Abstract
This article includes suggestions generated from input from people with autism spectrum disorder or Asperger’s syndrome for how to support the mental health needs and promote the participation and well-being of autistic adults in case of another pandemic or other unexpected event.
According to autistic people, autism is a neurodevelopmental disability characterized by strong interests, a preference for consistency in daily routines, and differences in social processing (Autism Self Advocacy Network, n.d.; Bury et al., 2020). Some of these characteristics have likely made the impacts of the coronavirus disease 2019 (COVID-19) pandemic particularly difficult for adults who are on the autism spectrum. For example, many states were concerned with reducing the spread of the virus; subsequently, officials in those states abruptly stopped in-person gatherings to reduce person-to-person contact. Prepandemic research has shown that social occupations are related to autistic people’s quality of life (Dallman et al., 2021; DaWalt et al., 2019); thus, sudden changes such as these may have negatively affected their preferred social occupations and led to an increased risk of mental illness, because autistic adults experienced high levels of pre-pandemic anxiety (van Steensel et al., 2011).
Experts have highlighted the need to attend to the mental health–related impacts of the pandemic on autistic adults, with uncertainty and social isolation indicated as key areas of concern (Cassidy et al., 2020), but little is known about the experiences and needs of autistic adults during the COVID-19 pandemic. Paradoxically, the emerging limited literature elucidating the experiences of autistic people during the pandemic has suggested that lessened social demands can reduce stress and fatigue for autistic adults while also increasing feelings of social isolation, anxiety, and depression (Nisticò et al., 2022; Oomen et al., 2021).
Changes to social participation during the COVID-19 pandemic are likely not the sole cause of decreased mental health among autistic adults because disruptions to and decreased participation in healthy routines may also be a key factor that has uniquely influenced them (Yarımkaya & Esentürk, 2020). Because physical health is linked to mental health, a reduction in behaviors promoting physical health may also increase the risk of worsening mental health (Goodwin, 2003; Salmon, 2001).
Before the COVID-19 pandemic, the coping strategies of autistic adults were varied and included social support from peers, teachers, and family members (Dachez & Ndobo, 2018; First et al., 2016). The pandemic disrupted these coping strategies, along with the support provided by community services and programs. Other research has shown that some autistic adults may have used social media as a coping method and outlet to seek friendships and social connection when in-person interactions were not possible; however, others experienced increased stress as a result of the constant presence of anxiety-provoking information regarding current events and health on social media platforms (Guessoum et al., 2020). Autistic adults have reported a desire for more autism-related advice and guidance amid the pandemic to guide behavior and coping strategies (Oomen et al., 2021). Similarly, clinicians (e.g., occupational therapists, speech-language pathologists) and other care providers are in need of guidance as they seek to support autistic adults during the pandemic. Increased mental health burdens and the reduction in coping opportunities among this population have been reported and predicted; however, limited information has been published about the firsthand experiences of autistic adults amid the COVID-19 pandemic.
The aims of the current study were to (1) understand the mental health experiences of autistic adults during the COVID-19 pandemic and (2) identify ways that occupational therapists and other clinicians can help autistic adults during unexpected and challenging times, as reported by autistic adults. We use identity-first language (e.g., autistic person) throughout this article because disability advocates and scholars assert that person-first language (e.g., person with autism) contributes to disability stigma (Collier, 2012; Gernsbacher, 2017).
Method
Research Design and Participants
Data collection and subsequent analysis was guided by an interpretative phenomenological analysis (IPA) framework (Smith et al., 2009). This sequential explanatory mixed-methods study was designed to gather quantitative and qualitative information from autistic adults to uncover how their daily lives, routines, and mental health changed as a result of the COVID-19 pandemic. Because the research questions were guided by IPA principles, the focus of the research questions was on understanding how the participants made sense of their individual experiences. In addition, it is worth noting that the COVID-19 pandemic was unfolding as data were collected for this study and that IPA methods are suitable for understanding events as they are lived in real time (Finlay, 2013).
To accomplish the study’s aims, participants (N = 34) were recruited who met the following inclusion criteria: self-reported diagnosis of autism spectrum disorder (ASD) or Asperger’s syndrome, ages 18–35 yr, residing in the Eastern United States, able to understand English, and able to participate in a one-on-one interview using either verbal or written communication. Recruitment was conducted via snowball sampling through local agencies serving autistic people.
