The Burden of COPD Due to Ozone Exposure in Germany
; ; ; ; ; ;
Background: The chronic effects of ozone have only rarely been investigated in disease burden studies to date. Our goal was to determine this disease burden in Germany over the period 2007–2016, with particular attention to estimation based on effect estimates adjusted for particulate matter (PM2.5) and nitrogen dioxide (NO2).
Methods: The nationwide, high-spatial-resolution (2 km × 2 km), population-based exposure to ozone in the summer months (“summer ozone”) was calculated on the basis of modeled ozone data and population counts in Germany. Next, risk estimates derived from cohort studies were used to quantify the burden of chronic obstructive pulmonary disease (COPD). Data on population counts, life expectancy, and mortality in Germany were used to reflect the situation across the country as accurately as possible.
Results: The estimates of years of life lost (YLL) due to summer ozone ranged from 18.33 per 100 000 people (95% confidence interval [14.02; 22.08]) in 2007 to 35.77 per 100 000 people [27.45; 42.98] in 2015. These findings indicate that ozone affects the COPD burden independently of other harmful components of the air. No clear secular trend in the COPD burden can be seen over the period 2007 to 2016.
Conclusion: Long-term exposure to ozone contributes to the COPD burden among the general population in Germany. As climate change may lead to a rise in the ozone concentration, more intensive research is required on the effects of ozone on health.
The concept of disease burden was developed to describe the loss of life as well as the loss of health due to disease, injury and risk factors in the population (1, 2, 3, 4, 5). This concept is based on the assumption that while all people can live for a certain number of years, a certain number of healthy years of life are lost as the result of limitations of health due to disease and premature death. With the help of the concept of environmental burden of disease, in turn, the disease burden related to the effect of an environmental risk factor on human health can be quantified (1). Apart from other measures, the years lost due to premature disease-related death as the result of environmental risk factors—the Years of Life Lost (YLL)—can be calculated to describe the environmental burden of disease (6, 7).
Due to its potent chemical reactivity, ozone is an irritant, causing oxidative damage to cells and the mucous membrane of the respiratory tract and immunoinflammatory reactions in the lungs, among other effects (8). It is scientifically proven and considered a causal factor that short-term ozone exposure is associated with increased mortality due to diseases of the respiratory system and with an increased number of respiratory disease-related emergency consultations and hospitalizations (8, 9, 10). Long-term exposure to ozone—the focus of this paper—is associated with compromised pulmonary function and deterioration of lung disease (11) as well as increased mortality due to diseases of the respiratory system (12, 13, 14). A comprehensive review of the literature, published by the U.S. Environmental Protection Agency in 2020, concluded that the association between long-term ozone exposure and respiratory mortality is likely to be causal (8). Relationships are referred to as likely to be causal if there are clear indications of causality, but the published data are considered not adequate to fulfill all the criteria for causality (10).
To date, only few studies have quantified the disease burden from long-term exposure to ozone (15, 16). Just recently, the latest estimation of disease burden from long-term ozone exposure on chronic obstructive pulmonary disease (COPD) mortality has been published as a result of the Global Burden of Disease (GBD) study (17). However, this estimation is based on modelled ozone data with a spatial resolution of 11 km × 11 km. In addition, the standardized life expectancy used in the GBD study is higher than the life expectancy documented in Germany.
The aim of this study was to determine the disease burden caused by ozone in Germany over the period 2007–2016. The calculation was based on the nationwide, high-spatial-resolution (2 km × 2 km) mean population-based exposure to ozone in the summer months (“summer ozone”) and data on population, life expectancy, and deaths in Germany to reflect the national situation as accurately as possible. In addition, further epidemiological evidence on the COPD mortality risk due to long-term exposure to ozone from the most recent cohort studies was taken into account. Another aim of this study was to estimate the disease burden from ozone after adjustment for additional air pollutants (particulate matter [PM2.5], nitrogen dioxide [NO2]) .
