Skip to main content
NCBI home page
EPA Author Manuscripts logoLink to EPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Sep 4.
Published in final edited form as: Cancer. 2017 May 8;123(15):2901–2908. doi: 10.1002/cncr.30709

County-level cumulative environmental quality associated with cancer incidence

JS Jagai 1,*, LC Messer 2, KM Rappazzo 3, CL Gray 4,5, SC Grabich 4,5, DT Lobdell 3
PMCID: PMC6121813  NIHMSID: NIHMS1500340  PMID: 28480506

Abstract

Background

Individual environmental exposures are associated with cancer development; however, environmental exposures occur simultaneously. The Environmental Quality Index (EQI) is a county-level measure of cumulative environmental exposures, occurring in five domains.

Methods

The EQI was linked to county-level annual age-adjusted cancer incidence rates from the Surveillance, Epidemiology, and End Results (SEER) Program State Cancer Profiles. All-site cancer and the top three site-specific cancers for males and females were considered. Incident rate differences (IRD, annual rate difference per 100,000 persons) and 95% confidence intervals (CI) were estimated using fixed-slope, random intercept multi-level linear regression models. Associations were assessed with domain-specific indices and analyses were stratified by rural-urban status.

Results

Comparing highest quintile/poorest environmental quality to lowest quintile/best environmental quality for overall EQI, all-site county-level cancer incidence rate was positively associated with poor environmental quality overall (IRD: 38.55, 95%CI 29.57,47.53) and for males (IRD: 32.60, 95%CI 16.28,48.91) and females (IRD: 30.34, 95%CI 20.47,40.21), indicating a potential increase in cancer incidence with decreasing environmental quality. Rural-urban stratified models demonstrated positive associations comparing highest to lowest quintiles for all strata, except the thinly populated/rural stratum and in the metropolitan-urbanized stratum. Prostate and breast cancer demonstrated the strongest positive associations with poor environmental quality.

Conclusion

We observed strong positive associations between the EQI and all-site cancer incidence rates and associations differed by rural-urban status and environmental domain. Research focusing on single environmental exposures in cancer development may not address the broader environmental context in which cancers develop and future research needs to consider cumulative environmental exposures.

Keywords: all-site cancer, cumulative environmental exposures, air, water, land, built, sociodemographic

Introduction

Cancer is a major public health problem in the United States (U.S.), causing one in four deaths 1. In 2014, an estimated 585,720 deaths, approximately 1,600 per day, were due to cancer 1. Estimated cancer costs in 2009 were $243.4 billion 2. The most common causes of cancer death in men are lung, prostate, and colorectal cancer and in women are lung, breast, and colorectal cancer 1. Cancer risk is affected by a combination of genetic factors and environmental exposures. Recent research suggests that genetic variations interact with harmful environmental exposures to exacerbate exposure effects and increase cancer risk 3–5.

Analysis of data on twins suggests that the genetic contribution to cancer is about 50%, indicating that exogenous factors play a significant role in cancer development 6–8. Environmental exposures can alter or interfere with a variety of biological processes, including hormone production and function, inflammation, DNA damage, and gene suppression or overexpression 4, 9. For example, lung cancer is associated with several environmental exposures including radon 10, 11, pesticides 12, 13, and diesel exhaust 14, 15. Breast and prostate cancers are also associated with environmental exposures, such as ionizing radiation, solvents, and environmental mutagens 16–18. Social exposures, such as poverty, and the built environment have also been associated with cancer outcomes 19–21. However, the interaction of multiple environmental exposures remains largely unstudied 22, 23; therefore, the burden of environmentally-induced cancer may be underestimated.

Epidemiologic research has traditionally focused on single environmental exposures because the empirical quantification of cumulative exposures, from various environmental sources, is difficult 24, 25. However, measuring a single environmental exposure does not fully capture environmental effects on health. Rather, several environmental exposures, including social exposures, occur simultaneously, working through multiple mechanisms to induce poor health outcomes, including cancer.

To capture multi-dimensional ambient environmental exposures, the Environmental Quality Index (EQI) was developed. The publically available EQI 26 is a county-level measure of cumulative ambient environmental exposures for the U.S. for the period 2000–2005 24, 27. The index was constructed to provide one unified EQI as well as domain-specific indices for all counties.

We used the EQI to assess the burden of cumulative environmental exposures on all-site and site-specific cancer incidence. We examined county-level cancer incidence rates for 2006–2010 in association with the EQI which represents the period 2000–2005. In order to assess which environmental domains drive the associations with cancer incidence, we also considered domain-specific indices. Factors influencing environmental quality vary in urban and rural areas 28; therefore, we also investigated associations between cancer incidence and the EQI and domain-specific indices stratified by rural-urban status.

