Secondary Logo

Journal Logo

Air Pollution

Particulate Matter Composition and Respiratory Health

The PIAMA Birth Cohort Study

Gehring, Ulrikea; Beelen, Roba; Eeftens, Marloesa,b,c; Hoek, Gerarda; de Hoogh, Keesb,c,d; de Jongste, Johan C.e; Keuken, Mennof; Koppelman, Gerard H.g,h; Meliefste, Keesa; Oldenwening, Mariekea; Postma, Dirkje S.h,i; van Rossem, Leniej; Wang, Menga; Smit, Henriette A.j; Brunekreef, Berta,j

Author Information
doi: 10.1097/EDE.0000000000000264
  • Free

Abstract

There is a vast body of evidence for the chronic adverse effects of ambient air pollution exposure on the respiratory health of children. Recent reviews conclude that there is consistent evidence for a positive association between traffic-related air pollution and asthma prevalence and incidence in children.1–3 Findings were more heterogeneous for asthma incidence than for asthma prevalence.2 Furthermore, there is evidence for adverse effects of outdoor air pollution on lung function of children.4 The association between air pollution exposure and allergic sensitization is less clear.2

To date, many studies on the health effects of ambient air pollution used nitrogen dioxide (NO2), mass concentrations of particulate matter (PM) with aerodynamic diameters of less than 10 μm (PM10) and less than 2.5 μm (PM2.5), and elemental carbon/filter reflectance as surrogates for the complex air pollution mixture. However, PM composition can vary considerably as PM is emitted from a large variety of sources including traffic, industry, biomass burning, and long-range transport.5 With the substantial reduction of tailpipe emissions from motorized traffic, non-tailpipe emissions from resuspension of road dust formed by abrasion of road surface material and wear of tires and brakes may become more important.5 There is increasing evidence that different constituents of PM affect health in different ways.5

We previously reported positive associations between traffic-related air pollution (NO2, PM2.5, and PM2.5 absorbance) and the incidence and prevalence of asthma during the first 8 years of life from our prospective Prevention and Incidence of Asthma and Mite Allergy (PIAMA) birth cohort study.6 Furthermore, as part of the European collaborative European Study of Cohorts for Air Pollution Effects (ESCAPE) study, we recently reported negative associations between air pollution (NO2, NOx, PM2.5, and PM2.5 absorbance) and lung function at age 6–8 years from five European birth cohort studies, including PIAMA.7 Since then, another follow-up of our cohort, including a questionnaire survey and a medical examination of a subset of the population, was completed when the participants were 11–12 years old. Furthermore, within the framework of collaborative Transport related Air Pollution and Health impacts-Integrated Methodologies for Assessing Particulate Matter project, land-use regression models have been developed for a selection of eight elements in PM2.5 and PM10 representing different sources such as non-exhaust emissions of traffic from brake linings and tires (copper, iron, and zinc), industrial (smelter) emissions (iron and zinc), crustal materials (soil; silicon, and potassium), fossil fuel combustion (nickel, vanadium, and sulfur) and biomass burning (potassium).8–10 We therefore investigated the effects of different PM characteristics and PM constituents on asthma, hay fever, allergic sensitization, and lung function of the participants of the PIAMA birth cohort study until the age of 11–12 years. Because we currently know very little about the relevance of the timing of the exposure in addition to the exposure level, we explored associations with early life exposures and with exposures later in life.

MATERIALS AND METHODS

Study Population

The PIAMA study is a prospective birth cohort study.11 Women were recruited in 1996–1997 during their second trimester of pregnancy from a series of communities in the North, West, and Centre of the Netherlands. Non-allergic pregnant women were invited to participate in a “natural history” study arm; allergic women, identified through a screening questionnaire were primarily allocated to an intervention arm (involving the use of mite-impermeable mattress and pillow-covers) with a random subset allocated to the natural history arm. The study population for this study consisted of participants in the intervention and natural history studies with data on at least one of the health outcomes studied and with data on air pollution exposure and potential confounders available to be included in at least one of the adjusted analyses (N = 3,702). The Institutional Review Boards of the participating institutes approved the study protocol, and written informed consent was obtained from the parents or legal guardians of all participants.