Procedure
Before the interview, participants completed a brief online survey via Qualtrics that included of the Depression Anxiety and Stress Scale (DASS–21; Lovibond & Lovibond, 1995), detailed in the next section. In the survey, all participants were invited to complete a 60- to 90-min virtual interview (interviews occurred between August 15, 2020, and May 1, 2021). Interviews were conducted by the first author (Aaron R. Dallman) or a graduate student trained in interview methods and supervised by Dallman. All interviews followed the same script (see the Appendix) and were supervised by Dallman to ensure that fidelity was maintained. Interviews were adapted to fit individual preferred communication styles. All interviews were conducted at least in part orally, although several participants responded to some questions in written format, and some requested to have questions presented both orally and in written format. Participants were also given ample opportunities for breaks. At the start of each interview, participants were told that they could skip questions if they did not want to answer them or if the questions made them uncomfortable, turn off their camera at any point in the interview, and take as many breaks as they desired. Before participation, all participants completed informed consent procedures and, after the interview, were compensated for their time with a $20 gift card. All procedures were approved by the Towson University Institutional Review Board.
Measures
To characterize participants’ emotional states and assess symptoms of depression, anxiety, and stress, participants completed the DASS–21 (Lovibond & Lovibond, 1995), a 21-item self-report scale that assesses depression, anxiety, and stress. It has three subscales: Stress, Anxiety, and Depression. Each subscale consists of seven items. Individual items are rated on a scale ranging from 0 (did not apply to me) to 3 (applied to me very much all of the time), with scores indicating normal to extremely severe levels of stress, anxiety, and depression (Lovibond & Lovibond, 1995). A sample item is “I experienced trembling (e.g., in the hands)” (Lovibond & Lovibond, 1995). The DASS–21 demonstrates strong internal consistency (α = .93; Henry & Crawford, 2005) and good convergent validity; scores on the DASS–21 subscales have a moderate correlation with scores on the Hospital Anxiety and Depression Scale (HADS; Zigmond & Snaith, 1983) Depression subscale (DASS–21 Depression, r = .66; Anxiety, r = .49; Stress, r = .54), as well as between the DASS–21 subscales and the HADS Anxiety subscale (Depression, r = .59; Anxiety, r = .62; Stress, r = .71; Crawford & Henry, 2003).
We also developed a nonstandardized demographic questionnaire to characterize participant demographic information. This demographic questionnaire included questions about participant age, ethnicity, income, and other relevant participant characteristics.
Data Analysis
To ensure our data were valid, we ceased collecting data once data saturation was achieved and began data analysis. Our conceptualization of saturation was grounded by Braun and Clarke (2021), and for our study, saturation was achieved when Dallman’s initial review of audio recordings from the interviews seemed to, on the surface, reflect no new ideas. That is, recordings from Interviews 29 to 34 revealed no new information from that obtained from the first 28 interviews. Saturation was then verified at the end of the data coding (discussed next). To analyze the qualitative data from the semistructured interviews, we followed the procedures outlined by Cascio et al. (2019). First, we generated verbatim transcripts. Zoom generated an initial transcript that was then checked and corrected by a graduate research assistant supervised by Dallman. Next, two coders each independently and iteratively coded each transcript using NVivo (Version 12). Dallman served as the master coder. All qualitative analyses were guided by the principles of IPA (Smith et al., 2009). Each coder read the transcripts multiple times, identified segments of meaningful text (a data-driven approach), and evaluated these segments of text for their value in providing information about the phenomenon of interest.
In the initial round of open coding, three team members each analyzed the same two transcripts; met to discuss, compare, and contrast codes; and identified an initial codebook. This initial codebook was a living document that was updated after each transcript was analyzed. Each subsequent transcript was analyzed by two team members, which resulted in an initial codebook consisting of a total of 160 codes. Because the coders analyzed data independently, disagreements in coding were expected. Throughout the analysis process, Dallman met with the coding team to discuss any discrepancies in an analyzed transcript until 100% agreement was reached. Once agreement was reached, overlapping codes were merged by Dallman, resulting in a total of 130 codes. Afterward, Dallman organized these codes into two themes according to the data analysis strategies discussed in Krueger and Casey (2015). Last, Dallman continued the iterative analysis by rereading each transcript to verify that all themes were applicable to the data. Throughout the data analysis, the research team met to discuss the research findings, codes, thoughts, and reactions in accordance with guidance on achieving rigor in qualitative research (Morse, 2015).
The DASS–21 and demographic questionnaire were used to characterize participants; thus, only summary descriptive statistics (e.g., means and standard deviations) were calculated.
Results
A total of 34 participants completed the survey and the interview. One participant completed the survey, but not the interview, and thus this person’s data are not included. Table 1 provides demographic information for the participants. Participants shared how changes to occupational performance during the COVID-19 pandemic both supported and were detrimental to their mental health. The two themes presented in this article are (1) autistic people’s social experiences during the COVID-19 pandemic paradoxically supported and hindered their mental health and (2) during the COVID-19 pandemic, autistic people’s mental health was tied to daily activity.
