Abstract
The association between long-term exposure to ambient air pollution and respiratory symptoms was investigated in a cross-sectional study in random population samples of adults (aged 18 to 60 yr, n = 9,651) at eight study sites in Switzerland. Information on respiratory symptoms was obtained with an extended version of the European Community Respiratory Health Survey questionnaire. The impact of annual mean concentrations of air pollutants was analyzed separately for never-, former, and current smokers. After controlling for age, body mass index, gender, parental asthma, parental atopy, low education, and foreign citizenship, we found positive associations between annual mean concentrations of NO2, total suspended particulates, and particulates of less than 10 μ m in aerodynamic diameter (PM10) and reported prevalences of chronic phlegm production, chronic cough or phlegm production, breathlessness at rest during the day, breathlessness during the day or at night, and dyspnea on exertion. We found no associations with wheezing without cold, current asthma, chest tightness, or chronic cough. Among never-smokers, the odds ratio (95% confidence interval) for a 10 μ g/ m3 increase in the annual mean concentration of PM10 was 1.35 (1.11 to 1.65) for chronic phlegm production, 1.27 (1.08 to 1.50) for chronic cough or phlegm production, 1.48 (1.23 to 1.78) for breathlessness during the day, 1.33 (1.14 to 1.55) for breathlessness during the day or at night, and 1.32 (1.18 to 1.46) for dyspnea on exertion. No associations were found with annual mean concentrations of O3. Similar associations were also found for former and current smokers, except for chronic phlegm production. The observed associations remained stable when further control was applied for environmental tobacco smoke exposure, past and current occupational exposures, atopy, and early childhood respiratory infections when restricting the analysis to long-term residents and to non- alpine areas, and when excluding subjects with physician-diagnosed asthma. The high correlation between the pollutants makes it difficult to sort out the effect of one single pollutant. This study provides further evidence that long-term exposure to air pollution of rather low levels is associated with higher prevalences of respiratory symptoms in adults.
Over the past decade, literature on short-term effects of air pollution has grown considerably (a summary of the past 15 yr has been provided [1]). Information on long-term effects of air pollution, especially for respiratory symptoms in adults, is comparatively more scarce. Community studies in different geographic regions of the United States and of Europe have shown associations of respiratory symptoms and conditions with long-term exposure to total suspended particulates (TSP) and SO2 (2-8), to particulate matter (9-11), to black smoke (12), and to NO2 (8). Increasingly, associations appear to be more consistent with bronchitic symptoms than with asthma/ asthmatic conditions (1, 13). Furthermore, studies of hospital admissions and mortality studies point to an association of short- and long-term exposure to air pollution with symptoms that are related both to pulmonary and to cardiac diseases (14-21). Biologic explanations for adverse health effects of air pollution (inflammatory pathways, toxic mechanisms, alterations in mucociliary clearance, structural changes) are still the subjects of toxicologic research (1). Air pollution may harm by mechanisms that can be relevant for short- and long-term effects.
The Swiss Study on Air Pollution and Lung Diseases in Adults (SAPALDIA) has found an association between long-term exposure to air pollutants and decrements in lung function across eight study areas, both in subjects reporting respiratory symptoms and in symptom-free subjects (22). The magnitude of the effect was similar in these two groups, as well as in smokers and nonsmokers. Furthermore, the effect could be observed across exposure gradients both among and within the eight study areas (23). Whether the symptoms themselves are associated with air pollution has not yet been addressed. In this report we present the results of an analysis of the association of long-term exposure to air pollution with three symptom groups: (1) bronchitic symptoms/bronchitis; (2) asthmatic symptoms/asthma; and (3) nonspecific cardiopulmonary symptoms. These associations were studied in random population samples of adults at eight sites in Switzerland with rather low levels of air pollution. To our knowledge, this is the first analysis of the association of long-term exposure to measured levels of particulates of less than 10 μm in aerodynamic diameter (PM10) and respiratory symptoms in adults.
SAPALDIA is a multicenter study with cross-sectional and longitudinal components, and is designed to examine the potential association of long-term exposure to air pollution with respiratory health (24). For the cross-sectional part of the study, random population samples of adults aged 18 to 60 yr were drawn in 1991 from the registry of inhabitants of eight areas in Switzerland (Aarau, Basel, Davos, Geneva, Lugano, Montana, Payerne, and Wald) representing a range of air-pollution exposure, urbanization, altitude, and meteorologic conditions. Subjects were required to have been resident in the respective area for at least 3 yr to be eligible for the study. Subjects were invited via a mailed letter to participate in the study. If no answer was received, at least five telephone calls were made until definite refusal or cooperation was ascertained. A total of 9,651 subjects (59% of those invited) completed a standardized questionnaire on respiratory health, with 290 items being asked by trained interviewers. The questionnaire was an extended version of the European Community Respiratory Health Survey (ECRHS) questionnaire, which is a modified version of the International Union Against Tuberculosis and Lung Disease (IUATLD) questionnaire (13). The questionnaire was available in three languages (German, French, and Italian), according to the language spoken in the pertinent region of Switzerland. The subjects' respiratory health was also assessed through spirometry, a bronchial challenge test with methacholine, and allergy tests (skin prick testing, blood sampling for total IgE, and a fluoroenzyme immunoassay [CAP Phadiatop; Kabi Pharmacia, Uppsala, Sweden] for individual allergens). A detailed description of the study methodology has been published (24). Subjects who declined to participate were asked a shorter set of questions on smoking and respiratory history (either through a mailed short ECRHS questionnaire or by phone), and 72% of nonparticipants provided this information (Table 1). The proportion of current smokers among participants was about the same as among nonparticipants (men and women). As compared with nonparticipants, participants were more likely to report asthma, wheezing without cold, and allergic rhinitis, but were less likely to report phlegm production (24). Characteristics of the study population and prevalences of major covariates are given in Table 2.
