Viral infections during pregnancy and exposure to maternal stress can have adverse effects on a child, including an increased risk of autism. Healthcare professionals have been concerned that in utero exposure to maternal COVID-19 disease, as well as the stresses of the pandemic, could result in this effect. In a new study published in JAMA Network Open, physicians at NewYork-Presbyterian and Columbia have shown for the first time that children born during the first year of the pandemic were no more likely to screen positive for autism than unexposed or pre-pandemic-born children.
The study is based on a treasure trove of data from NewYork-Presbyterian's electronic health records (EHR) as well as Columbia's COVID-19 Mother Baby Outcomes (COMBO) Initiative. Below, Dani Dumitriu, MD, PhD, a pediatrician at NewYork-Presbyterian and associate professor of pediatrics and psychiatry at Columbia, chair of the COMBO Initiative, and the study's senior author, elaborates on the implications of this pivotal study’s findings.
Research Background
We know very well from other viral exposures during pregnancy that they can lead to numerous adverse child outcomes. There are viruses that can pass from the mother to the baby if the mother is infected while she's pregnant. We learned early in the pandemic that this was not happening too often with SARS-CoV-2. But there are also many viruses with indirect effects, meaning the mother makes various inflammatory proteins and other compounds that can pass through the placenta and affect the baby.
Just the maternal stress experience during the pandemic could also have had adverse effects. This pandemic therefore presented a unique two-pronged mechanism for increasing neurodevelopmental risk, including risk of autism. This has been a worry of ours and a question for the last four-and-a-half years, and one we sought to answer through this study.
Research Methods
We examined data from the Modified Checklist for Autism in Toddlers, Revised (M-CHAT-R), a screening tool that asks a parent or caregiver about their toddler’s behavior, to assess autism risk. Serendipitously, all NewYork-Presbyterian and Columbia clinics started universal M-CHAT-R screening on February 1, 2020.
For this study, we obtained M-CHAT-R scores from children aged 16 to 30 months during routine clinical care at NewYork-Presbyterian and Columbia. The data were abstracted from EHRs for children born between January 2018 and September 2021 (COMBO-EHR cohort). Separately, the M-CHAT-R was administered at 18 months for children born between February 2020 and September 2021 through a prospective longitudinal study (COMBO-RSCH cohort). Prenatal pandemic exposure (birth after March 1, 2020) and maternal SARS-CoV-2 status during pregnancy were determined through EHRs, and data were analyzed from March 2022 to June 2024.
Prenatal pandemic exposure was not associated with a higher positive M-CHAT-R screening rate in either cohort, indicating that prenatal exposure to the COVID-19 pandemic did not increase the likelihood for those children to screen positive for autism.
— Dr. Dani Dumitriu
Key Findings
There were 1,664 children in the COMBO-EHR cohort and 385 in the COMBO-RSCH cohort. Prenatal pandemic exposure was not associated with a higher positive M-CHAT-R screening rate in either cohort, indicating that prenatal exposure to the COVID-19 pandemic did not increase the likelihood for those children to screen positive for autism.
Prenatal exposure to maternal SARS-CoV-2 infection was associated with a lower rate of M-CHAT-R positivity in the COMBO-EHR cohort. This finding does not mean that SARS-CoV-2 exposure is neuroprotective for babies. Rather, we believe this may be attributable to reporting bias. Mothers who did not have COVID-19 may have experienced higher stress — due to the constant worry of getting sick and the vigilance around preventing infection — and may have been more likely to report concerning child behaviors.
Future Implications
In this study, we did not see any increased risk of screening positive for autism with either in utero exposure to COVID-19 disease or to the pandemic environment. Across the field, doctors and scientists have been hypothesizing that we would see an increased risk of autism. Autism is highly responsive to early intervention, starting around age three-to-five. That’s partly why it was important to us to gather data very quickly: in case there was a worldwide need to implement wide-scale screening and intervention strategies. But as of now, we don’t seem to need those, so this is really good news. Physicians can reassure their patients that pandemic exposures during pregnancy are not likely to play a role in autism.
It's very important to note that M-CHAT-R is a screening tool and is not the same as making an autism diagnosis. It is used widely, not just in the United States but in many countries, to screen for behavioral aspects or symptoms of autism in children. Twenty questions are asked of the parent or caregiver and if the parent states there are at least two concerning behaviors, it places the child at higher risk of eventually having an autism diagnosis. But of all the children who screen positive on the M-CHAT-R, only a quarter or less actually receive an autism diagnosis later on. There's still a chance that we will see a slight difference in autism diagnoses in children born during the pandemic once we get enough data, but the current results indicate this is unlikely to be a huge difference.
This study could not have been conducted anywhere else other than NewYork-Presbyterian. We instituted universal M-CHAT-R screening shortly before the pandemic began as well as universal COVID-19 testing for all mothers admitted for labor and delivery starting in March 2020. Between our EHR and COMBO Initiative data, we have an unprecedented cohort of over 2,000 children. These data simply don't exist anywhere else in the world.
Various disorders have different developmental windows at which they appear, and some may not be diagnosed for decades. We will continue to monitor these children as they age into time points when autism, and other neurodevelopmental outcomes, are actually diagnosed. This pandemic has led to worse mental health for pregnant and postpartum individuals, and the social, economic, and educational system impacts of the pandemic may continue to influence developmental trajectories, with intergenerational consequences that could exacerbate existing inequalities, especially ones experienced by marginalized communities during the pandemic. Our aim is for the COMBO cohort to continue into adulthood so we can continue to learn about the effects of in utero exposure of these children on the development of other disorders and diseases later in life.
