1

Introduction

WASTE HAS BEEN A PRODUCT OF human activity since the dawn of human history. In the early stages of industrial development, workplace wastes were generated on site and swept, sent, or poured “away.” Occasionally, “away” meant literally out of the door and into the street or into local stoves or community incinerators. Later, waste materials were sold as fill for uneven ground and spread over large expanses of unsettled land that was subsequently urbanized. Waste oils were used as dust suppressants; unneeded products were poured down drains, or directly or indirectly dumped into streams, rivers, lakes, and oceans. Recognition that such wastes were potentially hazardous usually came long after they had been generated and distributed.

During the nineteenth century, improvements in basic sanitation, housing, nutrition, and sewage treatment substantially improved life expectancy throughout the industrial world by reducing deaths from such infectious diseases as tuberculosis, diphtheria, and pertussis (McKeown, 1976). Attention in the twentieth century has shifted to chronic illnesses, such as some kinds of cancer (NCI, 1990) and neurologic disease (Lilienfeld et al., 1989), that have become more common in industrial societies than before. Questions have come to be raised about the possible relationship of industrial waste and other aspects of modern life to chronic diseases.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 1 Introduction WASTE HAS BEEN A PRODUCT OF human activity since the dawn of human history. In the early stages of industrial development, workplace wastes were generated on site and swept, sent, or poured “away.” Occasionally, “away” meant literally out of the door and into the street or into local stoves or community incinerators. Later, waste materials were sold as fill for uneven ground and spread over large expanses of unsettled land that was subsequently urbanized. Waste oils were used as dust suppressants; unneeded products were poured down drains, or directly or indirectly dumped into streams, rivers, lakes, and oceans. Recognition that such wastes were potentially hazardous usually came long after they had been generated and distributed. During the nineteenth century, improvements in basic sanitation, housing, nutrition, and sewage treatment substantially improved life expectancy throughout the industrial world by reducing deaths from such infectious diseases as tuberculosis, diphtheria, and pertussis (McKeown, 1976). Attention in the twentieth century has shifted to chronic illnesses, such as some kinds of cancer (NCI, 1990) and neurologic disease (Lilienfeld et al., 1989), that have become more common in industrial societies than before. Questions have come to be raised about the possible relationship of industrial waste and other aspects of modern life to chronic diseases.

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 Part of our modern heritage is the increasing volume of waste created by all industrial societies. There also is an unprecedented concern over the potential consequences for public health and the environment caused by exposure to wastes that are deemed hazardous under a variety of regulatory regimes. Since the earliest days of industrialization, substantial volumes of wastes have been produced and sometimes disposed of in ways that could create problems for later generations. In the U.S. more than 6 billion tons of waste is produced annually—nearly 50,000 pounds per person (OTA, 1989). Some analyses indicate that in the U.S. racial and ethnic minorities are more likely than are non-minorities to live in areas where abandoned hazardous-waste dumps or operating waste disposal facilities are located (Bullard, 1990). One study noted that in communities with two or more commercial waste disposal facilities, the average minority percentage of the population was more than three times that of communities without such facilities (Commission for Racial Justice, 1987). In many industrial countries, a number of highly publicized episodes of pollution have made it clear that pollutants can migrate in complex and not completely understood ways. Accordingly, a variety of laws now require that public policy should provide for better waste disposal practices. The legacy of past practices, however, provides a series of difficult challenges to policy makers and scientists regarding how to analyze the public health and environmental effects of old methods of disposal, how to set appropriate policies to reduce harm in the future, and how much resources should be devoted to these issues. At the request of the Agency for Toxic Substances and Disease Registry (ATSDR), the National Research Council (NRC) convened the Committee on Environmental Epidemiology to review current knowledge of the human health effects caused by exposure to hazardous-waste sites and to suggest how to improve the scientific bases for evaluating the effects of environmental pollution on public health, including specifically the conduct of health assessments at Superfund sites. With additional support from the Environmental Protection Agency (EPA), the Committee also is examining the role of state health departments in generating relevant information on this topic. This first report of the committee reviews and assesses the published scientific literature on health effects that could be linked with exposure to hazardous-waste disposal sites, and makes recommendations about major data gaps that need to be filled as scientists go on to answer important questions in the field. A second report of the committee will identify research opportuni-

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 ties and issues in methodology for the general field of environmental epidemiology and will evaluate selected non-peer-reviewed reports on the subject of the epidemiologic study of hazardous wastes. This literature includes such sources as state health department reports and selected technical reports from the legal literature. While not accessible in the peer-reviewed literature, such reports can also be found in recent court decisions in which evidence about hazardous-wastes sites has been extensively reviewed and is at issue. To the extent feasible, the second report also will evaluate emerging reports from a variety of newly available international sources that bear on these questions, such as those from Eastern Europe (Environment and Health in Developing Countries, 1991). This first report, to be consistent with the sponsors' requests, focuses on an evaluation of the published literature on the health effects of exposures from hazardous-waste sites. Because of this limited scope and also because a number of other NRC committees are concerned with environmental issues, the Committee on Environmental Epidemiology is excluding from its consideration dietary factors and the effects of radiation, including the hazards of exposure to radon, low-level radioactive waste contamination, and electromagnetic fields. The first section of this chapter defines environmental epidemiology. The second section discusses conventional views of statistical significance and principles for inferring causation based on epidemiologic evidence. After that, the principles of statistical inference are evaluated in the context of constraints associated with the litigious and controversial world of hazardous-waste sites and toxic torts. Toxic torts are among the fastest growing field of litigation involving legal claims of alleged injuries caused by exposure to toxic chemicals. The next section describes the historical context for the committee's work. The chapter concludes with an outline of the rest of this volume. ENVIRONMENTAL EPIDEMIOLOGY In recent years the term “environmental epidemiology” has seen extensive use, although it has not been well defined. For example, Report 27 in the Environmental Health Criteria series, published under the joint sponsorship of the United Nations Environment Program, the International Labor Organization, and the World Health Organization, was entitled Guidelines on Studies in Environmental Epidemiology (WHO, 1983). The report considered “[The use of] . epidemiological methods for assessing the effects of environmental agents on human health.” Similarly, neither a compendium published as Environmental Epidemiology in 1986 (Kopfler and Craun, 1986)

