Abstract
BACKGROUND: Billions of dollars have been spent implementing regulations to reduce traffic-related air pollution (TRAP) from exhaust pipe emissions. However, few health studies have evaluated the change in TRAP emissions and associations with infant health outcomes. We hypothesize that the magnitude of association between vehicle exposure measures and adverse birth outcomes has decreased over time, parallelling regulatory improvements in exhaust pipe emissions. METHODS: Using birth records in Texas from 1996 to 2016, we calculated residential exposure measures related to TRAP: nitrogen dioxide (NO2, a marker of the TRAP mixture), vehicle miles travelled within 500 m of homes (VMT500), a measure of traffic volume, and highway proximity. Using an accountability study framework, our analysis examined term birthweight, term low birthweight (TLBW) (<2500 g), preterm birth (PTB) (<37 weeks) and very preterm birth (VPTB) (<32 weeks). We implemented linear and logistic regression models to examine overall and time-stratified associations, including trends by race/ethnicity and socioeconomic groups. RESULTS: Among exposures for 6 158 518 births, NO2 exposures decreased 59% over time but VMT500 remained relatively stable. TRAP-related exposure measures were persistently associated with harmful birth outcomes [e.g. OR1996-2016 of 1.07 (95% CI: 1.04, 1.08) for TLBW comparing the highest vs lowest NO2 quintile]. The magnitude of associations decreased for total VMT500 and TLBW (-60%, OR1996: 1.08 to OR2016: 1.03 for the highest vs lowest quintile) and PTB (-65%) and VTPT (-61%), but not for term birthweight. CONCLUSIONS: We observed evidence of small improvements in birth outcomes associated with reductions in exhaust pipe emissions over a 20-year period in Texas.