This study was designed as a prospective, randomized, single-blinded, open-label endpoint trial with two parallel intervention groups, the TAU group and the COMB group, consisting of a six-week treatment for both groups. The study was conducted as a multisite study. Qualified individuals who consented to this trial were randomly assigned to either the COMB group or the TAU group. We evaluated outcomes at the pre-intervention (0 week), post-intervention (6 weeks), and follow-up (10 weeks). The Chiba University Research Ethical Review Committee approved this trial (CRB3180015, No. G29027). All methods were carried out in accordance with relevant guidelines and regulations in the Declaration of Helsinki.
Randomization
Enrollment and randomization were based on the central registration system of the Clinical Research Data Centre at Chiba University Hospital, Chiba, Japan. After the baseline assessment, participants were randomly assigned to the COMB or TAU in a 1:1 ratio. Assignments were made using the minimization method and were balanced for baseline AKQ-C score (AKQ-C < 12) and gender. At the end of the baseline assessment, eligible participants (dyads of autistic adolescents and their parents/guardians) were randomly assigned to either the COMB arm or TAU arm at a ratio of 1:1. Therapists and participants were not blinded to their group assignment; however, two trained testers, MH and MS, who assessed all outcome measures except the self-assessment scale, were blinded to the group.
Participants
The adolescents and parents/guardians provided written and verbal informed consent using the form and protocol approved by the Clinical Research Ethical Committee of the affiliated university. The inclusion criteria were: (1) adolescents with ASD who: a) were diagnosed with ASD by a primary doctor using the DSM-5 and met the cut-off value for ASD measured using the Autism Diagnostic Observation Schedule (ADOS-2) [17] and Autism Diagnostic Interview-Revised (ADI-R) [18,19]; b) had a verbal IQ (VCI) of 90 or above as measured by the Wechsler Adult Intelligence Scale-III (WAIS-III) or Wechsler Intelligence Scale for Children, Fourth Edition (WISC-IV)([20]); c) had a social difficulty assessment of moderate and above, as measured using the Strengths and Difficulties Questionnaire (SDQ; Goodman [21]); d) were aged between 10 and 17 years, e) were outpatient at a psychiatry in a hospital. Since this study was recruited in a hospital setting, it included patients only attending the hospital's psychiatry department; (2) parents/guardians of ASD individuals who are living with the ASD individual and act as a guardian. We set inclusion criterion b) since the recommended IQ (FIQ or VCI) for providing CBT for children with ASD is typically approximately 70–100 [15]. In this study, it was set at 90, taking the middle ground. The exclusion criteria were: (1) autistic adolescents who had suicidal intentions at the time of the study enrollment, a repetitive antisocial behavior, and/or severe degenerative physical disorder, as identified by a primary psychiatrist of an autistic adolescent, or any other condition that would make participation not be in the best interest (e.g., compromise the well-being) and (2) parents/guardians of autistic individuals who met the diagnostic criteria for psychiatric disorders, assessed using the Mini-International Neuropsychiatric Interview [22]. We excluded parents/guardians with psychiatric disorders since they were asked to understand autistic adolescents to provide adequate support for reasonable accommodation and to be mentally healthy and stable enough to implement those assistances.
Sample size
According to a previous study by Gordon et al. [13], the change in autistic awareness on the AKQ was 2.26 with an effect size of .92. Based on the above report, we used G.Power 3.1.9.2 for detection, and under a significance level of 5% on both sides and a power of 80%, 20 cases per group were required. Considering dropout cases and 20% ineligible cases, the required number of cases was set at 24 per group, for a total of 48 cases.
The ACAT program
ACAT is an individual CBT session conducted once a week for 100 minutes each time ([16]). The ACAT consists of six sessions, starting with feedback on psychological assessments related to ASD at baseline and ending with a follow-up session one month after the completion of six sessions. There was a follow-up session one month after the completion of the sixth session.
