Associations between criteria air pollutants and asthma.

The evidence that asthma is increasing in prevalence is becoming increasingly compelling. This trend has been demonstrated not only in the United States, but also in the United Kingdom, New Zealand, Australia, and several other Western countries. In the United States, the increase is largest in the group under 18 years of age. There is mounting evidence that certain environmental air pollutants are involved in exacerbating asthma. This is based primarily on epidemiologic studies and more recent clinical studies. The U.S. Clean Air Act of 1970 provides special consideration to the class of outdoor air pollutants referred to as criteria pollutants, including O3, sulfur dioxide (SO2), particulate matter (PM), NOx, CO, and Pb. Standards for these pollutants are set by the U.S. Environmental Protection Agency with particular concern for populations at risk. Current evidence suggests that asthmatics are more sensitive to the effects of O3, SO2, PM, and NO2, and are therefore at risk. High SO2 and particulate concentrations have been associated with short-term increases in morbidity and mortality in the general population during dramatic air pollution episodes in the past. Controlled exposure studies have clearly shown that asthmatics are sensitive to low levels of SO2. Exercising asthmatics exposed to SO2 develop bronchoconstriction within minutes, even at levels of 0.25 ppm. Responses are modified by air temperature, humidity, and exercise level. Recent epidemiologic studies have suggested that exposure to PM is strongly associated with morbidity and mortality in the general population and that hospital admissions for bronchitis and asthma were associated with PM10 levels. In controlled clinical studies, asthmatics appear to be no more reactive to aerosols than healthy subjects. Consequently, it is difficult to attribute the increased mortality observed in epidemiologic studies to specific effects demonstrated in controlled human studies. Epidemiologic studies of hospital admissions for asthma have implicated O3 as contributing to the exacerbation of asthma; however, most study designs could not separate the O3 effects from the concomitant effects of acid aerosols and SO2. Controlled human clinical studies have suggested that asthmatics have similar changes in spirometry and airway reactivity in response to O3 exposure compared to healthy adults. However, a possible role of O3 in worsening atopic asthma has recently been suggested in studies combining allergen challenge following exposure to O3. Attempts at identification of factors that predispose asthmatics to responsiveness to NO2 has produced inconsistent results and requires further investigation. In summary, asthmatics have been shown to be a sensitive subpopulation relative to several of the criteria pollutants. Further research linking epidemiologic, clinical, and toxicologic approaches is required to better understand and characterize the risk of exposing asthmatics to these pollutants.


Introduction
Asthma is a diffuse, inflammatory chronic airway disease generally associated with hyperreactivity to nonspecific bronchoconstrictor drugs, in which eosinophils are prominent among inflammatory cells. The eosinophils and their specific products are responsible in part for bronchial hyperreac-This manuscript was presented at the Workshop on Air Toxics and Asthma-Impacts and End Points held 4 February 1994 in Houston, Texas. Manuscript received: January 17, 1995; accepted February 3, 1995. This article has been reviewed in accordance with the U.S. Environmental Protection Agency policy and approved for publication. It  tivity and for airway epithelial damage and desquamation.
Many patients with asthma are atopic, i.e., have a genetically determined type of immune reactivity that favors IgE response to multiple environmental antigens. These individuals may or may not develop clinical asthma despite the existence of IgE-mediated mucosal and skin reactivity to common environmental antigens. Allergic asthma is, most commonly, an immediate onset reaction (within 1 hr and often within minutes of exposure) and results from the local release of inflammatory mediators. Such reactions are usually, although not exclusively, affected by IgE antibody. Asthmatic reactions may also be persistent or have a late onset and it is possible that other types of immune reactions play a part (1)(2)(3). Recently it has been apparent that cytokines play a role of particular importance in the regulative antibody responses (4). Interleukin 4 (IL-4) and interferon y (IFNy) are the most important cytokines with respect to the regulation of IgE antibody. In mice, the initiation and maintenance of IgE responses is dependent on the availability of IL-4 (5). In contrast, IFNy inhibits IgE production (6). In humans, IL-4 and IFNy also have similar reciprocal effects on IgE antibody (7). Interestingly, there exists a functional heterogeneity among T helper (TH) cells, the class of T lymphocytes required for B lymphocytes to respond to antigen and develop into antibody-producing plasma cells. Two populations of TH cells have been described, designated TH. and TH2' which differ in respect to the spectrum of cytokines they produce following activation ( Figure 1). Although both populations secrete interleukin 3 (IL-3) and granulocyte/macrophage colony-stimulating factor (GM-CSF), only TH1 cells produce interleukin 2 (IL-2), tumor necrosis factor (TNF-,), and IFNy, and only TH2 cells produce interleukins 4, 5, 6, and 10 (IL-4, IL-4, IL-6, and IL-10) (8). Heterogeneity among TH cells has recently been confirmed in humans (9), and there is emerging evidence that immediate-onset allergic reactions in man are associated with the selective activation of TH2 cells. It appears that conditions that favor the activation of TH2 cells and IL-4 production will facilitate IgE antibody responses and the development of respiratory sensitization. Variables that might affect the induction of T.2-type responses include the nature of antigen, the route and duration of exposure, genetic predisposition, and possibly environmental factors.
