Maternal Exposure to Ambient Ozone and Birthweight Reduction During COVID-19 Epidemic in East China: A Sex- and Window-Specific Retrospective Cohort Study

Background: Low birthweight (LBW) remains a major global public health challenge, with long-term consequences for neonatal survival and chronic disease risk. Maternal exposure to ambient ozone (O3) has been implicated in fetal growth restriction, but evidence remains inconsistent. Methods: We conducted a retrospective cohort study of 14,381 term singleton live births in Zhejiang Province, East China, during the COVID-19 lockdown and quarantine period. Maternal O3 exposure during each trimester and the entire pregnancy was estimated using data from nearest atmospheric monitoring stations, with refined time-activity patterns accounting for home and workplace locations. Results: In multivariable linear regression adjusted for the pre-specified covariates, maternal O3 exposure averaged over the entire pregnancy was inversely associated with birthweight (pooled estimate across sex- and residence-stratified models: β = −2.1 g per μg/m3, 95% CI: −3.6 to −0.5; stratum-specific β ranged from −2.24 to −0.92). An interquartile range (IQR) of approximately 10–15 μg/m3 in entire-pregnancy O3 exposure increase was associated with an estimated birthweight reduction of about 20–33 g; the range reflected the combination of stratum-specific β estimates and stratum-specific IQRs. The explained variance was small as indicated by R2, showing that O3 was not a powerful individual-level predictor of birthweight; the precisely estimated β nonetheless reflected a systematic shift in the population distribution of birthweight. Conclusions: Maternal O3 exposure during pregnancy was associated with reduced birthweight in a sex- and window-specific manner. The third trimester was more stronger association period for female fetuses, whereas cumulative exposure across mid-to-whole pregnancy was more critical for males. Maternal nutritional intake is a plausible modifiable factor that may attenuate susceptibility, but the present study did not measure diet, food insecurity, or psychological stress. Future studies should incorporate dietary assessment and biomarkers to test this hypothesis.