Maternal COVID-19 in pregnancy not linked to ADHD or autism in 108,148 Japanese children
In a Japanese claims-database cohort, children of mothers with COVID-19 in pregnancy showed no higher rate of ADHD, autism or intellectual disability diagnoses over the available follow-up.
Yusuke Okubo and Kazuhiro Uda examined whether COVID-19 in a mother during pregnancy is associated with neurodevelopmental diagnoses in her child. According to the study abstract on ovid.com, it is a nationwide birth cohort of 108,148 singleton children in Japan, contributing 224,965 person-years of follow-up in an administrative claims database for 2020 to 2024. Maternal COVID-19 during pregnancy was recorded for 7797 children (7.2%) ovid.com
ovid.comovid.comOpen source ↗. The work was published in The Pediatric Infectious Disease Journal.
Design and cohort
The outcomes were clinically diagnosed attention-deficit/hyperactivity disorder (ADHD), autism spectrum disorder (ASD), intellectual disability (ID) and a composite of the three. Diagnoses were identified from diagnostic codes and relevant prescription records. The primary analysis used Cox proportional hazards models with time-dependent propensity score matching.
Matching produced 7679 pairs of exposed and unexposed children, contributing 22,360 person-years, with well-balanced covariates. Hospitalization for COVID-19 was documented in 939 of the 7797 exposed pregnancies (12.0%). Before matching, mothers with COVID-19 had higher rates of infection, fever and asthma during pregnancy; mean maternal age was 32.7 years in the exposed group and 32.5 years in the unexposed group.
Results and limitations
In the matched analysis, maternal COVID-19 was not associated with the composite outcome (hazard ratio [HR] 0.96; 95% confidence interval [CI] 0.56–1.66), ADHD (HR 0.75; 95% CI 0.32–1.78), ASD (HR 0.79; 95% CI 0.40–1.55) or ID (HR 1.06; 95% CI 0.55–2.01) ovid.com
ovid.comovid.comOpen source ↗. For the composite outcome, no association was found for infection in the first (HR 1.00; 95% CI 0.20–4.96), second (HR 1.00; 95% CI 0.40–2.52) or third trimester (HR 0.93; 95% CI 0.44–1.98).
Multivariable-adjusted models gave similar results, with an HR of 0.89 (95% CI 0.61–1.29) for the composite outcome. A sibling comparison of 2278 children also showed no association (composite HR 0.86; 95% CI 0.26–2.82). A sensitivity analysis with stricter diagnostic criteria gave a composite HR of 0.73 (95% CI 0.29–1.81).
The authors report no evidence of increased risk during the available follow-up. They list residual confounding, possible misclassification of exposure or outcome in claims-based diagnoses, and exclusion of pregnancy losses as limitations. The database mainly covers employees of medium- and large-sized companies and their dependents, which may limit generalizability. The authors note that longer follow-up remains important, because some neurodevelopmental conditions may become evident later in childhood.