Air pollution does not affect the health of exposed persons with equal severity. Certain subgroups of people potentially exposed to air pollution can be identified as particularly at-risk from the adverse health effects of airborne pollutants. The U.S. Environmental Protection Agency (U.S. EPA) has described the population at-risk as "...a segment of a defined population exhibiting characteristics associated with significantly higher probability of developing a condition, illness, or other abnormal status...." These subgroups have been identified through clinics demiologic studies of the the six criteria pollutants risk subgroups described I are based on information recent U.S. EPA criteria to set the National Amb Standards (NAAQS) an (Table 1). For example, tI scientific evidence that much more sensitive (i., symptoms at relatively lot to the effects of sulfur di the general healthy popuL In the United States, of 1970 established the p of the nation's effort to ( tion and established the I 108 requires the U.S. El pollutants that "...ma anticipated to endanger p station to issue air quality criteria documents for tory tract such pollutants that reflect "the latest scientific knowledge useful in indicating the kind and extent of all identifiable effects on public health and welfare which may be Bronchoconstriction expected from the presence of such pollutants in the ambient air." The six criteria pollutants, the primary standards, and their permissible levels are shown in Table  Inflammatory 1. In the 1990 Amendments to the Clean mediators Air Act, Congress gave the U.S. EPA authority to impose technology-based standards to control specific toxic substances (air toxics). The ultimate goal is to regulate the list of 189 pollutants based on known or anticipated health risks.
There is mounting evidence that asthma prevalence, morbidity, and mortality are )n increasing in the United States and many other western countries. The reasons for these trends are not clear but are probably complex and involve a number of factors. According to our present understanding, the development of clinical asthma requires al, field, and epi-the presence of host factors and environhealth effects of mental factors. Work in laboratory animals The specific at-(10) and epidemiologic evidence (11) both for each pollutant support the existence of an intimate relacontained in the tionship between atmospheric pollution documents used and IgE-mediated sensitization to environ-)ient Air Quality mental antigens. The issue of whether air id other sources pollution can increase the severity of iere is very strong asthma in an already sensitized population t asthmatics are is not resolved; this topic has recently been e., respond with reviewed (12). Respiratory viruses also w concentrations) seem to be an important cofactor, but their oxide (SO2) than role is unclear. The evidence that asthma is nation.
increasing in prevalence is becoming the Clean Air Act increasingly compelling based on trends 'ublic health basis demonstrated in the United States, across control air pollu-Europe, New Zealand, Australia, and sev-J.S. EPA. Section eral other countries (13). In the United ?A to identify air States, the increase is largest among chily reasonably be dren and adolescents (14). It is still not ublic health," and completely clear how much of the increase Abbreviations: 03, ozone; PM10, particulate matter less than 10 pm aerodynamic diameter; SO2, sulfur dioxide; NO2, nitrogen dioxide; CO, carbon monoxide; Pb, lead. 'The primary standard is to protect against adverse health effects. bChronic obstructive pulmonary disease, including emphysema and chronic bronchitis. CNot to be exceeded more than once a year.