Methods and Materials

Study population and outcome data

Population based cancer incidence rates were abstracted from National Cancer Institute, State Cancer Profiles 29 for all available counties in the United States for which data were available. This is a national county-level database of cancer data that are collected by state public health surveillance systems. All-site cancer is defined as any type of cancer that is captured in the state registry data, though non-melanoma skin cancer is not included. All-site age-adjusted cancer incidence rates were abstracted separately for males and females. County-level annual age-adjusted all-site cancer incidence rates for years 2006–2010 were available for 2687 of 3142 (85.5%) counties in the U.S. Counties for which there are fewer than 16 reported cases in a specific area-sex-race category are suppressed to ensure confidentiality and stability of rate estimates; this accounted for 14 counties in our study. Two states, Kansas and Virginia, do not provide data because of state legislation and regulations which prohibit the release of county level data to outside entities. Data from Michigan does not include cases diagnosed in other states because data exchange agreements prohibit the release of data to third parties. Finally, state data is not available for three states, Minnesota, Ohio, and Washington 29. The age-adjusted average annual incidence rate for all counties was 453.7 per 100,000 persons.

We selected 2006–2010 as it is subsequent in time to the EQI exposure data which was constructed to represent the years 2000–2005. We also gathered data for the three leading causes of cancer for males (lung, prostate, and colorectal) and females (lung, breast, and colorectal). These cancers account for a total of less than 50% of all cancer incidence (breast 14.0%, prostate 13.3%, lung 13.3%, and colorectal 8.0%) 30. Incidence rate data were limited for site-specific cancers due to the lower number of cases. Site-specific cancer incidence rates for males were available for 2322 (73.9%) counties for lung cancer, 2508 (79.8%) counties for prostate cancer, and 2133 (67.9%) counties for colorectal cancer. For females, site-specific cancer incidence rates were available for 2168 (69.0%) counties for lung cancer, 2460 (75.3%) counties for breast cancer, and 2020 (64.3%) counties for colorectal cancer.

Exposure data: The Environmental Quality Index (EQI)

The EQI was used as an exposure metric as an indicator of cumulative environmental exposures at the county-level representing the period 2000 to 2005. A complete description of the datasets used in the EQI are provided in Lobdell et al. 27 and methods used for index construction are described by Messer et al. 24. The EQI was developed for the period 2000–2005 because it was the time period for which the most recent data were available when index construction was initiated. The EQI includes variables representing each of the environmental domains. The air domain includes 87 variables representing criteria and hazardous air pollutants. The water domain includes 80 variables representing overall water quality, general water contamination, recreational water quality, drinking water quality, atmospheric deposition, drought, and chemical contamination. The land domain includes 26 variables representing agriculture, pesticides, contaminants, facilities, and radon. The built domain includes 14 variables representing roads, highway/road safety, public transit behavior, business environment, and subsidized housing environment. The sociodemographic environment includes 12 variables representing socioeconomics and crime. A table listing all variables in each domain by construct is provided as Supplemental Table 1.

Domain-specific indices (air index, water index, etc.) were created by retaining the first component of a principle components analysis (PCA) which included all of the domain-specific variables. The EQI was created by retaining the first component of a PCA which combined the domain-specific indices. Recognizing environments differ across the rural–urban continuum, the EQI and domain-specific index construction was also stratified by rural-urban continuum codes (RUCC) 31. We utilized four categories for which RUCC1 represents metropolitan urbanized; RUCC2 non-metro urbanized; RUCC3 less urbanized; and RUCC4 thinly populated which have been used in previous health analyses 32–35. Finally, we have six non-stratified indices (one overall EQI and five domain-specific indices) and six corresponding indices for each of the four RUCC strata. This allows for assessment of cumulative environmental exposure, domain-specific drivers, and rural-urban variations. The rural-urban stratified overall EQI is shown in Supplemental Figure 1; for each index, higher values correspond to poorer environmental quality.

Data analysis

We assessed relationships between county-level exposures representing 2000–2005 and cancer incidence rates for 2006–2010 to account for the lag in cancer development. Exposure variables used in the analysis were: non-stratified and RUCC-stratified EQI, and non-stratified and RUCC-stratified domain-specific indices. For cancer outcomes we considered county-level, age-adjusted, all-site cancer incidence rates and age-adjusted, site-specific cancer incidence rates for the leading cancer types by sex. Indices were developed as continuous variables and standardized to have a mean of zero and standard deviation (sd) of one. Therefore, analyses used quintiles of the indices to allow for more meaningful interpretation (between areas of good (1), moderate (3) and poor (5) environmental quality). The index quintiles were associated with county-level cancer incidence rates using fixed slope, random intercept multi-level linear regression models, with state as the random effect and county as the fixed effect, to estimate the fixed effects of index quintiles on cancer incidence rates. Correlations among the EQI domains range from 0.08 (air and water domains) to 0.40 (air and built domains) 24. Domain-specific analyses were adjusted for all other environmental domains. Given that the EQI includes variables from all domains of the broader environment, there are few potential confounders to adjust for in the analysis. However, analyses were adjusted by county percentage of population ever smoked which is available from the SEER database. Analyses of breast cancer incidence rates in females were additionally adjusted for county-level mammography screening rates. Results are reported as incidence rate difference (IRDs, annual rate difference per 100,000 persons) and 95% confidence intervals (CIs) comparing each quintile to the lowest quintile/best environmental quality for each index. We utilize a difference measure since, especially given the severity of the outcomes of interest, they are informative for assessing public health impact and to inform decision making 36.