| Participants*(n = 9,651) | Nonparticipants† | |||
|---|---|---|---|---|
| Male, % | 49.2 (44.5–51.4) | 49.5 | ||
| Age ⩽ 40 yr, % | 46.6 (40.8–51.4) | 43.0 | ||
| Current smokers, % | 33.5 (26.8–39.6) | 32.2 | ||
| Wheezing without cold, %‡ | 7.7 (5.1–11.2) | 5.5 | ||
| Regular phlegm, %§ | 7.8 (4.4–10.8) | 10.6 | ||
| Allergic rhinitis‖ | 18.0 | 15.2 |
| Never-Smokers (n = 4,229 ) | Former Smokers (n = 2,175 ) | Current Smokers (n = 3,232 ) | ||||
|---|---|---|---|---|---|---|
| Sex, % of females | 60.0 | 43.9 | 43.6 | |||
| Low education, %* | 16.7 | 14.9 | 18.4 | |||
| Nationality, % non-Swiss | 16.8 | 17.4 | 20.4 | |||
| Atopy, %† | 25.8 | 20.4 | 21.7 | |||
| Elevated total serum IgE, %‡ | 20.5 | 19.9 | 27.5 | |||
| Early childhood respiratory infections, % | 8.1 | 7.9 | 6.9 | |||
| Ever exposed to airbone irritants at work, %§ | 30.7 | 38.3 | 42.1 | |||
| Currently exposed to airbone irritants at work, %‖ | 27.2 | 32.0 | 37.4 | |||
| Gas cooking, % | 15.7 | 16.0 | 18.9 | |||
| Passive smoke exposure, %¶ | 29.9 | — | — | |||
| Exposure to maternal smoking, % | 10.5 | 10.7 | 16.5 | |||
| Exposure to paternal smoking, % | 48.2 | 56.9 | 60.8 | |||
| Age, mean (SD) | 40.3 (12.5) | 44.5 (10.0) | 39.9 (11.1) | |||
| Body mass index, (kg/m2), mean (SD) | 23.7 (3.8) | 24.5 (3.9) | 23.9 (3.9) | |||
| Pack-years, mean (SD) | — | 14.6 (17.8) | 22.6 (19.8) | |||
| Cigarettes/day, mean (SD) | — | — | 17.4 (12.2) | |||
| Years since quitting, mean (SD) | — | 11.9 (8.9) | — |
Air pollutants were monitored by local authorities in each of the eight regions included in the study. SO2, NO2, TSP, O3, and PM10 were measured continuously (26). SO2, NO2, and O3 concentrations were determined as half-hourly means, using fluorescence (Horiba; Monitor Labs, Environment SA, Japan), and chemiluminescence (Tecan, Horiba; Monitor Labs) assays and UV photometry (Horiba; Monitor Labs, Environment HA). TSP was determined gravimetrically through high-volume sampling (daily means) and through beta radiation analysis (half-hourly means) and using a tapered element oscillating monitor (TEOM). In addition, PM10 was measured with Harvard impactors during 1993.
A quality surveillance for gaseous pollutants was done across the study areas at the end of the measuring period in 1991. All monitors gave results within a range of ± 10%, with the exception of that for SO2 at Lugano (15% above the reference) and O3 at Montana and Geneva (21% and 29% above the reference, respectively). The annual pollution values for these locations were adjusted accordingly.
For NO2 and TSP, the annual average concentration of pollutants over the year 1991 was used as an indicator of long-term air pollution exposure. PM10 measurements were available only for 1993. Because of the better standardization of PM10 data, the analysis presented here uses the 1993 PM10 measurements as an indicator of particle exposure. It is unlikely that the annual means of PM10 concentrations would have changed significantly between 1991 and 1993. In fact, average TSP concentrations were stable over this 3-yr period.
For ozone exposure, three measures were used: annual average, summer daytime average (defined as the average ozone concentration between 10:00 a.m. and 6:00 p.m. during the months of May to September), and an index of cumulative exposure to ozone concentrations above 120 μg/m3 (“excess ozone,” defined as the sum of all half-hourly values of ozone concentration minus 120 μg/m3 for those half hours when the concentration exceeded 120 μg/m3).
Descriptive statistics for the environmental data are shown in Table 3. The annual mean concentrations of TSP, PM10, SO2, and NO2 were highly correlated, especially in the case of TSP, PM10, and SO2. All mean concentrations were inversely correlated with altitude, except in the case of ozone. The annual means of particulates and NO2 were negatively correlated with the annual mean concentrations of ozone; however, they were positively correlated with excess ozone. The coefficients of correlation for summer daytime ozone concentrations with particulates were positive but weak (0.11 to 0.40).
| Range Across Study Sites | Annual Mean | SD | Correlation Coefficients (r) | |||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| TSP | PM10 | SO2 | NO2 | O3 | Summer Daytime O3 | Excess O3 | Altitude | |||||||||||||||
| TSP, μg/m3 | 14.6–53.2 | 37.3 | 11.2 | 1 | 0.95 | 0.78 | 0.90 | −0.71 | 0.11 | 0.47 | −0.88 | |||||||||||
| PM10, μg/m3, 1993 | 10.1–33.4 | 21.2 | 7.4 | 1 | 0.93 | 0.91 | −0.55 | 0.31 | 0.67 | −0.77 | ||||||||||||
| SO2, μg/m3 | 2.5–25.5 | 11.7 | 7.1 | 1 | 0.86 | −0.39 | 0.40 | 0.73 | −0.55 | |||||||||||||
| NO2, μg/m3 | 9.2–57.7 | 35.6 | 16.0 | 1 | −0.78 | −0.02 | 0.36 | −0.73 | ||||||||||||||
| O3, μg/m3 | 31.5–55.2 | 43.1 | 9.5 | 1 | 0.54 | 0.21 | 0.64 | |||||||||||||||
| Summer daytime O3, μg/m3 | 79.2–118.2 | 92.0 | 12.9 | 1 | 0.89 | −0.15 | ||||||||||||||||
| Excess O3, μg year/m3 * | 0.015–4.72 | 1.21 | 1.47 | 1 | −0.39 | |||||||||||||||||
| Altitude, m | 278–1,558 | 581 | 425 | 1 | ||||||||||||||||||
Besides annual mean concentrations, we also calculated other statistical long-term parameters, reflecting repeated peak exposures (e.g., 95th percentile, numbers of hours above a certain level). The relative ranking of the eight sites for these parameters differed little for any pollutant except ozone.
With the exceptions of Wald and Montana, for which data only from 1991 were available, concentrations of air pollutants were available for the previous years. In comparing the years 1988 to 1990, a slightly decreasing trend was found for SO2. For NO2, TSP, and O3, the study year 1991 is representative of the concentrations of the previous years. Because there was no major change in emission sources in Wald and Montana, we also expect the measures in these two areas to be representative of long-term exposure in both areas.
Given the difficulties in clearly defining symptom complexes for respiratory diseases, especially for asthma, analyses in the study were primarily performed for individual symptoms. Respiratory symptoms were classified as bronchitic, asthmatic, and nonspecific cardiopulmonary (for definitions see Appendix).