Health Outcomes

Information on the children’s respiratory health was collected by questionnaire annually until age 8 years and between 11 and 12 years. From these questionnaires, the same health outcomes as in previous analyses,6 namely incident doctor-diagnosed asthma, asthma symptoms in the past 12 months, hay fever ever, and rhinitis in the past 12 months (see Supplemental Digital Content at https://links.lww.com/EDE/A883 for exact definitions) were studied.

At age 12 years, a blood sample was taken from all children who gave consent. Specific immunoglobulin E levels for house dust mite (Dermatophagoides pteronyssinus), cat, cocksfoot (Dactylis glomerata), and birch pollen were measured in blood samples by a radioallergosorbent test-like method used at the Sanquin Laboratories (Amsterdam, The Netherlands). Sensitization was defined as a positive reaction (specific IgE level ≥ 0.70 IU/ml) to at least one of the allergens tested.

Spirometry was performed when the children were approximately 8 and 12 years old following the recommendations of the American Thoracic Society (ATS)/European Respiratory Society (ERS).12 A detailed description of the tests can be found in the Supplemental Digital Content, https://links.lww.com/EDE/A883. Forced expiratory volume in 1 second (FEV1) and forced vital capacity (FVC) were measured at both ages; mid-expiratory flows (FEF25-75) were assessed at age 12 only. A total of 555 children had lung function measurements at both ages (8 and 12 years). Height and weight were measured as described in the Supplemental Digital Content, https://links.lww.com/EDE/A883.

Long-term Air Pollution Exposure Assessment

Air pollution concentrations at the participants’ birth addresses and current home addresses at the time the questionnaires were completed for questionnaire-based health outcomes and at the time of the medical examination were estimated by land-use regression.8–10 In brief, air pollution monitoring campaigns were performed between October 2008 and February 2010 in the study area. Three 2-week measurements of nitrogen dioxide (NO2) and nitrogen oxides (NOx) were performed at 80 sites within 1 year. In addition, simultaneous measurements of PM2.5, PM10, PMcoarse (2.5–10 μm), and soot (determined as the reflectance of PM2.5 filters) were performed at half of the sites.13,14 All PM2.5 and PM10 filters were analyzed for elemental composition using x-ray fluorescence.10

For each site, results from three measurements were averaged to estimate the annual average.14 Predictor variables on nearby traffic intensity, population/household density and land use were derived from Geographic Information Systems (GIS) to explain spatial variation of PM, NO2 and concentrations of copper, iron, potassium, nickel, sulfur, silicon, vanadium, and zinc. Elements were selected to represent major sources of air pollution. A brief description of the models and their performance is provided in eTable 1 (https://links.lww.com/EDE/A883); detailed descriptions have been published elsewhere.8–10 The models explained substantial fractions of the variability in annual average concentrations for NO2, PM2.5, PM10, PM2.5 absorbance, copper, iron, nickel, vanadium, and zinc (leave-one-out cross-validation R2, R2LOOCV = 0.58–0.89, eTable 1; https://links.lww.com/EDE/A883), but only limited fractions for PMcoarse, potassium, sulfur, and silicon (R2LOOCV = 0.25–0.45). The relatively poor performance of these models can be explained by the lack of specific GIS variables for sources other than traffic and low spatial variation (sulfur).10

Covariates

Information on covariates such as sex, maternal education (low: primary school, lower vocational, or lower secondary education; medium: intermediate vocational education or intermediate/higher secondary education; high: higher vocational education and university), parental allergies, older siblings (yes/no), breastfeeding for at least 3 months (yes/no), maternal smoking during pregnancy (yes/no), smoking at the child’s home (yes/no), use of gas for cooking (yes/no), mold/dampness (yes/no), furry pets in the child’s home (yes/no), and day-care center attendance (yes/no) was available from questionnaires. Information on the percentage of low-income households in the neighborhood was available from Statistics Netherlands. Average ambient levels of PM10 and NO2 for the 7 days preceding the lung function tests were retrieved from the National Air Quality Monitoring Network (see the Supplemental Digital Content, https://links.lww.com/EDE/A883).

