Participants
We recruited participants from the outpatient clinics of Showa University and the National Center for Child Health and Development, Tokyo, Japan. These institutes are located in the western part of Tokyo. We contacted all clients with a clinical diagnosis of ASD based on the DSM-5 diagnostic criteria who visited one of these institutes from November 2020 to March 2021. As a result, we collected data from 102 children with ASD. During the data collection period, the pandemic was overwhelming, and Tokyo was under a state of emergency. Participants received a maximum of ¥1,000 gratuity for completing questionnaires.
Among the 102 participants, 75 participants (73.5%) were male (Table 1). The mean ages of the participants and caregivers were 11.6 (standard deviation [SD] = 5.3) and 45.7 (SD = 6.2) years, respectively. Most of the caregivers were mothers (n = 92, 90.2%) or fathers (n = 7, 6.9%) of children with ASD, while three did not indicate their relationship with the participants. The psychiatric and neurological comorbidities included attention-deficit/hyperactivity disorder (n = 28), learning disorder (n = 5), epilepsy (n = 8), and intellectual disabilities (n = 22). On the basis of the low per capita patients who contracted COVID-19 in Japan, only two participants had a family member diagnosed with COVID-19 whereas 99 participants did not have any member diagnosed with COVID-19 (one participant did not answer). One person had contact with a person with symptoms potentially related to COVID-19, whereas the rest had no contact with anyone who had symptoms or diagnosis of COVID-19.
Questionnaires
We used the CoRonavIruS Health Impact Survey (CRISIS) – Adapted for Autism and Related Neurodevelopmental conditions (AFAR) [10]. This questionnaire, which is openly available online (http://www.crisissurvey.org/crisis-afar/), assesses daily life behaviors, others clinically relevant to autism including RRB, as well as service changes occurring during the pandemic, and COVID-19 worries. The survey has three versions depending on the participant’s age and reporters, such as a caregiver-report form for children and adolescents (3–21 years old), a self-report form for youth and adults (14 years and older), and a caregiver-report form for adults with ASD. In the present study, we used the caregiver-report form for children and adolescents. The caregiver-report form for children and adolescents consists of 93 questions. Some were Likert fashion, whereas others were discrete variables or descriptive (please see details on the original version in English). Some authors (YS, TY, FJ, and HT) translated the English version into Japanese. The Japanese version was then back-translated by different authors (YYA and TH) (All language version of the questionnaire is openly available online (http://www.crisissurvey.org/crisis-afar/). For the purpose of this study, we selected on three specific domains including Changes in Life, Mass media and social networking service (SNS) usage, and RRB symptoms, as detailed below.
Besides the CRISIS-AFAR, we added original questionnaires that focus on the mask-wearing culture in Japan (CRISIS-AFAR-J). It includes six questions of the CRISIS-AFAR-J as follows:
- Did your child find it difficult to wear masks and go out?
- Did your child feel an uncomfortable sensation while wearing a mask?
- Do you think your child finds it difficult to communicate because others are wearing masks?
- Do you think your child finds it difficult to refer to other people’s emotions because they are wearing masks?
- Do you think that your child’s wearing a mask makes it harder for people to hear him or her?
- Do you think your child finds it difficult to convey his or her emotions to others while wearing masks?
The answers for these questions were formulated in a Likert-type format: (1) yes, (2) sort of yes, (3) I cannot say either, (4) sort of no, and (5) no.
Changes in life
We focused on questionnaires that measure sleep and physical exercise as well as media and SNS usage in the CRISIS-AFAR. The questions are listed in the Supplement. In the pairs of the questions, the first asked regarding the participants’ behavior during the 3 months prior to the pandemic, and the second asked regarding the participants’ behavior during the past 2 weeks (i.e., during the pandemic).
RRB symptoms
We focused on questionnaires that measure RRB symptoms before and during the pandemic in the CRISIS-AFAR. The questionnaires are listed in the Supplement. In the pairs of questions, the first asked regarding the participants’ behavior during the 3 months prior to the pandemic, and the second asked regarding the participants’ behavior during the past 2 weeks (i.e., during the pandemic).
Statistical analyses
We computed the composite scores using answers of questions on RRB in our data. The combinations of the answers were derived from the original CRISIS-AFAR Parent/Caregiver survey [10]. Briefly, in the original CRISIS-AFAR Parent/Caregiver survey, exploratory and confirmatory factor analyses (i.e., EFA and CFA) were performed on separate split-half datasets matched for demographic information, including sample, sex, child age, full-scale intelligence quotient, and primary DSM-5 diagnosis. In the EFA, the questions with resulting factor loading ≥ 0.3 were retained. Then, the CFA was subsequently performed only on the retained questions. Using the questions identified by these procedures, we calculated the composite scores in RRB. The lower-order RRB symptoms include sensory seeking, repetitive motor mannerisms/movements, and rituals and routines. The higher-order RRB symptoms refer to request to family members to maintain specific routines, rituals, and habits as well as engaging in an activity related to a highly restricted and strong interest.
Changes in life
Sleep and exercise
To see the change before and during the pandemic, we conducted Wilcoxon signed-rank tests in each pair of questionnaires to compare the scores of each question before and after the pandemic. Statistical significance was set at P < 0.005 (=0.05/11) after correcting for multiple comparisons using the Bonferroni method.
Mass media and SNS usage
To see the change before and during the pandemic, we conducted Wilcoxon signed-rank tests in each pair of questionnaires. Statistical significance was set at P < 0.017 (=0.05/3) after correcting for multiple comparisons using the Bonferroni method.
Association between hypersensitivity and mask-wearing
To examine the relationships between RRB and mask-wearing, we focused on the results of factor analyses: the lower- and higher-order RRB symptoms (see Results section). We conducted Spearmans’ rank correlation analyses between both the lower- and higher-order RRB symptoms and mask-wearing impact of CRISIS-AFAR-J. Statistical significance was set at P < 0.0042 (=0.05/6/2, Bonferroni corrected for multiple comparisons).