Although it seems plausible that phlegm has to be brought up by coughing, more people reported phlegm production than cough (Table 4). Because of this inconsistency, the presence of either of these symptoms was also analyzed as an outcome measure. Attempting to be as specific as possible with regard to asthmatic symptoms, we chose “wheezing without cold” as an outcome measure, rather than any wheezing. Breathlessness during the day and at night was analyzed separately for each period, as well as being combined.
| Never-Smokers (n = 4,229 ) | Former Smokers (n = 2,175 ) | Current Smokers (n = 3,232 ) | ||||
|---|---|---|---|---|---|---|
| Chronic cough | 3.3 (2.8–3.8) | 3.0 (2.3–3.7) | 9.2 (8.2–10.2) | |||
| Chronic phlegm | 4.9 (4.2–5.5) | 5.5 (4.5–6.5) | 11.2 (10.1–12.3) | |||
| Chronic cough or phlegm | 7.0 (6.2–7.8) | 7.5 (6.3–8.6) | 16.7 (15.4–18.0) | |||
| Wheezing without colds | 4.8 (4.2–5.4) | 6.6 (5.5–7.6) | 12.2 (11.1–13.3) | |||
| Breathlessness during the day | 5.6 (4.9–6.3) | 7.2 (6.1–8.3) | 6.0 (5.2–6.8) | |||
| Breathlessness at night | 5.2 (4.6–5.9) | 5.7 (4.7–6.7) | 4.3 (3.6–5.0) | |||
| Breathlessness, day or night | 8.3 (7.5–9.2) | 10.7 (9.4–12.0) | 8.5 (7.6–9.5) | |||
| Current asthma | 3.2 (2.7–3.7) | 3.7 (2.9–4.5) | 1.9 (1.4–2.4) | |||
| Dyspnea on exertion | 23.5 (22.2–24.8) | 23.5 (21.7–25.3) | 31.5 (29.9–33.1) | |||
| Chest tightness | 13.4 (12.4–14.5) | 16.1 (14.5–17.6) | 14.4 (13.2–15.6) |
Logistic regression analysis was used to study associations between local prevalences of respiratory symptoms and local levels of air pollution. In a first step, analyses were restricted to lifetime nonsmokers to remove confounding by smoking.
For each of the respiratory symptom prevalences considered, the appropriate form of the logistic regression model in terms of basic individual and geographic predictors was established in a first step. The resulting preliminary logistic regression models contained the variables of age and body mass index (BMI), along with dummy variables for the study areas, gender (female sex), parental asthma, parental atopy, low level of education (lowest regular school level in Switzerland), and foreign citizenship. A dummy variable for high level of education was also tested, but proved to be insignificant for all the respiratory symptom prevalences considered. None of the models was improved by introducing a quadratic age or BMI term. However, in some of the models, terms for interaction between gender and age or gender and BMI were statistically significant. Therefore, such interaction terms were included in the basic model if they were marginally significant (i.e., if p < 0.1).
The air pollutants were then studied one by one, with control applied for all variables in the basic model other than the area indicators, (i.e., the dummy variables for the study areas were replaced, alternatively, by the annual area means of PM10, NO2, TSP, O3, and excess O3).
Because associations between symptom prevalences and average ambient air pollution levels might be confounded by other environmental exposures varying across study areas, a first sensitivity analysis included the following additional exposure variables: environmental tobacco smoke exposure (ETS) in the prior 12 mo, ETS exposure at work, gas cooking, paternal as well as maternal smoking during childhood, and former as well as current occupational exposure to dust, gas, vapor, fumes, or aerosols. A second sensitivity analysis included atopy (defined as a positive skin test) and severe respiratory infection before the age of 5 yr. Furthermore, analyses were restricted to long-term residents (subjects with at least 10 yr of residency in their area), to subjects without physician-diagnosed asthma, and to the study areas without an alpine climate (high altitude and long winter seasons). Given the high correlation between the different indicators of ambient air pollution, and since PM10 levels vary less within study areas than do other pollutants (23), these sensitivity analyses were done only for PM10.
In a further step, the analysis described here was also performed for current and former smokers. In these models, pack-years, daily number of cigarettes, and years since quitting were additionally controlled for.
Additionally, models including a random area effect were computed. These models took into account that regression residuals from the same area tend to be correlated because of the clustering of some uncontrolled causal factors associated with area. When this occurs, the variability of regression estimates is underestimated when considering residuals as independent. To adjust standard errors of regression estimates for this type of intercorrelation of residuals, a two-stage approach was used (27, 28). At first, fixed area effects and their covariance matrix, Σ, were estimated, and then these area effect estimates were regressed against the area means of air pollutant concentrations. In the second regression model, involving only one data point per area, the covariance matrix of residuals was assumed to be of the form Σ + σ2 I (where σ2 stands for the variance of the random area effect accounting for the cross-correlation of residuals within areas, and I denotes the identity matrix).
Raw prevalences of reported respiratory symptoms and 95% confidence intervals (CIs) are given by smoking status in Table 4. Current smokers reported higher prevalences of bronchitic symptoms, wheezing without cold, and dyspnea on ex– ertion than did never-smokers, whereas the rate of current asthma was lowest in current smokers.
The estimated changes in the prevalence of respiratory conditions associated with an increase of 10 μg/m3 in the annual mean concentration of air pollutants are shown in Table 5 for never-smokers. These estimates are adjusted for the individual risk factors according to the basic model, as indicated in the Methods section. Within the range of exposures given in Table 3, an increase of 10 μg/m3 in pollutant levels was associated with an increase in the prevalence of chronic phlegm (NO2, PM10), chronic cough or phlegm (NO2, PM10), breathlessness during the day (NO2, TSP, PM10), breathlessness during the day or at night (NO2, TSP, PM10), and dyspnea on exertion (NO2, TSP, PM10). An association was also observed with the prevalence of chronic cough or phlegm and with breathlessness occuring at rest during the day or at night. For chronic cough alone, the association was not significant. No association was seen for wheezing without cold, breathlessness at night, or chest tightness.
| NO2 | TSP | PM10 | O3 | Excess O3 † | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Chronic cough | 7.0 (−4.3–19.6) | 6.2 (−9.6–24.7) | 11.7 (−11.9–41.5) | −12.6 (−26.9–4.4) | 1.0 (−10.2–13.6) | |||||
| Chronic phlegm | 14.0 (3.7–25.3) | 13.9 (−0.7–30.7) | 35.2 (11.0–64.7) | −4.5 (−17.7–10.8) | 11.2 (1.7–21.5) | |||||
| Chronic cough or phlegm | 11.0 (2.6–20.1) | 12.1 (−0.7–25.7) | 27.4 (7.9–50.5) | −7.9 (−18.6–4.5) | 8.1 (−0.1–16.8) | |||||
| Wheezing without colds | 1.6 (−7.6–11.5) | −6.6 (−17.9–6.2) | −3.2 (−20.8–18.1) | −0.3 (−14.4–16.1) | −1.4 (−10.7–9.0) | |||||
| Breathlessness during the day | 18.7 (8.4–29.9) | 22.7 (7.3–40.2) | 47.2 (22.4–77.2) | −7.7 (−19.9–6.5) | 15.9 (7.0–25.4) | |||||
| Breathlessness at night | 6.1 (−3.0–16.2) | 4.8 (−7.9–19.3) | 10.8 (−8.4–34.0) | −4.1 (−17.1–11.1) | 4.9 (−4.1–14.8) | |||||
| Breathlessness, day or night | 12.8 (4.8–21.4) | 16.4 (4.4–29.6) | 32.8 (13.9–54.8) | −6.5 (−16.9–5.2) | 12.9 (5.4–20.9) | |||||
| Current asthma | −3.7 (−14.0–7.6) | −16.8 (−28.5– −3.1) | −21.6 (−38.9–0.5) | −3.0 (−19.4–16.6) | −15.9 (−27.8– −2.1) | |||||
| Dyspnea on exertion | 8.5 (3.2–14.1) | 12.7 (4.8–21.2) | 31.6 (18.2–46.4) | −3.7 (−11.3–4.6) | 11.8 (6.3–17.7) | |||||
| Chest tightness | −2.1 (−7.7–3.8) | −6.8 (−14.0–1.1) | −8.0 (−18.8–4.3) | 1 (−8.2–11.1) | −4.3 (−10.3–2.0) |
Current asthma was negatively associated with TSP.