Results

Population description

Of the 2687 counties in the analysis, 34% (921) were metropolitan-urbanized (RUCC1), 10% (274) were non-metropolitan urbanized (RUCC2), 34% (927) were less-urbanized (RUCC3), and 21% (564) were thinly-populated (RUCC4). This mirrors the RUCC distribution of all U.S. counties, which is also 34% RUCC1, 10% RUCC2, 34% RUCC3, and 21% RUCC4. The average annual county-level age-adjusted all-site cancer incidence rate was 451.03 cases per 100,000 population (sd: 59.43). The mean and standard deviations of annual county-level age adjusted incidence rates for all cancer outcomes varied across rural-urban strata (Table 1).

Table 1.

Mean and standard deviations (per 100,000 population) for annual county-level age-adjusted incidence rates, 2006–2010, of various cancer outcomes considered by rural-urban strata (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated) and the number of counties for which data were available.

All counties RUCC1 metropolitan-urbanized RUCC2 non-metropolitan urbanized RUCC3 less-urbanized RUCC4 thinly-populated
Outcome mean ± std (N counties) mean ± std (N counties) mean ± std (N counties) mean ± std (N counties) mean ± std (N counties)
All-site Cancer 451.03 ± 59.43 (2687) 462.80 ± 48.01 (921) 462.62 ± 44.98 (274) 448.48 ± 59.49 (927) 430.44 ± 74.51 (564)
Males
 All-site Cancer 525.26 ± 78.69 (2663) 537.23 ± 63.03 (920) 537.90 ± 62.10 (274) 521.80 ± 82.68 (927) 504.47 ± 96.47 (541)
 Lung Cancer 93.86 ± 26.39 (2323) 89.54 ± 22.49 (901) 89.75 ± 22.00 (274) 96.35 ± 29.20 (851) 103.68 ± 29.08 (296)
 Prostate Cancer 139.90 ± 32.31 (2508) 143.37 ± 29.39 (913) 141.35 ± 28.45 (274) 135.56 ± 32.26 (907) 140.85 ± 39.29 (413)
 Colorectal Cancer 56.07 ± 13.02 (2133) 52.77 ± 9.84 (874) 55.01 ± 10.83 (274) 58.41 ± 13.97 (784) 62.83 ± 18.61 (200)
Females
 All-site Cancer 399.78 ± 55.15 (2650) 409.97 ± 44.14 (919) 408.41 ± 41.59 (274) 395.57 ± 51.08 (927) 385.07 ± 77.02 (529)
 Lung Cancer 59.42 ± 14.33 (2167) 58.54 ± 12.81 (887) 58.43 ± 13.15 (273) 59.66 ± 15.23 (785) 63.40 ± 17.26 (221)
 Breast Cancer 114.22 ± 19.10 (2454) 119.16 ± 16.04 (909) 116.26 ± 14.99 (274) 109.60 ± 18.82 (903) 111.88 ± 25.40 (367)
 Colorectal Cancer 41.70 ± 9.35 (2020) 39.67 ± 7.38 (858) 40.63 ± 7.56 (272) 43.48 ± 10.18 (738) 46.40 ± 13.73 (151)
Open in a new tab

std = standard deviation; RUCC= Rural Urban Continuum Code; N = number

All-site Cancer Incidence Results

County-level models were used to assess environmental drivers of all-site cancer incidence rates, stratified by sex and RUCC. Comparing the highest quintile/poorest environmental quality to the lowest quintile/best environmental quality for the overall EQI, all-site county-level cancer incidence was positively associated with poor environmental quality overall (IRD: 38.55, 95%CI 29.57,47.53) and for males (IRD: 32.60, 95%CI 16.28,48.91) and females (IRD: 30.34, 95%CI 20.47,40.21). Considering all quintiles, the analysis demonstrated increasing trends in association of cancer as environmental quality declined (Figure 1). RUCC-stratified models demonstrated positive associations comparing highest to lowest quintiles for all strata, except the thinly-populated, and increasingly stronger associations by quintile for all-site cancer in the metropolitan-urbanized strata.

Figure 1.

Figure 1

Open in a new tab

Incidence Rate Differences (95% CI) for all site cancer, 2006–2010, and overall Environmental Quality Index (EQI), combined and separately for males and females by urban/rural continuum (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated) using quintile 1 (best environmental quality) as reference and adjusting for county percentage of population ever smoked.

Associations with the air index and all-site cancer incidence rates were positive and demonstrated an increasing trend by quintiles, indicating a potential increase in cancer incidence with decreasing air quality (Figure 3). Comparing the highest quintile to the lowest, the analysis demonstrated all-site county-level cancer incidence was positively associated with poor air quality (IRD: 44.19, 95%CI 34.84,53.54). Patterns of associations were similar across rural-urban strata, demonstrating an increase in all-site cancer incidence with worsening air quality.

