Testing a cognitive behavioural therapy program for anxiety in autistic adolescents: a feasibility study

Abstract

Background

Autism includes core symptoms affecting general and social development. Up to 60% of autistic adolescents experience co-occurring anxiety disorders negatively influencing educational, social, and general development together with quality of life. A manualised cognitive behavioural therapy (CBT) program ‘Cool Kids - Autism Spectrum Adaptation (ASA)’ has previously demonstrated efficacy in reducing anxiety in children with co-occurring autism. The current study investigates the feasibility of adapting this program for adolescents.

Methods

Fifteen autistic adolescents, aged 14–17 years, with co-occurring anxiety disorders were enrolled in the study. Outcome measures collected from both adolescents and parents pre-, post-treatment and at 3-month follow-up included participant evaluation of the program, scores from a semi-structured anxiety interview, and questionnaires on anxiety symptoms, life interference, and quality of life.

Results

92% of the families who completed the program found it useful and would recommend it to other families in a similar situation. At follow-up, 55% no longer met the criteria for their primary anxiety diagnosis and 34% of adolescents were free of all anxiety diagnoses. Of the five adolescents who did not attend school before treatment three (60%) had returned to school after treatment.

Conclusion

This study suggests that the adaptation of the program ‘Cool Kids - ASA’ into an adolescent version is feasible and has the potential to show good effects thus enhancing the possibility of education, development and independence in future life for this group. Larger RCTs studies are, however, needed to examine the efficacy of the adolescent version.

Keywords:

• Autism spectrum disorder
• anxiety disorder
• CBT
• adolescents

Background

The prevalence of autism has increased markedly over the last three decades [1–3] and is now thought to occur in up to 2% of children and young people worldwide [2]. The average age of diagnosis is around 5.3 years [4] but recent studies show a large increase in recognition of autism among adolescents [5]. This increase is attributed to autism being harder to identify in some individuals before the beginning of adolescence where demands on social communication skills, flexibility, and social interaction become more complex [5].

Approximately half of autistic adolescents experience significant psychosocial challenges in adulthood [6,7] affecting work prospects, friendships, and independence, and leaving them socially isolated and reliant on social, community, and family care [8]. Although outcome for autistic adults have improved over recent years, many remain highly dependent on others for support [9,10].

Co-occurring disorders in autistic individuals are common [11,12] and have a significant impact on the young person’s life [11,13]. Since many autistic youths are not recognised before adolescence, their concurrent difficulties might not have been dealt with earlier leaving them in need of disorder and age-appropriate treatment. Thus, treatment of co-occurring disorders may add considerably to improvement of the quality of life, general functioning, independence, and development in these young people [12].

Anxiety disorders [14–17] are estimated to occur in 11–84% of autistic individuals [18]. Due to a unique interaction between anxiety and core autism characteristics, the manifestation of anxiety in autistic children and adolescents differs in several ways from anxiety seen in neuro-typical youths. Anxiety in autistic individuals is associated with increased social avoidance, repetitive behaviours, and aggression. These maladaptive behaviours may be difficult to differentiate from symptoms of autism resulting in anxiety being underreported [18].

Cognitive Behavioural Therapy (CBT) has been shown to be very effective in the treatment of anxiety in neuro-typical children and adolescents [19–21]. CBT is typically conceptualized as a short-term, skills-focused treatment aimed at altering maladaptive emotional responses by changing the individual’s thoughts, behaviours, or both [22]. Research shows that CBT programs for autistic youths use the same mechanisms as in neurotypical individuals, and when the method is specifically adapted for autistic youths, are highly efficacious in treating anxiety disorder [18,23–26] reducing the anxiety symptoms that otherwise may lead to extra impairment in daily life skills. As seen in the meta-analysis by Ung et al. supporting CBT as an effective treatment at reducing anxiety in autistic youth, the majority of this evidence, is primarily from studies with younger children with participants in these studies being mainly school children and youth in their yearly to middle teenaged years (ages 7–17 with a mean age of 11.10) [27]. Some more recent studies have, nonetheless looked at the efficacy of such programs for adolescents [28,29]. Showing positive efficacy these studies have, however, concentrated on early adolescence with participants of 11–15 years (Mean age: 12.3) [29] and 11–16 years (Mean age: 12.79) [28]. Further, of the treatment studies that have targeted anxiety in autistic youth, the majority have focused mainly on adolescents diagnosed with high-functioning autism.

