Grantee Research Project Results
Final Report: An Integrated Geospatial and Epidemiological Study of Associations Between Birth Defects and Arsenic Exposure in New England
EPA Grant Number: R834599C003Subproject: this is subproject number 003 , established and managed by the Center Director under grant R834599
(EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
Center: Children's Environmental Health and Disease Prevention Center - Dartmouth College
Center Director: Karagas, Margaret Rita
Title: An Integrated Geospatial and Epidemiological Study of Associations Between Birth Defects and Arsenic Exposure in New England
Investigators: Shi, Xun , Purvis, Lisa A. , Moeschler, John B. , Onega, Tracy L. , Gui, Jiang , Rees, Judy , Miller, Stephanie
Institution: Dartmouth College , Dartmouth Medical School
EPA Project Officer: Callan, Richard
Project Period: February 15, 2010 through February 14, 2013 (Extended to February 14, 2014)
RFA: Children's Environmental Health and Disease Prevention Research Centers: Formative Centers (with NIEHS) (2009) RFA Text | Recipients Lists
Research Category: Children's Health , Human Health
Objective:
Summary/Accomplishments (Outputs/Outcomes):
New England and in particular, New Hampshire (NH) and Maine (ME) have some of the highest concentrations of groundwater arsenic in the country. In addition, NH and ME also have a relatively high proportion of residents that use private wells as a source of drinking water. Data from animal models has demonstrated that excessive exposure to arsenic causes birth defects and it is known that arsenic readily passes through the placental wall thus potentially impacting the developing fetus. Aims 1 and 2 of the project established and tested a methodology to characterize the spatial association between infants born with birth defects or with low birth weight, and their relation to arsenic exposures from drinking water in NH and ME, as well as developed a spatial distribution of arsenic exposure data using mathematical methods developed by researchers in the Department of Geography at Dartmouth College. This model estimated drinking water arsenic exposure using public databases combined with large data sets of private water systems collected through epidemiologic studies in both NH and ME. Aim 3, developed and pilot tested an epidemiologic study titled the “Prenatal Environmental Exposures and Child Health (PEECH) study." This study used case data obtained by the New Hampshire Birth Conditions Program and the Maine Birth Defects Program for study subject recruitment. Cases were notified by mail in NH and by mail and/or telephone in ME about the opportunity to participate in the study. Formal consents were obtained and study kits were mailed to recruited 143 subjects. Study kits included: materials for tap water collection, fingernails from mother and child, and saliva samples from mother and child. In addition, an extensive telephone interview was conducted by a subcontractor (University of New Hampshire Survey Center). The purpose of Aim 3 was to pilot the ability to collect relevant and useable human and environmental data from a small cohort of potentially exposed population in two separate New England states.
Specific results from Aims 1 & 2 include:
Data acquisition and database construction: The geospatial analysis for this project required geo-referenced data of birth defect cases, all births, general population (including its geography, demographics, lifestyle, and socioeconomic status), water supply, and environment (particularly arsenic in groundwater). We acquired birth defect data from the New Hampshire Birth Conditions Program (NHBCP) for all birth defect cases between 2003 through 2009 (N = 2,290 cases). Data quality checks were conducted upon receipt of each dataset. We conducted filtering processes to address issues including plurality, maternal age, and geographic location. Based on extensive research and discussion, we excluded fetal alcohol syndrome and chromosomal defects (i.e., Down syndrome, Trisomy 13, and Trisomy 18, accounting for about 8% of all the records) from analysis, because these conditions result from known causes. We extracted birth weight data from the 2003-2009 birth certificate database acquired from the NH Department of Health and Human Services (NH DHHS). We extracted gestational age data from the 2003-2009 birth certificate database acquired from the NH DHHS. We also acquired the 2003-2009 birth certificate database from the NH DHHS. We went through sophisticated and intensive processes to integrate the LandScan population data from the Oakridge National Lab and the demographic data from US Census. The outcome is a series of high-resolution (cell size = 100 m) GIS data layers of population, with demographic information attached. We have also acquired and compiled certain socioeconomic and lifestyle (smoking) data, and have associated them to the population data.
