1.

There are many pressing reasons why outdoor air pollution needs to be controlled. Polluted air damages plants and houses, obscures the sunlight, interferes with visibility and decreases the well-being of our citizens. In our country, a major effort is being made to abolish the deleterious effects of air pollution through a comprehensive program of controls imposed by law. These efforts have improved the quality of the air in many cities. I believe these efforts should continue, in particular because the outdoor air environment is one over which the individual citizen has little control. I appreciate the opportunity to appear before the Subcommittee, in order to submit some recent observations concerning air pollution and lung disease for your consideration.

The work summarized in this statement is being performed at the Yale University Lung Research Center with the support of the National Heart, Lung and Blood Institute, National Institutes of Health. The project concerns the prevalence of chronic respiratory disease among citizens of three communities in the states of Connecticut and South Carolina, in relation to air pollution outdoors and indoors.

Three specific issues are addressed in this statement:

(2)

(1) Sulfur dioxide vs. sulfate concentrations in the air in Connecticut;
Significance of indoor vs. outdoor air pollution in determining an indi-
vidual's total pollutant load;

(3) Chronic respiratory symptoms and lung function among rural and urban resi-
dents of Connecticut, and among residents of a semi-rural town in South
Carolina.

(1) Sulfur Dioxide and Sulfate Concentrations

In Connecticut, sulfate concentrations in urban (Ansonia, CT) and rural (Lebanon, CT) air were higher in the summer of 1973 than in the preceding and following winter months. Sulfur dioxide concentrations, in contrast, were high in winter and low in summer (Table 1). The highest sulfate levels in summer occurred when the wind direction was from Long Island Sound and the Atlantic Ocean. We believe (ref. 1) that sea water spray may be a significant contributor to sulfate levels in Connecticut.

The marked seasonal variations in levels of S02 and of sulfates suggest that seasonal sources of emission may be important for both pollutants. E.g., seasonal domestic fuel burning may account for the higher S02 levels in winter. The very low S02 levels in summer suggest that sources other than domestic fuel burning are minimal.

2.

Our data show S02 levels ranging from virtually zero on some summer days to 80 ug/m3 on some days in winter in rural Lebanon as well as in urban Ansonia, with mean winter levels slightly higher (21.6 μg/m3) in Ansonia than in Lebanon (17.2 μg/m3). In Winnsboro, S.C., SO2 never exceeded 30 ug/m3 (mean values in winter 4.4 5.8 μg/m3 on different sites). We have not found systematic differences in respiratory symptoms nor in lung function between residents of these towns. The details of these studies are described sub (3). Our results are consistent with the data included in a graph by Dr. B. G. Ferris, Jr., which characterizes combinations of S02 less than 100 μg/m3 and suspended particulates less than 80 ug/m3 as having "no effect" (graph reproduced in NAS report, volume 2, p. 425, ref. 2).

In addition, our data do not indicate differences in respiratory disease prevalence or lung function which might be attributed to airborne sulfates in the

range of concentrations (Table 1) encountered in our study.

(2) Indoor vs. Outdoor Air Pollution

Data on respiratory illness in children are frequently quoted in connection

with outdoor pollution levels. We monitored the daily activities of 20 boys (10 blacks, 10 whites, ages 12-17 years) in Ansonia, CT., and found that they were indoors (at home, in school, or elsewhere) from at least 60% to about 80% of an average school day. We monitored the personal environment of these boys with portable air-sampling equipment (ref. 3). We found that respirable particulates were much higher in the personal environment than total particulates outdoors on the same day (Table 2), while S02 and NO2 concentrations were lower. Boys who lived in homes with one or more smokers had significantly more exposure to respirable particulates (avg. 132 μg/m3). than boys in homes where no one smoked (avg. 93 μg/m3). NO2 levels were also slightly higher for boys in homes with smokers. The group of boys included 10 healthy boys and 10 boys with chronic bronchitis or asthma. When comparing the results of the environmental data on healthy and diseased boys we found no evidence for a relation between exposures and disease. We concluded that, at least for respirable particulates, a person's air pollutant load is usually determined primarily by indoor exposures, and that no significant improvement in respirable particulate loads can be expected to result from reduction of outdoor particulate levels alone, even in urban areas.

3.

(3) Respiratory Disease in Rural Residents (refs. 4,5)

Lebanon, CT., is a rural town with 74 inhabitants per square mile, away from the densely populated industrial zones in Connecticut. There are no factories, and only 4 commercial buildings on its 50 square mile surface area. No major highways run through the town. Part of the town is a summer resort area with lakes. Outdoor air monitoring at different sites (throughout 1973) showed that S02 was far below the annual mean AQS of 80 ug/m3, total suspended particulates averaged 39μg/m3, mean NO2 was 56 ug/m3, and peak 1 hr. 03 concentrations exceeded

the primary AQS of 160 ug/m3 in only 1 of 41 observations over the year 1973 (ref. 6). Thus, Lebanon, CT. may be considered a town with acceptable air quality throughout the year, according to the criteria of the primary air quality standards.