Selection of the health endpoint
This study estimated the disease burden only for COPD, because strong evidence from epidemiological studies as well as biological plausibility from experimental studies (animal experimental and mechanistic/in vitro studies) was available for this disease. In the study on which our paper is based, a systematic search of the literature and systematic mappings were conducted to review the evidence (18).
The population-based exposure to ozone across Germany was estimated based on nationwide data on the spatial distribution of ozone concentrations in ambient air (19). The calculation used data from the period 2007–2016, reflecting the pollution levels in both rural and urban backgrounds in a spatial resolution of approximately 2 km × 2 km (ozone grid cell). Using the chemical transport model REM/CALGRID, the data were generated and then combined with ozone measurement data of the nationwide air monitoring network of the German federal states and the Federal Environment Agency, using the optimal interpolation technique (20).
As an ozone exposure indicator, the summer ozone concentrations (in µg/m³) for each year were calculated as the mean over the daily maxima of the 8-hour moving averages from April to September for each ozone grid cell (19). These ozone data were combined with spatial population data for Germany of the 2011 census (German Federal and State Statistical Offices—version 1; 23 April 2015); then the population-weighted summer ozone exposure was calculated for each year of the period 2007–2016.
Quantification of the disease burden from ozone
For the quantification of the disease burden attributable to ozone, data on population counts, life expectancy, and mortality in Germany were used to reflect the situation across the country as accurately as possible. For a detailed description of the input data, refer to the eMethods section.
Estimation of the disease burden from ozone
Only risk estimates from cohort studies on long-term exposure (annual mean of summer ozone concentrations) were used to calculate the burden of disease from summer ozone exposure. If multiple robust effect estimates for the exposure–response relationship were available from different studies, these were pooled in a first step by performing a meta-analysis with random effects (21) (eMethods).
The population-weighted ozone exposure in Germany and the combined effect estimates were then used to calculate the so-called attributable fraction (3, 22)—the proportion of the COPD burden that can be attributed to the risk factor ozone.
In a further step, the number of years of life lost (YLL) due to premature mortality related to COPD were determined (7). These were calculated by multiplying the number of COPD-related deaths per year by the statistical residual life expectancy of the German population (1, 7) (eMethods). Finally, the ozone-related burden for COPD was estimated by multiplying the attributable fraction by the number of years of life lost (YLL) due to COPD mortality.
The uncertainty about whether the burden of disease can actually be attributed to a single pollutant (single pollutant effect estimate) is a fundamental problem of studies on pollutant-related burden of disease, since air pollutants always occur as a mixture of pollutants which often originate from the same sources. Thus, our study paid particular attention to the estimation of the disease burden from summer ozone based on effect estimates adjusted for particulate matter (PM2.5) and nitrogen dioxide (NO2).
In a sensitivity analysis, an exposure–response relationship after adjustment for PM2.5, NO2 and, additionally, temperature (12) was used as an alternative approach to calculate the attributable fraction and the YLL related to COPD burden.
The majority of the German population (mean 95.3%; minimum: 79.6% in 2015, maximum 99.0% in 2013) were exposed to summer ozone concentrations of 75–95 µg/m³ in the period 2007–2016 (Figure). On average, 1.1% of people lived in the least polluted regions with summer ozone concentrations below 75 µg/m³, from 0.0% in 2015 to 3.4% in 2007. On average, 3.6% of people lived in regions with summer ozone concentrations higher than 95 µg/m³.
COPD burden from long-term exposure to summer ozone
The results for COPD burden (based on deaths coded with J40–J44, ICD-10) due to long-term exposure to summer ozone in the period 2007–2016 are presented in Table 1. Based on the age range for the study populations, the age range was set at 30 years and older. An important point to note when interpreting these results is that the years of life lost are not equally distributed among all residents, but rather some residents lose more years of life than others.
Overall, no clear trend can be identified in the period 2007–2016; during the 10-year observation period, the disease burden varied by more than one-third from year to year. Moreover, variations in disease burden are caused by variations in COPD mortality. The highest attributable fraction and also the highest number of YLL was observed in 2015, reflecting the very high ozone concentrations recorded in that year.
