Health Effects of Gaseous Air Pollutants

JON G. AYRES

1 Introduction

Gaseous air pollutants constitute an important overall component of both outdoor and indoor air and are recognized to cause health effects, essentially in individuals with pre-existing disease. For the purposes of this chapter the gases, the primary pollutants sulfur dioxide (SO2), nitrogen dioxide (No2) and carbon monoxide (CO) with the secondary pollutant ozone, will be considered whereas acidic species will not as they are generally regarded as part of the particulate fraction. As it is likely to be the acidic nature of those species that are important in health terms, even those acids which are present in the air as a vapour phase will not be considered here.

The sources of these pollutants are important when considering health effects because sources relate to individual and population exposures. The main source of SO2 is from fossil fuel burning, the major contributors in the UK being coal-fired power stations. Nitrogen dioxide is derived from vehicle emissions, industrial sources (including power stations) and, in the indoor environment, from combustion of gas. Although for smokers of cigarettes the major contribution to their CO exposure by far comes from their habit, in ambient air the main source is again traffic derived. Ozone is formed by the action of ultraviolet light on oxides of nitrogen and hydrocarbons, so is essentially a pollutant of the summer months in climates such as the UK but may be more perennial in countries where sunlight is present all year round. Ozone levels are generally higher downwind from a city because of the atmospheric chemistry of the formation of ozone combined with the fact that ozone, a very reactive gas, is quickly neutralized by nitric oxide in urban areas.

The health effects of gaseous pollutants have been determined in a number of different ways:

1. By chamber (human challenge) studies

2. By studies of morbidity (e.g. symptoms, inhaler use), usually in panels of subjects perceived to be at risk

3. From studies of hospital admissions (i.e. routinely collected data)

4. From studies of mortality

Chamber studies enable the effects of individual pollutants to be studied alone or in combination with other pollutants on volunteers under strictly controlled conditions. The chief ad vantage of this type of study is that accurate doses can be delivered and the effects of selected co-factors assessed. However, the volunteers involved in such studies are usually normal subjects or patients with mild asthma who tend to be younger, in contrast to the older subjects who are more likely to be affected by air pollution. Additionally, in chamber studies, the duration of exposure is relatively short compared to outdoor, real-life exposures and consequently it may be difficult to extrapolate findings from these types of studies to the effects that would be seen in the overall population exposed to the outdoor environment. Children are not studied in these types of experiments for ethical reasons, which prevents study of an age group where asthma is very common and in whom the health effects of pollution are often perceived to be significant. However, despite these caveats, chamber studies have, in general, provided very useful information as to the presence or absence of effects of specific pollutants at specific doses and have provided useful insights into the mechanisms of these effects.

Epidemiological studies have been much more informative about health effects both at an individual and population level, studying as they do the real-life situation. The difficulty comes in deciding how large an effect may be and to what specific pollutant or pollutant mix such an effect may be attributable. On a day-to-day basis, exposure to air pollutants may have an immediate effect, either on the same day as a rise in air pollution or perhaps delayed, lagging two, three or more days after a rise. In some situations the cumulative or average exposure over a period of three days or more may be important in determining health outcome. It is even possible that longer lags may be more important for differing health end points, an area which is currently being explored.

There is no doubt that there is a range of sensitivities to pollutants across different 'at risk' groups in terms of health effects of air pollution. Patients with pre-existing lung and heart disease appear to be particularly at risk, notably patients with asthma and chronic obstructive pulmonary disease (COPD). More recently, the effects of particulate pollution on patients with coronary heart disease and cerebrovascular disease have been identified, but the role of gaseous pollutants in these two disease categories is not so clear. Asthma is a common condition, affecting around 6% of the total population of the UK. In this condition, the lining of the bronchial tree is inflamed and unduly sensitive to external triggers, such as allergens in those sensitized, viral infections or physical stimuli such as exercise or inhaling cold air. Consequently, these patients a re not only important as a risk group for the effects of air pollution but also act as a group where changes in lung function are frequent and measurable when trying to define the presence and size of an effect from an external stimulus. COPD is essentially a disease of cigarette smokers and although, like asthma, it is also an inflammatory condition, on a day-to-day basis these patients show no marked changes in lung function. Patients with either COPD or asthma develop symptoms because of the airway narrowing resulting from the inflammatory process. Where the baseline airway diameter is small, only minor reductions in diameter can produce marked reductions in airflow and hence symptoms. However, it is at least intuitively logical that, for respiratory diseases, inhalation of polluted air can lead to a deterioration in symptoms.

