Disease Detection: The Effects of Contaminants on Human Health

Understanding the effects chemical contaminants have on human health is not a simple or straightforward process. One reason for this challenge is because there are approximately 15,000 chemicals produced and less than half of those have ever been tested to determine their impact on human health. Add to the quantity of chemicals, the many possible interrelationships between all of the chemicals and the unique genetic and environmental circumstances of every human, predicting or confirming that a disease has been caused by a chemical contaminant is extremely complicated. This page will describe contaminant-based disease detection, discuss epidemiology, describe the weight-of-evidence cancer categories, and provide additional web links.

Contaminant-Based Disease Detection

One of the biggest fears of exposure to contaminants is cancer. Some contaminants are known carcinogens while others are only suspected. An example of a known carcinogen is asbestos; a suspected carcinogen is DEHP [di(2-ethylhexyl)phthalate] - a component of plastic. For many contaminants, scientists do not always know how much exposure will cause harm or disease. Scientists do not know about the risk to humans because many studies are conducted on rats and mice rather than on humans. (It would be unethical to purposefully expose humans to high doses of these chemicals.) In human studies, knowing how much exposure has occurred is a significant challenge.

Environmental contaminant exposures can be related to other diseases in addition to cancer. Such diseases include immune deficiencies (from polychlorinated biphenyls or PCBs), asthma (from particulates), reproductive disorders (from DEHP), neurological (brain and nerve) disorders (from lead), blood disorders (from benzene), liver disorders (from dioxins) and other health problems such as endocrine, gastrointestinal or skin conditions. There are many other conditions that are suspected as being caused by contaminants, but physicians and scientists are not certain of the cause. Most physicians are not trained in the complex analysis of chemical exposures. In addition, special testing by select laboratories is needed to test blood or tissue for many of the suspected causative agents (chemicals). This uncertainty can make citizens anxious and angry that their conditions are not being considered seriously. It is important to describe your concerns to your physician and ask for a referral to a specialist.

Some people might think hazardous chemicals should be eliminated entirely. However, it is difficult, to weigh the possible harm of exposure versus the need for the chemical. For example, asbestos is an important substance in fire protection and cannot be completely eliminated from use. Therefore, means to protect human health from exposure has become a priority.

The Centers for Disease Control and Prevention (CDC) has a department called the Agency for Toxic Substances and Disease Registry (ATSDR). ATSDR will actually conduct exposure investigations using an Exposure-Dose Reconstruction Analysis (EDRA). ATSDR takes environmental samples and conducts a computer model to estimate exposure levels. The State Health Department may also conduct these analyses under contract with the ATSDR.

ATSDR typically requires additional information in the following categories (ATSDR, 1994):

  1. Contaminant concentrations in all off-site media to which the public may be exposed;
  2. An appropriate detection limit and level of quality assurance/quality control (QA/QC) in samples to ensure the resulting data are adequate for assessing possible human exposures;
  3. Discrete samples that reflect the potential range of exposure of the public;
  4. Surface soil and sediment samples not deeper than 3 inches;
  5. More extensive biota studies, and analyses of edible portions only;
  6. More ambient and indoor air sampling; and
  7. Lists of physical hazards and barriers to site access.

Finally, citizens can help look for disease clusters themselves. A cluster is an unusually high number of health events found at similar times in a given location. If you suspect you or your community has been exposed, it is important to gather information to help identify a cluster. The ATSDR can provide guidance about the kind of environmental data needed to assess public health related to toxic substances. This website provides the details of an environmental public health assessment.


Disease clusters are often of interest to communities that have hazardous waste sites. It often seems that there is more cancer in the community than there should be. Public health departments and epidemiologists can help a community identify if there is a link between disease and environmental exposure. "Epidemiology is the study of the distribution of disease, or other health-related states and events in human populations, as related to age, sex, occupation, ethnicity, and economic status in order to identify and alleviate health problems and promote better health" (EPA). However, epidemiology is challenged by the uneven ways diseases and illnesses affect populations. Environmental epidemiologists have the added complexity of determining if the people that have the disease were ever exposed to an environmental contaminant.

Some of the unevenness in exposure can be related to individual susceptibility to contaminants, "dose" of exposure (length, frequency, amount), and geographic differences that can include climate or populations relocating. Therefore, tracking environmental diseases in populations is a challenge. It is especially challenging to understand how much exposure to a contaminant any individual has received. For example, in a location that is affected by air pollutants, some individuals may reside in the community all day while others may commute away from the area for work. Distance from the pollutant source may make a difference in the response to an exposure, as can age, sex, susceptibility or the presence of other disease processes. Combinations of unrelated exposures may also complicate epidemiology. Such combinations might include smoking, diet or workplace exposures that can add to an individual's risk from an environmental exposure.

Environmental epidemiology may begin when an unusual number of health problems (cluster) is identified in a population that seem unexplained. Epidemiology may also begin by recognizing that an exposure has occurred, for example, discovering arsenic in a population's wells. When the health problems are identified first, epidemiologists typically work "backwards" to find the people with the disease and obtain detailed information about past exposures. When an exposure is identified, epidemiologists may observe the population's health into the future to determine if disease occurs.

It is helpful to understand some of the terms that epidemiologists use in their attempts to understand the epidemiology of environmental hazards.

