < .0001 in both compulsive tendencies and physical injury fears.
of the intervention groups compared with Time 1. Whilst we had expected that the sample would
Whilst there was no difference between groups at exhibit anxiety, since they were recruited to particip-
Time 1, there was a significant difference between ate in an anxiety programme, it is nonetheless
each of the intervention groups and the wait-list interesting to note the pattern of anxiety reported by
group at Time 2, p < .0001 and at Time 3, p < .0001. parents.
 14697610, 2005, 11, Downloaded from https://acamh.onlinelibrary.wiley.com/doi/10.1111/j.1469-7610.2005.00411.x by University Of Trnava, Wiley Online Library on [22/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1158 Kate Sofronoff, Tony Attwood, and Sharon Hinton

The higher levels of OCD may be a reflection of the The child measure of ability to generate effective
general characteristics of Asperger syndrome (i.e., strategies to deal with an anxiety-provoking situ-
unusual or repetitive routines or behaviours). Pre- ation, ‘James and the Maths Test’, showed a signi-
vious literature has suggested that children with AS ficant increase in strategies from Time 1 to Time 3. It
use their repetitive, ritualistic behaviours as a is particularly interesting to note that a child’s ability
means of reducing anxiety (Despert, 1965) or that to improve on ‘James and the Maths Test’ requires
they occur as a consequence of experiencing anxiety the child to generalise strategies from one situation
(Jolliffe, Landsdown, & Robinson, 1992). to another and it is an important finding that the
The reason for the higher endorsement of phys- children whose parents received training demon-
ical injury fears may also be congruent with char- strated the best results on this measure.
acteristics of the disorder. Children with AS often
experience heightened physical sensitivity to noise
Effect of parent involvement
and touch and frequently find touch of any kind
distressing and threatening (Attwood, 1998). They There were several indicators that active parental
can become distressed by the normal jostling and participation in the anxiety programme produced
activity of the school environment and may inter- significant benefits both for parents themselves and
pret this in a negative way. Furthermore, many of for the children, with parent evaluations providing
these children experience bullying in the school overwhelming endorsement for the parent compon-
environment (Attwood, in press; Little, 2002) and ent of the programme. In the evaluations, parents
this may also cause anxiety related to the possibil- reported feeling competent with the content of the
ity of physical injury. programme and therefore able to assist their child,
empowered by meeting parents with similar experi-
ences, supported by the group and able to share not
Effectiveness of brief CBT with children diagnosed
only problems but solutions.
with Asperger syndrome
The results from the measures demonstrate sev-
The results demonstrated that the programme was eral significant differences found between the two
effective in reducing anxiety symptomatology repor- intervention groups. On the total score of the SCAS-P
ted by parents in children diagnosed with AS. The at Time 3 there was a significant difference between
results from the Spence Children’s Anxiety Scale – the two groups, with the combined group showing a
Parent form shows a significant reduction in parent- greater reduction in parent-reported symptoms. This
rated symptoms from pre-intervention to six-week is also the case for the separation anxiety subscale of
follow-up, both on the total score and on individual the SCAS-P. On the Social Worries Questionnaire the
subscales. It is important to note the change across difference between the two intervention groups ap-
the subscales because children presented with a proached significance at follow-up, with the com-
broad range of fears; they were not a homogeneous bined group showing better results. The most
group. It is also interesting to note that although in compelling difference between the two groups can be
some cases there was a significant change from pre- seen on the child measure ‘James and the Maths
intervention to immediately post-intervention (Time Test’. In the combined group it was possible for the
2), in many cases there was no significant change therapist to give much more encouragement to par-
until the six-week follow-up (Time 3). Whilst it is ents in their efforts to promote use of strategies by
common in clinical interventions to see a significant the children. Parents had time each week to bring
change from pre- to post-intervention and then a re- problems to the group and brainstorm ways to ap-
turn closer to baseline by follow-up, this did not occur proach their child and encourage use of new skills.
with the current sample. It seems that the uptake and
implementation of strategies took longer and there-
Qualitative findings
fore increased the time before parents reported any
symptom reduction. Whilst this lag could reflect an Whilst conducting intervention trials it is often not
intervening event to explain improvement, this is possible to capture or even to hypothesise some of
unlikely since it occurred in both intervention groups the findings in a quantitative medium. As part of the
but not in the wait-list group. evaluation of the programme we asked parents to
The second measure used showed a similar describe any changes in their child, either positive or
change across time. The Social Worries Question- negative, that they felt could be attributed to parti-
naire (Spence, 1995) demonstrated that parents in cipation in the programme. Many parents pointed to
both intervention groups reported significantly fewer the development of friendships among their children.