Environmental Health Perspectives in prevalence is due to a real increase and how much is due to increased recognition on the part of the lay public and health professionals. Asthma hospitalizations increased during the 1970s and levelled off during the 1980s for all but a few age groups (14). These increases come at a time when hospitalization for most conditions has been decreasing, and greater use has been made of extended emergency room units and short-stay admissions. Asthma mortality increased in the United States during the 1980s for all ages (14). It is important to note, however, that asthma mortality in the United States is among the lowest for the countries that have reliable mortality statistics (13). Nevertheless, a continuing trend toward increasing mortality is worrisome. It will take a number of years to sort out why these trends for asthma are occurring. Attention must be focused on finding the causes and risk factors for asthma so that appropriate intervention strategies can be developed.
The purpose of this report is to review the available data that provides evidence for an association between asthma and exposure to selected criteria pollutants and that focuses on clinical studies using physiological and biochemical end points.

Responses of Asthmatics to Criteria Pollutants
Not all criteria pollutants have been shown to have an effect on asthmatics (Table 1). Therefore, the discussion below will focus on those criteria pollutants that have been shown to have pulmonary effects on asthmatics, namely SO2, particulate matter less than 10 pm aerodynamic diameter (PM10), 03, and NO2.

Sulfur Dioxide
Sulfur dioxide (SO2) is emitted primarily by coal-and oil-fired plants and by industrial processes involving fossil fuel combustion, as sulfur is a natural contaminant of these fossil fuels. SO2 and particles are directly regulated under the Clean Air Act whereas the possibility of a standard for acid aerosols has only recently been considered by the U.S. EPA. Even though particles, sulfur oxides (SOX), and acid aerosols are chemically distinct, they are all released by fossil fuel combustion processes and are usually present together as components of a complex mixture.
Epidemiologic data indicate an association between particles and acid aerosols and increased mortality, although the association is notably weaker for SO2' The epidemiologic studies are limited in that the effects of these different pollutants cannot be assessed separately. This can, however, be addressed in controlled chamber exposure studies.
SO2 is a respiratory irritant that does not cause substantial acute or chronic adverse effects in animals exposed to ambient concentrations (15,16). Because SO2 is 50 times more soluble than CO2 in water at 300 C, it is mostly absorbed in the upper airways in subjects at rest, although increasing ventilation results in deposition in deeper parts of the lung. Controlled human studies of healthy subjects exposed to SO2 at rest or with exercise have failed to demonstrate effects on respiratory mechanics at levels up to 1.0 ppm (17)(18)(19)(20). In contrast, exposure to low levels of SO2 does alter the lung function of asthmatics (21,22). Asthmatic subjects exercising in air containing SO2 develop bronchoconstriction within minutes, even at levels as low as 0.25 ppm. When asthmatic subjects are briefly exposed to SO2, many develop significant increases in airway resistance, especially with oral breathing at a minute ventilation greater than 40 1/min (23)(24)(25)(26). Spirometric functions (e.g., forced expiratory volume in 1 sec; FEV1) were also decreased in response to exposure to 0.4 to 1.0 ppm SO2. The data indicate heterogeneity in the response to SO2 among asthmatics (27) and that individual responses are relatively reproducible (27,28). SO2 can be considered to have comparable effects to those of nonsensitizing bronchoconstrictors, such as histamine and methacholine, but are not comparable to those of specific antigens and certain occupational chemicals that can provoke so-called late reactions.
The predominant acid aerosols are sulfate and bisulfate ions that are found in solution in water droplets in the atmosphere. Because of the abundance of ammonia in the atmosphere, sulfate is usually found in one of the two above-mentioned forms. Some short-term controlled inhalation studies suggest that asthmatics are more sensitive than normal subjects to H2SO4 aerosols at concentrations in excess of the high range of observed ambient levels of acid aerosols in the United States (29,30). Other studies, however, have failed to confirm the susceptibility of asthmatics (31)(32)(33). One reason for the difference in the results obtained from different laboratories may be that endogenous respiratory NH3 provides protection against inhaled acidic pollutants. High oral respiratory levels of NH3 have been shown to inhibit bronchoconstriction induced by H2SO4 in exercising asthmatics (34). Taken together, these studies show that an effect of acid aerosols clearly occurs in some asthmatics, but with less consistency than the SO2 response. The variation in response of asthmatics to acid aerosols among various laboratories awaits explanation.

