Verifying locations of sources of historical environmental releases of dioxin-like compounds in the U.S.: implications for exposure assessment and epidemiologic inference.

Polychlorinated dibenzo-p-dioxin and dibenzofuran (PCDD/F) emissions from industrial sources contaminate the surrounding environment. Proximity-based exposure surrogates assume accuracy in the location of PCDD/F sources, but locations are not often verified. We manually reviewed locations (i.e., smokestack geo-coordinates) in a historical database of 4478 PCDD/F-emitting facilities in 2009 and 2016. Given potential changes in imagery and other resources over this period, we re-reviewed a random sample of 5% of facilities (n = 240) in 2016. Comparing the original and re-review of this sample, we evaluated agreement in verification (location confirmed or not) and distances between verified locations (verification error), overall and by facility type. Using the verified location from re-review as a gold standard, we estimated the accuracy of proximity-based exposure metrics and epidemiologic bias. Overall agreement in verification was high (>84%), and verification errors were small (median = 84 m) but varied by facility type. Accuracy of exposure classification (≥1 facility within 5 km) for a hypothetical study population also varied by facility type (sensitivity: 69-96%; specificity: 95-98%). Odds ratios were attenuated 11-69%, with the largest bias for rare facility types. We found good agreement between reviews of PCDD/F source locations, and that exposure prevalence and facility type may influence associations with exposures derived from this database. Our findings highlight the need to consider location error and other contextual factors when using proximity-based exposure metrics.

Farm residence and lymphohematopoietic cancers in the Iowa Women's Health Study.

BACKGROUND: Cancer incidence in male farmers has been studied extensively; however, less is known about risk among women residing on farms or in agricultural areas, who may be exposed to pesticides by their proximity to crop fields. We extended a previous follow-up of the Iowa Women's Health Study cohort to examine farm residence and the incidence of lymphohematopoietic cancers. Further, we investigated crop acreage within 750 m of residences, which has been associated with higher herbicide levels in Iowa homes. METHODS: We analyzed data for a cohort of 37,099 Iowa women aged 55-69 years who reported their residence location (farm, rural (not a farm), town size based on population) at enrollment in 1986. We identified incident lymphohematopoietic cancers (1986-2009) by linkage with the Iowa Cancer Registry. Using a geographic information system, we geocoded addresses and calculated acreage of pasture and row crops within 750 m of homes using the 1992 National Land Cover Database. Cox regression was used to estimate hazard ratios (HR) and 95% confidence intervals (CI) in multivariate analyses of cancer risk in relation to both residence location and crop acreage. RESULTS: As found in an earlier analysis of residence location, risk of acute myeloid leukemia (AML) was higher among women living on farms (HR=2.23, 95%CI: 1.25-3.99) or rural areas (but not on a farm) (HR=1.95, 95%CI: 0.89-4.29) compared with women living in towns of >10,000 population. We observed no association between farm or rural residence and non-Hodgkin lymphoma (NHL; overall or for major subtypes) or multiple myeloma. In analyses of crop acreage, we observed no association between pasture or row crop acreage within 750 m of homes and risk of leukemia overall or for the AML subtype. Chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma (SLL) risk was nonsignificantly elevated among women with pasture acreage within 750 m of their home (HRs for increasing tertiles=1.8, 1.8 and 1.5) and with row crop acreage within 750 m (HRs for increasing tertiles of acreage=1.4, 1.5 and 1.6) compared to women with no pasture or row crop acreage, respectively. CONCLUSIONS: Iowa women living on a farm or in a rural area were at increased risk of developing AML, which was not related to crop acreage near the home. Living near pasture or row crops may confer an increased risk of CLL/SLL regardless of residence location. Further investigation of specific farm-related exposures and these cancers among women living on farms and in agricultural areas is warranted.

Residential proximity to industrial combustion facilities and risk of non-Hodgkin lymphoma: a case-control study.

BACKGROUND: Residence near municipal solid waste incinerators, a major historical source of dioxin emissions, has been associated with increased risk of non-Hodgkin lymphoma (NHL) in European studies. The aim of our study was to evaluate residence near industrial combustion facilities and estimates of dioxin emissions in relation to NHL risk in the United States. METHODS: We conducted a population-based case-control study of NHL (1998-2000) in four National Cancer Institute-Surveillance Epidemiology and End Results centers (Detroit, Iowa, Los Angeles, Seattle). Residential histories 15 years before diagnosis (similar date for controls) were linked to an Environmental Protection Agency database of dioxin-emitting facilities for 969 cases and 749 controls. We evaluated proximity (3 and 5 km) to 10 facility types that accounted for >85% of U.S. emissions and a distance-weighted average emission index (AEI [ng toxic equivalency quotient (TEQ)/year]). RESULTS: Proximity to any dioxin-emitting facility was not associated with NHL risk (3 km OR = 1.0, 95% CI 0.8-1.3). Risk was elevated for residence near cement kilns (5 km OR = 1.7, 95% CI 0.8-3.3; 3 km OR = 3.8, 95% CI 1.1-14.0) and reduced for residence near municipal solid waste incinerators (5 km OR = 0.5, 95% CI 0.3-0.9; 3 km OR = 0.3, 95% CI 0.1-1.4). The AEI was not associated with risk of NHL overall. Risk for marginal zone lymphoma was increased for the highest versus lowest quartile (5 km OR = 2.6, 95% CI 1.0-6.8; 3 km OR = 3.0, 95% CI 1.1-8.3). CONCLUSIONS: Overall, we found no association with residential exposure to dioxins and NHL risk. However, findings for high emissions and marginal zone lymphoma and for specific facility types and all NHL provide some evidence of an association and deserve future study.

