Objective: We have conducted a study to assess the role of environment on the burden of maternal morbidities and mortalities among women using an external exposome approach for the purpose of developing targeted public health interventions to decrease disparities.
Methods: We identified counties in the 48 contiguous USA where observed low birthweight (LBW) rates were higher than expected during a five-year study period. The identification was conducted using a retrospective space-time analysis scan for statistically significant clusters with high or low rates by a Discrete Poisson Model.
Results: We observed statistically significant associations of LBW rate with a set of predictive variables. However, in one of the two spatiotemporal models we discovered LBW to be associated with five predictive variables (teen birth rate, adult obesity, uninsured adults, physically unhealthy days, and percent of adults who smoke) in two counties situated in Alabama after adjusting for location changes. Counties with higher than expected LBW rates were similarly associated with two environmental variables (ozone and fine particulate matter).
Conclusions: The county-level predictive measures of LBW offer new insights into spatiotemporal patterns relative to key contributory factors. An external framework provides a promising place-based approach for identifying "hotspots" with implications for designing targeted interventions and control measures to reduce and eliminate health disparities.
Keywords: discrete Poisson model; external exposome; geographic information systems; health disparities; infant mortality; low birthweight; maternal mortality; pregnancy outcomes; premature birth.