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 nor a didactic volume with the same title (Goldsmith, 1986) presented a definition of the field of environmental epidemiology. The recently established International Society for Environmental Epidemiology devised a definition in its charter in 1988: epidemiologic studies on the effects of environmental exposures of human populations. The Committee on Environmental Epidemiology has adopted the following definition: Environmental epidemiology is the study of the effect on human health of physical, biologic, and chemical factors in the external environment, broadly conceived. By examining specific populations or communities exposed to different ambient environments, it seeks to clarify the relationship between physical, biologic or chemical factors and human health. One challenging question that confronts environmental epidemiologists is how to estimate the health effects associated with past patterns of disposal of hazardous chemicals and effects that could occur in the future as a result of continued or projected exposure from failures to clean up sites, or from proposed remediation plans. Investigating these problems is technically difficult, time consuming, and expensive (Ozonoff and Boden, 1987). As part of its project on environmental epidemiology, the committee elected to focus first on an evaluation of available scientific and technical literature that concerns the health effects of exposure to materials found in and issuing from hazardous-waste sites. In using this focal point, the committee has not restricted itself to sites officially listed under various state and federal laws, but has undertaken a broad review of available evidence on the human health effects that could be linked to exposures from materials at sites where disposal of hazardous wastes has taken place. The committee's members acknowledge that the published literature regarding toxic chemical waste disposal sites is limited and uneven and that profound methodological and practical problems attend the field, as others have noted (Grisham, 1986). However, the committee members believe that a deliberate and systematic assessment of current knowledge will provide a useful foundation for their later work in developing and extending the intellectual framework of the larger field of environmental epidemiology. EPIDEMIOLOGIC RESEARCH In general, epidemiologists conduct two major types of studies to assess relationships between suspected risk factors and disease: descriptive and analytic. Descriptive studies portray disease patterns

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 in populations according to person, time, and place and include time-series analyses and prevalence studies that analyze large sets of data and are usually used to generate hypotheses. Analytic studies include case-control (retrospective) and cohort (prospective) studies and typically test hypotheses. In case-control studies, comparable series of cases of a disease and controls drawn from the same population are investigated to determine past exposures that could have resulted in the development of the disease. In cohort studies, comparable series of exposed and unexposed persons are followed to ascertain the incidence of disease or mortality caused by disease in association with the exposure. This traditional delineation between descriptive and analytic studies has fostered the notion that distinct research principles apply to each type of study. In fact, both descriptive and analytic studies can generate and test hypotheses. It is readily apparent that studies of hazardous-waste sites pose some special practical and ethical challenges. Long-term cohort studies of continued exposures cannot ethically be conducted on persons who have reasons for assuming they are at risk of chronic disease as a consequence of exposure. For instance, persons living near most hazardous-waste sites have in common a measured or estimated exposure to toxic substances in the area. Researchers cannot both verify this exposure and expect people to remain near the sites and continue to be exposed. Moreover, at many sites, citizens groups and neighbors have provided the first information about the existence of a suspected health problem associated with exposure to hazardous wastes. Once suspicions are expressed publicly, residents often leave the area if they can, and the study becomes mired in public fears and expectations. Who can be expected to wait patiently for scientists to gather and analyze data when they fear for their own and their children's safety —even if these fears later prove unfounded? Because all the major methods of epidemiology are essentially observational and nonexperimental, drawing inferences about causation is considerably more difficult than it is for those controlled experiments that use random samples and controls. People move around, eat different foods, engage in different social and recreational activities, have different genetic backgrounds, and live their lives with the full diversity of the human experience. Yet, all of these factors can directly or indirectly influence their health at any given time. To sort out the relative role of such factors, epidemiologists, like other scientists who study human events, must rely on inductive methods for drawing inferences about their data. The committee acknowledges that experimental (e.g., toxicologic) studies and epidemiologic studies each have their strong points and

