Introduction

The catastrophic impacts — both direct and indirect — of the COVID-19 pandemic on the lives of millions of people across the world are becoming increasingly clear. Some of these impacts have been the direct consequence of the virus itself, including a significant rise in serious acute ill-health and early death, to which disabled people [1], including Autistic people,Footnote 1 may have been particularly at risk [5, 6]. Other impacts have been caused by the measures put in place to try to curb the spread of the virus, including restricted access to face-to-face health and other social services; the closure of schools and other educational institutions; and the frequent requirement for people to stay at, and work from, home to dramatically constrain their physical and social interactions — measures that have varied substantially across and within regions [7]. Some of these measures may have had occasional positive effects above and beyond their impact on slowing the spread of the virus [8]. Yet, there are good reasons to believe that these measures imposed a particularly heavy burden on Autistic people [5, 6], especially as this group is at greater risk of experiencing mental health problems [9, 10] and may depend significantly on formal and informal support from others to maintain quality of life.

The current scoping review sought to synthesise evidence of the impact of the COVID-19 pandemic on the lives of Autistic adults. We focus on Autistic adults specifically given that they are an under-represented group in current autism research [11, 12]; are more likely to have co-occurring physical and mental health conditions [10, 13•], which place them at risk of health complications following COVID-19 infection; and have poorer life outcomes across many domains, often due to systemic and social failures to provide the requisite support [14].

Method

We addressed the following research questions:

  1. 1.

    What has been the impact of the COVID-19 pandemic on Autistic adults?

  2. 2.

    Which individual and/or contextual factors during the COVID-19 pandemic are associated with a more positive outcome for Autistic adults?

Based on research on the impact of the pandemic, we sought to identify lessons for the future care and support of Autistic adults. The review protocol was registered on Open Science Framework (https://doi.org/10.17605/OSF.IO/B73QN).

Eligibility Criteria

We used the following criteria to determine study eligibility:

  1. 1.

    An empirical quantitative and/or qualitative study on the impact of the COVID-19 pandemic and its associated restrictions on the mental health or wellbeing of at least one Autistic adult (> 18 years), based on self-report and/or observation/proxy-report. Autistic adult(s) should be mentioned explicitly in the paper’s “Method” section.

  2. 2.

    The sample should consist of adult(s) with a clinical diagnosis of autism or who self-identify as Autistic, and/or consists of informants for the Autistic adults. In cases where the sample also included Autistic children (< 18 years) or people with diagnoses other than autism (e.g. intellectual disability), the study is included only where Autistic adults’ data (a) are described separately in the paper or (b) make up ≥ 50% of the total sample.

  3. 3.

    English language.

  4. 4.

    Published in a peer-reviewed journal or as a pre-print on a pre-print service provider.

Search Strategy

On 24–25th May 2022, we searched through the electronic databases of PubMed, EMBASE, CYNAHL, Psychinfo and Science Citation Index (Web of Science) using the following search terms: [1] COVID-19 OR coronavirus OR SARS-CoV-2 OR pandemic AND [2] autis* OR developmental disorder AND [3] adult. Additionally, we searched through the major autism-related journals, including Autism, Autism Research, Journal of Autism and Developmental Disorders, Molecular Autism, Autism in Adulthood, Research in Autism Spectrum Disorders and Research in Developmental Disabilities using the following search terms: [1] COVID-19 OR coronavirus OR SARS-CoV-2 OR pandemic AND [2] adult. On 8th August 2022, we screened the same databases and journals for any newly published journal articles or pre-prints in the period from 26th May to 8th August.

Data Extraction

Table 1 shows key information extracted from eligible papers (where available).

Table 1 Data extraction form for selected papers
Full size table

Results

Study Selection

During the initial search (May 2022), we identified 747 abstracts. All abstracts were screened for eligibility by AMS. Almost half (n = 337; 45%) were independently screened by both AMS and LC, for which there was excellent inter-rater agreement (98%). Disagreements (n = 6; 2%) were resolved by discussion. Of the 747 abstracts, we selected 105 (14%) for a full paper read. The same two researchers read all 105 papers to determine final eligibility. Disagreements (n = 6 papers; 6%) were resolved through discussion, leading to a final selection of 47 (45%) papers. During the second search (August 2022), we identified an additional eight eligible papers, yielding a final list of 55 studies (see Fig. 1 [15, 16]).

