April 4, 1973
Page 11002
AIR POLLUTION AND PUBLIC HEALTH
Mr. MUSKIE. Mr. President, recently there has been a good deal of discussion regarding the health effects of air pollution. Recognizing that it was the responsibility of those of us who have advocated air pollution control legislation to continually monitor its impact, I initiated, several months ago, a review of recent studies relating to air pollution and public health with particular emphasis on automotive air pollution.
Today I would like to introduce into the RECORD some of the studies which I have received as a result of this reexamination. These studies will aid all of us to evaluate the claims that are often made regarding the relationship, or non-relationship, of air pollution to public health. The materials I am introducing today are as follows:
First, a summary of a study done by the California State Department of Public Health indicating that auto accidents occur more frequently in times of higher concentrations of automotive air pollutants.
Second, a summary of results of research on air pollution health effects being conducted by the Environmental Protection Agency. With regard to individual pollutants, this study reaches the following conclusions:
PARTICULATE MATTER
All recent evidence has strengthened the research base upon which the Administrator made his judgment. Thus, the present standards are not deemed unreasonable and it is not recommended that they be changed at this time.
PHOTOCHEMICAL OXIDANTS
In light of old and new evidence to date, the present primary national air quality standard for photochemical oxidants of 160 µg/m²(1.08 ppm) maximum concentration for one hour is reasonable. The evidence does not warrant a different standard.
NITROGEN OXIDES
Based on ... existing health intelligence, the margin of safety provided by the primary NOx standard of 100 µg/m²(0.05 ppm) is adequate. Present evidence does not compel us to suggest a change in the existing standard.
CARBON MONOXIDE
In view of [the] relatively small margins between present standards and observed health effects, it does not seem that existing standards are unduly restrictive and it does not seem prudent that they should be raised. Although the margins are relatively small, the carboxyhemoglobin levels that will result from adherence to the standard is sufficiently close to background, that no lowering of the standard is recommended either.
Third, a study done at the Veterans' Administration hospital in Long Beach, Calif., indicating that air pollution in normal rush hour traffic can have severe adverse effects on persons susceptible to angina and other circulatory problems.
Fourth, a document prepared by the Clean Air Now Information Bureau discussing a number of recent studies relating carbon monoxide pollution and human health.
Finally, a report on the status of research on health effects of air pollution which was prepared for the Subcommittee on Air and Water Pollution by the Library of Congress.
There being no objection, the matter was ordered to be printed in the RECORD as follows:
LETHAL GAS LINKED TO ACCIDENT TOLL
Motor vehicle accidents in Los Angeles occur more frequently during periods when carbon monoxide and oxidant levels are highest, according to findings of a California health research team.
Norman Perkins, head of a State Department of Public Health air pollution project, said that recent findings supplement an earlier analysis pointing to a strong statistical link between accidents and oxidant levels, but not carbon monoxide (CO). Oxidants are pollutants in the atmosphere caused by the photochemical reaction of sunlight on motor vehicle exhaust.
Perkins said further analysis of two four month periods of above-average CO levels in Los Angeles showed that the highest number of auto accidents tended to occur one and two hours after the daily CO peak hours. Peak periods varied seasonally because of different meteorological conditions. The winter CO high points were found between 8 p.m. and midnight; in the summer, peaks occurred from 3 a.m. to 5 p.m., but primarily between 8 and 10 o'clock in the morning.
The association between accidents and pollutant levels is most likely caused, Perkins said, by the effects of oxidants or CO on the "response capacity" of the drivers.
"Pollutants in concentrations that are frequently experienced in Los Angeles can impair vision and affect judgment or performance", he explained, adding that impairment of any physical function may increase the potential for accidents.
Since carbon monoxide builds up slowly in the body, the time lag noted between high atmospheric levels of CO and peak accident periods is consistent with the gases effects, Perkins said.
He added:
"We need quantitative measurements of carbon monoxide in the expired breath and in the blood of people in highway accidents that can be compared with CO levels in the air. This can be done most effectively at receiving hospitals."
Because oxidant levels in humans cannot be measured, similar studies with this pollutant are not possible, Perkins said. However, experimental studies should be made on the effects of oxidants on human performance, he added.
Perkins said additional statistical studies could be made to determine possible linkages between the type or severity of accidents and high pollution levels. Better estimates of the extra number of accidents that could be expected to occur during heavy pollution periods might also be obtained.
Perkins called for similar investigations of carbon monoxide and oxidants in other parts of the United States and in at least two or three European cities that have serious air pollution problems, in order to make a scientific comparison of findings.
Co-researcher in the project is Dr. John R. Goldsmith, also of the public health department. Hans K. Ury, Ph.D., now with the Permanent Medical Group in Oakland, initiated the study.
THE LIBRARY OF CONGRESS,
CONGRESSIONAL RESEARCH SERVICE
Washington, D.C., January 24, 1973.
To Subcommittee on Air and Water Pollution
Att.: Leon Billings
From James A.T. McCullough
Specialist, Life Sciences Via Charles S. Sheldon II
Chief, Science Policy Research Division
Subj. Health effects of air pollution – additional information
We have recently received a summary of the results of reviews and research being conducted under the CHESS program (an acronym for the Community Health and Environmental Surveillance System) at the National Environmental Research Center, Research Triangle Park, North Carolina. Since we did not know whether you had received this informal report, we are enclosing a copy as a matter of interest to you in your considerations.
The enclosed summary, as far as I can tell, provides the results of literature reviews and research under the CHESS program since its establishment. As you will note, the conclusions provided within the summary statements (I have marked the appropriate sections) provide essentially the same information forwarded by our previous memorandum to you; namely, that there is more confidence that the current standards for sulfur dioxides and particulates are readily supportable by medical evidence; that more information is needed with regard to oxides of nitrogen; that the photoxidant levels also seem reasonable; and that the opinions regarding carbon monoxide standards, while supported by some research, still require additional information to substantiate claims regarding low level effects. With regard to carbon monoxide, the problem still seems to center around the significance of effects being noted at levels of 3 to 5% carboxyhemoglobin. Susceptible groups continue to be individual with various circulatory disorders (heart deficiencies, etc.) young children and older persons with respiratory disorders, asthmatics, and persons allergic to air pollutants.
We are continuing our efforts to provide you with other information which you have requested.
PARTICULATE MATTER
The present national primary ambient air quality standards for particulate matter are as follows:
(a) 75 micrograms per cubic meter – annual geometric mean.
(b) 260 micrograms per cubic meter – maximum 24-hour concentration not to be exceeded more than once per year.
These standards were promulgated in the Federal Register, Volume 36, No. 84, Part II: pp. 8185-7, April 30, 1971.
I. Evidence upon which the standard was based:
The basis for the development of these standards was "Air Quality Criteria for Particulate Matter," (NAPCA Publication No. AP-49). This monograph critically reviewed all pertinent health studies discussing their individual strengths and weaknesses.
(a) Evidence for the maximum 24-hour concentration not to he exceeded more than once per year
1. Epidemiologic studies of populations exposed 200-300 µg/m² of particulates accompanied by 530-1060 µg/m² (.20-40 ppm) of sulfur dioxide for 24-48 hours showed that an increase above normal general mortality levels was likely and suggested that rises in infant mortality, increased admissions to hospital clinics for treatment of upper respiratory tract illness and cardiac and aggravation of symptoms in patients with chronic bronchitis would occur.
2. Studies of populations exposed to 150-200 of particulate matter accompanied by 265-530 µg/m² (.10-20 ppm) of sulfur dioxide for 24-48 hours suggested that an increase above normal general mortality level would occur.
No data on immediate health effects of exposure to less than 150 µg/m² of particulate matter accompanied by less than 2.55 µg/m²(.10 ppm) of sulfur dioxide had been reported.
(3) Evidence for the annual geometric mean level:
Several epidemiologic studies ascertained the health effects of long-term exposure to average annual particulate concentrations of 125-175 µg/m² accompanied by 00-120 µg/m² (.035– .045 ppm) of sulfur dioxide. At these levels increased chronic respiratory disease(chronic bronchitis and emphysema) prevalence in adults and excess chronic bronchitis mortality was well documented. Likely health effects resulting from long-term exposure to these average annual levels included increased frequency of acute lower respiratory tract disease (pneumonia, chest colds, bronchitis) in children, an increased frequency of common colds and minor respiratory ailments, a diminished pulmonary function (measured by spirometry), and an excess mortality above normal levels. There were suggestive findings of excess mortality from lung cancer at these exposure levels.
2. Studies of populations exposed to average concentrations of 75-125 µg/m² of particulates accompanied by 65-10 µg/m²(.025-.035 ppm) of sulfur dioxide indicated a likely impairment in pulmonary function and suggested both an increased frequency of upper and lower respiratory tract disease in children and an increase above normal levels of mortality.
3. No data on health effects from exposure to average annual concentrations of less than 75 µg/m² of particulates accompanied by less than 65 µg/m² (.025 ppm) of sulfur dioxide had been reported.
(c) Laboratory Evidence:
1. Toxicologic studies in laboratory animals were valuable in delineating adverse health effects of individual components of particulate matter and in determining effects of crude air particulate. Furthermore, some mechanisms of the toxicological action of particulate matter had been determined and interactions with irritant gases were studied. However, most of the toxicologic studies in animals employed concentrations of particulate far in excess of ambient concentrations. Furthermore, many of these studies did not employ natural modes of exposure.
Because of these shortcomings, as well as the difficulty in extrapolation from animal exposure to man, toxicologic data was not used directly in recommending the standard for particulates.
2. Toxicologic studies of particulates had shown the results of exposure to include (1) pathophysiologic changes in pulmonary mechanics; (2) anatomic pathologic changes in animal tissues including carcinogenesis, fibrogenesis, inflammation, and necrosis; and (3) pathologic changes in pulmonary clearance mechanisms. Moreover, the importance of particle size in producing biologic effects was established.
3. A series of papers by Amdur and her associates utilized an increase in pulmonary air flow resistance of guinea pigs as an assay tool of lung mechanics. Particles of sulfuric acid, ammonium sulfate, zinc sulfate, and zinc ammonium sulfate have been shown to produce a response in this system. However, aerosols of the following alone did not produce a response: spectrographic carbon at 2 and 8 µg/m², activated carbon at 8.7 µg/m², manganese dioxide at 9.7 µg/m², open hearth dust at 7.0 µg/m², iron oxide at 11.7 and 21.0 µg/m², manganous chloride and ferrous sulfate at 1 µg/m², and sodium orthovanadate at 0.7 µg/m². The particle sizes were less than 0.7 µm. Ferric sulfate at 1 mg/m² produced a 77% increase in flow resistance. In the presence of SO2, aerosols of soluble salts of ferrous iron, manganese, and vanadium potentiated the irritant response, but aerosols of carbon, iron oxide fume, fly ash, or open hearth dust do not potentiate the irritant effect of SO2. This work is interpreted to show that SO2 reacts in some way with low concentrations of metallic particulate and water vapor to form a potent respiratory irritant aerosol.
4. The health effects of carcinogenic ambient particulate matter have been extensively reviewed in a recent separate document.
II. Evidence:
Several foreign studies as well as studies in this country have been published since the particulate criteria document was written. Results of CHESS studies constitute the majority of epidemiologic evidence relating to health effects of particulates since that time, but these studies, like most other studies reported to date, generally could not fully distinguish the individual effects of SO2, and particulates.
(a) Evidence for the maximum 24-hour concentration not to be exceeded more than once a year.
1. Panels of healthy families in Birmingham reported an excess of irritation symptoms during a short-term exposure to 100-269 µg/m² of particulate matter accompanied by low levels of sulfur dioxide (13-16 µg/m²) for twenty-four hours. Cough, chest discomfort, and restricted activity were most consistently elevated. Non-smokers had lower baseline reporting rates than smokers, but both were equally affected during the episode.
2. CHESS studies of health effects of short term exposure are summarized in table 1 and discussed in detail in the monograph Health Consequences of Sulfur Oxides: A Report from CHESS 1970-1971. Health studies included panels of patients with pre-existing asthma, chronic respiratory, and/or cardiac disease. All studies suggested health effects at or below the present standard. However, levels of suspended sulfates were more consistently associated with the observed adverse health effects and were generally one to two orders of magnitude lower than those of sulfur dioxide or total suspended particulates.
(b) Evidence for the average annual geometric mean.
New evidence was available from CHESS studies of pulmonary function in children, acute lower respiratory disease in children, and chronic respiratory disease in adults (Table 2).
1. School children exposed to average annual particulate levels f 110 µg/m² accompanied by 230 µg/m² of sulfur dioxide and 10 µg/m² suspended sulfates exhibited subtle decreases in respiratory function.
2. An increase in the frequency and severity of acute lower respiratory disease was reported for school children exposed to average annual concentrations of 100 µg/m² of total suspended particulates accompanied by 91 µg/m² of sulfur dioxide.
3. Increased chronic respiratory disease symptom prevalence in smokers and nonsmokers was reported for adults exposed to average annual concentrations of 100 µg/m² of total suspended particulates accompanied by 95 µg/m² of sulfur dioxide.
(c) New toxicologic evidence:
Two types of toxicologic studies have emerged since the standard was recommended in 1969.
First, greater appreciation has developed for the role of individual compounds. These compounds acting alone have been shown to be toxic or carcinogenic and in most cases interactions with biologic systems in the presence of other complex factors of the atmosphere remains to be determined. Second, controlled studies of animals exposed to urban air and filtered air have been done. The animals exposed to urban air tend to show a greater retention of particulate material in the lung, but the short exposure time employed has usually not been adequate to determine long-term effects. Also a study of animals exposed to street atmospheres in Detroit showed an increased in white blood count in the exposed animals as well as an increase in pulmonary emphysema.
III. Recommendation concerning existing standard:
The present primary air quality standards for particulate matter were established on the basis of scientifically defensible evidence. Although the individual health effects of particulate matter and sulfur oxides could not be entirely tested apart, setting separate standards was wholly consistent with the Administrator's legislated obligation to provide adequate safety margins in promulgating ambient air quality standards.
All recent evidence has strengthened the research base upon which the Administrator made his judgement. Thus, the present standards are not deemed unreasonable and it is not recommended that they be changed at this time.
IV. Improving the basis for air quality standards:
(a) Original evidence:
Numerous studies had clearly indicated an association between air pollution as measured by particulate matter accompanied by sulfur dioxide and health effects of varying severity.
Although these epidemiologic studies were the best available at the time their limitations were recognized and Chapter 11 of the Particulate Criteria Document discusses interpretative cautions fully and completely. In brief, the measurement methods were often not directly comparable, the physics of particle size and chemistry of particulate matter had almost always been ignored in field studies, and pollution and health indices were not always measured over the same time periods. Few studies identified or adjusted for all other known or possible morbidity determinants such as cigarette smoking, occupational and other past exposures, past medical history, socio-economic status, residential mobility and indoor pollution exposure. Furthermore, almost no epidemiologic studies had been possible in areas characterized solely by particulate matter. On the other hand, almost no studies of highly susceptible individuals had been done.
