Accumulating evidence since the onset of the coronovirus disease 2019 (COVID-19) pandemic has demonstrated that adults with autism spectrum disorder (ASD) face increased risk for infection and worse health outcomes if infected.^1-5 Although evidence also suggests that children with ASD and other developmental disabilities (DD) have poorer health outcomes associated with COVID-19 infection,^6,7 few studies have evaluated whether youth with ASD and other DD are also at increased risk for initial COVID-19 infection. One study leveraged a large health record dataset from Israel and identified increased odds of infection for autistic youth ages 0 to 16 years, although no significantly increased risk for hospitalization.¹ Another study compared the prevalence of COVID-19 infection among youth age ≤19 years from the South Korean National Health Insurance Service⁸; the overall prevalence of COVID-19 was lower among youth with ASD compared with neurotypical peers, but when stratified by age, it was higher among children with ASD in the 0 to 4 years age group. Although ample evidence suggests that youth with ASD were more vulnerable to pandemic-related disruptions in critical supports such as dental care, health care, and educational and daycare services than youth with other disabilities,^9-11 evidence related to COVID-19 mitigation and infection risk is still lacking in this population.

During the pandemic, the Centers for Disease Control and Prevention (CDC) and other public health entities engaged in widespread education efforts focused on key mitigation behaviors to reduce risk of infection, including handwashing/sanitization, the use of masks that cover the nose and mouth, and physical distancing at least 6 feet from others.^12,13 Some reports suggest that children with ASD often had difficulty adhering to these mitigation strategies, amplifying concerns for possible increased infection risk. In a survey of parents of youth and young adults with ASD, Mutluer et al.¹¹ found that many reported that their child did not understand much about COVID-19 (58%), was not able to follow COVID-19 prevention measures like keeping physical distance and staying at home (45%), and experienced sensory problems when using disinfectants, washing hands, or wearing masks (37%). Similarly, Stadheim et al.¹⁴ found in a qualitative analysis that parents of youth with ASD reported themes of their children struggling to accept the “new normal” of restrictions. Although informative, these and similar studies did not include a comparison group of parents of children without ASD, making it unclear the extent to which difficulties with COVID-19 mitigation efforts affected children with ASD compared with children who were typically developing or with other developmental disabilities.

To address these gaps, the goals of this study were (1) to investigate positive tests for COVID-19 infection reported in a large, well-characterized sample of young children with ASD, with other DD, and those selected as a population comparison group (POP); and (2) to compare the groups regarding prevalence of difficulties experienced with COVID-19 mitigation strategies. Specifically, we sought to examine both individual and combined difficulties with handwashing, masking, and physical distancing, and the key health behavior of nasal swabbing (required for COVID-19 testing).

Data from the Study to Explore Early Development—Phase 3 (SEED3) and the SEED3 COVID-19 Impact Assessment were used for these analyses. SEED3 is a Centers for Disease Control and Prevention (CDC) funded multisite case-control study of the risk factors and behavioral characteristics associated with autism spectrum disorder (ASD) in early childhood. Three groups of children between 2 to 5 years of age were recruited: children diagnosed with ASD, children diagnosed with a DD other than ASD, and children from a population comparison group (POP). Children with ASD and other DD were recruited from clinic and education partners at each study site and children in the POP group were selected randomly from state vital records.

All children received a Social Communication Questionnaire (SCQ) and participated in a developmental evaluation; those who demonstrated ASD risk (SCQ score ≥11) received 2 gold-standard ASD diagnostic measures administered by research-reliable clinicians. After the evaluation, children were placed into 3 study groups based on the results of the developmental evaluation: ASD, other DD, and POP. A complete description of SEED methods and study classification procedures can be found in Schendel et al.¹⁵ and Wiggins et al.¹⁶ SEED3 was reviewed and approved by the CDC and local site Institutional Review Boards (IRBs).

A total of 1027 SEED3 participants (autism spectrum disorder group [ASD] = 274, other DD = 368, population comparison group [POP] = 385) completed the SEED COVID-19 Impact Assessment. Characteristics of the sample are provided in Table 1. A larger percentage of ASD as compared with other DD and POP respondents reported they were from a race/ethnicity group other than non-Hispanic White, caregivers of male study children, and from households with lower household incomes relative to the federal poverty level.

Table 2 contains a summary of survey responses to questions related to COVID-19 mitigation strategies and COVID-19 testing and results. A similar percentage of respondents in the ASD (36%), other DD (39%), and POP groups (36%) (p = 0.580) indicated the study child ever received a COVID-19 test. However, the prevalence of having 1 or more positive test for children in the ASD and other DD groups was approximately 3 times higher (11% and 13%, respectively) than for children in the POP group (<5%) (p = 0.032).

Among the entire sample of children, physical distancing was the most frequently endorsed mitigation challenge (45%) followed by masking (39%) and handwashing (21%). For the question related to COVID-19 nasal swabbing, we restricted the sample to children who had a COVID-19 test and answered yes or no to difficulties with nasal swabbing (n = 365). Of those, 68% endorsed the study child had difficulty with a COVID-19 test nasal swab. A greater percentage of the ASD group endorsed difficulty for each individual mitigation measure and nasal swabbing compared with the other DD and POP groups. In addition, a greater percentage of the ASD group endorsed difficulty with 2 or more of the 3 mitigation measures (Table 2). Over half (53%) of the POP group endorsed no mitigation measure difficulties compared with 42% for the other DD group and 21% for the ASD group. Conversely, 22% of ASD respondents reported difficulties with all 3 mitigation measures compared with 11% among other DD respondents and 5% among POP respondents (Fig. 1).