Fig. 1
figure 1

PRISMA flow diagram of the paper selection process. PRISMA flow diagram derived from: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71

Full size image

Critical Appraisal

We assessed the quality of eligible studies using the Mixed Methods Appraisal Tool (MMAT, [17]). Each study was assessed in terms of five methodological quality criteria, which differed according to study methodology (quantitative, qualitative and mixed-methods). For each item, the study was rated ‘yes meets criteria’, ‘no does not meet criteria’ or ‘can’t tell’, where relevant information was not reported. EP rated all eligible studies (n = 55), half of which (n = 27; 49%) were independently coded by MH, with excellent agreement (88%) (LC rated papers for which EP and MH were also authors). Disagreements were discussed and resolved by consensus. The MMAT discourages allocating a numerical score, preferring a narrative description of a paper’s quality (see Supplementary Table 1). No studies were excluded based on their MMAT assessment, but study quality is applied in the interpretation of the findings.

Study Themes and Types

We categorised the 55 eligible papers as follows: 1) direct effects of the COVID-19 virus on Autistic adults including infection risk, hospitalisation, death and vaccination rates (n = 10); 2) effects of COVID-19 restrictive measures on care, support and services for Autistic adults (n = 11); 3) COVID-19-related effects on the wellbeing of Autistic adults (n = 32); and 4) other (n = 2) (see Table 1). Of the 55 studies, 30 (55%) were quantitative, 13 qualitative (24%) and 12 reported mixed-methods (22%). Thirty-five studies (64%) were, at least partly, based on Autistic adults’ self-reports, 10 (18%) on proxy-reports (caregivers or clinicians) and the remaining 10 (18%) on secondary analyses of medical/insurance databases. Eighteen studies (33%) included a non-autistic comparison group. Thirteen studies (24%) included data collected prior to pandemic onset. Most studies (n = 34; 62%) did not report on Autistic community involvement in the research process, and of those studies that did, most reported modest (n = 11; e.g. review of the survey) or no (n = 2) involvement of community members (see Table 1). In eight studies (15%), it was reported that Autistic researchers contributed to the research.

Demographic Data

Most included studies took place predominantly in the USA (n = 13; 24%) and the UK (n = 12; 22%), followed by Italy (n = 5; 9%) and Australia (n = 4; 7%). Thirty-one studies (56%) collected data in (multiple) European countries (including the 17 UK and Italian studies). Three studies took place in Israel, one in Canada, one in South-Africa, one in Turkey and one in Hong Kong (China). No studies presented data on Autistic adults in Central or South-America. Most (n = 34; 62%) collected data at least partly between March and May 2020 (i.e. during the earliest phase of the COVID-19 pandemic and before COVID-19 vaccines were available). The following policy rules applied in most jurisdictions during this time, although their implementation and enforcement varied substantially [7]: physical distancing; wearing face masks; disinfecting hands; cancellation of public events; restriction of international travel; closure of schools, non-essential shops, and restaurants and staying at, and working from, home. Seventeen studies (32%) collected data during a later phase, and three studies did not report timing of data collection. One study, examining risk for COVID-19 infection [13•], based their conclusions solely on pre-pandemic data.

Only nine studies (16%) focussed, at least in part, on Autistic adults with intellectual disability [13•, 18, 19, 20, 21, 22, 23, 24, 25]. For the 35 studies (64%) including Autistic adults’ self-report data, it is likely that participating adults did not have a co-occurring intellectual disability, since self-report requires the ability to understand and respond to questions (orally or in writing), usually without support. The male-to-female gender/sex ratio varied greatly across studies, with the lowest proportion of women (17%) in an employment study [26] and the highest proportion (94%) in a study with Autistic parents [27]. In studies including self-reports (n = 35), there was a preponderance of Autistic females (either female gender and/or sex) with 19 (54%) sampling > 50% Autistic females. In the 20 studies without any direct participation by Autistic adults (e.g. through the use of proxy-reports or medical records), the highest proportion of Autistic females was 39% [22] (excluding one study, where the single Autistic participant was female [25]). In 12 of 55 studies (22%), Autistic adults’ gender/sex was not reported, mostly because either professionals were the primary informant [28, 29, 30] or the sex ratio was described only for a larger group of participants with an intellectual or developmental disability (IDD) [18, 31].