(b) New evidence:
All new evidence strengthened and extended the initial observations on health effects. The CHESS studies were able to identify or adjust for the effects of other possible disease determinants in addition to ambient air pollutants. The limitations of the CHESS studies differ for the short-term and long-term health effects studies. For the studies of short-term health effects, the estimates of exposure are very reliable, but the indicators used to measure health effects require further quantitative refinement. For the studies of long-term health effects, previous exposures had to be estimated because of incomplete exposure monitoring data. Hence, the estimates of exposure were somewhat less accurate although the observed health effects were quite consistent.
(c) Additional information which would strengthen the standards:
The inability to differentiate adequately the effects of particulate pollution from those of SO2 pollution and the indication of the strong association of illness with particulate sulfates and other particulates in the respirable size range suggest the need for the following information:
1. Development of methods to identify specific particulate compounds.
2. Toxicity screening of particulate compounds especially with regard to particle size and chemical characteristics.
3. Additional epidemiologic studies of effects associated with particulate size or specific particulate compounds both in the presence and absence of sulfur dioxide.
4. Epidemiologic studies of particulate matter with known or suspected carcinogenic properties.
5. Inhalation studies on the carcinogenics of low concentrations of both crude particulates and pure compounds in complex atmospheres.
SULFUR OXIDES
National Primary Air Quality Standard
– 365 µg/m² – 24 hr. mean
– 80 µg/m² – annual mean
I. Adequacy of research base for air quality standards for SO2:
Many people have studied major air pollution episodes and have shown that adverse health effects result from the very high levels of pollution experienced. Few studies have been designed to determine minimum levels at which adverse effects occur. Information available in 1969 did indicate that increased mortality may he associated with 24-hour mean SO2 levels in the range of 500 µg/m² (0.19 ppm) with relatively low TSP levels, and increased hospital admissions with respiratory illness may be associated with 24-hour mean SO2 levels of 300-500 µg/m² (0.11 – 0.19 ppm). Long-term exposure to SO2 concentrations of about 120 µg/m² (0.046 ppm) appeared to be associated with increased frequency and severity of respiratory disease and with some types of increased mortality.
These studies represented the best information available. They did have limitations, however, that were recognized and acknowledged. Most prominent among these were: the inability to differentiate clearly between the effects of SO2 and the effects of particulate matter, measurement techniques both for effects and for pollution believed to be less accurate than those now recommended, non-recognition of co-variates which might have affected study results and the almost total lack of replication to confirm associations suggested by a single study. In spite of these limitations, the evidence available gave indications of adverse health effects associated with exposure to SO2 pollution that were believed to be sufficiently strong to warrant immediate control action. Consequently, the standards were established that were believed to be adequate to protect the health of even sensitive segments of our population.
II. Recently developed evidence related to the adverse effects of SO2 on pollution:
In addition to several foreign studies which have recently been translated, the major quantitative studies published since the time of the SO2 criteria document have been the results emanating from the CHESS program. These studies characterized short-term and long-term exposures to sulfur oxides and related these exposures to adverse health effects through dose response functions (Tables 1 & 2). In the four studies devoted to acute respiratory disease, there was a consistent excess of disease in children (ranging from 14 to 64 percent) among those exposed for 3 years or longer to SO2 levels of 90-95 µg/m² in the presence of 80-100 µg/TSPm² . In the area of chronic respiratory disease a significant and consistent excess bronchitis morbidity occurred over an exposure range of 100-350 µg/m² SO2 with associated TSP levels of 66-365 µg/m².
These long-term health effects were observed in populations having mean exposure only slightly above the present primary standard for SO2.
The strongest relationships in the CHESS studies appear to be the short-term effects between the exacerbation of symptoms in respiratory cripple, and daily suspended sulfates (Table 1). The levels of suspended sulfates necessary to cause adverse health effects were numerically one to two orders of magnitude lower than the levels of sulfur dioxide or total suspended particulates.
The findings f the CHESS studies with respect to the adverse health effects of suspended sulfates are supported by laboratory findings. Studies conducted in animals have shown that in terms of comparative toxicity sulfuric acid and some metallic sulfate compounds such as zinc ammonium sulfate are the more potent irritants than sulfur dioxide. These studies have also shown that the size of irritant particulate material affects the degree of response and that within the range of 0.3 to 2.5 µ MMD, the smaller the particle size the greater the potency.
III. Recommendation concerning existing standard
Judgmental decisions made to develop the current air quality standards were based on inadequate but the best available information and there are few criticisms that can be raised now which were not considered previously. More recently collected data have strengthened the available defense of the existing standards for SO2, but in addition. these same studies have tended to add new and important dimensions to the problem. The support for the standards stems primarily from the consistency of positive results produced in the CHESS program which showed increases in adverse health effects associated with SO2 levels not exceptionally higher than the standards.
The new dimension relates extensively to the stronger associations of adverse effects with suspended particulate sulfate than with SO2, and the total lack of knowledge about the relationship of SO2, levels to suspended sulfate levels.
Evidence now available provides adequate support for the position that present air quality standards for sulfur oxides are not too high. In fact, the new sulfate data indicate that the short term standard may not be low enough to be fully protective. Before reasonable decisions concerning the need to reduce the short-term SO2 standard can be made, additional data must be obtained relative to the rate at which SO2 is converted to sulfate, the health significance of sulfate exposure and the relationship of chronic effects to repeated short-term exposures.
When additional data covering these points are obtained it will be necessary to review again the basis for the SO2 standards and possibly recommend revisions. The fact that adverse health effects may relate more closely to sulfate than to SO2 indicates that simple control of SO2 may not be the most efficient way of protecting human health.
IV. Improving the basis for air quality standards:
Major problems remaining to be solved include: determinations of how extensively SO2 acts as a precursor for suspended particulate sulfates, determination of the relative toxicities of individual sulfates, development of methods for measuring the individual compounds in ambient situations, determination of the atmospheric transformation of SO2, particularly the significant interactions with other pollutants as hydrocarbons and oxidants, and the carrying out of well-planned, designed and executed health effect studies to determine the adverse health effects associated with the interaction of pollutants rather than individual substances.
The information obtained should indicate the significance of sulfate pollution in addition to SO2 pollution, and the atmospheric relationships between SO2 and particulate sulfate. These data must be available to permit a final determination of the adequacy of the present national ambient air quality standards for sulfur oxides.
PHOTOCHEMICAL OXIDANTS
The national primary standard for photochemical oxidants is 160 µg/m² (0.08 ppm) maximum one hour concentration not to be exceeded more than once per year (Reference 1).
I. Evidence upon which standard was based:
1. At short-term photochemical oxidant concentrations in excess of 139 µg/m² (0.07 ppm) for one hour, performance of student athletes was impaired (Reference 2).
2. At short-term ozone concentrations of 160 µg/m² (0.08 ppm) for 3 hours, experimental animals (mice) exhibited increased susceptibility to laboratory induced bacterial infections. A firm dose-response relationship was established for this effect (References 3-4).
3. At short-term peak oxidant concentrations of 196 µg/m² (0.10 ppm) and above, humans begin to experience eye irritation. (Reference 5) .
4. At short-term ozone concentrations of 392 µg/m² (0.20 ppm) for one-half to one hour, experimental animals had increased sphering of red blood cells (Reference 6). After 6 hours of exposure to 392 µg/m² (0.20 ppm), running activity of experimental mice was decreased (Reference 7).
5. At short-term maximum peak daily oxidant concentrations of 490 µg/m² (0.25 ppm) subjects with asthma begin to experience significantly more attacks. These maximum daily peak values may occur with a maximum hourly average concentration as low as 300 µg/m² (0.15 ppm) (Reference 7).
6. At long-term ozone concentrations of 392 µg/m² (0.20 ppm) for 5 hours daily for 3 weeks structural changes in cell membranes and in the nuclei of heart muscle fibers were produced in mice. These changes reverted to normal one month after exposure (Reference 8).
II. New evidence:
1. At short-term ozone concentrations of 196 µg/m² (0.10 ppm) for 2½ hours, the stability of those cells which protect the deep lung against infection (alveolar macrophages) was reduced in experimental animals (rabbits) (Reference 10).
2. At long-term ozone concentrations of 196 µg/m² (0.10 ppm) for 6 hours per day for 2.68 exposures, experimental animals exhibited chronic bronchitis and emphysema and a decrease of normal lipid tissue in the adrenal gland (Reference 9).
3. At short-term ozone concentrations of 392 µg/m² (0.20 ppm) for 5 hours, experimental hamsters exhibited mutagenic changes (chromosome breaks in circulating white blood cells) (Reference 11).
III. Recommendation concerning existing standard:
The studies demonstrate that photochemical oxidants, ozone in particular, are primary irritants causing aggravation of asthma and impairment of performance. They alter the body's defenses against infection, cause various structural changes, and can produce mutagenic effects in animals even at ambient concentrations.
In the light of old and new evidence to date, the present primary national air quality standard for photochemical oxidants of 160 µg/m² (0.08 ppm) maximum concentration for one hour is reasonable. The evidence does not warrant a different standard.
IV. Improving the basis for air quality standards:
Air quality standards should be based on complete and reliable dose-response relationships. The great majority of quantitative health intelligence to date has been obtained from animal models.
Naturally, direct extrapolation of this research to humans is questionable. Until comprehensive data about acute and chronic human exposures are gathered, the margins of safety will necessarily remain obscure.
Exposures of other vulnerable populations such as infants, pregnant women and persons with heart and lung disease have not been investigated. These groups may well require more stringent air pollution controls than healthy young adults.
Finally, photochemical oxidants occur with other atmospheric pollutants. Nearly all epidemiologic findings to date could possibly reflect interactions between oxidants and other environmental factors like ionizing radiation, sulfur oxides, or nitrogen oxides. These interactions remain to be discovered in the laboratory and in the field. Because interactions have not been characterized either in laboratory or epidemiologic studies, existing evidence can he applied to community exposures only with reservations. Considerable human dose-response data, both under controlled laboratory conditions and in natural community settings, are urgently needed to establish protective air quality standards for photochemical oxidants with adequate but not excessive margins of safety.
Discussion
The existing primary NOx standard of 100 µg/m² (0.05 ppm) was based on the Chattanooga School Children studies. Methods used in these studies (Jacobs-Hochheiser technique) to monitor exposure do not provide accurate measurements of NOx concentrations. However, NOx measurements taken independently by the U.S. Army (continuous Saltzman technique) at the same time and at sites adjacent to the location of monitors used for the health studies confirm the NOx levels reported in the Chattanooga School Children study (Reference 9). Additional pollutants emitted by the point source of exposure in Chattanooga have now been identified (Reference 10). Among these pollutants are substantial quantities of nitric acid mist, sulfuric acid mist and suspended nitrates and sulfates. Vegetation injury attributable to acid mist occurred within two miles of the point source (Reference 11). Although no direct measurements of acid mist were taken, measured pollutants other than gaseous NOx were closely correlated with observed health effects. These relationships are shown in the following table:
[TABLES OMITTED]
To the extent that nitric acid and sulfuric acid mist are reflected in suspended nitrate and sulfate measurements respectively, excess respiratory illness rates could be accounted for by acid mist exposures. Furthermore, recent EPA CHESS studies have identified adverse respiratory effects attributable to suspended sulfate concentrations exceeding 10 µg/m².
Therefore health effects observed in the Chattanooga School Children studies could have been produce by NOx alone, acid mist alone, nitrates alone, sulfates alone or combinations of these compounds. In view of these uncertainties, the results of the Chattanooga studies do not provide the health intelligence required by EPA to determine what NOx level represents the lowest threshold for adverse health effects. Animal studies demonstrate adverse effects, similar to those observed in the Chattanooga studies, at long-term NOx levels of 940 µg/m² (0.5ppm). Since animal exposure levels cannot be directly extrapolated to humans, these data do not provide the necessary health intelligence concerning adequate margins of safety for an air quality standard.
At the same time, our re-evaluation of the evidence does not suggest that an alternate standard is now indicated.
II. New evidence:
1. In the same Chattanooga areas, as above, exposed to a point source of nitrogen oxides and sulfur oxides, long-term exposure to 94 µg/m² (0.05 ppm) for 1 year at concurrent levels and to 117 to 250 µg/m² (0.062 to 0.109 ppm) during the previous 2 years was associated with slight but statistically significant reduction of lung function (FEV) in adults aged 30-55 years (Reference 12).
2. At long-term NOx exposure for 6 months to 564 to 940 µg/m²3 (0.3-0.5 ppm), mice had alterations in lung chemistry (reduced glutathione level). This change was accompanied by similar decrease in liver chemistries (Reference 13).
3. At long-term NOx exposures of 40 days at concentration of 940 µg/m² (0.5 ppm), rats exhibited an increased hematocrit and total blood specific gravity (Reference 14).
4. Structural changes were seen in the lungs of mice exposed to 940 µg/m² (0.5 ppm) for 30 days (edematous swelling, hyperplastic foci, proliferation of Clara cells, and damage to cilia) (Reference 15).
5. Preliminary findings indicate that users of gas stoves experience excess acute respiratory illness. Use of gas stoves for cooking results in repeated daily exposures for one and more hours to 940 µg/m² (0.5 ppm) NOx accompanied by NOx concentrations of 1230 µg/m² (1.0 ppm) (Reference 16).
III. Recommendation concerning existing standard:
Current knowledge concerning the biological consequences of NOx exposure alone demonstrates consistent adverse effects in animals and humans at concentrations of 940 µg/m² (0.5 ppm). The evidence suggests that at lower concentrations, nitrogen dioxide may cause adverse health effects, but these effects may be a result of exposure to other pollutants alone or in combination with NOx.
Based on these qualifications and existing health intelligence, the margin of safety provided by the primary NOx standard of 100 µg/m² (0.05 ppm) is adequate when pollutant mixes similar to those of Chattanooga occur. When isolated exposure to NOx alone exists, the standard may represent a generous margin of safety, comparable to safety margins for radiation and other environmental hazards. Existing data reveal that the standard for isolated exposure to NOx should be under 940 µg/m² (0.5 ppm) annual average, but the lower threshold for health impairment has yet to be determined. Present evidence does not compel us to suggest a charge in the existing standard. At the same time, additional health intelligence is required to determine the most appropriate margin of safety and to define the role of other compounds in producing the adverse health effects attributed to NOx.
IV. Improving the basis for air quality standards:
The biological consequences of short and long-term animal and human exposures to nitrogen dioxide, suspended nitrates and sulfates, nitric acid mist and combinations of these compounds with other common pollutants must he delineated. The following health effects of nitrogen oxides should be investigated:
(1) impaired resistance to infection;
(2) impairment of lung function;
(3) increased frequency of chronic bronchitis and emphysema; and
(4) structural and biochemical changes associated with blood cells and lipid membranes.
These effects have already been observed in animal or human studies, but dose-response information required for adequate margins of safety has not been obtained. A substantial effort in this direction is clearly mandated.
CARBON MONOXIDE
The national primary and secondary ambient air quality standard for carbon monoxide was established at 10 µg/m² (9 ppm) for a maximum eight-hour average and 40 µg/m² (35 ppm) for a maximum one-hour average, not to be exceeded more than once per year (Reference 1).
