Socioeconomic movement between successive deliveries: impact on preterm delivery in a diverse racial and ethnic population

BACKGROUND: Numerous studies have shown that women who live in more deprived neighborhoods are at higher risk of preterm delivery than those in wealthier neighborhoods, even after controlling for individual socioeconomic status. Since socioeconomic status and personal characteristics affect the choice of neighborhood of residence, these studies are likely to be confounded by unmeasured differences between the woman living in the better and worse neighborhoods. This dissertation has attempted to examine the effect of neighborhood socioeconomic status (NSES) using a better study design by following up women who moved between successive deliveries. METHODS: All newborns to the same mother in the Electronic Birth Certificate database were linked to create a longitudinal dataset. Neighborhood information was obtained from Census 2000 to calculate a neighborhood deprivation score. Quintiles of this score formed the socioeconomic neighborhood strata, the highest score being most deprived. Lowest three quintiles were combined to form one neighborhood stratum the other two quintile formed two more strata. Logistic regression was used to estimate the change in risk of preterm delivery associated with change in NSES for women who moved between successive pregnancies. Stratified Analysis by race/ethnicity was also done. RESULTS: Moving to a worse socioeconomic neighborhood increased the risk of preterm delivery but moving to a better neighborhood did not reduce the risk, when comparing women who moved to better or worse neighborhood to those who remained in the same neighborhood. On stratifying by race/ethnicity this was observed for Non-Hispanic (NH) Blacks only. However, when comparing a woman after she relocated to her outcome before relocating, the change of neighborhood made no difference. On stratifying by race/ ethnicity, NH Blacks alone had an increase in risk of preterm delivery for those who moved to a worse neighborhood. Similarly teenagers who moved to worse neighborhoods were also at increased risk of preterm delivery. CONCLUSION: Neighborhood effect on preterm delivery is very small in this cohort. Improvements in the externalities of neighborhood environments would be unlikely to have a short term effect on preterm birth rates. Nevertheless there is a suggestion in this data that at-risk women may do worse in deprived neighborhoods than they would do in more favorable circumstances.Ph.D.Includes bibliographical referencesby Alice Davi