Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Childhood allergic rhinitis : the role of the environment and genetics Fuertes, Elaine Isabelle 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2014_november_fuertes_elaine.pdf [ 3.2MB ]
Metadata
JSON: 24-1.0135589.json
JSON-LD: 24-1.0135589-ld.json
RDF/XML (Pretty): 24-1.0135589-rdf.xml
RDF/JSON: 24-1.0135589-rdf.json
Turtle: 24-1.0135589-turtle.txt
N-Triples: 24-1.0135589-rdf-ntriples.txt
Original Record: 24-1.0135589-source.json
Full Text
24-1.0135589-fulltext.txt
Citation
24-1.0135589.ris

Full Text

CHILDHOOD ALLERGIC RHINITIS: THE ROLE OF THE ENVIRONMENT AND GENETICS byElaine Isabelle FuertesB.Sc., Queen's University, 2007M.Sc., The University of British Columbia, 2009A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Population and Public Health)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)October 2014© Elaine Isabelle Fuertes, 2014AbstractAllergic rhinitis is a global health problem that causes major illness and disability. Inherited and environmental factors influence its development. This thesis examined the role of traffic-related air pollution, genetic variants and their potential interactions, on childhood allergic rhinitis. Global spatial associations with climatic factors known to influence aeroallergen distributions were also studied. Data from two Canadian (CAPPS and SAGE) and four European birth cohorts (BAMSE, GINIplus, LISAplus and PIAMA) participating in the Traffic, Asthma and Genetics collaborationwere pooled. No consistent associations between individual-level traffic-related air pollutants (NO2, PM2.5 mass, PM2.5 absorbance and ozone) estimated to the home address and childhood allergic rhinitis were observed in a longitudinal analysis (up to ten years) of two cohorts (GINIplus and LISAplus; N=6,604) and a pooled analysis of all six cohorts (N=15,299). These latter null associations were not modified by ten tested single nucleotide polymorphisms in the GSTP1, TNF, TLR2 and TLR4 genes. Although these results do not support an adverse role of traffic-related air pollution on childhood allergic rhinitis, much remains to be learned regarding for whom, when and how air pollution may impact disease.In further analyses, genetic variants in the TNF and TLR4 genes and at the 17q21 gene locus were found to be associated with childhood allergic rhinitis in pooled analyses of the six cohorts. As genetic variability in these regions has also been linked to asthma, the observed associations support the hypothesis of shared genetic susceptibility between asthma and allergic rhinitis. These results may be important for public health given the large proportion of the population carrying the studied risk variants.Lastly, using cross-sectional data from 6-7 and 13-14 year-olds participating in the International Study of Asthma and Allergies in Childhood, several ecological spatial associations between climatic factors (temperature, precipitation and vapour pressure) and intermittent and persistent iirhinitis symptom prevalences were identified. Although not conclusive, these results represent a first step in investigating how climate change may affect rhinitis symptom prevalence.Collectively, this dissertation contributes to our understanding of the effects of air pollution, genetic variability and climate on childhood allergic rhinitis. iiiPrefaceUnder the guidance of my thesis committee, I developed the overall research plan for this dissertation. Four research chapters (Chapters 2-5) were written as independent manuscripts intended for publication in peer-reviewed journals. For each research chapter, I developed the study design, conducted all analyses, interpreted the results and prepared and revised the manuscript in accordance with the suggestions provided by my committee and co-authors, whoserespective contributions are detailed below. As the research incorporated in this dissertation brings together many different and large data sources, several co-authors are included on each manuscript. This research was approved by the University of British Columbia Behavioural Research Ethics Board (certificate number: H09-02495; project title: “Traffic-related air pollution as a risk factor for the development of childhood asthma”). A version of Chapter 2 has been published [Fuertes E, Standl M, Cyrys J, Berdel D, von Berg A, Bauer CP, Krämer U, Sugiri D, Lehmann I, Koletzko S, Carlsten C, Brauer M, and Heinrich J. 2013. A longitudinal analysis of associations between traffic-related air pollution with asthma, allergies and sensitization in the GINIplus and LISAplus birth cohorts. PeerJ. 1:e193; doi:10.7717/peerj.193]. Joachim Heinrich and Marie Standl were involved in the conceptualization of the project and contributed to data collection and manuscript edits. Michael Brauer and Chris Carlsten were involved in the conceptualization of the project and manuscript edits. All other co-authors contributed to data collection and manuscript edits. A version of Chapter 3 has been published [Fuertes E, Brauer M, MacIntyre E, Bauer M, Bellander T, von Berg A, Berdel D, Brunekreef B, Chan-Yeung M, Cramer C, Gehring U, Herbarth O, Hoffmann B, Kerkhof M, Koletzko S, Kozyrskyj A, Kull I, Heinrich J, Melén E, Pershagen G, Postma D, Tiesler CM, Carlsten C. 2013. Childhood allergic rhinitis, traffic-relatedair pollution, and variability in the GSTP1, TNF, TLR2, and TLR4 genes: Results from the TAG Study. J Allergy Clin Immunol. 132(2):342–352; doi:10.1016/j.jaci.2013.03.007]. Michael Brauer, Elaina MacIntyre and Chris Carlsten were involved in the conceptualization of the ivproject and contributed to data collection and manuscript edits. All other co-authors contributed to data collection and manuscript edits. A version of Chapter 4 has been published [Fuertes E, Söderhäll C, Acevedo N, Becker A, Brauer M, Chan-Yeung M, Dijk FN, Heinrich J, de Jongste J, Koppelman GH, Postma DS, Kere J, Kozyrskyj AL, Pershagen G, Sandford AJ, Standl M, Tiesler CMT, Waldenberger M, WestmanM, Carlsten C, Melén E. 2014. Associations between the 17q21 region and allergic rhinitis in five birth cohorts. J Allergy Clin Immunol. [in press] doi:10.1016/j.jaci.2014.08.016]. Michael Brauer, Chris Carlsten, Göran Pershagen and Erik Melén were involved in the conceptualization of the project and contributed to data collection and manuscript edits. All other co-authors contributed to data collection and manuscript edits. A version of Chapter 5 has been published [Fuertes E, Butland BK, Anderson HR, Carlsten C, Strachan DP, Brauer M and the ISAAC Phase Three study group. 2014. Childhood intermittent and persistent rhinitis prevalence and climate and vegetation: a global ecologic analysis. Ann. Allergy, Asthma, Immunol. 113(4):386-392e9; doi:10.1016/j.anai.2014.06.021]. H Ross Anderson and David P Strachan were involved in the conceptualization of the project and contributed to data collection and manuscript edits. Chris Carlsten and Micheal Brauer were involved in the conceptualization of the project and contributed to manuscript edits. Barbara K Butland contributed to data collection, the analysis and manuscript edits. Sarah Henderson is alsoacknowledged for contributing to the assignment of some environmental exposures to the study centres. vTable of contentsAbstract......................................................................................................................................iiPreface.......................................................................................................................................ivTable of contents.......................................................................................................................viList of tables...............................................................................................................................xList of figures...........................................................................................................................xiiList of abbreviations..............................................................................................................xivAcknowledgements.................................................................................................................xvi1   Introduction..........................................................................................................................11.1   Literature review............................................................................................................11.1.1   Childhood allergic rhinitis......................................................................................11.1.1.1   Clinical presentation and treatment.................................................................11.1.1.2   Prevalences and burden of disease..................................................................21.1.1.3   Known and suspected risk factors ..................................................................21.1.1.4   