Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
CONCLUSIONS AND RECOMMENDATIONS 191 6 CONCLUSIONS AND RECOMMENDATIONS The Committee reviewed the current status of epidemiologic research on air pollution in the United States. It sought to identify the types of epidemiologic activities that could feasibly be applied to determine the size and nature of the health risks of air pollution, given current magnitudes and patterns. It observed that new approaches will be needed to quantify adverse health effects associated with low concentrations of air pollutants and to separate the impacts of individual pollutants more effectively. The Committee concluded that efforts to specify current and future research questions have been insufficient and that additional investment is needed in research on and development of appropriate investigative tools. It expects the development of new techniques and strategies to make it possible to overcome some of the barriers to the application of epidemiologic approaches. The Committee also concluded that many previous studies had generally failed to recognize the complexity of air pollution research and to integrate all the relevant sciences, including epidemiology, into coherent research plans and programs. The Committee finds that current air pollution can cause acute and perhaps chronic health effects, particularly respiratory effects, in the population of the United States. Respiratory disease is a major cause of work loss and disability. Even though exposure to some major types of pollution has decreased substantially in the last 15 years, exposure to other types has persisted. Improvement has not been uniform in all parts of the country, and some new patterns of air pollution are
CONCLUSIONS AND RECOMMENDATIONS 192 emerging. The entire population is affected by low concentrations of both indoor and outdoor air contaminants. Even if only a small proportion of very prevalent disease is due to air pollution, the absolute amount of illness that could be prevented by reducing air pollution would be large. Epidemiologic studies can be used to accomplish several things that cannot be accomplished in other ways: ⢠Directly determine whether a public health problem exists. ⢠Estimate the magnitude of the existing public health problem. ⢠Evaluate the impact (health and economic) of decreases in exposure. ⢠Define characteristics of the problem that can guide intervention even before mechanisms are understood. The results of well-conducted epidemiologic research are important in the development and revision of air quality standards and other means of regulating air pollution. It is recognized, however, that regulatory decisions sometimes have to be made before complete observational evidence is available. Although regulatory requirements must influence the definition of research questions and the design of studies, they should not be the sole basis for the development of broad research programs. Regulatory concerns, when narrowly interpreted, revolve around identification of âsafeâ exposures to individual pollutants--a question that presents few opportunities for studies of their effects on free-living populations. The Committee concludes that epidemiology, if it addresses this question and others mentioned above, can make a unique contribution to the prevention of illness related to air pollution. In view of the above conclusions, the Committee offers the following major recommendation: The Environmental Protection Agency should develop a long-term plan for research on air pollution; and population-based studies, in the form of a program in epidemiology, should be an integral part of that plan.
CONCLUSIONS AND RECOMMENDATIONS 193 CHARACTERISTICS OF A PROGRAM IN AIR POLLUTION EPIDEMIOLOGY To be productive and cost-effective, the epidemiologic program should have the following four major characteristics: ⢠Maintenance of a capability to interpret and synthesize current knowledge about air pollution and, accordingly, to plan relevant short-term epidemiologic research and make appropriate changes in long-term research. ⢠Inclusion of mechanisms for the creation of multidisciplinary teams to plan and conduct population-based epidemiologic research on the adverse health effects of air pollution. ⢠Means of ensuring stable, long-term epidemiologic research in a context that supports the career development of the researchers and thus limits disruptive changes in personnel. ⢠Exploration of collaboration with agencies that collect or could collect relevant data, after careful consideration of overall long-term data needs. A productive epidemiologic research program must have a dual character with respect to sensitivity to outside forces. Part of the program must be dedicated to responding to rapidly changing conditions that offer important opportunities for. study, to the varying concerns of regulators, and to new information from technologic development and parallel disciplines. In particular, advances in toxicology and clinical research, and epidemiology should be parts of the same framework; advances in any one can drive new efforts in the others. Another part of the program, engaged in long-term research strategies or longitudinal studies, must remain relatively isolated and free from abrupt change, although the incorporation of new elements will sometimes be desirable. It is important that the Environmental Protection Agency (EPA) program staff at all levels include highly trained and experienced epidemiologists, regardless of whether research is conducted intramurally or extramurally.