Statistical Analysis

Associations between air pollution and asthma incidence until age 11 years were analyzed by means of discrete-time hazard models,15 associations with repeated yearly questionnaire reports of asthma symptoms, hay fever, and rhinitis until age 11 years by generalized estimation equations with a logit-link using a six-dependent correlation matrix.16 Analyses were performed with exposures at the birth address (constant) and exposures at the address at the time of the follow-up (time-varying), representing exposure during the past 12 months. Associations of air pollution levels at the birth and current address with allergic sensitization were analyzed by means of log-binomial regression because of the high prevalence. All analyses were performed with and without adjustment for the potential confounding variables mentioned above except short-term air pollution exposures. Linear regression analyses with natural log (ln) transformed lung function parameters (FEV1, FVC, FEF25-75) as dependent variables were used to analyze associations between air pollution and continuous lung function parameters.17 Associations were calculated for exposures at the birth address and exposures at the address at the time of spirometry, with adjustment for sex and natural log-transformed age, height, and weight only (crude model), and with adjustment for the potential confounding variables mentioned above and respiratory infections during the past 3 weeks (adjusted model). In addition, we performed linear regression analyses of the association between annual changes in lung function from age 8 to 12 years and average air pollution exposure during that period. Time-varying covariates were defined for the first year of life in analyses with birth address exposures and for the year of the follow-up otherwise to coincide with exposure.

Because the intervention with mite-impermeable mattress covers did not reduce the risk of asthma and allergic sensitization,18 no adjustment was made for study arm (ie, intervention or natural history study). Confounding by birth weight, which may be on the causal pathway between air pollution and the health outcomes studied, was explored as part of a sensitivity analysis. Air pollution levels were entered as continuous variables without transformation in all models. Associations are presented for an interquartile range increase in exposure to facilitate comparison of effect estimates between exposures. For lung function, potential effect modification by moving (defined as any change of address since birth) was explored in stratified analyses to investigate the relevance of chronic exposures (non-movers) versus more recent exposures (movers). In addition, potential effect modification by sex was explored in stratified analyses. Analyses were performed with the Statistical Analysis System (SAS 9.4, Cary, NC) for Windows. Effects are presented as relative risks for allergic sensitization, as odds ratio for all other binary outcomes, as percent change in lung function parameter and as mean change in annual lung function growth with 95% confidence intervals for a given increase in exposure.

RESULTS

Population Characteristics

General characteristics and distributions of the health outcomes are presented in Table 1 and frequency distributions of questionnaire-based health outcomes are presented in Figure 1. Participants who were included in this analysis more often had highly educated parents and more often lived in non-smoking homes than non-participants (eTable 2; https://links.lww.com/EDE/A883). For all other characteristics, differences between participants and non-participants were small.

T1-3
TABLE 1:
Distribution of Population Characteristics and Health Outcomes (Total N = 3,702)
F1-3
FIGURE:
Percentage of children with incident asthma, hay fever, and related symptoms at ages 1 to 11 years. NA indicates not available.

Air Pollution Exposure

Distributions of annual average concentrations of NO2, PM2.5 absorbance, PM2.5, PM10, and PMcoarse mass as well as elemental composition of PM2.5 and PM10 at the participants’ birth addresses are presented in Table 2. Spatial variability was larger for NO2 and PM2.5 absorbance (maximum–minimum concentration ratios of 4–7, data not shown) than for PM2.5 and PM10, and largest for copper and iron in PM2.5 (maximum–minimum ratios of 8–10). Some elements were mainly contained in the coarse fraction (>2.5μm) of PM10 (copper, iron, and silicon), whereas others (sulfur, nickel, vanadium, and zinc) were mainly contained in PM2.5. Correlations between air pollutants at the birth address are presented in eTable 3 (https://links.lww.com/EDE/A883). Correlation patterns were complex. Some elements (eg, potassium in PM2.5) had a low correlation with most of the other air pollutants suggesting different sources. Distributions of concentrations, and correlations between air pollutants, for the 11-year addresses were similar to those for the birth addresses (data not shown). Correlations between levels of the same pollutant at the birth and 11-year address were moderate to high, ranging from 0.56 (potassium in PM10) to 0.91 (sulfur in PM2.5, eTable 4; https://links.lww.com/EDE/A883). Correlations of long-term exposures with short-term exposures were all low for short-term PM10 and varied between -0.09 and 0.60 for short-term NO2 (eTable 5; https://links.lww.com/EDE/A883).