For ozone, no association was seen of the annual mean concentrations with the prevalences of symptoms. However, excess ozone was positively associated with chronic phlegm, breathlessness during the day, breathlessness during the day or at night, and dyspnea on exertion. A negative association was seen with current asthma.
Table 6 shows the odd ratios (ORs) for the adjusted prevalences of the respiratory symptoms and conditions for a 10 μg/m3 increase in the annual mean concentrations of PM10. Figures 1A to 1E display the prevalences of respiratory symptoms versus the average annual concentrations of PM10 for chronic cough or chronic phlegm, breathlessness during the day, breathlessness occurring during the day or at night, wheezing without cold, and dyspnea on exertion, respectively. Within the range of 10.1 to 33.4 μg/m3 of PM10 (Table 3), the OR for an increase of 10 μg/m3 in the annual mean was 1.35 (CI: 1.11 to 1.65) for chronic phlegm among never-smokers, 1.27(CI: 1.08 to 1.50) for chronic cough or phlegm, 1.48 (CI: 1.23 to 1.78) for breathlessness during the day, 1.33 (CI: 1.14 to 1.55) for breathlessness during the day or at night, and 1.32 (CI: 1.18 to 1.46) for dyspnea on exertion.
| Never-Smokers†OR (95% CI) | Former Smokers‡OR (95% CI) | Current Smokers§OR (95% CI) | ||||
|---|---|---|---|---|---|---|
| Chronic cough | 1.11 (0.88–1.41) | 1.08 (0.76–1.53) | 0.85 (0.71–1.01) | |||
| Chronic phlegm | 1.35 (1.11–1.65) | 0.95 (0.73–1.23) | 1.04 (0.89–1.23) | |||
| Chronic cough or phlegm | 1.27 (1.08–1.50) | 1.00 (0.80–1.23) | 1.00 (0.87–1.15) | |||
| Wheezing without colds | 0.97 (0.79–1.18) | 0.99 (0.77–1.23) | 1.17 (1.00–1.36) | |||
| Breathlessness during the day | 1.48 (1.23–1.78) | 1.53 (1.21–1.93) | 1.47 (1.19–1.81) | |||
| Breathlessness at night | 1.11 (0.92–1.35) | 0.97 (0.75–1.26) | 0.995 (0.78–1.27) | |||
| Breathlessness, day or night | 1.33 (1.14–1.55) | 1.28 (1.06–1.56) | 1.30 (1.09–1.56) | |||
| Current asthma | 0.78 (0.61–1.01) | 0.85 (0.61–1.17) | 1.19 (0.83–1.70) | |||
| Dyspnea on exertion | 1.32 (1.18–1.46) | 1.32 (1.13–1.53) | 1.22 (1.10–1.37) | |||
| Chest tightness | 0.92 (0.81–1.04) | 0.87 (0.73–1.02) | 0.93 (0.81–1.06) |
Fig. 1. Annual mean level of PM10 and adjusted prevalences of chronic cough or phlegm (A), breathlessness during the day (B), breathlessness, day or night (C ), wheezing without cold (D), and dyspnea on exertion (E ). For exact definition of symptoms see Appendix.
When running the models for former and current smokers, the associations of the annual means of PM10 with breathlessness during the day, breathlessness during the day or at night, and dyspnea on exertion remained significant (Table 6). For current smokers, an association was seen for wheezing without cold.
To evaluate whether further exposures confounded the association between long-term exposure to air pollution and respiratory conditions, we repeated the analyses, including environmental tobacco smoke exposure, past and current workplace exposures, maternal and paternal smoking, and gas cooking in the model (Table 7, Model 2). The associations decreased a little, but remained stable. Nor did controlling for atopy and early childhood respiratory infections (Table 7, Model 3) alter the results. Including all potential confounders from Models 2 and 3 in the same model yielded similar results (Table 7, Model 4). The two least polluted areas in our study, Montana and Davos, have the highest altitude (Davos at 1,560 m above sea level and Montana at 1,531 m, whereas the range of altitudes of the other study sites is 277 to 616 m above sea level) and the longest winter seasons. To see whether altitude or duration of the winter season might confound the association with symptom prevalence, we performed the analyses without the Montana or Davos areas. The picture remained the same; the associations with breathlessness during the day, dyspnea on exertion, and chronic cough or phlegm production were stronger (data not shown).