Phenotype definitions in epidemiological studies...........................................31.1.1.5   Pathophysiology..............................................................................................41.1.1.6   Comorbidities .................................................................................................51.1.2   Traffic-related air pollution ....................................................................................61.1.2.1   Potential mechanisms linking air pollution to allergic rhinitis.......................71.1.2.2   Review of the literature on traffic-related air pollution and allergic rhinitis. .81.1.3   Genetic risk factors...............................................................................................111.1.3.1   Genetic association studies...........................................................................111.1.3.2   Review of the literature on genetic risk factors and allergic rhinitis............121.1.4   Climate..................................................................................................................151.1.4.1   Potential mechanisms linking climate to allergic rhinitis.............................151.1.4.2   Review of the literature on climate and allergic rhinitis...............................161.2   Data sources and study population...............................................................................171.2.1   Traffic, Asthma and Genetics collaboration ........................................................171.2.1.1   Study population and outcome data..............................................................17vi1.2.1.2   Air pollution estimates..................................................................................211.2.1.3   Genetic data...................................................................................................221.2.2   International Study of Asthma and Allergies in Childhood .................................221.2.2.1   Study population and outcome data..............................................................221.2.2.2   Environmental factors and covariates...........................................................231.3   Dissertation objectives..................................................................................................231.4   Dissertation structure....................................................................................................242   A longitudinal analysis of associations between traffic-related air pollution with asthma, allergies and sensitization in the GINIplus and LISAplus birth cohorts ...........252.1   Introduction..................................................................................................................252.2   Methods........................................................................................................................272.2.1   Study population...................................................................................................272.2.2   Questionnaire data................................................................................................272.2.3   Air pollution estimates..........................................................................................282.2.4   Statistical analysis.................................................................................................302.3   Results..........................................................................................................................312.3.1   Distribution of outcomes......................................................................................342.3.2   Air pollution estimates..........................................................................................352.3.3   Total and area-specific associations......................................................................362.3.4   Sensitivity analyses...............................................................................................392.4   Discussion.....................................................................................................................433   Childhood allergic rhinitis, traffic-related air pollution, and variability in the GSTP1,TNF, TLR2 and TLR4 genes .................................................................................................473.1   Introduction..................................................................................................................473.2   Methods........................................................................................................................483.2.1   Study population...................................................................................................483.2.2   Health outcomes...................................................................................................493.2.3   Air pollution estimates..........................................................................................493.2.4   Genetic data..........................................................................................................503.2.5   Statistical analysis.................................................................................................51vii3.3   Results..........................................................................................................................523.3.1   Main environmental associations.........................................................................563.3.2   Main genetic associations.....................................................................................603.3.3   Genotype stratification and interaction associations............................................643.4   Discussion.....................................................................................................................664   Associations between the 17q21 region and allergic rhinitis in five birth cohorts ......714.1   Introduction..................................................................................................................714.2   Methods........................................................................................................................724.2.1   Study population...................................................................................................724.2.2   Health outcomes...................................................................................................734.2.3   Genetic data..........................................................................................................734.2.4   Statistical analysis.................................................................................................744.2.5   Sensitivity analyses ..............................................................................................744.3   Results..........................................................................................................................754.3.1   Distribution of genotypes and outcomes..............................................................754.3.2   Longitudinal associations.....................................................................................784.3.3   Age-specific cross-sectional associations ............................................................834.4   Discussion.....................................................................................................................845   Childhood intermittent and persistent rhinitis prevalence and climate and vegetation:a global ecologic analysis .......................................................................................................895.1   Introduction..................................................................................................................895.2   Methods........................................................................................................................905.2.1   Study population...................................................................................................905.2.2   Health outcomes...................................................................................................915.2.3   Environmental factors and covariates...................................................................915.2.4   Statistical analysis.................................................................................................935.3   Results..........................................................................................................................945.3.1   Distribution of intermittent and persistent rhinitis prevalences ...........................945.3.2   Distribution and correlation of environmental factors..........................................965.3.3   Associations between rhinitis symptoms and environmental factors ..................98viii5.4   Discussion...................................................................................................................1096   