We calculated both unadjusted and adjusted prevalence ratios (PRs) for difficulty with each mitigation measure and nasal swabbing for ASD and other DD compared with POP (Supplemental Digital Content 1, http://links.lww.com/JDBP/A507, which provides unadjusted and adjusted PRs for mitigation measure difficulty and study group). In adjusted models, controlling for child’s race, sex, household income relative to the Federal Poverty Level, and study site, both ASD and other DD children were more likely to have difficulty with physical distancing and handwashing compared with POP children. Children with ASD (aPR = 2.29 [1.89–2.78]) were more likely to have difficulties with masking compared with children in the POP group. Children with ASD were more likely than POP children to have difficulty with nasal swabs related to COVID-19 testing (aPR = 1.54 [1.26–1.88]) but the other DD and POP groups reported similar levels of difficulty with nasal swabs (Fig. 2).

We believe our data are the first from the United States to document higher prevalence of at least 1 positive COVID-19 test in young children with autism spectrum disorder group (ASD) and other developmental disabilities (DD) compared with the general population, and the first ever examination of comparative COVID-19-positive test prevalence in a large, diverse sample characterized by direct developmental assessment. We found the prevalence of at least 1 positive COVID-19 test during 2020 among young children with ASD and other DDs was approximately 3 times that reported by caregivers in the population comparison group (POP) group. This finding in young children is consistent with the literature in adults suggesting an increased risk for COVID-19 infection with developmental disabilities.³

Our finding of higher prevalence of 1 or more positive COVID-19 test among young children with ASD and other DD provides indirect evidence that children with developmental conditions may be at increased risk for COVID-19 infection, bolstering the importance of recommendations for targeted respiratory illness prevention and mitigation strategies specifically for youth with ASD and other DD.^17,18 Infection prevention in this population is particularly important given the evidence that children with developmental conditions that contract COVID-19 have poorer outcomes,¹⁹ and the importance of helping youth with disabilities maintain access to critical educational and therapeutic settings.^18,20,21

The existing COVD-19 literature generally references infection mitigation challenges experienced by individuals with disabilities^17,22; however, the current analysis is the first to quantitatively demonstrate the challenges experienced with COVID-19 infection mitigation strategies for youth with developmental disabilities, and especially those with ASD. Compared with caregivers in the other DD and POP groups, caregivers of children with ASD were more likely to endorse that their children experienced 2 or more mitigation challenges. In addition to the considerations raised by these data in anticipation of future public health emergencies, our findings also suggest opportunities for supporting the health of youth with ASD and other DD in relation to seasonal illnesses (e.g., COVID-19, influenza, respiratory syncytial virus). Training approaches to improve infection mitigation behaviors like masking that emerged during the early pandemic^22-26 continue to be important in infection prevention, particularly in settings and programs focused on improving health behaviors. Especially for youth with additional risk factors in addition to their developmental disability, routine and anticipatory coaching to improve mitigation adherence among children with ASD and other DD may have substantial benefit. This training could also be embedded in school-based individualized education programs, disseminated in primary care, and through tertiary care developmental disability providers. Increasing awareness of the potential for increased infection risk and heightened disease severity will be a key component to motivate both providers and families to support this type of training. More robust research is needed to ensure that infection mitigation coaching programs are effective and can be successfully implemented in school and community-based settings.

Strengths of our study include the use of a large, diverse sample of youth who were directly evaluated for ASD (vs identified administratively by health records). We also were able to include comparison groups of youth with other DD and youth from the general population. Several limitations should be noted, including the potential selection bias among participants in SEED3 and the subsequent COVID-19 survey. The distributions of variables including race/ethnicity, child’s sex, and family income were not balanced across the 3 groups, and despite adjusting for these variables in our models, it is possible that residual confounding could have affected our findings.

Although other large-scale reports on COVID-19 use similar methodologies, another limitation is the reliance on caregiver report for the infection and mitigation variables used in this analysis. Respondents were asked about events and behavior that occurred up to a year before completing the survey, which could diminish accuracy of recall. In addition, respondents could only respond “yes” or “no” to the mitigation difficulty questions. Future investigations should seek more information about the contexts and extent of the experienced difficulties and ideally collect data contemporaneously to minimize recall bias. Relatively few respondents endorsed at least 1 positive COVID-19 test for the study child, which prevented us from adjusting the association between at least 1 positive COVID-19 test and study group for potential confounding variables. Finally, we did not have data on the total number of COVID-19 tests administered per child, and therefore, we could not calculate and compare cumulative COVID-19 test positivity or test positivity rates across the 3 groups. The percentage of caregivers reporting that their child received at least 1 COVID-19 test was similar (36%–39%) for all 3 groups, but this might obscure differences in the total number of tests used, reflect the availability of tests when data were collected, and/or be a function of other unmeasured factors.

In summary, in a large-scale examination of COVID-19 infection for young children with autism spectrum disorder (ASD) group and other developmental disabilities (DD) in the United States, we found that the prevalence of having at least 1 positive COVID-19 test was 3 times higher for young children with ASD or other DD compared to the POP group. We also found that children with DD and ASD had a significantly higher prevalence of difficulties with key infection mitigation strategies compared with children from the general population. Health care providers, educational professionals, and policy makers can maintain awareness of these disparities and pursue opportunities to accommodate and support youth with disabilities and their families to improve mitigation effectiveness.