Disease burden from ozone after adjustment for other air pollutants
Table 2 shows the results for the disease burden from long-term exposure to summer ozone after adjustment for PM2.5 and NO2. Compared to the results in Table 1, the disease burden of COPD is even higher after adjustment for PM2.5 and NO2. These findings indicate that ozone affects the COPD burden independent of other pollutants.
The use of the exposure-response relationship after adjustment for particulate matter, nitrogen dioxide and, in addition, temperature is illustrated in Table 3 for 2016 as an example. A corresponding effect estimate is available from one study only (12). For this reason, disease burden estimates based on single-pollutant effect estimates and after adjustment for particulate matter and NO2 are presented in Table 3 to enhance comparability. While no significant changes in the attributable fraction and the YLL were observed, it was found that the confidence intervals were very wide. All in all, the results are indicative of an independent effect of summer ozone on COPD burden, even after adjustment for temperature.
Long-term exposure to ozone contributes to the disease burden among the general population in Germany. For the COPD burden, the estimates of years of life lost ranged from 18.33 per 100 000 people (95% confidence interval [14.02; 22.08]) (year 2007) to 35.77 per 100 000 people [27.45; 42.98] (year 2015). Overall, no clear temporal trend can be identified in the disease burden over the period 2007–2016; during the 10-year observation period, the relative disease burden varied by more than one-third from year to year. After additional adjustment of the effect estimate for PM2.5 and NO2, the COPD burden attributable to ozone was slightly higher compared to the unadjusted results.
To date, there are only few studies that have quantified the disease burden from long-term exposure to ozone (15, 16). Just recently, the latest estimation of disease burden from long-term ozone exposure on COPD mortality has been published as a result of the Global Burden of Disease (GBD) study (17). Unlike previous burden of disease studies, this latest estimation was based on a risk estimate averaged from three cohort studies. For the year 2016, 43.05 ozone-related YLL per 100 000 people [18.14; 74.06] were published for Germany. This figure is higher than the estimate of 27.92 YLL [21.38; 33.60] for 2016 calculated in our study. However, it should be noted that the estimation of the GBD study is based on modeled ozone data with a spatial resolution of 11 km × 11 km, while we used a high resolution of 2 km × 2 km in our study. In addition, the standardized life expectancy used in the GBD study is higher than the life expectancy documented in Germany. Furthermore, in the GBD study the disease burden was estimated for all age groups 25 years and older, while our study set the lower age limit at age 30 years.
With regard to the biological plausibility of a link between long-term ozone exposure and COPD mortality, there is evidence from experimental studies (animal experimental and mechanistic/in vitro studies), indicating that chronic exposure to ozone can trigger pathophysiological processes similar to those seen with COPD (8, 11).These include persistent inflammatory processes, oxidative stress as well as airway damage and structural remodeling, resulting in irreversible changes, including fibrotic and emphysematous changes of the lung (8).
Of special interest are the estimates of COPD burden from ozone which were calculated using effect estimates adjusted for other air pollutants (PM2.5 and NO2) and are presented for the first time. These results suggest that the effect of ozone on COPD burden is largely independent of other air pollutants. However, it should be noted that the level of ozone concentrations and the mix of pollutants in the United States (all studies covered were US studies) are not necessarily comparable with those in Germany, mainly due to differences in the structure of the emitters, the climatic conditions and, consequently, the atmospheric transformation processes.
Any comparisons of disease burden estimates in this paper with estimates for particulate matter (23, 24) and NO2 (25) should take into account the complexity of the data and methods underlying these estimates. First, there are considerable differences in the health endpoints investigated and the risk estimates used. Further differences also arise for the time periods considered and the spatial resolutions of the nationwide exposure data. Despite all this uncertainty, and when the results of the GBD study are also included (17), a clear gradation of disease burdens appears to be emerging for the three air pollutants studied. Long-term ozone exposure shows the lowest disease burden, presumably followed by NO2 and PM2.5 with the by far highest burden of disease.