T2-3
TABLE 2:
Distribution of Air Pollution Concentration Levels at the Birth Address (N = 3,682)

Air Pollution and Health

Differences between crude (eTable 6; https://links.lww.com/EDE/A883) and adjusted (Table 3) associations of annual average air pollution levels at the birth address with asthma, hay fever, and allergic sensitization were mostly small. NO2, zinc in PM10, sulfur in PM2.5, and potassium in PM2.5 and PM10 at the birth address were positively associated with asthma incidence during the first 11 years of life, after adjustment for potential confounders. NO2, zinc, silicon, and potassium in PM10 and potassium in PM2.5 at the birth address were positively associated with asthma symptoms. NO2 at the birth address was positively associated with hay fever. NO2, zinc and potassium in PM2.5 and in PM10, and sulfur in PM2.5 were positively associated with rhinitis. PM10 and copper and iron in PM10 were positively associated with allergic sensitization. We found few associations of annual average air pollution levels at the current address with asthma, hay fever, and allergic sensitization; they were largely limited to copper and potassium in PM10 (eTable 7; https://links.lww.com/EDE/A883). Effect sizes were similar for NO2, particle mass, and element concentrations. Effect sizes for exposures at the birth address (ranging from 1.07 for asthma symptoms and silicon in PM10 to 1.29 for rhinitis and sulfur in PM2.5) generally were somewhat larger than effects for exposures at the current address (all estimated risks between 1.06 and 1.11), after adjustment for confounders.

T3-3
TABLE 3:
Adjusteda Associations of Asthma, Hay Fever, and Allergic Sensitization with Exposure at the Birth Address

Estimated effects of air pollution exposures on lung function were generally largest for FEF25-75 and lowest for FVC. After adjustment for potential confounders, FEV1 was associated with NO2, PM2.5 absorbance, copper and iron in PM2.5, and FEF75-50 was associated with PMcoarse and copper in PM10 at the current address (Table 4, crude associations are presented in eTable 10; https://links.lww.com/EDE/A883). Few associations were found between lung function and exposure at the birth address (eTable 11; https://links.lww.com/EDE/A883). Estimated reductions in lung function were around 1%–2% for all pollutants. No association was found between lung function growth from 8 to 12 years of age and air pollution exposure during the follow-up (eTable 10; https://links.lww.com/EDE/A883).

T4-3
TABLE 4:
Adjusteda Associations Between Lung Function and Exposure at the Current Address

Sensitivity Analyses

Associations with symptoms and with lung function remained unchanged after additional adjustment for birth weight (data not shown). Associations with lung function remained largely unchanged when we restricted our analysis to measurements that fulfilled the ATS/ERS criteria or excluded children (n = 73) who took asthma medication during the 48 hours before the lung function measurement (data not shown).

Because of the high correlations between exposure levels for some elements (in particular copper, iron, sulfur, and silicon) and PM mass, the associations found for these elements may actually be partly attributable to PM mass. We explored to what extent the observed associations with elemental composition of PM2.5 and PM10 were independent of PM mass by additionally adjusting models with elemental composition for PM2.5 and PM10 mass, respectively. Our findings indicate that observed associations with elemental composition are largely independent of PM mass (eFigures 1 and 2; https://links.lww.com/EDE/A883).

For FEF25-75, and to a lesser extent for FEV1, but not for FVC, the negative associations with air pollution exposures tended to be stronger in children who lived at the same address since birth (eFigure 3; https://links.lww.com/EDE/A883). Confidence intervals for movers and non-movers, however, largely overlap. The adverse effects of air pollution exposures on asthma (symptoms), rhinitis, and FVC tended to be stronger in girls than in boys (eFigures 4 and 5; https://links.lww.com/EDE/A883). Confidence intervals of effect estimates for boys and girls largely overlapped.

DISCUSSION

We investigated the effects of NO2 and different PM characteristics and PM constituents on asthma, hay fever, allergic sensitization, and lung function until the age of 11–12 years. We found adverse effects of air pollution on asthma, rhinitis, allergic sensitization, and lung function most consistently for exposure to NO2, copper, iron, zinc, and potassium. We found fewer associations for these outcomes with PM2.5 absorbance and particle mass.