Irrespective of the presence of neurodiversity and the state or severity of autism, treatment of anxiety in adolescents is especially important since this is a vulnerable time of development. The adolescent years are when social relationships become most important, youngsters start dating and get sexually interested, become more independent, and ideas and motivations are forming for future careers and work [30]. Given the additional challenges faced by autistic individuals in these areas, reducing the impact of anxiety during the adolescent period is especially critical for long-term functioning, independence, and ability to become self-sufficient [31]. Adolescence is also the time when complex anxiety disorders such as social anxiety disorder most commonly begin [32] and in turn these disorders have proven the hardest to treat [33]. In order to support the best possible development and future possibilities for adolescents it is vitally important that interfering conditions such as anxiety can be treated.

The original CBT program ‘Cool Kids Anxiety Program’ [34] developed at the Centre for Emotional Health, Macquarie University, Australia was adapted specifically for autistic children and was shown to significantly decrease anxiety in children 8–13 years [23]. The adaptations included early introduction of relaxation techniques, simplified cognitive restructuring skills, discussion of how autistic characteristics can interact with anxiety, and a structured approach to planning the implementation of exposure steps. In addition, the written content within the workbooks was reduced and visuals used were possible. These visual included the use of two hand puppets to assist with externalise the anxiety issues and modelling skills. The program was also examined in a randomized controlled trial (RCT) to evaluate its efficacy in a non-English general hospital setting with satisfactory results [24]. While these results were positive, this program was aimed at children between 7 and 13 years of age and was not designed for older adolescents. Even though this program showed good efficacy in younger children it was unclear whether a similar program would work for adolescents due to the differences in age-related challenges and needs, in common anxiety diagnoses [32] and the method of conveyance.

Aim

The primary objective was to investigate the feasibility of a newly developed manualised Cognitive Behavioural Therapy (CBT) program for anxiety in older autistic adolescents (aged 14–17 years) when delivered in a group format. Our aim was to focus primarily on acceptability and compliance with the program with some investigation of treatment effects on anxiety diagnosis, anxiety symptoms and general quality of life.

Method

Study population

The study took place at Aarhus University Hospital, Department of Child and Adolescent Psychiatry (DCAP). Fifteen adolescents aged 14–17 years (Mean age: 15.8) diagnosed with autism according to the International Classification of Diseases, 10th edition (ICD 10) [35] by a Private Psychiatric Clinic or by one of the Regional Child and Adolescent Psychiatric Departments in Denmark (following the Gold Standard of diagnostics involving specific autism tests and ‘best practices’ As a minimum standard the diagnosis followed the regional guidelines entailing; an amnestic interview with the parents, a medical examination (height, weight, vision, hearing, gross and fine motor skills), observation in either school or home, an IQ test, and SRS. Only adolescents diagnosed by one of the Regional Child and Adolescent Psychiatric Departments additionally completed an ADOS assessment [36]), and who were experiencing life interference from anxiety symptoms participated in the study. All participants fulfilled the diagnostic criteria for one or more of any anxiety diagnoses on the Anxiety Disorders Interview Schedule for DSM-IV: Parent & Child interview schedule (ADIS C/P) [37].

Exclusion criteria for entering the study included intellectual disability (IQ < 70), active psychosis, untreated hyperkinetic disorder, and families considered not able to follow the CBT program e.g. due to inability to speak Danish or to meet for attendance due to family circumstances. Participants were recruited through information conveyed to different Regional Centres of Child and Adolescent Psychiatry, and the Danish and Regional Autism Association.

Diagnostic assessment of anxiety

Between 1–3 weeks before start of treatment the adolescents’ diagnostic status of anxiety was assessed using the Anxiety Disorders Interview Schedule for DSM-IV: Parent & Child interview schedule (ADIS C/P) [37], for both child and parents. The ADIS/CP assessment was performed by trained psychology graduate students under supervision of a senior assessor at the Centre for Psychological Treatment of Children and Adolescents, (CEBU) Department of Psychology and Behavioural Science, Aarhus University, which is separate from the DCAP, in order to secure blinding.

Intervention

Development

The original CBT program ‘Cool Kids Anxiety Program - Autism Spectrum Adaptation’ [38] was revised in line with the 2^nd edition of the standard (neuro-typical) ‘Cool Kids Anxiety Program’ and extended to incorporate workbooks for adolescents and options for both group and individual sessions (Cool Kids - ASA 2^nd edition [39]). For adolescents the modifications included replacement of the animal puppets with teen characters who have specific visual characteristics (such as one wears a cap and another has red hair), replacement of examples with common age relevant experiences (friendships, dating, exams) and age-consistent illustrations and layouts. The workbook for teens maintained the nickname ‘Chilled’ as used in the standard Cool Kids teen workbooks.