Estimating water supply arsenic levels in the New England Bladder Cancer Study.

BACKGROUND: Ingestion of inorganic arsenic in drinking water is recognized as a cause of bladder cancer when levels are relatively high (≥ 150 µg/L). The epidemiologic evidence is less clear at the low-to-moderate concentrations typically observed in the United States. Accurate retrospective exposure assessment over a long time period is a major challenge in conducting epidemiologic studies of environmental factors and diseases with long latency, such as cancer. OBJECTIVE: We estimated arsenic concentrations in the water supplies of 2,611 participants in a population-based case-control study in northern New England. METHODS: Estimates covered the lifetimes of most study participants and were based on a combination of arsenic measurements at the homes of the participants and statistical modeling of arsenic concentrations in the water supply of both past and current homes. We assigned a residential water supply arsenic concentration for 165,138 (95%) of the total 173,361 lifetime exposure years (EYs) and a workplace water supply arsenic level for 85,195 EYs (86% of reported occupational years). RESULTS: Three methods accounted for 93% of the residential estimates of arsenic concentration: direct measurement of water samples (27%; median, 0.3 µg/L; range, 0.1-11.5), statistical models of water utility measurement data (49%; median, 0.4 µg/L; range, 0.3-3.3), and statistical models of arsenic concentrations in wells using aquifers in New England (17%; median, 1.6 µg/L; range, 0.6-22.4). CONCLUSIONS: We used a different validation procedure for each of the three methods, and found our estimated levels to be comparable with available measured concentrations. This methodology allowed us to calculate potential drinking water exposure over long periods.

Determinants of agricultural pesticide concentrations in carpet dust.

BACKGROUND: Residential proximity to agricultural pesticide applications has been used as a surrogate for exposure in epidemiologic studies, although little is known about the relationship with levels of pesticides in homes. OBJECTIVE: We identified determinants of concentrations of agricultural pesticides in dust. METHODS: We collected samples of carpet dust and mapped crops within 1,250 m of 89 residences in California. We measured concentrations of seven pesticides used extensively in agriculture (carbaryl, chlorpyrifos, chlorthal-dimethyl, diazinon, iprodione, phosmet, and simazine). We estimated use of agricultural pesticides near residences from a statewide database alone and by linking the database with crop maps. We calculated the density of pesticide use within 500 and 1,250 m of residences for 180, 365, and 730 days before collection of dust and evaluated relationships between agricultural pesticide use estimates and pesticide concentrations in carpet dust. RESULTS: For five of the seven pesticides evaluated, residences with use of agricultural pesticides within 1,250 m during the previous 365 days had significantly higher concentrations of pesticides than did residences with no nearby use. The highest correlation with concentrations of pesticides was generally for use reported within 1,250 m of the residence and 730 days before sample collection. Regression models that also accounted for occupational and home use of pesticides explained only a modest amount of the variability in pesticide concentrations (4-28%). CONCLUSIONS: Agricultural pesticide use near residences was a significant determinant of concentrations of pesticides in carpet dust for five of seven pesticides evaluated.

Using Landsat satellite data to support pesticide exposure assessment in California.

BACKGROUND: The recent U.S. Geological Survey policy offering Landsat satellite data at no cost provides researchers new opportunities to explore relationships between environment and health. The purpose of this study was to examine the potential for using Landsat satellite data to support pesticide exposure assessment in California. METHODS AND RESULTS: We collected a dense time series of 24 Landsat 5 and 7 images spanning the year 2000 for an agricultural region in Fresno County. We intersected the Landsat time series with the California Department of Water Resources (CDWR) land use map and selected field samples to define the phenological characteristics of 17 major crop types or crop groups. We found the frequent overpass of Landsat enabled detection of crop field conditions (e.g., bare soil, vegetated) over most of the year. However, images were limited during the winter months due to cloud cover. Many samples designated as single-cropped in the CDWR map had phenological patterns that represented multi-cropped or non-cropped fields, indicating they may have been misclassified. CONCLUSIONS: We found the combination of Landsat 5 and 7 image data would clearly benefit pesticide exposure assessment in this region by 1) providing information on crop field conditions at or near the time when pesticides are applied, and 2) providing information for validating the CDWR map. The Landsat image time-series was useful for identifying idle, single-, and multi-cropped fields. Landsat data will be limited during the winter months due to cloud cover, and for years prior to the Landsat 7 launch (1999) when only one satellite was operational at any given time. We suggest additional research to determine the feasibility of integrating CDWR land use maps and Landsat data to derive crop maps in locations and time periods where maps are not available, which will allow for substantial improvements to chemical exposure estimation.

Positional accuracy of two methods of geocoding.

BACKGROUND: Geocoding is often used in epidemiologic studies to map residences with geographic information systems (GIS). The accuracy of the method is usually not determined. METHODS: We collected global positioning system (GPS) measurements at homes in a case-control study of non-Hodgkin lymphoma in Iowa. We geocoded the addresses by 2 methods: (1) in-house, using ArcView 3.2 software and the U.S. Census Bureau TIGER 2000 street database; and (2) automated geocoding by a commercial firm. We calculated the distance between the geocoded and GPS location (positional error) overall and separately for homes within towns and outside (rural). We evaluated the error in classifying homes with respect to their proximity to crop fields. RESULTS: Overall, the majority of homes were geocoded with positional errors of less than 100 m by both methods (ArcView/TIGER 2000, median = 62 m [interquartile range = 39-103]; commercial firm, median = 61 m [interquartile range = 35-137]). For town residences, the percent geocoded with errors of