We have acquired, from NH Department of Environmental Services, “Water Well Inventory," "Wellhead Protection Area," "Public Water Supplies," and "Public Water Supply Pipelines." We have conducted GIS analyses using these data, trying to accurately delineate areas in NH with public or private water supplies.
We acquired two sets of data for the arsenic concentration in NH groundwater. The first is from the epidemiologic studies from NH at Dartmouth, containing water sample measurements from thousands of wells across NH, in the format of GIS point data. The second is USGS NH-VT Water Science Center, containing modeled arsenic levels in groundwater, in the format of GIS raster data layers. For the USGS data, we acquired three raster layers, with the cell values represent the probabilities of arsenic concentration > 1, 5, and 10 µg/L, respectively, at each location (represented by 30 m pixel) in NH.
Methodology development: We developed and improved the so-called RCMC-UCMC approach (restricted-and-controlled-Monte-Carlo and unrestricted-and-controlled-Monte-Carlo) for mapping spatial variation in disease distribution. The strengths of this approach include mapping at an approximately individual level using aggregate data, and at the same time evaluate the spatial uncertainty in the process. We applied it to birth defect mapping and created high-resolution risk maps of birth defects in NH, along with quantified evaluation of spatial uncertainty of the results. We also developed a novel local t-Test method to detect not only overall but also local spatial association between a disease and an environmental factor. This method can use aggregate data (e.g., town-level) to create a high-resolution map showing spatial variation in the association between a disease and an environmental factor, and at the same time evaluate the spatial uncertainty in the process.
Analysis and results: We conducted an extensive exploration for detecting spatial association between groundwater arsenic and birth defects in NH and did not detect a significant positive spatial associations between groundwater arsenic and birth defects. For NH, we conducted town-level analysis on spatial association between low birth weight (LBW) and groundwater arsenic, for younger mother (≤ 24), we observed a modestly positive correlation between LBW and groundwater arsenic at the state level. To increase statistical stability of the calculated LBW rates, we applied a polygon-based adaptive locally-weighted-averaging smoothing to the rates which strengthen correlations. For the original rate, r = 0.16; for smoothing threshold = 2,200 births, r = 0.25.Among the 10 counties of NH, strongest associations were among the four most populous counties.
Ongoing research: GIS analysis developing in the pilot phase will be applied in in the new Dartmouth Children’s Center research (NIH P01 ES022832, EPA RD83544201) as part of the Administrative Core, including by expanding our data to the State of Maine.
Major achievements for Aim 3 include: The Prenatal Environmental Exposures and Child Health (PEECH) Study Pilot sought to evaluate whether or not a larger scale epidemiological study evaluating the relationship between birth defects and arsenic exposure in groundwater can be successfully conducted in NH and ME. Cases were obtained from the NH Birth Conditions Program (NHBCP) and the Maine Birth Defects Program.
We adapted an existing survey instrument originally developed by the National Birth Defects Prevention Study (NBDPS), the largest population-based study ever conducted on the causes of birth defects and risk factors in the US, to fit the needs of the PEECH study. We used the NBDPS questionnaire because it had already been used for several years in another study and had been thoroughly reliability and validity tested. The survey gathered detailed family history including: pregnancy history, illnesses, medications taken, life style factors, environmental exposures, residences, and demographic information on subjects. We adapted it by adding specific questions to the existing survey related to drinking water and intake of rice. Our survey took approximately 45-minutes, by telephone, to complete. The surveys were conducted by trained interviewers at the University of New Hampshire’s Survey Center (UNH), a center that utilizes Computer-Assisted Technology Interviewing (CATI) technology.
