You are here

Article Text

Article menu

Download PDFPDF

Air pollution
Effects of long-term exposure to PM10 and NO2 on asthma and wheeze in a prospective birth cohort
  1. Anna Mölter1,
  2. Raymond Agius1,
  3. Frank de Vocht1,
  4. Sarah Lindley2,
  5. William Gerrard3,
  6. Adnan Custovic4,
  7. Angela Simpson4
  1. 1Centre for Occupational & Environmental Health, Centre for Epidemiology, Institute of Population Health, Manchester Academic Health Sciences Centre, The University of Manchester, Manchester, UK
  2. 2School of Environment and Development (Geography), The University of Manchester, Manchester, UK
  3. 3Salford Lung Study, North West e-Health, Salford, UK
  4. 4The University of Manchester, Manchester Academic Health Science Centre, Institute of Inflammation and Repair, University Hospital of South Manchester NHS Foundation Trust, Wythenshawe Hospital, Manchester, UK
  1. Correspondence to Dr Anna Mölter, Centre for Occupational & Environmental Health, Centre for Epidemiology, Institute of Population Health, Manchester Academic Health Sciences Centre, Room C4.19 Ellen Wilkinson Building, Oxford Road, Manchester M13 9PL, UK; anna.molter{at}manchester.ac.uk

Abstract

Background Epidemiological studies on the effect of urban air pollution on childhood asthma have shown conflicting results and so far no consistent association has emerged. However, a common limitation in previous studies has been exposure misclassification leading to uncertainties in risk estimates.The aim of this study was to analyse the effects of long-term exposure to particulate matter (PM10) and nitrogen dioxide (NO2) on the prevalence of asthma and wheeze within a population-based birth cohort—the Manchester Asthma and Allergy Study (MAAS).

Methods The prevalence of asthma and current wheeze within the cohort (N=1185) was determined through parental questionnaires at ages 3, 5, 8 and 11 years. The typical monthly PM10 and NO2 exposure of each child was estimated through a novel microenvironmental exposure model from birth to age 11. The association between exposure and asthma or wheeze was analysed using generalised estimating equations and multiple logistic regression.

Results The range of asthma prevalence was 15.2–23.3%, with the lowest prevalence at age 3 and the highest at age 5. The prevalence of current wheeze decreased from ages 3 to 8 (23.7–18%). The mean NO2 exposure decreased from the 1st year of life (21.7 µg/m3) to the 11th year of life (16.0 µg/m3). The mean PM10 exposure showed a smaller decrease (12.8 –10.7 µg/m3). The statistical analysis showed no significant association between the exposures and either outcome.

Conclusions No evidence of a significant association between long-term exposure to PM10 and NO2 and the prevalence of either asthma or wheeze was found.

  • Asthma
  • Air Pollution
  • Child Health
  • Environmental Epidemiology

https://doi.org/10.1136/jech-2013-202681

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Asthma is the most common chronic disease during childhood with multiple environmental and genetic risk factors.1 The UK has the highest prevalence of childhood asthma within Europe2 and physician diagnosed asthma in children has increased twofold to threefold from 1970 to 2000.3 The amount of road traffic in the UK has increased fivefold from 1950 to 2008,4 resulting in an increase in air pollutants such as nitrogen dioxide (NO2) and particulate matter (PM) until the 1990s, when the introduction of catalytic converters led to a decrease in air pollution in the UK. This parallel rise in asthma prevalence and road traffic makes traffic-related pollution a potential risk factor for the development of asthma.

    Epidemiological studies of air pollution and asthma have been heterogeneous in design, health outcomes and exposure assessment used and have provided conflicting results.5 Within asthma sufferers, increasing exposure to NO2 and PM is associated with more severe disease.6 However, most studies found no association between these pollutants and asthma prevalence or incidence during childhood,7–11 although a few did report associations, often in post hoc analysis of subgroups. Therefore, whether air pollution exposure is associated with childhood asthma prevalence remains unclear.