Pr iculat Matter
The initial reference method for total suspended particles (TSP) measured mass of all suspended particles. In 1987, the EPA restricted the NAAQS to the mass concentration of inhalable particles less than 10 pm aerodynamic diameter (PM10) (35) ( Table 1). The NAAQS for PM does not specify chemical composition of particles. The 10 pm size cutoff focuses monitoring and regulatory efforts on particles of a size that would be deposited in and could damage the lower airways and gas-exchanging portions of the lung. So far, most of the human health effects data of PM is based on epidemiologic data. High particulate and SO2 concentrations were associated with substantial short-term increases in morbidity and mortality during dramatic air pollution episodes of the past (36)(37)(38)(39)(40)(41).
The obvious increases in morbidity and mortality that accompanied these episodes and the frequency and severity of respiratory complaints left little doubt that pollution exposure caused the adverse effects. Exposure to particulates was associated with increases in respiratory symptoms during the 1960s in Londoners with chronic obstructive lung disease (COPD) (37). Daily total mortality has recently been reported to be positively associated with TSP concentrations. An analysis of data from Steubenville, Ohio, for 11 years (1974-1984) showed a statistically significant increase in daily total mortality associated with TSP concentrations of the previous day (42). A similar association was shown in a study conducted in Philadelphia (43). Analysis of causespecific mortality in Philadelphia showed the strongest association with respiratory mortality and, secondarily, with cardiovascular mortality. Relatively few studies have specifically assessed the association between PM and asthma. Daily reports of respiratory symptoms were collected among a panel of 209 adult asthmatics residing in Denver, Colorado (44). Hydrogen ion and PM2 5 levels were associated with moderate to severe cough and shortness of breath, both of which are indicators of asthma. Volume (45,46).
The biological effect of particles is determined by the physical and chemical nature of the particle, the physics of the deposition in the respiratory tract, and the biologic events occurring in response to the particle. There is a general belief that a disturbing aspect of the epidemiologic findings thus far is that effects appear independent of the chemical composition of the particles. Approaches to understanding the mechanisms) responsible for the biologic effects of particles include in vivo instillation and inhalation of particles followed by analyses of tissues and fluids for toxicological changes indicative of a disease and in vitro exposures of particles to pulmonary cell types including macrophages and epithelial cells. Intratracheal instillation of particle suspensions directly into the lung of rodents has been used to study direct cytotoxic effects (47). Bronchopulmonary lavage in animals at various times postexposure has enabled the performance of assays of inflammation and edema for various particles (48,47). In vitro study of particles is a useful approach to assess the response of cells that may contribute to the inflammatory process as it relates to interactions with particles. Several studies describing the responses of alveolar macrophages (49,50) to particles such as SiO2 and TiO2 have been described. So far none of these studies involved the use of pulmonary cells obtained from asthmatic individuals. It is noteworthy that recent studies with normal human subjects exposed to instilled particles followed by bronchoalveolar lavage (BAL) at different times revealed an inflammatory response for 2 days following the instillation of the particles as indicated by neutrophilic infiltrates (51). Inflammation may be exacerbated by alteration in the deposition characteristics of particles in compromised lungs. Again, no such experiments have so far been performed with asthmatic subjects.
Collectively, the epidemiologic studies provide provocative evidence for adverse pulmonary health effects associated with particulate pollution. The association between PM and acute mortality and morbidity has primarily been demonstrated with people who have cardiopulmonary disease and asthma. Clearly, there is a need to better understand the underlying mechanisms responsible for these effects by performing in vivo and in vitro exposure studies with a variety of particles comparing the responses of asthmatics to normal subjects. Ozone Ozone (03) is a gas that occurs with other photochemical oxidants and fine particles in the complex mixture called smog. Ozone is formed by a series of sun-driven reactions involving nitrogen oxides (NOX) and volatile organic compounds (VOC) arising largely from mobile and stationary combustion sources (52)(53)(54). As a potent oxidant, 03 is capable of reacting with a variety of extracellular and intracellular molecules, particularly those containing thiol or amine groups or unsaturated C=C bonds (55).
Bates and Sizto (56) studied hospital admissions in southern Ontario, Canada, an area with a population of 7 million people and observed increased rates of admissions for asthmatic subjects in the summer, which correlated with both 03 and suspended sulfates. These results implicate 03 as a contributing cause of asthma admissions; however, the study design could not separate the 03 effects from concomitant effects of acidic aerosol and SO2 (56,57). A more recent study by Spektor et al. (58) found significant decreases in peak flow and FEV1 in children in summer camp, although the 03 concentration during the study period did not exceed 0.12 ppm. None of the children were reported to have a history oflung disease or atopy.