Conclusions of dissertation..............................................................................................1146.1   Summary of research and contributions.....................................................................1146.2   Methodological considerations...................................................................................1166.2.1   Traffic, Asthma and Genetics collaboration .......................................................1166.2.1.1   Study design................................................................................................1166.2.1.2   Air pollution exposure assessment..............................................................1176.2.1.3   Genetic data.................................................................................................1186.2.2   International Study of Asthma and Allergies in Childhood ...............................1196.2.2.1   Study design................................................................................................1196.2.2.2   Environmental assessment..........................................................................1206.3   Recommendations for future research........................................................................1206.4   Summary.....................................................................................................................124References..............................................................................................................................125Appendices.............................................................................................................................147Appendix A: Genotype imputation ....................................................................................147Appendix B: International Study of Asthma and Allergies in Childhood Phase Three study groups.................................................................................................................................149Appendix C: Methods for assigning the Normalized Difference Vegetation Index and population density values...................................................................................................154Appendix D: Descriptive statistics for the 13-14 age-group International Study of Asthma and Allergies in Childhood centres.....................................................................................156Appendix E: Descriptive statistics for the 6-7 age-group International Study of Asthma and Allergies in Childhood centres...........................................................................................164ixList of tablesTable  1: Classification of allergic rhinitis symptoms proposed by the Allergic Rhinitis and its Impact on Asthma collaboration......................................................................................4Table 2:  Characteristics of cohorts participating in the Traffic, Asthma and Genetics collaboration..................................................................................................................20Table 3:  Characteristics of models used to estimate air pollution exposures...............................29Table 4:  Characteristics of study participants...............................................................................32Table 5:  Characteristics of the total study participants with available serology data ..................33Table 6:  Distribution of estimated annual average concentrations of NO2, PM2.5 mass, PM2.5 absorbance and ozone at the birth addresses in the total dataset and per area...............36Table 7:  Total and area-specific associations between air pollutants estimated to the birth addressand health outcomes during the first ten years of life ...................................................38Table 8:  Characteristics of the study population...........................................................................53Table 9:  Single nucleotide polymorphism characteristics and genotype frequencies in the study population......................................................................................................................55Table 10: Pooled and cohort specific associations between allergic rhinitis and aeroallergen sensitization with air pollutants.....................................................................................59Table 11: Pooled and cohort specific associations between allergic rhinitis and ten single nucleotide polymorphisms.............................................................................................62Table 12: Pooled and cohort specific associations between aeroallergen sensitization and ten single nucleotide polymorphisms..................................................................................63Table 13: Associations between air pollutants and allergic rhinitis among homozygous major and heterozygous/homozygous minor allele carriers...........................................................64Table 14: Genetic information for seven single nucleotide polymorphisms at the 17q21 locus.. .76Table 15: Pooled and cohort-specific numbers of cases and controls [cases/controls].................77Table 16: Pooled and cohort specific associations between seven single nucleotide polymorphisms at the 17q21 locus and allergic rhinitis................................................79Table 17: Pooled longitudinal associations between seven single nucleotide polymorphisms at the 17q21 locus and allergic rhinitis, stratified by history of asthma ...........................82xTable 18: Intermittent and persistent rhinitis symptom prevalence by climate type for the 13-14 age-group centres...........................................................................................................94Table 19: Correlations between modelled variables for the 13-14 age-group centres...................97Table 20: Between-country associations between intermittent and persistent rhinitis prevalence and environmental factors for the 13-14 age-group centres..........................................99Table 21: Between-country associations between intermittent and persistent rhinitis prevalence and environmental factors for the 6-7 age-group centres............................................100Table 22: Within-country associations between intermittent and persistent rhinitis prevalence andenvironmental factors among countries with two or more centres per country for the 13-14 age-group centres...............................................................................................102Table 23: Within-country associations between intermittent and persistent rhinitis prevalence andenvironmental factors among countries with two or more centres per country for the 6-7 age-group centres......................................................................................................103Table 24: Between- and within-country associations between environmental factors and the prevalence of intermittent rhinitis and itchy-eyes and persistent rhinitis and itchy-eyes for the 13-14 age-group centres...................................................................................105Table 25: Between- and within-country associations between intermittent and persistent rhinitis prevalence and environmental factors for the 13-14 age-group centres, independently adjusted for PM2.5 mass and NO2, and mutually adjusted for meteorological and vegetation factors. .......................................................................................................107xiList of figuresFigure 1: Follow-up time points for each cohort participating in the Traffic, Asthma and Geneticscollaboration..................................................................................................................19Figure 2: Flow chart of study population.......................................................................................31Figure 3: Period prevalence of children with doctor diagnosed asthma (bars with diagonal lines) or allergic rhinitis (filled bars) at ages three to ten years in the total population (A) and stratified by area: GINI/LISA South (B), GINI/LISA North (C) and LISA East (D)....35Figure 4: Total and area-specific associations between NO2 (A), PM2.5 mass (B), PM2.5 absorbance (C) and ozone (D) estimated to the six-year address with doctor diagnosed asthma (red squares), doctor diagnosed allergic rhinitis (purple triangles), eye and nosesymptoms (blue stars) and aeroallergen sensitization (black diamonds). .....................40Figure 5: Total and area-specific associations between NO2 (A), PM2.5 mass (B) and PM2.5 absorbance (C) estimated to the ten year address with doctor diagnosed asthma (red squares), doctor diagnosed allergic rhinitis (purple triangles), eye and nose symptoms (blue stars) and aeroallergen sensitization (black diamonds). ......................................41Figure 