observed in populations. (Appendix A discusses these and other related issues in greater depth.) Researchers need to design environmental health studies that will provide adequate measurement, classification, and reporting of data on race, ethnicity, and relevant socioeconomic variables and to develop improved methods of descriptive, analytic, clinical, and molecular epidemiology that are accurate and practical for investigating relationships between environmental exposures and disease in low-income and minority populations.
One tool that can help epidemiologic studies is geographic information systems (GISs). Geographic data can be used to relate the location of a known or a suspected environmental health hazard to public health trends and racial distributions, among other factors. Because GISs can provide powerful summaries of relationships that may be lost in numerical analyses, they have been found to provide clues to relationships that can then be investigated by quantitative techniques (Elliott et al., 1996). Such techniques can also merge environmental and public health data collected from many different sources.
Because a central focus of environmental justice is on disparities among racial groups, it is important that studies and research take account of race and socioeconomic factors. However, the committee took cognizance of concerns being raised about conventional definitions and classifications.
The use of racial and ethnic categories for health surveillance is often confounded with other differences, such as geography, economic status, culture, lifestyle, or behavior. In his synopsis of a workshop, Health Surveillance and Communities of Color, sponsored by the Centers for Disease Control and Prevention and the Agency for Toxic Substances and Disease Registry, Rabin (1994) stated that surveillance needs to "pay more careful attention to differentiation within minority populations regarding year of migration, family status, income, age, daily work habits, religion, [and] media habits" (p. 45).
The categorization of individuals simply by race ignores other variables that can lead to valuable insights into predictors of risk. Many groups, including the American Anthropological Association and the Institute of Medicine (IOM) Committee on Cancer Research Among Minorities and the Medically Underserved, have been critical of the use of the term "race" in health research, primarily because the term implies the existence of distinct human subgroups that differ fundamentally in biological makeup and origin and ignores the tremendous heterogeneity within such groups (American Anthropological Association, 1997; Institute of Medicine, 1999). In reality, "genetic diversity appears to be on a continuum, with no clear breaks delineating racial groups" (Marshall, 1998, p. 654).
The vast majority of health research on human population groups in the United States, however, has categorized populations according to familiar terms such as "white," "African American" or "black," "Hispanic," ''Asian American," and other terms. Such categorizations may be reinforced by federal research