    Recent reviews have highlighted the need for improved exposure assessment methods in epidemiological studies.5 ,12 Although ideally one would monitor personal exposure throughout life, this is clearly not feasible and alternative methods of assigning an index of exposure have been used, for example, central outdoor monitors, land use regression (LUR) models or temporary indoor static monitors. Each methodology has its strengths and weaknesses, which can lead to exposure misclassification, adding uncertainty to the findings of epidemiological studies: using only outdoor estimates may overestimate exposure of children living in industrialised countries as they spend the majority of their time indoors.13 Indoor NO2 and PM10 concentrations tend to be lower than outdoor concentrations in UK homes.14 However, previous studies have shown large increases in indoor NO2 and PM10 concentrations, if indoor sources such as gas cookers and cigarette smoke are present.15–17 At least 60% of homes in the UK contain gas cooking appliances14 and the prevalence of adult smokers in the UK was 20% in 2010;18 therefore, indoor concentrations may make an important contribution to the overall exposure of individuals. We have developed and validated a novel exposure assessment method, the microenvironmental exposure model (MEEM),19–21 which models indoor and outdoor exposures longitudinally. A performance evaluation of this model against personal monitoring data showed a lower mean prediction error compared to methods such as LUR or central monitors.21

    The aims of this study were to estimate time-weighted long-term exposure to PM10 and NO2 of children participating in the Manchester Asthma and Allergy Study (MAAS)22 and to analyse effects of exposure on asthma and current wheeze prevalence at ages 3, 5, 8 and 11 years. We hypothesised that children exposed to higher PM10 or NO2 concentrations may be more likely to have asthma than children exposed to low concentrations.

    Methods

    Study population

    MAAS is a prospective birth cohort based in Greater Manchester, located in the Northwest of England22 (see online supplementary figure A1, appendix 1). Participants were recruited while attending antenatal clinics at two local hospitals between 1995 and 1997. Children were reviewed at ages 3, 5, 8 and 11 years. At each review, parents completed a nurse-administered validated questionnaire. Approval to conduct MAAS has been granted by the local Research Ethics Committee (SOU/00/258; SOU/00/259) and the study is registered as ISRCTN72673620. Parents of all children participating in the study have provided written informed consent.

    Clinical outcomes

    Asthma: defined as at least two positive answers to the following three questions: (1) doctor diagnosis of asthma ever; (2) child having wheezed during the previous 12 months and (3) child having received asthma medication during the previous 12 months.

    Current wheeze: a positive answer to “Has your child ever had wheezing or whistling in the chest in the last 12 months?”.

    Modelled exposure to PM10 and NO2

    Children's exposure was estimated using a novel model ‘MEEM’ described in a previous publication.21 The model assumes that children spend the majority of their time in three types of microenvironments (MEs): the child's home (kitchen, living room, bedroom), the child's school and the journey between home and school (see online supplementary table A1, appendix 1). Each of these MEs can be in an indoor or an outdoor ME and different methods were used to model air pollution in each ME: PM10 and NO2 concentrations in all outdoor MEs were estimated through land use regression models described in detail in previous publications.19 ,20 Concentrations in the kitchen, living room and child's bedroom were estimated through the INDAIR model, which is a validated mass balance model, specifically designed for residential buildings within the UK.17 Concentrations in the journey indoor and school MEs were estimated using published indoor to outdoor ratios23 ,24 (see online supplementary table A1, appendix 1).

    A timeline of each child's home and school addresses, as well as the mode of transport between the two, was collected through parental questionnaires at the age 11 review. The route of the journey from home to school was estimated using a shortest path analysis in ArcMap 9.2 (ESRI Inc). The time a child typically spent in each ME was derived from a parental questionnaire at the age 11 review.

    This questionnaire also collected essential input information for the INDAIR model, including room sizes, air exchange rates, presence of gas cooking appliances, typical cooking duration and the number of cigarettes smoked in the home. However, this information was only applicable for the most recent home of each child and it was not feasible to collect such detailed input information for the entire lifetime of each child. Therefore, MEEM was used to estimate exposure for the most recent exposure time period, that is, the summer and winter prior to the age 11 review. Separate exposure estimates were calculated for summer and winter, because we assumed that the air exchange rates in children's homes would differ in summer and winter. We hypothesized that during the winter, the windows would be mainly closed, while during the summer they might be open for extended periods of time. The parental questionnaire described above contained separate questions for winter and summer regarding the windows being open or closed.