6: Total and area-specific associations between NO2 (A), PM2.5 mass (B) and PM2.5 absorbance (C) averaged over the birth, six and ten year address estimates, and ozone (D) averaged over the birth and six year address estimates with doctor diagnosed asthma (red squares), doctor diagnosed allergic rhinitis (purple triangles), eye and nosesymptoms (blue stars) and aeroallergen sensitization (black diamonds). .....................42Figure 7: Distribution of air pollutants pooled and by cohort. .....................................................57Figure 8: Associations between allergic rhinitis (stars) and aeroallergen sensitization (dark squares) with NO2, ozone, PM2.5 mass and PM2.5 absorbance. ................................58Figure 9: Associations between allergic rhinitis (stars) and aeroallergen sensitization (dark squares) with ten single nucleotide polymorphisms. ....................................................61Figure 10: Pooled longitudinal associations between seven single nucleotide polymorphisms at the 17q21 locus and allergic rhinitis (black squares), asthma (red stars), allergic rhinitisand concomitant asthma (blue triangles) and allergic rhinitis without concomitant asthma (purple circles). .................................................................................................80xiiFigure 11: Cohort- and age-specific cross-sectional associations between seven single nucleotide polymorphisms at the 17q21 locus and allergic rhinitis for the three cohorts with available longitudinal data: BAMSE (A), GINI/LISA (B) and PIAMA (C). ...............83Figure 12: Data availability for the International Study of Asthma and Allergies in Childhood Phase Three study population........................................................................................91Figure 13: World map showing the centre prevalence of intermittent rhinitis symptoms for the 13-14 age-group centres.................................................................................................95Figure 14: World map showing the centre prevalence of persistent rhinitis symptoms for the 13-14 age-group centres......................................................................................................95xiiiList of abbreviationsAPMoSPHERE Air Pollution Modelling for Support to Policy on Health and Environmental Risk in EuropeBAMSE Children, Allergy, Milieu, Stockholm, Epidemiological SurveyCAPPS Canadian Asthma Primary Prevention StudyCI confidence intervalsESCAPE European Study of Cohorts for Air Pollution Effects GINIplus German Infant study on the influence of Nutritional Intervention plus environmental and genetic influences on allergy developmentGNI gross national incomeGWAS genome-wide association studyIgE immunoglobulin EIL interleukinISAAC International Study of Asthma and Allergies in ChildhoodLD linkage disequilibriumLISAplus Lifestyle related factors, Immune System and the development of Allergiesin East and West Germany plus the influence of traffic emissions and genetics studyLUR land-use regressionNDVI Normalized Difference Vegetation IndexNO2 nitrogen dioxideOR odds ratioPIAMA Prevention and Incidence of Asthma and Mite AllergyPM2.5 particulate matter with an aerodynamic diameter less than 2.5 µmPM10 particulate matter with an aerodynamic diameter less than 10 µmSAGE Study of Asthma, Genes and EnvironmentSNP single nucleotide polymorphismTAG Traffic, Asthma and GeneticsxivTRAPCA Traffic Related Air Pollution and Childhood AsthmaTh2 type 2 helperxvAcknowledgementsThis research would not have been possible without the knowledge, guidance and unwavering support of my supervisors Drs. Michael Brauer and Chris Carlsten. I have immensely enjoyed working with you both. Such a positive experience would not have been possible without your dedicated efforts and great ideas. I would also like to thank my committee members, Dr. JoachimHeinrich, for welcoming me into this team and giving me room to grow, and Dr. Andrew Sandford, whose expertise in genetics helped guide my way. A special thanks to the many students and researchers with whom I have crossed paths, both within my School and abroad. To those at Helmholtz who helped transform Munich into my home, Danke für Alles! Ich habe unglaublich viel von Euch gelernt. Thank you to the groups at St George's in London and at the Karolinska Institutet in Stockholm with whom I closely collaborated. Working with you always was and continues to be a pleasure. I also thank the manyco-authors who have contributed to this work. Your input and suggestions not only improved the quality of my dissertation but also helped me develop into a better researcher. I would like to thank the University of British Columbia, Allergy, Genes and Environment Network of Centres of Excellence (AllerGen), Michael Smith Foundation for Health Research, Canadian Institutes of Health Research and Helmholtz Zentrum München for their generous financial support. A huge thanks to my family, whose support I've never doubted, and my friends, for reminding me of life outside of work. Finally, thank you Patrick - for everything.xvi1   Introduction1.1   Literature review1.1.1   Childhood allergic rhinitis1.1.1.1   Clinical presentation and treatmentAllergic rhinitis is characterized as inflammation of the membranes lining the nose. The inflammation is induced by an immunoglobulin E (IgE)-mediated response to allergen exposure and leads to increased mucus production and the development of one or more cardinal symptoms: rhinorrhoea, nasal obstruction, nasal itching and sneezing. Ocular symptoms also frequently co-occur. These symptoms can differ in duration and severity. Individuals with non-allergic rhinitis also present with similar symptoms, but these symptoms are triggered by exposure to factors other than allergens, such as medications, hormones and physical and chemical agents (Fokkens 2002). Infectious rhinitis, also known as the common cold or rhinosinusitis, is typically caused by a virus. This thesis is focused entirely on allergic rhinitis. There are three main treatment strategies for reducing allergic rhinitis symptoms (Kay 2001). First, individuals may try to reduce their exposure to allergens. Although intuitive, indoor allergen avoidance strategies have not yet proven to effectively reduce symptoms (Custovic et al.2002; Sheikh et al. 2010). Second, medications such as antihistamines, decongestants and nasal corticosteroid sprays can be used to reverse symptoms. Although generally effective, individuals who experience symptoms often may need to medicate frequently as these medications do not have sustained long-term effects. Third, treatment by allergen-specific immunotherapy can be pursued, during which increasing doses of an allergen extract are administered to an allergic individual in controlled settings over a long period of time. This type of treatment can induce clinical and immunological tolerance, has long-term efficacy and may prevent allergic disease progression. However, a precise diagnosis of IgE-mediated allergy is needed and a risk of anaphylaxis exists for all patients, especially during “up-dosing” phases. 11.1.1.2   Prevalences and burden of diseaseAllergic rhinitis affects individuals from all ages, countries and ethnic groups and of every socioeconomic status. Despite being often undiagnosed, recent global estimates indicate that 8.5% and 14.6% of children aged 6-7 years and 13-14 years, respectively, reported currently suffering from this disease (Aït Khaled et al. 2009)‐ . Allergic rhinitis prevalence is increasing in some regions of the world, especially those undergoing rapid socio-economic development (Björkstén et al. 2008). The rate of disease onset appears to be plateauing or decreasing in areas with high prevalences (Björkstén et al. 2008). Disease prevalence tends to be greatest in higher income countries but the prevalence of severe disease is higher in middle and low income countries (Aït Khaled et al. 2009)‐ . Interestingly, disease rates within countries can vary considerably. Given that allergic rhinitis greatly impacts quality of life, leads to significant costs to school performance, work productivity and social life, contributes to sleep disturbance and learning disabilities and is associated with substantial indirect economic costs, the growing global prevalence of this disease is concerning (Bousquet et al. 2008).1.1.1.3   Known and suspected risk factors Allergen exposure is the primary and required environmental risk factor for allergic rhinitis onset. Pollen, fresh grasses and (outdoor) moulds are the major sources of outdoor aeroallergens. These outdoor aeroallergens, especially those from pollen, are largely responsible for the exacerbation of seasonal allergies, perhaps the most common manifestation of allergic rhinitis. Dust mites, animal dander, cockroaches and (indoor) moulds are the major sources of indoor aeroallergens. Most individuals are sensitized to several