Over the last 15 years, controlled human exposure studies have clearly demonstrated that the lung responds to 03 exposure by irritative cough and substernal chest pain on inspiration; decrements in forced vital capacity (FVC) and FEV1; increase in specific airway resistance (SRaw) and airway responsiveness; and neutrophilic inflammation of the airway submucosa accompanied by increased level of mediators and cytokines in the BAL (59). Even among homogeneous study populations, there has been a wide range of susceptibility to these effects (60,61).
Many clinical studies have failed to show that subjects with asthma are more sensitive to 03 than are healthy subjects (62)(63)(64). These studies have typically involved subjects with clinically mild asthma performing mild exercise, and those studies focused on spirometric abnormalities. Recent data suggest that if more intense exercise is imposed, asthmatic subjects show increased SRaw to a 2-hr exposure to 0.4 ppm 03 (65,66), which is in excess of the response seen in nonasthmatics.
A number of studies have addressed the effects of 03 following an exposure to or in the presence of other pollutants. A few controlled human exposure studies addressing the interaction between 03 and other criteria pollutants have been performed. Prior exposure of healthy individuals to other pollutants may modify the response to 03 (67). Similarly, preexposure to 03 may change a response to other pollutants. Recently, preexposure to 0.12 ppm 03 for 45 min followed by 0.10 ppm SO2 for 15 min elicited greater bronchial hyperreactivity in adolescent asthmatics than SO2 alone or 03 exposure followed by 03 (68). Exposures to mixtures containing acid aerosols and 03 have shown modestly increased effects (20,69). For example, lung function, as measured by spirometry was worse (70) when 03 was inhaled with SO2 than with either alone, although this study has not been replicated. Very few studies focusing on the effects of exposure to mixtures on asthmatics have been carried out so far. However, the concept of influencing the asthmatic response by combining exposure to 03 with a challenge of a specific allergen has created a lot of interest in the potential indirect effects of 03 exposure. In one study, individuals with allergic rhinitis were initially exposed to clean air or 0.5 ppm 03 for 4 hr (71). The high level exposure to 03 did not enhance the acute response to antigen in the nose under these experimental conditions. In a more recent study, Molfino et al. (72) examined the effect of 03 without exercise on the airway response to inhaled allergen in adult subjects with asthma. They reported 03induced increases in bronchial responsiveness to inhaled ragweed or grass pollen as measured by allergen bronchoprovocation tests. This is a very provocative study that needs to be confirmed. Along the same line of experiments, studies are currently being conducted examining the effects of preexposure to 03 and exercise followed by a nasal allergen challenge (using house dust mite as antigen) on atopic asthmatics (73). Preliminary data from those studies suggest that the exposure to 03 before the nasal challenge caused a shift in the dose of allergen needed to induce symptoms and caused an increase in the levels of inflammatory cytokines detected in the nasal lavage.

Environmental Health Perspectives
In developing future programs intended to better understand the effects of 03 on asthma, there is a need to do longitudinal and cross-sectional population studies, clinical exposure studies including nasal and bronchial challenges with allergen, and in vitro studies that will be primarily done to address mechanistic aspects of environmental asthma and how the epithelial cells from asthmatics may differ from normal epithelial cells. Future studies should also recognize the fact that 03 is not only an outdoor air pollutant, but is also present in the indoor air environment (albeit at lower concentrations) where we spend most of our time. Nitrogen Dioxide Nitrogen dioxide (NO2) is an oxidant that contaminates ambient air in many urban and industrial locations and indoor air in homes with unvented combustion appliances. The present NAAQS refers to the annual average concentration; a short-term standard is not in place. However, some evidence, largely from studies of the effects of acute exposure to NO2 on lung function and airway responsiveness of asthmatics, suggests that a new standard governing short-term concentrations may be warranted.