I. Evidence upon which standard was eased
A. Background
The most important toxic property of carbon monoxide is its reversible reaction with hemoglobin to form carboxyhemoglobin. Carboxyhemoglobin level is determined by a large number of physiologic and environmental variables, the most important of which are: 1) concentration of carbon monoxide in the air, 2) duration of exposure, and 3) level of ventilation, which is related to the level of physical and metabolic activity. Approximate carboxyhemoglobin levels to be expected for varying levels of activity after exposure to those levels of carbon monoxide defined in the standards are: [Table Omitted].
It is seen that the effect of exercise on resultant carboxyhemoglobin levels is not appreciable for longer, low level exposure conditions, but is quite significant for short term, higher level exposures.
B. Central nervous system effects
l. Beard and Wertheim (Reference 2) found that university students exposed to a carbon monoxide concentration of 58 µg/m² (50 ppm) for only 90 minutes experienced a decrement in their ability to discriminate between time intervals presented to them in the form of an auditory signal. Though direct blood carboxyhemoglobin measurements were unfortunately not available, such levels can be estimated at about 2 per cent from the concentration and duration of exposure and the sedentary activity.
2. Ray and Rockwell (Reference 3) found that young male automobile drivers required increased time to respond to brake-light signals and to adjust to changes in speed during car-following tests when tested at the 10 per cent carboxyhemoglobin level compared with the background exposure level.
3. In a study involving both carbon monoxide and altitude, McFarland and others (Reference 4) showed that visual threshold was increased at carboxyhemoglobin levels as low as 5 per cent. The magnitude of this increase was the same as that experienced at a simulated altitude of 8,000 feet.
4. Schulte (Reference 5) found that healthy male subjects experienced impairment in performance of some psychomotor tests at a 5 per cent carboxyhemoglobin level. Blood measurements were not accurate enough to ascertain effects at lower carboxyhemoglobin levels.
5. Stewart and others (Reference 6) investigated the effects of carbon monoxide on several aspects of human task performance, including time interval discrimination among healthy male subjects, and found virtually no decrement in performance, even at relatively high carboxyhemoglobin levels.
6. Mikulka and colleagues (Reference 7) also investigated carbon monoxide effects on human performance tasks similar to those examined by Beard and Stewart and found no consistent decrement in performance.
C. Cardiovascular system effects
1. Ayres and associates (References 8, 9) found that carboxyhemoglobin levels as low as 5 per cent were associated with some decreases in the ability of the heart, as measured by abnormal left ventricular function and decreased venous oxygen tension.
II. New evidence:
A. Horvath et al. (Reference 10) have seen diminished performance during a test of visual vigilance at carboxyhemoglobin levels of approximately 6 per cent.
B. Aronow and his group (Reference 11) have reported striking changes in electrocardiograms and exercise tolerance of men who have heart disease after they were exposed to 90 minutes of Los Angeles freeway traffic during the morning rush hour. The carboxyhemoglobin levels of these men rose to approximately 5 per cent at the end of this exposure.
C. EPA researchers (Reference 12) have obtained similar results in carefully controlled laboratory exposures of men with heart disease to carbon monoxide. These men had significantly decreased exercise tolerance as well as electrocardiogram changes, at carboxyhemoglobin levels of 3%. This EPA research is still in progress, but the initial findings have been presented at scientific meetings and will be published soon.
III. Recommendation concerning existing standard:
Research on the effects of carbon monoxide on human behavior and task performance has still not provided a reassuring degree of uniform opinion on this topic. Differences in the method for testing time estimation and differences in conditions under which the test was administered are possible reasons for the apparently contradictory results of these studies. A greater degree of consistency is evident in the results of work on the cardiovascular effects of carbon monoxide exposure.
Although the scientific basis for the present carbon monoxide air quality standard is far from complete, the most recent research results show that a large number of susceptible individuals in our population suffer impairment of their health at carboxyhemoglobin levels as low as 3%. If 3% carboxyhemoglobin is defined as the lowest level at which serious human health effects have been reliably observed, the 2% level cited in the Federal Register affords a relatively small safety margin for these individuals. The standard of 10 µg/m² for 8 hours actually gives about a two-fold safety margin at sea level. The standard of 40 µg/m² for one hour gives less than a 50% safety margin with moderate activity. In view of these relatively small margins between present standards and observed health effects, it does not seem that existing standards are unduly restrictive and it does not seem prudent that they should be raised. Although the margins are relatively small, the carboxyhemoglobin levels that will result from adherence to the standard is sufficiently close to background that no lowering of the standard is recommended either.
IV. Improving the basis for air quality standards:
It is evident that research results from EPA laboratories, as well as others only now becoming available have given a somewhat better estimate of the health consequences of low-level carbon monoxide exposure but that many questions still remain unanswered. The effects of ambient carbon monoxide on normal as well as cardiac disease remain largely unexplored. The lack of agreement on behavioral effects must eventually be resolved. Finally, the effects of chronic exposure – such as city dwellers receive year after year – have been superficially investigated only to the extent that gives rise to real concern. Astrup and his colleagues have found striking changes in the structure of arterial walls and development of arteriosclerosis after only a few weeks exposure (References 13, 14). Since altitude as well as carbon monoxide can cause oxygen deprivation, better quantitation of the altitude effect is needed particularly to clarify the added risk of carbon monoxide exposure to high altitude urban populations. These additional studies suggested here should be carried out and most of the other investigations cited here should be carried out and most of the other investigations cited here must be extended and corroborated by others before one can consider the standards to be based on indisputable scientific evidence.
From the Annals of Internal Medicine, November 1972
EFFECT of FREEWAY TRAVEL ON ANGINA PECTORIS
(By Wilbert S. Aranow, M.D., F.A.C.P., Clifford N. Harris, M.D., Michael W. Isbell, Stanley N. Rokaw, M.D., and Bruno Imparato, M.D.)
(Ten patients with angina had cardiopulmonary tests done in the control state, after being driven for 90 minutes during heavy morning freeway traffic, and 2 hours after return. The cardiac tests were repeated on a subsequent morning, with the patients breathing compressed, purified air during freeway travel. The expired-air carbon monoxide and arterial carboxyhemoglobin levels significantly increased after breathing freeway air. There was a significant decrease in exercise performance until angina, in systolic blood pressure at angina, in heart rate at angina, in systolic blood pressure times heart rate at angina, and in the FEV/FVC (forced expiratory volume in the
second forced vital capacity) after breathing freeway air. Ischemic ST-segment depression occurred in 3 of 10 patients while breathing freeway air. No significant change from control values occurred in the above variables after breathing compressed, purified air. Exposure to heavy freeway traffic increased carboxyhemoglobin levels, causing angina to develop sooner after less cardiac work.)
Smoking cigarettes causes a significant increase in carboxyhemoglobin levels in normal subjects and in patients with angina pectoris caused by coronary heart disease (1-3). The increased carboxyhemoglobin levels caused by smoking non-nicotine cigarettes decreases the rate of oxygen delivery to the myocardium, with angina pectoris developing sooner, after less cardiac work (3). In Los Angeles, an association between atmospheric carbon monoxide pollution and case fatality rates for myocardial infarction has also been observed (4). Carbon monoxide exposure has also been implicated in the pathogenesis of atherosclerosis (5).
A major source of carbon monoxide in urban atmosphere is automobile exhaust. Twenty million pounds of carbon monoxide per day were estimated to have been emitted by motor vehicles in Los Angeles during 1967 (4). Ambient carbon monoxide concentrations increase in the winter because the average inversion level occurs at lower altitudes (4).
Therefore this study was performed during the winter, to determine the effects of breathing ambient air, while subjects were being driven in heavy freeway traffic, on the carboxyhemoglobin levels and cardiopulmonary function of patients with normal pulmonary function (6) and with angina pectoris caused by documented coronary artery disease.
SUBJECTS AND METHODS
Ten men, between the ages of 40 and 56 years (mean age, 48 years), with classical exertional angina pectoris were studied. Six men had coronary artery disease documented by previous coronary angiography, with 50% or greater narrowing of the lumen of at least one major vessel. The other four patients had a previously documented transmural myocardial infarction, at least 1 year old. Seven patients were not smokers. Patients 1, 2, and 8 were instructed not to smoke for at least 8 hours before the performance of the study or during the study. Informed consent was obtained from the 10 men who participated in this study.
The patients were brought to the laboratory and familiarized with the equipment and the procedures before the study was done. The patients practiced exercising upright on a Collins bicycle ergometer and practiced having measurements made of their forced vital capacity (FVC) and of their second forced expiratory volume (FEV,). All patients were hospitalized overnight before each of the two study mornings.
On the first study morning, after 20-sec breath holding (7, 8), an expired-air sample was collected in a rubber-breathing bag and analyzed for carbon monoxide content, using a Beckman IR-215 nondispersive, infrared, carbon monoxide analyzers (7, 8). Then blood was drawn from a brachial artery and analyzed for carboxyhemoglobin and hemoglobin levels with a 182 Co-Oximeter s and for arterial Po, Pco, and pH with a Beckman 160 physiological gas analyzer.
All determinations were made in duplicate. Duplicate measurements of the PVC and FEV, were next obtained on each subject, with a Collins Respirometer-13 Liter. After this the resting blood pressure was measured with a mercury sphygmomanometer and the resting heart rate with an electrocardiograph, while the patient sat upright on the Collins bicycle ergometer.
Each patient then exercised upright on the constant-load bicycle ergometer, with a progressive work load (9) until the onset of angina, and the duration of exercise was recorded. Blood pressure and heart rate at the onset of angina pectoris were recorded. Electrocardiograms with a modified lead V., were made at rest and at the onset of angina. A Holter electrocardiocorder; model 350 G, was then attached to the subject. After this, the patient was then driven in a station wagon with windows open during early morning, heavy Los Angeles County freeway traffic for 90 minutes.
The subjects were dressed warmly to avoid chilling. Periodic blood pressure and heart rate measurements were made by a physician who accompanied the patients during the trip.
Immediately after return from the freeway and 2 hours later, the above protocol was repeated.
Approximately 3 weeks later the study was repeated for each patient, with two modifications.
Spirometric tests were omitted. During their 90-minutes freeway trip, the patients breathed compressed air from a tank through a mask, using a Bird Mark 8 Respirator with pressure settings and flow rates reduced and a built-in expiratory leak, so that significant positive pressure was not applied. The arterial blood gas determinations from these studies do not indicate hyperventilation effects or changes in Po, which would suggest improvement in distribution of ventilation. The time of day and freeway routes were identical for each patient during the two study mornings.
Ambient air measurements for carbon monoxide were made in the laboratory during each control period of this study. On four mornings, while the subjects were breathing compressed, purified air, the ambient air in the moving car during heavy traffic was collected twice, in a sample bag fabricated from Aluminized Scotchpack, and brought back for analysis for carbon monoxide. The study was done during January through March, 1972, in the south coastal area of Los Angeles County. An analysis of variance tests was done to analyze the data (10).
RESULTS
Table 1 shows the expired-air carbon monoxide and arterial carboxyhemoglobin levels for each patient in the control period, after breathing freeway air for 90 minutes, and 2 hours after return from the trip. There was a significant increase in the mean expired air carbon monoxide level, from the control period to immediately after breathing freeway air (mean difference, 16.2 ppm; critical difference, 8.7 ppm; P<0.001). There was also a significant increase in the mean expired air carbon monoxide level from the control period to 2 hours after breathing freeway air (mean difference, 8.3 ppm; critical difference, 3.7 ppm; P<0.001) .
There was a significant decrease in the mean expired-air carbon monoxide level immediately after breathing freeway air to 2 hours later (mean difference, 7.9 ppm; critical difference, 3.7 ppm; P<0.001). A significant increase was found in the mean arterial carboxyhemoglobin level from the control period to immediately after breathing freeway air (mean difference, 4.96%; critical difference, 1.03%; P<0.001). A significant increase was also found in the mean arterial
carboxyhemoglobin level, from the control period to 2 hours after breathing freeway air (mean difference, 1.79%; critical difference, 1.03 % ; P<0.001). The mean arterial carboxyhemoglobin level, from immediately after breathing freeway air to 2 hours later, was significantly decreased (mean difference, 117%; critical difference, 1.03%; P<0.001). All the hemoglobin values were within normal limits and showed no significant change.
Table 2 indicates the expired-air carbon monoxide and arterial carboxyhemoglobin levels for each patient in the control period, immediately after breathing compressed, purified air during a freeway trip for 90 minutes, and 2 hours after return from travel. There was no significant difference between the mean expired-air carbon monoxide levels in the control period, after breathing compressed, purified air during the freeway trip, and 2 hours later. Neither was there a significant difference between the mean arterial carboxyhemoglobin levels in the control period, after breathing compressed purified air during the freeway trip and 2 hours later.
The ambient carbon monoxide in the laboratory ranged between 1 and 3 ppm in the control periods. The mean, ambient carbon monoxide in the laboratory was 2+1 ppm for the control period before breathing freeway air and 2+1 ppm for the control period before breathing compressed, purified air. The ambient carbon monoxide in the moving car ranged between 42 and 63 ppm, with a mean of 53+6 ppm, on the four mornings it was sampled while the patients were breathing freeway air during peak early-morning traffic. The ambient carbon monoxide in the moving car ranged between 37 and 61 ppm, with a mean of 47+8 ppm, on the five mornings it was sampled while the patients were breathing compressed, purified air during early-morning traffic.
Table 3 illustrates the exercise performance until the onset of angina for each patient in the control periods, after breathing freeway air for 90 minutes, after breathing compressed, purified air for 90 minutes during the freeway trip, and 2 hours after return from the freeway trips. The mean exercise performance, from the control period to after breathing freeway air, was significantly decreased (mean difference. 75.1; critical difference, 63.5; P<0.001). The mean exercise performance, from the control period to 2 hours after breathing freeway air, was significantly decreased (mean difference, 39.1; critical difference, 30.1; P<0.05). The mean exercise performance, from immediately after breathing freeway air to 2 hours later, was significantly increased (mean difference, 36.3 critical difference, 30.1: P<005) The mean resting systolic blood pressure was 115.8+8.6 mm Hg and 116.6+9.0 mm Hg in the control periods, 119.0+-11.8 mm Hg after breathing freeway air for 90 minutes, 118.0+8.3 mm Hg after breathing compressed, purified air for 90 minutes, and 117.4 +9.6 mm Hg and 119.0+8.3 mm Hg 2 hours after return from the freeway trips. These values were not significantly different.
The mean resting diastolic blood pressure was 79.4+5.7 nun Hg and 79.0+4.3 mm Hg in the control periods, 79.2+6.5 mm Hg after breathing freeway air for 90 minutes, 79.2 +-3.8 mm Hg after breathing compressed, purified air for 90 minutes, and 79.6+4.6 min Hg and 78.0-3.6 mm Hg 2 hours after return from the freeway trips. These values were not significantly different.
The mean resting heart rate was 80.6+5.0 and 81.8 – 6.6 in the control periods, 78.4+5.2 after breathing freeway air for 90 minutes, 79.0+5.8 after breathing compressed, purified air for 90 minutes, 77.6+5.6 and 79.0 – 5.8 2 hours after return from the freeway trips. These values were not significantly different.