    To estimate exposure for the whole lifetime of the children, MEEM was slightly simplified by replacing the INDAIR model with an indoor to outdoor ratio (see online supplementary table A1, appendix 1) and by assuming that the children attended school from 9:00 to 15:00 on weekdays. For clarity, this simplified version of MEEM is referred to as the ‘Lifetime model’; however, it should be noted that the two models are almost identical and that the concentration estimates only differ in the home indoor environment.

    For analytical purposes, the exposure estimates from the ‘Lifetime model’ were averaged into the following aetiologically relevant time windows: first year of life (age 0–1), birth to review (age 0–3, 0–5, 0–8, 0–11), 1 year prior to review (ages 2–3, 4–5, 7–8, 10–11).

    Statistical analysis

    The association between PM10 and NO2 exposure and asthma and current wheeze prevalence at each review was analysed using multiple logistic regression. To analyse the association between the above exposures and health outcomes over time, generalised estimating equations were used to adjust for the within-subject correlation of repeated measures. Gender and age were included in all analyses; at age 11, the Tanner stage was included to adjust for pubertal development. A list of potential confounding variables (identified in MAAS and the published literature) is shown in online supplementary appendix 1. Initially, we carried out univariate analyses of the potential confounding variables, the clinical outcomes and the exposures to identify variables of further interest. This was followed by a stepwise process to assess the variability explained by each potential confounder within the exposure-outcome models. We decided a priori to include gender and age in all models. Of the remaining variables, only those with a significant effect (p<0.05) were included in the final models. All statistical analyses were carried out in SPSS V.16.0.

    Results

    Characteristics of the study population

    The number of children participating in each review, the corresponding asthma and current wheeze prevalence and the number of exposure estimates available at each time window are shown in figure 1. The children included in this analysis did not differ from those who were excluded, apart from being more likely to have older siblings (55% vs 52.3%) and less likely to have a dog (13.2% vs 15.4% at age 5; 16.8% vs 19.3% at age 8 years; table 1).

    View this table:
    Table 1

    Overview and descriptive statistics of potential confounding variables

    Figure 1
    Figure 1

    Overview of the Manchester Asthma and Allergy Study cohort showing participation rates at each review, the prevalence of asthma and current wheeze and the number of exposure estimates available at different time windows.

    Exposure to pollutants

    The mean PM10 and NO2 concentrations for the different exposure windows are shown in table 2. Annual mean exposure estimates decreased from the 1st year of life to the 11th year of life (ΔPM10=−2.1 µg/m3; ΔNO2=−5.7 µg/m3), reflecting a general decrease in the annual mean NO2 and PM10 concentrations in the study area.25 The SD of most exposure estimates were relatively small, indicating relatively low variability within the lifetime exposure data. However, the mean and SD of the summer and winter exposures estimates were higher, indicating greater variability due to peak exposures in the home indoor ME estimated by INDAIR. The PM10 and NO2 concentrations were moderately to strongly correlated across all exposure windows (Pearson's r=0.59–0.89).

    View this table:
    Table 2

    Distribution of PM10 and NO2 exposures (in µg/m3)

    Pollution and asthma or current wheeze

    The final models were adjusted for gender, age, body mass index, paternal income at birth, sensitisation, family history of asthma, hospitalisation during the first 2 years of life and smoking within the child's home during the first year of life. At age 11, the mean Tanner stage was used as an additional covariate to identify pubescent children. Asthma prevalence was not associated with exposure to PM10 and NO2 during the first year of life (figure 2A), during the lifetime (figure 2B) or during the year prior to the review (figure 2C), in either the cross-sectional or longitudinal models. Furthermore, there was no association between a child having asthma at age 11 and their PM10 and NO2 exposure during the winter and summer prior to the age 11 review (figure 3).

