Abstract
This study examined COVID-19-related distress among adults with autism spectrum disorder and its associations with demographic and mental health information. A sample of 275 adults (ages 18–35) in the United States, all capable of providing self-reports, was recruited from a large national registry. Adults completed online surveys of mental health symptoms just before COVID-19 was declared a pandemic. Two months later, they completed the same surveys of mental health symptoms, and also reported about distress they had experienced as a result of the COVID-19 pandemic. Nearly two-thirds (65.9%) endorsed some form of COVID-related distress with a little over half reporting any individual form of distress. Although depressive and anxiety symptoms did not change from Time 1 to Time 2 on average, higher levels of COVID-related distress was associated with an increase in depressive and anxiety symptoms between Time 1 and Time 2. In addition, women and those who reported higher anxiety symptoms before the pandemic were more likely to report some form of COVID-related distress. The findings suggest that as this global crisis continues to unfold, it will be critical to ensure that adults with autism spectrum disorder who were experiencing mental health challenges prior to the pandemic receive appropriate monitoring.
Lay abstract
This study used data collected from 275 adults in the United States with autism spectrum disorder both before the pandemic and then 10 weeks into the pandemic to assess COVID-19-related distress and its impact. Two-thirds of those surveyed reported some type of distress related to the pandemic (i.e. difficulty coping or negative impact on emotional and mental health). While there were no changes in depressive and anxiety symptoms from prior to COVID-19 to 10 week later in the group as a whole, self-reported distress predicted increases in both anxiety and depression across the two timepoints. Furthermore, adults with higher levels of anxiety prior to the pandemic were more likely to report distress, and women were more likely to report a negative impact of the pandemic on their emotional and mental health. Findings highlight the importance of monitoring with adults with autism spectrum disorder to assess their need for mental health support, and providing ongoing support to those who already experience anxiety even before the pandemic.
Keywords
Events surrounding the COVID-19 pandemic are causing significant psychological distress and increased rates of mental health problems for people across the globe (McGinty et al., 2020; United Nations, 2020). Those already vulnerable to poor mental health are likely to be at particularly high risk during this time (Huremović, 2019; United Nations, 2020). Adults with autism spectrum disorder (ASD) fall in this heightened risk category due to their already high rates of depression and anxiety (Hollocks et al., 2019; Hudson et al., 2019; Lugnegaard et al., 2011), as well as negative life circumstances known to further impact mental health (e.g. negative social interactions/isolation (Liptak et al., 2011), unemployment/underemployment (Hendricks, 2010). Even more concerning, preliminary reports from early in the pandemic indicated that COVID-19 was exacerbating problems that were already common in this group, with many adults with ASD experiencing interruptions to their services/therapies, employment, and financial security during the initial months of the crisis (Chung, 2020; National Autistic Society, 2020; Pellicano et al., 2020).
While it is known that adults with ASD are experiencing increased rates of negative life circumstances during this time (Chung, 2020; National Autistic Society, 2020; Pellicano et al., 2020), little is known about their level of distress directly related to COVID-19 and if this distress affects mental health symptoms comparing to their symptom levels pre-pandemic. This study fills this gap by examining the impact of COVID-related distress felt by adults with ASD during the first 10 weeks of the pandemic in the United States on changes in mental health symptoms. An additional goal of this study was to identify characteristics that may place certain adults with ASD at risk for feeling COVID-related distress. To accomplish these objectives, this study capitalized on a natural experimental design opportunity that arose when, as part of a larger study, adults with ASD provided information about their mental health just prior to COVID-19 restrictions in the United States. Two months later, these same adults were again asked to provide information about their mental health, as well as specific information about COVID-related distress. The availability of a pre-COVID data point allows us to extend the existing work on COVID-19 in adults with ASD (Bal et al., 2021; Pellicano et al., 2020), most of which only included data collected during the pandemic.
Method
Participants
Data for Time 1 were drawn from a sample of 315 adults from the Simons Foundation Powering Autism Research for Knowledge (SPARK) registry (SPARK Consortium, 2018), recruited for a study on mental health services in adults with ASD. Email invitations were sent to 700 SPARK adult registrants who met the following inclusion criteria: (a) age between 18 and 35 years old; (b) able to complete self-report surveys and consent for themselves; (c) received an ASD diagnosis from a professional before age 18 years; and (d) completed the Background History Questionnaire at initial SPARK registration. This study had a recruitment goal of 300 participants. Typically, SPARK allots a 16-day recruitment period, but we exceeded our goal of 300 participants in only 8 days. Thus, recruitment was closed at that time: 322 participants consented and 315 completed all the online surveys at Time 1. At Time 2, 275 participants out of the 315 (87.3%) completed online surveys again. These 275 adults who completed both Time 1 and Time 2 comprise the final analytic sample (see Table 1 for baseline demographic characteristics).