Effect estimates for the water index were negative or near-null for all-site cancer incidence. For example, for all-site cancer comparing the highest to lowest quintile the IRD was −0.75 (95%CI −14.21,12.70) and there was no trend by quintile. Results for RUCC-stratified analyses were similar, demonstrating negative or near-null effect estimates without a trend by quintile. The metropolitan-urbanized strata, however, demonstrated negative results across all quintiles with the highest effect estimates seen with the 4th and 5th quintiles (IRD: −19.68, 95%CI −35.45, −3.91 and IRD: −20.45, 95%CI −35.24, −5.85, respectively), indicating a decrease in all-site cancer with worsening water quality.

Effect estimates for the land index varied by quintile and by RUCC. All-site cancer demonstrated an increasing association with quintile of land domain index; the highest effect estimate seen in the highest quintile (poor environmental quality) (IRD: 6.27, 95%CI −3.93,16.47). RUCC-stratified analyses demonstrated generally positive effect estimates in the metropolitan-urbanized and non-metropolitan urbanized strata and negative effect estimates in the less-urbanized and thinly-populated strata.

Associations with the built index were positive for all quintiles and demonstrated an increasing trend by quintiles in non-stratified analyses. The highest effect estimate for the built index was in the 4th quintile (IRD: 29.82, 95%CI 21.27,38.38). Patterns of association were positive across all rural-urban strata with the highest associations seen in the thinly-populated strata.

Effect estimates for the sociodemographic index were also positive for all quintiles of the EQI, though a trend by quintiles was not seen in the non-stratified analysis. The highest effect estimate for the sociodemographic index was in the 3rd quintile (IRD: 20.59, 95%CI 11.30,29.88). Similar to the built domain, patterns of association were positive across all rural-urban strata; the highest associations seen in the thinly-populated strata.

Site-specific Cancer Incidence Results

County-level models were used to assess environmental drivers of incidence rates of the top three site-specific cancers for males and females. In general, effect estimates for site-specific cancers for both males and females were negative or near-null (Figure 2). However, both prostate cancer and breast cancer demonstrated positive associations when comparing the highest quintile/poorest environmental quality to the lowest quintile/best environmental quality for the overall EQI (IRD: 10.17, 95%CI 0.84, 19.50 and IRD: 7.29, 95%CI 3.05, 11.54, respectively). Lung cancer in males was negatively associated with environmental quality (IRD: −8.66, 95%CI −12.95, −4.83). RUCC stratified analyses demonstrated similar trends.

Figure 2.

Figure 2

Open in a new tab

Incidence Rate Differences (95% CI) for all site cancer, 2006–2010, for domain-specific indices (Air, Water, Land, Built, and Sociodemographic (SD) domains) by urban/rural continuum (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated) using quintile 1 (best environmental domain quality) as reference and adjusting for county percentage of population ever smoked and all other environmental domain indices.

When considering domain-specific associations with site-specific cancers, again comparing the highest quintile to the lowest quintile, prostate cancer demonstrated positive associations in the air domain (IRD: 10.09, 95%CI 1.84, 18.34), built domain (IRD: 8.98, 95%CI 2.54, 15.41), and sociodemographic domain (IRD: 10.39, 95%CI 3.38, 17.41). Prostate cancer also demonstrated positive associations with the air and built domains in all RUCC strata except the thinly-populated (Supplemental Figure 3). Prostate cancer was negatively associated with the sociodemographic domain in the metropolitan urban stratum but positively associated in all other RUCC strata.

Breast cancer in females demonstrated trends similar to prostate cancer in males. Breast cancer was positively associated with poor air quality (IRD: 3.69, 95%CI −0.35, 7.73), poor built environment (IRD: 5.59, 95%CI 2.04, 9.14), and poor sociodemographic environment (IRD: 5.01, 95%CI 0.82, 9.20). Again, considering RUCC stratification, breast cancer demonstrated positive associations with the air and built domains in all RUCC strata except the thinly populated and was negatively associated with the sociodemographic domain in the metropolitan urban stratum but positively associated in all other RUCC strata (Supplemental Figure 4).

The analysis demonstrated that lung cancer was positively associated with the air domain for both males and females. These associations were highest in the non-metropolitan urban and less urban strata (IRD: 12.78, 95%CI 3.95, 21.61 and IRD: 8.50, 95%CI 2.12, 14.88, respectively) for males and the metropolitan urban stratum (IRD: 5.01, 95%CI 1.31, 8.71) for females. Otherwise, effect estimates for lung cancer for both males and females were negative or near null for all other domains and varied greatly for all domains when analyses were stratified by RUCC (Supplemental Figures 3 and 4).