The epidemiologic evidence on the effects of specific PM constituents on respiratory and allergic disease is currently limited and the comparison with the findings of this study is hampered by differences in health outcomes, constituents, study design, and age of the participants. In a birth-cohort study from New York City, a positive association was found between PM metal content, in particular nickel, and wheeze during the first 2 years of life.19 Very recently, as part of the European collaborative ESCAPE study, negative associations were reported between nickel and sulfur in PM10 and lung function at age 6–8 years from five European birth cohort studies, including PIAMA.20 In a time-series study in the Greater Baltimore area (US), higher levels of zinc in PM2.5 were associated with asthma exacerbations in children ages 0–17 years.21 A European multicenter panel study found acute adverse effects of iron and silicon in PM10 on peak expiratory flow (PEF) and prevalence of phlegm in children ages 6–12 years.22

Indirect evidence comes from a study in Eastern Germany. In a cross-sectional analysis, higher lifetime prevalence of asthma or wheezy bronchitis, bronchitis, wheeze, shortness of breath and cough without cold, and increased rates of allergic sensitization were found in children ages 5–14 years living in two counties impacted by industrial pollution (chemical and power plants in Bitterfeld, mining and smelting of non-ferrous metals including lead and copper in Hettstedt) than in children living in a neighboring county without industrial pollution.23 In a subsequent longitudinal analysis, a decline in air pollution levels was associated with a decrease in bronchitis and sinusitis and improved lung function (FVC and to a lesser extent FEV1) in these children.24,25 Only particle mass, but not particle metal composition, was measured in the East German study. However, later experimental work with dust collected in the same areas indicated that higher concentrations of transition metals, in particular copper and zinc, may have been partly responsible for the observed associations with PM.26

In vivo and in vitro studies with PM collected in the Utah Valley area (US) before and after closure of a steel mill provide further evidence for a contribution of metals (iron, copper, nickel, and zinc among others) to the increase in hospital admissions for respiratory illnesses including pneumonia, pleurisy, bronchitis, and asthma in children and adults associated with PM exposure.27–29 Experimental studies in mice also suggest that transition metals in PM, including zinc, copper, and cadmium may be involved in allergic responses.30 Transition metals, such as iron and copper, are believed to contribute to particle-induced formation of reactive oxygen species through the Fenton reaction, whereas zinc may trigger effects more directly by interacting with cellular proteins.5

We investigated the effects of eight different elements in PM on asthma, hay fever, allergic sensitization, and lung function of children. A review of European studies dealing with source apportionment of PM identified the following sources for these eight elements: traffic (copper, iron, and zinc), crustal material (silicon and iron), and industrial/fuel oil combustion (vanadium and nickel).31 Within the source types, copper and zinc were associated with tire and brake wear and iron was mostly associated with brake abrasion and road dust; silicon was associated with resuspended road dust, and vanadium and nickel were mainly derived from shipping emissions.31 The observed associations of asthma symptoms, rhinitis, allergic sensitization, and lung function with copper, iron, and zinc therefore suggest that traffic affects health not only through exhaust emissions but also through non-exhaust emissions.

Potassium has been suggested as a tracer for biomass burning, like residential wood burning, but potassium is also present in soil.32 The land-use regression models for potassium for the Netherlands, however, were dominated by traffic variables, reflecting resuspension of road dust.10 Therefore, associations of health outcomes with potassium in this study most likely represent effects of traffic-related air pollution.

Also, the associations of health outcomes with PMcoarse, potassium, sulfur, and silicon we observed should be interpreted with caution as the performance of the land-use regression models for these pollutants was relatively poor and consequently observed effect estimates are most likely underestimates of the true effects.

Non-exhaust emissions from traffic arising from abrasion of brake, tire, and road surface material, as well as from resuspension of PM by passing traffic, in contrast to exhaust emissions are currently not regulated by the European Union, despite their considerable contribution to traffic-related ambient PM10.33 As substantial efforts to reduce exhaust emissions have been successful, non-exhaust emissions from traffic are gaining importance.34

Iron and copper in proximity to busy roads have been associated with brake wear,35 and zinc may derive from tire wear.34 In our study, iron and copper, but not zinc, were mainly contained in the coarse fraction of PM10. Exposure contrasts for three elements in PM10 were similar to those of other traffic-related pollutants such as NO2 and PM2.5 absorbance. Asthma symptoms and allergic sensitization were associated with higher iron and copper in PM10 and lower FEF25-75 was associated with higher copper in PM10. Moreover, asthma incidence, symptoms of asthma, and rhinitis were positively associated with zinc in PM10.