    Figure 2
    Figure 2

    Adjusted cross-sectional and longitudinal associations between asthma prevalence and long-term particulate matter and nitrogen dioxide exposure during different time windows (‘Lifetime model’). ORs were calculated per 1 µg/m3 increase in exposure.

    Figure 3
    Figure 3

    Adjusted association between asthma at age 11 and particulate matter and nitrogen dioxide exposure during the summer and winter before the age 11 review (microenvironmental exposure model). ORs were calculated per 1 µg/m3 increase in exposure.

    The results of the analyses of current wheeze prevalence and PM10 and NO2 exposure are shown in online supplementary figures A2 and A3 (appendix 2). These figures show similar trends as the figures for asthma and none of the analyses with current wheeze indicated a significant association with air pollution exposure. In addition, we have also included results of analyses of asthma incidence and PM10 and NO2 exposure in online supplementary appendix 2 (figures A4 and A5). However, these were not materially different from the above results for asthma prevalence.

    Discussion

    This is the first study to model indoor and outdoor air pollution exposure, providing time-weighted estimates of PM10 and NO2 exposure for multiple time windows throughout primary school age, and to analyse the effects of exposure on childhood asthma and wheeze prevalence from birth to age 11. Using this comprehensive exposure model, we found no association between PM10 or NO2 and asthma or wheeze within a population-based birth cohort located in the Northwest of England.

    Previous cohort studies on air pollution and childhood asthma have reported conflicting findings. As in our study, several cohort studies on the prevalence of asthma and wheeze in children also found no association with outdoor NO2 exposure.9 ,10 ,26 ,27 However, while some studies found an association between outdoor NO2 and asthma or current wheeze prevalence in schoolchildren,28 ,29 others only found this in long-term resident girls30 or when relative humidity31 was included in the model. Similar to our study, a birth cohort study in Oslo10 also found no association between modelled outdoor NO2 exposure during the first year of life and asthma prevalence at age 9–11 years. Furthermore, a birth cohort study in Stockholm, the Barn Allergy Milieu Stockholm Epidemiology (BAMSE), found no significant association between modelled NOx exposure during the first year of life or lifetime and asthma prevalence at age 1, 2, 4, 8 or 12 years.32 However, in contrast to our findings, a birth cohort study in the Netherlands, the Prevention and Incidence of Asthma and Mite Allergy (PIAMA) study,33 found a significant association between exposure at the birth address and asthma prevalence at age 5 years and longitudinal asthma (age 1–8 years). Fewer studies have been carried out on the association between outdoor PM concentrations and asthma or current wheeze in children. In keeping with our results, most of these studies found no association with asthma or wheeze.8 ,11 ,27 ,30 ,34–36 Of the European birth cohort studies, the BAMSE study also found no significant association between PM10 exposure during the first year of life or lifetime and asthma prevalence at age 1, 2, 4, 8 or 12 years,32 further supporting our results. However, the PIAMA study found an association between outdoor particulate matter exposure and asthma and current wheeze prevalence;26 ,33 but it should be noted that they analysed the effects of smaller particles (PM2.5), which may explain some of the discrepancy compared to our results Nevertheless, a previous cohort study in the UK of 1-year-old to 5-year-old children found a significant association between modelled outdoor PM10 exposure and current wheeze prevalence.37 Statistically significant associations were only present in adjusted regression models, possibly indicating a strong influence of a covariate within the models. Outside of Europe, a Canadian case–control study of young children (age 0–4 years old)38 found an association between outdoor PM10 exposure and asthma prevalence. This study used health records, and therefore information on potential confounding variables such as smoking or parental asthma was limited.