The analysis demonstrated that colorectal cancer was negatively associated with the air, built, and sociodemographic domains for both males and females. However, effect estimates varied greatly for all domains when analyses were stratified by RUCC (Supplemental Figures 3 and 4). Colorectal cancer was negatively associated with poor water quality for both males and females (IRD: −4.45, 95%CI −7.73, −1.16 and IRD: −3.38, 95%CI −6.00, −0.76, respectively). Associations were also negative for all RUCC strata, except the thinly-populated, for both males and females.

Discussion

We observed positive associations between the Environmental Quality Index (EQI), a metric of cumulative environmental exposure, and all-site cancer incidence rates, overall and in both males and females. Associations differed by rural-urban status and by the five environmental domains considered. The highest associations were seen in the air, built, and sociodemographic domains, suggesting these domains are driving the associations with cancer outcomes. Associations in the most urbanized areas were highest for both males and females and across the domain-specific indices. When site-specific cancers were considered the highest positive associations were seen for prostate cancer in males and breast cancer in females.

Genetic variation alone does not account for all cancer outcomes, but instead interacts with harmful environmental exposures, modifying the effects of these exposures and risk of cancer 3–5. Environmental exposures can alter or interfere with a variety of biological processes, including hormone production and function, inflammation, DNA damage, and gene suppression or overexpression 4, 9. Analysis of data on twins suggests that the genetic contribution to cancer is about 50%, allowing for a significant role of environmental exposures 6, 8. Both breast and prostate cancers are associated with environmental exposures such as ionizing radiation and solvents 16–18. Our findings also show positive relationships between incidence of breast and prostate cancer and environmental quality. Lung cancer is associated with air pollution exposures such as diesel exhaust 14, 15 and polycyclic aromatic hydrocarbons (PAHs) 37. Our findings demonstrate positive associations between lung cancer incidence in males and poor air quality. Lung cancer has also been shown to be associated with land pollutants such as radon 10, 11 and pesticides 12, 13; however, we did not observe associations with lung cancer and the land domain index.

Environmental health research has utilized indices to represent multiple variables with a single quantitative measure. This methodology has most commonly been used to represent the built and social environments to describe neighborhood differences 38, 39. Air pollution studies have also utilized an index to examine complex mixtures of air pollutants 40. These cumulative metrics have been associated with various health outcomes, including preterm birth 38–40. Measuring a single environmental exposure does not fully capture the health effects resulting from the overall burden of environmental exposures. Rather, environmental exposures occur simultaneously and work through multiple mechanisms to result in cancer. This is the first study, of which we are aware, to utilize an index of environmental quality to assess the burden of cumulative environmental exposures on cancer incidence.

Index development methods have typically been used within a single environmental domain but not to assess simultaneous burden across environmental domains. The EQI is a novel metric of cumulative environmental exposures which was developed utilizing publically available data. However, combining data across domains is challenging for several reasons. Environmental data are often collected for administrative and regulatory purposes and therefore may not provide the spatial and/or temporal coverage to properly assess health outcomes 41. For both the development of the EQI and this analysis, data were better represented in urban areas compared to suburban and rural areas.

The use of an ecological exposure metric is both a limitation and strength of this analysis. The EQI represents the period 2000–2005 and reflects exposures occurring prior to the cancer incidence assessed in this analysis, but there are varying and long lag periods associated with the development of cancer. The EQI is a rank index and can be representative of environmental quality over time. Sensitivity analyses have shown little change in county rankings over time; however, we did not assess changes over more than 10 years, which may be possible lag periods for cancer development. In addition, the ecological nature of this analysis does not allow us to account for individual level confounders, such as alcohol use, physical activity, and nutrition, which may bias results. However, the use of a broad ecological exposure metric is also a strength of this analysis because of the ability to assess the cumulative environment. The EQI considers hundreds of environmental exposures simultaneously across multiple environmental domains, including the sociodemographic environment, which is often neglected when considering environmental exposures. In addition, we were able to leverage publically available exposure and outcome data to assess relationships between environmental quality and cancer incidence on a national level.

Conclusion

Our county-level analyses demonstrated positive associations between cumulative environmental quality and cancer incidence for all rural-urban strata. The results were mixed for domain-specific indices; for breast and prostate cancer incidence, results were strongest in the air domain. Our analyses suggest that cumulative environmental quality can influence cancer risk and that associations vary by urbanicity. Our study addresses the current appeal for research that expands beyond single exposures by using the Environmental Quality Index, a novel index representing five environmental domains 42, 43. This study demonstrates that focusing on single environmental exposures in cancer development, while necessary to understand specific mechanisms, may not address the broader environmental context in which cancers develop and future research needs to also consider the impact of cumulative environmental exposures.

Supplementary Material

Sup 1 Table
Sup Fig Legend
Supplement1

Supplemental Figure 1. Map of Overall Environmental Quality Index, 2000–2005, by county stratified by four rural urban strata (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated).

Supplement2

Supplemental Figure 2. Incidence Rate Differences (95% CI) for site-specific cancers, 2006–2010, and overall Environmental Quality Index (EQI) for each sex by urban/rural continuum (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated) using quintile 1 (best environmental quality) as reference. All analyses are adjusted for county percentage of population ever smoked, analyses of breast cancer rates are also adjusted for county-level mammography screening rates.