Effects of markers of non-exhaust emissions were of similar magnitude as effects of exhaust emission markers (NO2) in this study. If confirmed by other studies, this means that the impact of non-exhaust emissions from traffic on lung function, asthma, and allergy of children may be as large as the impact of exhaust emissions from traffic.

We investigated associations with exposure in early life and more recent exposure. Asthma and hay fever and related symptoms were mainly associated with early life exposure, whereas allergic sensitization was associated with both early life and current exposure, and lung function with current exposure rather than early life exposure. The relevance of exposures in early life as well as later childhood for the development of asthma and allergies has been suggested by earlier analyses in the same cohort.6 However, correlations between levels of the same pollutant at the birth and current address were moderate to high despite the fact that more than 60% of the population had moved at least once during the follow-up. The greater relevance of current exposures for lung function is consistent with earlier findings from this and other cohorts for FEV1, FVC, and PEF7 and is supported by the findings from studies that suggested reversibility of the air pollution effects on FVC, FEF25-75, and PEF.36 The stronger association with mid-expiratory flows than with FEV1 and the lack of an association with FVC in this study suggest that air pollution exposure exerts its effects especially on the small airways rather than the larger airways.

The stronger associations with lung function in non-movers compared with movers suggest that those with the most constant exposure are at greatest risk. However, these findings disagree with findings for lung function at age 6–8 years from this and other cohorts, where associations were slightly stronger in movers.7

The prospective design, the availability of the participants’ residential histories and the detailed air pollution exposure assessment, in particular the availability of individual estimates of residential exposure to PM constituents, are important strengths of our study. Children of highly educated and to a lesser extent children of atopic mothers were overrepresented in the study population and probably as a consequence of this, participants were slightly more often breastfed and less often exposed to environmental tobacco smoke and pets. As the effect estimates remained largely unchanged after adjustment for potential confounders including parental education and parental atopy, this most likely did not result in any bias of the observed associations. Another limitation of this study is that exposure models were based on air pollution measurements performed in 2008–2010 and that this does not capture temporal variability. The measurements were conducted close to the 11- to 12-year follow-up conducted in 2008–2009 (questionnaire) and 2008–2011 (medical examination). Applying the models to the children’s historical addresses, however, we implicitly assume that the spatial patterns did reflect the baseline period of the cohort, ie, 1996–1997. This assumption is supported by four studies that showed that spatial contrasts in measured and modeled annual average NO2 levels were stable over periods of 7–12 years.37–40 Further support for the validity of the estimated air pollution levels for the historical addresses comes from the National Air Quality Monitoring Network: annual average NO2 and PM10 levels were relatively stable between 2000 and 2007, which covers most of the study period.41 For the other pollutants, no such information is available. The findings for NO2 may be applicable to traffic-related constituents such as copper, iron, and zinc. It is, however, not clear to what extent this applies to the other constituents that derive from sources other than traffic. As metals largely relate to non-exhaust emissions of traffic, which are less regulated than exhaust emissions (affecting NO2), changes in concentrations over time are likely less for metals than for NO2. Concentrations of copper, iron, and zinc may even have increased slightly with increasing traffic intensity. Furthermore, we consider it highly unlikely that measurement error for the baseline addresses is differential and would lead to false-positive associations with asthma and allergy. Rather, we would expect the use of the exposure models for historical addresses to introduce random noise and bias exposure––response relations toward the null. Although we performed fairly large numbers of comparisons, we did not adjust for multiple testing as it reduces the probability of a false rejection of the null hypothesis—that there is no association between exposure and outcome—at the cost of increasing the frequency of incorrect statements about statistical non-significance of associations and missing possibly important findings.42 Moreover, multiple testing in situations like the present with several levels of multiplicity, more than one (related) endpoint and several (related) exposures, is complex and no adequate adjustments exist.43 Instead of adjusting for multiple comparisons, we interpreted results on the basis of consistency of effect estimates across elements and PM size fractions.

In conclusion, our study provides evidence for an adverse effect of PM constituents, in particular copper, iron, potassium, silicon, and zinc, on asthma, rhinitis, allergic sensitization, and lung function of schoolchildren. Associations with copper, iron, and zinc may indicate the health relevance of non-exhaust emissions of traffic. This finding is important because the importance of non-exhaust emissions from traffic is increasing because of the substantial efforts to reduce exhaust emissions.