The mean product of resting systolic blood pressure times heart rate was 9345+997 and 9565+1325 in the control periods, 9341+1197 after breathing freeway air for 90 minutes. 9322+1013 after breathing compressed. purified air for 90 minutes, and 9130+1147 and 9422+1174 2 hours after return from the freeway trips. These values were not significantly different.
Table 4 indicates the systolic and diastolic blood pressure at the time of angina, the heart rate at the time of angina, and the product of systolic blood pressure times heart rate at the time of angina for each patient in the control period, after breathing freeway air for 90 minutes, and 2 hours after returning from the trip. The following significant differences were observed. There was a significant decrease in the mean systolic blood pressure at the time of angina from the control period to immediately after breathing freeway air (mean difference, 4.8; critical difference, 3.9; P less than 0.01). There was also a significant decrease in the mean heart rate at the time of angina from the control period to immediately after breathing freeway air (mean difference, 19.5; critical difference, 6.9 P less than 0.001) ; and to 2 hours after breathing freeway air (mean difference, 8.9; critical difference, 6.9; P less than 0.001). The mean heart rate at the time of angina, from immediately after breathing freeway air to 2 hours later, was significantly increased (mean difference, 10.6; critical difference, 6.9; P less than 0.001). The mean product of systolic blood pressure times heart rate at the time of angina, from the control period to after breathing freeway air, was significantly decreased (mean difference, 3122; critical difference. 1205; P less than 0.001). There was also significant decrease in the mean product of systolic blood pressure times heart rate at the time of angina, from the control period to 2 hours after breathing freeway air (mean difference, 1635; critical difference, 1205; P less than 0.001). A significant increase in the mean product of systolic blood pressure times heart rate at the time of angina, from after breathing freeway air to 2 hours later, was found (mean difference, 1687; critical difference, 1205; P less than 0.001).
Table 5 shows the systolic and diastolic blood pressure at the time of angina, the heart rate at the time of angina, and the product of systolic blood pressure times heart rate at the time of angina for each patient in the control period, after breathing compressed, purified air during a freeway trip for 90 minutes, and 2 hours after return from travel. No significant difference was found in the mean systolic and mean diastolic blood pressure at the time of angina, in the mean heart rate at the time of angina, and in the mean product of systolic blood pressure times heart rate at the time of angina in the control period, after breathing compressed, purified air during the freeway trip, and 2 hours later.
Table 6 shows the forced vital capacity (FVC), the forced expiratory volume in 1 second (FEV), and FEV X 100/FVC for each patient in the control period, after breathing freeway air for 90 minutes, and 2 hours later. The following significant difference was found: a significant decrease in the mean FEV X 100/FVC, from 81.55 in the control period to 79.5% after breathing freeway air for 90 minutes (mean difference, 2.0; critical difference. 1.4; P <0.05). This significant difference was due to a marked decrease in FEV, X 100 /FVC in Patient 8, from 77 to 67% from the control period and after breathing freeway air for 90 minutes. No other subject showed a significant change in pulmonary function after breathing freeway air.
The mean arterial Po. levels were 89.1 + 4.2 mm Hg in the control period, 88.3 + 5.4 mm Hg after breathing freeway air for 90 minutes, and 88.9 + 3.9 mm Hg 2 hours after returning from the freeway trip; for the second trip the levels were 90.4+ 3.0 mm Hg in the control period, 90.7 + 2.3 mm Hg after breathing compressed, purified air for 90 minutes, and 90.8 + 2.5 mm Hg 2 hours after return from the freeway trip. These values were not significantly different.
The mean arterial Pco., levels were 38.7 + 0.6 and 38.9 + 0.8 mm Hg in the control periods, 38.3 + 1.3 mm Hg after breathing freeway air for 90 minutes, 39.2 + 0.8 mm Hg after breathing compressed, purified air for 90 minutes, and 38.6 + 0.9 and 38.8 + 0.7 mm Hg 2 hours after return from the freeway trips. These values were not significantly different.
The mean arterial pH levels were 7.40+ 0.01 in the control periods, after breathing freeway air for 90 minutes, after breathing compressed, purified air for 90 minutes, and 2 hours after return from the freeway trips.
While breathing freeway air, three (Patients 1, 8, and 10) of 10 patients (30% ) developed ischemic ST-segment depression of at least 1 mm greater amplitude than in the control Holter electrocardiographic recordings. In other words, significant ST-segment depression occurred only during actual inhalation of freeway air. None of 10 patients developed ischemic ST-segment depression greater than that observed in the control Holter electrocardiographic recordings while breathing the compressed, purified air during the freeway trip. A fourth patient developed very frequent premature ventricular beats while breathing freeway air and again while breathing compressed air during the freeway trip.
Figure 1 illustrates the Holter electrocardiographic recording of Patient 1, with ischemic ST-segment depression while breathing freeway air.
No significant differences were observed in the electrocardiograms at rest and at the onset of angina in the control periods, on return to the laboratory immediately after breathing freeway air for 90 minutes, on return to the laboratory immediately after breathing compressed, purified air during a freeway trip, and 2 hours after return from the freeway trips.
DISCUSSION
Our data show that the mean expired-air carbon monoxide level and mean arterial, carboxyhemoglobin level significantly increased after travel in Los Angeles County heavy, early- morning freeway traffic during winter months. This increase in mean arterial carboxyhemoglobin level was less than that found after heavy cigarette smoking (1-3) but was, nevertheless, of sufficient magnitude to cause angina pectoris to develop sooner after less work in our patients.
Chevalier, Krumholz, and Ross (11) have reported that nonsmokers who inhaled carbon monoxide to raise their carboxyhemoglobin level to the range seen in a control group of smokers developed an increased oxygen debt with exercise. Ayres and his associates (12) have shown that acute elevation of carboxyhemoglobin levels causes an increase in coronary blood flow in patients with non-coronary heart disease but not in patients with coronary heart disease.
Myocardial oxygen extraction and extraction ratios significantly decreased in both groups of patients, but the myocardial lactate extraction ratio significantly changed to production only for the patients with coronary heart disease (12). Ayres and his co-workers (13) have also stated that carbon monoxide decreases myocardial oxygen tension by three mechanisms: (1) decreased oxygen extraction, (2) decreased capillary oxygen tension because of the leftward shift of the oxyhemoglobin dissociation curve, and (3) increased ventricular work oxygen demand owing to stimulation of the adrenergic system.
Since our patients with documented coronary artery disease could not adequately increase their coronary blood flow while exercising and since the elevated carboxyhemoglobin levels caused by freeway exposure made less oxygen deliverable to the myocardium, their myocardial oxygen demand exceeded their myocardial oxygen supply, inducing angina pectoris sooner after equivalent exercise. That the angina-inducing effect. was related to carbon monoxide acquisition rather than to the stress of freeway travel is supported by the absence of this effect when compressed, purified air was supplied during the journey. It is also compatible with angina pectoris after equivalent exercise developing sooner after the acquisition of carbon monoxide from smoking non-nicotine cigarettes (3).
Sarnoff and his associates (14) have shown that the primary hemodynamic determinant of myocardial oxygen consumption is the total tension developed by the myocardium (heart rate times the area under the systolic study shows that the product of systolic blood portion of the aortic pressure curve). Our pressure times heart rate at the time of exercise-induced angina was significantly less after breathing freeway air for 90 minutes than in the control period or after breathing compressed air during a freeway trip. This result strongly suggests that less myocardial work can be done before the onset of exercise induced angina in patients with elevated carboxyhemoglobin levels because less oxygen is available to the myocardium.
This study was not double-blind, and some unconscious bias may have been introduced by the awareness of the exposure of the investigators and the subjects. But the consistent decrease in exercise performance until the onset of angina in every patient after breathing freeway air and the consistent decrease in product of systolic blood pressure times heart rate at the onset of angina in every patient after exposure to freeway air, associated with a consistent rise in expired air carbon monoxide and carboxyhemoglobin levels in every patient after breathing freeway air, suggests that unconscious bias did not significantly affect our results.
Only one of our 10 patients (10%) had a significant decrease in FEVx100/FVC from 77% to 67% after being exposed to freeway air for 90 minutes. This patient complained during his freeway trip that the atmosphere was “very smoggy”. The airway obstruction he developed was accompanied by a drop in his arterial Po, from 92 mm in the control period to 85 mm immediately after his return from freeway travel, which suggests, in the absence of under-ventilation, a coincident disturbance of ventilation perfusion relationships. Several of the chemical agents known to pollute the Los Angeles County atmosphere – nonoxidants, SO2 and some particulates – may have produced the ventilatory abnormalities in this patient. No assessment was made of the possible contribution of these pollutants to the effects on cardiac function observed in our study patients.
The observation that 3 of our 10 patients (30%) developed ischemic ST-segment depression during heavy freeway traffic while breathing freeway air whereas none of our 10 patients developed ischemia ST-segment abnormalities during heavy freeway traffic while breathing compressed, purified air was a disturbing finding. The increased carboxyhemoglobin levels, plus exposure to other pollutants and the stress of being driven during heavy freeway traffic may have precipitated the electrocardiographic abnormalities.
(ACKNOWLEDGMENTS: The authors express their appreciation to Reed Boswell, Ph.D., for biostatistical analysis of the data and to Jewell Kietzman for performing the arterial blood gas determinations.
(To be presented in part October 1972 at the annual meeting of the American College of Chest Physicians in Denver, Colorado.
(Received 30 May 1972; revision accepted 28 July 1972.)
OF CARS ... CARBON MONOXIDE AND HUMAN HEALTH
BACKGROUND MEMORANDUM,
JUNE 1972
"The new technological man carries strontium-90 in his bones, iodine-131 in his thyroid, DDT in his fat, and asbestos in his lungs."
That was the graphic picture drawn by Dr. Barry Commoner of Washington University in testifying several years ago before a Senate subcommittee holding hearings on the multiple threats to the environment.
It was a vivid enough but, of course, not a complete picture, nor was it meant to be.
Increasingly, in recent years, the public has become concerned over an almost-fantastic morass of environmental pollution problems: the poisoning of lakes, contamination of rivers, mounting accumulations of solid waste, threats from unsafe foods, and, most recently, even electronic pollution as electric gadgets in homes and factories emit stray signals that jam air-to-ground communication at airports.
Not the least of all pollution is that of the air. It may, in fact, deserve to rank at or very near the top in importance. It is a problem in all large American cities; it is a problem in more and more small towns. Each year over 200 million tons of manmade waste products are released into the air over the U.S.
It is no new knowledge that at levels common in urban areas, air pollution contributes to the incidence of such chronic diseases as emphysema, bronchitis and asthma, which has increased dramatically inn recent decades.
But it now appears that air pollution may threaten health in many other ways:
That it may contribute significantly to coronary heart disease and heart attacks;
That it may have undesirable effects on unborn babies;
That it may affect the nervous system and impair mental performance in young and old:
That it may, through impairing the performance of drivers, contribute to highway fatalities.
Increased understanding of these and other diverse effects of air pollution has been emerging from studies of the subtle effects of one ingredient in such pollution: carbon monoxide. CO has ranked, in terms of sheer weight, as the biggest of air pollutants. Its importance now is growing on the basis of findings about what low-level, long-term exposure to it may provoke.
And if, as the scientific literature increasingly indicates, such exposure can help explain some of our most important health problems, there is a bright side: the availability of the technology to make a deep dent in the CO pall.
CLEOPATRA'S NEEDLE AND OXYGEN BUMPING
On a knoll in Central Park stands a 224-ton granite obelisk called Cleopatra's Needle, carved in 1600 B.C., a gift to New York City by the government of Egypt. Originally, the obelisk had hieroglyphic characters cut into all four of its sides; they were visible when the Needle arrived in New York 90 years ago. Now the markings have been obliterated from – and, in fact several inches of granite have been chewed off – the south and west sides which face prevailing winds and air pollution concentrations. In less than a century in New York more damage has been done to Cleopatra's Needle than in 3½ millennia in Egypt.
It has occurred to some to point out that if polluted air can eat away rock, it can hurt people.
There is much in polluted air that can be injurious: sulphur dioxide which in sufficient quantities can cause men's eyes to sting and women's stockings to disintegrate; nitrogen oxides which help cause smog; particulates, known to laymen as dust, which generally are more nuisance than health menace because for the most part they are too large to get into lung tissues (but some smaller particles, including asbestos fibers have been linked to lung disease and cancer).
But it is carbon monoxide which is the most abundant air pollutant.
Of the more than 200 million tons of toxic materials annually released into U.S. air, 100 million tons are CO – almost as much as the rest, the 33.2 million tons of sulfur oxides, 20.6 million of nitrogen oxides, 32 million of hydrocarbons, and 28.3 million of particulates, combined.
Many sources – factories, power plants, municipal dumps, private incinerators – contribute, but 60% of total pollutants come from internal combustion engines. And, according to the Public Health Service, the 100 million motor vehicles in the country produce two-thirds of the tonnage of carbon monoxide.
CO is dangerous because it hampers delivery of oxygen to body tissues.
The gas combines with hemoglobin in blood, forming carboxyhemoglobin (COHb). Hemoglobin ordinarily carries oxygen from the lungs; COHb cannot. Because hemoglobin's affinity for CO is 210 times greater than for oxygen, a small quantity of CO can inactivate a substantial percentage of oxygen-carrying capacity.
When a normal person breathes air – completely devoid of CO – he still has about 0.4% COHb in his blood. This is a background level resulting from the body's own production of small amounts of CO.
When air contains CO, additional COHb is formed in the blood. With a concentration of 30 parts per million (ppm) of CO in air, 5%. more COHb will be formed, inactivating that percentage of total hemoglobin. A 60 ppm CO concentration will inactivate 10% of the hemoglobin. If an individual is a smoker, another 5% of his hemoglobin will be inactivated by the CO in cigarette smoke.
CO is taken up rapidly but given off slowly.
After exposure to it is ended and pure air breathed again, only 15% of the CO combined with hemoglobin is eliminated in exhaled air per hour. It takes 3 to 4 hours to eliminate one-half of whatever amount of CO may be in the blood.
CO is insidious. It has no warning characteristic, no odor or color.
At increasing concentrations, it does begin to produce warning signs and symptoms of injury, usually in this sequence: headache, dizziness, lassitude, flickering before the eyes, ringing in the ears, nausea, vomiting, palpitations, pressure on the chest, difficulty in breathing, apathy, muscular weakness, collapse, unconsciousness, and death.
Actually, decades ago, when people were exposed to CO from coal-burning heating devices and leaky illuminating-gas fixtures, studies of CO effects showed that a healthy person might survive acute exposure, even of several days, to moderately high levels of the gas producing 20 to 40% of COHb in the blood.
Today the concern is not so much over acute exposure to high levels. It is over the harmful effects of long exposure to low-level CO concentrations in a range that would produce even less than 10% COHb, concentrations increasingly common in the air Americans breathe.
BEHAVIORAL CONSEQUENCES
As far back as World War II, some investigators had found some indications that vision might be affected when COHb concentration reached about 5%.
