Grant Number:	5U01CA259376-02 Interpret this number
Primary Investigator:	Wheeler, David
Organization:	Virginia Commonwealth University
Project Title:	Modeling Cancer Risk and Environmental and Socio-Spatial Exposures Using Residential Histories
Fiscal Year:	2022

In many studies on the risk of disease, investigators analyze the spatial patterns of health outcomes and evaluate residential environmental exposures at the time of study enrollment in hopes of identifying potential causal environmental factors. However, for cancers with long latencies like bladder cancer and non-Hodgkin lymphoma (NHL) residential locations many years prior to diagnosis are important for determining where and when relevant environmental exposures occurred in mobile study populations. Many environmental factors are distributed unevenly over space and time and several decades may have elapsed between exposure to a relevant risk factor and diagnosis. While some investigators have begun to consider residential histories in studies of cancer, a number of research challenges remain. Statistical methods are currently lacking for modeling cumulative spatial risk of cancer over time using residential histories in epidemiologic studies. There is also a need for methods that estimate environmental and socio-spatial exposure effects over time while modeling cumulative spatial risk. Exposure data are increasingly becoming multivariate and there is a need to develop statistical methods for handling multivariate exposures such as chemical mixtures over time. In the consideration of residential histories, more investigators are proposing to use public record databases such as LexisNexis to acquire historic residential locations for study subjects. However, it is currently unknown what impact using residential histories from public record databases in place of subject-reported residential histories has when studying environmental cancer risk over time. Measurement error and therefore bias could result from using public record databases particularly going several decades back in time. In this project, we aim to develop a comprehensive set of methods that incorporate residential histories into cancer studies to estimate both cumulative spatial risk and health effects of many environmental and socio-spatial exposures over time. We will apply these methods to the New England Bladder Cancer Study and the NCI-SEER NHL study to better understand environmental factors for bladder cancer and NHL. We will also assess the impact of using residential histories from public record databases on the ability of methods to identify spatial areas of risk and estimate environmental and socio-spatial exposure effects. The expected outcomes of this research will be 1) new statistical methods for estimating cumulative spatial risk and health effects for many environmental and socio-spatial exposures over time, and 2) identification of areas of significantly elevated risk over time for bladder cancer and NHL risk, 3) estimates of effects for mixtures of historic environmental and socio-spatial exposures and bladder cancer and NHL risk, and 4) an assessment of using residential histories from public record databases to estimate historic exposure effects and detect areas of elevated cancer risk. The methodological approaches developed will be applicable to many studies of cancer and environmental risk. In addition, the findings from the assessment of public record database residential histories will be useful for many investigators considering using them in environmental risk studies.




Modeling Historic Arsenic Exposures and Spatial Risk for Bladder Cancer.
Authors: Boyle J. , Ward M.H. , Koutros S. , Karagas M.R. , Schwenn M. , Johnson A.T. , Silverman D.T. , Wheeler D.C. .
Source: Statistics In Biosciences, 2024 Jul; 16(2), p. 377-394.
EPub date: 2023-12-17 00:00:00.0.
PMID: 39247147
Related Citations

Assessing and attenuating the impact of selection bias on spatial cluster detection studies.
Authors: Boyle J. , Ward M.H. , Cerhan J.R. , Rothman N. , Wheeler D.C. .
Source: Spatial And Spatio-temporal Epidemiology, 2024 Jun; 49, p. 100659.
EPub date: 2024-05-12 00:00:00.0.
PMID: 38876558
Related Citations

Modeling variation in mixture effects over space with a Bayesian spatially varying mixture model.
Authors: Boyle J. , Ward M.H. , Cerhan J.R. , Rothman N. , Wheeler D.C. .
Source: Statistics In Medicine, 2024-02-02 00:00:00.0; , .
EPub date: 2024-02-02 00:00:00.0.
PMID: 38303638
Related Citations

Modeling historic neighborhood deprivation and non-Hodgkin lymphoma risk.
Authors: Boyle J. , Ward M.H. , Cerhan J.R. , Rothman N. , Wheeler D.C. .
Source: Environmental Research, 2023-09-01 00:00:00.0; 232, p. 116361.
EPub date: 2023-06-08 00:00:00.0.
PMID: 37295583
Related Citations

Modeling historic environmental pollutant exposures and non-Hodgkin lymphoma risk.
Authors: Boyle J. , Ward M.H. , Cerhan J.R. , Rothman N. , Wheeler D.C. .
Source: Environmental Research, 2023-05-01 00:00:00.0; 224, p. 115506.
EPub date: 2023-02-18 00:00:00.0.
PMID: 36805898
Related Citations

Estimating mixture effects and cumulative spatial risk over time simultaneously using a Bayesian index low-rank kriging multiple membership model.
Authors: Boyle J. , Ward M.H. , Cerhan J.R. , Rothman N. , Wheeler D.C. .
Source: Statistics In Medicine, 2022-12-20 00:00:00.0; 41(29), p. 5679-5697.
EPub date: 2022-09-25 00:00:00.0.
PMID: 36161724
Related Citations

Estimating cumulative spatial risk over time with low-rank kriging multiple membership models.
Authors: Boyle J. , Ward M.H. , Koutros S. , Karagas M.R. , Schwenn M. , Silverman D. , Wheeler D.C. .
Source: Statistics In Medicine, 2022-07-11 00:00:00.0; , .
EPub date: 2022-07-11 00:00:00.0.
PMID: 35816955
Related Citations

Knot selection for low-rank kriging models of spatial risk in case-control studies.
Authors: Boyle J. , Wheeler D.C. .
Source: Spatial And Spatio-temporal Epidemiology, 2022 06; 41, p. 100483.
EPub date: 2022-01-21 00:00:00.0.
PMID: 35691650
Related Citations