allergens and thus may experience symptoms under a wide range of conditions. The number and types of allergens present in the environment are region-specific and depend on a host of climatic factors (further discussed in section 1.1.4 and Chapter 5 of this thesis). Host (sex and age) and early-life factors (young maternal age, multiple gestation and low birth weight) are known to influence disease (Bousquet et al. 2008). Additionally, aspects of a “Western-country lifestyle” are believed to increase the risk of disease incidence. This increased risk is well exemplified in historical prevalence data in East and West Germany. Prior to 2reunification in Germany (1990), asthma and allergy rates were substantially lower in East Germany compared to West Germany, despite the much higher air pollution levels in East Germany. After reunification, the adaptation of a “Western-country lifestyle” was hypothesized to lead to a large increase in the prevalence of asthma and allergies in East Germany (Nowak et al. 1998). Today, few differences in lifestyle, living conditions and allergic disease prevalences exist between the two German areas. Which aspect of a “Western-country lifestyle” promotes allergic disease development continues to be investigated. Changes in nutrition, pet ownership, housing, daycare attendance (higher respiratory infection exposures) as well as increased urbanization and traffic-related air pollution exposure may all play a role. For example, the character of the air pollution in East Germany changed dramatically after reunification due to a drop in “classical” air pollutants (total suspended particles and sulphur dioxide) but an increase in very small ultra-fine particles (Pitz et al. 2001). The impact of traffic-related air pollution on allergic rhinitis is an integral part of this dissertation and is further discussed in section 1.1.2. Finally, there is strong evidence indicating that an individual's genotype plays a role in determining disease susceptibility. Genetic variability, as a risk factor for allergic rhinitis, is discussed in detail in section 1.1.3. 1.1.1.4   Phenotype definitions in epidemiological studiesAllergic rhinitis is a highly heterogeneous disease. Consequently, creating a common phenotype definition to be used in research settings has proven challenging. In epidemiological studies, allergic rhinitis is typically defined based on a report of nasal symptoms in the absence of a cold, often with consideration for the presence of ocular symptoms. A stricter definition of allergic rhinitis may also include a positive test for sensitization to one or more allergens. However, this latter information is not always available in epidemiological settings as it is often impractical to visit all participants to measure immune responses. Allergic rhinitis has traditionally been categorized as seasonal or perennial according to the presumed timing of allergen exposure. This classification is imperfect for several reasons: pollens and moulds may be perennial allergens in some areas, perennial symptoms may not be 3present all year round, symptoms do not necessarily occur in strict conjunction with the allergen season, the majority of patients are poly-sensitized, etc (Bousquet et al. 2008). To address these limitations, the Allergic Rhinitis and its Impact on Asthma collaboration (www.whiar.org) has proposed new definitions which classify allergic rhinitis symptoms as either intermittent or persistent (Bousquet et al. 2001). This classification is based on symptom duration (and thereforedoes not depend on the allergens' origins), better reflects a patient's true experiences, is applicable worldwide (which facilitates comparisons between epidemiological studies) and can be further divided into mild or moderate/severe based on symptom severity (Table 1). Table  1: Classification of allergic rhinitis symptoms proposed by the Allergic Rhinitis and its Impact on Asthma collaborationClassification DefinitionDuration Intermittent Less than four days a week or for less than four consecutive weeksPersistent More than four days a week and for more than four consecutive weeksSeverity Mild None of the following are present: sleep disturbance, impairment of school, work or daily activities (leisure and/or sport), symptoms present but not troublesomeModerate/severe One or more of the following are present: sleep disturbance, impairment of school, work or daily activities (leisure and/or sport), troublesome symptoms Adapted from (Bousquet et al. 2008).1.1.1.5   PathophysiologyAllergen sensitization is the fundamental step that leads to further symptom development (Skoner 2001). After exposure to threshold concentrations of an allergen for a given period of time, antigen presenting cells take up the allergen and present it to CD4+ T lymphocytes. Once activated, these T lymphocytes release Type 2 helper (Th2) pro-inflammatory cytokines, such as interleukin (IL)-4 and IL-13, that stimulate the production of antigen-specific IgE that binds to effector cells, such as mast cells and basophils. In non-atopic individuals, allergen exposure leadsto a low-grade immunologic response and subsequent release of cytokines produced mainly by Type 1 helper cells, rather than the overproduction of Th2 cytokines. An over-expression of a Th2 response is a hallmark feature of allergic disease.4Once sensitized, subsequent allergen exposure leads to an allergic response that can be divided into an early and late phase. The early phase develops within minutes of allergen exposure. Uponallergen recognition and binding to IgE-coated mast cells, the mast cells degranulate and release preformed and de novo inflammatory mediators, such as histamines and lipid mediators. This release causes watery rhinorrhea, occlusion and congestion of nasal air passages as well as nasal itch and congestion. In addition to this early phase response, other inflammatory cells are recruited to the nasal mucosa, such as basophils, eosinophils, neutrophils, T lymphocytes and newly synthesized mast cells. These additional inflammatory cells become activated four to eighthours after allergen exposure and cause additional mediator release and recruitment and a sustained inflammatory reaction. Similar symptoms are experienced during the late phase as during the early phase, with possible increased nasal congestion. Allergen “priming”, the term used to describe an increase in allergen reactivity after repeated allergen exposure (Connell 1969), is believed to be a consequence of the additional inflammatory cells released during the late phase. It has been demonstrated that in addition to the pollen allergen, which leads to the allergic cascade described above, pollen grains may also release bioactive lipid mediators that activate human neutrophils and eosinophils in vitro, and likely other cells involved in the allergic cascade(Traidl-Hoffmann et al. 2003). Furthermore, like other non-specific irritants or adjuvants (dust, tobacco smoke or air pollutants), the pollen grain and other allergen sources, and the allergens themselves, may enhance nasal responsiveness to other allergens by, for example, increasing epithelial permeability (Takai and Shigaku 2011). 1.1.1.6   Comorbidities There is substantial evidence supporting the existence of a united airways disease (Passalacqua 2000), also referred to as the “one airway one disease” concept. This hypothesis postulates that allergic rhinitis may be an early stage manifestation of airway disease and asthma a latter stage phenotype (or vice versa). Atopic dermatitis is often also considered an early indicator of allergicdisease (Spergel 2005; Zheng et al. 2011).5Indeed, the comorbidity between allergic rhinitis and asthma is high (Aït Khaled et al. 2009; ‐Bousquet et al. 2008) and appears to involve a dynamic process (Ballardini et al. 2012). However, it remains unclear how much of disease comorbidity is attributable to the fact that these diseases may be causal for one another (that is, early-life asthma increases the risk of later-life allergic rhinitis), that they have common environmental and genetic risk factors, or more likely, that a combination of both of these reasons is at work. Efforts targeting the biological mechanisms of allergic diseases will help elucidate how environmental and genetic influences contribute to the many distinct and complex (comorbid) allergic phenotypes. 