Much more recently, Drs. Rodney R. Beard and George A. Wertheim of Stanford University School of Medicine, aware of how little was know about the effects of low level CO exposure, undertook intensive studies looking for any subtle behavioral changes. They started with rats, then went on to humans.
Rats can be trained to press levers to obtain food pellets. They can even be trained to work on a fixed-interval schedule under which a food pellet is provided only when a lever is pressed, say, three minutes after a previous pellet has been delivered. The animals learn that pressing any sooner is useless.
Beard and Wertheim trained their rats, then went on to note the effects of introducing increasing concentrations of CO. They soon found that after only 11 minutes of exposure to CO at 100 ppm concentration, the time discrimination of the rats went haywire.
They studied the effects of CO exposure on the ability of healthy young human adults to discriminate short intervals of time. The subjects were presented with pairs of signals. The first tone of the pair was always one second in duration; the second varied between 0.675 and 1.325 seconds. The subjects were supposed to indicate whether the second tone was shorter, identical, or longer than the first. After as little as 90 minutes exposure to 50 ppm CO, there was a marked reduction in the proportion of correct responses.
In vision tests with healthy young adults, it turned out that after 27 minutes exposure to 50 ppm CO, there was a 5½% impairment of visual acuity, and after an hour a 17% impairment.
As Dr. Beard recently reported: "It appears that a little COHb can cause a significant reduction of oxygen supply to a tissue with high oxygen demand such as the brain (which) only weighs 2-2.5% of total body weight yet gets about 15% of the blood and its need is for about 20% of the oxygen requirement of the whole body."
THE DRIVING BRAIN
As evidence of sensory impairment from low-level CO exposures has begun to accumulate, investigators have begun to think it important to determine whether such effects in motor vehicle drivers could predispose them to accidents.
Actually, during the period 1959 to 1963, one investigator had analyzed blood samples from three population groups: 1,672 from drivers thought to be responsible for accidents, 3,818 from workers sometimes exposed to CO on the job, and 1,518 from individuals suspected of having been exposed to domestic sources of CO. He found that drivers involved in accidents had the highest COHb levels.
More recently, Dr. Thomas H. Rockwell, director of the Ohio State University driving research laboratory, and an associate, A. M. Day, set out to determine under actual road conditions what the influence of CO might be.
They were aware of surveys which had shown that about 4 % of highway drivers have vehicle cabin concentrations greater than 100 ppm of CO (some as high as 300), and about 8% have cabin concentrations greater than 50 ppm. Such concentrations would result, if the drivers were non-smokers, in 8 % having greater than 7 % COHb concentrations and in 4% having greater than 14%. Smokers would have considerably higher COHb levels.
Rockwell and Ray studied drivers operating instrumented vehicles on the highway under known and controlled CO concentration levels. They found effects, at levels of COHb below 10%, that they called "definitely detrimental to driving performance," and they emphasized that "These are at levels found in the driving population and at which drivers are unaware of any symptoms such as headache or other CO warning signals."
Among the effects were reduced ability of a driver to respond to changes in velocity of the car in front of him, reduced ability to detect and respond to tail light intensities, decreases in ability to estimate time.
In view of the results of their exploratory study and the potential benefits for highway safety, Rockwell and Ray urged further research into the effects of CO on driving performance.
THE HEART
Can CO be a significant factor in heart disease? More and more studies now suggest so.
Recently, at Johns Hopkins University, Dr. Thomas J. Preziosi and associates found abnormal electrocardiograms appearing in all dogs exposed to 100 ppm CO concentrations. In another study, when 15 dogs were exposed to 50 ppm CO concentration for six weeks – 7 of them exposed for 6 hours a day for 5 days a week, the others continuously – all showed EKG changes and, at autopsy, dilatation of the right side of the heart, with scarring of the heart muscle in some cases and fatty degeneration in others.
Some years ago, a study was made of 3,080 patients with heart attacks admitted to 35 hospitals in Los Angeles County. No significant association could be found between carbon monoxide air pollution and the occurrence of the heart attacks – but there was a significant correlation between fatality rates from the heart attacks and ambient CO levels during the week of admission. Patients admitted to hospitals in areas where the weekly CO concentration ranged from 8 to 14 ppm had markedly greater death rates than others.
Recently, at St. Vincent's Hospital and Medical Center, New York City, Dr. Stephen M. Ayres and associates carried out studies in dogs and humans exposed to various CO concentrations. At concentrations raising COHb levels to between 5 and 10%, they found many changes, including alterations in blood flow through the coronary arteries feeding the heart.
Among 41 patients with heart disease, the hearts of half of them suffered a serious oxygen shortage when 5 to 10% of their hemoglobin was tied up as COHb. For a patient with a heart already deprived of plentiful oxygen because of diseased, fat-choked coronary arteries, Ayres and his group report, even extremely low levels of COHb could make for a critical enough shortage of oxygen to fatally disrupt heart rhythm.
They also considered one of medicine's major mysteries: Why does angina pectoris, the chest pain associated with coronary artery disease, sometimes appear in patients who show no X-ray evidence of coronary artery disease? Lack of enough oxygen to meet the demands of physical activity, with the lack caused by relatively large concentrations of COHb, might well be the answer, they report.
Actually, as the St. Vincent's researchers have noted, the effects of CO inhalation on the heart are similar to the reported effects of tobacco smoking.
And, indeed, recent studies on the previously overlooked importance of CO as one of the most harmful ingredients in cigarette smoke are helping shed new light on the importance of CO from all sources. (The 1972 report of the Surgeon General of the Public Health Service on "The Health Consequences of Smoking" places emphasis on CO in cigarette smoke, smoke as a hazard not only for smokers but for nonsmokers exposed to CO in rooms filled with tobacco smoke.)
At a New York Academy of Sciences 3-day meeting in 1970 on CO, Dr. Poul Astrup of the University of Copenhagen reported that he had found high COHb concentrations in young smokers with severe hardening of the arteries. And, spurred by this, he had undertaken animal experiments to determine the significance of CO exposure in artery hardening. He had been able to see injurious effects with small COHb concentrations. With levels of 10 to 20% COHb, cholesterol deposits in rabbits' arteries increased by 3 to 5 times, and there were many injuries to artery walls.
Urging further studies, Dr. Astrup emphasized that "If the results of such studies confirm a role for CO in development of arteriosclerosis in smokers, they will be of interest from a preventive point of view concerning all types of exposure of man to this toxic gas."
At the same meeting, Dr. Lawrence D. Longo of Loma Linda University reported that CO in the blood of women who smoke during pregnancy may explain at least in part why such mothers tend to have smaller babies. It may explain, too, why investigators have been finding that in infants of smoking mothers the onset of crying immediately after birth is delayed and some infants show definite evidence of asphyxia with irregular breathing and blueness. In addition to making less hemoglobin available for oxygen transport, Dr. Longo noted, increased concentration of CO fetal blood may act to inhibit vital enzyme systems.
AN EXTRA-MYSTERIOUS-MECHANISM
In fact many investigators, impressed by the effects of small concentrations of CO, think that the gas must do something more than interfere with oxygen transport.
Says Dr. R. E. Forster of the University of Pennsylvania School of Medicine: "Detectable effects upon certain. human mental and bodily processes have been reported at such low concentrations that it is difficult to provide an explanation. We should, therefore, not neglect other possible mechanisms (beyond oxygen transport) by which CO call affect physiological processes."
Forster thinks that inasmuch as investigators have found decreases in visual sensitivity caused by 3 to 5 % COHb in blood, with no immediate improvement when the COHb is washed out, it could be that CO may react with some other compound in the central nervous system with even greater affinity than it does with hemoglobin.
Dr. Arthur B. Otis of the University of Florida echoes the need to "remain alert to the possible effects of CO that are not attributable to COHb formation."
Some investigators like Dr. Britton Chance of the University of Pennsylvania believe that CO may have effects at the cell level, producing changes in the mitochondria, small bodies within cells that are involved in cell breathing. That might help explain why when CO levels in the blood diminish after an exposure and more hemoglobin once again becomes available for oxygen transport, visual impairment or other disturbances may persist for long periods.
THE SPECIALLY SUSCEPTIBLE
People who might be expected to be most susceptible to CO include those with severe anemia, chronic lung disease, or impairment of circulation to the heart or other organs; and unborn babies and newborn infants.
Given perfect health, an individual may tolerate some increase of CO concentration in air although at the cost of reserve capacity for oxygen transport. He may adapt through mechanisms such as increased work by the heart to pump the blood more often through the body and by changes in lung ventilation.
But, as a National Academy of Sciences-National Academy of Engineering Environmental Studies Board reported in 1969, the adaptive processes themselves may impose a continuing burden on physical reserves. And if normal people can compensate for some increase in CO levels, these with pre-existing medical conditions might be more susceptible to the effects of even relatively small increases.
A study which may reflect this increased susceptibility was reported in April, 1971 by Drs. Alfred C. Hexter and John R. Goldsmith. Carried out in Los Angeles County over a 4-year period, it showed a significant association between community CO concentrations and overall mortality. As compared with a 7.3 ppm CO concentration, the lowest level observed, when the concentration reached 20.2 ppm, the CO contribution to mortality for that day was 11 deaths.
When, late in 1971, Dr. John J. Hanlon, Assistant Surgeon General, Public Health Service, surveyed the situation in a report in Modern Medicine, he noted that as little as 10 ppm of CO in the atmosphere, "a level common in many cities," had been shown to impair mental performance; that elevated CO levels had been associated with increased probability of motor vehicle accidents and with the ability of individuals to survive heart attacks.
"While admittedly," he wrote, "several other factors are involved, it is notable that death rates from coronary heart disease are 37% higher for men and 46% higher for women in metropolitan areas with high atmospheric pollution levels than they are in non-metropolitan areas. Excessive levels of carbon monoxide have also been associated with decreased ability of patients with cirrhosis of the liver to survive.
"While many other examples might be presented," he wrote, "the seriousness of the situation is perhaps best illustrated by the recent recommendation of the California Department of Health that physicians should be prepared to estimate the contribution air pollution may make to the outcome of a patient's illness and that in areas with high air pollution some patients may benefit from a preoperative period in a clean-air room or chamber before being administered general anesthesia."
HOW MUCH CO DO WE BREATHE?
The exposure to CO in the air is indeed worrisome.
Says Dr. W. H. Forbes of Harvard: "The streets of our big cities vary greatly in the amount of carbon monoxide they contain, but not infrequently may reach peaks of 100 ppm ... On occasion certain streets may go even higher, and in the region just behind a large, slow moving or stationary truck, it may reach ten times this figure or 1,000 ppm."
Dr. A. J. Haagen-Smit of the California Institute of Technology, using a new carbon monoxide analyzer, set out to determine the CO concentration which a Los Angeles rush hour commuter breathes while driving on freeways. He made eight trips on the Pasadena and Harbor freeway between the Institute and Exposition Park, passing through the congested area in the heart of Los Angeles. In the best trips made, the overall average level of CO measured was 37 ppm. In heavy traffic, moving at less than 20 mph, the level rose to an average of 54 ppm. with peaks up to 120 ppm.
"A commuter spending two hours at the higher exposure levels found in this study," he reported, "would have a CO level in his blood which would be approximately that cited by the Health Department of the State of California for the 'serious' level. A concentration of 30 ppm of the gas will inactivate 5% of a person's hemoglobin and 60 ppm will inactivate 10% of it. If a person is smoking, another 5 % of his hemoglobin in inactivated."
According to Dr. John R. Goldsmith of California's Department of Public Health, there is enough CO in Los Angeles air at times to reduce the blood's oxygen carrying capacity by 20%. He also notes that air monitoring in Chicago, Denver, Philadelphia and Washington has found each with an 8-hour average in excess of 30 ppm of CO.
A recent study of commuter CO exposures in 15 large American cities found that in 11 of the 15 people in moving vehicles, particularly those in heavy traffic, are at times exposed to sustained levels of 50 ppm or more, with peaks as high as 147 ppm in Los Angeles arterial traffic and 141 ppm in New York expressway traffic.
New York State guidelines warn that CO levels should not exceed 15 ppm more than 15% of the time during an 8-hour period. But after four months of measurements in midtown Manhattan, according to a Wall Street Journal report in mid-1970, it was determined that CO levels remained above that all day every day. During daytime hours when traffic is heaviest, the CO level in Manhattan often soars to between 25 and 30. In some areas, such as the Lincoln Tunnel and approaches to the George Washington Bridge, the level reaches 100 ppm – nearly seven times the "safe" level. If some authorities are right, it is CO exposure which may account for the reputed surliness of both New York cops and cab drivers.
As might be expected, during periods of air stagnation which occur in most urban communities, the levels of CO and other air pollutants build up. On one such occasion in downtown London, a peak level of 235 ppm CO was measured.
Traffic slowdowns also greatly augment CO emissions from cars. For example, while an average car emits about one-ninth of a pound of CO per mile of travel at 25 mph, it pours out about one-third of a pound at 10 mph.
Adding to the worrisomeness of the automotive CO emission problems is the fact that, as the National Academy of Sciences-National Academy of Engineering Committee has noted, "because of the rapidly growing number of motor vehicles, CO is produced in such large quantities that it is no longer a problem only in the immediate vicinity of traffic, but has become a factor within entire communities."
The spreading nature of the problem is underscored sharply by studies projecting future CO emissions and concentrations from urban traffic data.
In one of them, investigators calculated present and future CO emissions and ambient CO concentrations in Washington, D.C. They found that without emission control the total CO emitted in the city would double by 1985, and would do so in a nonuniform manner. The smallest increases would occur downtown, the greatest in the outskirts. Not only would the total area of the city exposed to high emission densities be increased but in the downwind parts of the area people would be exposed to high concentrations for longer periods of time than previously because cleaner air would have to come from further away.
It's hardly any wonder, then, that the Environmental Protection Agency, as called for under the Clean Air Act Amendments passed by Congress in 1970, has issued stringent standards requiring a 90% reduction in CO emission by 1975 model automobiles. Also for such models, hydrocarbon emissions are to be reduced by 90%.
Is that possible?
THE ANSWER IN THE CATALYSIS PHENOMENON
It's a phenomenon known to anyone who ever had a course in high school chemistry.
Bring together two materials that don't react readily. No reaction. But add a third material of the right kind and the original two now react with each other while the third just sits there, unchanged by the events. The third material is a catalyst.
Catalysts are widely used in industry. They are also vital materials in human physiology, the hundreds upon hundreds of enzymes in the body are catalysts – they make possible all the chemical reactions that go on at the relatively low body temperature (so far as chemical reactions are concerned) of 98.6º F.
Catalysis provides a low energy path for generating or assisting a chemical reaction. In effect, according to one theory, a catalyst takes one reluctant chemical by the hand and leads it to another and, having performed the introduction and urged a mating, bows out of the picture while the two bond. And the catalyst remains intact, ready to repeat its function.
CO is the result of incomplete combustion. In the combustion, carbon unites with only one atom of oxygen to produce the toxic CO. If it unites with a second atom of oxygen, the result is harmless carbon dioxide.
Once carbon monoxide is formed, it won't ordinarily pick up an extra oxygen atom no matter how plentiful such atoms are in the air. But introduce the right catalyst when there is a mix of CO and air and immediately the CO is converted to the harmless CO2.