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 that they complement each other with respect to making causal inferences. To a large extent, all empirical scientists rely on inductive methods. Moreover, while one can frame and often answer precise questions experimentally, experimental constraints may make it very difficult to generalize from them. In this regard, continued support for epidemiologic studies constitutes a linchpin of public health research. CAUSAL INFERENCE As we expect to describe more fully in the second report, an optimal investigation of potential adverse health effects from hazardous-waste sites would proceed from an adequate assessment of past as well as current exposures to chemicals at a site (see Chapter 3 of this report) to the formulation of testable hypotheses of effects to be studied in a specific population. Then, an assessment would be made of adverse health effects in exposed and unexposed persons and would take account of all potential confounders. No study that fits this ideal has been published, and it seems unlikely that any such study could be conducted in the immediate future. Accordingly, the committee must rely on a combination of evidence from different sources to reach any conclusion in accordance with its mandate to estimate health effects associated with hazardous wastes. Figure 1-1 illustrates the types of information on which the committee has relied. Knowledge of potential exposures is derived from studies that characterize the substances present in or migrating from hazardous-waste sites. As discussed more fully in Chapter 3, these must be described in terms of their toxicity—including their carcinogenicity and other effects studied experimentally in animals; and where the knowledge is available, effects studied on humans. Information about the nature of toxic substances is derived from the general scientific literature. Knowledge of health risks to humans from potential exposures can be obtained from other sources, including, sometimes, related epidemiologic studies involving analogous exposures. For some chemicals such sources will include published studies of occupational risks, usually involving higher exposures than those in the general environment. For others, especially for airborne exposures, it will come from studies of the general effects of specific pollutants and may be extended to circumstances where such pollutants are emitted from hazardous-waste sites.

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 FIGURE 1-1 Sources of evidence for inferring whether exposures from hazardous-waste sites cause an impact on public health. Knowledge of symptomatology or disease occurrence has in some instances been derived from studies of populations exposed to hazardous-waste sites. Often, these have not described exposures accurately, or they have failed completely to identify a specific causal factor. Nevertheless, with the knowledge that is available about exposure elsewhere, and from the knowledge that some of these exposures can result in the observed symptomatology or diseases found in excess in those exposed to hazardous-waste sites when compared to suitable controls, sufficient indirect evidence of causality might have been accumulated to justify remedial action for purposes of protecting public health. In adopting the above framework, the committee does not follow the approach traditionally used by epidemiologists in deriving inferences of causality (Hill, 1953; USDHHS, 1976). Historically, discussions on causality have proceeded once a statistically significant relationship between a potential causal factor and a disease has been found, as is discussed below. However, what constitutes the best

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 means of evaluating statistical significance itself is evolving, as are the grounds for inferring causation in some circumstances. Small numbers, rare events, or small populations are often involved in hazardous-waste sites. Consequently, the committee does not adhere strictly to conventional approaches to establishing causality only after a finding of statistical significance has been made. Before detailing the committee's reasons for relying on an inferential approach in developing an understanding of causation in environmental epidemiology, it is useful to consider the function and limits of statistically significant findings in studies of the health effects of hazardous wastes. STATISTICAL SIGNIFICANCE The requirement that a finding be statistically significant has been a convention of epidemiologic research. If results have a likelihood of only 5 percent or less of occurring by chance, then they are usually considered statistically significant, as measured by a number of customary tests, such as p and t values. Under some circumstances, this stipulation can stifle innovations in research when studies that fail to meet the conventional criteria for a positive finding are prematurely dismissed. Thus, a study of a common disease in a small number of people might not achieve a level of statistical significance, even though a causal association could, in fact, exist. Several analysts maintain that the indiscriminate application of tests for statistical significance to epidemiologic studies has discouraged advances in research and conferred undue importance on negative findings. Rothman (1986) argues that conventionally applied tests of statistical significance, such as p values, are inadequate and subject to extensive misinterpretation. He favors the broad application of confidence intervals, so that results are depicted as ranging over a set of possible values, viz., there is a 90 percent chance that a given finding falls between some high value and some low value. Ahlbom et al. (1990) describe two general categories of negative studies that can result from an overreliance on traditional tests of statistical significance: those that actually suggest that a given exposure lacks an effect of a detectable size on the studied disease risk, and those that might miss such an effect because of inadequate sample size, random error, or because systematic error biases the study toward finding no such effect. Random error increases the chance that inaccurate measures of the effect will imply that there is no difference between those exposed and unexposed. Discussions of negative studies must recognize the importance of the size and detectability of the effects being missed.

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 According to some philosophers of science, the hypothesis that a given exposure has no effect on increasing the risk of a particular disease can be rejected (Buck, 1975), but can never be proved (Bunge, 1963). Greenland (1988) has criticized this strict application of philosophy for its failure to meet the requirements of epidemiology regarding public health. Several analysts have noted that an inductive approach can be more appropriately applied to epidemiologic study, considering epidemiologic study a measurement exercise with which relevant measures of effect are estimated (Miettinen, 1985). Even the inductive approach to causation in epidemiology is vulnerable to random or systematic error. Where the size of a study is small, random error can overwhelm a finding. Whether the level of statistical significance exceeds or fails to meet the 0.05 level does not necessarily bear on whether the effect parameter is biologically important or is equal to the null value, that is, does not differ significantly from what is expected to occur by chance. A better indication of the statistically plausible range of values can be provided by identifying the estimated confidence interval, that is, the range within which there is a 90 percent chance that the true value is contained. The confidence interval brackets the interval or range of values that may occur and provides a clearer indication of the significance of a study than does strict application of p values and other measures of statistical significance. Systematic error in classifying disease or exposure produces invalid results. Error arising from a misclassification of exposure can occur under a number of conditions, including the following: if the exposure measurement is random or subject to error; if an invalid or systematically inaccurate proxy for exposure is used, such as distance from a hazardous-waste site independent of relevant wind patterns or sources of domestic water; if a biologically inaccurate indication of exposure is applied, such as the use of a point-in-time exposure intensity rather than a cumulative dose; or if people either do not know the amount of exposure or exaggerate it. The problem of reconstructing exposures is especially subject to recall bias. Recall bias occurs where persons who have learned that they may be at risk from an exposure associate nonspecific health problems with the exposure or develop a health problem that they seek to attribute to the exposure and then “remember ” specific symptoms better than do non-exposed persons. Error also can be introduced through misclassification of disease. For instance, including persons who do not, in fact, have a given disease along with those who do have the disease, produces low specificity in the results. Ahlbom et al. (1990) warn that over-interpretation of epidemiologic

