Epidemiologist Ann Aschengrau admits to never having fulfilled her childhood dream of becoming a "mad" scientist surrounded by beakers and test tubes in a big, bubbling laboratory. However, she derives great satisfaction from her career in academia, particularly from mentoring future epidemiologists and conducting research on the health effects of environmental pollution. Her work in environmental epidemiology began with a doctoral thesis on the risk of birth defects among Vietnam War veterans who may have been exposed to the herbicide Agent Orange. It has continued with investigations of the potential health effects of many other environmental exposures, including the risk of breast cancer among women exposed to pesticides and the risk of poor pregnancy outcomes among women exposed to drinking water contaminated with tetrachloroethylene.
Research summary
Tetrachloroethylene, a solvent widely used in dry cleaning, degreasing, and other industrial operations, has become a common contaminant of drinking water. Also called perchlorethylene or PCE, it is a nonflammable liquid that evaporates easily. While studies of adult PCE exposure have demonstrated numerous effects to the nervous system, little is known about the long-term neurotoxic effects of exposures during pregnancy and childhood. In her current SBRP project, Ann Aschengrau is testing the hypothesis that PCE found in the public drinking water supplies in Cape Cod, Massachusetts, is associated with an increased risk of nervous system abnormalities among men and women exposed during gestation and childhood. The source of the water contamination was a vinyl liner that was applied to hundreds of miles of asbestos cement water distribution pipes from the late 1960s through 1980.
The epidemiological study design Dr. Aschengrau is using to test this hypothesis is known as a "cohort study." In a cohort study, groups of exposed and unexposed individuals are followed over time to determine their incidence of disease. The current study includes a group of 3,000 Cape Cod residents who were exposed to PCE-contaminated drinking water in eight towns between 1969 and1983 and a comparable group of residents from that time period who were not exposed. The exposed residents include individuals with both prenatal and childhood exposures and their older siblings, who were only exposed during childhood. Dr. Aschengrau and her team hypothesize that exposed individuals will have increased rates of nervous system effects. They are measuring these effects by looking at evidence of vision disorders and learning and attention developmental disabilities; performance on neuropsychological and vision tests; and neuroimaging studies, which can reveal brain disorders and dysfunction. They expect their results to show that the nervous system effects are time dependent, so that individuals who had both prenatal and childhood exposures will show greater effects compared with those exposed only during childhood.
If you have a question for Ann, email it to This e-mail address is being protected from spambots, you need JavaScript enabled to view it
Q. My wife was a child during her exposure to the contaminated drinking water at camp LeJeune. We have three children, the first was diagnosed with Oxidative Phosphorylation disorder. Have there been any studies looking at the offspring of females who had long term exposure to tetrachloroethylene (PCE) and associated by-products as children?
A.There are a few studies on the health effects of prenatal and early childhood exposure to tetrachloroethylene, but, to the best of our knowledge, no studies on possible health effects among their offspring. This means that there have been a few studies to determine if there are ill effects among people like your wife but there have not been any studies (yet) on individuals like your children. So far, our own studies on
Q. In trying to understand whether there is a relationship between PCE exposure and nervous system abnormalities, how do you take into account other variables that could explain the results? These might include things that you can measure (e.g. smoking) as well as things that you might not even have thought of or are more difficult to measure (genetic differences; socio-economic factors; diet; exposure to other chemicals). It seems extraordinarily difficult—almost impossible—to consider all of these things.
A. Epidemiologists spend a great deal of time designing and conducting studies that address confounding—the technical term for the problem you are describing. The first thing we do is make sure that our exposed and unexposed study groups are as comparable as possible. In the PCE study, we did this by selecting both exposed and unexposed mothers from the same geographic area and time period. This selection process made the groups nearly identical on race, socioeconomic status, and parental age.
Epidemiologists also address the problem of confounding by using sophisticated statistical techniques. For example, in the PCE study we collected information on smoking during pregnancy and used this information to “adjust” the associations in the data analysis. While elaborate computer programs allow us to control for many confounding variables simultaneously, some confounding factors cannot be controlled in this manner because it is impossible to gather the necessary information. For example, we could not collect information on prenatal folic acid consumption, because the women could not accurately recall this dietary information for the pregnancies we were studying, which had occurred 20 to 25 years earlier. While the lack of these data could be considered a flaw in the PCE study, it would only have biased the results if women exposed to PCE had eaten a different diet than unexposed women. We believed that this unequal pattern was unlikely, given the known similarities between the groups. Thus, confounding is less common and easier to control than you might think.
