Background There is increasing evidence of the role of arsenic in the etiology of adverse human reproductive outcomes. Since drinking water can be a major source of arsenic to pregnant women, the effect of arsenic exposure through drinking water on human birth may be revealed by a geospatial association between arsenic concentration in groundwater and birth problems, particularly in a region where private wells substantially account for water supply, like New Hampshire, US. Methods We calculated town-level rates of preterm birth and term low birth weight (term LBW) for New Hampshire, using data for 1997-2009 and stratified by maternal age. We smoothed the rates using a locally-weighted averaging method to increase the statistical stability. The town-level groundwater arsenic values are from three GIS data layers generated by the US Geological Survey: probability of local groundwater arsenic concentration > 1 μg/L, probability > 5 μg/L, and probability > 10 μg/L. We calculated Pearson's correlation coefficients (r) between the reproductive outcomes (preterm birth and term LBW) and the arsenic values, at both state and county levels. Results For preterm birth, younger mothers (maternal age < 20) have a statewide r = 0.70 between the rates smoothed with a threshold = 2,000 births and the town mean arsenic level based on the data of probability > 10 μg/L; For older mothers, r = 0.19 when the smoothing threshold = 3,500; A majority of county level r values are positive based on the arsenic data of probability > 10 μg/L. For term LBW, younger mothers (maternal age < 25) have a statewide r = 0.44 between the rates smoothed with a threshold = 3,500 and town minimum arsenic level based on the data of probability > 1 μg/L; For older mothers, r = 0.14 when the rates are smoothed with a threshold = 1,000 births and also adjusted by town median household income in 1999, and the arsenic values are the town minimum based on probability > 10 μg/L. At the county level, for younger mothers positive r values prevail, but for older mothers it is a mix. For both birth problems, the several most populous counties - with 60-80% of the state's population and clustering at the southwest corner of the state – are largely consistent in having a positive r across different smoothing thresholds. Conclusion We found evident spatial associations between the two adverse human reproductive outcomes and groundwater arsenic in New Hampshire, US. However, the degree of associations and their sensitivity to different representations of arsenic level are variable. Generally, preterm birth has a stronger spatial association with groundwater arsenic than term LBW, suggesting an inconsistency in the impact of arsenic on the two reproductive outcomes. For both outcomes, younger maternal age has stronger spatial associations with groundwater arsenic.

Background: Limited by data availability, most disease maps in the literature are for relatively large and subjectively-defined areal units, which are subject to problems associated with polygon maps. High resolution maps based on objective spatial units are needed to more precisely detect associations between disease and environmental factors. Method: We propose to use a Restricted and Controlled Monte Carlo (RCMC) process to disaggregate polygon-level location data to achieve mapping aggregate data at an approximated individual level. RCMC assigns a random point location to a polygon-level location, in which the randomization is restricted by the polygon and controlled by the background (e.g., population at risk). RCMC allows analytical processes designed for individual data to be applied, and generates high-resolution raster maps. Results: We applied RCMC to the town-level birth defect data for New Hampshire and generated raster maps at the resolution of 100 m. Besides the map of significance of birth defect risk represented by p-value, the output also includes a map of spatial uncertainty and a map of hot spots. Conclusions: RCMC is an effective method to disaggregate aggregate data. An RCMC-based disease mapping maximizes the use of available spatial information, and explicitly estimates the spatial uncertainty resulting from aggregation.