Epidemiologisch onderzoek naar een beter begrip van de relatie tussen luchtverontreiniging en gezondheid.

Background In the past few decades, exposure to air pollution has been found to be associated with all-cause mortality, cardiovascular and respiratory morbidity, both in the short term (acute exposure) and the long term (chronic exposure). According to the most recent report by the World Health Organisation (WHO, 2014), 3.7 million deaths worldwide and per year are attributed to ambient air pollution, placing it in the top 10 of risk factors. Ambient air pollution consists of particulate matter (PM) and gases, such as NO2, SO2, and ozone. Of all pollutants, PM is most reliably associated with human disease. It is usually classified according to particle size, with PM10 (particles smaller than 10 µm) as the most commonly studied fraction. The European Union (EU) has set two limit values for PM10 concentrations: annual mean levels of PM10 must not exceed 40 μg/m³ (25 µg/m³ for PM2.5), and daily averages must not exceed 50 μg/m³ on more than 35 days/year, for any monitoring station in the EU member states. In contrast, the WHO advises that annual averages of PM10 levels should not exceed 20 μg/m³ (10 µg/m³ for PM2.5) and that daily averages should not exceed 50 μg/m³ on more than 3 day/year. Objectives The general objective of this PhD project was to gain more insight in the relationship between PM and human health in susceptible subgroups, such as infants, lung-transplanted patients and elderly. Also, in two of the four studies conducted within the scope of this thesis, I investigated the possible biological mechanisms involved in the pathway from PM inhalation to disease. Finally, I aimed to evaluate EU limit values and WHO guidelines for ambient PM concentrations, based on our study results. Main study results In a first study (Chapter 1), I investigated effects of daily variation in environmental PM10 on risk of infant mortality (<1y of age) in Flanders. 2382 infants died during the study period (1998-2006). The PM10 concentration averaged 31.9 µg/m³, and there were 321 days (an average of 35.7 days per year) with a mean daily concentration exceeding 50 µg/m³. This means that the EU air quality standards were met for the yearly average, but barely for the daily average. It is clear, however, that the more stringent WHO guideline values were not met at all in Flanders. Using a bidirectional time-stratified case-crossover (CCO) design, I found that PM10 was associated with infant mortality, especially in the late neonatal stage (i.e. between one week and one month of age, N=372). For each 10 µg/m³ increment of PM10, the risk of late neonatal mortality increased with 11% (95% CI 1-22%). On days with average PM10 levels exceeding the EU limit value of 50 µg/m³, the risk of mortality was 74% higher (95% CI 18-158%) than on days below that value. The current EU limit value for PM10 is not protective to prevent triggering infant mortality. Moreover, the linear shape of the association between exposure and mortality gives no evidence for a threshold or a save level. In a second study (Chapter 2), we investigated whether graft rejection after lung transplantation (LTx) could be linked to recent exposure to PM air pollution, and which underlying mechanisms are involved. In the period 2001-2011, transbronchial biopsies were repeatedly executed in 397 LTx recipients at the UZ Leuven. I linked estimated PM10 levels for each patient’s home address with symptoms of graft rejection and related physiological parameters. We found that a 10 µg/m³ increase in PM10 concentration 3 days before biopsy increased the risk of lymphocytic airway disease (LAD) with 12% (95% CI 1-25%). LAD is the pathological correlate of acute graft rejection after LTx. Additionally, PM10 exposure was positively associated with BAL counts of neutrophils and lymphocytes, indicating that inflammation plays a part in the physiological pathway. Preventive treatment with the antibiotic azithromycin appeared to block the effect of PM10 on acute graft rejection. In chapter 3, I investigated whether a decrease or increase of exposure to air pollution during several days (compared to a person’s ‘normal’ level) can already result in changes in biomarkers of cardiovascular health. During the course of one year, we measured air pollution exposure, carotid