| Model 1 Basic Model | Model 2 OR (95% CI) | Model 3 OR (95% CI) | Model 4 OR (95% CI) | Model 5 OR (95% CI) | Model 6 OR (95% CI) | Model 7 OR (95% CI) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Chronic cough | 1.11 (0.88–1.41) | 1.07 (0.82–1.40) | 1.13 (0.88–1.43) | 1.07 (0.81–1.41) | 1.10 (0.84–1.43) | 1.19 (0.91–1.55) | 1.11 (0.86–1.43) | |||||||
| Chronic phlegm | 1.35 (1.11–1.65) | 1.30 (1.05–1.62) | 1.35 (1.10–1.65) | 1.30 (1.04–1.63) | 1.41 (1.13–1.75) | 1.43 (1.15–1.77) | 1.28 (0.94–1.76) | |||||||
| Chronic cough or phlegm | 1.27 (1.08–1.50) | 1.22 (1.01–1.47) | 1.26 (1.07–1.50) | 1.21 (0.99–1.46) | 1.27 (1.06–1.53) | 1.32 (1.10–1.58) | 1.24 (0.998–1.54) | |||||||
| Wheezing without colds | 0.97 (0.79–1.18) | 0.88 (0.70–1.11) | 0.99 (0.81–1.22) | 0.93 (0.74–1.18) | 0.99 (0.79–1.23) | 1.07 (0.83–1.39) | 0.95 (0.76–1.20) | |||||||
| Breathlessness during the day | 1.48 (1.23–1.78) | 1.36 (1.10–1.68) | 1.53 (1.26–1.86) | 1.41 (1.13–1.76) | 1.50 (1.23–1.84) | 1.73 (1.39–2.15) | 1.44 (1.16–1.78) | |||||||
| Breathlessness at night | 1.11 (0.92–1.35) | 1.14 (0.92–1.42) | 1.12 (0.92–1.36) | 1.14 (0.91–1.42) | 1.11 (0.90–1.36) | 1.27 (1.01–1.59) | 1.09 (0.89–1.34) | |||||||
| Breathlessness, day or night | 1.33 (1.14–1.55) | 1.30 (1.10–1.55) | 1.35 (1.15–1.58) | 1.31 (1.10–1.57) | 1.33 (1.12–1.57) | 1.52 (1.27–1.80) | 1.31 (1.10–1.57) | |||||||
| Current asthma | 0.78 (0.61–1.01) | 0.81 (0.61–1.06) | 0.78 (0.60–1.01) | 0.82 (0.62–1.10) | 0.82 (0.61–1.10) | — | 0.77 (0.55–1.07) | |||||||
| Dyspnea on exertion | 1.32 (1.18–1.46) | 1.24 (1.10–1.40) | 1.32 (1.18–1.47) | 1.23 (1.09–1.39) | 1.32 (1.17–1.48) | 1.40 (1.25–1.57) | 1.26 (0.87–1.82) | |||||||
| Chest tightness | 0.92 (0.81–1.04) | 0.87 (0.76–1.01) | 0.92 (0.81–1.05) | 0.87 (0.75–1.01) | 0.88 (0.76–1.01) | 0.96 (0.84–1.10) | 0.92 (0.81–1.04) |
To examine associations in long-term residents, a separate analysis was restricted to subjects with at least 10 yr of residency in the respective areas (Table 7, Model 5). The overall pattern did not change, but the associations with breathlessness during the day and with chronic phlegm production were stronger.
Subjects affected by respiratory diseases, and especially those with diagnosed asthma, may tend to move to less polluted areas. Therefore, we tested whether excluding subjects with physician-diagnosed asthma would affect the associations (Table 7, Model 6). Again, the pattern remained unchanged; the associations with breathlessness during the day, dyspnea on exertion, and chronic phlegm production were stronger.
Because subjects living in the same areas may be more alike for some uncontrolled covariates, a random area-effect model was run. The effect size estimates showed little change; the CIs became wider, as was expected. The associations with chronic phlegm production and with dyspnea on exertion were no longer significant (Table 7, Model 7).
In the SAPALDIA cross-sectional study, average annual ambient concentrations of NO2, TSP, and PM10 were positively associated with adjusted 1-yr-prevalences of chronic phlegm production, chronic cough or chronic phlegm production, breathlessness during the day, breathlessness during the day or at night, and dyspnea on exertion in never-smokers. However, no associations were seen with asthma or asthma symptoms. In both current and former smokers the association between PM10 and chronic phlegm production was not significant. In former smokers, however, an association was seen between PM10 and wheezing apart from colds. No associations were found with the annual mean concentration of O3.
The predicted effect of a 10 μg/m3 increase in annual mean concentrations of PM10 was substantial: an increase of 35% in the prevalence of chronic phlegm production, of 27% in the prevalence of chronic cough or phlegm production, of almost 50% for breathlessness during the day, of 33% for breathlessness during the day or at night, and of 32% for dyspnea on exertion.
Given the nonparticipation rate of 40% in the study, there was a potential for selection bias. The relatively high differences in participation between the study centers might have been due to the different sizes, social structures, and cultural backgrounds of the study areas, as well as to the differing promotional strategies applied locally. To quantify the potential for selection bias, additional information was gained from nonparticipants by telephone interviews during the first contact and through a mailed, short questionnaire. As shown in Table 1, participants were more likely to report wheezing without cold and allergic rhinitis, but less likely to report phlegm production. In a logistic regression model in which the influences of age, sex, and smoking status were controlled and a center–participation interaction term was included the strength and direction of the association was compared between centers (24). The effects of participation did not vary significantly for asthma, wheezing, or phlegm production. Therefore, it seems unlikely that selection bias was responsible for the associations with the symptoms considered in this study.
Major efforts were made to minimize confounding. Known individual cofactors for respiratory health were assessed and controlled in the analyses. Restricting the analyses to never-smokers removed confounding by smoking. In a series of sensitivity analyses (Table 7), the observed associations remained stable. The exposure measures used did not include individual factors that influence personal exposure, such as time spent outdoors. In fact, there is large variability in time–activity patterns between subjects and over time (33). Our results, however, would not be biased by this variability unless the area mean time–activity patterns were correlated with long-term air pollution. Given the tendency of residents to spend less time outdoors in more urban SAPALDIA areas, our results may underestimate rather than overestimate the impact of ambient air pollutants with limited indoor penetration. Because fine particulate pollution efficiently penetrates indoors, the lack of time–activity data may be of limited importance in our study. Other factors affecting respiratory symptoms may vary across geographic areas and correlate with area mean exposure, and may therefore confound this association. Given the extensive control for individual cofactors and the consideration of a random area-effect model, it is difficult to envision further uncontrolled and strong confounders to explain our results. However, we cannot completely exclude this possibility.
Some of the SAPALDIA centers were known for their clean air or as health resorts, and physicians in Switzerland and abroad tend to counsel asthma patients to choose a dry alpine climate. If persons affected by respiratory diseases (such as asthma), move to less polluted areas, an overrepresentation of such persons may interfere with the association of air pollution and respiratory conditions. Such selection would lead to an underestimation of the studied associations. Participants in the present study were asked whether they had moved into their areas, and if so, the reasons why they moved there. Less than 1% of the participants in all centers had moved to the respective area for health reasons, with the exception of Davos, where this was the case for 5.7% (34 of 596 participants) (24). In fact, the group of people who had moved to Davos because of health problems represented 44% (24 of 54) of the participants reporting physician-diagnosed asthma in this study area, and 14 of the 18 participants reporting current asthma symptoms. This selection explains the remarkably high prevalence of wheezing without cold in Davos (Figure 1D). The negative associations between current asthma and TSP (PM10), and between current asthma and excess ozone (Table 5), were actually also due to this single area (figure not shown). When excluding subjects from the analysis who moved to their area because of respiratory diseases, the predicted changes in the prevalence of current asthma were no more significant (−8.4 [−22.2 to 7.8] per 10 μg/m3 increase in the annual mean of TSP, and −12.7 [−32.6 to 13.1] for PM10). No such selection could be observed for bronchitis.