ACKNOWLEDGMENTS

The authors thank all the children and their parents for their cooperation. The authors also thank all the field workers and laboratory personnel involved for their efforts, and Marjan Tewis for data management.

REFERENCES

1. Rückerl R, Schneider A, Breitner S, Cyrys J, Peters A. Health effects of particulate air pollution: a review of epidemiological evidence. Inhal Toxicol. 2011;23:555–592
2. HEI Panel on the Health Effects of Traffic-Related Air Pollution. Traffic-Related Air Pollution: A Critical Review of the Literature on Emissions, Exposure, and Health Effects. Special Report 17. 2010 Boston, MA Health Effects Institute
3. Bråbäck L, Forsberg B. Does traffic exhaust contribute to the development of asthma and allergic sensitization in children: findings from recent cohort studies. Environ Health. 2009;8:17
    4. Götschi T, Heinrich J, Sunyer J, Künzli N. Long-term effects of ambient air pollution on lung function: a review. Epidemiology. 2008;19:690–701
    5. Schwarze PE, Ovrevik J, Låg M, et al. Particulate matter properties and health effects: consistency of epidemiological and toxicological studies. Hum Exp Toxicol. 2006;25:559–579
    6. Gehring U, Wijga AH, Brauer M, et al. Traffic-related air pollution and the development of asthma and allergies during the first 8 years of life. Am J Respir Crit Care Med. 2010;181:596–603
    7. Gehring U, Gruzieva O, Agius RM, et al. Air pollution exposure and lung function in children: the ESCAPE project. Environ Health Perspect. 2013;121:1357–1364
    8. Beelen R, Hoek G, Vienneau D, et al. Development of NO2 and NOx land use regression models for estimating air pollution exposure in 36 study areas in Europe: The ESCAPE project. Atmos Environ.. 2013;72:10–23
    9. Eeftens M, Beelen R, de Hoogh K, et al. Development of land use regression models for PM(2.5), PM(2.5) absorbance, PM(10) and PM(coarse) in 20 European study areas; results of the ESCAPE project. Environ Sci Technol. 2012;46:11195–11205
      10. de Hoogh K, Wang M, Adam M, et al. Development of land use regression models for particle composition in twenty study areas in Europe. Environ Sci Technol. 2013;47:5778–5786
      11. Wijga AH, Kerkhof M, Gehring U, et al. Cohort profile: the prevention and incidence of asthma and mite allergy (PIAMA) birth cohort. Int J Epidemiol.. 2014;43:527–535
      12. Miller MR, Hankinson J, Brusasco V, et al.ATS/ERS Task Force. Standardisation of spirometry. Eur Respir J. 2005;26:319–338
      13. Cyrys J, Eeftens M, Heinrich J, et al. Variation of NO2 and NOx concentrations between and within 36 European study areas: results from the ESCAPE study. Atmos Environ.. 2012;62:374–390
      14. Eeftens M, Tsai MY, Ampe C, et al. Spatial variation of PM2.5, PM10, PM2.5 absorbance and PMcoarse concentrations between and within 20 European study areas and the relationship with NO2: results of the ESCAPE project. Atmos Environ.. 2012;62:303–17
      15. Singer J, Willett J Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. 2003 New York Oxford University Press
      16. Diggle P, Liang K, Zeger SL Analysis of Longitudinal Data. 1994 Oxford Clarendon Press
      17. Moshammer H, Hoek G, Luttmann-Gibson H, et al. Parental smoking and lung function in children: an international study. Am J Respir Crit Care Med. 2006;173:1255–1263
      18. Gehring U, de Jongste JC, Kerkhof M, et al. The 8-year follow-up of the PIAMA intervention study assessing the effect of mite-impermeable mattress covers. Allergy. 2012;67:248–256
      19. Patel MM, Hoepner L, Garfinkel R, et al. Ambient metals, elemental carbon, and wheeze and cough in New York City children through 24 months of age. Am J Respir Crit Care Med. 2009;180:1107–1113
      20. Eeftens M, Hoek G, Gruzieva O, et al. Elemental composition of particulate matter and the association with lung function. Epidemiology. 2014;25:648–657
      21. Hirshon JM, Shardell M, Alles S, et al. Elevated ambient air zinc increases pediatric asthma morbidity. Environ Health Perspect. 2008;116:826–831