social worries in their children at Time 3 compared Some parents noted that children seemed more
with Time 1. Whilst the SCAS-P yields information confident in their day-to-day interactions and sug-
about specific types of anxiety and focuses on gested that the time spent with children similar to
internal feeling states, the SWQ taps information themselves, and with very positive therapists, had
dealing with day-to-day social worries surrounding helped in this respect. Other parents reported that
more overt behaviours. whilst issues still arose on a regular basis for their
 14697610, 2005, 11, Downloaded from https://acamh.onlinelibrary.wiley.com/doi/10.1111/j.1469-7610.2005.00411.x by University Of Trnava, Wiley Online Library on [22/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
CBT anxiety intervention for children with Asperger syndrome 1159

child, they found that the children became dis- trial. Journal of Consulting and Clinical Psychology,
tressed more slowly and recovered more quickly, 64, 333–342.
especially if encouraged to use strategies that they Barrett, P.M., Duffy, A.L., Dadds, M.R., & Rapee, R.M.
had learned. Interestingly, most of the children en- (2001). Cognitive-behavioral treatment of anxiety
joyed attending the university each week and many disorders in children: Long-term (6-year) follow-up.
Journal of Consulting and Clinical Psychology, 69,
were quite dismayed when the programme ended.
135–141.
Bauminger, N. (2002). The facilitation of social-emo-
Clinical implications tional understanding and social interaction in
high-functioning children with autism: Intervention
The study showed good results with the targeted outcomes. Journal of Autism and Developmental
population of young people with a diagnosis of AS Disorders, 31, 461–469.
and was well accepted by parents and children alike. Capps, L., Yirmiya, N., & Sigman, M. (1992). Under-
The study was designed to be of value to clinicians, standing of simple and complex emotions in non-
and, at the behavioural treatment level, there is little retarded children with autism. Journal of Child
to distinguish between AS and HFA. The programme Psychology and Psychiatry, 33, 1169–1182.
is applicable to both groups. It is also the case that Despert, J.L. (1965). The emotionally disturbed child:
Then and now. New York: Robert Brunner.
with very little modification the programme would be
Frith, U. (1991). Autism and Asperger syndrome. Cam-
useful to more severely impaired young people and bridge: Cambridge University Press.
their parents. Graham, P. (1998). Cognitive behaviour therapy for
children and families. Cambridge: Cambridge Univer-
Limitations and future research sity Press.
Gray, C. (1998). Social stories and comic strip conver-
Although the sample recruited for the current sations with students with Asperger syndrome and
study was adequate, it was not a large sample and high functioning autism. In E. Schopler, G.B. Mes-
the findings will need to be replicated in future ibov & L.J. Kunce (Eds.), Asperger syndrome or high
work. It must be acknowledged that the data col- functioning autism. New York: Plenum Press.
lected was largely through parent-report and par- Green, J., Gilchrist, A., Burton, D., & Cox, A. (2000).
Social and psychiatric functioning in adolescents
ents may have had expectations of improvement
with Asperger syndrome compared with conduct
following their participation. It is important that disorder. Journal of Autism and Developmental Dis-
future research collects data from multiple sources orders, 30, 279–293.
and accesses behavioural change as well as par- Happé, F., & Frith, U. (1996). The neuropsychology of
ent-report information. Literature on parental autism. Brain, 119, 1377–1400.
involvement suggests an impact of parents’ own Hare, D.J., & Paine, C. (1997). Developing cognitive
anxiety and whilst this was not addressed in this behavioural treatments for people with Asperger’s
study, it should be considered in future works. The syndrome. Clinical Psychology Forum, 110, 5–8.
use of a parent-only group would also be valuable Howlin, P. (1998). Children with autism and Asperger
to compare with the findings from the extant child syndrome: A guide for practitioners and carers. New
anxiety literature. York: Wiley.
Jolliffe, T., Landsdown, R., & Robinson, T. (1992).
Autism: A personal account. London: The National
Correspondence to Autistic Society.
Kendall, P.C. (2000). Child and adolescent cognitive
Kate Sofronoff, School of Psychology, University of behavioural therapy procedures. New York: Guilford.
