Love Canal Follow-up Health Study - April 1999

Study Plan and Progress

In 1997 the New York State Department of Health (DOH) was awarded a grant from the federal Agency for Toxic Substances and Disease Registry (ATSDR) to conduct a follow-up health study of the families who lived near Love Canal before 1979. With help from the community, DOH selected a panel of leading experts to give advice about technical aspects of the study. This panel, named the Love Canal Expert Advisory Committee, met twice in 1998 and offered several suggestions to help improve the study. Even though the study proposal is already approved and funded by ATSDR, changes are possible at this time because the full study will take six years. Since ATSDR can fund the study for only three years at a time, we will submit additional plans and a new budget this June. The Love Canal community now has an opportunity to participate in upcoming decisions about the study's direction for the next three years before DOH asks ATSDR for continued funding.


Planning for the study started in 1988, when we met with the community to see which health outcomes they were concerned about. In addition to community concerns, we looked at which health effects are believed to be associated with exposure to chemicals disposed of at the Canal. Using this information and common epidemiologic techniques, we developed a study proposal which was ultimately funded by ATSDR. There are two phases to the study. The first phase of work is nearly completed. We began with an existing registry of about 6300 people who were interviewed for earlier health studies (people who lived near the Canal from 1940 up to and including June of 1978). We continued efforts to locate and contact all of this group. To date, 95% of this group have been located. We informed them about the study and asked them to verify their identifying information (birthdates, social security numbers, etc). We also asked to be informed of any new addresses during the study so we can keep them apprised.

Phase I

Some health outcome information is already available to us because it is reported to DOH routinely. This includes birth and death records, cancer incidence, and birth defects reports. The Department is in the process of checking its list of Love Canal residents from the registry against the larger list of all New Yorkers who experienced these events. We will look for unusual trends in birth experiences, cancer incidence, and causes of death. We will also see if the rates of cancer and death among Canal residents are different from others in Niagara Falls who didn't live near the Canal, Niagara County residents, and upstate New Yorkers. This is called passive data collection because it doesn't require any participation on the part of the residents.

Phase II

In addition to offering suggestions about how to interpret the findings from the passive data collection, the Committee is helping us to plan the next phase of the health study. We will "actively collect" health data by identifying which residents were most likely to have higher exposures, determining health outcomes for which they might be at increased risk, and contacting them about their health experience. We will discuss options and approaches with the Committee in April, and keep you informed about what was discussed.

Exposure Assessment

One of the most challenging tasks faced by the health study team is to describe each resident's likelihood of past exposure to chemicals that were disposed of at the Canal. If we were currently measuring exposure to people living near a landfill, we could go into the homes and measure chemicals in the air, water, or dust, and we could sample soil and vegetables from their yards (environmental samples). We also have tests available to measure chemicals inside the body of someone who might have been exposed to chemicals. For example, lindane and other compounds can be measured in blood or in mother's breastmilk (biological samples).

Of course, we can't go back in time to collect these samples. We have some information about environmental conditions near the Canal from sampling that began in 1977. We are placing all available environmental sampling results (about 14,000 samples of air, soil, and water from state, private and federal agencies) into a unified database. The data collected from 1978 to 1980 were analyzed using equipment and procedures that, although appropriate for the time, are considered outdated today. Data collected from 1986 to 1988 are considered more reliable measurements of persistent chemical contaminants, but the samples were collected outside of the remediated area and do not include samples taken from inside the fence. This limits the usefulness of these measurements for assessing the likelihood of exposure before the remedial work was completed. However, the data provide information about chemical concentrations at a given point in time, and about relative concentrations in different areas at the same time.

In 1978 the Health Department tested blood for more than 4,000 people who wished to be tested. Routine medical screening methods (blood chemistry and liver function tests) were used and copies of the results were sent to the individuals' physicians for interpretation. At the time analytical methods had not been developed for many of the Canal chemicals. As we did with environmental samples, we froze any remaining specimens when the tests were done. We archived approximately 1,800 vials of blood. The Expert Advisory Committee suggested we analyze these samples for persistent Canal chemicals and use the information to help assess individuals' likelihood of exposure. Residents who lived inside the area now fenced did participate in this testing program, and the samples were collected at the time of possible exposure. Running new tests on old samples raises questions about how the samples were stored over the years and whether the individuals involved would object to such a procedure. The samples are now being evaluated at the Wadsworth Center for Laboratories and Research to see if this procedure is possible before we address the larger questions of residents' preferences.