Eligible cases were selected from the entire cohort of cases obtained by each program for calendar years 2011 and 2012 and study subject recruitment in NH and ME was initiated in May 2012. Study subjects were based on birth defect cases only, comparing types of birth defects in “arsenic exposed” households, to those same types of defects in "non-exposed" households. Exposure classification consisted of subjects’ self-reporting of prenatal water usage, results of the analysis of fingernail samples, and tap water samples taken at the time of this study.
The first subject recruitment strategy initiated for the PEECH study consisted of the NHBCP contacting 400 subjects (both ME and NH) by mail about the opportunity to participate in the study with formal consents included in this notification. After consents were received, full study kits were mailed to recruited subjects. Study kits consisted of study instruction materials; water, fingernail and saliva collection kits; and mailers for all kits. Subjects performed all collection processes themselves and then mailed samples to study office. Reminder letters and telephone calls to all subjects who did not return samples when indicated were implemented. Overall, this strategy was easy to implement in both NH and ME however after 2 months of recruitment efforts, we evaluated current recruitment methodology and determined that the recruitment strategy would need to be changed in order to obtain larger response rates – therefore, we redesigned the recruitment strategy in collaboration with the Maine Birth Defects Program. The NHBCP Coordinating Center contacted potential subjects by telephone within 10 days of the initial direct mail study invitation. It was anticipated that this change in strategy would net an increase in recruitment given contact with a person to promote the study as opposed to relying on the direct mail invitation only. After all necessary IRB and state approval for ME were obtained; recruitment using the new strategy in ME was initiated in January 2013 but only on a few subjects. Additional funding will be sought to carry out additional pilot work and interviews that require further resources than the pilot study could provide.
All subject recruitment for the epidemiological study in New Hampshire and Maine ended in February 2013. As part of this pilot, 54 maternal/child pairs were recruited (N = 35 from NH and N = 19 from ME). We obtained water samples, fingernail samples and DNA samples from participating mothers and children and completed analysis on participant water and fingernail samples. In general, all study samples arrived intact and were not damaged by the postal method for transporting back to the coordinating center. Of the 54 mothers who met study eligibility, 38 completed the telephone survey. Residential water samples were obtained for N = 42 and paired maternal and infant nails were obtained for N = 43.
We are currently in the process of evaluating all study data and protocols, as well as data from Aims 1 & 2 of to plan additional grant applications to support expansion of this project to a larger cohort.
Conclusions:
Journal Articles on this Report : 1 Displayed | Download in RIS Format
Other subproject views: | All 9 publications | 1 publications in selected types | All 1 journal articles |
---|---|---|---|
Other center views: | All 76 publications | 29 publications in selected types | All 29 journal articles |
Type | Citation | ||
---|---|---|---|
|
Shi X, Miller S, Mwenda K, Onda A, Rees J, Onega T, Gui J, Karagas M, Demidenko E, Moeschler J. Mapping disease at an approximated individual level using aggregate data: a case study of mapping New Hampshire birth defects. International Journal of Environmental Research and Public Health 2013;10(9):4161-4174. |
R834599 (Final) R834599C001 (Final) R834599C003 (Final) |
Exit Exit Exit |
Supplemental Keywords:
Relevant Websites:
https://www.dartmouth.edu/~childrenshealth/index.php Exit
Progress and Final Reports:
Original AbstractMain Center Abstract and Reports:
R834599 Children's Environmental Health and Disease Prevention Center - Dartmouth College Subprojects under this Center: (EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
R834599C001 Arsenic and Maternal and Infant Immune Function
R834599C002 Food Borne Exposure to Arsenic During the First Year of Life
R834599C003 An Integrated Geospatial and Epidemiological Study of Associations Between Birth Defects and Arsenic Exposure in New England
R834599C004 Determining How Arsenic (As) Modulates Sonic Hedgehog (Shh) Signaling During Development
The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.
Project Research Results
1 journal articles for this subproject
Main Center: R834599
76 publications for this center
29 journal articles for this center