Demographics
Note. N = 34. ADHD = attention deficit/hyperactivity disorder; COVID-19 = coronavirus disease 2019; DASS–21 = Depression Anxiety Stress Scale; GED = general equivalency diploma.
Participants could have more than one diagnosis.
Percentage of participants meeting the cutoff for severe or extremely severe depression, anxiety, or stress.
Autistic People’s Social Experiences During the COVID-19 Pandemic Paradoxically Supported and Hindered Their Mental Health
All of the participants discussed how the COVID-19 pandemic changed their social participation. Although participants described finding new ways of connecting with others when in-person social interactions were not possible, they experienced these changes to participation as both positive and negative. For example, many participants used social media, virtual social groups, and other virtual meeting spaces to find new connections. However, these virtual social connections were often experienced as uncomfortable. Participants’ descriptions of their experiences made it clear that these changes to social participation affected their mental health in several important ways.
Changes That Positively Affected Autistic Adults’ Mental Health
The most endorsed positive change to social interactions during the pandemic was the shift in social norms. Changes to social routines better matched the preferred interaction styles of autistic adults in several ways. For example, several participants noted that social handshakes give them feelings of anxiety, but that changes to social norms and the belief that “handshakes are just not going to come back” relieved them. Similarly, participants were pleased when hugs and other physical contact gestures were stopped as a result of the pandemic. Participants also preferred other protectionary practices; they reported enjoying social distancing when talking to friends, smaller crowds in stores, and fewer social gatherings. These practices, which most people performed to reduce the spread of COVID-19, were part of the participants’ preferred social interaction style. For example, participants shared how shopping before the pandemic was exhausting; however, crowd-limiting measures implemented by stores reduced this exhaustion and made them feel more content.
The pandemic also provided an opportunity for positive changes in social experiences among participants because they could now express agency in their social interactions. For example, one participant shared a story about being autistic in college: Everyone knows each other really well and so there is that pressure of, like, if you see someone in the hallway you have to say hi to them, and because it’s such a small program in such a small building, you have to say hi to them all the fucking time.
Virtual classes provided this participant an opportunity to avoid these unwanted interactions. They could turn off their camera, mute themselves, and take necessary mental health breaks without drawing undue attention. This stark contrast in social participation via the virtual platform enabled participants to express agency in social situations and reduce the frequency of occupations that negatively influence their mental health, such as expected social connection in the higher education classroom. Another participant noted how virtual interactions allowed them to use coping mechanisms that may not be socially accepted in different contexts (e.g., stimming).
Changes That Negatively Affected Autistic Adults’ Mental Health
Participants also noted several negative social experiences. They frequently mentioned how virtual interactions felt different. In person, social cues grounded participants in their social interactions: “For me, talking face to face with someone, it’s almost like checkboxes go off in my brain. It’s like you’ve gotten real social interaction. You’ve gotten the physical [connection, and when that interaction happens virtually] it’s not a genuine exchange.” Participants offered some explanations for why these virtual interactions might feel different. For example, “there’s either these highly pregnant silences that occur and I don’t, and people don’t use their webcams and everyone’s a bit timid. No one really speaks.” These seemingly unnatural pauses felt awkward to participants and made virtual interactions anxiety provoking.
Although participants sought social agency, the ability to choose social participation, they also grieved the loss of fortuitous social meetings. Before the pandemic, participants’ daily routines provided them with the opportunities to meet friends and other social acquaintances. As one participant said, “I’m [used to having] something that I need to go to and seeing people in person. . . . I’ve always been, you know, bad at just, like, calling someone up and saying, hey, you want to go out and do something?” The pandemic reduced the number of opportunities to have spontaneous social interactions, a primary source of social contact for many participants.
During the COVID-19 Pandemic, Autistic People’s Mental Health Was Tied to Daily Activity
Participants described increased anxiety during the COVID-19 pandemic. As aptly stated by one participant, “My anxiety is through the roof right now because of the pandemic.” Frequently, participants increased their previous anxiety-like behaviors to reduce their chances of catching COVID-19. As one participant described, “I ended up doubling down on a lot of things that I need to be doing. I wash my hands, probably more than I should. . . . It makes anxiety issues in certain, in certain instances, a lot worse.”
One cause for anxiety and depression among participants was the lack of routine, which led to changes in participants’ mental health. A lot of the time my mood is lower with just this lack of routine, and you know I felt that way, you know, on and off for the, for the past several months. . . . And when I’m feeling down, it’s, [it] definitely tends to be a bit more of an anxious . . . down than it has been a lot of other times. You know, it’s not something that I haven’t experienced before.