CONCLUSIONS AND RECOMMENDATIONS 194 Continuous access to both advice and review from scientists outside the institution will also be greatly needed. Given the constraining and difficult nature of today's questions about air pollution, stable research teams with expertise in several critical fields will be most productive. Epidemiologists in the teams should collaborate with atmospheric scientists, statisticians, and health effects scientists in all phases of research, from study design through interpretation. The present structure of many universities and government research agencies often makes it difficult to arrange such collaboration. Financial and administrative mechanisms that encourage development of these teams must be implemented; and the EPA program staff itself must have a multidisciplinary composition. The value of large data systems developed for reasons other than air pollution studies (such as the National Health and Nutrition Examination Survey, the National Health Interview Survey, and the National Ambulatory Medical Care Survey) as resources in air pollution research should be assessed. Whether modifications in some of them and linkage to air pollution exposure data constitute a feasible and cost-effective research approach that affords very large samples should also be explored. Modification in the routinely collected air sampling data alone might be appropriate, to facilitate use of these data in some types of epidemiologic studies. THE FOCUS FOR RESEARCH For the immediate future, the epidemiologic research program should focus on the following exposures and effects of concern: ⢠Persistent air pollution problems, including the health effects of acid sulfate particles, ozone, nitrogen dioxide, carbon monoxide, lead, and radon. It should be flexible enough to address emerging problems, such as the health effects of products of incomplete combustion and volatile organic chemicals.
CONCLUSIONS AND RECOMMENDATIONS 195 ⢠Lung disorders in which air pollution might play a role, including chronic obstructive pulmonary disease, asthma, decreased rate of lung growth or increased rate of lung decline with aging, and increased susceptibility to acute respiratory infections. ⢠The quantitative contribution of air pollutants to lung cancer in human populations. For this purpose, it could take advantage of the existing funding arrangement between EPA and the National Cancer Institute for the support of epidemiologic studies of this problem. Many important questions about the âtraditionalâ pollutants remain unanswered. For example, the acute and chronic respiratory effects of acid aerosol and ozone exposures, which might result only from outdoor sources, are not well understood. An extensive population, perhaps in excess of 100 million, is exposed to ambient ozone at high concentrations during the spring and summer. Shifts in coal combustion to the south central states will result in an increase in the area and population exposed to acid aerosols. No systematic epidemiologic study has been designed to assess either the acute or chronic effects of this type of shift in exposure. Exposures to acid aerosols are likely to increase more in rural areas than in urban areas, so it is recommended that consideration be given to locating baseline and followup studies in rural areas. As new automotive fuels, new industries, new fuel sources, new commercial products for the home, and new building ventilation patterns are introduced, they might yield new pollutants and pollution patterns that require epidemiologic evaluation. These changes will result in increased exposures to volatile and particulate organic compounds, radon, carbon monoxide, and other potentially hazardous materials. Indoor air pollution can be a major factor--in some instances the principal factor--in determining total personal exposure (averaged and acute) to air pollutants. Epidemiologic studies must therefore consider indoor or outdoor concentrations, or both, depending on the health response and pollutant being examined. For example, average and peak exposures to nitrogen dioxide are determined primarily by the presence of unvented indoor combustion sources. Peak exposures higher than those in
CONCLUSIONS AND RECOMMENDATIONS 196 most urban outdoor environments occur often in residences that use unvented gas or kerosene as a cooking or heating fuel. Therefore, to assess the health effects of nitrogen dioxide, EPA should study respiratory infection, pulmonary function changes, and, as soon as possible, biochemical indicators in association with indoor exposures and simultaneous outdoor exposures. There have been important advances in air monitoring instruments and inferential data analysis techniques. Air pollution is a complex mixture of gases, vapors, and particles; epidemiologic studies will often benefit from detailed characterization of its components. For instance, the chemical composition and acidity of size-fractionated particles should be characterized where appropriate. In some cases, new techniques for biologic characterization are also appropriate. Detailed characterization can help to determine the air pollution components most closely associated with health outcomes, potential confounders, and the relative contributions of various sources. The most readily observed health effects of air pollution are in the respiratory system. The proportion of the overall disease burden from common respiratory diseases that is attributable to air pollution has not yet been established. This part of the disease burden includes both the development of disease de novo and the exacerbation of pre-existing disease. Air pollution might have nonrespiratory effects that, although not emphasized in this report, also deserve study. These include neurobehavioral deficits and essential hypertension related to lead exposure, ischemic heart disease related to carbon monoxide, and carcinogenesis and mutagenesis related to various volatile hydrocarbons. Respiratory cancers, which are the most common cause of cancer death in men and will soon be in women, are attributable largely to cigarette smoking. However, attempts to assign proportions of this disease burden to separate causal agents, such as air pollution, are frustrated by the multifactorial and interactive etiology of these cancers. It is particularly important to understand the role of air pollutants in lung cancer, inasmuch