That's the principle behind catalytic exhaust purifiers which are expected to provide the means for meeting the government standards for 90% reduction in CO emission – and also for 90% reduction in hydrocarbon emission. For catalysis can further not only the reaction of carbon monoxide with oxygen in the air but also the reaction of hydrocarbons with the oxygen, with the end results being carbon dioxide and water vapor.
One such device is a stainless steel cylinder about 12" long, somewhat resembling an artillery shell casing. It fits on an auto's exhaust pipe, near the manifold, and funnels the exhaust through a ceramic honeycomb impregnated with a catalyst. The catalyst is platinum. Only a tiny amount of that precious metal, a remarkably effective and durable catalytic agent, is needed.
For more than half a dozen years, the same type of device has been used on fork lift trucks operating in confined spaces such as warehouses, ship holds, and mines. It also is widely used on auxiliary generators located close to air conditioning intakes, on repair equipment in subway systems, and in other situations where purity of exhaust is, beyond esthetics and comfort, a matter of life or death.
But the internal combustion engines on which the device has been used have differed in one critical respect from automobiles.
They are fueled by liquid propane gas, natural gas, diesel oil or unleaded gasoline. The device cannot be used effectively on an engine, fueled, as automobiles have been, by leaded gas. The lead rapidly poisons the catalyst and renders it inactive.
Almost from the beginning, Engelhard Minerals & Chemicals Corporation, which developed the device (called PTX© catalytic exhaust purifier) tried to adapt it for automobiles. There were similar efforts by other companies and researchers working with catalytic agents of many types.
But the efforts came to nothing because of leaded gasoline.
Then, early in 1970, stimulated by the Congress' obviously serious intent to cut automotive air pollution and by a drive in the automotive and petroleum industry toward producing and using lead-free fuel, Engelhard reactivated a research and development program. Other companies have done the same.
Also in 1970, college students preparing cars for entry in a Clear Air Car Race, a cross-country run under the sponsorship of the Massachusetts Institute of Technology and the California Institute of Technology, sought catalytic purifiers. When the race was run, 26 of the 37 entries which started and finished were equipped with the PTX purifier in various configurations and catalyst loadings, including the overall winner of the 3,600 mile race and the winners in all the internal combustion classes.
The overall winner was a 1971 Ford Capri modified by four students at Wayne State University. It used unleaded gasoline. Upon reaching California tests showed that where an uncontrolled car engine can spew 73 grams of CO and 11.2 of hydrocarbons per mile, the Wayne State car emitted 1.48 of CO and 0.19 of hydrocarbons per mile, more than meeting the 1975 standards.
In announcing "the race winners, David Ragone, Dean of the D. Thayer School of Engineering at Dartmouth and chairman of the judging committee, observed: "The public is another of the winners in this race."
Development work has been stepped up since then. Federal standards require that the catalyst remain effective for a minimum of 50,000 miles. In one Automobile Manufacturers Association driving schedule test of a station wagon equipped with PTX, the catalyst was removing 86% of CO and 96% of hydrocarbons at 570 miles and, at 48,307 miles, was removing even more CO (97%) and 82% of hydrocarbons.
A Federal Test Procedure, issued after the first test work began, calls for a maximum emission of 3.4 grams of CO and 0.41 of hydrocarbons per vehicle mile. Data on various PTX catalytic converters designed to meet the standards under the procedure show a range down to 1.2 grams of CO and 0.1 gram of hydrocarbons per vehicle mile. And currently PTX-equipped cars are being driven scores of thousands of miles on New Jersey roads by three shifts of drivers.
The likelihood that a platinum catalyst, given unleaded gasoline, can stand up not only for 50,000 miles but for several times that is strengthened by experience on forklift trucks. For such vehicles, 2,000 hours of operation is equivalent to 50,000 miles of car use. On fork lift trucks using lead-free fuel the platinum lasts up to 10,000 hours.
At $120 an ounce, platinum is expensive. But because it is so catalytically active, only a very small fraction of an ounce – no more than $15 worth – is needed and it can be recovered through the same commercial channels as lead in a used auto battery is recovered. A platinum catalyst is favored over basic metal types because the latter necessarily are much bulkier, difficult to place on a car because of the mass, and difficult if not impractical to recover and recycle.
With mass production, platinum catalytic converters are expected to cost automobile companies less than $50 per car.
As of now, catalytic devices are not available to solve the problem of oxides of nitrogen (NOx) pollution. The present approach to the NOx problem involves the recirculation of exhaust gases which is expected to meet standards for 1975 (3.0 grams per vehicle mile). But the much more stringent 1976 standards (0.4 gram per vehicle mile), many researchers believe that a catalyst, most likely a precious metal catalyst, will be required. Such two-step catalysis could make for a compact and less complicated control system.
Cleaning up automobile exhausts is one important part of the total challenge of cleaning up the air we breathe. It's a vast challenge, it requires intensive effort, there are costs. But there are returns to be expected – perhaps more than most of us have fully realized.
Through ill health and related losses alone, according to Dr. Paul Kotin, director of the National Institute of Environmental Health Sciences, man's misuse of the environment is costing Americans $35 billion a year. Of the total national bill of $70 billion a year for health services, $7 billion goes for treating illnesses directly resulting from the environment. Americans lose $25 billion a year more through misused wages and services attributable to environmentally caused illnesses. The rest of the $35 billion consists of costs of compensation and rehabilitation.
Also, notes Dr. Kotin, the remaining life expectancy of a 40-year-old man has increased only moderately since 1900 and very little at all from 1920 to now. The apparent reason is the emergence of such chronic degenerative diseases as heart, lung, and some forms of cancer as major health hazards, but environmental factors, Dr. Kotin suggests, contributes significantly to these illnesses.
Recently, too, Dr. John J. Hanlon, Assistant Surgeon General, told an American Medical Association meeting: "It is frustrating to hear sometimes from fellow professionals that the predominant illnesses of the day – malignancies, cardiovascular diseases, chronic respiratory diseases, and the like – are more prevalent because people live longer and have time to acquire these conditions. This is a gross oversimplification. Chronic and degenerative physical and mental ailments have increased, probably in large part because of environmental changes which have resulted from extensive and ever-increasing industrialization and urbanization."
Dr. Hanlon went on to add:
"Members of the medical profession must become more cognizant of the fact that a very significant proportion of the so-called chronic, noncommunicable diseases, as well as mental illness, have their genesis in the environment, and that not only can more attention to the environment prevent many of these illnesses, but also that the expenditure of money and professional effort merely to treat these clinical conditions without concern for their environmental source is fruitless.
A BRIEF REPORT ON THE STATUS OF RESEARCH ON THE HEALTH EFFECTS OF AIR POLLUTION
INTRODUCTION
Scope
This report has been prepared in response to a request for a review of current literature (post- 1970) to determine whether there have been any significant advancements in knowledge of the effects of air pollution on human health. Of more specific interest, there is a need to determine whether any studies have been published on significant new data which correlate measured levels of air pollution with measurable public health effects. Any additional reports which attempt to evaluate the health cost/benefit effects of reductions of air pollution levels are also of interest. In keeping with the "progress report" nature of this report, only very recent data have been examined. In addition to the limited reports immediately available, several comprehensive reviews have been examined for other data in order to compensate for the lack of time available for the completion of an examination of all original research.
Background
The determination of the health effects of air pollution is an essential but increasingly controversial part of the total process of developing air quality standards. In considering this problem, two categories of health effects are of concern. In the first instance, some pollutants seem to produce no discernable permanent physical deterioration of health but contribute to public reaction because of the immediate unpleasant effects. These are the effects such as eye and nose irritation, odors, and obscuration of vision (smoke).
The more serious health effects are exemplified by air pollution disasters with sudden onset of illness and fatalities or the more insidious and chronic effect of continuous exposure to lower levels of air pollutants which are believed to contribute to the rising incidence of such diseases as lung cancer, emphysema, bronchitis and asthma. These diseases generally develop over a long period of time and it is very difficult to establish on a scientific basis the direct cause and effect relationship between exposure to the air pollutants and the specific diseases in question (a similar problem has existed for years with regard to the health effects of smoking). While a direct cause and effect relationship is difficult to determine, the collection of epidemiological data, not only within the United States but within the other Nations of the world as well, is leading to the compilation of a rather impressive foundation of knowledge. This knowledge suggests some definite associations between certain chronic disease and continuous exposure to a polluted atmosphere.
The data on health effects are collected in several ways. Direct evidence is difficult to collect since experimental and deliberate long-term exposure of human beings to potentially toxic substances is not a permissible procedure. As with other problems of this nature, the alternative of long term experimentation with animals is producing much data on the specific effects of air pollutants. For example, it has been demonstrated in animal experiments that ozone (2.5-4.5 ppm) is harmful to the mammalian lung (1). Unfortunately, it is not possible to make a direct extrapolation from animal experiments to effects on human beings.
The collection of data from occupational exposure to pollutants has been of great value. Clinical studies of human beings have produced data which suggest how the course of a disease proceeds, for example, black lung disease in miners exposed to coal dust has enabled the researchers to gain knowledge about the effects of certain kinds of particulate pollution.
However, the most significant data on health effects of air pollutants has been developed through epidemiological studies. In these types of studies an attempt is made to identify correlations between differences in mortality (and/or morbidity) and differences in air pollution levels for different areas. The areas compared may be different urban areas as well as different urban and rural communities. The studies are very dependent upon accurate health statistics, that is, reliable identification and classification of diseases as well as adequate information concerning patient case history with regard to smoking habits, occupation, place of residence, period of residence in each area, and other data necessary to understand contributing factors for a disease condition which may take years to develop. In addition, there must be available, for the population being studied, reasonably reliable information on the air pollutant levels present at the same (or preceding) time that the mortality and morbidity statistics were being collated. In many instances, air pollution data are very poor. Many studies depend upon measurements taken from a single station within a community. The measurements may have been taken infrequently or may have missed a considerable period of time. Additionally. there may be doubt as to whether a representative location had been used for establishing the station. Further, there may not have been adequate data collected on all pollutants of interest. For example, frequently only smoke index and particulate or sulfur dioxide measurements were taken when in fact a number of other pollutants, such as carbon monoxide or nitrogen oxides, were also of interest. There are inadequate physiological data available on the exact mechanism by which most air pollutants produce the adverse health effects being studied. (This fact adds to the difficulty of establishing accurate dose/response levels essential for the determination of air quality standards).
There are a number of other reasons why data on the health effect of air pollution are scarce or subject to so much criticism. Among the problems in these areas are the following factors:
1. Pollutants are not isolated in effect. The pollutants may be gases or small particles. Mixtures of particles and gases actually occur in an environment and it is extremely difficult to pinpoint the actual individual pollutant effects. Since the levels of pollution are variable from area to area, differences in health effects may actually be due as much to the differences in concentrations of mixtures as due to the total level of any particular substance in the air at a particular time. Synergistic effects, where a combination of two or more pollutants is more injurious than the sum of individual effects, are often observed.
2. Weekly and seasonal activities give rise to mortality and morbidity data which are totally unrelated to air pollution levels. For example, snow storms may increase instances of heart attack or respiratory distress; hospitals may have days for elective surgery which may increase mortality in a cyclic manner; seasonal activities of other types may affect mortality rates at the same time that fluctuations in air pollution are occurring which seem to be related to the changes in mortality rates.
3. There are direct links between meteorological events (weather) and air pollution and between economic activities such as weekend vacations – which in themselves increase morbidity and mortality data.
4. Frequently, air pollution simply seems to aggravate an acute preexisting condition; it has been suggested that increases in mortality which seem to be related to air pollution, may actually be due to a moving forward to an earlier date of deaths which would have occurred without the air pollution event. It is difficult to separate the deaths from acute episodes from other deaths attributable to chronic exposure to air pollutants.
5. Another argument, particularly with regard to the significant difference between urban and rural death rates is that the urban dweller smokes more, is more overweight, less exercised, more tense, and generally less healthy. Since it is true that various air pollutants do occur in higher concentration in cities, it has been suggested that it may just be an incidental fact that air pollution is linked statistically to mortality and increased morbidity in many epidemiological studies. The real health hazard may be the total effect of urban living habits rather than the differences in air pollution alone.
6. Other factors frequently related to studies of the effects of air pollution are social and economic status, smoking, water borne factors, genetic differences, radiation, employment factors, and all the other variables produced by differences in life style.
In spite of all these arguments, there is a trend which is detectable in all of the research reports and this trend is one which indicates that acute air pollution episodes are positively associated with significant increase in mortality. There also seems to be a strong correlation between air pollution generally and chronic disease of certain types. While it is very desirable from the standpoint of regulation to be able to define the specific point or level or contamination at which health effects will occur, it is not possible at this time to define precisely that point. There have been some efforts to indicate the levels at which certain effects will occur. When the effects are definite and positive, as for the attack rate of asthmatics with increases of pollution or for increases of a specific type of cancer definitely associated with a particular chemical pollutant (for example lung cancer and benzopyrenes) then some degree of reliability between indicated levels of pollutants and increases in morbidity or mortality can be developed. When the diseases are chronic, as with bronchitis or emphysema, or cardiovascular disease, then the health effects/
air pollution levels data are not so unquestionable and a determination of the portion of the total morbidity or mortality attributable to the levels of air pollution is difficult. This is particularly true when it is kept in mind that many of these diseases can be induced by a number of factors.
For example, the relationship of bronchitis or lung cancer to smoking, industrial gases, genetic factors, cold, or allergenic factors is suspected but not easily quantified. When an attempt its made to explain psychological, behavioral, visual, or psychomotor effects, in terms of increased carbon monoxide, for example, it is even more difficult to attribute air pollution levels to observed effects.
Since so many environmental and genetic factors may be at work in the development of chronic disease, and since the observed data are frequently not near the desired level of accuracy, very complex statistical methods of analysis have evolved for the purpose of attempting to measure not only the presence of a correlation between air pollution levels and health but also the precise contribution of specific levels of air pollutants to the health effects being noted. There are many different methods of statistical analysis. Therefore, it should not be surprising to learn that there are different estimates of the degree of risk associated with specific levels of pollutants. In some instances, the estimates may indicate that no risk to health may be detectable at certain levels of pollution; by other methods, a significant degree of risk may be involved.
Air Quality Standards
As a result of analyses of available data on estimates of health effects versus pollution levels the Environmental Protection Agency published National Primary Ambient Air Quality Standards. These Standards are associated with Air Quality Criteria and can be compared with data provided by the World Health Organization and Threshold Limit Values developed by industrial hygienists for occupational situations. It may be of some interest at this time to briefly summarize these different values:
All of these standards are subject to certain other limiting criteria not presented in the table. These differences depend somewhat upon time, period of exposure, simultaneous exposure to multiple insult, and other factors such as number of times it is permissible to exceed standards per year. For the purposes of this report, however, the preceding data provide some order of magnitude for the dangers being estimated so that the more recent literature reports may be examined with some perspective.