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 results can occur when results that show no effect are believed to prove no effect, even though they are actually inconclusive. Among the factors that can contribute to this over-interpretation of negative studies are failure to achieve statistical significance, too small a sample size, the poor assessment of exposure, the presence of confounding factors, and the lack of known biologic mechanisms that may account for the particular relationship between exposure and disease. CAUSATION IN EPIDEMIOLOGY The world of epidemiology, as that of any human science, seldom permits elegant inferences to be drawn about causation. The object domain of epidemiology consists of numerous uncontrollable aspects, with considerable variations in precedents, so that we cannot vary only one factor at a time. With human sciences, causation usually must be inferred, and is never proved absolutely. Human minds seem to be more credulous than skeptical, and most people need protection against being gulled. Undue skepticism, however, can be as dangerous as credulity to scientific progress and the improvement of health. Only judgment can prevent the hypercritical rejection of useful results. (Susser, 1973, p. 141) Susser's statement reminds us that the judgment of experts is a critical component for interpreting any findings in epidemiology. A fundamental dilemma for epidemiologic research on hazardous-waste sites, or any other topic involving multiple causes and results, derives from the fact that the statistical correlation of variables does not necessarily indicate any causal relationship among them, even where tests of statistical significance may be met. Mere coincident occurrence of variables says nothing about their essential connection. Moreover, partial correlations between variables that exclude other relevant variables can be misleading. To estimate the relationship between exposure and health status it is necessary to include relevant variables or their appropriate proxies, to the extent that these can be determined. Efficient use of that information requires the choice of a functional form that is compatible with the health-related practices and decisions of the individuals who are under study. No matter how carefully such proxy variables are estimated, causal inference should not be equated with statistical inference. Nor can statistical expertise alone establish causation. In order to facilitate the inference of causation from statistical information, contemporary epidemiologists have developed guidelines based on the view that absolute truth cannot be determined scientifically

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 (Mill, 1865). The relative likelihood that a finding is true must be inferred from careful, systematic, and repeated observations of recurring phenomena. Thus, association can be proved beyond a reasonable doubt, but not refuted, while causation can be refuted, but cannot be proved. To make a reasonable inference of causation in environmental epidemiology, eight basic characteristics of the findings should be considered: the strength, specificity, and consistency of the association; the period of exposure; the biologic gradient or the relationship between the dose and the response; the effects of the removal of the suggested cause; the biologic plausibility of the association (Hill, 1953; USDHHS, 1986), including how well it coheres with other findings. Strength of the Association How great is the risk of disease apparently induced by a given factor (exposure)? This is often expressed as relative risk (RR), standard mortality ratio (SMR), odds ratio (OR), or standard fertility ratio (SFR), each of which compares the risk of disease incurred by exposed persons with that of unexposed persons. The greater the RR, SFR, or SMR, the stronger the inferred link for exposed individuals. Of equal concern for public health, however, is the attributable risk, which might be much harder to detect, study, and estimate in environmental epidemiology, given the problems of evaluating baseline rates for a disease of interest. An RR of 3 for a lifetime that affects 1 of 100 persons in a small population produces a much smaller impact on public health than does a lifetime RR of 1.1 that affects several million persons. Epidemiologists have long appreciated that high RRs are relatively easy to detect. Thus, evidence linking lung cancer and cigarette smoking is strong; active smokers have a tenfold or greater risk of contracting lung cancer than non-smokers do. In contrast, evidence linking lung cancer and passive smoking is less firmly established; a variety of studies (NRC, 1986a) place the RR between 1.2 and 2.0, with the 95 percent confidence interval for a summary of the case-control studies ranging from 0.9 to more than 2. The difficulty with the use of this criterion of the strength of the association in environmental epidemiology is that misclassification of exposure can greatly attenuate the strength of a relatively weak observed association. Other sources can contribute to a specific chemical exposure, and the same health effect can be caused by different pollutants. Most of the results of concern are common, chronic diseases, for which the baselines—their normally expected rates—are not clearly

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 Santa Clara County, CA 1985 CA Dept. of Health Services Retrospective follow-up 1980-1981 1981-1982 1980-1981: Pregnancies in one census tract served by contaminated water Referent: Pregnancies in one census tract not served by contaminated water 1981-1982: live births in a 7 census tract study area served by contaminated water Referent: live births in the rest of the county Surrogate: residence in households served by contaminated water at the time of chemical leak 1980-1981: pregnancy outcomes 1981-1982: congenital cardiac defects 1980-1981: significant excess of spontaneous abortions and congenital malformations 1981-1982: excess incidence of cardiac defects within and outside the study area. No support for an association with the chemical leak Santa Clara County, CA 1989 Swan et al. Retrospective follow-up 1981-1983 106 babies with diagnosis of cardiac anomaly, born in county during period county exposed to contaminated water Referent: babies born in unexposed area and during unexposed time Surrogate: residence in households served by contaminated water Cardiac anomalies Increased prevalence of cardiac anomalies but temporal distribution suggests solvent leak not responsible