    Of the studies focusing on indoor air pollution exposure, a study in Japan39 found an association between indoor NO2 exposure and the prevalence of asthma in children, but several other published studies found no significant association between measured indoor NO2 concentrations and childhood asthma or wheeze prevalence.40–43 It is known that children living in industrialised countries spend more than 80% of their time indoors.13 Therefore, indoor air pollution exposure may be a better proxy of personal exposure than outdoor exposure and associations found in previous studies based on outdoor exposure may have been affected by exposure misclassification. Although the aim of our study was to analyse time-weighted exposures, we did carry out a small number of ‘spot checks’ with modelled residential outdoor exposure estimates (see online supplementary table A2, appendix 2). These ‘spot checks’ did not indicate an association between outdoor exposure and asthma prevalence either. Therefore, it remains unclear to what extent previous findings based solely on outdoor exposure estimates may have been affected by exposure misclassification.

    A strength of this study was the novel exposure model used. Although we cannot conclude that exposure misclassification was completely absent from this study, our exposure assessment method eliminated common sources of exposure misclassification by including home and school addresses and by modelling indoor and outdoor exposures. A further advantage of the exposure model was that it provided exposure estimates for different exposure time windows. There is currently no scientific consensus on which exposure window is most likely to have an effect on asthma in children.33 While some studies suggest that exposure during early life is most important,38 others found that lifetime exposure had an effect.33 However, none of the exposure windows used in this study showed a significant association with asthma or wheeze and no trend was apparent.

    A limitation of this study was the number of children with a full set of exposure estimates available for the duration of the study (eg, 373 children for the 0–11 years exposure window) compared to the total size of the MAAS cohort (eg, 927 children at age 11, figure 1). This limited availability was due to the following two reasons: (1) the children's address history and school history were collected retrospectively at the age 11 clinical follow-up. However, some of the address/school histories provided were incomplete, resulting in some missing data. (2) Children's home and/or school addresses were outside Greater Manchester and therefore the land use regression model19 ,20 could not be used. Approximately 20% of the children had lived outside this geographic area for part of the follow-up period and hence had to be excluded from the analysis during these time periods. Nevertheless, to the author's knowledge, this is the largest study until now to use a detailed microenvironmental model to estimate PM10 and NO2 exposure in children over an 11 year time period.

    A limitation in previous studies may have been the sole use of the question ‘has a doctor ever diagnosed asthma in your child?’ to identify the asthma phenotype. In children under the age of 3, it can be difficult to distinguish asthma from other respiratory illnesses.27 Therefore, this question may misclassify some children. In contrast, our study used three variables to identify asthmatic children, which is thought to reduce misclassification and to improve the reliability of our findings.44

    The aim of the current study was to investigate the association between the modelled exposure to two pollutants (NO2 and PM10) and asthma prevalence at a population level. Real-life exposures to pollutants are more complex, as many sources emit a range of pollutants and mixtures will vary over space and time, particularly between the outdoor and the indoor environment. While it was beyond the remit of the current study to model all exposures with a high degree of resolution, we acknowledge that other unmeasured pollutants may have an effect on asthma prevalence. In addition, we chose prevalence rather than incidence as our outcome measure; we emphasise that for incident asthma the results are not materially different (see online supplementary appendix 2, figure A4 and A5). As MAAS is a population-based birth cohort, asthma exacerbations were relatively infrequent events precluding any assessment of the association between exposure and exacerbations. We contrast our findings with those seen in asthma populations where increasing exposure is frequently associated with an increase in risk of exacerbations.45

    In conclusion, using a comprehensive exposure assessment method, modelling indoor and outdoor exposures to NO2 and PM10 longitudinally with time weighting, we found no association between pollutant exposure and prevalence of asthma or wheeze. However, these results should not be extrapolated to geographical areas, where the range of exposures is much higher, and they do not imply that in those with pre-existing asthma, exposures at UK levels are safe. Replication of this comprehensive exposure assessment method should now be performed in larger populations.

    What is already known on this subject

    • Previous studies on air pollution exposure and childhood asthma have shown conflicting results and no consistent association has emerged. Exposure misclassification is thought to be a common limitation and may have contributed to inconsistencies in previous findings.