During high temperature combustion, oxygen reacts with nitrogen to generate nitric oxide (NO) and, to a lesser extent nitrogen dioxide (NO2) and other nitrogen oxides. NO2 is highly reactive and, in the presence of sunlight, it participates with hydrocarbons and oxygen in the complex reactions that form 03 and other photochemical oxidants. The principal source of NO and NO2 in outdoor air is motor vehicle emissions, but power plants and fossilfuel-burning industries also contribute. In most U.S. cities, ambient levels of NO2 vary with traffic density. Of 163 counties reporting NO2 monitoring data, only Los Angeles County did not meet the NAAQS of 0.053 ppm annual arithmetic mean. In contrast to other criteria pollutants, NO2 is a widespread contaminant of indoor as well as outdoor air, and indoor levels can exceed those outdoors. Indoor sources of NO2 include cooking ranges (74) and kerosene space heaters (75). Because over one-half of the residences in the United States have gas cooking stoves and Americans spend a large proportion of time in their homes, the residential environment has generally been found to be the most important contributor to the population's total exposure to NO2 (76,77). The toxicity of NO2 is generally attributed to oxidative capabilities (78,79), although as an oxidant it is less reactive than 03 (78,80). Empirical and theoretical studies indicate that NO2 penetrates to the lung periphery, with the centriacinar region as its primary deposition site, and has a somewhat greater airway deposition than 03 due to its higher solubility in water (81,82).
Only a few epidemiologic studies have directly addressed the association between NO2 concentration in outdoor air and respiratory illness. Much more extensive information is available from studies examining the effects of indoor exposures resulting from emissions from gas cooking stoves and space heaters (83)(84)(85). The evidence for a relationship between exposure to NO2 and respiratory illness from studies of indoor exposure does not consistently indicate an increased incidence of more severe illness in children and adults classified as having higher exposures to NO2 (85)(86)(87). The evidence is more abundant for children. While several studies showed significantly increased risk of respiratory illness for children living in homes with gas stoves compared with children living in homes with electric stoves (88,89), other studies did not (84,(90)(91)(92)(93). Metaanalysis summarizing data from 11 epidemiologic studies ofhealth effects associated with NO2 showed a significant association between estimated NO2 exposure and illness (94).
Studies examining responses of healthy volunteers to acute exposure to NO2 (< 1 ppm) have generally failed to show alterations in lung mechanics of healthy volunteers (95,96). Several recent observations indicate that NO2 exposures in the range of 1.5 to 2.0 ppm cause small but significant increases in airway reactivity. Mohsenin (97) found that a 1-hr exposure to 2 ppm 02 increased responsiveness to methacholine, as measured by changes in airway responsiveness, without directly affecting lung function.
Early findings (98) brought attention to the intriguing possibility that a relatively brief exposure of asthmatics to low-level NO2 (0.1 ppm) might enhance subsequent responsiveness to bronchial challenge with a bronchoconstricting drug, although these results could not be replicated in a later study (99). Kleinman et al. (100) have shown that inhalation of 0.2 ppm NO2 for 2 hr, although not causing alterations in flow rates or airway resistance, resulted in increased reactivity to methacholine. Bauer et al. (101) reported that NO2 inhalation produced significant decrements in forced expiratory flow rates after exercise and increased responsiveness to cold air challenge. Despite this evidence of hyperresponsiveness of asthmatics to low levels of NO2, considerable controversy remains because of the inconsistency of the results (59). The effects of NO2 exposure on S02-induced bronchoconstriction have been examined, but with inconsistent results. Collectively, the findings reported above differ from those of Linn et al. (102,103), for asthmatics inhaling concentrations of NO2 as high as 4.0 ppm. It is evident that a wide range of responses occur among asthmatics exposed to NO2. This variation may in part reflect differences in subjects and exposure protocols. However, the consistency of responses of asthmatics to NO2 across a 1-year interval suggests that some asthmatics are inherently more responsive and implies a need for better understanding of NO2 and asthma (104).
NO2 appears to be much less potent than 03 in eliciting a neutrophilic inflammatory response in normal subjects. Two preliminary reports (105,106) describe increases in numbers of PMNs obtained by BAL following a single 4or 6-hr exposure to 2.0 ppm N02 Devlin et al. (105), observed the increase in PMN only in the cells recovered from the bronchial lavage aliquot. The effect of NO2 on challenges with a specific allergen in atopic asthmatics is of considerable interest considering the recent data by Molfino et al. (72) and Peden et al. (73).
Responses to NO2 exposure in clinical studies are characterized by marked variability, which directs attention toward identifying determinants of susceptibility. Characterizing these and other factors may hold the key to understanding the risks of NO2 exposure, especially to sensitive subpopulations such as asthmatics.