Although direct comparisons with other findings are limited because of different definitions of respiratory symptoms and conditions, and differences in exposure assessment, our results are in line with those in studies conducted in different geographic regions of the United States and Europe: in an early study, long-term exposure to TSP and SO2 was related to bronchitis (2). Later, SO2 was found to be associated with respiratory symptoms in Pennsylvania (4), with the development of chronic bronchitis in California (5), and with chronic respiratory conditions and chronic obstructive pulmonary disease (COPD) in the U.S. Health Interview Survey (6). Exposure to TSP was also found to be related to self-reported chronic bronchitis as well as to physician-diagnosed respiratory disease (but not with dyspnea) in the First National Health and Nutrition Education Survey (NHANES I) (7). In the Seventh-day Adventist cohort study in California, TSP was shown to be associated with the development of COPD, chronic bronchitis, and asthma (9), indirect estimates of PM10 were shown to be associated with the development of symptoms of overall airway obstructive disease, chronic productive cough, increased severity of airway obstructive disease and asthma (11), and retrospectively estimated concentrations of PM2.5 were shown to be associated with increasing severity of respiratory diseases (10). In Europe, long-term exposure to black smoke was related to phlegm production in a British study (12), exposure to NO2 was found to be associated with cough, and nasal and throat irritations in Sweden (8), and exposure to SO2 was found to be associated with cough, nasal irritation, and phlegm production in Sweden (8), and with the prevalence of lower respiratory symptoms in France (3).
It is notable that an increased risk for wheezing and asthma but not for cough and phlegm production has been found in current smokers, whereas for never-smokers an increased risk has been observed for cough and phlegm production but not for asthma or asthma symptoms. This may indicate an effect modification by smoking status. Effects of exposure to air pollution regarding bronchitis symptoms may be hidden among smokers, whereas such exposure may make them more susceptible to asthma symptoms.
No clear-cut picture emerged from the attempt to classify the symptoms in our study as bronchitic, asthmatic, or nonspecific cardiopulmonary symptoms (see Appendix): with regard to bronchitic symptoms, associations were seen with chronic phlegm production and with the combination of chronic phlegm production or chronic cough, yet not with chronic cough alone. Among the symptoms classified as asthmatic, associations were found with those related to breathlessness but not with wheezing without cold or with current asthma. The symptom of breathlessness during the day, which was associated with air quality measures, was classified as asthmatic, but may be considered less specific than wheezing, and may also be considered as related to bronchitic symptoms. Notably, the associations of PM10 with phlegm and breathlessness clearly increased when subjects with asthma were excluded from the analysis (Table 7, Model 6). This is what one would expect if air pollution enhances bronchitic diseases without having an impact on asthma. Understanding of the meaning and classification of respiratory symptoms—not only of wheezing, but also of breathlessness and dyspnea—awaits further efforts. Thus, as in some studies (7, 29, 30), but in contrast to others (6, 9, 10), our findings give no evidence for an association of the prevalence of asthma (physician-diagnosed asthma) or of key symptoms of asthma with the level of long-term air pollution.
Among the symptoms classified as nonspecific cardiopulmonary symptoms, we found significant associations of air pollution with dyspnea on exertion but not with chest tightness. This finding is in line with both short-term and long-term effects of air pollution on respiratory and cardiovascular diseases. In the United States, short-term changes in air pollution levels were found to be associated with an increase in cardiovascular mortality and hospital admissions (20, 21). In cohort studies, air pollution was also related to long-term mortality from cardiopulmonary causes (16, 19).
The analysis reported here is based on the SAPALDIA cross-sectional study. The cross-sectional design is commonly considered to be of limited value for assessing causal relationships, owing to the contemporaneous assessment of health and exposure factors. This criticism, however, cannot be generalized and does not apply for the semiindividual cross-sectional design (31), in which exposure information is neither asked of nor modified by subjects, but is instead defined through residential history. In fact, the impact of objective environmental conditions on long-term chronic health outcomes is efficiently addressed with such cross-sectional approaches (32). SAPALDIA participants had been living for at least 3 yr—and often much longer—in the local areas involved in the study. The measures of the cross-sectional year were representative of measures for previous years. In an analysis restricted to subjects who had resided for at least 10 yr in their areas, the associations were indeed slightly stronger than in the overall study population.
The annual average concentrations of air pollutants in the year 1991 were taken as indicators of long-term air-pollution exposure. Other long-term statistical parameters, reflecting peak exposures (e.g., 95th percentiles, numbers of hours above a certain pollutant level) had also been calculated. The relative ranking of the eight sites in the study differs little for any of the pollutants except for ozone.
The measurements in 1991 were representative for the years 1988 to 1990 for NO2, TSP, and O3 for all study sites, where these data are known (with the exceptions of Wald and Montana, for which only data for 1991 are available, but where no major change in emission sources has occurred). For NO2, measurements are available from 1986 on, and the annual means are highly correlated (R = 0,996) in five study sites (23). For SO2, a slightly decreasing trend was found in the documented period. The differences between the study sites remain similar. We feel therefore that the measured exposures can reliably represent exposures dating back to the late 1980s.
The induction time for long-term effects of air pollutants on respiratory health have not yet been established. From what we have documented with regard to past exposure, and given the population inclusion criterion of 3 yr of residency in the respective study sites, our findings are compatible with a time frame of at least 3 yr. However, we can certainly not exclude effects from exposures dating back to far earlier periods. In fact, the results of some of our analyses are compatible with longer-term effects or with effects that might result from some cumulative exposure over time: The analysis that was restricted to residents living in their respective areas for at least 10 yr yielded stronger associations (see Model 5 in Table 7). In an analysis of air pollution and lung function (23), the association between FVC and average NO2 exposure within study communities was stronger in the older half of the sample (people over 40 yr of age, [i.e., born between 1931 and 1950]).
The small number of areas in our study, the strong intercorrelation of air pollutants, and the similar ranking of the study areas with regard to different air pollution measures make it difficult to relate the associations found in the study to a single pollutant. Although the associations were strongest with PM10, the single indicators stand for the mixture of outdoor pollutants rather than for one specific component.