We reviewed a variety of information sources to learn more about possible exposure, such as analytical data and trial testimony. Many reports showed how some people in the area came in contact with Love Canal chemicals, and this tells us how others in the area might also have been exposed to chemicals. Ways people could have been exposed include:

  • students walking to the 99th Street school or across the school property
  • children playing on the school property when chemicals had surfaced
  • people playing or swimming in the open canal after disposal began
  • air transport as evidenced by odor complaints during open dumping
  • chemicals seeping into people's yards and homes, often identified by odor complaints.

We decided to use broad exposure categories instead of trying to quantify each person's exposure to each of the chemicals in the Canal. Each resident will be assigned a "likelihood of exposure" (high, medium, or low), as compared to other Canal residents.

Given all the possible pathways, exposure will be determined by years of residence, address, gender and age at the time of residence. This allows us to account for changing conditions at the Canal over time (open dumping, partially covered, chemicals surfacing) and to account for the ways people could come in contact with chemicals during various time periods. For example, some teenagers living near the Canal were likely to swim in it when the Canal was open, so they would get a high likelihood of exposure rating for this time period. A toddler living at the same address as the teenager would get a lower likelihood of exposure for the open period unless there was some information to suggest otherwise. Neither age group would get a high rating once the Canal was covered, unless they were attending the 99th Street school, lived very close to the Canal, or had some expected behavior that would make them likely to encounter chemicals. Using this method, which we call "behavior grouping", we propose to assign all 99th Street school children a high likelihood of exposure, regardless of their address. We intend to refine the classifications for residents as indicated by their address (Ring 1, etc.), available sampling data for that address, historical documentation of exposures, and computer modeling about where chemicals may have moved.

What is Exposure - Information Sheet

How you can help

  • Tell us your concerns
  • Provide health information if asked
  • Participate in the study design
  • Tell us if you move
  • Comment on this newsletter

Send comments to:

Center for Environmental Health
Flanigan Square, Room 316
547 River Street
Troy, NY 12180-2216
Fax: (518) 458-6402
Phone: (518) 402-7530

Understanding Epidemiology

Health study results are often criticized and their usefulness questioned by scientists, the press and the public. Press report and emphasize newsworthy studies--those which find new or unusual associations. Ironically, these are the most untested (and they might later be found erroneous). Scientists rely on the principle that observing something once or failing to observe it is only the beginning of the discovery process. They do not draw conclusions from a single study.

A newly found association needs more study or follow-up to verify its conclusions, but should not be interpreted as a finding by itself. Scientists do not draw conclusions from a single study, especially an epidemiology study, because of the nature of the field (conditions aren't clearly controlled or understood). Discussion and debate about the meaning of a particular study is normal. It helps identify the limitations of the study and other unknown factors that should be considered. Discussion, including critique of reported studies, helps strengthen the reasoning and should not be interpreted as an inability of the scientific method to discover causes of illness and prevent future disease. Even though the press and the public are involved in the discussions, a study's limits are often unstated. The limitations of the study design or the framework in which scientists view their findings are not obvious.

A health study can be more accurately interpreted and explained if people know the current scientific understanding and the assumptions and errors that are part of the human health study design. Knowing what human health studies in general can and can't tell us also helps us to interpret the findings accurately without expecting something that is not possible. People who know the basics of epidemiology will be better prepared for entering into the discussions at any stage of the study, and they will be able to effectively participate in the study's design.

There are two primary sources of information about health effects from chemical exposure: animal and human studies. Animal studies provide information about health effects on the animal after it is exposed to the chemical in question. Scientists expose a group of animals to a measured dose of a specific chemical. A control group of the same type of animal is treated in exactly the same way, except that the controls are not exposed to the chemical being studied. Observing the differences between the study group and the control group is fairly straightforward. Any differences are generally accepted as being caused by the chemical exposure. What is not so straightforward is predicting whether human bodies will react the same way if exposed, especially at different levels. Typically, doses used in animal studies are much more concentrated than what humans take in from their environment. Animal studies are valuable, even though they provide limited information, because they can flag possibly harmful substances in a way that studying humans cannot. Human studies are limited, too, but for different reasons. Studies on groups of humans (epidemiology) do not confine all the study participants to the same environment. Participants each have their own activities and exposures including other factors which may affect the results. So observing a difference between the study group (who were exposed to a chemical) and the control group does not mean the chemical in question is the cause. Epidemiology study results are usually reported as

Of course, the purpose of human studies is to discover possible causes of disease and the ways in which they can be modified. Many "causal factors" can be identified by critically thinking about the associations that were found. Eliminating or reducing one or more of these factors can sometimes reduce the occurrence of disease, even before the entire process of disease causation is understood.

