%A Donna Charissa Almario Doebler
%T Understanding racial disparities in low birthweight in Pittsburgh, Pennsylvania: The role of area-level socioeconomic position and individual-level factors
%X Background: Low birthweight (LBW, <2500g) is a leading cause of infant mortality, and disparities exist between Blacks and Whites. About 11% of Pittsburgh births in 2003 were LBW, and the racial difference was wide: 8.4% of LBW infants were born to Whites, whereas 16.0% were born to Blacks. Studies suggest an association between contextual factors and LBW?lower levels of area-level socioeconomic position (SEP) are associated with increased LBW risk. The dissertation's main research hypotheses are whether 1) area-level SEP predicts LBW, 2) racial difference in LBW is partially explained by area-level SEP, and 3) racial difference is explained after controlling for area-level SEP and individual-level factors.Methods: Using U.S. Census 2000 data, area-level SEP measures were created for Pittsburgh: overall neighborhood disadvantage (ONDijk), material and economic deprivation (MEDij), and concentrated disadvantage (CDij). LBW and other individual-level data from 10,830 birth records were obtained from the 2003-2006 Allegheny County birth registry. Multilevel logistic regression was utilized to examine the association between SEP measures and LBW. Results: ONDijk was a significant predictor of LBW (OR: 1.306, p<0.001), remained significant after controlling for race (OR: 1.10, p<0.03), but was no longer significant after controlling for individual-level disadvantage (OR: 1.05, p=0.27). In addition, 74% of Blacks resided in disadvantaged neighborhoods, compared to 13% of Whites. In the unadjusted race model, Blacks were at increased odds of LBW compared to Whites (OR: 2.119, p<0.001), and the race OR decreased after adjusting for ONDijk (OR: 1.917, p<0.001) and individual-level disadvantage (OR: 1.56, p<0.001). Due to the lack of variability of LBW at the block group level, there was insufficient power to test the association between LBW and CDij and MEDij. Conclusions: Findings suggest that contextual factors are associated with LBW: knowing one's race and neighborhood may help predict one's risk for LBW. Public health significance includes using ONDijk as an indicator of areas with higher levels of LBW risk and targeting these neighborhoods for interventions to improve birth outcomes. In addition, understanding racial differences in neighborhood conditions may help further understand the social determinants that contribute to health disparities in LBW between Blacks and Whites.
%D 2010
%K  mapping; neighborhood disadvantage; low birthweight; multilevel modeling; Pittsburgh
%I University of Pittsburgh
%L pittir6868