Changes in the spatial pattern of low birth weight in a southern California county: the role of individual and neighborhood level factors

Individual-level maternal risk factors have been able to explain only a small proportion of low birth weight (LBW) births in the US to date and neighborhood-level factors have not been recognized as important predictors of reproductive outcomes. At the US/Mexico border, tremendous demographic changes have taken place between 1980 and 1990. Whether high population growth, economic pressures, and community instability have affected reproductive health in this region is not known. We used spatial epidemiologic methods to investigate the importance of changes in neighborhood and individual-level risk factors in predicting changes in two measures of adverse reproductive health: term and preterm LBW. Using approximately 16,000 births in 1980 and 24,000 births in 1990, we produced a continuous surface of LBW change (Z scores for difference in proportions) for southern San Diego County, CA, and identified statistically significant "hot spots" of elevated risk. Using a geographic information system, data from the 1980 and 1990 US Census were smoothed to link neighborhood-level variables with individual-level data from the 1980 and 1990 California birth certificates. Multivariate regression models were developed to identify individual- and neighborhood-level variables of 1980-1990 changes which were predictive of Z scores of LBW change. Results of conditional autoregressive models were used to assess effects of spatial autocorrelation. The continuous surface of LBW identified areas with statistically significant increases (17 areas for term LBW and 14 areas for preterm LBW). While individual- and neighborhood-level variables explained about an equal amount of variance in term LBW, only neighborhood-level variables were significant predictors of preterm LBW. While changes in maternal race/ethnicity composition and a measure of stability were related to term LBW, measures of affluence were related to preterm LBW. This study highlights the importance of socioeconomic and demographic changes in a woman's neighborhood on reproductive health.