Examining individual and neighborhood-level risk factors for delivering a preterm infant.

摘要

This dissertation examines the relationship between several neighborhood and individual-level factors and the probability of delivering a preterm infant. Infants born prematurely, or at less than 37 weeks of gestation, have a greater risk of long-term developmental disabilities, such as cerebral palsy, than infants born at full term. This research draws on various perspectives on the social causes of health to illustrate pathways that may account for neighborhood socioeconomic and racial differences in preterm birth. I hypothesize that neighborhood concentrated disadvantage influences preterm birth indirectly through smoking, medical risk, and perceived neighborhood support. Traditionally, the birth outcome literature has focused on individual-level risk factors. Neighborhood concentrated disadvantage is expected to affect preterm birth, in part, by patterning exposure to stress, which can compromise overall health by increasing the likelihood of having a medical risk, such as hypertension, or smoking during pregnancy. In turn, such illnesses and smoking can lead to the onset of preterm labor. As a community-level social resource, perceived neighborhood support is expected to function by buffering the impact of adverse neighborhood conditions. Living in a community where residents believe that their neighbors are supportive and can be depended upon in times of need should reduce the likelihood of delivering preterm, especially in disadvantaged communities. Maternal black race and older maternal age---also included in the model---are hypothesized to influence preterm birth indirectly through medical risk.;This study made use of data from the U.S. Census and Ohio birth certificate files as well as the Greater Cincinnati Community Health Status Survey. The path model's dependent variables were binary or continuous and information was combined from two-level logistic and linear regression analysis to compute a comparable metric for the model, which allowed for a comparison of the relative strength of the predictors. Neighborhood concentrated disadvantage had significant indirect effects on preterm birth through medical risk and smoking, but not perceived neighborhood support. An indirect relationship between black maternal race (compared to white) and preterm birth---through medical risk---was also found, and the overall effect of black maternal race on preterm birth was suppressed somewhat by smoking since black mothers were less likely than white mothers to smoke while pregnant. Medical risk during pregnancy operated as a causal mechanism linking older maternal age (e.g., 35 years of age and older) with an increased likelihood of delivering preterm. Although statistically significant, the indirect effects were relatively weak. Compared to other predictors, neighborhood concentrated disadvantage had the strongest effect on preterm birth, medical risk, and maternal smoking---the three individual-level outcomes in this study. This dissertation extends the literature on neighborhood effects and birth outcomes by providing support for causal mechanisms linking neighborhood socioeconomic status, and the findings demonstrate the need to contextualize individual-level risk factors for preterm birth. Given the relatively weak indirect effects, however, this research also demonstrates that these multilevel relationships are complex and additional theoretical and empirical work is needed to better specify the causal mechanisms.