What Has Been the Impact of the COVID-19 Pandemic on Autistic Adults?

Direct Effects of the COVID-19 Virus

A large-scale study in Israel (n = 32,812; 44% ≥ 16 years) reported that Autistic adults were two times more likely to be infected by COVID-19, compared to sex- and age-matched adults and also two times more likely to develop serious illness, resulting in higher hospitalisation rates [32]. Similarly, Autistic adult COVID-19 patients in the USA — just like those with an IDD — were shown to have a higher risk of Intensive Care Unit admittance and a significantly longer hospital stay, compared to their baseline group of COVID-19 patients without an IDD [33]. The causes for higher infection rates and more severe outcomes among Autistic versus non-autistic adults are not definitive but appear to be linked to contextual factors (e.g. living in a residential facility, receiving services in the home from outside caregivers) [13•, 18].

A review of COVID-19 policies in multiple European countries [34] demonstrated that, despite concerns about an increased risk, most countries did not prioritize COVID-19 testing for Autistic people. Furthermore, survey data showed that 55% of Autistic adults and 51% of caregivers of Autistic adults reported that reasonable adjustments to COVID-19 testing (e.g. a support person or sensory accommodations) were not made possible [34]. Moreover, it appears that Autistic people with an intellectual disability may have been discriminated against, as some countries used cognitive impairments as an exclusion criterion for Intensive Care Unit admittance in case of triage [34].

Given Autistic adults’ increased risk of infection, some advocates called for prioritization of COVID-19 vaccination in this group. During the early phase of a mass vaccination campaign in Israel, Autistic adults were reported as being more likely to be vaccinated, compared to a sex- and age-matched comparison group, particularly in the 21–40 age group (Autistic: 62%; non-autistic: 29%; [35]). In a US-based study, most Autistic adults (78%) were said to have had received or planned to get a COVID-19 vaccine [36].

In sum, Autistic adults appear to be at increased risk for COVID-19 infection and show poorer outcomes following infection compared to non-autistic adults. Despite this risk, COVID-19 testing has not been prioritized for this group. Given that only two studies in this review reported about COVID-19 vaccinations for Autistic adults, it remains uncertain whether COVID-19 vaccinations were consistently prioritized.

Impact on Services and Supports

Autistic adults, caregivers and professionals reported major disruption of services for themselves or those they supported during the COVID-19 pandemic [29, 34, 37]. The barriers to accessing healthcare, which were already excessive for many Autistic people prior to the pandemic [38, 39, 40], were deemed even greater during the pandemic [41•]. Due to physical distancing and stay-at-home orders, many services shifted to remote delivery to ensure continuity of care. Experiences of so-called ‘telehealth’ services varied substantially, linked to the nature of the service. Telehealth services were reported to be an acceptable alternative to in-person service delivery for primary care [42] and diagnostic assessments [43], largely because virtual appointments can minimise barriers to care for Autistic adults (e.g. less stress negotiating travel and waiting rooms; [39]). Yet, remote mental health services were not accepted or preferred by everyone [44••, 45, 46] and were advised against in cases of high psychological distress or intellectual disability [41•]. Autistic adults who accessed remote mental health services reported challenges with sensory issues [44••, 46], body awareness [44••, 46], technological issues [41•, 42, 43, 46] and effective communication with their therapist [44••, 46]. Some reported experiencing discontinuity of care as a result [46], and those who felt they did not benefit from telehealth services early in the pandemic were more likely to experience mental health distress 2 months later, compared to those who felt they benefited [47•]. These mixed findings suggest that, even though telehealth services may be an acceptable and welcome alternative for some Autistic adults, the specific nature of service delivery should depend on individual preference.