Baseline characteristics of the adult sample who completed both Time 1 and T2 (N = 275).
DASS: Depression Anxiety Stress Scale; GED: General Educational Development.
Measures
The Depression Anxiety Stress Scale (DASS; Lovibond & Lovibond, 1995) was administered at both time points. Participants were asked to rate 42 statements describing emotional symptoms for three subscales with possible scores ranging from 0 to 42: the depression scale includes items assessing depressed mood and lack of interests and hope in life, the anxiety scale consists of items measuring anxious feelings and somatic experiences related to anxiety, and the general stress scale includes items describing stress-induced actions and emotions. In the current sample, Cronbach’s Alphas for the subscales were good (across each sub-scales and time points, range = 0.92–0.97) and test–retest reliability was good (across each sub-scales and time points, range = 0.73–0.77). The DASS provides clinical screening cutoffs based on a large non-clinical sample where the depression cut-off of 14 corresponded to the 88th percentile, the anxiety cut-off of 10 to the 90th percentile, and the stress cut-off of 19 to the 89th percentile (Crawford & Henry, 2003). Table 1 provides the percentage of participants in the current sample meeting these clinical screening cut-off scores.
At Time 2 only, participants responded to two questions related to feelings of distress due to the COVID-19 pandemic: “How well are you coping with the changes happening in your life due to COVID-19?” and “To what extent has your emotional or mental health been negatively impacted by COVID-19?.” Coping responses were re-coded into two categories: coping well (i.e. I am coping very well, I am coping well) and difficulty coping (i.e. I am having some difficulty coping, and I am having lots of difficulty coping) and impact responses were re-coded into two categories: low negative impact (i.e. minimally, not at all) and high negative impact (i.e. severely and moderately).
Procedure
Data collection happened in two waves: the first wave was collected between 11 March and 20 March 2020 (Time 1) right before COVID-19 restrictions were implemented across the United States; the Time 2 surveys were completed between 18 May and 27 May 2020, 10 weeks after COVID-19 was declared a pandemic.
For each wave of data collection, participants were directed to a webpage to provide electronic informed consent and then complete the online surveys, after which they received a $25 e-gift card for their participation. This study was approved by the institutional review board at the authors’ institutes (Cincinnati Children’s Hospital Medical Center, IRB # 2020-0338; University of California, San Francisco, IRB # 19-28191; Vanderbilt University Medical Center, IRB # IRB 200787).
Community involvement
A Community Advisory Council (CAC) of self-advocates and family members from the ASD community organized by the SPARK registry reviewed the study application and approved the study protocol.
Plan of analysis
Percentages reporting each type of COVID-related distress were calculated. Repeated measures ANCOVAs, controlling for age and sex, tested mean differences between Time 1 and Time 2 measures of depressive and anxiety symptoms as well as general distress. Multiple regression models predicted changes in depressive and anxiety symptoms from Time 1 to Time 2. For each model, either Time 2 depressive or anxiety symptoms was the outcome variable. Age, sex, Time 1 general stress (i.e. DASS stress subscale), and the Time 1 score of the outcome variable were entered on Step 1, and both distress variables were simultaneously entered on Step 2. As these models predicted Time 2 of the outcomes while controlling for Time 1 of the same outcome, the COVID-19 distress variables are actually predicting changes in the outcome variables from Time 1 to Time 2. In addition, controlling for Time 1 general stress increased the likelihood that the distress measured by the COVID-19 distress variables was unique to COVID-19 and not reflective of pre-COVID-19 general stress. Finally, logistic regression models tested individual characteristics (i.e. age and sex) and Time 1 depression and anxiety symptoms and stress as predictors of the two COVID-related distress questions separately. All analyses were performed using SPSS version 26.
Results
Of the 275 participants, 270 answered the two questions about COVID-related distress. A total of 65.9% (n = 178) endorsed some form of COVID-related distress, with 51.1% (n = 138) reporting difficulty coping with COVID-19-related changes, and 58.1% (n = 157) reporting that their emotional/mental health had been negatively impacted. Repeated measures ANCOVAs found no significant changes in depressive symptoms, anxiety symptoms, or general distress from Time 1 to Time 2 (all Fs (1, 270) < 1.88, p = ns), or in percentages meeting clinical screening cutoffs from Time 1 to Time 2 (all McNemar test p values > .13, see Table 1).