Strengths and limitations
This study has a number of strengths. For example, the 2 km × 2 km grid size resolution of the spatial ozone concentration data used in this study is higher than any of the resolutions used in other comparable studies. The use of the current effect estimates in studies of the American Cancer Society (ACS) (13) and the Canadian Census Health and Environment Cohort (CanCHEC) (26) as well as other effect estimates published for the first time is considered an additional strength of this study. Effect estimates adjusted for particulate matter, NO2 and, additionally, temperature were used to estimate the disease burden due to ozone in further analyses. This is of special importance since the use of estimates derived from single-pollutant models has been criticized for years (27), in many cases, however, without remedying this shortcoming.
On the other hand, our study has some limitations. With regard to the calculation of ozone exposure, it must be clearly stated that these ozone concentrations are only modeled concentrations in a specific (small-scale) polygon across Germany. Although this is a common approach to estimate exposure, it is a greatly simplified method, especially because the length of stay differs between people. Furthermore, there is potential for misclassification of exposure because movements of people in and out of an area are not or only incompletely taken into account; this applies especially to migrants where the COPD rate can be particularly high as the result of the situation in their home countries (28).
In addition, it is important to mention that COPD is a disease with progression over decades. The development or rather the cause underlying the development of COPD predates the exposure years assessed in this study. Apart from ozone, other factors are known to contribute to the development of COPD (e.g., smoking, genetic disposition, occupational inhalation of dust, respiratory tract infections in childhood, and prematurity).
As with all environmental burden of disease studies, the results of this study are to be used exclusively to draw conclusions on the population level. Information about the health status of specific individuals cannot be derived from it (29). Furthermore, these estimates are based on calculations. In order to perform these calculations, various assumptions must be made, such as, for example, assumptions about the exposure–response relationship or the residual life expectancy at the time of death. The effect estimates used in this study are from large North American studies. Their applicability to German conditions is to be understood as a necessary, pragmatic approach, reflecting the lack of pertinent German or European data.
The ranges of uncertainty of the estimates of disease burden from ozone are large, as shown, for example, by the confidence intervals. Therefore, we explicitly warn against an uncritical use of the point estimate without stating the confidence interval.
Long-term exposure to ozone contributes to disease burden among the general population in Germany. Since climate change may lead to a rise in ozone concentrations (30, 31), more intensive research is required on the effects of ozone on health and on the disease burden attributable to ozone. This means that more cohort studies on the long-term effects of ozone in Europe and especially in Germany are generally needed. Moreover, the fact that regulated air pollutants, such as particulate matter, NO2 and ozone, always occur as a mixture of pollutants has not been adequately taken into account to date. Furthermore, the health effects of interactions between long-term ozone exposure on the one hand and temperature, volatile organic compounds, ultrafine particles and green spaces on the other hand have not yet been studied sufficiently.
This project was funded by the German Federal Environment Agency within the scope of the Departmental Research Plan (“Ressortforschungsplan”) 2018 of the German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (project code number 3718 62 208 0).
Conflict of interest statement
The authors declare that no conflict of interest exists.
Manuscript received on 9 March 2021; revised version accepted on 1 June 2021
Translated from the original German by Ralf Thoene, MD.