Impact on Wellbeing

Of the 32 studies examining the pandemic’s effects on wellbeing, 13 (41%) were quantitative, 10 (31%) qualitative and nine (28%) used a combination of qualitative and quantitative methodologies. Most studies assessed the pandemic’s effects on the wellbeing of Autistic adults (either via self-report or proxy-report) in a retrospective way, that is, by asking participants to reflect on and estimate the impact of the pandemic, compared to their memory of pre-pandemic experience. Almost all retrospective studies indicated an overall decrease in wellbeing during the pandemic. Only 10 (31%) studies were prospective, where the wellbeing of Autistic adults was assessed repeatedly over time. These prospective studies revealed mixed experiences, demonstrating either overall stability of psychopathology symptoms, loneliness and stress in Autistic adults during the first months of the pandemic [47•, 48, 49], an overall decrease in psychopathology symptoms [44••] or an increase in acute psychiatric distress, as evidenced by increased psychiatric emergency admissions of Autistic adults before and during lockdown [50].

Perhaps unsurprisingly, there were large individual differences reported within these prospective studies, even for those reporting negligible overall effects — suggesting that, for some Autistic adults, the COVID-19 pandemic and its restrictions may have had positive effects on their wellbeing, while for others, it had negative effects. This variability was confirmed by studies adopting in-depth qualitative retrospective approaches, which also revealed potential reasons for this variability. Most studies reported differences between individuals, with some Autistic adults highlighting benefits of the restrictions (e.g. relief of pressures from the external world, control over sensory stimulation, ability to spend quality time with family) and others reporting serious and damaging effects of these same restrictions (e.g. loss of social contacts and friendships, no access to professional support, and loss of independence) [44••, 45, 46, 51, 52••]. Unsurprisingly, some of these studies also reported substantial differences within individuals, with some Autistic adults experiencing both positive and negative effects of the stay-at-home orders at different points during the pandemic [27, 44••, 46, 52••, 53]. This latter finding might provide one explanation for the apparent stability in psychological wellbeing reported by several prospective, quantitative studies [47•, 48, 49]. Another possibility is that such stability is driven more by the consistently deeply unsatisfactory quality of life and poor mental health experienced by many Autistic adults pre-pandemic (see [14], for review) than the impact of the pandemic itself.

Which Individual and/or Contextual Factors During the COVID-19 Pandemic Are Associated with a More Positive Outcome for Autistic Adults?

Individual Factors

Many studies identified individual characteristics that were reportedly predictive of better mental health in Autistic adults during the pandemic, including being male gender/sex [47•, 49], older [47•, 48], having relatively good mental health prior to the pandemic [44••, 47•, 48, 49], low COVID-19-related stress [27, 48, 49], high perceived social support [46, 48, 52••] and being able to maintain routines or develop new ones [44••, 52••, 54]. Caution is warranted, however, as these effects (particularly regarding gender/sex and age) were not consistently replicated [44••, 48].

Contextual Factors

There were several contextual factors that appear to have had a positive effect on Autistic people’s mental health or wellbeing during the pandemic, including (continued) access to (professional) support [27, 44••, 52••, 53], reduced pressure to conform to societal rules [27, 46, 51, 53], reduced sensory overload or greater control over the sensory environment [44••, 52••, 55], absence of COVID-19 infection [47•], strong social connections and activities (including maintaining contact with friends, family and community online) [44••, 46] and clear and accessible public health messaging with regard to COVID-19-related information and rules [52••]. Societal inclusion and acceptance are also likely protective factors of Autistic adults’ mental health in general [53].

Discussion

Through our scoping review, we have demonstrated that Autistic adults are likely to have been disproportionately impacted by the COVID-19 pandemic in several ways: [1] they were at increased risk of contracting COVID-19 and developing serious illness; [2] they often struggled to access critical services and supports as a result of stay-at-home restrictions implemented in many jurisdictions; and [3] despite some positive uplift due to reduced external pressures, many experienced challenges to their mental health — particularly those who were female, had pre-existing mental health issues and had limited social connections and support. These findings are consistent with the pandemic’s detrimental effects on disabled people more broadly, referred to as the “triple jeopardy” [1].