1.1.2   Traffic-related air pollution Air pollution is a mixture of hundreds of gaseous compounds and particles of complex compositions arising from various sources. Traffic-related air pollution consists primarily of carbon dioxide, carbon monoxide, hydrocarbons, nitrogen oxides and particulate matter. Nitrogen dioxide (NO2) is often used as a marker for traffic-related air pollution in epidemiological health studies as it can be easily and affordably measured. Also frequently investigated are the health effects of exposure to particulate matter with aerodynamic diameters less than 2.5 µm (PM2.5) and less than 10 µm (PM10), which are released from several sources and have compositions that vary temporally and spatially across climatological and geographical areas (Querol et al. 2004).Despite large improvements in engine technologies and cleaner fuels, traffic-related air pollution is likely to continue having important adverse impacts on health given the expansion of metropolitan areas and increasing number of vehicles on roads and individuals living and working near busy roads. Its relative importance to health compared to air pollution released from other sources is also likely to remain large given that traffic-related air pollution has a high intake fraction, as it is emitted in close proximity to individuals, and several other air pollution sources (for example, industry) have undergone large reductions in emissions. The ubiquitous nature of traffic-related air pollution, combined with the fact that no “safe-concentration” has yet been identified, renders research into its associated health effects of high importance. 61.1.2.1   Potential mechanisms linking air pollution to allergic rhinitisMost anthropogenic air pollutants are themselves not allergenic materials. Rather, air pollutants are hypothesized to exert their effects by acting directly upon the individual and/or indirectly by affecting the allergen-carrier (that is, the plant and its pollen). Regarding the former mechanism, air pollutants appear to act as mucosal adjuvants and to interact with the innate and adaptive immune cells to skew the immune response to inhaled allergens towards a preferential activation of Th2 cells (Saxon and Diaz-Sanchez 2005). The generation of oxidative stress induced by traffic-related air pollution exposure is one of the main underlying mechanisms by which this Th2-dominated response is believed to be generated, with the strongest evidence emerging from experimental studies on diesel exhaust particles. Briefly, upon impact and internalization of diesel exhaust particles within the airway epithelium and macrophages, the production of reactive oxygen species is increased. In response, the release of pro-inflammatory cytokines and IgE is also increased. Although poorly understood, under these pollutant-induced conditions, a Th2-like response leads to inflammation, and when antioxidant defences are overwhelmed, to further reactive oxidative stress production and subsequent downstream inflammatory effects (Saxon and Diaz-Sanchez 2005). Although it has been suggested that fine or ultrafine particles may be most detrimental because of their suspected high pro-inflammatory potential, the oxidative stress hypothesis also extends to gaseous pollutants, especially ozone, which is a highly potent oxidant. Supplementation with anti-oxidants may be one way to block or alleviate air-pollution enhanced IgE production and mitigate the inflammatory and adjutant effects, although the evidence supporting the efficacy of such interventions is limited (Romieu et al. 2008). The oxidative stress hypothesis is further supported by gene-environment studies on asthma (MacIntyre et al. 2014a; Romieu et al. 2006; Salam et al. 2007; Wu et al. 2007) and sensitization (Melén et al. 2008) which report that the effect of traffic-related air pollution on respiratory health may be modified by variants in genes involved in controlling the antioxidant system (for example, GSTP1) and inflammatory response (TNF) in allergy (further discussed in section 1.1.3.2 and Chapter 3). Traffic-related air pollutants may also exert adverse effects via pathways other than through the generation of oxidative stress, as previously reviewed (D’Amato and Cecchi 2008; D’Amato et 7al. 2010, 2007). First, air pollution exposure has been shown to have an inflammatory effect on the airways of susceptible subjects when exposed. This inflammatory effect can lead to increasedpermeability of the epithelium, greater access and penetration of allergens to the immune system and easier interactions with immune system cells. This air pollution induced inflammation may also increase “priming” of allergen responses, leading to the development of symptoms at lower allergen exposures. Second, combustion particles appear to act as physical carriers of allergens, thereby potentially facilitating the entrance of allergens into the airways when the air pollutants are respired (Namork et al. 2006). Finally, as previously mentioned at the beginning of this section, air pollutants may also exert adverse effects indirectly by acting on the allergens. Allergen release, morphology and allergenicity have all been shown to be modified following interactions with air pollutants (Beck et al. 2013; Ghiani et al. 2012; Motta et al. 2006).1.1.2.2   Review of the literature on traffic-related air pollution and allergic rhinitisExperimental studies on humans, animals or in vitro test systems suggest that traffic-related air pollution exposure leads to biological effects. The strongest evidence comes from studies on diesel exhaust particles which have consistently demonstrated that exposure leads to altered immunological responses to allergens and inflammatory effects in the airways (reviewed in (Saxon and Diaz-Sanchez 2005)). Gilliland et al. (2004) also experimentally demonstrated that the effect of diesel exhaust particles on allergic inflammation may be modified by variability in two oxidative stress genes (GSTM1 and GSTP1). Although less consistent, there is also some evidence which suggests that exposure to NO2 (Strand et al. 1997) and ozone (Chen et al. 2004), the latter only among a subgroup, may also modify allergen responses among asthmatics.Interestingly, despite possible biological mechanisms and experimental and toxicological studies supporting the existence of an association between traffic-related air pollution exposure and allergic disease, particularly for diesel exhaust particles, the evidence from epidemiological studies remains largely inconclusive, especially for non-asthma allergic phenotypes (Heinrich and Wichmann 2004).8Studies examining associations between allergic rhinitis and air pollution in general, usually by comparing between-community air pollution levels, have yielded inconsistent results. Two large cross-sectional Taiwanese studies reported positive associations between allergic rhinitis prevalence among schoolchildren and community-averaged exposure to some air pollutants estimated from air-monitoring data derived from monitors less than two kilometres from the schools of the children (Hwang et al. 2006; Lee et al. 2003). However, a large global study foundno association between ambient air pollution and allergic rhinitis prevalence (Anderson et al. 2009). Furthermore, an earlier study conducted in East Germany showed an inverse relationship between atopic disease prevalences and air pollution levels during the 1990s (especially total suspended particulates and sulphur dioxide) (Heinrich et al. 2002b). The authors of this latter study cautioned against concluding that air pollution does not influence allergic disease onset, but rather suggested that lifestyle changes which occurred during the time-span of the study weremore influential. Although informative, studies which compare air pollution levels between communities are hindered by their inability to capture within-city air pollution variability, which can vary by several magnitudes and may thus be more important (Heinrich 2010). Interestingly, the body of evidence from studies examining the effects of traffic-related air pollution (within-community air pollution variability) is not much stronger. A review conducted by the Health Effects Institute in 2010 concluded that the evidence was “inadequate and insufficient” to determine the existence (or not) of a causal relationship between traffic-related air pollution and IgE-mediated allergies (Tager et al. 2010). Of the 16 studies considered to be of sufficient quality to be included in this review, only two showed consistent associations. Positive associations were reported between several allergic outcomes and sensitization with outdoor NO2concentrations at the home addresses of nine year-old children living in West Germany (Krämer et al. 2000) and between hayfever and sensitization to outdoor aeroallergens with indicators of traffic-related air pollution at the home addresses of six year-old children living in Munich, Germany (Morgenstern et al. 2008).Since this review was published, a multicentre study including more than 7,000 children, conducted as part of the European Study of Cohorts for Air Pollution Effects (ESCAPE; 9www.escapeproject.eu), found no evidence to suggest that traffic-related air pollution exposure was a risk factor for sensitization at four to six and eight to ten years of age (Gruzieva et al. 2014). This multicentre study had harmonized outcome and covariate definitions and utilized a standardized protocol to assess individual-level estimates of various