Modeling Low Birth Weight Incidence in the State of Georgia Using Spatial Regression

Low birth weight (LBW) is a prevalent public health issue in the state of Georgia and the country. This condition is a significant predictor of infant mortality and associated with numerous developmental abnormalities of affected children. Previous studies have documented a strong positive association between an increase risk of LBW and many factors including socioeconomic status, behavioral patterns, demographic characteristics, and environmental conditions. Furthermore, these risk factors occur at both the individual-level and the neighborhood-level and complex interactions between risk factors at different levels are reported. Using the birth data in Georgia for the year 2000, two-level logistical models are built to investigate the association between LBW risk with explanatory variables at both the individual level and the county level. These models are also used to examine how/to what extent county-level variables translate into differences in individual-level risk. A multi-step sequential modeling strategy is adopted with complexity being increased in every successive model.