Are associations between gestational exposure of air pollution and preterm birth modified by neighborhood deprivation in a North Carolina birth cohort?
On this page:
Background/Aims
Exposure to air pollutants is associated with preterm birth (PTB, <37 weeks completed gestation); however, there is limited research on whether these associations are modified by neighborhood-level risk factors. We evaluated whether associations between gestational exposure to fine particulate matter (PM2.5), nitrogen dioxide (NO2), and ozone (O3) and PTB were modified by neighborhood deprivation in a North Carolina birth cohort from 2011-2015.
Methods
Modeled daily PM2.5, NO2, and O3 concentrations were aggregated to census tract, linked to residential address at delivery, and averaged across each trimester of pregnancy. The Neighborhood Deprivation Index (NDI) was created using principal component analysis with census variables representing income/poverty, education, employment, housing, and occupation; with NDI dichotomized into high and low deprivation areas. Modified Poisson regression models were used to estimate risk differences (RDs) per 10,000 births stratified by NDI categories, adjusted for gestational parent marital status, race/ethnicity, age at delivery, Medicaid status, and month of conception.
Results
Of the 543,086 births, 7.7% were PTB. The daily median (IQR) concentration was 9.1 (1.2) µg/m3 PM2.5, 14.0 (7.4) ppb for NO2, and 40.4 (4.2) ppb for O3. The RDs for PM2.5 and PTB ranged from -17 (95% CI:-25,9) to 1 (-6,8) across NDI strata, and did not show evidence of modification. RDs for NO2 were generally null and did not differ by NDI stratum. The O3 RDs were small but there was some evidence of separation for first [4 (1,7) and 0 (-2, 2) in low and high NDI] and third [1 (-3, 5) and 5 (2, 7)] trimesters across strata.
Conclusion
Associations with air pollutant exposure and PTB were generally null across trimesters, but there was some evidence of effect modification by NDI with first and third trimester O3 exposures. While changes were small, it is important to consider social context of environmental exposures.