traffic-related air pollutants to the home address of all participants using area-specific land-use regression (LUR) models. Additionally, the only two prospective epidemiological studies to examine associations between traffic-related air pollution exposure and allergic rhinitis from birth up to the age of eight years were also published. Gehring et al. (2010) reported positive associations between traffic-related air pollutants and the prevalence of asthma and asthma symptoms in 3,863 Dutch children. Positive associations with allergic rhinitis were found only among non-movers. Interestingly, thisstudy found no association between traffic-related air pollution and aeroallergen sensitization at eight years despite a previously documented association at four years (Brauer et al. 2007). Similarly, no association was found between indicators of traffic-related air pollution and aeroallergen sensitization at eight-years among more than 2,500 Swedish children (Gruzieva et al. 2012) despite a previously published association (especially to pollen) at four years (Nordlinget al. 2008). These latter results may suggest that the timing of air pollution exposure relative to disease onset may be important. In contrast, the evidence for an adverse effect of traffic-related air pollution on asthma is much stronger. Recent reviews (Guarnieri and Balmes 2014; Tager et al. 2010) and a meta-analysis (Anderson et al. 2013) concluded that there is sufficient/strong evidence supporting a causal role for traffic-related air pollution on asthma exacerbation and that the evidence for a contribution to new-onset asthma is building. These recent reports did not differentiate between allergic and non-allergic asthma. Why traffic-related air pollution may affect asthma more strongly than non-asthmatic allergic phenotypes is unknown, but this might point to potentially different biological mechanisms. For example, one hypothesis may be that different traffic-related air pollution components deposit in certain areas of the respiratory system and subsequently have different effects on the development of asthma and other non-asthmatic allergic diseases. 101.1.3   Genetic risk factorsThe completion of the human genome sequence (Venter et al. 2001), the initiatives of the International HapMap Project (Gibbs et al. 2003) and rapid advances in bioinformatics have led to an explosion of studies investigating the impact of genetics on health. The very high familial clustering of cases could suggest the importance of genetic constitution on allergic disease development, although it may also be explained by shared environmental factors. Some of the most convincing evidence for genetic variability as a risk factor for allergic rhinitis comes from twin studies, which have reported heritability estimates of 71-96% (Beijsterveldt and Boomsma 2007; Räsäsnen et al. 1998; Thomsen et al. 2006). Given these very high estimates, genetic variability and the effect of this variability on the response to environmental factors (gene-environment interactions) cannot be ignored in the search for etiological risk factors of allergic rhinitis. 1.1.3.1   Genetic association studiesGenetic association studies are broadly categorized as candidate gene or genome-wide studies. Although different in some design aspects, all genetic association studies aim to identify genetic variability that is more frequent among those with the disease compared to healthy individuals. The vast majority of genetic studies to date have focused on identifying single nucleotide polymorphisms (SNPs) as risk factors for disease. Although SNPs are typically ascribed to their most proximal genes, in most cases, there is no functional evidence for a specific target. A SNP may in fact affect the regulation of several genes including those more distant.Candidate gene association studies are a hypothesis testing method; genes of interest are selectedfor analysis based on location and/or function. These studies tend to be inexpensive and sufficiently powered to detect common variants of modest effects (Hirschhorn and Daly 2005). However, these studies are restricted to the effects of one (or a few) genetic risk factors and thus their potential for discovery depends on previous knowledge.In contrast, genome-wide studies provide an unbiased and comprehensive assessment of the impact of all types of genetic variants (SNPs, deletions, copy number variations, inversions) on 11disease development and can implicate previously unidentified genes and chromosome regions (Manolio et al. 2009). Genome-wide studies include linkage mapping studies, in which allele-disease associations are examined within families, and genome-wide association studies (GWASs), which consider associations in the broader population (Cordell and Clayton 2005). Linkage mapping studies have been very successful in identifying genes that cause large effects in Mendelian diseases. However, as these types of studies often lack sufficient statistical power to detect small effects, their use is more limited for common diseases (Cordell and Clayton 2005). Properly designed GWASs have the power to detect common alleles of modest effects, although the need for stringent multiple testing correction, such as by using the false discovery rate (Benjamini and Hochberg 1995), and the potential for population stratification (Marchini et al. 2004) are important methodological challenges. Gene-environment studies aim to investigate interactions between genetic and environmental riskfactors which, when analyzed separately, may yield different or no associations with disease. For example, the effects of genetic variants may only be apparent in the context of certain environmental exposures (Thomas 2010), as appears to be the case for risk variants in IL13, which only increased the risk of childhood allergic rhinitis when a child was exposed to mould inthe home during the first year of life in a Korean study (Kim et al. 2012). Gene-environment interactions can be tested using both candidate gene and GWAS study designs, although the latterrequires a very large number of participants to achieve sufficient statistical power. 1.1.3.2   Review of the literature on genetic risk factors and allergic rhinitisThe first GWAS on allergic rhinitis was published in 2011 (Andiappan et al. 2011). This study was conducted on an ethnic Chinese population in Singapore with a mean age of 21.4 years, included 456 and 676 allergic rhinitis cases (in the discovery and replication cohorts, respectively) and found no significant associations at the genome-wide level. However, consistent suggestive evidence implicating the MRPL4 and BCAP genes was observed, both of which had not yet been previously associated with allergic rhinitis. In a larger genome-wide association meta-analysis including 3,933 self-reported allergic rhinitis cases from four Europeanadult cohorts, only a few loci were associated with prevalent allergic rhinitis (Ramasamy et al. 122011). A SNP near C11orf30 and LRRC32, a region previously associated with allergic disorders of the skin (Esparza-Gordillo et al. 2010; Marenholz et al. 2011; O’Regan et al. 2010), reached genome-wide significance and six others showed suggestive evidence of an association. Two years later, Hinds et al. (2013) reported that 11 of 23 variants discovered in a GWAS of self-reported allergy were specifically associated with self-reported hayfever symptoms at the genome-wide level (variants were in or near ETS1, HLA-C, HLA-DQA1, IL1RL1, IL33, LPP, LRRC32, PLCL1, SMAD3, TLR1 and TSLP). Combined, these efforts are beginning to identify regions of the genome suspected of influencing disease onset, many of which appear common to asthma and/or atopy. The last GWAS on allergic rhinitis published at the time of this dissertation capitalized on this concept and examined associations with a combined phenotype of asthma and hayfever (6,685 cases with both conditions) (Ferreira et al. 2014). This study reported genome-wide significant associations between 11 genetic variants and the combined phenotype, nine of which were located in or near established risk loci for allergic disease (GSDMA, HLA-DQB1, LRRC32, IL1RL1, IL33, SMAD3, TLR1, TSLP and WDR36) and two of which were novel (CLEC16A and ZBTB10). These 11 variants were also associated with both asthma and hayfever when modelled as independent outcomes, but associations were consistently strongest for the combined phenotype. The authors of this study concluded that using a combined phenotype may increase the efficiency of genetic studies aiming to identify novel risk variants. In addition to these (hypothesis generating) GWASs, candidate gene studies have identified associations between more than 100 SNPs and allergic rhinitis. In 2013, the first systematic studyof the general reproducibility of reported SNPs for allergic rhinitis was published (Nilsson et al. 2013). The authors investigated associations between 49 SNPs (of the 116 