BRIEF REVIEW OF CURRENT LITERATURE
Carbon monoxide (CO)
In order to interpret the significance of the reports on health effects of carbon monoxide it must be understood that varying the concentration of carbon monoxide within the environment results in a variation in the equilibrium concentration of carboxyhemoglobin (COHb) in the blood. (Carboxyhemoglobin is the compound formed by the union of CO with blood hemoglobin. This reaction reduces the availability of hemoglobin for transport of oxygen in the respiratory exchange). The higher the concentration of COHb the greater will be the effect on the human being in terms of adverse health effects. The following table is provided as an indication of the approximate effect of varying the CO concentration:
As lower levels of CO are reached, the data on time of exposure required to read equilibrium concentrations become more controversial. For example, the data presented within the WHO report for 1972 (2), indicate a difference of opinion for CO concentrations of 100 ppm as compared with the preceding table (which was developed by industrial hygienists). The WHO data provided in the following table are of greater interest with regard to air quality standards which deal with very low concentrations of CO.
It is also of interest to be aware that smokers frequently have COHb levels in the general range of 4% and may actually have COHb levels higher than this level. As may be noted from the literature, blood levels slightly above 4% COHb are suspected as having some risk for patients with cardiovascular disease. As may also be noted, the level established as a National standard (9 ppm) would produce an equilibrium concentration near 1.7%o COHb. Individuals will reach an equilibrium blood COHb level depending upon the environmental conditions. For example, smokers in an environment with less than 10 ppm CO, and without further smoking, will reach an equilibrium level below that which is produced by smoking. A reduction in equilibrium concentration will take longer in smokers than in non-smokers. The higher the concentration of CO the less time will be required to reach a particular COHh level; for example, at 25 ppm about 24 hours are required to reach 4% COHb; this same concentration could be reached in 11 hours at 100 ppm.
The data on carbon monoxide effects as summarized in the National Academy of Sciences study (3), are not conclusive with regard to health effects. The evidence reviewed within the study indicates that there is no doubt that exposure to environmentally encountered levels of carbon monoxide will result in an increase in blood carboxyhemoglobin. Exposure to 10-12 ppm CO for 4-5 hours can add to the normal body burden (depending upon the history of previous exposure).
However, their conclusions indicate that there is insufficient evidence to demonstrate that air pollution levels encountered in typical community environments will produce chronic health effects. (For another example, in a recent House report on hearings on the Clear Air Act (4), an industrial representative from Chrysler Corporation reported results of one of their studies in non-smokers for 14 cities which indicated that the average COHb for over 21,000 observations was well below 2% COHb.)
However, the Academy report does indicate that there is some evidence associating daily levels of more than 10 ppm with an increase in mortality in hospitalized patients with myocardial infarction. The report also includes a conclusion that there is some evidence that CO exposure "may have some effect on the biochemical processes underlying atherosclerosis, but this evidence at this point is also inconclusive." There is a suggestion in the Academy report, based on reports of limited studies of human drivers and more evidence with laboratory experiments, that as little as 2 % COHb may affect visual sensitivity, psychological testing and time interval estimation.
Low levels of carbon monoxide have been associated with mortality in Los Angeles. In this study (5), evidence is presented that carbon monoxide concentrations varying between approximately 10 and 18 ppm produce an increase in mortality. The data indicate that a CO mean concentration of 20.0 ppm as compared with a mean concentration of 7.3 ppm will contribute 11 deaths for each day. In other words, an increase in mortality per day has been correlated with an increased level of CO. An evaluation of oxidants in this same study indicates that these pollutants are not producing a similar effect. The mortality and CO data were evaluated for the years 1961-1965.
This report suggests a desirability for maintaining CO concentrations below 10 ppm. Other reports, as summarized within the 1972 report of the World Health Organization (2), indicate that headache and impaired coordination were noted when COHh concentrations exceed 10 % saturation (1970). At a saturation of 7.2%, significant decrease occurred in visual perception, manual dexterity, and ability to learn intellectual tasks (1971). A 1970 study cited in the WHO report indicated no effect on sensitive visual and auditory thresholds after 500 ppm CO for 1 hour (8% saturation). Headache, impaired manual coordination and modification of visual evoked responses were noted when COHb measured in excess of 20% saturation (exposure to more than 100 ppm CO would be required in these test conditions). Coronary arteriovenous oxygen differences increase with COHb above 5-10% saturation (1970); myocardial changes were observed in patients with above 6 % saturation (1976) ; and saturation levels above 4% were reported to be a risk for patients with cardiovascular disease. (Note that many smokers are exposed to levels of this magnitude.)
In another recent study (6), concern was expressed about the potential danger from carbon monoxide concentrations encountered in traffic. It is pointed out in this study that, if it is accepted that a 2% increase in blood carboxyhemoglobin is sufficient to impair psychomotor standards, then the concentrations encountered in traffic may be sufficiently high to impair performance. Groups who regularly drive in the city (bus and truck drivers and police) as well as individuals working in the city (police, inspectors, parking lot attendants, etc.) may be at some risk. These investigators found that final blood COHb concentrations of 3.2% were correlated with atmospheric levels of 20 ppm; COHb levels of 9.7 % were produced at 60 ppm concentrations – regularly encountered in their sample survey during traffic hours in Toronto.
Their data indicate that a 40 minute drive with CO concentrations at 20 ppm would increase blood COHb by 0.4%; on congested days with CO at 60 ppm a one hour drive would increase COHb by 1.8%. Walking for ten minutes would increase blood COHb from 0.1% to 0.8% depending on CO levels (20-160 ppm). Their conclusion indicates that the city dweller is always close to the point of impaired psychomotor performance and that short term exposure to moderate levels of CO might easily lead to impaired performance. Ury and his coworkers (7)
however, were not able to demonstrate a definite correlation between increases in motor vehicle accidents (one evidence of impaired psychomotor performance) and increases in CO concentration. They were able to show a correlation between increases in oxidant levels and increases in motor vehicle accidents.
It has been reported that the Biomedical Research Branch, EPA (Chapel Hill) observed that exposure to CO at 50-100 ppm induced angina after exercise earlier than when "pure" air preceded exercise. At Stanford University, it was observed that CO at 50 ppm for one hour produced about 17 percent impairment of visual acuity in poor light.
In a recent report (17) of results of experiments with monkeys, Eckardt and others describe a two year exposure (22 hours per day, 7 days per week) for two different levels of CO (19.86 ppm and 65.46 ppm). Their conclusions indicate that these levels of CO did not lead to any biologically significant changes in the test monkeys. If these results can be extrapolated to human beings, it would suggest that the current standards of 9 ppm are too stringent, that is, current levels of CO are probably not producing any significant damage. Perhaps it should be kept in mind, however, that no psychological tests were conducted and these were physiologically normal monkeys. This is significant for at the lower levels of CO, psychological impairment and stress on defective systems (cardiovascular deficiencies, for example) have been emphasized in other reports.
Theodore (18) referred to the two other studies which indicated that COHb levels as low as 2-5% caused changes in performance ability. (These are the studies most frequently referred to as indicating a low level CO effect; one by Schulte, a 1963 study, indicated impairment of cognitive and psychomotor performance at 5% COHb, with some slight indication of associated changes at
2% COHb; the other study by Beard and Wertheim, a 1967 study, showed an impairment of auditory discrimination at 4-5.7 COHb.) In their studies, Theodore and his coworkers studied low level effects in human beings and attempted to measure the same types of effects and reached the conclusion that "no performance decrements were found in humans during three-hour exposures of 50-250 ppm CO".
In one of the few well controlled human experiments, Stewart and others (35) exposed human volunteers to carbon monoxide at concentrations of less than 1, 25, 50, 100, 500 and 1,000 ppm for periods of one-half to 21 hours. They observed "no untoward subjective symptoms or objective signs of illness ... in the 24-hour period following the exposures to 25, 50 and 100 ppm CO". Hosko, in reporting his work with these volunteers (38), indicated that changes in visual evoked response were noted when COHb levels exceeded 20%. These authors consider the fact that an 8-hour exposure to 100 ppm CO (11-13% COHb levels) produced no impairment in performance of tests to be their most important finding. The tests included evaluations of visual and auditory acuity, coordination, reaction time, manual dexterity, and time estimation. The volunteers were normal. The investigators carefully indicate that the effects of similar levels of CO on persons with cardiovascular disease, alcohol blood levels from drinking, or individuals using central nervous system depressant drugs remains to be determined.
In a recent British report (37), evidence is reviewed which suggests very strongly that carboxyhemoglobin levels up to 20% exert measurable physiological effects. Reference also is made to the various reports which indicate that 5-10% COHb levels affect the circulatory system. This same report states "It should be stressed, however, that non-significantly raised carboxyhaemoglobin levels differ from significantly raised carboxhaemoglobin levels from car exhaust in the streets, which has been shown in many studies, and clearly demonstrated recently by the findings of carboxyhaemoglobin levels of only 1.4-3.0% in non-smoking taxi drivers in London".
None of these reports seem to provide any significant new information as compared with the data available at the time that the air quality criteria were being developed. Several reports indicate no measurable effects at 5-10% COHb, others continue to cite and support evidence that levels near this concentration will have effects on the circulatory system or produce interference with psychomotor processes. The problem is one of determining the real long term significance of the effects which are being reported for these low levels of COHb (in the range of 25-50 ppm CO). Groups at risk from CO would include smokers, cardiovascular patients with impaired blood transport systems (anemics, etc.) and individuals on drugs which affect behavior response time.
Sulfur Oxides (SO2, etc.)
Sulfur oxides have been of particular concern in air pollution studies since this is the agent most suspect in acute pollution episodes, particularly where fossil fuels such as coal are being consumed. Glasser and Greenberg (8), in a study relating air pollution to mortality in New York City examined daily SO2 and smoke levels for the years 1960-1964. They found that at daily mean SO2 levels of less than 0.10 ppm the mean number of daily deaths decreased; at levels between 0.20 ppm and 0.30 ppm the mean number of daily deaths rose to 1.8; as levels rose to 0.30 ppm to 0.40 ppm deaths increased to about 9.4 deaths per day and at levels of 0.60 ppm and above daily deaths rose to 12.6 deaths per clay. Their analysis of these data supports a very strong correlation of SO2 levels with increase in daily mortality.
In general, the difference in mean number of daily deaths on days with 0.20 ppm SO2 levels and days with mean levels of 0.40 ppm or more was in the range of 10-20 deaths per day. A later report by Schimmel and Greenberg (9) related daily pollution to daily mortality rates in New York City for the period 1963-1968. Their purpose in this retrospective study was to try to determine whether well-documented excesses of mortality over a long period of time could be correlated with daily air pollution levels. Although the study admittedly deals with short term levels of pollution, it is interesting in that the pollution levels considered are less than the levels ordinarily detected in crisis events. These investigators were examining SO2 concentrations in the general range of 0.17 ppm mean daily levels (still well above standards). The measurements were taken at East 121st Street. Manhattan, an area of high pollution levels.
These authors indicate that approximately 10,000 deaths per year occurred which would not have occurred when they did if there had not been air pollution on the day of death or on immediately preceding days. Their data indicate that about 40% of the respiratory disease deaths were associated with short term pollution effects. Coronary excess deaths were actually twice as great as respiratory deaths associated with pollution although in percentage term excess coronary deaths were only 12% of total deaths examined. Precise data on daily levels of SO2 and smoke shade pollution levels are not presented in the study although as noted a mean level of SO2 of 0.17 ppm was cited.
A very limited study of the effect of air pollution on asthmatics was completed in New Cumberland, West Virginia (10). This study showed a very high correlation between increases in air pollution and increased incidence of asthmatic attacks. This correlation was still present after temperature effects were removed (by statistical analysis techniques). It was not possible to determine which of the several pollutants were associated with the attacks. Sulfur dioxide, suspended sulfates, particulate matter, suspended nitrates, and soiling index were all factors evaluated (as well as temperature effects). These pollution effects occurred at concentrations similar to levels measured in many cities. The attack rates noted in this study were: 40-50% attack rates at SO2 concentrations near 0.12 ppm with suspended particulate matter at 500 µg/m² a soiling index of 1.3 and suspended nitrates at 5.5 µg/m²". All concentrations varied somewhat with temperature changes.
In their review of research data on S02, Ayres and Buehler state "although statistical changes in pulmonary function are found following inhalation of sulfur dioxide, it is difficult to imagine how community air pollution with average levels of 0.1 to 0.4 ppm of SO2, and occasional peaks to 1.0 to 1.5 ppm can produce significant health effects suggested by epidemiological studies" (15). The inability to reproduce clinical effects at levels encountered in the urban atmosphere has focused attention on other components – particularly particulate material found in polluted environments. In other sections of their comprehensive review of the literature on epidemiological studies, however, Ayres and Buehler provide conclusions which strengthen the general observations that air pollution and adverse health effects are related.
Winthrop (16) points out that adverse health effects associated with SO2 have been observed in individuals with disease when SO2 levels exceed 0.11 ppm for a 24 hour average for periods of 3 to 4 days. These effects also are noted when annual mean levels exceed 0.04 ppm. (Note that in this case, Winthrop's 0.11 ppm 24 hour average is below the National standard of 0.14 ppm for a 24 hour average.)
The recent WHO report indicates that an excess of mortality occurs when the levels of suspended particulates and sulfur oxides have both exceeded 500 µg/m² for 24 hours. This effect has been noted particularly for individuals with cardiac or pulmonary disease. The report refers to unpublished preliminary findings that persons with respiratory disease may exhibit an increase in symptoms with levels of SO2 of about 250 µg/m². Higher frequencies of respiratory disorders may be found in children in areas where annual average levels for smoke and SO2 are both above 100 µg/m². A British report it can be detected with 24-hour means of 250 µg/m² smoke together with 500 µg/m² SO2 (36).
Particulates, polycyclic organic matter, hydrocarbons
The National Academy of Sciences recently published an in-depth review of the state of our knowledge with regard to the health effects of polycyclic organic matter. While the present brief report cannot adequately summarize the comprehensive Academy study, there are certain data within the Academy report which are germane. Selected data from the Academy report are presented in the following paragraphs.
In the report on particulate polycyclic organic matter (POM) (11), such as benzopyrene, it is pointed out that there is convincing evidence that the incidence of lung cancer is about twice as high in urban areas as in rural areas. This difference is not entirely explained by the known epidemiological association of cigarette smoking with lung cancer. The tentative conclusion has been drawn that this higher urban incidence of cancer is due to the presence of particulate polycyclic organic matter where air pollutants are higher – that is, in urban areas and particularly where fossil-fuel products from industrial usage are concentrated in the air. A "working hypothesis" is presented which compares epidemiological data on mortality with the benzopyrene content of the urban atmosphere. This hypothesis suggests that there is a casual relation between air pollution and the lung cancer death rate. It is further suggested in the report that if the hypothesis is valid, there will be a 5% increase in death rate with each increment of 1 microgram of benzopyrene per 1,000 cubic meters of air. This hypothesis is extended to estimate that a reduction in air pollution from 6 µg/ 1000 m² air to 2 µg 1000 m² air would reduce the lung cancer death rate by 20%.
The report provides an extensive review of clinical and epidemiological studies which support this summary statement. The authors emphasize that the association between urbanization and lung cancer is not in question; the subject of controversy is the identification of the causal agents.
