Skip to main content
         

Vol. 08 Issue 1, Early Spring 2003

GIS Approaches to Studying Variations in Breast Cancer Incidence in California
The Ribbon 

Peggy Reynolds and Susan Hurley
California Department of Health Services

Observed regional variations in breast cancer incidence have been a source of public health concern as well as, traditionally, a source of hypothesis generation for factors which might contribute to differences in risk.

The rates of breast cancer among women in the urban San Francisco Bay and Los Angeles areas of California have been historically higher than those in many other areas of the U.S.and the world (1). Rates in these areas have also been notably higher than rates in other areas of California (2,3). Recently, a dramatic increase in rates for residents of Marin County, a small affluent area north of San Francisco, has been a topic much discussed in both the popular press and scientific literature (4).

The reasons for these variations are not well understood. Public concern over possible environmental contributors to such differences has prompted a series of studies undertaken by the California Department of Health Services (CDHS) designed to assess the influence of sociodemographic and environmental factors.  These studies are built on the availability of a high-quality, population-based cancer surveillance system, improving information on environmental toxicants, a large well-defined cohort study and more widely available geographic information system (GIS) tools. Complementary funding from the National Cancer Institute and National Institute for Environmental Health Sciences (Grants No.U01- CA81789 and R01-CA77398), and from the California Breast Cancer Research Program (Grant No.6JB- 0111), support a multidisciplinary team of researchers with expertise in epidemiology, environmental health, statistics, geographic information systems and survey research to conduct these studies.

Designed to utilize the tools of GIS, researchers at the CDHS are currently investigating regional variations in breast cancer incidence at three different levels of analysis. They are designed to incorporate both large-scale, population-based patterns and individual level exposure potential into a comprehensive assessment of risk relationships and are illustrated in Figure 1. The largest of these studies is focused on investigating statewide patterns in breast cancer incidence using GIS to unite cancer surveillance data, demographic information from the census and data on potential environmental exposures from various statewide environmental datasets. For this study, more than 180,000 cases of invasive breast cancer have been identified in California during 1988 to 1997 and the address at diagnosis for all cases has been eocoded to a census blockgroup. GIS is being used to overlay sociodemographic and environmental characteristics and statistical models are being constructed to evaluate whether areas with high breast cancer incidence are associated with area measures of sociodemographic characteristics GIS Approaches to Studying Variations in Breast Cancer Incidence in California and environmental contamination. Because this study is limited by case information available in the California Cancer Registry surveillance data, it does not have the opportunity to evaluate the degree to which differences in prevalences of established breast cancer risk factors, such as age at menarche or age at first live birth, which typically are not available at the population-level, may explain the observed geographic differences in breast cancer incidence.

In an effort to incorporate such individual-level factors in an examination of patterns of breast cancer incidence, CDHS is also conducting an analysis of regional variations of breast cancer incidence within the California Teachers Study (CTS) cohort, a large study of over 133,000 California professional school employees (5). Established in 1995, the CTS gathers extensive information on breast cancer risk factors and has followed participants for breast cancer incidence since 1995. Designed to build on the statewide population analysis, this study has identified over 1,500 cases of breast cancer occurring in the cohort in the first four years since its inception. Similar to the statewide population analysis, GIS has been used to pinpoint the address of all study members living in California at the time they joined the cohort and statewide census and environmental data are being used to characterize sociodemographic and potential environmental contamination. With regional patterns of breast cancer incidence similar to what is observed in the statewide population, the CTS provides a unique opportunity to incorporate both individual-level and population-level risk factor information in an examination of the factors underpinning the observed geographic patterns in breast cancer rates.

Finally, the last level of analysis and the smallest of the three studies, Census block group (1990) focuses on molecular markers of exposure as they might relate to potential urban-based sources of environmental exposures.Nested within the CTS cohort, this pilot study is designed to evaluate a number of questions regarding exposure assessment in these types of studies. Approximately 150 urban-based and 150 rural-based cohort members are enrolled in this study in which participants provided a 24-hour urine sample and answered a number of questions regarding diet, environmental xposures and residential history. Its objectives are to:(1) evaluate whether biomarkers of selected exposures of emerging interest in breast cancer etiology (e.g.,traffic, pesticides) differ in urban versus rural women; (2) evaluate the degree to which these biomarkers correlate with GIS-derived measures of these exposures; (3) evaluate the degree to which self-reported exposures correlate with levels of urinary biomarkers; (4) evaluate the ability of CTS members to provide a life-time residential history; and (5) evaluate the impact that residential mobility may have on studies examining regional patterns of reast cancer that rely on address at diagnosis.

Initial analyses have focused on examining relationships with the pattern of factors influencing regional differences in breast cancer rates in California and on environmental data for agricultural pesticide use and air pollution. The question of whether breast cancer rates are higher in areas of agricultural pesticide use, as one example, has been of particular interest to many Californians. As the largest agribusiness state in the U.S., pesticide use is quite intensive in some areas but not uniformly so (see Map 1). California's Pesticide Use Reporting (PUR) system is quite unique, with detailed information on all agricultural pesticide applications by date of application, by method, by crop, by chemical and by amount. These, as illustrated in a sample map for San Joaquin County (Map 2), are reported at the level of approximately one square mile (based on the Public Land Service System designation of Township, Range and Section). This information provides us with a opportunity to evaluate pesticide use density by block group (see Map 3) for studies of population rates, or within a specified distance for individuals in the CTS cohort study (see Figure 2). Our environmental scientists have reviewed the more than 850 different chemicals reported by the PUR and grouped them for these analyses into agents thought to be probable or possible carcinogens, agents thought to be endocrine disruptors and agents with evidence of producing mammary tumors in laboratory studies. We prioritized the pesticides for study based on a variety of factors including cancer classification and potency, environmental fate and persistence, amount used, the distribution of potentially exposed populations and correlations among chemicals. Validation studies are also under way to assess exposure potential for people living near treated fields.

For these kinds of efforts, new and evolving GIS tools combined with traditional epidemiologic study designs offer the opportunity to better assess environmental influences for a number of disease outcomes. As this work is currently in development, it will not represent definitive evidence for environmental influences on health but it should provide us with some initial steps to better formulate our study questions and to better address the kinds of questions we get from the public.

When results from this study become available they will be posted on our website at www.ehib.org.

References:
1. Parkin DM, Muir CS, Whelan SL, et al, eds. Cancer Incidence in Five Continents. Vol VI. Lyon, France:World Health Organization, International Agency for Research on Cancer, 1992 (IARC Publication No.120).
2. Perkins C, Morris C, Wright W, Young J. Cancer incidence and mortality in California by detailed race/ethnicity, 1988-1992. Sacramento, CA: California Department of Health Services, Cancer Surveillance Section, 1995.
3. Robbins AS, Brescianini S, Kelsey JF. Regional differences in known risk factors and the higher incidence of breast cancer in San Francisco. J Natl Cancer Inst 1997; 89: 960-5.
4. Clarke CA, Glaser SL, West DW et al. Breast cancer incidence and mortality trends in an affluent population: Marin County, California, USA, 1990-1999. Breast Cancer Research 2002;4:R13.
5. Bernstein L, Allen M, Anton-Culver H et al. High breast cancer incidence rates among California teachers: results from the California Teachers Study (United States). Cancer Causes Control 2002;13:625-35.

Back to the top

Last Update 04.04.03