CBT intervention

The manualised CBT program Cool Kids – Autism Spectrum Adaptation 2^nd Edition was implemented in a group setting with the ‘Chilled - ASA’ teen workbooks utilized. The program was conducted in 10 sessions over approximately 13 weeks. Each group included four to six adolescent and parent dyads and sessions were conducted weekly, with a one week break after sessions 3, 6, and 8 to allow participants time to implement strategies in daily life. Each session lasted 2 h and included time spent working with the adolescents alone, time with the parents alone, and time with parents and adolescents together. The adolescent and at least one parent attended the groups. The main group therapist and the co-therapist were trained psychologists, with at least one accredited in the Cool Kids programs. A graduate psychologist was also involved as a practical helper.

Measures

To assess feasibility both adolescents and parents were post-treatment asked to complete a short questionnaire adapted from The Experience of Service Questionnaire (ESQ) [40,41] measuring their satisfaction with the treatment. On a Likert scale they were asked to rate positive and negative statements as ‘not true’ (1), ‘partly true’ (2), or ‘true’ (3) and in an open section freely comment on their experience with the treatment.

Anxiety Disorders Interview Schedule for DSM-IV: Parent & Child interview schedule (ADIS/CP) [37], was used to measure anxiety disorder diagnoses. ADIS/CP is a semi-structured interview with separate interviews for both child/adolescent (age range 6–17 years) and parents designed to assess current episodes of anxiety disorders, and to permit differential diagnoses among the anxiety disorders according to DSM-IV criteria. Clinical severity ratings (CSR) from 0 (no interference) to 8 (extreme interference) are applied to each disorder. Severity ratings of 4 or above indicate the presence of a disorder. Adolescents, parents, and the clinician give separate CSRs. In the current study only the CSRs provided by the clinician are reported with the most impairing diagnosis considered the primary diagnosis. In other studies, the ADIS/CP has demonstrated good-to-excellent 7–14 days test–retest reliability for the presence of specific anxiety diagnoses (Cohen’s Kappa [κ] range for different diagnoses = 0.71–0.84 for adolescent interviews and 0.73–0.92 for parent interviews) [42]. High concurrent validity against the Multidimensional Anxiety Scale for Children (MASC) [43] has also been demonstrated [44].

Additional assessments administered to adolescent and parents of anxiety symptoms, quality of life, socio economic status (SES - obtained by questions concerning parent’s level of education, employment, household income, etc.), were collected by using an on-line data collection platform (RedCap). The adolescent’s IQ score received at the time of ASD diagnosis by using either Wechsler Intelligence Scale for children-V (WISC) [45] or Reynolds Intellectual Assessment Scales (RIAS) [46] was obtained from the original diagnostic report.

Spence Children’s Anxiety Scale (SCAS) [47] is a questionnaire for children and adolescents (age range 8–15 years) and parents, assessing the severity of anxiety symptoms broadly in line with the dimensions of anxiety disorder proposed by the DSM-IV. SCAS consists of 44 items rated from 0 (never) to 3 (always). The six subscales of different anxiety disorders may be scored separately as well as added together for a score of overall anxiety symptoms. The Danish version of SCAS has shown excellent internal consistency for the total scale (α = 0.89) in a sample of children and adolescents with anxiety disorders and good test–retest reliability after 2 weeks (r = 0.84) and 3 months (r = 0.83) [48]. This scale has further shown excellent internal consistency and fair-good parent–child agreement in an autistic population [49].

Child Anxiety Life Inference Scale (CALIS) [50] is designed to assess life interference attributed to fears and worries from child (age range 6–17 years) and parent perspectives. The measure targets interference with the child’s life and with the parent’s/family’s life. The items (9 for adolescents and 16 for parents) are rated from 0 (not at all) to 4 (a great deal) and scores can be reported combined or divided into subscales. Subscales consist of the child’s and the parents’ view of life interference at home or outside the home together with nine items for parents rating interference with their own life (e.g. stress and relationship with spouse, friends, and family). The scale has demonstrated satisfactory internal consistency on subscales for both children and adolescents (α range = 0.70–0.84) and parent ratings (α range = 0.75–0.90) and moderate stability for a 2-month retest period (r range = 0.62–0.91) [50].