Supplement3

Supplemental Figure 3. Incidence Rate Differences (95% CI) for site-specific cancers, 2006–2010, for males for domain-specific indices by urban/rural continuum (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated) using quintile 1 (best environmental domain quality) as reference and adjusting for county percentage of population ever smoked.

Supplement4

Supplemental Figure 4. Incidence Rate Differences (95% CI) for site-specific cancers, 2006–2010, for females for domain-specific indices by urban/rural continuum (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated) using quintile 1 (best environmental domain quality) as reference. All analyses are adjusted for county percentage of population ever smoked, analyses of breast cancer rates are also adjusted for county-level mammography screening rates.

Acknowledgments

The Office of Research and Development (ORD), U.S. Environmental Protection Agency (EPA), partially funded the research with L. C. Messer (Contracts EP09D000003 and EP12D000264) and also supported in part by an appointment to the Internship/Research Participation Program at Office of Research and Development (National Health and Environmental Effects Research Laboratory), U.S. Environmental Protection Agency, administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the U.S. Department of Energy and EPA. The authors wish to thank Therese Dolecek and Radhika Dhingra for their insightful reviews and suggestions to improve this manuscript.

Footnotes

Conflicts of Interest: The authors do not have any conflicts of interest.

Author Contributions: JSJ conceived the study, led the analysis and writing of the manuscript. LCM, KMP, CLG, SCG, and DTL all participated in the development of the exposure metric, study design, interpretation of results, and contributed to the writing of the manuscript.

Disclaimer

The views expressed in this manuscript are those of the authors and do not necessarily reflect the views or policies of the U.S. Environmental Protection Agency. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.