Similarly, one participant described how routine changes in combination with avolition reduced their mental health. Sometimes it’s very hard to get that motivation to actually do something when, when I know that I can’t leave the house or do anything and knowing that my, you know, my social, you know, I guess my social channels aren’t exactly open at the moment.
This participant sought social support yet lacked the opportunities or possibly the strategies, given the opportunities they did have, to seek that social support.
Most participants also reported a connection between social support and mental health. Before the pandemic, autistic adults connected with friends when they needed social support. However, during the pandemic, many relied on their families for mental support, which placed a strain on their familial relations: “I feel like I walk on eggshells, a little bit more than I did before this, because there were more outlets to be able to talk with people instead of having to rely just heavily on my family.” In another case, an autistic college student lived with her partner and her partner’s family: I ended up relying on him for basically all the cooking and the cleaning, like him and his mom, because my energy levels were just like shot to hell with, like, online classes, because we were adapting to that, and I just wouldn’t have any spoons left for basically anything.
When self-care occupations were negatively affected, social support became an important facilitator of well-being.
Discussion
Autistic adults were uniquely challenged by the COVID-19 pandemic. Some changes to routines were difficult for participants; however, other changes (e.g., adaptations to social routines) were supportive of participants’ mental health. The COVID-19 pandemic highlighted the importance of social agency to the well-being of autistic adults. This finding may explain why previous literature has found a paradoxical relationship between the desires for socialization and for isolation among autistic people (Haertl et al., 2013). Before the pandemic, participants felt they lacked agency in their social interactions; they were frequently required to participate in unwanted social interactions as a result of societal norms. Previous research has found that autistic college students find the volume of social interactions during college difficult (Bailey et al., 2020). In our study, autistic adults reported challenges with the volume of unsolicited social contact. When participants were in public spaces, they felt compelled to interact with others in that space to meet expectations. They feared that if they did not respond to a peer’s social overture or showed their reluctance to interact, they would be viewed as mean or unsocial.
In our study, participants described experiences of gaslighting, in which an autistic person is made to “question his or her own sense of reality” (Raymaker et al., 2020, p. 140). Although our study was conducted only during the pandemic, the autistic adults in our study described experiences of gaslighting outside of the pandemic; these were moments in which nonautistic people failed to empathize with autistic adults’ social experience and thereby forced them to accommodate nonautistic people’s communication style. The changes to social routines associated with the COVID-19 pandemic provided space for autistic adults to express agency in these social interactions in ways that did not elicit the gaslighting response to which they were accustomed. When joining virtual meetings, it was easy for participants to avoid unwanted initial pleasantries by joining meetings right at meeting start times. In addition, the culture of video-based interactions is different from that of in-person social interactions. Participants shared that many people on video calls are on mute, so the typical loud chatter that happens in meeting spaces is reduced, improving their experience.
Although the participants reported an aversion to unwanted social contact, they also emphasized a strong desire for social interaction. All of the participants indicated an interest in connecting with their peers in person. Participants described interest in returning to social occupations such as social groups, in-person classes, and informal gatherings. This important finding contrasts with theories of autism, such as the social motivation theory of autism (Chevallier et al., 2012), that suggest that limited social motivation is part of the autism phenotype, and it is in line with other self-reported research findings on the social motivation of autistic adolescents (Dallman et al., 2021). This paradox suggests that autistic people want the agency and opportunity to engage in social opportunities, but not the expectation.
An interesting finding was that participants noted a strong preference for in vivo versus video-based interactions. When participants described video-based interactions, they noted that virtual interactions feel different and lack the “essence” of a person, making them robotic and uncomfortable. Although efforts to use virtual reality to understand autistic social behaviors and increase more typical (i.e., more neurotypical) communication are increasing (see Parsons, 2016, for a review), our findings suggest that autistic adults may experience virtual social interactions as less pleasant social overtures. This differs from past findings that autistic people prefer virtual interactions, specifically when using chat functions (Johnson, 2014). Our findings suggest that, for many autistic people, face-to-face interactions are preferred over online interactions. This echoes results from another study conducted during the pandemic in which autistic participants who were receiving ongoing support indicated that they preferred in-person to phone-based support (Oomen et al., 2021).