Children’s Automatic Thoughts Scale (CATS) [51] is a ­developmentally sensitive, general measure of negative self-statements across both internalising and externalising problems (age range 8–17 years). The 40 items are rated from 0 (not at all) to 4 (all the time) and can be combined into a single score. Four separate subscales of cognitive content can, in addition, be assessed including physical threat, social threat, personal failure, and hostility. Internal consistency of the subscales was high, (α range 0.85-0.92) and test-retest demonstrated adequate reliability at both 1 month (r = 0.79) and 3 months (r = 0.76) [51].

Parents’ quality of life was assessed with the WHO-five Well-being Index [52] and the adolescents quality of life with the Children Quality of Life Questionnaire – Revised (KINDL-R). The KINDL-R is a generic instrument comprising 24 items for assessing Health-Related Quality of Life in children and adolescents [53,54]. Psychometric testing of the KINDL-R questionnaire reveals good scale utilisation and scale fit as well as moderate internal consistency (α range = 0.54–0.73 for subscales and an α = 0.82 for the total score) [55].

Finally, data on proportion of attended sessions, completion of homework and change in status of school attendance was collected.

Diagnostic interviews and all questionnaires were completed before the intervention (Pre), 1–2 weeks after the intervention (Post) and at 3 months follow-up (Fu).

Analysis

The satisfaction questionnaire consisted of 7 statements for the adolescents and 10 for the parents and was analysed by calculating the percentage of ‘not true’, ‘partly true’, or ‘true’ answers for each of the statements. All qualitative feedback from open ended questions was reported in full.

The percentage of adolescents free of all anxiety diagnoses or no longer fulfilling the criteria for their primary diagnosis on the ADIS/CP after the treatment and at FU were calculated. Not all children recover completely from what is an often prolonged experience of anxiety in the relatively short time from pre- to post-treatment or even at the 3-month follow-up. Thus, information regarding smaller improvements can also be investigated by analysing the potential reduction in clinical severity rating (CSR) [24,25,28,29,37,56]. Outcomes regarding changes in diagnostic status, change in CSR scores for main diagnosis and in mean CSR scores across all diagnoses, severity of anxiety symptoms, life interference attributed to fears and worries and youth’s negative self-statements measured with ADIS/CP, SCAS, CALIS and CATS were investigated. Further, the quality of life (QoL) for both adolescents and parents were investigated together with possible change in the adolescents’ school attendance.

Mean scores (M) with standard deviations (SD) were calculated Pre- and Post-treatment and at FU for the above measurement followed by the use of Cohen’s d for repeated measures to calculate within-group effect sizes at the different data points – pre to post-treatment and post-treatment to FU. The effect size was interpreted according to Cohen’s standard interpretation [57] with.8 = large (8/10 of a standard deviation unit), .5 = moderate (1/2 of a standard deviation) and.2 = small (1/5 of a standard deviation).

Statistical analyses were conducted by the use of STATA [58]. Only unadjusted analysis was performed and no significance testing was employed due to the small number of participants.

Results

A total of 15 adolescents between 14 and 17 years of age (mean age = 15.8) were enrolled in the study with a preponderance of females (n = 9) compared to males (n = 6). Mean IQ was 101.8 (SD = 14.97) and two thirds of the adolescents were living with both biological parents. Parents’ years of education (mean = 14.9, SD = 1.93) and household income in DKK (mean = 803.333, SD = 463.090) further fitted the average level in Denmark (Table 1). The families were encouraged not to seek other treatment during the study which all of the participants obliged.

Table 1. Demographics and diagnostics characteristics of the participants.

N = 15
Age at time of treatment (Mean, SD)	15.8 (1.07)
Adolescents’ total IQ (Mean, SD)	101.8 (14.97)
Parents average education in years (Mean, SD)	14.9 (1.93)
Household income in DKK (Mean, SD)	803.333 (463.090)
Gender	(n=)
Female	9
Male	6
ASD diagnosis (ICD-10)
F84.0	5
F84.1	3
F84.5	1
F84.8	4
F84.12	2
On psychopharmacological medication
Yes*	5
No	10
Living with both biological parents
Yes	10
No	5
Primary anxiety diagnosis (N = 11)
SoP	9
GAD	1
SP	1
Number of anxiety disorders (N = 11)
1	4
2	3
3	3
4	1

Note: F84.0 – Infantile Autism; F 84.1 – Atypical Autism; F84.5 – Asperger’s Syndrome; F84.8 – Childhood Disintegrative Disorder; F98.8.

