Does area level socioeconomic status modify the effects of PM₁₀ on preterm delivery?

Background: Air pollution has been one of the focal points for investigations of maternal health and birth, and an association has been found between air pollution and birth outcomes. Some individual-level risk factors of mothers have been identified as strong predictors of preterm delivery (PTD). Recent literature has shown that the neighborhood environment of the mother and child has an independent influence on birth outcomes that was not explained by individual-level risk factors. Objective: This study explored the effects of maternal exposure to particulate air pollution (PM₁₀) and its relationship to the probability of preterm delivery (PTD) in Seoul, Korea. Our hypothesis was that PM₁₀ increased the probability of PTD, and that the area level socioeconomic status (SES), measured by monthly average household income, modified the PM₁₀ effects on PTD after controlling for individual-level risk factors such as individual-level SES. Data and Method: Birth data from the Korean National Statistics Office consisted of records on 433,173 singleton births in the years 2000-2003, and included individual characteristics. Area-level characteristics were also controlled for in 25 administrative regions in Seoul. Maternal exposure to PM₁₀ was calculated at an individual level for each trimester using information from monitoring sites in residential areas. Hourly data on PM₁₀ for 1999-2003 were acquired from the National Institute of Environment Research. Birth outcome was dichotomously coded: term birth (gestation ≥37 weeks) or PTD (gestation <37 weeks). For multilevel analysis, we conducted a hierarchical logistic regression with a random intercept for each administrative area to account for the hierarchical nature of the data. Results: We found that area-level SES modified the PM₁₀ effects on PTD, specifically among those who were in low SES areas. Results of multilevel analyses showed that an increase in PM₁₀ by 10 μg/m³ led to a statistically significant increase in the probability of PTD [3.12% (95% CI, 0.17-6.15)] during the 2nd trimester in the low-income group. Analyses using the stratified population found a 3.4% increase (95% CI, 0.31-6.58) in the probability of PTD during the 2nd trimester in areas with low income. Conclusion: Mothers exposed to higher PM₁₀ in Seoul, Korea, had a higher probability of PTD. Individual-level characteristics had different effects on the probability of PTD depending on the area-level SES. Area-level SES modified the effect of PM₁₀ concentration on PTD.