Coronary heart disease and cerebrovascular disease share a common pathogenesis characterized by the formation of atheroma in the arteries supplying the heart or brain, respectively. In contrast to diseases of the respiratory tract, it is not entirely clear at present how inhalation of air pollutants can lead to vascular health effects, but associations have been shown between ischaemic heart disease deaths and ozone, although the major impacts in this disease area appear to derive from particulate exposure, so further discussion falls outside the remit of this chapter.

It is important to recognize that there may well be interactions between different elements of the pollutant mix in determining health effects. The statistical analysis of time series data (i.e. following individuals over long periods of time or considering hospital admissions and mortality over periods of time) will regard each pollutant as a separate entity acting on its own behalf. Because the possible degree of interaction of different pollutants is not known it is impossible to analyse separately for any combined effects. The studies therefore allow for the effects of all other pollutant and non-pollutant factors on that health outcome before determining a residual effect which is then attributed to that pollutant.

2 Quantification of Effect

Quantification of these effects is not easy, but certain guidelines can be used when trying to determine how much of an impact air pollution may have on the public health. A basic concept is that of a threshold. For all the gases considered here (with the probable exception of CO) the assumption has been made that at a population level the effect of the pollutant on health is linearly related and that the relationship passes through zero. Consequently, once the effect size coefficient is known for that pollutant, an estimate of overall effect on the population under consideration can be determined. These quantification estimates will vary from country to country (and almost certainly from area to area within a country) and so we will not consider this further in numerical terms here.

3 Chronic Effects

These discussions apply to the effects of short-term changes in health outcomes which can, in theory, be relatively easily recognized. In contrast, the question of whether long-term exposure over years to particular pollutants or pollutant mixes can lead to long-term health effects as yet remains to be convincingly answered, but may be more important in public health terms. The evidence for such a chronic effect with respect to gaseous air pollutants is scant, whereas there are some data with respect to long-term exposure to particulate pollution which are discussed elsewhere in this volume. Determination of chronic effects is largely dependent upon acquiring data prospectively over a matter of years (longitudinal or cohort studies). Cross-sectional studies where prevalence rates are compared between different areas at the same point in time can contribute to this question to some extent, although they are regarded as being less powerful studies and more likely to be open to uncorrectable confounding. There are no satisfactory longitudinal studies which have considered the effects of gaseous pollutants like the Six Cities Study and the American Cancer Association Study have considered the effects of particulate pollution in this regard. One series of studies of Seventh Day Adventists (an attractive study group as these individuals do not smoke cigarettes, thus removing the major complicating cause of respiratory and heart disease) has suggested that long-term exposure to ozone is associated, in men only, with an increased risk of developing asthma. However, this is an unusual group in an unusual setting and it is not easy to extra polate these findings to other populations. Consequently, we will only consider the short-term effects in this chapter.

4 Sulfur Dioxide

Controlled Challenge Studies

Normal Subjects. There is consistent evidence that normal subjects are much less sensitive to the effects of inhaled SO2 than are patients with asthma. Although one study showed small increases in airways resistance at exposure of 1000 ppb (2860 µg m-3) after a short (ten minute) exposure, other studies have failed to confirm this. At exposures of 4 ppm or greater (11 440 µg m-3), clear effects on airway size have been noted both at rest and with intermittent light exercise. However, within these averaged group findings a wide range of individual responses can be found, suggesting that there may be individual sub-sets of normal subjects who show a greater response on exposure to this pollutant gas. The clinical significance of these effects is far from clear at present.

There are a number of different factors which may help to explain these variations in response, chief amongst which is the amount of gas entering the lower airways. It is always assumed that SO2 is a very soluble gas and that if nasal breathing is predominant then doses to the lower respiratory tract will be much reduced because of the nasal trapping of the gas at normal ambient concentrations. A second factor is that some subjects appear to breathe more deeply on exposure to SO2, thus increasing the dose to the lower respiratory tract. Temperature and humidity can also have a bearing in this regard, in particular cold air which can cause a degree of airway narrowing, although this is only a very small effect in normal individuals, being much more marked in subjects with asthma. However, in challenge studies the effects of these factors should be able to be kept constant between individuals and between exposures within given individuals. It is not known whether cigarette smoking enhances or inhibits any effects of sulfur dioxide on normal subjects.

