Portal de Periódicos da CAPES

496: Crime and property vacancy are not risk factors of preterm birth in Philadelphia, Pennsylvania

2012; Elsevier BV; Volume: 208; Issue: 1 Linguagem: Inglês

10.1016/j.ajog.2012.10.662

1097-6868

Jamie Bastek, Mary D. Sammel, Sindhu K. Srinivas, Tara Jackson, Meghan A. McShea, Meghan E. Ryan, Michal A. Elovitz,

Preterm Birth and Chorioamnionitis

While there are undoubtedly some genetic predispositions to preterm birth (PTB), recent GWAS studies have been relatively unrevealing as to specific genetic risks for PTB. This argues that environmental factors may be important modifiers of PTB. Our objective was to determine whether census-level crime and property vacancy data are associated with PTB in a large diverse city with a high rate of PTB. A prospective cohort of women with singleton pregnancies with preterm labor (2009-11) (N=598) was used for this study. Individual level data from the primary cohort including home address zip code were merged with crime and property vacancy census data (2005-09). Associations between PTB <37 weeks with individual and census level data were determined using chi square tests, and generalized estimating equations to account for clustering of participants within zip codes. Areas under the receiver-operator curve (AUC) were calculated and compared. Individual study-level risk factors were significantly associated with PTB (TABLE). The most parsimonious model of study-level variables to predict PTB included GA at PTL, race, prior PTB, cocaine, and PTL admission (AUC 0.85). Racial demographics of study participants varied significantly by zip code (p<0.001), and zip codes with the most study-level PTBs were more predominantly black according to census-level data (FIGURE). However, when census-level data were studied both individually and through factor analysis, neither zip code, nor census-level crime nor property vacancy data predicted study-level PTB outcomes.Tabled 1Individual-level risk factors of PTBStudy-level PTB count in Philadelphia by zip codeView Large Image Figure ViewerDownload Hi-res image Download (PPT) Our success in decreasing the PTB rate on a population level may depend on our ability to identify potentially modifiable environmental risk factors of PTB. Despite significant racial differences, census-level crime and property vacancy data do not appear informative on an individual level. Other census-level economic, educational, and social capital data may prove to be important modifiers of PTB (K12HD001265, MOD#21FY08-539).