References

  • 1.Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin. 2014 Jan;63(1):11–30. doi: 10.3322/caac.21166. [DOI] [PubMed] [Google Scholar]
  • 2.National Heart L, and Blood Institute. Fact book: fiscal year 2008. National Institues of Health; 2009. [Google Scholar]
  • 3.Tomasetti C, Vogelstein B. Cancer etiology. Variation in cancer risk among tissues can be explained by the number of stem cell divisions. Science. 2015 Jan 2;347(6217):78–81. doi: 10.1126/science.1260825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Wallace TA, Martin DN, Ambs S. Interactions among genes, tumor biology and the environment in cancer health disparities: examining the evidence on a national and global scale. Carcinogenesis. 2011 Aug;32(8):1107–1121. doi: 10.1093/carcin/bgr066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Wu S, Powers S, Zhu W, Hannun YA. Substantial contribution of extrinsic risk factors to cancer development. Nature. 2016 Jan 7;529(7584):43–47. doi: 10.1038/nature16166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Amundadottir LT, Thorvaldsson S, Gudbjartsson DF, et al. Cancer as a complex phenotype: pattern of cancer distribution within and beyond the nuclear family. PLoS Med. 2004 Dec;1(3):e65. doi: 10.1371/journal.pmed.0010065. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Lichtenstein P, Holm NV, Verkasalo PK, et al. Environmental and heritable factors in the causation of cancer--analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med. 2000 Jul 13;343(2):78–85. doi: 10.1056/NEJM200007133430201. [DOI] [PubMed] [Google Scholar]
  • 8.Risch N. The genetic epidemiology of cancer: interpreting family and twin studies and their implications for molecular genetic approaches. Cancer Epidemiol Biomarkers Prev. 2001 Jul;10(7):733–741. [PubMed] [Google Scholar]
  • 9.Irigaray P, Newby JA, Clapp R, et al. Lifestyle-related factors and environmental agents causing cancer: an overview. Biomed Pharmacother. 2007 Dec;61(10):640–658. doi: 10.1016/j.biopha.2007.10.006. [DOI] [PubMed] [Google Scholar]
  • 10.de Groot P, Munden RF. Lung cancer epidemiology, risk factors, and prevention. Radiol Clin North Am. 2012 Sep;50(5):863–876. doi: 10.1016/j.rcl.2012.06.006. [DOI] [PubMed] [Google Scholar]
  • 11.Torres-Duran M, Barros-Dios JM, Fernandez-Villar A, Ruano-Ravina A. Residential radon and lung cancer in never smokers. A systematic review. Cancer Lett. 2014 Apr 1;345(1):21–26. doi: 10.1016/j.canlet.2013.12.010. [DOI] [PubMed] [Google Scholar]
  • 12.Bassil KL, Vakil C, Sanborn M, Cole DC, Kaur JS, Kerr KJ. Cancer health effects of pesticides: systematic review. Can Fam Physician. 2007 Oct;53(10):1704–1711. [PMC free article] [PubMed] [Google Scholar]
  • 13.Blair A, Freeman LB. Epidemiologic studies in agricultural populations: observations and future directions. J Agromedicine. 2009;14(2):125–131. doi: 10.1080/10599240902779436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Sun Y, Bochmann F, Nold A, Mattenklott M. Diesel exhaust exposure and the risk of lung cancer--a review of the epidemiological evidence. Int J Environ Res Public Health. 2014 Feb;11(2):1312–1340. doi: 10.3390/ijerph110201312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Tsoi CT, Tse LA. Professional drivers and lung cancer: a systematic review and meta-analysis. Occup Environ Med. 2012 Nov;69(11):831–836. doi: 10.1136/oemed-2012-100666. [DOI] [PubMed] [Google Scholar]
  • 16.Clapp RW, Jacobs MM, Loechler EL. Environmental and occupational causes of cancer: new evidence 2005–2007. Rev Environ Health. 2008 Jan-Mar;23(1):1–37. doi: 10.1515/reveh.2008.23.1.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Golubicic I, Borojevic N, Pavlovic T. Risk factors for breast cancer: is ionizing radiation among them? J Buon. 2008 Oct-Dec;13(4):487–494. [PubMed] [Google Scholar]
  • 18.Gray J, Evans N, Taylor B, Rizzo J, Walker M. State of the evidence: the connection between breast cancer and the environment. Int J Occup Environ Health. 2009 Jan-Mar;15(1):43–78. doi: 10.1179/107735209799449761. [DOI] [PubMed] [Google Scholar]
  • 19.Boscoe FP, Johnson CJ, Sherman RL, Stinchcomb DG, Lin G, Henry KA. The relationship between area poverty rate and site-specific cancer incidence in the United States. Cancer. 2014 Jul 15;120(14):2191–2198. doi: 10.1002/cncr.28632. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Henry KA, Sherman R, Farber S, Cockburn M, Goldberg DW, Stroup AM. The joint effects of census tract poverty and geographic access on late-stage breast cancer diagnosis in 10 US States. Health Place. 2013 May;21:110–121. doi: 10.1016/j.healthplace.2013.01.007. [DOI] [PubMed] [Google Scholar]
  • 21.Erhunmwunsee L, Joshi MB, Conlon DH, Harpole DH., Jr Neighborhood-level socioeconomic determinants impact outcomes in nonsmall cell lung cancer patients in the Southeastern United States. Cancer. 2012 Oct 15;118(20):5117–5123. doi: 10.1002/cncr.26185. [DOI] [PubMed] [Google Scholar]
  • 22.Belpomme D, Irigaray P, Hardell L, et al. The multitude and diversity of environmental carcinogens. Environ Res. 2007 Nov;105(3):414–429. doi: 10.1016/j.envres.2007.07.002. [DOI] [PubMed] [Google Scholar]
  • 23.Felter SP, Conolly RB, Bercu JP, et al. A proposed framework for assessing risk from less-than-lifetime exposures to carcinogens. Crit Rev Toxicol. 2011 Jul;41(6):507–544. doi: 10.3109/10408444.2011.552063. [DOI] [PubMed] [Google Scholar]
  • 24.Messer LC, Jagai JS, Rappazzo KM, Lobdell DT. Construction of an environmental quality index for public health research. Environ Health. 2013;13(1):39. doi: 10.1186/1476-069X-13-39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Peters A, Hoek G, Katsouyanni K. Understanding the link between environmental exposures and health: does the exposome promise too much? J Epidemiol Community Health. 2012 Feb;66(2):103–105. doi: 10.1136/jech-2011-200643. [DOI] [PubMed] [Google Scholar]
  • 26.U.S. Environmental Protection Agency (USEPA) Environmental Quality Index. 2015 [Google Scholar]
  • 27.Lobdell DT, Jagai JS, Rappazzo K, Messer LC. Data sources for an environmental quality index: availability, quality, and utility. Am J Public Health. 2011 Dec;101(Suppl 1):S277–285. doi: 10.2105/AJPH.2011.300184. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Hall SA, Kaufman JS, Ricketts TC. Defining urban and rural areas in U.S. epidemiologic studies. J Urban Health. 2006 Mar;83(2):162–175. doi: 10.1007/s11524-005-9016-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.National Cancer Institute (NCI) [Accessed June 10, 2015, 2015];State Cancer Profiles. https://statecancerprofiles.cancer.gov/
  • 30.National Cancer Institute (NCI) [Accessed November 11, 2015, 2015];Cancer Stat Fact Sheets. http://seer.cancer.gov/statfacts/
  • 31.U.S. Department of Agriculture (USDA) Rural-Urban Continuum Codes: Economic Research Service. 2015 [Google Scholar]
  • 32.Langlois PH, Jandle L, Scheuerle A, Horel SA, Carozza SE. Occurrence of conotruncal heart birth defects in Texas: a comparison of urban/rural classifications. J Rural Health. 2010 Summer;26(2):164–174. doi: 10.1111/j.1748-0361.2010.00278.x. [DOI] [PubMed] [Google Scholar]
  • 33.Langlois PH, Scheuerle A, Horel SA, Carozza SE. Urban versus rural residence and occurrence of septal heart defects in Texas. Birth Defects Res A Clin Mol Teratol. 2009 Sep;85(9):764–772. doi: 10.1002/bdra.20586. [DOI] [PubMed] [Google Scholar]
  • 34.Luben TJ, Messer LC, Mendola P, Carozza SE, Horel SA, Langlois PH. Urban-rural residence and the occurrence of neural tube defects in Texas, 1999–2003. Health Place. 2009 Sep;15(3):848–854. doi: 10.1016/j.healthplace.2009.02.006. [DOI] [PubMed] [Google Scholar]
  • 35.Messer LC, Luben TJ, Mendola P, Carozza SE, Horel SA, Langlois PH. Urban-rural residence and the occurrence of cleft lip and cleft palate in Texas, 1999–2003. Ann Epidemiol. 2010 Jan;20(1):32–39. doi: 10.1016/j.annepidem.2009.09.006. [DOI] [PubMed] [Google Scholar]
  • 36.Poole C. On the origin of risk relativism. Epidemiology. 2009 Jan;21(1):3–9. doi: 10.1097/EDE.0b013e3181c30eba. [DOI] [PubMed] [Google Scholar]
  • 37.Wagner M, Bolm-Audorff U, Hegewald J, et al. Occupational polycyclic aromatic hydrocarbon exposure and risk of larynx cancer: a systematic review and meta-analysis. Occup Environ Med. 2014 Nov 14; doi: 10.1136/oemed-2014-102317. [DOI] [PubMed] [Google Scholar]
  • 38.Messer LC, Laraia BA, Kaufman JS, et al. The development of a standardized neighborhood deprivation index. J Urban Health. 2006 Nov;83(6):1041–1062. doi: 10.1007/s11524-006-9094-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Miranda ML, Messer LC, Kroeger GL. Associations between the quality of the residential built environment and pregnancy outcomes among women in North Carolina. Environ Health Perspect. 2012 Mar;120(3):471–477. doi: 10.1289/ehp.1103578. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Wilhelm M, Ghosh JK, Su J, Cockburn M, Jerrett M, Ritz B. Traffic-related air toxics and term low birth weight in Los Angeles County, California. Environ Health Perspect. 2011 Jan;120(1):132–138. doi: 10.1289/ehp.1103408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Jagai JS, Rosenbaum BJ, Pierson SM, et al. Putting regulatory data to work at the service of public health: utilizing data collected under the clean water act. Water Qual Expo Health. 2013;5:117–125. [Google Scholar]
  • 42.Morello-Frosch R, Zuk M, Jerrett M, Shamasunder B, Kyle AD. Understanding the cumulative impacts of inequalities in environmental health: implications for policy. Health Aff (Millwood) 2011 May;30(5):879–887. doi: 10.1377/hlthaff.2011.0153. [DOI] [PubMed] [Google Scholar]
  • 43.Olden K, White SL. Health-related disparities: influence of environmental factors. Med Clin North Am. 2005 Jul;89(4):721–738. doi: 10.1016/j.mcna.2005.02.001. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Sup 1 Table
NIHMS1500340-supplement-Sup_1_Table.docx (26.1KB, docx)
Sup Fig Legend
NIHMS1500340-supplement-Sup_Fig_Legend.docx (12.1KB, docx)
Supplement1