Incidence Rate: An incidence rate is the change in health status in a population over time. A rate is generally measured and reported annually. For example a death rate or birth rate is the number of deaths or births in a year. In epidemiology, the number of "cases" in a year would be the rate. It is important to know how the rate was calculated because rates use different population sizes for reference. For instance, the difference between 5 cases in 1000 population is quite different from 5 cases in 1,000,000 population. Rates allow comparisons between populations because the population size is controlled. If people recover from the disease or die from the disease, they are still counted in the incidence rate. One example of a rate is the breast cancer rate in the US population - 134 cases/100,000 in 1999 http://www.cdc.gov/cancer/npcr/npcrpdfs/USCSreport.pdf. This figure means that for every 100,000 people in the US in 1999, 134 of them had breast cancer.

Prevalence: Prevalence is the percentage of a disease in a population. This measure only provides a description at a particular point in time. The prevalence will decline if people recover (or die) from the disease because of the "snapshot" nature of the prevalence measurement. For example, the prevalence of cigarette smoking among those 18 years of age and older was 25% in 1990, and has declined since then http://www.cancer.org/downloads/STT/CAFF2003PWSecured.pdf. This figure means that among all people 18 and older in the US, 25% of them smoked in 1990.

Mortality Rate: Similar to an incidence rate, a mortality rate provides a rate of deaths (mortality) over time (typically a year). As with incidence rate, it is important to know the population size, such as 100,000 or 1,000, 000. Often the mortality rate will be described in relation to a particular disease, such as the mortality rate for breast cancer. For example, the breast cancer mortality rate for African American females is 37.3/100,000 population from 1992-1999 http://www.cancer.org/downloads/STT/CAFF2003PWSecured.pdf. This figure means that every for every 100,000 African American women, 37.3 of them died from breast cancer every year between 1992-1999.

Epidemiology often considers relationships between environmental exposures and disease. Establishing the cause of disease is difficult because exposure to the contaminant is so uncertain. To try to establish causal relationships, exposure must be certain. Animal models are used for this purpose and the animal studies must then be "translated" or extrapolated so they apply to human populations. These studies have wide variations in exposure conditions as well. Therefore, the EPA uses the "weight of evidence" to determine exposure levels that will likely cause (or not cause) disease. This process is described below.

Weight of Evidence for Cancer Risk Assessment

The EPA reviews the scientific literature and studies to determine how to classify chemicals in terms of their potential for causing cancer. These classifications are determined based on the "weight of evidence" that examines laboratory, animal and human epidemiological studies. In 1986 an alpha-numeric system was used and many web sources still display this system. This system considered primarily the chemical's ability to cause tumors. In 1996, a three-category system was approved. It considers additional information, such as knowledge about similar chemicals and chemical behavior as part of the classification. Both systems are described below.

The new system uses the categories of Known/Likely, Cannot Be Determined, or Not Likely. A narrative description accompanies the category that describes the evidence available and how the conclusion was reached to categorize the chemical in a certain way. Chemicals may be Known/Likely carcinogens for an inhalation route of exposure, but Not Likely for an oral route of exposure. This narrative system helps make such distinctions clear. Limitations in the data are also part of the narrative. A lengthy EPA document about this change can be found on their website.

  • Known/Likely: The data are convincing that the agent may cause cancer in humans. Tumor evidence and epidemiologic data are used. A chemical may be classified Known/ Likely for one route of exposure, but not another.
  • Cannot Be Determined: Data about tumor effects are incomplete, but the evidence does not confirm the risk. If there are no data, the agent would be classified here.
  • Not Likely: There is adequate evidence to demonstrate there is no basis for human cancer concern. Agents may be classified Not Likely if the evidence shows cancer in animals only at very high doses or by certain routes. A chemical may be classified Not Likely for one route of exposure, but not another.

The older system is described at the Penn State web site. These categories are:

  • Group A: Human Carcinogen. Used when there is sufficient evidence from epidemiologic studies to support a causal association between exposure to the agent and cancer.
  • Group B: Probable Human Carcinogen. Used when epidemiologic data are limited and when the weight of evidence of carcinogenicity from animal studies is sufficient.
    • Group B1: Used when agents have limited evidence from epidemiologic studies. Evidence of carcinogenicity in animals is considered supporting evidence of risk in humans.
    • Group B2: Used when there is sufficient evidence from animal studies but inadequate or no evidence from epidemiologic studies.
  • Group C: Possible Human Carcinogen. Used when there is limited animal evidence and no human evidence to support carcinogenicity. Problems in experimental design or statistical significance make the evidence less certain.
  • Group D: Not Classifiable as to Human Carcinogenicity. Used when there is inadequate evidence in humans or animals about carcinogenicity.
  • Group E: Evidence of Non-Carcinogenicity for Humans. This group is used for agents that show no evidence for carcinogenicity in at least two animal tests in different species or in both adequate epidemiologic and animal studies. It is based on available evidence and is not a definitive conclusion that the agent will not be a carcinogen under any circumstances.
<< Exposure Pathways Steps to Cleanup >>

To find resources related to HEALTH, click on the following icons:

EnviroTools.org is created at Michigan State University and is sponsored by National Institute of Environmental Health Sciences.  If you have comments or questions, please contact us.