* Sertraline (4) and Ritalin Uno (1).

GAD: Generalized Anxiety Disorder; SoP: Social Phobia; SP: Specific Phobia.

Feasibility of the chilled - ASA

Of the 15 adolescents who were offered the Chilled - ASA, one dropped out before start of the treatment due to change of address and one dropped out after attending 2 sessions due to the youth’s reluctance to participate in a group setting. No differences in baseline characteristics between the two non-completers and the completers were evident. Thirteen attended 7 or more sessions (mean = 8.5; SD = 2.10) - the minimum number of sessions required to have completed the program. All participants completed homework between sessions with 78% of the adolescents completing homework for 7 or more of the 9 sessions (no homework was required before session 1).

Following treatment, the ESQ was completed by all adolescents and parents and showed overall general satisfaction with the program. For each item on the ESQ, the majority of adolescents (76%) and the majority of parents (92%) gave ratings indicating positive or semi-positive associations with the program. The responses from adolescents and parents to each of the statements can be seen in Table 2.

Table 2. Results from the experience of service questionnaire in order of most to least endorsed statements.

Adolescents N = 13	True	Partly true	Not true
The treatment made me feel worse.	1	1	11
The treatment helped me.	7	5	1
If a friend needed this kind of help I would recommend the treatment.	7	3	3
I trusted the therapists.	6	5	2
The therapist understood the most important parts of my problems.	6	5	2
The wellbeing of the family has improved after treatment.	3	4	6
I feel more motivated to be with friend after the treatment.	3	4	6
Parents N = 13	True	Partly true	Not true
I trusted the therapists.	12	1	0
I feel I have been rightly informed about meaning, purpose, and process of the treatment.	12	1	0
The treatment made me feel worse.	0	1	12
The treatment made my child worse.	0	2	11
If a friend needed this kind of help I would recommend the treatment.	11	1	1
During the treatment I acquired a better understanding of my child’s psychological state.	10	2	1
During the treatment I became able to change my behaviour towards my child in a positive way.	9	3	1
The treatment helped my child.	8	5	0
The treatment helped me.	8	4	1
The wellbeing of the family has improved after treatment.	7	5	1

Adolescents and parents were in addition provided the opportunity to comment on satisfaction with the treatment. Seven adolescents and twelve parents took this opportunity to do so with some of the comments overlapping (see Table 3).

Table 3. Comments from parents and adolescents regarding satisfaction of the program.

Adolescents Positive statements

The treatment helped me.

I am feeling better. I can now do things I couldn’t before and I am less anxious now.

It was good to meet others in the same situation. It has helped me and given me the courage to try new things.

10/10

It was good and has helped me quite a lot.

Negative statements

Felt physically ill when there was conversation around the table for us young people.

I didn’t like the book.

Parents Positive statements

Help to self-help. Tools that can be used going forward in all situations that create anxiety. Our approach has changed and our child’s approach to trying out ‘new country’ has changed.

Useful tools. better knowledge of anxiety in relation to Autism. Good to meet other families.

It has contributed to a greater awareness of one’s own behaviour and approach as parents when something is difficult and anxiety-provoking for one’s child.

We have gained a common language which helps us understand each other.

Common language, systematic plan of attack, ladders, exposure, meeting others in a similar situation.

The process, the therapists and that we were a group.

Liked the focus on methods and focus on the parental role.

We got a common language.

Insight into our son’s perception of reality. The book seems easily accessible.

Good to be in a group with others in the same situation.

Negative statements

Consider some kind of reward system other than stickers, most people think it’s silly. Perhaps you should set a check mark - so that the young people themselves can follow with their active efforts.

Focus on process rather than pushing the goal.

Adapted more to young people. Table of contents in the material.

The manual is way too compressed.

The book could be aimed more at the youth group.

For me, there was a lack of clarity about exactly which pages/homework we had to do from time to time.

The layout of the written material is too messy.

Enhancement of the program

Feedback from adolescents and parents indicated that the material was well set-up and easy to understand. However, some found the adolescent’s workbook to be too childish and would like a more mature set-up and language. Further, it was recommended that the program lasted longer than 13 weeks to allow more time for exercises between sessions.

