Using satellite derived land cover information for a multi-temporal study of self-reported recall of proximity to farmland

Abstract

Exposure misclassification is a major concern in epidemiologic studies. The potential for misclassification becomes even more problematic when participants are asked to recall historical information. Yet, historical information is important in cancer studies, where latency is long and causative exposures may have occurred years or even decades prior to diagnosis. Even though self-reported proximity to farmland is a commonly used exposure measure, the accuracy of recall is seldom, if ever validated. Geographic Information Systems (GISs) and land cover information derived from satellite imagery can allow researchers to assess the accuracy of this exposure measure, and to quantify the extent and importance of exposure misclassification. As part of a bladder cancer case-control study in Michigan, participants were asked whether they lived on a farm, or within a distance of 1/4, 1/4-1, 1-5, or 45 miles from farmland for each residence over their lifespan. Responses from 531 participants over two time periods F 1978 and 2001 F were investigated. Self reported proximity to farmland was compared to a ''gold standard'' derived from Michigan land cover files for the same time periods. Logistic regression and other statistical measures including sensitivity, specificity, and percentage matching were evaluated. In comparing self-reported and land cover-derived proximity to farmland, cases exhibited better agreement than controls in 2001 (adjusted OR ¼ 1.74; 95% CI ¼ 1.01, 2.99) and worse agreement in 1978, although not significantly (adjusted OR ¼ 0.74; 95% CI ¼ 0.47, 1.16). When comparing 2001 with 1978, both cases and controls showed better agreement in 2001, but only cases showed a significant difference (adjusted OR ¼ 2.36; 95% CI ¼ 1.33, 4.18). These differences in agreement may be influenced by differences in educational attainment between cases and controls, although adjustment for education did not diminish the association. Gender, age, number of years at residence, and geocoding accuracy did not influence agreement between the proximity approaches. This study suggests that proximity measures taken from satellite-derived land cover imagery may be useful for assessing proximity to farmland, and it raises some concerns about the use of self-reported proximity to farmland in exposure assessments.

Identifying environmental risk factors and gene–environment interactions in holoprosencephaly

“…A significant strength of this study is its use of a unique control group affected by congenital malformations. Previous studies have shown that biased odds ratios can result in self‐report exposure studies utilizing healthy control groups where cases more accurately recall exposures than controls (Avruskin, Meliker, & Jacquez, 2008; Rull et al, 2006). To mitigate such biases, we specifically used a control group (Williams‐Beuren syndrome) that also has congenital anomalies with a disease pathophysiology and etiology that is entirely different than HPE (Kruszka, Porras, et al, 2018; Morris, 2017; Solomon et al, 2013).…”

Prenatal exposure to pesticides and risk for holoprosencephaly: a case-control study

“…This study design is important to rule out potential recall bias inherent in retrospective studies for traumatic health outcomes including birth defects. The specific benefit of having a control group with a congenital anomaly can be seen considering the biased odds ratios that can result in selfreport studies on exposures utilizing healthy control groups [57]. For example, Rull and colleagues [17] explored the association between residential proximity to agricultural crops during pregnancy and neural tube defects by comparing self-reports and land use maps of actual proximity.…”

Use of land surface remotely sensed satellite and airborne data for environmental exposure assessment in cancer research

“…Seven papers used satellite imagery in exposure science applications (no health end points). These included validation studies of models derived from satellite imagery (Matthijsen et al, 2000; Maxwell et al, 2006; Rull et al, 2006; AvRuskin et al, 2008), constructing a model of atrazine in a drinking water reservoir using land cover data (Atkinson et al, 2001), estimating deterioration of asbestos cement roofing using hyperspectral imagery (Bassani et al, 2007), and using satellite imagery to document flooding in schools as an indicator of building dampness and potential respiratory risk (Guidry and Margolis, 2005). One paper used satellite imagery and GIS to estimate proximity to different crop types in a study of birth weight (Xiang et al, 2000).…”