We studied men, women and children (age 7 and over; 75% of total population) with rigidly standardized, computer-controlled methods (ref 5). We recorded respiratory symptoms and performed a lung function test designed to detect minor degrees of airway obstruction such as occurs in asthma, chronic bronchitis and emphysema. For example, this test (flow-volume curve) is able to detect lung function loss due to cigarette smoking even in teenagers who have smoked only a few years.

Table 3 shows that lifetime rural residents of Lebanon, who have never smoked, still have appreciable prevalences of symptoms and respiratory illness. Moreover, these prevalences were not significantly different in urban Ansonia, CT., where over 1300 residents were examined with identical methods. Lung function tests showed no differences between nonsmoking lifetime rural men or women on the one hand, and nonsmoking lifetime urban men or women on the other hand (see example for women in Fig. 1). In both towns, smoking men and women had more symptoms and lower lung function than nonsmokers of the same sex, age and race (black or white). In Ansonia, the mean annual particulate level (60 μg/m3) was close to the annual primary AQS of 75 ug/m3 and reached 160 ug/m3 in some samples. S02 and 03 levels were similar to those in Lebanon, but NO2 levels were significantly higher (annual mean, 86 μg/m3; higher levels 160-170 μg/m3). We have concluded that the urban environment of Ansonia, CT., where past pollutant levels were probably higher than those we measured in 1973, is not associated with a demonstrable excess morbidity of chronic airway obstruction (asthma, chronic bronchitis, emphysema).

Air quality in Ansonia as well as Lebanon might be unfavorably individual by long-range pollutant transport, e.g. from the New York City area or beyond (ref. 7).

If this were a significant factor in lung disease, one would expect lower prevalences of disease in areas not subject to transport of pollutants from major metropolitan areas. Such areas have not been clearly identified since long-range pollutant transport is a phenomenon which has only recently been described in detail. It may be of interest that our study, with identical methods and the same key personnel, in Winnsboro, S.C., showed disease prevalence and lung function similar to those in the two Connecticut towns, when sex, race, age and smoking were taken into account. Particulates, S02 and NO2 were all lower in Winnsboro, S.C., than in Ansonia, CT., while ozone levels were more frequently above the 1 hour primary AQS (11% of all samples in Winnsboro vs. 2% in Ansonia, CT.). Since S02 levels were very low in all three towns we have no information on S02 exposures near or above current AQS. However, a comparison of symptoms in nonsmoking women (65 years and over) in Lebanon, CT., and in suburban Genoa, Italy (ref. 8) suggests that decreases of SO2 below about 90-110 μg/m3 are not associated with decreases in symptoms in chronic bronchitis. E.g., cough in winter was just

as common among elderly women in Lebanon, CT. (mean winter SO2 level 21.6 μg/m3) as among similar women in suburban Genoa (mean S02 at the time 86 μg/m3) (ref. 9). Studies on telephone workers on the U.S. East Coast (refs. 8,9) suggest a similar conclusion.

The data presented here do not allow one to conclude that pollutant levels in the ranges we found in Connecticut and South Carolina have no significant effects on human health. Short-lasting effects may not be detected by our methods, nor do we have data on risks of other diseases (e.g., lung cancer) in relation to air pollution. However, our data support the view that any excess morbidity of chronic lung disease, if it occurs at all, due to air pollution in a typical U.S. industrial town (Ansonia, CT.) is exceedingly difficult to document with objective data. While other explanations are possible, the similarities in symptom prevalence and lung function (once sex, race, age and smoking are accounted for) in three U.S. communities in two states with contrasting climates are most simply explained as due to effects of risk factors other than air pollution, which are distributed among the general population. In smokers, the risk of smoking is a predominant factor, in rural as well as urban residents. Among non-smokers, occupational exposures, indoor exposures at home, and exposures to naturally occurring allergens such as pollens, as well as genetic factors (e.g., susceptibility to asthma) are probably all contributing to the prevalence of disease.

5.

In conclusion, we have found no evidence that outdoor air pollution with particulates, NO2 and 03 at levels ranging from well below to slightly higher than current primary air quality standards bears a demonstrable relation to morbidity of chronic lung diseases in three communities. Studies by other investigators (ref. 8,9,10) suggest that the same may be true for S02. A recent study of Comstock et al (ref. 11) has reached conclusions similar to ours (1.e., no demonstrable differences attributable to outdoor air pollution) with respect to symptoms and lung function of 40-64 year old men in different cities and towns on the U.S. East Coast, including Manhattan.

Even if air quality throughout the United States were improved to match that found in rural Lebanon, Connecticut, any resulting improvement in respiratory health might be minimal, and objective evidence of such an improvement will be very difficult to obtain. For these reasons, I believe that the scientific basis for air pollution control to promote "respiratory health" will remain tenuous. Present evidence suggests that, while there are many compelling reasons to control air pollution, ill health from lung disease is not foremost among them. Thus, the secondary standards for ambient air quality those "requisite to protect the general welfare" may offer the strongest grounds for pollution control.