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 Galena, KS 1990 Neuberger et al. Retrospective follow-up 1980-1985 White residents of Galena exposed to heavy metal mining Superfund site Referent: Two unexposed towns Residence in towns for at least 5 years prior to 1980 Age and sex-specific illnesses Significant associations of stroke, anemia, hypertension, heart disease, skin cancer with exposure Lowell, MA 1987 Ozonoff et al. Cross-sectional 1983 1049 potentially exposed Referent: 948 presumably unexposed Surrogate: residence in households within a given distance from site Self-reported health problems Increased prevalence of minor symptoms, irregular heart beat, fatigue, bowel complaints Woburn, MA 1986 Lagakos et al. Case-control 1964-1983 20 childhood leukemia cases Referent: 164 children resident in Woburn Surrogate: residence in households served by contaminated wells Childhood leukemia Significant association with estimated exposure Woburn, MA 1986 Lagakos et al. Retrospective follow-up 1960-1982 4936 pregnancies among Woburn residents 5018 residents 18 or younger Referent: internal Surrogate: residence in households served by contaminated wells Adverse pregnancy outcomes; childhood disorders Association with perinatal deaths; eye/ear anomalies, CNS anomalies; association with kidney/urinary tract infection

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 Woburn, MA 1988 Feldman et al. Clinical case-control 1987(?) 28 members of 8 families with suspected neurotoxicity due to chronic exposure to TCE contaminated water Referent: 27 subjects evidencing no sign of neurologic disease or exposure to neurotoxins Surrogate: residence in households served by contaminated wells Blink reflex measurement as indicator of neurotoxic effects of TCE exposure Significant differences in blink reflex function when means were compared Rutherford, NJ 1980, Burke et al. Halperin et al. Case-control 1973-1978 13 leukemia cases, 9 Hodgkin's cases Referent: 25 sixth graders and 17 community controls (leukemia); 17 age-sex-race matched cases from random digit dialing (Hodgkin 's) Surrogate: residence in the area Possible etiologic risk factors for leukemia and Hodgkin's Reduced prevalence of rubella vaccination in leukemia cases. Excess of prior vaccinations and tonsillectomies in Hodgkin's cases Hyde Park, NY 1981 Rothenburg Cross-sectional 1979 246 persons working in the area Referent: 492 persons from HANES National Survey Surrogate: employment in plants near site Health problem, urine and blood tests Increased prevalence of hiatus hernia and other minor gastrointestinal problems

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 Love Canal, NY 1981 Janerich et al. Retrospective follow-up (census tract) 1955-1977 700 census tract residents Referent: NY state population Surrogate: proximity to dump site Cancer: liver lymphomas leukemias Incidence: no increase Love Canal, NY 1984 Heath et al. Cross-sectional 1982 45 residents in houses potentially contaminated by organic chemicals Referent: 46 residents in adjacent census tract Surrogate: testing of chemicals (two years before) in the house of exposed Cytogenic: SCE chromosomal aberrations No difference Love Canal, NY 1984 Vianna and Polan Retrospective follow-up 1941-1978 174 live births in swale areas near dump site Referent: 1. 443 live births in the rest of Canal area 2. all live births in upstate NY Surrogate: proximity to dumpsite and at least 5 months residence Low birth weight Elevated incidence among exposed Love Canal, NY 1985 Paigen et al. Cross-sectional 1980 523 children residents of L.C. neighborhood Referent: 440 children of adjacent census Surrogate: proximity to dump site Health problems: seizures, learning problems, hyperactivity, eye irritation, skin rash, abdominal Increased prevalence

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 Love Canal, NY 1987 Paigen et al. Cross-sectional 1980 172 children born and 75% of life in Love Canal area Referent: 404 children born in adjacent census tract Surrogate: proximity to dump site Anthropometric measurements Increased prevalence of shorter stature Hamilton, Ontario 1987 Hertzman et al. Retrospective follow-up Workers: 1965-1980 Residents: 1976-80 Workers: 197 workers at site Referent: 235 nonlandfill outdoor workers from Hamilton Wentworth Region Residents: 614 households within 750 m of edge of dumpsite Referent: 636 households in same air pollution region as landfill site Workers: outdoor employment on or adjacent to site Residents: long/short-term residence in area during 1976-1980 Self-reported health outcomes Workers: clusters of respiratory, skin, narcotic, and mood disorders Residents: confirmed association between landfill site exposure and mood, narcotic, skin, and respiratory conditions Clinton County, PA 1984 Budnick et al. Mortality 1950-1979 Clinton County and three adjacent counties, PA Referent: 1. State of Pennsylvania 2. U.S.A. Surrogate: residence in the area Bladder cancer mortality Increased bladder cancer mortality in male resident population after 1970