Asthmatic Subjects. In patients with asthma, effects on lung function are seen at much lower concentrations. The study by Sheppard et al., while not demonstrating much of an overall effect, showed increases in airways resistance in two very sensitive subjects at an exposure of 100 ppb (286 µg m-3). Other workers, exposing asthmatic subjects to 200 ppb (572 µg m-3, showed small symptom changes but, in a further study, no changes in lung function at 200 ppb exposures associated with heavy exercise. The same group of subjects were exposed to 400 ppb (572 µg m-3) SO2 while undergoing heavy exercise and produced small changes in lung function, but it is not until exposures to levels of around 500 ppb (1430 µg m-3) are employed that there is clear evidence of sulfur dioxide enhancement of exercise-induced airway narrowing. These responses were seen after exposures of a matter of minutes, whereas, in other studies, longer exposures (up to hours) appeared to be needed to produce an effect. These differences in effect size may be due to differing volunteer characteristics, habituation to the individual's usual air pollutant exposure resulting in tolerance to these levels of laboratory exposure (a recognized phenomenon in studies of ozone challenge) or to methodological differences.

The size of the effect in the challenge studies can be determined by measurement of lung function, the usual measures being those obtained from spirometry, namely the FEV1 (the forced expired volume in 1 second) and the FVC (forced vital capacity). For SO2 the results of exposure on lung function are reasonably consistent across studies, with falls of the order of 50 mL in FEV1 from an approximate start volume of 3 litres for an exposure dose of 200 ppb of SO2. These changes are easily reversible and the size of the effect is small, although if repeated over time these changes may become clinically significant, particularly if the pattern of induced inflammatory change was seen to be relevant to the type of inflammatory change associated with chronic asthma.

Another method of assessing the airway response to a pollutant gas is to measure the change in bronchial responsiveness (bronchial hyper-reactivity) of the individual. This can involve measuring the effect of gas exposure on the response to a non-specific irritant such as methacholine or histamine, a standard method used to characterize the severity of subjects with asthma. Alternatively, the subjects can be exposed to a range of doses of the specific gas and a curve of lung function responses constructed. The study by Horstman et al. represents an example of the latter, taking as its main outcome measure the PC100 sRaw, the provocative concentration of SO2 causing a 100% increase in specific airways resistance, a sensitive index of flow through larger intrapulmonary airways. Their results show, in a group of asthmatic subjects, a range of responses to sulfur dioxide (Figure 1) with a medium PC100 sRaw of 750ppb (2145 µg m-3). However, the usual concentrations of sulfur dioxide seen in ambient air in the UK rarely exceed 120 ppb (343 µg m-3) nowadays, although occasionally, during episodes, levels in excess of 200 ppb have been recorded.

Extrapolating the findings from the chamber studies to the effects on public health is, therefore, somewhat difficult, particularly as those most susceptible to the effects of air pollution (i.e. those with more severe disease) are not used in challenge studies. It is likely that these more severely affected individuals have a much lower threshold for developing symptoms or changes in lung function on exposure to air pollution. Consequently, controlled chamber studies can be used to show whether effects in response to a pollutant challenge could occur in a given group of subjects, but extrapolation to all potential members of such a group in real life would be unwise.

Mechanisms. The way in which sulfur dioxide can result in these pathological changes in the airway are likely to be multiple and in some individuals a particular mechanism may be more important than in others. Animal studies show that SO2 can activate mucosal sensory nerves, leading to airflow obstruction both by central neural reflex and by local axon reflex changes (neurogenic inflammation). Although it is likely that these effects are also true for man, there is no direct work to confirm this. SO2 may also act by non-neural mechanisms with mucosal damage leading to release of inflammatory mediators, perhaps attracting inflammatory cells, notably neutrophils and eosinophils, to the airway wall.

Morbidity Studies

Studies on morbidity (i.e. changes in symptoms and treatment use) are conducted by establishing a cohort (panel) of susceptible individuals and following them prospectively over a period of weeks or months. Over this time the individual will record symptoms twice daily and, in most studies, a measure of lung function such as peak expiratory flow. This approach produces a large amount of data over time at both an individual and a group level. A second way of assessing the response to ambient changes in air pollution is to study individuals during an air pollution episode. However, the quality of information in the latter situation is generally less good as the study is by definition retrospective and opportunistic.