While these findings are already deeply concerning, our review likely significantly underestimates the negative impact of COVID-19 on Autistic adults’ health and wellbeing, since most reviewed studies relied on selective samples. For instance, almost all studies were conducted in high-income countries, meaning we know virtually nothing of the experiences of Autistic adults residing in low- and middle-income countries, many of whom may be undiagnosed and with limited or no access to appropriate mental health care [56, 57]. We also know little of the experiences of Autistic adults from seldom-heard groups in autism research, including those who face economic hardship, are of minority racial/ethnic backgrounds, use non-traditional forms of communication, and/or have an intellectual disability. Autistic adults identifying with one or more of these groups were vastly under-represented in the studies we identified, which makes it difficult to draw conclusions about their experiences of COVID-19. It is likely, however, that those who are multiply marginalized and those living in low- and middle-income countries will have responded far worse than the identified studies suggest.

Aside from the sampling issues outlined above, many of the identified studies were also of relatively poor methodological quality. While there was an urgent need to understand the impact of COVID-19 on the Autistic and autism communities, many studies relied, often for logistical and practical reasons, on retrospective designs as well as measures of often questionable reliability, validity and generalisability. For instance, several studies relied on bespoke rating scales, single items or failed to report a measure’s reliability within an Autistic sample (see also Supplementary Table 1). Many studies also did not account for potential confounding factors, especially the extent and nature of COVID-19-related restrictions at the time of study. Despite clear research recommendations formulated early on in the pandemic [58], concerns regarding the scientific rigor of the COVID-19 literature are not limited to autism research [59, 60]. Such concerns are worrying nevertheless because conclusions based on less-than-sound science can pose significant challenges to clinical and policy-relevant decision making and may also reduce trust in autism science.

Our review may also be limited in other ways. We acknowledge that our search strategies may not have been exhaustive, especially regarding the exclusion of non-English language studies, and that our review included several studies of our own, which may have increased the risk of bias. It also may have over-represented issues faced during the first phase of the pandemic (likely due to lags between study completion and publication), rendering it important to examine later and long-term impacts of COVID-19 and its associated restrictions. Further limitations of the reviewed studies likely transcend the pandemic situation. For example, level of community involvement was rarely reported or was very modest (e.g. questionnaire review), indicating many missed opportunities to actively learn from and be guided by the needs and expertise of the Autistic community [58]. Such opportunities may have been particularly pertinent during the unprecedented pandemic situation.

Lessons Learned

Despite the aforementioned concerns, there are several important lessons we can draw from this review on the future care and support of Autistic adults. First, Autistic adults have been disproportionately affected by the COVID-19 pandemic — both by the effects of the virus itself and the social measures used by governments to curtail it. We therefore need a distinct strategy to support Autistic adults during times of crisis and beyond. Such a strategy must be deeply informed by, and actively co-produced with, the Autistic community to ensure it is responsive, respectful and relevant to Autistic adults’ lives.

Second, continued access to effective, knowledgeable and tailored care and services was vital for Autistic adults’ wellbeing during the pandemic. We need greater investment in the development — and rigorous testing — of services and supports that are tailored to the needs, preferences and abilities of individual Autistic adults, particularly those that optimise the person-environment fit [61]. Autistic adults should also be actively involved in making decisions about specific, formal supports (in terms of personnel and delivery) and be able to shape the systems and policies enabling access to these supports, even during health crises.

Third, although our review revealed much variability in the extent and nature of the impact of the pandemic on Autistic adults’ mental health, many experienced acute loneliness and social isolation, with some reporting severe mental health distress. This finding challenges prevailing stereotypes of autism — that is, that Autistic people prefer a life of self-isolation. In future, more concerted efforts, ideally led by Autistic-led organisations, must be made to implement and maintain social support structures for the Autistic community [46, 48], especially during times when social supports are otherwise curtailed.

Finally, few studies in our review reported actively including Autistic adults or their allies in the research process. Participatory autism research is still in its infancy, with researchers often reporting that it can be time-consuming and challenging [62]. The pandemic may well have exacerbated these challenges [63]. Researchers should seek to deepen their involvement with the Autistic community so that they can respond rapidly and flexibly in time of crises such as this one. Doing so will ensure that future research, clinical services and care are aligned with the needs and priorities of Autistic people [27], accelerate translational research and enable Autistic people to have their perspectives and experiences fully taken into account.