Anxiety diagnoses

Of the 13 participants ADIS/CP-CSR scores were obtained from 11 families. Mean scores, standard deviations, and within-group effect sizes (Cohen’s d) for pre-, post-treatment and follow-up (FU) are presented in Table 4. No other differences in the baseline criteria than geography was found between the ones who participated in the ADIS/CP and the ones who did not with the non-participants living in the Copenhagen area.

Table 4. Mean scores, standard deviations, and effect sizes for all continuous measures.

	Pre M (SD)	Post M (SD)	FU M (SD)	Effect sizes Pre to Post	Effect sizes Post to FU
ADIS-CRS Primary diagnosis	6.73 (1.10) n = 11	4.73 (1.95) n = 11	3.91 (1.81) n = 11	d = 1.26	d = .43
ADIS-CRS All diagnosis	6.19 (1.11) n = 11	4.46 (1.81) n = 11	3.91 (1.42) n = 11	d = 1.15	d = .34
ADIS No. of anxiety diagnosis	2.09 (SD = 1.04) n = 11	1.55 (SD = 1.21) n = 11	1.18 (SD = 1.08) n = 11	d = .48	d = .32
SCAS-C Adolescents	40.69 (16.62) n = 13	31.85 (16.92) n = 13	28.77 (15.18) n = 13	d = .53	d = .19
SCAS-P Parents	42.77 (15.18) n = 13	30.38 (13.33) n = 13	28.38 (12.03) n = 13	d = .87	d = .16
CALIS-C Adolescents	18.62 (5.77) n = 13	12.23 (7.91) n = 13	11.77 (7.13) n = 13	d = .92	d = .06
CALIS-P Parent	24.23 (6.07) n = 13	21.46 (5.19) n = 13	16.62 (8.52) n = 13	d = .49	d = .69
CATS	44.23 (30.71) n = 13	26.92 (29.99) n = 13	26.77 (26.96) n = 13	d = .57	d = .005
KINDL-R* Adolescents	39.08 (8.64) n = 13	44.31 (9.30) n = 13	43.23 (9.97) n = 13	d = −0.58	d = .11
WHO-5* Parent	48 (18.48) n = 13	53.54 (19.22) n = 13	56.31 (21.26) n = 13	d = −0.29	d = −0.14
Attending school*	62% n = 13	77% n = 13	85% n = 13	d = −0.32	d = −0.18

Note: ADIS-CRS: Anxiety Disorder Interview Schedule for DSM-IV, Clinical Severity Rating (Clinicians rating); SCAS-C/P: Spence Children’s Anxiety Scale, Child/Parent version; CALIS-C/P: Child Anxiety Life Interference Scale, Child/Parent version; CATS: Children’s Automatic Thoughts Scale; KINDL-R: Quality of Life, Adolescent version; WHO-5: Quality of Life, Adult version.

* Negative effect sizes indicate improvement.

Results from baseline (pre-) to post-treatment showed an improvement on the severity of the primary anxiety diagnosis with a large effect size (d = 1.26). Continued reduction in CSR from post-treatment to FU showed a medium effect size (d = .43). For all anxiety diagnoses, a large effect size in the overall mean CSR scores was evident from pre- to post-treatment (d = 1.15) whereas the continued effect size post-treatment to FU (d = .34) was small. (Individual CSR scores for each of the participating adolescents taking part in the ADIS/CP interview can be seen in Table S1).

Following treatment 45.5% of the adolescents had recovered from their primary anxiety diagnosis and at FU this number rose to 54.5%. The percentage of adolescents recovered and no longer meeting the criteria for any anxiety diagnosis on the ADIS/CP after treatment was 18.2%. This number rose to 34.4% at FU. The number of anxiety diagnosis was reduced from pre to post with a medium effect (d=.48) and continued the reduction at FU with a smaller effect (d=.32).

Anxiety symptoms, impact on life and adolescents’ measure of negative self-statements

As seen in Table 4 the results from the self-reported questionnaire on anxiety symptoms (SCAS) (mean scores, standard deviations, and effect sizes for all continuous measures) showed an improvement from pre- to post-treatment in both the adolescents’ responses (d = .54) and the parents’ responses (d = .87). There was further minor improvement from post-treatment to FU according to both the adolescents (d = .19) and the parent questionnaire (d = .16).

For the adolescents, results for the anxiety’s impact on life (CALIS) showed a large positive effect (d=.92) from pre to post-treatment whereas the parents experienced a medium positive effect (d=.49). A further very small positive effect was seen from post-treatment to FU in the adolescents (d = .06) whereas the parents experienced a larger medium continued effect (d = .69).