Prof. Dr. Joachim Heinrich
Institut und Poliklinik für Arbeits-,
Sozial- und Umweltmedizin
Ziemssenstraße 1, 80336 München, Germany
Cite this as:
Breitner S, Steckling-Muschack N, Markevych I, Zhao T, Mertes H, Nowak D, Heinrich J: The burden of COPD due to ozone exposure in Germany. Dtsch Arztebl Int 2021; 118: 491–6. DOI: 10.3238/arztebl.m2021.0258
Institute of Epidemiology, Helmholtz Zentrum München GmbH – German Research Center for Environmental Health, Neuherberg, Germany: Dr. rer. nat. Susanne Breitner, Dr. hum. biol. Iana Markevych, Tianyu Zhao MSc
Institute and Outpatient Clinic for Occupational, Social and Environmental Medicine, LMU Hospital, Comprehensive Pneumology Center (CPC) Munich, German Center for Lung Research (DZL), Munich, Germany: Dr. PH Nadine Steckling-Muschak, Dr. hum. biol. Iana Markevych, Tianyu Zhao MSc, Hanna Mertes MPH, Prof. Dr. med. Dennis Nowak, Prof. Dr. Joachim Heinrich
Institute of Psychology, Jagiellonian University, Krakow, Poland: Dr. hum. biol. Iana Markevych
|1.||Tobollik M, Plass D, Steckling N, Zeeb H, Wintermeyer D, Hornberg C: [The environmental burden of disease concept]. Gesundheitswesen 2018; 80: 154–9 CrossRef MEDLINE|
|2.||Prüss-Üstin A, Mathers C, Corvalán C, Woodward A: Assessing the environmental burden of disease at national and local levels: introduction and methods. Geneva: World Health Organization 2003.|
|3.||Prüss-Üstin A, Corvalán C, Bos R, Neira M: Preventing disease through healthy environments: a global assessment of the burden of disease from environmental risks. Geneva: World Health Organization 2016.|
|4.||Plass D, Vos T, Hornberg C, Scheidt-Nave C, Zeeb H, Krämer A: Trends in disease burden in Germany—results, implications and limitations of the Global Burden of Disease Study. Dtsch Arztebl Int 2014; 111: 629–38 VOLLTEXT|
|5.||Rommel A, von der Lippe E, Plass D, et al. on behalf of the BURDEN 2020 Study Group: The COVID-19 disease burden in Germany in 2020—years of life lost to death and disease over the course of the pandemic. Dtsch Arztebl Int 2021; 118: 145–51 VOLLTEXT|
|6.||Martinez R, Soliz P, Caixeta R, Ordunez P: Reflection on modern methods: years of life lost due to premature mortality—a versatile and comprehensive measure for monitoring non-communicable disease mortality. Int J Epidemiol 2019; 48: 1367–76 CrossRef MEDLINE PubMed Central|
|7.||Wengler A, Rommel A, Plaß D, et al. on behalf of the BURDEN 2020 Study Group: Years of life lost to death—a comprehensive analysis of mortality in Germany conducted as part of the BURDEN 2020 project. Dtsch Arztebl Int 2021; 118: 137– 44 VOLLTEXT|
|8.||U.S. EPA: Integrated science assessment (ISA) for ozone and related photochemical oxidants (Final Report). Washington, DC: U.S. Environmental Protection Agency 2020.|
|9.||Peters A, Hoffmann B, Brunekreef B, et al.: Die Rolle der Luftschadstoffe für die Gesundheit. Umweltmed – Hygiene – Arbeitsmed 2019; 24: 9–14 CrossRef|
|10.||Ritz B, Hoffmann B, Peters A: The effects of fine dust, ozone, and nitrogen dioxide on health. Dtsch Arztebl Int 2019; 116: 881–6 VOLLTEXT|
|11.||Wang M, Aaron CP, Madrigano J, et al.: Association between long-term exposure to ambient air pollution and change in quantitatively assessed emphysema and lung function. JAMA 2019; 322: 546–56 CrossRef MEDLINE PubMed Central|
|12.||Lim CC, Hayes RB, Ahn J, et al.: Long-term exposure to ozone and cause-specific mortality risk in the United States. Am J Respir Crit Care Med 2019; 200: 1022–31 CrossRef MEDLINE PubMed Central|
|13.||Turner MC, Jerrett M, Pope CA 3rd, et al.: Long-term ozone exposure and mortality in a large prospective study. Am J Respir Crit Care Med 2016; 193: 1134–42 CrossRef MEDLINE PubMed Central|