    What this study adds

    • This study aimed to reduce exposure misclassification by modelling indoor and outdoor air pollution exposure in multiple microenvironments and through the use of time-weighted estimates of particulate matter and nitrogen dioxide exposure. The comprehensive exposure model provided longitudinal exposure for multiple time windows throughout primary school age, which enabled repeated cross-sectional and longitudinal analyses of air pollution and childhood asthma prevalence.

    Acknowledgments

    The authors would like to thank the families who participated in MAAS and all members of the MAAS study team for their tireless effort, in particular Joseph Nathan and Mandy Mycock who carried out the data quality checks. The exposure model used in this study was partly based on the Greater Manchester air dispersion modelling study carried out by the former Atmospheric Research and Information Centre on behalf of the local authorities of Greater Manchester. Therefore, they would also like to thank the Manchester Area Pollution Advisory Council for permitting us to access this data. Furthermore, they are very grateful to Prof M Ashmore, Dr S Dimitroulopoulou and Dr A Terry for permitting us access to the INDAIR model17 and for their helpful advice on the use of this model.

    References

    1. ↵
      Global strategy for asthma management and prevention. Global Initiative for Asthma, 2010. http://www.ginasthma.org/local/uploads/files/GINA_Report_2010_1.pdf
    2. ↵
      1. Asher MI,
      2. Montefort S,
      3. Björkstén B,
      4. et al
      . Worldwide time trends in the prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and eczema in childhood: ISAAC phases one and three repeat multicountry cross-sectional surveys. Lancet 2006;368:733–43.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Anderson HR,
      2. Gupta R,
      3. Strachan DP,
      4. et al
      . 50 years of asthma: UK trends from 1955 to 2004. Thorax 2007;62:85–90.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      Road statistics 2008: traffic, speeds and congestion. Department for Transport, 2009. http://www.dft.gov.uk/adobepdf/162469/221412/221546/226956/261695/roadstats08tsc.pdf
    5. ↵
      1. Braback L,
      2. Forsberg B
      . Does traffic exhaust contribute to the development of asthma and allergic sensitization in children: findings from recent cohort studies. Environ Health 2009;8:17–28.
      OpenUrlCrossRefPubMed
    6. ↵
      1. McCormack MC,
      2. Breysse PN,
      3. Matsui EC,
      4. et al
      . In-home particle concentrations and childhood asthma morbidity. Environ Health Perspect 2009;117:294–8.
      OpenUrlPubMedWeb of Science
    7. ↵
      1. Hirsch T,
      2. Weiland SK,
      3. von Mutius E,
      4. et al
      . Inner city air pollution and respiratory health and atopy in children. Eur Respir J 1999;14:669–77.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Janssen NAH,
      2. Brunekreef B,
      3. van Vliet P,
      4. et al
      . The relationship between air pollution from heavy traffic and allergic sensitization, bronchial hyperresponsiveness, and respiratory symptoms in Dutch schoolchildren. Environ Health Perspect 2003;111:1512–18.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Kramer U,
      2. Sugiri D,
      3. Ranft U,
      4. et al
      . Eczema, respiratory allergies, and traffic-related air pollution in birth cohorts from small-town areas. J Dermatol Sci 2009;56:99–105.
      OpenUrlCrossRefPubMed
    10. ↵
      1. Oftedal B,
      2. Nystad W,
      3. Brunekreef B,
      4. et al
      . Long-term traffic-related exposures and asthma onset in schoolchildren in Oslo, Norway. Environ Health Perspect 2009;117:839–44.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Penard-Morand C,
      2. Charpin D,
      3. Raherison C,
      4. et al
      . Long-term exposure to background air pollution related to respiratory and allergic health in schoolchildren. Clin Exp Allergy 2005;35:1279–87.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Arrandale VH,
      2. Brauer M,
      3. Brook JR,
      4. et al