Supplemental Figure 1. Map of Overall Environmental Quality Index, 2000–2005, by county stratified by four rural urban strata (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated).

NIHMS1500340-supplement-Supplement1.pptx (579.2KB, pptx)
Supplement2

Supplemental Figure 2. Incidence Rate Differences (95% CI) for site-specific cancers, 2006–2010, and overall Environmental Quality Index (EQI) for each sex by urban/rural continuum (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated) using quintile 1 (best environmental quality) as reference. All analyses are adjusted for county percentage of population ever smoked, analyses of breast cancer rates are also adjusted for county-level mammography screening rates.

NIHMS1500340-supplement-Supplement2.pptx (283.3KB, pptx)
Supplement3

Supplemental Figure 3. Incidence Rate Differences (95% CI) for site-specific cancers, 2006–2010, for males for domain-specific indices by urban/rural continuum (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated) using quintile 1 (best environmental domain quality) as reference and adjusting for county percentage of population ever smoked.

NIHMS1500340-supplement-Supplement3.pptx (338.6KB, pptx)
Supplement4

Supplemental Figure 4. Incidence Rate Differences (95% CI) for site-specific cancers, 2006–2010, for females for domain-specific indices by urban/rural continuum (RUCC1, metropolitan urbanized; RUCC2, non-metro urbanized; RUCC3, less urbanized; and RUCC4, thinly populated) using quintile 1 (best environmental domain quality) as reference. All analyses are adjusted for county percentage of population ever smoked, analyses of breast cancer rates are also adjusted for county-level mammography screening rates.

NIHMS1500340-supplement-Supplement4.pptx (336.3KB, pptx)

RESOURCES