Epidemiologists can use one of two approaches to their task:

Observational studies can be either descriptive or analytical. A descriptive study simply describes the occurrence of health effects in a population. It is a first step in the process of epidemiological investigation. Analytical studies analyze the relationships between health effects and other factors. There are four analytical study designs: ecological, cross-sectional, case-control and cohort. While no single study could prove causation, the cohort study is the most powerful of the designs for identifying possible causes. The Love Canal health study is a cohort study because it follows a defined group of people (the cohort) through time to see if they develop disease differently when exposed. Even though it is a powerful tool, it still has limitations which need to be considered when interpreting its findings. No single study can be conclusive. A positive association between exposure and health effects can point to possible causes of disease, but the finding is only a theory (also called a hypothesis) needing further investigation.

The finding could be a "false positive" (finding a relationship that really does not exist) if there is:

Another limitation in epidemiology is the "false negative" finding. There is always the possibility that an exposure really caused a health effect, but that researchers did not observe it. Usually, failure to observe an effect should not be interpreted to mean there is no association. It merely means that an association was not observed. We can increase the likelihood that an association will be seen by increasing the accuracy of our measurements of exposure and health effects. A large number of exposed people also increases the likelihood of seeing a real association. That is why it is important for as many exposed people as possible to participate in the study. If the health effects are greater, they are easier to see as well.

Epidemiologists account for these problems as much as possible by carefully designing the study and using statistical methods to help them interpret their observations. They also ask these questions:

  • Were the observed health effects present before the exposure?
  • Does it make biological sense that the exposure could cause the health effect?
  • Is the result consistent with those seen in other studies?
  • Is there a strong association between the exposure and the health effect?
  • Do people with higher exposure tend to have more health effects?
  • If the exposure is removed, does this lower the chance of developing health effects?
  • Was the study well designed?

As you might expect, different researchers might arrive at different opinions about cause after answering these questions. The debate process helps to refine our understanding of possible causes of disease and ways to prevent further illness. Scientists generally accept the idea that several consistent studies are needed to suggest causation.

The Love Canal study is a retrospective cohort study. Retrospective means that scientists are looking back in time, in this case, to exposures that occurred between 1940 and 1978. We are greatly limited by our ability to measure those exposures. However, the cohort study design is the strongest observational study available. The Love Canal Expert Advisory Committee is concerned that the study may not be powerful enough to detect health effects from exposure to Love Canal chemicals. We won't know for sure whether it will detect any health effects unless we conduct the study. The Committee feels that we should proceed with the study, being careful not to overstate or understate the reliability of the results. They made suggestions to improve our chances of finding health effects, such as trying to include 99th street schoolchildren in the study, and encouraging the community to participate. They are also helping us to make the most accurate exposure estimates possible.

While the Love Canal study can not prove that exposure to Love Canal chemicals caused certain health effects, we might observe differences in the health of people who lived near the Canal or in their offspring. If this happens, even if we are not sure about causation, we can still inform Love Canal families that they seem to be more likely to develop the observed health effect. This information can help residents and their health care providers take appropriate preventive action. Another possibility is that we will not see a difference between likelihood of exposure and the health effects selected for study. This will

Expert Advisory Committee

Investigating the health effects in a community exposed to toxic waste requires a multi-disciplinary approach, bringing together expertise in medicine, pharmacology, epidemiology, statistics, and engineering. We are fortunate to be guided by a prestigious advisory board with a broad range of talent.

  • Patty's Industrial Hygiene and Toxicology and the ILO Encyclopaedia of Occupational Safety and Health.
  • Everyone's Backyard " and in CHEJ monographs. He has served on numerous boards and committees, including several concerned with Superfund activities and environmental justice.