The ACAT program consisted of three main phases. The details are given in Table 2. The first step involved helping ASD individuals externalize and attain metacognitive awareness of their ASD traits. This way, they practiced “noticing” their autistic traits in daily life. The second step involved using the CBT model (see Figure S1) to understand their autistic responses to environmental or social stressors by drawing diagrams. Through this process, autistic individuals and their parents/guardians were able to understand how their autistic traits affected their daily lives. In the third stage, autistic individuals wrote down coping strategies that they thought they could use and tried to apply said strategies in their daily lives. Furthermore, an adolescent, their parent/guardian, and a therapist discussed what kind of reasonable accommodations could be sought to improve the adolescent’s adjustment and planned how they could utilize and implement them in daily life (see Figure S2). The therapists were qualified clinical psychologists with clinical experience working with adolescents; the two interviewers of the primary outcomes were graduate students. The participants in the TAU group received no intervention from the research team. Throughout the study, all participants in the COMB and TAU groups received medical care from psychiatric hospitals. The male to female ratio of the participants was 33:14, with more men than women. Average age when starting the program was 13.5 years, and the average age of diagnosis of ASD was 10.3 years. Additionally, the average age of the parents/guardians was 45.7 years, with the mothers to fathers ratio of 1:44, with more mothers participating than fathers. Their average years of schooling was 14.3 years. Detailed information regarding demographics is provided in Tables 3 and 4.. The COMB and TAU groups did not differ at baseline (p = .11) in terms of the number of additional services. For parents/guardians, there was no significant difference in the use of support such as parent training (p = .47).
<Table 2 here>
<Table 3 here>
<Table 4 here>
Assessment and outcomes
All the questionnaires used in this study, except for the AKQ-C, were standardized in Japanese and were published.
Primary outcome
Adolescents
Autism Knowledge Quiz-Child (AKQ-C)
The English version of AKQ-C [13] was translated into Japanese by a bilingual postgraduate student. The Japanese version was further back-translated by English native speakers who are also fluent in Japanese, for verification. The author of English version of AKQ-C confirmed that the contents of the back-translated version of the AKQ-C and the original version are consistent in content. The AKQ-C has two sections: a five-item structured interview on “awareness of autistic characteristics” and a 15-item general ASD knowledge quiz. The first section is designed to measure ASD self-knowledge, and it consists of open-ended questions for adolescents about their strengths and difficulties. The second section is designed to measure respondents’ general knowledge about ASD and includes questions on the nature, prevalence, causes of ASD, and the strengths and difficulties commonly observed in autistic people. We only used section 1 of the AKQ, index of awareness of autistic characteristics, as a primary outcome since the ACAT was developed to increase the understanding of one’s own autistic traits rather than improve their general knowledge of ASD. The details of the scoring of the AKQ-C section 1 are explained in the supplemental material. Briefly, an assessor asked an autistic adolescent about their strengths and weaknesses, and the assessor scored each descriptive response from the adolescent in a binary manner (1 as related to autistic traits, and 0 as not related to autistic traits). The assessor was blinded to the group assignment, and scoring was supervised by W.M.
The AKQ-C was administered to autistic adolescents at pre-intervention (week 0: prior to starting the first session), post-intervention (week 6: when they completed six sessions), and follow-up (week 10: one month after the completion of sessions). The change in AKQ-C scores from pre- to post-intervention was considered a primary outcome.
Secondary outcomes
Adolescents
Barriers to Access to Care Evaluation scale version 3 (BACE-3)
The BACE‐3 [23] is a 30-item questionnaire that assesses barriers to mental health care for people with mental health problems. Each item is scored from 0 (not at all) to 3 (a lot). The BACE-3 comprises two subscales: “barriers to treatment-related stigma” (12 items) and “barriers to treatment-unrelated stigma” (18 items). We only used the “barriers to treatment-related stigma” subscale to measure treatment stigma. A higher score indicates a greater barrier to treatment-seeking behaviors. Cronbach’s alpha for the barriers to treatment-related stigma subscale is α = 0.90, indicating good internal consistency. Cronbach’s alpha for the barriers to treatment-unrelated stigma subscale is α = 0.83, indicating good internal consistency [6]. Adolescents responded to the BACE-3 at three time points: pre-intervention, post-intervention, and follow-up.