arterial stiffness (a well-established marker of cardiovascular disease), and other biomarkers in a panel of 20 retired and healthy men and women in different locations: during a 10-day stay in Milan (Italy, PM10 > 50 µg/m³), during a similar 10-day stay in Vindeln (rural area in northern Sweden, PM10 < 10 µg/m³) and at regular time points in Leuven (reference location, PM10 ≈ 30 µg/m³). Compared with Leuven, exposure to pollutants was higher in Milan and lower in Vindeln, with the highest contrast found for NO2 (averages: Milan 63.7 µg/m³; Vindeln 4.4 µg/m³).We found strong associations between 7-days exposure to air pollution and arterial stiffness, e.g. a 4.4% decrease in compliance (i.e. an increase in stiffness) for a 10 µg/m³ increment in PM10. However, no direct inflammatory effects, measured as concentration of plasma CRP and leukocytes, were detected. Stroke, or cerebrovascular accident, is a prominent cause of mortality and it has been linked with exposure to air pollution. I performed a meta-analysis of the current literature to quantify the pooled association between stroke and long-term exposure to PM10 or PM2.5 (Chapter 4). I identified 20 studies on long-term PM exposure and stroke. The association between PM and stroke was positive in North America (N=7 studies), Europe (N=8), and China (N=3, with extremely high exposures), and negative in Japan (N=2). The estimated effect of PM2.5 [6% increase (95% CI 2-11%) in stroke risk for a 5 µg/m³ increment in PM2.5] was higher than the corresponding result using PM10 [2% increase (95% CI -2 to 7%) in stroke risk for a 10 µg/m³ increment in PM10). This indicates the importance of measuring PM2.5 directly and confirms the hypothesis that PM2.5 is more hazardous than the coarse fraction (PM2.5-10). Conclusions In this thesis, I found detrimental health effects of air pollution in different susceptible populations and for different time windows of exposure. Both short-term exposure (1 to 3 days) and long-term exposure (several months to years) can trigger acute events, such as stroke, lung graft rejection or infant mortality, but long-term exposure can also accelerate the development of chronic cardiovascular or respiratory diseases, such as atherosclerosis or lung cancer. The two panel studies (LTx and health elderly) provided mixed results concerning the role of inflammation in the process, as measured by levels of leukocytes and plasma CRP. Compared to other environmental factors, such as smoking, diet and physical activity, individual risk estimates are rather small, but exposure to air pollution is involuntary and ubiquitous. Given the fact that the whole population is exposed, our studies demonstrated that air pollution is an important public health issue. In Belgium (and in Western Europe in general), yearly ambient concentrations of PM10 have decreased over the past 10 years to levels well below the EU limit value of 40 µg/m³, but still above the WHO guideline value of 20 µg/m³, especially in urban areas, where most people live. In accordance with other studies, we found no indications for a threshold or ‘save’ level of air pollution, so public health will benefit from every single µg/m³ decrease in concentration. This is a shared responsibility of policymakers, industry, and the general population (since road traffic and household combustion of wood are important and even increasing sources of air pollution) alike.VOORWOORD 7 TABLE OF CONTENTS 11 LIST OF ABBREVIATIONS 13 INTRODUCTION 17 AIR POLLUTION 19 HEALTH EFFECTS OF AIR POLLUTION 22 METHODS IN AIR POLLUTION RESEARCH 30 PUBLIC HEALTH ASPECTS 33 AIMS 40 REFERENCES 41 CHAPTER 1 DOES AIR POLLUTION TRIGGER INFANT MORTALITY IN WESTERN EUROPE? A CASE-CROSSOVER STUDY 49 CHAPTER 2 LYMPHOCYTIC BRONCHIOLITIS AFTER LUNG TRANSPLANTATION IS ASSOCIATED WITH DAILY CHANGES IN AIR POLLUTION 71 CHAPTER 3 CHANGING PLACES TO STUDY ACUTE AND SUBACUTE EFFECTS OF AIR POLLUTION ON CARDIOVASCULAR HEALTH 93 CHAPTER 4 LONG-TERM EXPOSURE TO PARTICULATE MATTER AIR POLLUTION IS A RISK FACTOR FOR STROKE: META-ANALYTICAL EVIDENCE 127 DISCUSSION 165 EXPOSURE ASSESSMENT 167 HEALTH EFFECTS OF EXPOSURE TO AIR POLLUTION 172 PUBLIC HEALTH RELEVANCE 176 FUTURE PERSPECTIVES AND RECOMMENDATIONS 176 REFERENCES 181 SUMMARY 187 SAMENVATTING 193 SHORT CURRICULUM VITAE AND LIST OF PUBLICATIONS 201nrpages: 206status: publishe