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 Clinton County, PA 1986 Logue and Fox Cross-sectional 1983 179 long-term residents in the area near waste site Referent: 151 residents of surrounding communities Surrogate: residence in the area Self-reported health problems Increased prevalence of skin problems and sleepiness Dauphin County, PA 1985 Logue et al. Cross-sectional 1983 65 potentially exposed Referent: 64 presumably unexposed Surrogate: residence in households with past contamination of water with TCE Self-reported health problems Increased prevalence of eye irritation, diarrhea, and sleepiness Hardeman County, TN 1982 Clark et al. Meyer, 1983 Harris et al., 1984 Cross-sectional 1978 49 residents at high exposure and 33 at intermediate exposure Referent: 57 unexposed local residents Carbon tetrachloride in well water >150 µg/l (high exposure) <45 µg/l (intermediate exposure) Liver functions Transient abnormalities of liver functions in exposed Source: Expanded and adapted from Upton et al., 1989, with permission.

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 department reports on this subject, emerging international reports, and case studies of legal decisions that have evaluated epidemiologic evidence not otherwise available in the published literature. On the basis of this review, we will recommend research opportunities and developments for the field of environmental epidemiology. REFERENCES Ahlbom, A., O. Axelson, E.S. Hansen, C. Hogstedt, U.J. Jensen, and J. Olsen. 1990. Interpretation of “negative” studies in occupational epidemiology. Scand. J. Work Environ. Health: 153-157 Baker, D.B., S. Greenland, J. Mendlein, and P. Harmon. 1988. A health study of two communities near the Stringfellow Waste Disposal site. Arch. Environ. Health 43: 325-334 Black, B. 1990. Matching evidence about clustered health events with tort law requirements Am. J. Epidemiol. 132: S79-S86 Brady, J., F. Liberatore, P. Harper, P. Greenwald, W. Burnett, J.N. Davies, M. Bishop, A. Polan, and N. Vianna. 1977. Angiosarcoma of the liver: An epidemiologic survey. J. Natl. Cancer Inst. 59: 1383-1385 Buck, C. 1975. Popper's philosophy for epidemiologists. Int. J. Epidemiol. 4: 159-168 Budnick, L.D., D.C. Sokal, H. Falk, J.N. Logue, and J.M. Fox. 1984. Cancer and birth defects near the Drake Superfund site, Pennsylvania Arch. Environ. Health 39: 409-413 Bullard, R.D. 1990. Dumping in Dixie: Race, Class, and Environmental Quality. Boulder, Colo.: Westview Press. Bunge, M.A. 1963. Causality: The Place of the Causal Principle in Modern Science. Cleveland: World. Burke, T.A., S. Gray, C.M. Krawiec, R.J. Katz, P.W. Preuss, and G. Paulson. 1980. An environmental investigation of clusters of leukemia and Hodgkin's disease in Rutherford, New Jersey. J. Med. Soc. N.J. 77: 259-264 California Department of Health Services. 1985. Pregnancy Outcomes in Santa Clara County 1980-1982: Reports of Two Epidemiological Studies. Berkeley: California Department of Health Services. Chalmers, T.C., H. Levin, H.S. Sacks, D. Reitman, J. Berrier, and R. Nagalingam. 1987. Meta-analysis of clinical trials as a scientific discipline. I. Control of bias and comparison with large co-operative trials. Stat. Med. 6: 315-328 Clark, C.S., C.R. Meyer, P.S. Gartside, V.A. Majeti, B. Specker, W.F. Balisteri, and V.J. Elia. 1982. An environmental health survey of drinking water contamination by leachate from a pesticide waste dump in Hardeman County, Tennessee Arch. Environ. Health 37: 9-18 CMA (Chemical Manufacturers Association). 1991. Guidelines for Good Epidemiology Practices for Occupational and Environmental Epidemiology Research. Washington, D.C.: Chemical Manufacturers Association. Commission for Racial Justice, United Church of Christ. 1987. Toxic Wastes and Race in the United States: A National Report on the Racial and Socio-Economic Characteristics of Communities with Hazardous Waste Sites. [ New York]: Public Data Access. CRS (Congressional Research Service). 1980. Six Case Studies of Compensation for