|14.||Kazemiparkouhi F, Eum KD, Wang B, Manjourides J, Suh HH: Long-term ozone exposures and cause-specific mortality in a US Medicare cohort. J Expo Sci Environ Epidemiol 2020; 30: 650–8 CrossRef MEDLINE PubMed Central|
|15.||Malley CS, Henze DK, Kuylenstierna JCI, et al.: Updated global estimates of respiratory mortality in adults ≥ 30 years of age attributable to long-term ozone exposure. Environ Health Perspect 2017; 125: 087021 CrossRef MEDLINE PubMed Central|
|16.||Cohen AJ, Brauer M, Burnett R, et al.: Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: an analysis of data from the Global Burden of Diseases Study 2015. Lancet 2017; 389: 1907–18 CrossRef|
|17.||GBD 2019 Risk Factor Collaborators: Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 2020; 396: 1223–49 CrossRef|
|18.||Zhao T, Markevych I, Janßen C, Nowak D, Steckling-Muschack N, Heinrich J: Ozone exposure and health effects: a protocol for an umbrella review and effect-specific systematic maps. BMJ Open 2020; 10: e034854 CrossRef MEDLINE PubMed Central|
|19.||Breitner S, Steckling-Muschack N, Markevych I, et al.: Quantifizierung der Krankheitslast verursacht durch Ozon-Exposition in Deutschland für die Jahre 2007–2016. Umwelt & Gesundheit. Dessau-Roßlau: Umweltbundesamt, in Druck.|
|20.||Flemming J, van Loon M, Stern R: Data assimilation for ctm based on optimum interpolation and Kalman filter. Boston, MA: Springer US 2004: 373–82 CrossRef|
|21.||Veroniki AA, Jackson D, Bender R, et al.: Methods to calculate uncertainty in the estimated overall effect size from a random-effects meta-analysis. Res Synth Methods 2019; 10: 23–43 CrossRef MEDLINE|
|22.||Ostro B, World Health Organization (WHO): Outdoor air pollution: assessing the environmental burden of disease at national and local levels. Geneva: World Health Organization 2004.|
|23.||Kallweit D, Wintermeyer D: Berechnung der gesundheitlichen Belastung der Bevölkerung in Deutschland durch Feinstaub (PM10). UMID: Umwelt und Mensch – Informationsdienst: Umweltbundesamt 2013; 4: 18–24.|
|24.||Umweltbundesamt: Gesundheitsrisiken durch Feinstaub. www.umweltbundesamt.de/daten/umwelt-gesundheit/gesundheitsrisiken-durch-feinstaub#ermittlung-der-feinstaubbelastung-exposition-am-wohnort (last accessed on 31 January 2021).|
|25.||Schneider A, Cyrys J, Breitner S et al.: Quantifizierung von umweltbedingten Krankheitslasten aufgrund der Stickstoffdioxid- Exposition in Deutschland. Umwelt & Gesundheit. Dessau-Roßlau: Umweltbundesamt 2018.|
|26.||Weichenthal S, Pinault LL, Burnett RT: Impact of oxidant gases on the relationship between outdoor fine particulate air pollution and nonaccidental, cardiovascular, and respiratory mortality. Sci Rep 2017; 7: 16401 CrossRef MEDLINE PubMed Central|
|27.||Anenberg SC, Henze DK, Tinney V, et al.: Estimates of the global burden of ambient PM2.5, ozone, and NO2 on asthma incidence and emergency room visits. Environ Health Perspect 2018; 126: 107004 CrossRef MEDLINE PubMed Central|
|28.||Olaru ID, Van Den Broucke S, Rosser AJ, et al: Pulmonary diseases in refugees and migrants in Europe. Respiration 2018; 95: 273–86 CrossRef MEDLINE|
|29.||Gruhl H, Tobollik M, Wengler A, et al.: Schätzung der umweltbedingten Krankheitslast im Rahmen des Projektes BURDEN 2020 – Projekthintergrund und methodisches Vorgehen. UMID: Umwelt und Mensch – Informationsdienst: Umweltbundesamt 2019; 2: 37–50.|
|30.||Doherty RM, Heal MR, O’Connor FM: Climate change impacts on human health over Europe through its effect on air quality. Environ Health 2017; 16: 118 CrossRef MEDLINE PubMed Central|
|31.||Fortems-Cheiney A, Foret G, Siour G, et al.: A 3 °C global RCP8.5 emission trajectory cancels benefits of European emission reductions on air quality. Nature Communications 2017; 8: 89 CrossRef MEDLINE PubMed Central|