You are here:
MODELING BLACK-WHITE PRETERM BIRTH DISPARITY: ECOLOGIC AND MULTILEVEL MODELS
Citation:
MESSER, L. C. AND J. S. KAUFMAN. MODELING BLACK-WHITE PRETERM BIRTH DISPARITY: ECOLOGIC AND MULTILEVEL MODELS. Presented at International Conference on Urban Health, Toronto, ON, CANADA, October 26 - 28, 2005.
Description:
INTRODUCTION. The disparity between black and white women's adverse birth outcomes has been subject to much investigation, yet the factors underlying its persistence remain elusive, which has encouraged research on neighborhood-level influences. This work considers two main questions: 1) to what extent are neighborhood-level factors (violent crime, deprivation, suboptimal housing) associated with black-white (B-W) preterm birth (PTB) disparity? 2) Is neighborhood-level B-W PTB disparity associated with PTB risk for the women residing in these neighborhoods? METHODS. The authors examine B-W PTB disparity in Raleigh NC census tracts (1999-2001) using linear and logistic models. Geocoded vital records, city crime reports and US Census data were analyzed. PTB was defined as birth at < 37weeks (& < 3888g) gestational age. B-W disparity was defined as the difference in proportions between black PTB and white PTB per tract. A 10-item standardized index estimated neighborhood deprivation. Tertiles of deprivation, suboptimal housing and violent crime were modeled. Race-stratified random effects logistic models with fixed slope predictor values and randomly distributed tract-specific intercepts were fit; analyses were restricted to tracts with >50 black and white births. B-W PTB disparity was scaled so each 0.1 unit increase in disparity corresponded to a 1% increase in PTB. RESULTS. Black and white non-Hispanic women live in different neighborhood environments. Black non-Hispanic women live in tracts with more crime (mean=67 crimes, standard deviation [sd] 49) and deprivation (mean=1.4, sd 3.5), than white non-Hispanic women (mean=29 crimes, sd 35; mean deprivation= -1.1, sd 1.3). At the ecologic-level, high deprivation (third tertile) (difference =0.07; 95%CI: 0.00, 0.15), suboptimal housing (beta=0.10; 95% CI: 0.03, 0.16) and violent crime (beta=0.07; 95%CI: 0.01, 0.14) were associated with increased B-W disparity after adjustment for tract maternal age, education and marital status structures, compared with the first tertiles. Multilevel models adjusted for maternal age, education and tract deprivation found B-W PTB disparity associated with decreased preterm birth odds in white women (Odds Ratio [OR] = 0.98; 95% Confidence Interval [CI]: 0.98, 0.99) and increased odds in black women (OR=1.05; 95% CI: 1.03, 1.06). CONCLUSIONS. The neighborhood environment may influence preterm birth disparity. Exposure to preterm birth disparity at the tract-level appears differentially associated with preterm birth odds. IMPLICATIONS. PTB disproportionately affects disadvantaged populations in the U.S.; understanding how exposure to PTB disparity influences individual risk is important for planning effective policy interventions.