The adolescents’ negative self-statements measured with CATS showed a medium positive effect from pre- to post-treatment (d = .57) but no effect from post-treatment to FU (d=.005).

Both measures on adolescents’ and parents’ quality of life showed improvement with the adolescents showing a medium positive effect from pre to post-treatment (d=–0.58) on the KINDL-R and a further small effect at FU (d=–0.11). The parents showed a small effect on the WHO-5 at post-treatment ­(d=–0.29) and another small improvement at FU (d=–0.14).

Five of the 13 adolescents did not attend school at the beginning of the program due to anxiety. At post-treatment this number was down to three and at FU only 2 were not attending a public school.

Discussion

Recent findings provide evidence for the benefit of manualised CBT programs for autistic children and co-occurring anxiety disorders with an average medium effect size (g = 0.66) [59]. In order to investigate whether the CBT program Cool Kids Anxiety Program: Autism Spectrum Disorder Adaption [38] could be adapted for adolescents, following the positive findings regarding the efficacy of the program in children [23,24,56], this study aimed to determine the feasibility of an adolescent version, Chilled - ASA, for young people.

Feasibility

Two adolescents dropped out - one before treatment and one after attending two sessions. The 13 adolescents participating in the Chilled - ASA completed a mean of 8.5 out of 10 sessions. One child managed only to participate in 2 out of 10 sessions due to difficulties in being in a group setting. In contrast, many of the adolescents expressed the benefits of being in a group to increase awareness of other peers in the same situation and thus not feel alone with their anxiety problems. Since social skills deficits have been shown to be a significant predictor of social anxiety [60] and autistic adolescents often encounter problems in this area, being in a group setting can be a particular challenge for some of these adolescents. In addition, sensory sensitivities may make group settings uncomfortable for some autistics. Thus, in order to achieve the best compliance, it is important to assess whether the adolescent will be able to attend group sessions or whether individual treatment may be more acceptable.

The majority of adolescents and parents were generally satisfied with the treatment, expressing that it had helped the youngsters and that they would recommend the treatment to others. However, this questionnaire was originally developed for neuro-typical adolescents and some of the statements might thus have been too abstract for autistic adolescents, e.g. ‘The wellbeing of the family has improved after treatment?’. Nevertheless, the overall satisfaction with the program seemed evident.

Program enhancements

This study helped to focus on changes and improvements that might be considered before initiating a larger randomised study. As seen in Table 3, feedback after sessions from participants indicated that the adolescent’s workbook was found to be a little childish especially for the older adolescents. However, since the program is aimed at adolescents from 13 years of age it is important that the workbook is comprehensible to all ages and leaving it up to the therapist to use it in an age-appropriate way.

The duration of the program can, in addition, be up to the therapist’s discretion adding extra weeks for training between sessions. This is a more practical issue and will depend on the available resources and flexibility of the treating clinic. However, our clinical experience from clinics in Denmark and Australia is that the families do not necessarily do more homework between sessions if the breaks between sessions are longer.

Limited efficacy results

Following treatment with the Chilled - ASA, almost half of the adolescents were free of their primary anxiety diagnosis and a third were free of all anxiety diagnoses at FU. The severity of the anxiety across all anxiety diagnoses was also greatly reduced, especially from pre- to post-treatment with an additional decrease to FU. In addition, the number of anxiety diagnoses was reduced where a fall occurred from pre- to post-treatment. However, even though this indicates that the program is promising, due to a very low number of participants in this study and the uncontrolled design, the efficacy of the program cannot be concluded.

A positive effect was revealed from pre- to post-treatment for anxiety symptoms (SCAS), impact on life (CALIS), and the adolescents’ frequency of cognitive distress (CATS). Over the longer term, these results maintained or even increased. An increase in quality of life was found from pre- to post-treatment according to both the adolescents and their parents and a further small improvement at FU. This, together with the fact that more adolescents attended school following treatment, indicates that the treatment of the anxiety aided the adolescents in becoming more independent and able to pursue an education.

Thus, the results and the satisfaction with the program point in the direction of the Chilled - ASA being a feasible and potential effective treatment for reducing anxiety among autistic adolescents.

Strengths of this study

One of the strengths in this feasibility study was that the adolescents all had undergone a thorough standardised clinical assessment of their autism diagnosis in accordance with the International Classification of Diseases, 10th edition [35] The diagnostic method followed the Gold standard and was either performed by an interdisciplinary team consisting of experienced psychologists, psychiatric nurses, educationalists, and child psychiatrists from the Children and Adolescents Psychiatric Hospital in Aarhus, Denmark or by experienced private child and adolescent psychiatrists. In addition, all outcome measures were collected by the use of validated clinical instruments even though these instruments are only validated with typically developing adolescents.