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 Toxic Substances Pollution: Alabama, California, Michigan, Missouri, New Jersey, and Texas. Committee Print, 96th Congress, 2d Session, Serial No. 96-13. Washington, D.C.: U.S. Government Printing Office. Davis, D.L. 1985. The “shotgun wedding” of science and law: Risk assessment and judicial review. Columbia J. Environ. Law 10: 67-109 Davis, D.L. 1988. Changing policy roles of environmental epidemiology. Stat. Sci. 3: 281-285 Doll, R., and R. Peto. 1978. Cigarette smoking and bronchial carcinoma: Dose and time relationships among regular smokers and lifelong non-smokers. J. Epidemiol. Community Health 32: 303-313 Environment and Health in Developing Countries. 1991. A newsletter published by Program on Environment and Health at Management Sciences for Health, Boston, Mass. Feldman, R.G., J. Chirico-Post, and S.P. Proctor. 1988. Blink reflex latency after exposure to trichloroethylene in well water. Arch. Environ. Health 43: 143-148 Goldberg, S.J., M.D. Lebowitz, E.J. Graver, and S. Hicks. 1990. An association of human congenital cardiac malformations and drinking water contaminants. J. Am. Coll. Cardiol. 16: 155-164 Goldman, L.R., B. Paigen, M.M. Magnant, and J.H. Highland. 1985. Low birth weight, prematurity and birth defects in children living near the hazardous waste site, Love Canal. Haz. Waste Haz. Materials 2: 209-223 Goldsmith, J.R., ed. 1986. Environmental Epidemiology: Epidemiological Investigation of Community Environmental Health Problems. Boca Raton: CRC Press. Greenland, S. 1988. Probability versus Popper: An elaboration of the insufficiency of current Popperian approaches for epidemiologic analysis. Pp. 95-104 in Causal Inference, K.J. Rothman, ed. Chestnut Hill, Mass.: Epidemiology Resources. Greenland, S. 1990. Divergent Biases in Ecologic and Individual-Level Studies. Paper presented at the Second Annual Meeting of the International Society for Environmental Epidemiology, August 12-15, 1990, Berkeley, Calif. Grisham, J.W., ed. 1986. Health Aspects of the Disposal of Waste Chemicals. New York: Pergamon. Halperin, W., R. Altman, A. Stemhagen, A.W. Iaci, G. Caldwell, T. Mason, J. Bill, T. Abe, and J.F. Clark. 1980. Epidemiologic investigations of cluster of leukemia and Hodgkin's disease in Rutherford, New Jersey. J. Med. Soc. N.J. 77: 267-273 Harris, R.H., J.H. Highland, J.V. Rodricks, and S.S. Papadopulos. 1984. Adverse effects at a Tennessee hazardous waste disposal site. Hazardous Waste 1: 183-204 Heath, C.W., Jr., M.A. Nadel, M.M. Zack, Jr., A.T.L. Chen, M.A. Bender, and J. Preston. 1984. Cytogenic findings in persons living near the Love Canal. J. Am. Med. Assoc. 251: 1437-1440 Henderson, T.W. 1990. Toxic tort litigation: Medical and scientific principles in causation Am. J. Epidemiol. 132(Suppl. 1): S69-S78 Hertzman, C., M. Hayes, J. Singer, and J. Highland. 1987. Upper Ottawa Street Landfill Site health study. Environ. Health Perspect. 75: 173-195 Hill, A.B. 1953. Observation and experiment. N. Engl. J. Med. 248: 995-1001 Janerich, D.T., W.S. Burnett, G. Feck, M. Hoff, P. Nasca, A.P. Polednak, P. Greenwald, and N. Vianna. 1981. Cancer incidence in the Love Canal area. Science 212: 1404-1407 Johnson, B.L. 1988. Health effects of hazardous waste: The expanding functions of the Agency for Toxic Substances and Disease Registry. Environ. Law Reporter 18: 10132-10138 Johnson, B.L. 1990. Implementation of Superfund's health-related provisions by the

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 Agency for Toxic Substances and Disease Registry. Environ. Law Reporter 20: 10277-10282 Kopfler, F.C., and G.F. Craun. 1986. Environmental Epidemiology. Chelsea, Mich.: Lewis. Lagakos, S.W., B.J. Wessen, and M. Zelen. 1986. An analysis of contaminated well water and health effects in Woburn, Massachusetts. J. Am. Statist. Assoc. 81: 583-596 Lilienfeld, A.M., and D.E. Lilienfeld. 1980. Foundations of Epidemiology, 2nd ed. New York: Oxford University Press. Lilienfeld, D.E., E. Chan, J. Ehland, J. Godbold, P.J. Landrigan, G. Marsh, and D.P. Perl. 1989. Rising mortality from motoneuron disease in the USA, 1962-84 [see Comment in: Lancet 1989 Apr. 29; 1(8644):958]. Lancet 1(8640): 710-713 Logue, J.N., and J.M. Fox. 1986. Residential health study of families living near the Drake Chemical Superfund site in Lock Haven, Pennsylvania. Arch. Environ. Health 41: 222-228 Logue, J.N., R.M. Stroman, D. Reid, C.W. Hayes, and K. Sivarajah. 1985. Investigation of potential health effects associated with well water chemical contamination in Londonderry township, Pennsylvania. Arch. Environ. Health 40: 155-160 McKeown, T. 1976. The Modern Rise of Population. New York: Academic Press. 168 pp. Meyer, C.R. 1983. Liver dysfunction in residents exposed to leachate from a toxic waste dump. Environ. Health Perspect. 48: 9-13 Miettinen, O.S. 1985. Theoretical Epidemiology: Principles of Occurrence Research in Medicine New York: John Wiley & Sons. Mill, J.S. 1865. An Examination of Sir William Hamilton's Philosophy. London: Longman. NCI (National Cancer Institute). 1990. Cancer Statistics Review, 1973-1987. NIH Publ. No. 90-2789. Bethesda, Md.: U.S. Department of Health and Human Services. Neuberger, J.S., M. Mulhall, M.C. Pomatto, J. Sheverbush, and R.S. Hassanein. 1990. Health problems in Galena, Kansas (USA): A heavy metal mining Superfund site. Sci. Total Environ. 94: 261-272 Neutra, R.R. 1990. Counterpoint from a cluster buster. Am. J. Epidemiol. 132: 1-8 NJDOH (New Jersey Department of Health) 1989. A Report on the Health Study of Residents Living Near the Lipari Landfill. Environmental Health Service. 120 pp. NRC (National Research Council). 1983. Risk Assessment in the Federal Government. Managing the Process. Washington, D.C.: National Academy Press. NRC (National Research Council). 1984. Assigned Share for Radiation as a Cause of Cancer. Review of Radioepidemiologic Tables Assigning Probabilities of Causation Washington, D.C.: National Academy Press. NRC (National Research Council). 1986a. Environmental Tobacco Smoke: Measuring Exposures and Assessing Health Effects. Washington, D.C.: National Academy Press. NRC (National Research Council). 1986b. Drinking Water and Health, Vol. 6. Washington, D.C.: National Academy Press. NRC (National Research Council). 1989. Biologic Markers in Reproductive Toxicology. Washington, D.C.: National Academy Press. OTA (U.S. Congress, Office of Technology Assessment). 1989. Coming Clean: Superfund Problems Can Be Solved. OTA-ITE-433. Washington, D.C.: U.S. Government Printing Office. 223 pp. Ozonoff, D., and L.I. Boden. 1987. Truth and consequences: Health agency responses to environmental health problems. Sci. Technol. Human Values 12: 70-77 Ozonoff, D., M.E. Colten, A. Cupples, T. Heeren, A. Schatzkin, T. Mangione, M. Dresner,