Autistic masking, hiding one’s autistic behaviors to present as neurotypical, has been related to exhaustion (Hull et al., 2017) and depression (Cage et al., 2018). Our participants described a reduction in masking behaviors during the pandemic, primarily because most of their time was spent at home. Our participants also provided strong evidence that this reduction in the perceived need to mask promoted their mental health. We consider these adaptations to normal routines to be neurodiversity-affirming approaches—that is, approaches that allow for people with differences in their ways of being to be successful. Features such as smaller crowds, less physical touching, more space between people at in-person gatherings, social agency, and the ability to remain in one’s own home or space where autistic behaviors were accepted are all important features of the pandemic. Therapists, clinicians, and educators should take note of these components and consider how these supportive environmental features can be implemented in nonpandemic times.
Lack of routines during the pandemic also had a strong impact on mental health. Given that preferences for structure and routine are hallmarks of ASD (Condy et al., 2019), this finding was not surprising. In pre-pandemic times, a lack of structure was shown to affect autistic adults’ motivation and engagement in occupations (Haertl et al., 2013). The results of the current study show that changes to routine were especially difficult because participants experienced limited social connections and were forced to spend more time with family, leading to conflict. Respondents noted that this was partly due to restricted access to social supports outside of their living situation. Social support has been associated with higher levels of quality of life among autistic adults (Bishop-Fitzpatrick et al., 2018), whereas there is a negative relationship between loneliness and life satisfaction and a positive relationship between loneliness and depression and anxiety (Mazurek, 2014). This suggests that rather than requiring autistic people to seek support on their own, therapists can increase the accessibility of support by connecting them with resources or creating social groups for this population.
Limitations
Although we believe these findings are significant and relevant for occupational therapy practitioners, several limitations should be considered when interpreting these findings. First, diagnoses of ASD were all self- reported. We chose to not administer any diagnostic instruments because we aimed to reduce participant burden. In addition, the majority of the participants communicated using verbal language. Future studies should make a concerted effort to include other autistic adults who communicate using nonverbal means. Last, we did not collect any demographic information on participants’ current living situation (e.g., with whom a participant was currently living). Although this information was elicited in most interviews, it was not systematically collected. It may be that participants’ experiences of the pandemic were influenced by their present living situation.
Implications for Occupational Therapy Practice
Our findings clearly indicate that occupational therapy services are needed among autistic adults. Occupational therapists can help autistic adults find social experiences that support their neurodiversity and cope with mental health struggles. Although the focus of our study was on experiences during the pandemic, the findings suggest a strong need for services into adulthood, specifically by addressing the accessibility of social supports, creating routines that support mental health, and advocating for a shift toward neurodiversity-affirming social interactions. Although pandemics and shutdowns are not frequent occurrences, autistic people may face other events that interfere with their social, sensory, and mental health experiences, as well as their occupational engagement. Occupational therapists should consider consulting with autistic clients to ensure that these services are ready when challenges or a need for adaptations arise. The pandemic has helped to reveal what supports and hinders autistic people’s participation; now, occupational therapists must incorporate these supports and recommendations into everyday practice.
Conclusion
This qualitative study provides an understanding of how the changes to routine and occupations during the COVID-19 pandemic affected the mental health of autistic adults. We found that autistic adults felt as though they are rarely given social agency, the ability to choose when and how to participate socially; however, the pandemic altered the conditions of their social interactions, allowing them to focus on participating on their own terms. The change in and lack of routines also affected mental health, resulting in participants expressing difficulty in seeking social support and adapting to these changes. Therapists should acknowledge these preferences as institutions prepare to return to normal, or a “new normal,” and social expectations are again placed on autistic people. Interventions should be tailored to address both the toll the pandemic took on autistic people’s mental health and the creation of spaces that fit their specific social needs.
Footnotes
Acknowledgments
This project was partially funded through support provided by the Dean of the College of Health Professions at Towson University.
Appendix. Interview Guide
How did you first learn about COVID-19?
When you first learned about COVID-19, how did you feel?
For many individuals, their routines and schedules have changed a lot as states have shut down. For example, many programs that were offering in-person activities moved to virtual activities. How have routines changed in response to COVID-19?
Probe: Were there any positive routine changes?
Probe: Were there any negative routine changes?
What has been your experience obtaining social support during this time?
Probe: What did your social support look like (friends, family members)?
Probe: Have you kept in contact with you friends you used to meet in person?
Probe: Before COVID, did you have online friends?
Do you still keep in contact with those friends?
What has been the most difficult for you during the pandemic?
What could researchers and clinicians do to make the time easier for you?
In hindsight, if you had more time to prepare, what, if anything, would you have done differently?
Probe: Why would you have done that?
Note. All questions were adapted based on the language style or preferences of the participant. Often, questions were made more concrete through examples. COVID-19 = coronavirus disease 2019.