In our study, mean ozone concentrations were negatively associated with those of other air pollutants. The inconsistent results reported for ozone should be interpreted with caution. SAPALDIA does not allow an assessment of long-term effects of ambient ozone levels, since the following points make it inherently difficult to address potential effects: (1) Although the range of excess ozone appears to have been quite wide, we basically had a two-area comparison, with high excess ozone exposure in Lugano and rather low and similar values in the other regions. Thus, confounding due to area-specific covariates could not have been sufficiently controlled. (2) In contrast to particulate pollution, personal ozone exposure is strongly determined by personal factors such as time–activity patterns or indoor air ventilation systems. Therefore, the exposure variability within a given area is very high (33), limiting the power of our study. (3) The observed range of mean ozone exposures across the eight study areas was small, whereas large between-area exposure variability is a prerequisite for efficiently assessing the effects of exposure. (4) Ozone mean concentrations were negatively associated with other air pollutants. The resulting strong confounding cannot be adequately addressed in this analysis. Moreover, the negative association that was seen between excess ozone and current asthma (as well as that between TSP and current asthma) seems to have been due to a selection of asthmatic subjects moving to one alpine site (Davos; Figure 1D).
We conclude that long-term exposure to air pollution even at a low level is associated with higher prevalences of respiratory symptoms, especially of breathlessness, and of symptom combinations probably related to chronic bronchitis. Thus, in the SAPALDIA study, two components of environmental exposures—namely environmental tobacco smoke (34) and air pollution—were found to be unfavorably associated with respiratory health.
The reported associations, together with the results of studies performed in the United States and in European regions and having various objective and reported respiratory endpoints, add to the evidence that long-term exposure to air pollution is causally associated with respiratory illness in adults.
1. Chronic cough. Positive answer to the question “Do you usually cough during the day, or at night, on most days for as much as three months each year?,” and an answer of ⩾ 2 to the question “For how many years have you coughed like this?”
2. Chronic phlegm. Positive answer to the question “Do you usually bring up any phlegm from your chest during the day, or at night, on most days for as much as 3 months each year?,” and an answer of ⩾ 2 to the question “For how many years have you brought up phlegm like this?”
3. Chronic cough or phlegm. Chronic cough (1.) or chronic phlegm (2.).
4. Wheezing without cold. Positive answer to both of the following questions: “Have you had wheezing or whistling in your chest at any time in the last 12 months?” “Have you had this wheezing or whistling when you did not have a cold?”
5. Breathlessness during the day. “Have you had an attack of shortness of breath that came on during the day when you were at rest at any time in the last 12 months?”
6. Breathlessness at night. “Have you been awakened by an attack of shortness of breath at any time in the last 12 months?”
7. Breathlessness, day or night. Breathless during the day (5.) or at night (6.)
8. Current asthma. Positive answer to both “Have you ever had asthma?” and “Was this confirmed by a doctor?,” and a positive answer to at least one of the following questions: “Are you currently taking any medicine for asthma?,” and “Have you had an attack of asthma in the last 12 months?”
9. Dyspnea on exertion. “Are you troubled by shortness of breath when hurrying on level ground or walking up a slight hill?”
10. Chest tightness. “Have you woken with a feeling of tightness in your chest at any time in the last 12 months?”
SAPALDIA Team: Study Director: P. Leuenberger; Programme Director: U. Ackermann-Liebrich; P. Alean, K. Blaser, G. Bolognini, J. P. Bongard, O. Brändli, P. Braun, C. Bron, M. Brutsche, C. Defila, G. Domenighetti, S. Elsasser, L. Grize, P. Guldimann, P. Hufschmid, W. Karrer, H. Keller-Wossidlo, R. Keller, N. Künzli, J. C. Luthy, B. W. Martin, T. C. Medici, C. Monn, A. G. Peeters, A. P. Perruchoud, A. Radaelli, C. Schindler, J. Schwartz, G. Solari, M. H. Schöni, J. M. Tschopp, B. Villiger, B. Wüthrich, J. P. Zellweger, and E. Zemp.
The authors acknowledge the help of many people who made the study possible. Apart from the support of the Swiss National Science Foundation and the Federal Office for Science and Education, logistic and financial support was received from cantonal governments (BS, GE, GR, VD, VS, TI, ZH, and VS) and from the cantonal offices for air hygiene measurements. The Federal Office for Forest and Environment supported the quality control of air hygiene measurements and the PM10 measurements. Further financial support was received from the Swiss Society of Pulmonology, the Lega Ticinese contro la tubercolosi e le malattie polmonari, and various cantonal offices of public health. The study could not have been done without the help of the field workers of the local medical teams: in Aarau, C. Persoz-Borer, C. Wettstein, G. Giger, H. Grob-Stalder, J. Lohmüller, K. Häfeli, and U. Rippstein; in Basel, V. Fluri, M. Herrous, G. Imboden, and L. Joos; in Davos, K. D'Alberti and A. Sönnichsen; in Genf, I. Barbey, K. Gegerle, and N. Penay; in Lugano, M. Astone, E. Haechler, E. Riesen, and B. Viscardi; in Montana, Dr. Ch. Hollenstein, E. Borgeat, and I. Clivaz; in Payerne, S. Menétrey-Jaques, C. Gilomen-Pages, and M.C. Collaud; and in Wald, B. Salzmann, V. Kienast, H. Astone, V. Keller, and C. Schwalm. Throughout the study the investigators received much appreciated advice from Prof. Frank Speizer of the Harvard Medical School, for which they are very grateful, and they also thank the persons who agreed to participate in the study. The authors also thank Renata Sandroni for secretarial assistance.