      . Exposure assessment in cohort studies of childhood asthma. Environ Health Perspect 2011;119:591–7.
      OpenUrlPubMedWeb of Science
    13. ↵
      1. Infante-Rivard C
      . Childhood asthma and indoor environmental risk factors. Am J Epidemiol 1993;137:834–44.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Coward SKD,
      2. Llewellyn JW,
      3. Raw GJ,
      4. et al
      . Indoor air quality in homes in England. CRC Ltd, 2001.
    15. ↵
      1. Dimitroulopoulou C,
      2. Ashmore MR,
      3. Byrne MA,
      4. et al
      . Modelling of indoor exposure to nitrogen dioxide in the UK. Atmospheric Environ 2001;35:269–79.
      OpenUrlCrossRefWeb of Science
    16. ↵
      1. Dimitroulopoulou C,
      2. Ashmore MR,
      3. Byrne MA
      . Modelling the contribution of passive smoking to exposure to PM10 in UK Homes. Indoor Built Environ 2001;10:209–13.
      OpenUrlCrossRefWeb of Science
    17. ↵
      1. Dimitroulopoulou C,
      2. Ashmore MR,
      3. Hill MTR,
      4. et al
      . INDAIR: a probabilistic model of indoor air pollution in UK homes. Atmospheric Environ 2006;40:6362–79.
      OpenUrlCrossRefWeb of Science
    18. ↵
    19. ↵
      1. Mölter A,
      2. Lindley S,
      3. de Vocht F,
      4. et al
      . Modelling air pollution for epidemiologic research—Part I: A novel approach combining land use regression and air dispersion. Sci Total Environ 2010;408:5862–9.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Mölter A,
      2. Lindley S,
      3. de Vocht F,
      4. et al
      . Modelling air pollution for epidemiologic research—Part II: predicting temporal variation through land use regression. Sci Total Environ 2010;409:211–17.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Mölter A,
      2. Lindley S,
      3. de Vocht F,
      4. et al
      . Performance of a microenviromental model for estimating personal NO2 exposure in children. Atmospheric Environ 2012;51:225–33.
      OpenUrlCrossRefWeb of Science
    22. ↵
      1. Custovic A,
      2. Simpson BM,
      3. Murray CS,
      4. et al
      . The National Asthma Campaign Manchester Asthma and Allergy Study. Pediatr Allergy Immunol 2002;13:32–7.
      OpenUrlCrossRefPubMed
    23. ↵
      In-car air pollution: the hidden threat to automobile drivers. Washington, DC: International Center For Technology Assessment, 2000. http://www.andrewkimbrell.org/doc/In-car%20pollution%20report.pdf
    24. ↵
      1. Stranger M,
      2. Krata A,
      3. Kontozova-Deutsch V,
      4. et al
      . Monitoring of NO2 in the ambient air with passive samplers before and after a road reconstruction event. Microchem J 2008;90:93–8.
      OpenUrlCrossRefWeb of Science
    25. ↵
      DEFRA. Automatic Monitoring Data. http://uk-air.defra.gov.uk/data/data_selector (accessed 12/01/2012).
    26. ↵
      1. Brauer M,
      2. Hoek G,
      3. Smit HA,
      4. et al
      . Air pollution and development of asthma, allergy and infections in a birth cohort. Eur Respir J 2007;29:879–88.
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Brauer M,
      2. Hoek G,
      3. van Vliet P,
      4. et al
      . Air Pollution from traffic and the development of respiratory infections and asthmatic and allergic symptoms in children. Am J Respir Crit Care Med 2002;166:1092–8.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Gauderman WJ,
      2. Avol E,
      3. Lurmann F,
      4. et al
      . Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology 2005;16:737–43.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Nicolai T,
      2. Carr D,
      3. Weiland SK,
      4. et al
      . Urban traffic and pollutant exposure related to respiratory outcomes and atopy in a large sample of children. Eur Respir J 2003;21:956–63.
      OpenUrlAbstract/FREE Full Text
    30. ↵
      1. Kim JJ,
      2. Smorodinsky S,
      3. Lipsett M,
      4. et al
      . Traffic-related air pollution near busy roads—the East Bay children's respiratory health study. Am J Respir Crit Care Med 2004;170:520–6.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Jerrett M,
      2. Shankardass K,
      3. Berhane K,