The Depression Self-Rating Scale for Children (DSRS-C)
The DSRS-C [24] is a self-assessment scale with 18 items. It assesses the depression level among children by enquiring about their feelings over a week. Each item is scored on a three-point scale: “always,” “sometimes,” and “never.” A higher score indicates increased severity of depression. Adolescents with autistic answered DSRS-C at three time points: pre-intervention, post-intervention, and follow-up.
The Strengths and Difficulties Questionnaire (SDQ)
The SDQ [20,25] is a parent-rated short screening instrument that addresses the positive and the negative behavioral attributes of adolescents. It includes 25 items, and each item can be marked as “not true,” “somewhat true,” or “certainly true.” A higher score indicates more severe multifaceted behavioral problems. The SDQ was completed by parents/guardians at three time points: pre-intervention, post-intervention, and follow-up.
The Vineland Adaptive Behavior Scales Second Edition (Vineland-Ⅱ)
The Vineland-II [26] is a semi-structured interview for parents/guardians to capture the developmental norm of adaptive behavior in individuals aged 0–92 years old. We interviewed parents/guardians about their adolescents at three time points: pre-intervention, post-intervention, and follow-up.
Parents/Guardians
Autism Knowledge Quiz-Parents (AKQ-P)
The AKQ-P (for parents/guardians) [13] is the parent version of the AKQ-C. Same as the AKQ-C [13], the English version of the AKQ-P was translated into Japanese and validated. We also only used the first section “awareness of autistic characteristics” to measure parents/guardians’ awareness of adolescents’ strengths and difficulties related to their autistic traits. The scoring for the AKQ-P is the same as that for the AKQ-C. Parents/guardians responded to the AKQ-P at three time points: pre-intervention, post-intervention, and follow-up.
Barriers to Access to Care Evaluation scale version 3 (BACE-3)
Similar to the assessment of adolescents, the BACE‐3 was administered to their parents/guardians at three time points: pre-intervention, post-intervention, and follow-up.
General Health Questionnaire (GHQ-12)
The GHQ-12 has 12 items [27], and it measures current mental health and psychological distress. Each item is rated on a four-point scale (less than usual, no more than usual, more than usual, or much more than usual). A higher score indicates greater psychological distress. Parents/guardians responded to the GHQ-12 at three time points: pre-intervention, post-intervention, and follow-up.
Parenting Resilience Elements Questionnaire (PREQ)
The PREQ (for parents/guardians) [28] measures the degree to which parents/guardians possess elements that aid in adapting to challenges and difficulties related to children with developmental disorders. It comprises 29 items across three factors: “knowledge of the child’s characteristics,” “perceived social support,” and “positive perceptions of parenting.” Each item is scored on a seven-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). Higher scores on each subscale indicate a greater capacity or attitude to cope with challenges and difficulties related to their children. Parents/guardians responded to the PREQ at three time points: pre-intervention, post-intervention, and follow-up.
Statistical analysis
All the statistical analyses were performed as we pre-registered in our previous paper ([16]). The primary outcome was the AKQ-C score, and we examined differences in score variations from baseline (week 0) to post-intervention (week 6) between the COMB and TAU groups, with allocation factors (i.e., AKQ-C scores and sex) and age as covariates (analysis of covariance: ANCOVA). The differences in score variations from baseline (week 0) to follow-up (week 10) between the two groups were the secondary outcome. Other secondary outcomes included psychometric variables of adolescents and parents/guardians (described above), and we evaluated longitudinal data on differences in scores from baseline (week 0) to post-intervention (week 6) and baseline (week 0) to follow-up (week 10) between the two groups. Student’s t-test was used to compare the differences between the two groups. Paired t-tests were used to compare the differences between baseline to post-intervention and baseline to follow-up (between week 0 to week 6 and between week 0 to week 10). All analyses were performed using the software SAS version 9.4.