| 1. | Committee of the Environmental and Occupational Health Assembly of the American Thoracic SocietyHealth effects of outdoor air pollution. Part 1: Am. J. Respir. Crit. Care Med. 153:3–50; and Part 2: Am. J. Respir. Crit. Care Med.1531996477498 |
| 2. | Holland W. W., Reid D. D.The urban factor in chronic bronchitis. Lancet11965445448 |
| 3. | PAARC: Groupe Cooperative /Lelouche JPollution atmosphérique et affections respiratoires chroniques ou à répétition. Bull. Eur. Physiopathol. Respir.18198287116 |
| 4. | Schenker M. B., Samet J. M., Speizer F. E., Gruhl J., Batterman S.Health effects of air pollution due to coal combustion in the Chestnut Ridge region of Pennsylvania: results of cross-sectional analysis in adults. Arch. Environ. Health381983325330 |
| 5. | Euler G. L., Abbey D. E., Magie A. R., Hodlkin J. E.Chronic obstructive pulmonary disease symptom effects of long term cumulative exposure to ambient levels of total suspended particulates and sulfur dioxide in California Seventh-Day Adventist residents. Arch. Environ. Health421987213222 |
| 6. | Portney P., Mullahy J.Urban air quality and respiratory disease. Reg. Sci. Urban Econ.201990407418 |
| 7. | Schwartz J.Particulate air pollution and chronic respiratory disease. Environ. Res.621993713 |
| 8. | Forsberg, B., N. Stjernberg, and S. Wall. 1997. Prevalence of respiratory and hyperreactivity symptoms in relation to levels of criteria air pollutants in Sweden. Eur. J. Public Health 7/3:291–296. |
| 9. | Abbey D. E., Lebowitz M. D., Mills P. K., Petersen F. F., Beeson W. L., Burchette R. J.Long-term ambient concentrations of particulates and oxidants and development of chronic disease in a cohort of nonsmoking California residents. Inhal. Toxicol.119952134 |
| 10. | Abbey, D. E., B. E. Ostro, F. Petersen, and R. J. Burchette. 1995. Chronic respiratory symptoms associated with estimated long-term ambient concentrations of fine particulates less than 2.5 microns in aerodynamic diameter (PM2.5) and other air pollutants. J. Exp. Anal. Environ. Epidemiol. 5/2:137–159. |
| 11. | Abbey D. E., Hwang B. L., Burchette R. J.Estimated long term ambient concentrations of PM10 and development of respiratory symptoms in a nonsmoking population. Arch. Environ. Health501995139150 |
| 12. | Scarlett J. F., Griffiths J. M., Strachan D. P., Anderson H. R.Effect of ambient levels of smoke and sulphur dioxide on the health of a national sample of 23-year-old subjects in 1981. Thorax501995764768 |
| 13. | Burney P. G. J., Luczynska C., Chinn S., Jarvis D.The European Community Respiratory Health Survey. Eur. Respir. J.71994954960 |
| 14. | Schwartz J., Dockery D. W.Increased mortality in Philadelphia associated with daily air pollution concentrations. Am. Rev. Respir. Dis.1451992600604 |
| 15. | Spix C., Heinrich J., Dockery D., Schwartz J., Volksch G., Schwinkowski K., Collen C., Wichmann H. E.Air pollution and daily mortality in Erfurt, East Germany, 1980–1989. Environ. Health Perspect.1011993518526 |
| 16. | Dockery D., Pope A., Xu X., Spengler J. D., Ware J. D., Fay M. E., Ferris B. J., Speizer F. E.An association between air pollution and mortality in six U.S. cities. N. Engl. J. Med.329199317531759 |
| 17. | Touloumi G., Pocock S. J., Katsouyanni K., Trichopoulos D.Short-term effects of air pollution on daily mortality in Athens—a time-series analysis. Int. J. Epidemiol.231994957967 |
| 18. | Schwartz J.Air pollution and daily mortality: a review and meta-analysis. Environ. Res.6419943652 |
| 19. | Pope A., Thun M., Namboodiri M., Dockery H. D. W., Evans J. S., Speizer F. E., Heath C. W.Particulate air pollution as a predictor of mortality in a prospective study of U.S. adults. Am. J. Respir. Crit. Care Med.1511995669674 |
| 20. | Schwartz J., Morris R.Air pollution and hospital admissions for cardiovascular disease in Detroit, Michigan. Am. J. Epidemiol.14219952325 |
| 21. | Burnett R., Dales R., Krewski D., Vincent R., Dann T., Brook J.Associations between ambient particulate sulfate and admissions to Ontario Hospitals for cardiac and respiratory diseases. Am. J. Epidemiol.14219951522 |
| 22. | Ackermann-Liebrich U., Leuenberger P., Schwartz J., Schindler C., Monn C., Bolognini G., Bongard J. P., Brändli O., Domenighetti G., Elsasser S., Grize L., Karrer W., Keller R., Keller-Wossidlo H., Künzli N., Martin B. W., Medici T. C., Perruchoud A. P., Schöni M. H., Tschopp J. M., Villiger B., Wüthrich B., Zellweger J. P., Zemp E.the SAPALDIA TeamLung function and long term exposure to air pollutants in Switzerland. Am. J. Respir. Crit. Care Med.1551997122129 |
| 23. | Schindler C., Ackermann-Liebrich U., Leuenberger P., Monn C., Rapp R., Bolognini G., Bongard J. P., Brändli O., Domenighetti G., Karrer W., Keller R., Medici T. G., Perruchoud A. P., Schöni M. H., Tschopp J. M., Villiger B., Zellweger J. P.the SAPALDIA TeamAssociations between lung function and estimated average exposure to NO2 in 8 areas of Switzerland (SAPALDIA). Epidemiology91998405411 |
| 24. | Martin B. W., Ackermann-Liebrich U., Leuenberger P., Künzli N., Zemp E., Stutz, et al.SAPALDIA: Methods and participation in the cross-sectional part of the Swiss Study on Air Pollutants and Lung Diseases in Adults. Soz. Präventivmed.4219976784 |
| 25. | Wüthrich, B., C. Schindler, T. C. Medici, J. P. Zellweger, P. Leuenberger, and the SAPALDIA Team. 1996. IgE levels, atopy markers and hay fever in relation to age, sex and smoking status in a normal adult Swiss population. Int. Arch. Allergy Immunol. III:396–402. |
| 26. | Monn C., Braendli O., Schaeppi G., Schindler C., Ackermann-Liebrich U., Leuenberger P.Particulate matter < 10 micrometer (PM10) and total suspended particulates (TSP) in urban, rural and alpine regions in Switzerland. Atmospher. Environ.19199525652573 |
| 27. | Ware J. H., Ferris B. G., Dockery D. W., Spengler J. D., Stram D. O., Speizer F. E.Effects of ambient sulfur oxides and suspended particles on respiratory health of preadolescent children. Am. Rev. Respir. Dis.1331986824842 |
| 28. | Wong G. Y., Mason W. M.The hierarchical logistic regression model for multilevel analysis. J. Am. Stat. Assoc.801985513524 |
| 29. | Braun-Fahrländer C., Ackermann-Liebrich U., Schwartz J., Gnehm H. P., Rutishauser M., Wanner H. U.Air pollution and respiratory symptoms in preschool children. Am. Rev. Respir. Dis.14519924247 |
| 30. | Von Mutius E., Martinez F. D., Fritsch E., Nicolai Th., Roell G., Theimann H. H.Prevalence of asthma and atopy in two areas of west and east Germany. Am. J. Respir. Crit. Care Med.1491994358364 |
| 31. | Künzli N., Tager I. B.The Semi-individual Study in Air Pollution Epidemiology: a valid design as compared to ecologic studies. Environ. Health Perspect.10199710781083 |
| 32. | Dockery, D., and B. Brunekreef. 1996. Longitudinal studies of air pollution effects on lung function. Am. J. Respir. Crit. Care Med. 154 (Suppl. 6):S250–S256. |
| 33. | Künzli N., Ackermann-Liebrich U., Leuenberger P., Brändli O.SAPALDIA TeamBetween-area variation of time spent outdoors—a theoretical determinant of ambient air pollution exposure (abstract). Epidemiology61995S45 |
| 34. | Leuenberger P., Schwartz J., Ackermann-Liebrich U., et al.Passive smoking exposure in adults and chronic respiratory symptoms (SAPALDIA study). Am. J. Respir. Crit. Care Med.150199412221228 |