Limitations of this study

A limitation of this study was the small number of participants and lack of control group which limits the comparison with previous studies from university clinics. As mentioned above, some of the instruments used in the diagnostic assessment together with the outcome measures were not adapted for autistic populations and thus do not account for potential symptom differences that arise in the context of neuro-diversity. However, at the time of the study no instruments investigating anxiety, modified for autistic adolescents and translated into Danish existed. Further, the ESQ was not anonymous and responses might thus have been biased.

Conclusion

Results from this study preliminarily indicate that the manualised CBT program, Chilled - ASA may provide a feasible group treatment for anxiety in Danish autistic adolescents and may work as a treatment in both public psychiatric health clinics and private clinics. Not only does there seem to be clinical improvement in the adolescents’ anxiety level measured with both parents’ and adolescents’ questionnaires, but in addition high attendance and a general satisfaction with the experience of the program. Being able to provide easy access to effective treatment for anxiety for more autistic adolescents is of great importance. This study indicates that the Chilled - ASA might be recommended as one such intervention, although a fully randomised trial would be important.

Further, a manualized CBT program like the Chilled - ASA with detailed therapist instructions and workbooks aimed at adolescents might facilitate the transportability of feasible and efficacious treatments to care settings where therapists may have less opportunity for extensive CBT training and expertise development. This, in turn, might improve implementation and provide age relevant treatment to a large sample of autistic adolescents enhancing their future development, education, and quality of life.

The socio-economic cost of autism is high [10] since many autistic individuals rely on public and family support. Decreasing difficulties associated with autism, such as co-occurring disorders and thereby increasing independence could support the possibility of reducing these costs on a long-term basis.

Larger RCTs are, however, needed to examine the efficacy of Chilled - ASA and the possibility of implementation in daily clinical practice in non-research settings together with examining the possible impacts that reducing anxiety may have on other challenges that autistic adolescents experience.

Ethical statement

The study involved human participants and is registered and approved by Clinicaltrials.gov, ID.nr: NCT04838912. The Regional Ethical Committee: ID.nr:1–10-72-24-21. All parents gave informed consent to participate in the study and to the use of their data.

Acknowledgments

We wish to thank the TRYG Foundation for funding this study and all the families for participating.

Disclosure statement

The authors report no conflicts of interest. Ron M Rapee and Tina R. Kilburn are co-authors of the treatment program - no royalties and Non-remunerative. The authors alone are responsible for the content and writing of the paper.

Data availability statement

The datasets generated and analysed during the current study are not publicly available due to the General Data Protection Regulations (GPDR) but are available in anonymised form from the corresponding author on reasonable request.

Additional information

Funding

Funding for the project was received from the TRYG Foundation (Grant 151763).

Notes on contributors

Tina R. Kilburn

Tina R. Kilburn is clinical psychologist and senior researcher at the research department at Department of Child and Adolescent Psychiatry at Aarhus University Hospital, Denmark. Her research has been concentrated on anxiety in autistic children and adolescents and how to develop effective and accessible treatment for this group.

Ronald M. Rapee

Ronald M. Rapee, PhD, is professor of psychology at Macquarie University, in Sydney, Australia, and director of the University’s Centre for Emotional Health. He is known for his theoretical models of the development of anxiety disorders and his creation of empirically validated intervention programs that are widely used internationally.

Heidi J. Lyneham

Heidi J. Lyneham is a clinical psychologist and researcher who focuses on assessment and treatment of emotional difficulties in children and families, particularly in improving access to services. She is clinic director at the Centre for Emotional Health at Macquarie University, Sydney, Australia.

Mikael Thastum

Mikael Thastum is professor of clinical child psychology and health psychology at the Aarhus University, Denmark. In his research, teaching and treatment, Mikael Thastum works on how to best understand and help children and young people with different kinds of challenges such as anxiety and chronic diseases.

Per Hove Thomsen

Per Hove Thomsen is doctor, MD, professor of child and adolescent psychiatry at Aarhus University, Denmark, and head of the research department at Department of Child and Adolescent Psychiatry at Aarhus University Hospital, Denmark with research areas including ADHD, eating disorders, autism, tics and Tourette’s syndrome