Their review of the literature supports the conclusion that there are twice as many deaths due to lung cancer in urban populations as with non-urban populations. They cite other reports which do not support suggested associations of lung cancer with air pollution. Of particular concern in their evaluation is the possibility that air pollution is actually correlated with some entirely different and unknown factor which is the actual cause of the difference in lung cancer death rates. (If air pollution is correlated closely with some unknown factor which is actually causing the cancer, then air pollution would also show up in the analysis as correlated with cancer; this is one of the dangers with this type of statistical analysis of epidemiological data).
In another study, data on the incidence of lung cancer during the period 1955-1966 were evaluated to determine whether any change in incidence of disease could be noted (12). It was believed that the reduction in the use of fuels such as coal, which produce large quantities of particulate matter, might be followed by a decrease in lung cancer incidence if indeed there is a correlation between these two events. The data from this study indicate that lung cancer rates are declining in the young, stabilizing in the middle aged, and increasing only in the older age groups. These authors suggest that there is a strong evidence to support a clear relationship of reductions in air pollution with beneficial health effects and "These observations which point strongly to some close relationship between lung cancer and soot-borne benzopyrene, as well as the soot per se, provide added impetus to the necessity for speedy implementation of an effective clear air program". No data on measurements of air quality are included in the study.
A report from the Charleston area suggests that the results of reduction in serious air pollution will be immediately apparent. In this report (13), a significant decrease in deaths due to cardiovascular disease was noted during the period 1698-1970. During this time, the imposition of pollution controls reduced particulate matter from an annual average of 227.6 µg/m² in 1968 to 120.5 µg/m² in 1970. Mortality from cardiovascular disease was dramatically reduced from an annual rate of 253.6 per 100,000 population in 1968 to 188.2 per 100,000 in 1970. The Charleston study involved a comparison of two communities, one with a high level of particulate pollution, the other with a low level (74.4 µg/m² annual average in 1968 and 55.3 µg/m² in 1970). The data indicate a significantly higher mortality in the population in the area of higher levels of pollution. Reduction of pollution resulted in a dramatic reduction in differences in death rate; 90 percent of the decrease in deaths was largely due to decreases in deaths due to cardiovascular disease. Although the investigators are exploring several alternate possible explanations for the decrease in deaths. it is believed at this time that a reduction in the number of inversions during 1970 combined with a significant reduction in suspended particulate matter as a result of the imposition of pollution controls is probably the most acceptable explanation.
Nitrogen Dioxide (NOx)
There is very little data available on the health effects of the oxides of nitrogen. As noted elsewhere in this report, the data are generally considered to be insufficient for developing any estimates on health effects due specifically to this agent. There are a number of reports in the literature, however, which have been published in the post-1970 period. Most of these reports are on the results of animal experiments and extrapolation to man is difficult. Stephens and his coworkers (28), for example, found that rat lungs suffered considerable damage when exposed to NOx at 17 ppm for 72 hours. Changes in lung tissue continued to occur at 2 ppm levels. In all of the references to studies of NOx effects on human beings, the Chattanooga studies by Shy and his associates (29) are cited most frequently. In these studies, the effects on school children of exposure to NOx released from a nearby TNT plant are compared with other areas with lower concentrations. These studies, which led to a conclusion that NOx adversely affects the ventilatory performance of the children are the subject of some continuing controversy in interpreting the results. The average NOx level in the high area of 0.10 ppm was exceeded 40% of the time in measurements taken at one of the stations; this level was only exceeded 17% of the time in the area of lower concentration. In addition, the investigators found what they believed to be a greater incidence of respiratory infections in the population exposed to higher concentrations of NOx. The EPA standards for oxides of nitrogen are based primarily upon the results of the Chattanooga studies. As pointed out in an EPA study on the oxides of nitrogen (30), there is some concern that the Chattanooga studies are not sufficiently accurate to firmly establish a basis for the standards. The nitrogen studies in Chattanooga were based on emissions from a single source and there are wide variations in the data base utilized for deriving the conclusions regarding effects.
Goldstein (31) reviewed the available literatures on studies of NOx, effects on human beings, making reference to the Chattanooga studies as well as other studies, and also indicated the inconclusive nature of these studies. According to his review acute exposure to NOx does affect respiratory function. Eye and nasal irritation occurs at 15 ppm; pulmonary discomfort occurs at 25 ppm; and bronchiolitis with focal pneumonitis occurs at 25-75 ppm levels (less than one hour duration of exposures). Increased resistance to respiration has been noted after 10 minutes at 4-5 ppm NOx. One minute at 50 ppm produces significant nasal irritation and pulmonary disruption.
The National Academy of Sciences considered that individuals with asthma, chronic bronchitis, emphysema, and other forms of obstructive disease will be particularly susceptible to high levels of NOx. Their report (32) indicates that 10 ppm or more NOx will produce additive effects with ozone. Guidelines for short term exposure of the public suggest limits of 1 ppm for periods of 10-60 minutes; 0.5 ppm for 5 hours/day, 3-4 days per month; and 1 ppm for 1 hour/day/year.
According to their review, persons with chronic bronchitis did not exhibit significant reactions after exposures to 1.5 ppm for 15 minutes. A number of other reports are reviewed within the Academy study.
The WHO report cited previously (2) indicates that "The published data on the effects of nitrogen oxides on human health are limited. While biological activity in animals and plants at low concentrations has been demonstrated, the Committee believes that there is insufficient information upon which to base specific air quality guides at this time.
Cost/Benefit Analysis of Health Effects
The research literature on the subject of cost/benefit analysis of the effects on health of air pollution is quite limited. Among those reports which are cited within the various studies of health effects of air pollutants, those by Lave and Seskin are the most striking in terms of comprehensiveness and currency. Their reports will certainly be a main focus of any arguments about cost/benefit effects with regard to health effects and are summarized very briefly in this report as being of primary interest. Lave has as one of his major goals in this research the estimation of the quantitative effect of air pollution on health.
As pointed out by Lave in one of his most recent papers "it is extremely difficult to prove that 'air pollution causes ill health,' much less to quantify the nature of the relation." Nevertheless, he has examined in great detail the mortality rates for 117 Standard Metropolitan Statistical Areas. In his studies over the past several years, he has responded to criticism of his conclusions by reexamining the data for other possible causal factors for the health effects predicted. He has considered the possible influence of such factors as race, income, population density, age, types of home heating equipment, fuel differences, use of air conditioners, meteorological factors, and occupational mix (21) (22) (23). He still reaches the conclusion that "air pollution has a marked effect on the mortality rate." He has then proceeded on the basis of these studies to project an estimate of the dollar value of this health effect. From these estimates, he has arrived at these conclusions (23)
1. A 25-50% reduction in morbidity and mortality would occur if air pollution in major urban areas were abated by about 50%. Based on an estimated cost for bronchitis of $950 million per year, a savings of $250-$500 million per year would be realized.
2. About 25% mortality from lung cancer could be saved by a 50% reduction in air pollution. This reduction would produce a savings of about $33 million per year.
3. A 25% reduction of morbidity and mortality from all respiratory disease would be achieved by a 50% reduction in air pollution. Based on an annual cost of $4887 million for respiratory disease, a savings of $1222 million would be realized.
4. About 10% reduction in cardiovascular morbidity and mortality could be achieved by a 50% reduction in air pollution. A savings of $468 million would be realized.
5. All mortality from cancer could be reduced by about 15% with a 50% reduction in air pollution for a savings of $390 million.
6. A total annual savings (health effects) of $2080 million would be secured by a 50% reduction in air pollution.
Lave and Seskin's estimates were based upon air pollution levels and mortality data available in 1960-1961. In several of their papers, savings were indicated as being based upon very conservative estimates. In a recent response to a Wall Street Journal editorial criticizing EPA data on health effects savings (as indicated in (24) ), Mr. Ruckelshaus wrote that in his opinion (25) : "the $6.1 billion estimated for 1968 health costs of air pollution which was singled out in your editorial (of December 4, 1972) is a conservative statement of all health damages of air pollution in 1968, even if linear extrapolation is not assumed. For example, the Lave-Seskin approach estimated a present value of future earnings lost by early death (of) wage earners, but did not estimate the true value of lost services of homemakers, persons over 65, and young children. The medical costs of increased morbidity were not included in the study and, in addition, Barrett-Waddell (26) assumed that health costs remained a constant percentage of GNP although health costs are rising faster than the average price level."
It should also be noted that in neither the Lave-Seskin estimates nor the Barrett-Waddell studies (as summarized in (24) ) were the effect of photoxidants, oxides of nitrogen or carbon monoxide included. As noted in the Barrett-Waddell study, the data on these compounds were considered insufficient to prepare any analyses.
As noted in earlier discussions, in the absence of controlled experimental data, complex statistical techniques have been utilized to ascertain whether air pollution factors are producing significant effects on health. The selection of the method of analysis may influence the conclusions to a considerable degree. For example, if meteorological factors are discarded in favor of SO2 concentrations as a major variable influencing mortality by one technique, the opposite conclusion – or a less conclusive measurement – might be determined if a different technique of analysis is utilized. Criticism of Lave-Seskin method of statistical analysis has appeared (27) which does bring into question the significance of some of the conclusions developed by Lave-Seskin and others by their methods of analysis. An evaluation of the relative merits of these different methods of statistical analysis is not possible within the context of this report but this problem is certainly an important part of any consideration of cost/benefit analysis of the health effects of air pollution, not only the effects upon human beings but in any other analyses as well.
SUMMARY
The status of our knowledge of health effects as supported current research does not seem to have progressed significantly beyond the position challenged in 1971 by Heuss, Nebel and Colucci (33). In this paper, presented at the 64th Annual Meeting of the Air Pollution Control Association, the following criticism were presented:
The carbon monoxide standard is based on a blood carboxyhemoglobin level below that associated with any physical or mental impairment. (a standard of 15 ppm for 120 hour exposure; 50 ppm for a 1 hour exposure is suggested as an alternative).
The nitrogen dioxide standard is based upon a questionable epidemiological study that needs further verification (the California standard of 0.25 ppm for 1 hour is suggested as an alternative).
The hydrocarbon standard is orders of magnitude below the levels associated with any health effects and is unnecessary (the suggestion is made that no hydrocarbon standard is necessary since an oxidant standard has been set).
The photochemical oxidant standard is based on a questionable extrapolation of the results of a single study (a standard of 0.10 ppm average over 1 hour oxidant level is suggested as an alternative).
In replying to this criticism the EPA representatives indicated that (33):
"National standards in such an important matter as air pollution must be established at levels that contain a sufficient safety margin to provide reasonable certainty that all sensitive population groups in the United States will be protected. Economic factors must be secondary in this primary consideration".
The estimates of Lave and Seskin seem conservative. If the statistical analysis used by them is valid in attributing the degree of health effect to pollution as they described, then the quantitative data on the benefits (in dollars saved) to be derived from reductions in air pollution are probably quite conservative since so many cost factors in health care were not included in the analysis. It is necessary to observe however, that a determination as to whether these savings are actually being achieved will be dependent upon whether the levels of reduction in pollution are actually occurring. The calculations by Lave were based upon air pollution levels observed in 1960-1961. He estimated the savings based upon a 50% reduction in these levels. He also used cost figures which were based upon earlier studies without updating to account for inflation and other effects on health costs.
As may be noted from the literature review within this report, no really significant or startling breakthroughs in knowledge were identified. There is still considerable uncertainty regarding the validity of health effects noted at the lower levels of carbon monoxide and the data are still considered insufficient with regard to nitrogen dioxide health effects at lower or ambient levels. There does seem to be an increasing agreement regarding the dangers of particulate and sulfur dioxide effects. Most of the investigators still express considerable uncertainty regarding the data on pollution levels – with regard to the position and number of samplers as well as the accuracy and area representation of the samples. Unfortunately, mortality statistics still lag significantly so that current evaluations are slowed. Additionally, morbidity data still suffer from a lack of standardization of reporting for many of the chronic diseases. The number of individuals "at risk" to the effects of air pollution is quite large – although not adequately described quantitatively – and are collectively a problem of continuing concern with regard to air pollutant health effects.
SUMMARY OF RESEARCH AND PROBLEM AREAS
Additional information is required with regard to the health effects of photo-oxidants, nitrogen dioxides, and carbon monoxide (particularly at CO levels below 25 ppm). Although few additional reports are available on the effects of particulates and SO2, there seems to be less disagreement regarding the effect of these materials even though the specific relationships and physiological mechanisms of the effects caused by these pollutants are not fully understood.
There is a compelling need for more epidemiological studies and long term studies of chronic effects of air pollutants. In this regard, it has been suggested that one obstacle to the completion f these types of studies is a natural reluctance of scientific personnel to become involved in long term (life time) studies. It has been suggested that some consideration might be given to the establishment of an institute or program within an existing institute with the objective of completing these types of life time studies. Scientific personnel could be assigned to the task, replaced as necessary, reassigned, or other personnel changes made, but the program itself would be institutional in nature with personnel continuously maintaining the continuity of the studies.
Such long term studies of populations to determine effects of environmental factors would require some assurance of continuous funding and support in order to attract competent investigators willing to invest the time and effort necessary. Investigations of this nature could focus attention upon such problems as perinatal mortality and air pollution; the relationship of cigarette smoking and urbanization (particularly with regard to lung cancer); and the contribution of air pollution to accidents. More accurate statistics on air pollutants by region and on morbidity and mortality data are required.
Studies of the mutagenic effect of airborne chemicals are needed. For example, some evidence is available that SO2, dissolved in water reacts adversely with genetic compounds. Additionally, more information is needed on the effects of airborne carcinogens. The National Academy of Sciences report on polycyclic organic materials indicates the significant risk from these substances. There is also a need to evaluate the chronic toxic hazard of atmospheric pollutants. particularly the various metals of interest. The data on metals have been mentioned as inadequate for establishing firm air standards. Of specific interest are cadmium, nickel, beryllium and lead.
More studies need to he completed on the health benefits (in dollars saved) to be derived from reductions in air pollution. The data from monitoring changes of levels of air pollutants need to be continuously reexamined in order to determine whether these benefits are actually being realized as the costs are incurred to effect the reduction in pollution. Further, the value of the benefits requires regular updating in order that the validity of the estimates may be assured for cost benefit analyses necessary when evaluating the need for further modification of standards.
Costs and benefits need to be related more directly to specific pollutants rather than to the broad category of air pollution. More information needs to be gathered on the impact of air pollution on susceptible members of the population.
Cost-benefit analyses would benefit from the inclusion of considerations of the non-disease health effects of air pollution. New techniques are evolving which permit sensitive measurements to be taken to predict the onset in changes in ventilatory capacity, for example. It would also be profitable to consider changes associated with increased storage of harmful pollutants in the body. This concept, addressed by John Goldsmith (34), would require the consideration of such reactions as: sensory irritation, odor and reactions to odor, central nervous system reactions, including visual threshold, psychomotor test performance, temporal discrimination, altered response under conditions demanding maximal performance, biochemical changes and blood changes, and effects on aging. Such evaluations would serve to strengthen our understanding of chronic effects as well as improving a capability to prevent the onset of disease caused by air pollution.