Association of ambient particulate matter exposure with preterm birth and gestational age in a retrospective sibling-matched Chinese cohort

The association of perinatal exposure to airborne particulate matter (PM) with preterm birth (PTB) remains unclear. This study aimed to investigate possible associations between PM exposure and risk of PTB. There were 61,620 sibling pairs from a retrospective sibling matched cohort study were included. PM exposure was estimated by space-time extremely randomized trees models. Within-group differences in 2.5 and 10 μm particulate (ΔPM_2.5 and ΔPM₁₀) maternal exposure levels between corresponding trimesters of two consecutive pregnancies were tested for associations with moderate, very, or all PTB using fixed-effect regression models. Gestational age and difference in gestational age (Δ gestational age) between siblings were treated as secondary outcomes. A total of 1301 (2.1 %) incident PTB cases were confirmed in our study. The exposure-response analyses suggested that each 10 μg/m³ increase in ΔPM_2.5 or ΔPM₁₀ were associated with the increased risk of PTB, with significant effect in first trimester (ΔPM_2.5 adjusted OR = 1.375, 95 % CI: 1.141, 1.658; ΔPM₁₀ adjusted OR = 1.241, 95 %CI: 1.073, 1.435). Increase in ΔPM_2.5 or ΔPM₁₀ in first, second, and third trimester were significantly associated with a shortened gestational age. PTB risk and gestational age shared a near linear exposure-response relationship with increased ΔPM_2.5 and ΔPM₁₀ in whole pregnancy. In conclusion, increased ambient PM_2.5 exposure is positively associated with reduced gestational age, while increased PM_2.5 and PM₁₀ concentration is associated with PTB risk.