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 and T. Colton. 1987. Health problems reported by residents of a neighborhood contaminated by a hazardous waste facility. Am. J. Ind. Med. 11: 581-597 Paigen, B., L.R. Goldman, J.H. Highland, M.M. Magnant, and A.T. Steegman, Jr. 1985. Prevalence of health problems in children living near Love Canal. Haz. Waste Haz. Materials 2: 23-43 Paigen, B. L.R. Goldman, M.M. Magnant, J.H. Highland, and A.T. Steegman, Jr. 1987. Growth of children living near the hazardous waste site, Love Canal Hum. Biol. 59: 489-508 Rall, D.P. 1991. Carcinogens and human health: Part 2 [letter]. Science 251(4989): 10-13 Robbins, J. 1988. Alternative approaches to cluster investigations. Unabstracted presentation at the 21st Annual Meeting of the Society for Epidemiologic Research, Vancouver, British Columbia, Canada, June 1988 Rosenman, K.D., J.E. Rizzo, M.G. Conomos, and G.J. Halpin. 1989. Central nervous system malformations in relation to two polyvinyl chloride production facilities. Arch. Environ. Health 44: 279-282 Rothenberg, R. 1981. Morbidity study at a chemical dump — New York. Morbid. Mortal. Week. Rep. 30: 293-294 Rothenberg, R.B., K.K. Steinberg, and S.B. Thacker. 1990. The public health importance of clusters: A note from the Centers for Disease Control. Am. J. Epidemiol. 132: S3-S5 Rothman, K. 1986. Modern Epidemiology. Boston: Little, Brown. Schuck, P.H. 1991. Two causation conundrums: Mass exposure and social causes. Courts Health Sci. Law 1: 305-319 Siegel, M.R. 1990. Integrating public health into Superfund: What has been the impact of the Agency for Toxic Substances and Disease Registry? Environ. Law Reporter 20: 10013-10020 Susser, M. 1973. Causal Thinking in the Health Sciences: Concepts and Strategies of Epidemiology. New York: Oxford University Press. Swan, S.H., G. Shaw, J.A. Harris, and R.R. Neutra. 1989. Congenital cardiac anomalies in relation to water contamination, Santa Clara County, California, 1981-1983. Am. J. Epidemiol. 129: 885-893 Upton, A.C., T. Kneip, and P. Toniolo. 1989. Public health aspects of toxic chemical disposal sites. Annu. Rev. Public Health 10: 1-25 USDHHS (U.S. Department of Health and Human Services). 1976. The Health Consequences of Smoking. Report of the Surgeon General Publ. No. CDC 78-8357. Washington, D.C.: Office of Smoking and Health, U.S. Department of Health and Human Services. USDHHS (U.S. Department of Health and Human Services). 1986. The Health Consequences of Involuntary Smoking: A Report of the Surgeon General. DHHS Publication No. (CDC) 87-8398. Washington, D.C.: U.S. Department of Health and Human Services. Utidjian, H.M.D. 1988. The interaction between epidemiology and animal studies in industrial toxicology. Pp. 309-329 in Perspectives in Basic and Applied Toxicology, B. Ballantyne, ed. London: Wright. Vesselinovitch, S.D. 1969. Significance of the stage of perinatal development in the induction of tumors by a radiomimetic multicarcinogen. Pp. 511-515 in Radiation Biology of the Fetal and Juvenile Mammal. Oak Ridge, Tennessee: U.S. Atomic Energy Commission. Vianna, N.J., and A.K. Polan. 1984. Incidence of low birth weight among Love Canal residents. Science 226(4679): 1217-1219

OCR for page 25
ENVIRONMENTAL EPIDEMIOLOGY: Volume 1 WHO (World Health Organization). 1983. Guidelines on Studies in Environmental Epidemiology. Environmental Health Criteria 27. Geneva: World Health Organization. Wong, O., M.D. Whorton, N. Gordon, and R.W. Morgan. 1988. An epidemiologic investigation of the relationship between DBCP contamination in drinking water and birth rates in Fresno County, California Am. J. Public Health 78: 43-46