Integrating spatiotemporal behavior, indoor-outdoor penetration, and ventilation rates to assess prenatal PM2.5 exposure and the association with birth weight

Abstract

Previous studies that evaluated the association of PM_2.5 with birth outcomes usually assessed personal exposure as outdoor PM_2.5 concentrations of home address (home-based exposure), overlooking factors such as individual spatiotemporal activities, which may result in exposure error. In a prospective birth cohort conducted in Guangzhou, China during 2017–2020, personal PM_2.5 exposure assessment was updated. We incorporated spatiotemporal activities into the exposure assessment by estimating PM_2.5 exposure for each activity based on its specific location and duration. Additionally, an infiltration factor was applied to estimate indoor-outdoor penetration, and ventilation rates (different age groups and activity levels) were used to better adjust individual exposure levels. Logistic regression and distributed lag non-liner model with Cox proportional hazard model were used to assess the associations of prenatal PM_2.5 exposure with low birth weight (LBW) and small for gestational age at a trimester and weekly level, respectively. Updated personal PM_2.5 exposure was lower than the home-based PM_2.5. Per interquartile range increase in PM_2.5 during the third trimester was associated with increased risk of LBW, with ORs (95 % CIs) was 2.17 (1.14–4.14) for updated personal exposure and 2.30 (1.17–4.55) for home-based exposure. Updated personal PM_2.5 in the 6th-7th, home-based PM_2.5 in the 5th-7th, and both PM_2.5 exposure in the 35th week later was associated with LBW. Our findings suggest that spatiotemporal activities, indoor-outdoor penetration, ventilation rate should be taken into account of exposure assessment, otherwise PM_2.5 exposure and the association with adverse birth outcomes may be overestimated.