Queensland, Brisbane, Australia. Tel: +61 7 3365 Kendall, P.C., & Choudhury, M.S. (2003). Children and
6411; Fax: +61 7 3365 4466; Email: kate@psy. adolescents in cognitive-behavioral therapy: Some
uq.edu.au past efforts and current advances, and the challenges
in our future. Cognitive Therapy and Research, 27,
89–104.
Kim, J.A., Szatmari, P., Bryson, S.E., Streiner, D.L., &
References
Wilson, F.J. (2000). The prevalence of anxiety and
Attwood, T. (1998). Asperger’s syndrome: A guide for mood problems among children with autism and
parents and professionals. London: Jessica Kingsley. Asperger syndrome. Autism, 4, 117–132.
Attwood, T. (2002). James and the Maths Test. Unpub- Kovacs, M. (1992). The Children’s Depression Inventory
lished manuscript. Manual. New York: Multi-Health Systems Incorpor-
Attwood, T. (2003). Frameworks for behavioral inter- ated.
ventions. Child and Adolescent Psychiatric Clinics of Little, L. (2002). Middle-class mothers’ perceptions of
North America, 12, 65–86. peer and sibling victimization among children with
Attwood, T. (in press). Children with Asperger’s syn- Asperger’s syndrome and non-verbal learning disor-
drome are more likely than their peers to be the target ders. Issues in Comprehensive Pediatric Nursing, 25,
for bullying. Australian Journal of Early Childhood. 43–57.
Barrett, P.M., Dadds, M.R., & Rapee, R.M. (1996). Mendlowitz, S.L., Manassis, K., Bradley, S., Scapillato,
Family treatment of childhood anxiety: A controlled D., Miezitis, S., & Shaw, B. (1999). Cognitive-behavioral
 14697610, 2005, 11, Downloaded from https://acamh.onlinelibrary.wiley.com/doi/10.1111/j.1469-7610.2005.00411.x by University Of Trnava, Wiley Online Library on [22/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1160 Kate Sofronoff, Tony Attwood, and Sharon Hinton

group treatments in childhood anxiety disorders: The Wing, L. (1991). The relationship between Asperger’s
role of parental involvement. Journal of the American syndrome and Kanner’s autism. In U. Frith (Ed.),
Academy of Child and Adolescent Psychiatry, 38, 1223– Autism and Asperger syndrome. Cambridge: Cam-
1229. bridge University Press.
Myles, B., & Simpson, R.L. (2002). Asperger syndrome:
An overview of characteristics. Focus on Autism and Manuscript accepted 25 August 2004
Other Developmental Disabilities, 17, 132–138.
Nauta, M.H., Scholing, A., Rapee, R.M., Abbott, M.,
Spence, S.H., & Waters, A. (2004). A parent-report Appendix: James and the Maths Test
measure of children’s anxiety: Psychometric proper-
ties and comparison with child-report in a clinic and James’ teacher is Mrs. Smith. She is a nice and kind
normal sample. Behaviour Research and Therapy, 42, teacher. He really likes the way she manages the
813–839. class. She makes the classroom quiet with no teasing
Ozonoff, S., Dawson, G., & McPartland, J. (2002). A between the children. She helps James with his dif-
parent’s guide to Asperger syndrome and high func- ficulty understanding maths.
tioning autism. New York: Guilford. She has set a difficult maths test for the class on
Reaven, J., & Hepburn, S. (2003). Cognitive-behavioral
Tuesday and James is worried that he won’t do well
treatment of obsessive-compulsive disorder in a child
and the other children will think he is stupid. On the
with Asperger syndrome. Autism, 7, 145–164.
Schopler, E., Mesibov, G.B., & Kunce, L.J. (1998). day of the big maths test, the school principal comes
Asperger syndrome or high functioning autism? New into the class and says that Mrs. Smith is ill today
York: Plenum Press. and that a new teacher will take the class, but they
Scott, F.J., Baron-Cohen, S., Bolton, P., & Brayne, C. still have the maths test. James is very anxious be-
(2002). The CAST (Childhood Asperger Syndrome cause he has a new teacher that day and the children
Test). Preliminary development of a UK screen for become very noisy and silly with a replacement
mainstream primary school age children. Autism, 6, teacher. James is also worried they could tease him
9–31. especially if he doesn’t do well in the maths test.
Spence, S.H. (1995). The Social Worries Questionnaire. Write down what you think James could do and
Social skills training: Enhancing social competence
think to feel less anxious.
with children and adolescents. Windsor: NFER-Nelson.
Wing, L. (1981). Asperger’s syndrome: A clinical
account. Psychological Medicine, 11, 115–129.