CONCLUSIONS AND RECOMMENDATIONS 197 as interactive effects might multiply the number of cases that could be prevented by reducing exposures. The limits of an epidemiologic research program depend on the questions under consideration. Only by considering each research question carefully can one understand the limits of investigative methods and discuss them productively. In general, studies of chronic health effects are more difficult than studies of acute effects. Specific reasons have been discussed in the preceding chapters; they include uncertainties in measurements of long-term exposure, the relative rarity of chronic diseases (which strains statistical power), and limitations in our understanding of the biology underlying the gradual evolution of chronic damage. Epidemiologic studies can show whether exposure to a complex pattern of polluted air increases the risk of adverse health effects in human populations, but studies are often limited in their ability to delineate the quantitative relationships between concentrations or sources of specific air pollutants and health. Such delineation requires interpretation of epidemiologic data in conjunction with toxicologic and clinical research. Susceptibility to the effects of air pollution varies widely; studies that focus on sensitive subgroups, necessary in themselves, are an important part of strategies to detect and measure the effects of air pollution in the general population. The sensitive-subgroup approach to increasing the effectiveness of air pollution epidemiology must be furthered by broad-based efforts (toxicology, clinical research, and epidemiology) to clarify the precise nature and degree of sensitivity of such groups as the young, the elderly, those with increased airway reactivity, and those with particular pre-existing diseases. Opportunities for carrying out epidemiologic studies on populations exposed to unusual magnitudes or patterns of air pollution must be specifically sought. By circumventing various methodologic constraints, these studies might provide information that would not otherwise be easily obtained. The opportunities include the use of occupational cohorts exposed to high concentrations, groups living in pristine environments or in highly polluted environments, and groups subjected to marked temporal changes in pollution. Collaboration with researchers in other countries might be necessary.
CONCLUSIONS AND RECOMMENDATIONS 198 NEW RESEARCH TOOLS AND OPPORTUNITIES Some constraints in air pollution epidemiology could be removed by complementary research to develop and explore the application of new tools for measuring exposure and effect. Such research and development is especially needed in two categories: ⢠New methods for assessing personal exposure and response to air pollution to be selectively incorporated into epidemiologic studies, with attention to the cost-effectiveness of these methods. ⢠Epidemiologic studies designed to determine the characteristics and the predictive value of potentially useful physiologic, biochemical, and morphologic markers of subclinical effects. Personal exposure monitoring and modeling are sometimes useful in epidemiologic research in defining study populations, optimal sample sizes, relationships of surrogate measures to exposures, and the extent of exposure misclassification associated with the use of central monitoring data. Depending on the design of a given study, only a sample of the study population might require such detailed monitoring; various strategies need to be explored and their performance documented. Markers of physiologic, biochemical, and cellular morphologic changes will be increasingly important in air pollution studies. FEV1, a physiologic test of lung function, has been used successfully to measure differences related to air pollution between populations. Serial measurement of FEV1 has also proved to be sensitive and effective in following the growth and decline of lung function in large populations. In adults, a more rapid than normal decline in FEV1 predicts premature death from pulmonary failure. Although FEV1 is a simple and highly reproducible test, its interpretation in terms of organ or cellular pathology is complex and subject to some judgment. Some biochemical indicators of air pollutant exposure or early effects--such as blood lead and carboxyhemoglobin concentrations, urinary mutagens, and indexes of genotoxic damage--have already been successfully applied in population studies. Other biochemical indicators, designed to
CONCLUSIONS AND RECOMMENDATIONS 199 detect early pathologic processes in the lung, have recently shown promise and require further development and validation. Insights into the pathogenesis of emphysema have been particularly fruitful in opening possibilities for biochemical markers related to the breakdown of connective tissue in the lung. Development of biochemical markers for epidemiologic studies requires particular attention to constraints imposed by the need to study large groups of relatively healthy people.
CONCLUSIONS AND RECOMMENDATIONS 200