      4. et al
      . Traffic-related air pollution and asthma onset in children: a prospective cohort study with individual exposure measurement. Environ Health Perspect 2008;116:1433–8.
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      1. Gruzieva O,
      2. Bergstrom A,
      3. Hulchiy O,
      4. et al
      . Exposure to air pollution from traffic and childhood asthma until 12 years of age. Epidemiology 2013;24:54–61.
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Gehring U,
      2. Wijga AH,
      3. Brauer M,
      4. 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.
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Gehring U,
      2. Cyrys J,
      3. Sedlmeir G,
      4. et al
      . Traffic-related air pollution and respiratory health during the first 2 yrs of life. Eur Respir J 2002;19:690–8.
      OpenUrlAbstract/FREE Full Text
    35. ↵
      1. McConnell R,
      2. Islam T,
      3. Shankardass K,
      4. et al
      . Childhood incident asthma and traffic-related air pollution at home and school. Environ Health Perspect 2010;118:1021–6.
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. Morgenstern V,
      2. Zutavern A,
      3. Cyrys J,
      4. et al
      . Respiratory health and individual estimated exposure to traffic-related air pollutants in a cohort of young children. Occup Environ Med 2007;64:8–16.
      OpenUrlAbstract/FREE Full Text
    37. ↵
      1. Pierse N,
      2. Rushton L,
      3. Harris RS,
      4. et al
      . Locally generated particulate pollution and respiratory symptoms in young children. Thorax 2006;61:216–20.
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. Clark NA,
      2. Demers PA,
      3. Karr CJ,
      4. et al
      . Effect of early life exposure to air pollution on development of childhood asthma. Environ Health Perspect 2010;118:284–90.
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      1. Shima M,
      2. Adachi M
      . Effect of outdoor and indoor nitrogen dioxide on respiratory symptoms in schoolchildren. Int J Epidemiol 2000;29:862–70.
      OpenUrlAbstract/FREE Full Text
    40. ↵
      1. Kramer U,
      2. Koch T,
      3. Ranft U,
      4. et al
      . Traffic-related air pollution is associated with atopy in children living in urban areas. Epidemiology 2000;11:64–70.
      OpenUrlCrossRefPubMedWeb of Science
    41. ↵
      1. Neas LM,
      2. Dockery DW,
      3. Ware JH,
      4. et al
      . Association of indoor nitrogen dioxide with respiratory symptoms and pulmonary function in children. Am J Epidemiol 1991;134:204–19.
      OpenUrlAbstract/FREE Full Text
    42. ↵
      1. Pilotto LS,
      2. Douglas RM,
      3. Attewell RG,
      4. et al
      . Respiratory effects associated with indoor nitrogen dioxide exposure in children. Int J Epidemiol 1997;26:788–96.
      OpenUrlAbstract/FREE Full Text
    43. ↵
      1. Raaschou-Nielsen O,
      2. Hermansen MN,
      3. Loland L,
      4. et al
      . Long-term exposure to indoor air pollution and wheezing symptoms in infants. Indoor Air 2010;20:159–67.
      OpenUrlCrossRefPubMedWeb of Science
    44. ↵
      1. Haland G,
      2. Carlsen KCL,
      3. Sandvik L,
      4. et al
      . Reduced lung function at birth and the risk of asthma at 10 Years of age. New England Journal of Medicine 2006;355:1682–9.
      OpenUrlCrossRefPubMedWeb of Science
    45. ↵
      1. Belanger K,
      2. Holford TR,
      3. Gent JF,
      4. et al
      . Household levels of nitrogen dioxide and pediatric asthma severity. Epidemiology 2013;24:320–30.
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • ).

    • Contributors RA, AS and AC contributed to the study conception and design and obtained funding; AM, SL, RA and FdV developed the exposure model; AM and WG contributed to the acquisition of data; AM and AS analysed the data; AM and AS drafted the manuscript and all authors contributed to the revision of the report and approved the version submitted.

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval Local Research Ethics Committee.

    • Provenance and peer review Not commissioned; externally peer reviewed.