As seen in Table 2, multiple regression analyses found that adults who reported difficulty coping with COVID-19 changes and/or who reported negative impact of COVID-19 on emotional and mental health were more likely to exhibit increases in depressive and anxiety symptoms from Time 1 to Time 2. This association held even after accounting for general levels of stress prior to the pandemic, as well as age and sex. Also, as seen in Table 3, logistic regressions found that baseline anxiety symptoms predicted both difficulty coping and negative impact of COVID-19 on emotional and mental health. In addition, women were twice as likely as males to report that their emotional and mental health was negatively impacted by COVID-19.
Findings from multiple regressions predicting changes in depressive and anxiety symptoms from Time 1 to Time 2.
The findings reported for the main effects are from the final step of the regression model. CI: confidence interval.*p < .05, **p < .01, ***p < .001.
Findings from logistic regressions predicting COVID-19 distress measures.
The findings reported for the main effects are from the final step of the regression model. CI: confidence interval.
p < .05, **p < .01, ***p < .001.
Discussion
Less than 3 months after COVID-19 was declared a pandemic, nearly two-thirds of our sample of adults with ASD reported COVID-related distress. While there were no significant changes in depressive and anxiety symptoms from prior to COVID-19 to after the start of the pandemic, those who reported difficulty coping with COVID-related changes, and those who reported their emotional/mental health was negatively impacted by COVID-19, were more likely to experience increases in depressive and anxiety symptoms over the course of only 10 weeks. In addition, the analyses of the predictors of distress suggest that adults with ASD who have more anxiety symptoms to begin with may be at the highest risk for COVID-19-related distress and worsening mental health due to that distress. Our findings also suggest that women with ASD might be experiencing more distress than males during the pandemic, especially in respect to their own reports that their emotional or mental health was negatively impacted by COVID-19. Findings from this study extend existing work examining COVID-19-related distress in adults with ASD (Bal et al., 2021) in two important ways: by demonstrating relationships between pre-COVID mental health and COVID-specific distress, and by taking steps to separate out general, pre-existing stress from distress specific to the pandemic. Moreover, the current sample was recruited prior to the pandemic and not specifically recruited for studying its impact. Therefore, our findings are less likely to be biased by sampling issues of potential over-representation of those experienced COVID-specific distress.
These findings should be interpreted in light of these study methods. In the absence of a non-ASD comparison group, it is not known whether the effects of COVID-related distress are specific to adults with ASD. Furthermore, this sample reported experiencing high levels of internalizing problems before the pandemic: the percentage of the sample meeting clinical screening cutoffs is around three times higher than the percentage of the normative sample upon which the cutoffs are based (see measures section; Crawford & Henry, 2003). In addition, given that this study only included legally independent young adults who were capable of self-reporting and that the participation rate (i.e. number of participants compared to the number of individuals invited to participate) was relatively low, our findings may not generalize to adults with ASD in general. Finally, it will be important for future studies to replicate these findings using more comprehensive measures of COVID-related distress, rather than the single item measures used in this study, in order to provide more nuanced information about specific stressors that could be targeted for intervention.
Conclusion
In general, these findings highlight the need to directly query adults with ASD about distress related to the pandemic. The findings suggest a worrisome cycle of mental health problems, wherein adults who had more anxiety symptoms prior to COVID-19 were also at risk for experiencing the most COVID-related distress, placing them at further risk for mental health challenges. This was especially pronounced for females as seen in many previous studies (Mirowsky & Ross, 2003). While such a finding of increased risk for those who are already vulnerable may not be surprising, it is an important reminder for clinicians to take extra steps to ensure that adults with ASD who were experiencing anxiety and/or other mental health challenges prior to the pandemic, are carefully monitored. It will be important for continued research to further pinpoint those adults with ASD who are most impacted by the pandemic and thus most in need of additional services and supports.
Footnotes
Acknowledgements
The authors are grateful to all of the families in SPARK, the SPARK clinical sites and SPARK staff. They appreciate obtaining access to demographic data on SFARI Base. Approved researchers can obtain the SPARK population data set described in this study by applying at 
. We appreciate obtaining access to recruit participants through SPARK research match on SFARI Base.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the FAR Fund.