Q. How do you conduct exposure assessments?
A. We assess our subjects’ exposures to PCE in public water supplies using a model developed in the early 1990s by Thomas Webler and Halina Brown at the Center for Technology, Environment, and Development at Clark University in Worcester, Massachusetts. The model, which is a series of mathematical equations, describes how PCE leached from the vinyl lining of pipes into the water distribution system on Cape Cod. It assumes that the concentration of PCE in a water pipe liner dropped exponentially over time after the pipes were installed. The rate of decline is thought to have varied with the volume of water flowing through the pipe; greater volumes reduced the amount of PCE more quickly. Thus, the input variables to the model are:
- those that determined the volume of water moving through a pipe — the pipe length and diameter;
- those that determined what volume of water was reaching the subject residence — the location of residences along the pipes and the location of subjects’ residences; and
- those indicating the timing of the subjects’ exposures — the years that the pipes were installed and the beginning and ending years of subjects’ residency.
For each subject, the model output is termed a "relative delivered dose," or RDD. The RDD is defined as the mass of PCE that entered the home as a solute in drinking water during a specific period of time. Because of the complexity of the water distribution system, Webler and Brown emphasize that their model serves only to rank the relative exposures that subjects received, not calculate the absolute doses.
To use the Webler-Brown model in our early studies, we first gathered information from the local water companies on the vinyl-lined asbestos cement pipes, including their location, length, diameter, and year of installation. Next, we interviewed the study subjects to obtain their residential history over the past 40 years. This history included the exact address; names of cross-streets and landmarks; and the years of residence of all Cape Cod addresses. Next, we mapped the locations of all subjects’ Cape Cod residences in relation to the locations of the vinyl-lined pipes. (In the early days, we actually mapped the residences on paper maps, but now we do the mapping entirely by computer using a Geographic Information System.) Last, we ran a computer program that incorporated Webler and Brown's equations and produced the RDD.
Over the years, we found that Webler and Brown's model had many simplifying assumptions that likely led to inaccuracies about the direction of the water flow, particularly in the areas where the geometries of the water distribution pipes were highly complex. Our more recent studies have incorporated the Webler-Brown PCE leaching model into a software package for modeling water flow called EPANET. EPANET was developed by the U.S. Environmental Protection Agency to assist towns and researchers in both maintaining and evaluating water distribution systems. Given the proper information, the program simulates the flow of water through the thousands of pipe segments in a town's entire water distribution system.
For each town with vinyl-lined pipes, we created a representation of water sources, pipes, and water users (residences) in EPANET based on information from the town water departments and the Massachusetts Department of Environmental Protection. The simulation was based on system operating conditions that were typical of each town during the study period. The combined Webler-Brown/EPANET model then determined the amount of PCE that was delivered to each residence. This modeling technique reduces the number of assumptions required for assessing water flow and improves the accuracy of the exposure measure.
Q. How did you become interested in PCE exposures on Cape Cod?
A. I began working on Cape Cod when I was a newly minted Assistant Professor at the BU School of Public Health. For quite some time, the population in the Upper Cape Cod region of Massachusetts-in the towns of Barnstable, Bourne, Falmouth, Mashpee, and Sandwich-had been experiencing elevated mortality rates of lung cancer and excesses in the incidence (new cases) of breast, colon/rectum, lung, and blood cancers. In 1987, the Massachusetts Department of Public Health issued a Request for Proposals (RFP) to investigate possible environmental causes for these elevated rates.
David Ozonoff and I decided to respond to the RFP. The current Director of the Superfund Basic Research Program at the BU School of Public Health, Dave then was the chair of the Department of Environmental Health. Our goal was to provide the state Public Health Department and Cape Cod citizens with sound scientific information on the cancer risks associated with environmental hazards affecting this region. The Department of Public Health and the community group looking into the situation wanted the investigation to focus mainly on the pollution associated with the Massachusetts Military Reservation. As epidemiologists, we thought that it was important to conduct a comprehensive study that included all current and historical environmental hazards. Before we began, we reviewed many data sources. We looked at information about the area's hazardous waste sites and manufacturing activities. We reviewed data documenting contamination associated with the Massachusetts Military Reservation, the Barnstable Airport, the Canal Electric Power Plant, and pesticides applied by cranberry growers. We also looked at the quality of the public drinking water supplies. Ultimately, we decided to investigate more than 40 different sources of air and water pollution to which a significant portion of the population may have been exposed.
Dave had already been aware of the PCE contamination episode, because he had been called upon to help remediate the problem when it was discovered in the early 1980s. The PCE contamination warranted inclusion in the epidemiological study because PCE is considered a possible carcinogen, and there were high levels of PCE contamination in many of the towns. In addition, a considerable portion of the population had been exposed: PCE had been leaching from hundreds of miles of water pipes for more than a decade, exposing tens of thousands of residents.
Thus began our research on the health effects of PCE-contaminated drinking water. Over the past 20 years, our work has helped advance our knowledge of the health effects of environmental pollution in several ways. It has provided information on the long-term effects PCE can have on many different types of diseases. We have investigated PCE in relation to cancer, reproductive and developmental disorders, and neurotoxic effects. We are now particularly interested in contributing to knowledge of the fetal basis of adult disease, a new and exciting research focus in our field. Our research has also helped improve methods of assessing a person's exposure to historical sources of water pollution.
If you have a question for Ann, email it to This e-mail address is being protected from spambots, you need JavaScript enabled to view it