      22. Roemer W, Hoek G, Brunekreef B, et al. PM10 elemental composition and acute respiratory health effects in European children (PEACE project): pollution effects on asthmatic children in Europe. Eur Respir J. 2000;15:553–559
      23. Heinrich J, Hoelscher B, Wjst M, Ritz B, Cyrys J, Wichmann H. Respiratory diseases and allergies in two polluted areas in East Germany. Environ Health Perspect. 1999;107:53–62
      24. Heinrich J, Hoelscher B, Frye C, et al. Improved air quality in reunified Germany and decreases in respiratory symptoms. Epidemiology. 2002;13:394–401
      25. Frye C, Hoelscher B, Cyrys J, Wjst M, Wichmann HE, Heinrich J. Association of lung function with declining ambient air pollution. Environ Health Perspect. 2003;111:383–387
      26. Schaumann F, Borm PJ, Herbrich A, et al. Metal-rich ambient particles (particulate matter 2.5) cause airway inflammation in healthy subjects. Am J Respir Crit Care Med. 2004;170:898–903
      27. Dye JA, Lehmann JR, McGee JK, et al. Acute pulmonary toxicity of particulate matter filter extracts in rats: coherence with epidemiologic studies in Utah Valley residents. Environ Health Perspect. 2001;109(Suppl 3):395–403
      28. Frampton MW, Ghio AJ, Samet JM, Carson JL, Carter JD, Devlin RB. Effects of aqueous extracts of PM(10) filters from the Utah valley on human airway epithelial cells. Am J Physiol. 1999;277(5 Pt 1):L960–L967
        29. Ghio AJ, Churg A, Roggli VL. Ferruginous bodies: implications in the mechanism of fiber and particle toxicity. Toxicol Pathol. 2004;32:643–649
          30. Gavett SH, Haykal-Coates N, Copeland LB, Heinrich J, Gilmour MI. Metal composition of ambient PM2.5 influences severity of allergic airways disease in mice. Environ Health Perspect. 2003;111:1471–1477
          31. Viana M, Kuhlbusch T, Querol X, et al. Source apportionment of particulate matter in Europe: a review of methods and results. J Aerosol Sci.. 2008;39:827–49
          32. Reche C, Viana M, Pandolfi M, et al. Urban NH3 levels and sources in a Mediterranean environment. Atmos Environ.. 2012;57:153–164
          33. Bukowiecki N, Lienemann P, Hill M, et al. PM10 emission factors for non-exhaust particles generated by road traffic in an urban street canyon and along a freeway in Switzerland. Atmos Environ.. 2010;44:2330–2340
          34. Thorpe A, Harrison RM. Sources and properties of non-exhaust particulate matter from road traffic: a review. Sci Total Environ. 2008;400:270–282
          35. Majestic BJ, Anbar AD, Herckes P. Elemental and iron isotopic composition of aerosols collected in a parking structure. Sci Total Environ. 2009;407:5104–5109
          36. Avol EL, Gauderman WJ, Tan SM, London SJ, Peters JM. Respiratory effects of relocating to areas of differing air pollution levels. Am J Respir Crit Care Med. 2001;164:2067–2072
          37. Wang R, Henderson SB, Sbihi H, et al. Temporal stability of land use regression models for traffic-related air pollution. Atmos Environ.. 2013;64:312–319
          38. Eeftens M, Beelen R, Fischer P, Brunekreef B, Meliefste K, Hoek G. Stability of measured and modelled spatial contrasts in NO(2) over time. Occup Environ Med. 2011;68:765–770
            39. Cesaroni G, Porta D, Badaloni C, et al. Nitrogen dioxide levels estimated from land use regression models several years apart and association with mortality in a large cohort study. Environ Health. 2012;11:48
              40. Gulliver J, de Hoogh K, Hansell A, Vienneau D. Development and back-extrapolation of NO2 land use regression models for historic exposure assessment in Great Britain. Environ Sci Technol. 2013;47:7804–7811
                41. Beijk R, Mooibroek D, Hoogerbrugge R Air quality in the Netherlands 2007. RIVM report 680704005. 2008 Bilthoven, The Netherlands: National Institute for Public Health and the Environment
                42. Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology. 1990;1:43–46
                43. Bender R, Lange S. Adjusting for multiple testing–when and how? J Clin Epidemiol. 2001;54:343–349

                Supplemental Digital Content

                Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.