previously identified) and allergic rhinitis in two study populations (Swedish and Singapore Chinese) and found remarkably poor reproducibility. The authors identified IL13 as the best candidate for future research as there was complete concordance in the direction of the odds ratios (ORs) across investigations. Only one of the association signals reported in this systematic study (IL33) was significant in the last two GWASs mentioned above (Ferreira et al. 2014; Hinds et al. 2013).13Results from candidate gene studies and GWASs on allergic rhinitis and other complex diseases may be challenging to reconcile because of limited/inadequate sample sizes, heterogeneity of outcome definitions and study designs and stringent multiple testing corrections. Larger population sizes and careful phenotyping are needed to improve reproducibility. Additionally, theapparent lack of coherence between studies may suggest that certain associations are only present in specific populations, at least some of the reported associations represent false positivesor that some effect-sizes may be much lower than reported (Nilsson et al. 2013). However, in addition to these study design and/or analysis-related obstacles, biological mechanisms should also be considered, such as true phenotypic disease heterogeneity as well as gene-gene, gene-environment and epigenetic-environment interactions (Vercelli 2008). Indeed, one possible reason why genetic hits may be poorly reproducible or may represent only mild increased risks may be that a permissive environmental context (or particular host characteristic) is required for the development of major phenotypic effects.Of particular interest with regard to air pollution and allergic rhinitis is the role of genetic variants in the body's oxidative stress pathway. As described in section 1.1.2.1, traffic-related air pollution exposure may heighten the response to allergens by increasing oxidative stress and inflammation in the airways. This proposed biological mechanism suggests that individuals who have diminished pulmonary antioxidant defences may be more susceptible to the adverse effects of traffic-related air pollution. At the time the work for this dissertation began, no gene-air pollution interaction study had been published on allergic rhinitis, but interaction effects with asthma had been clearly demonstrated for endotoxin, secondhand smoke and ozone exposure (London 2007). Gene-air pollution interactions for sensitization had also been documented for nitrogen oxides in a four year-old Swedish birth cohort (Melén et al. 2008) and for diesel exhaustparticles (Gilliland et al. 2004) and secondhand smoke (Gilliland et al. 2006) in experimental studies on adult patients. During the course of this dissertation, additional gene-air pollution studies were published, primarily for asthma (for example, (Carlsten et al. 2011; MacIntyre et al. 2014a)).141.1.4   ClimateThere remains little doubt that climate change is occurring. The most recent Intergovernmental Panel on Climate Change Summary for policy makers, published in 2013, highlights the fact thatthe amount of scientific evidence on the impacts of warming has almost doubled since the last report in 2007 (IPCC et al. 2013). Atmospheric concentrations of greenhouse gases and surface temperatures continue to increase and many additional changes in the climate system are predicted, including increased precipitation at high latitudes, decreased precipitation in subtropical regions, more intense and frequent weather phenomena and changes in wind, precipitation and temperature patterns. These meteorological changes will have important consequences for water and food availability, coastlines, ecosystems and human health – most of which will be negative. A major potential indirect effect of climate change on public health is predicted to arise via climate-induced changes in aeroallergens, the primary risk factors for allergic rhinitis. 1.1.4.1   Potential mechanisms linking climate to allergic rhinitisClimate change is predicted to affect allergic rhinitis by causing an overall increase in the length of pollen seasons, and thus the duration of allergen exposure. The geographic distribution of allergenic vegetation is also likely to be altered, leading to the introduction of new species in currently unaffected areas. These effects may lead to an increase in sensitization among non-sensitized individuals and in the duration of symptoms among those already suffering from the disease, thereby affecting both the incidence and prevalence of allergic rhinitis. Climate change may also lead to higher pollen counts, thus increasing the severity of allergic symptoms experienced (United States Environmental Protection Agency 2008). To support these predictions, there is already consistent evidence that changes in climate have and continue to affect the regional distribution and allergenicity of aeroallergens, especially, but not limited to, pollen and mould (Beggs 2004; D’Amato et al. 2007; Shea et al. 2008). Studies conducted in “real-world natural settings” may be most informative for assessing the extent to which climate change will affect aeroallergens. For example, a link between recent warming and increased duration of the ragweed pollen season was recently identified on a 15continental (North America) scale (Ziska et al. 2011). Further, ragweed in an urban site with higher temperatures and carbon dioxide concentrations, similar to those associated with projectedclimate change, grew faster, flowered earlier and produced significantly greater above-ground biomass and ragweed pollen compared to ragweed grown in a rural area (Ziska 2003). The authors of this latter study suggested that environmental conditions associated with projected climate change may already be present in small scale urbanized areas and may consequently already be inducing public health consequences. This is supported by a study conducted in natural conditions in which ozone, an environmental factor likely to increase in concentration as a result of climate change, increased the allergenicity of birch pollen (Beck et al. 2013). The atmospheric factors believed to have the most influence on the growth and distribution of aeroallergens, namely carbon dioxide concentrations, temperature, precipitation, vapour pressure (humidity) and wind speed and direction, are also predicted to affect air pollutants, for example, by increasing the frequency of urban air pollution episodes. Air pollutants appear to interact with pollen allergens to increase their allergenicity (Beck et al. 2013; Ghiani et al. 2012; Motta et al. 2006). Indeed, there are now several recent reviews that summarize the potential effects of climate change and air pollution on respiratory health (Barnes et al. 2013; D’Amato et al. 2010, 2014; Sario et al. 2013; Shea et al. 2008).1.1.4.2   Review of the literature on climate and allergic rhinitisDespite the strong evidence demonstrating climate-induced changes on aeroallergens and the known causal relationship between aeroallergens and allergic rhinitis onset and prevalence, studies examining the broader picture, that is, associations between climatic factors and allergic rhinitis, have yielded mostly inconsistent results. Positive associations between allergic symptoms and temperature, vapour pressure and precipitation have been reported by several geographically-limited (within one country) studies (Ariano et al. 2010; De Marco et al. 2002; Kim et al. 2011; Lee et al. 2003; Newhouse and Levetin 2004). Others, however, report null findings (Bhattacharyya 2009; Breton et al. 2006; Zanolin et al. 2004). One possible reason for these inconsistent results may be that climatic effects on aeroallergens (and presumably on allergic rhinitis) likely vary by geography and vegetation type (D’Amato and Cecchi 2008). 16Prior to the work conducted in Chapter 5 of this dissertation, only two studies had examined associations between climatic factors and allergic rhinitis using data from more than one country.The first study found that hayfever prevalence (in the last 12 months) was only associated with temperature during the hottest month using data on adults aged 20-44 years from 48 European centres (Verlato et al. 2002). The second study, based on children from 144 centres in 81 countries participating in the International Study of Asthma and Allergies in Childhood (ISAAC), reported no consistent relationships between allergic rhinitis symptom prevalence (in the last 12 months) and latitude, altitude, indoor relative humidity at 20 oC and measures of outdoor temperature and relative humidity (Weiland et al. 2004b). These primarily null results may be in part attributable to effect heterogeneity across areas, but possibly also to the fact that these studies only considered annual disease prevalence as the outcome. Climate change has and will continue to affect certain aeroallergens (pollens and moulds) more strongly than others (dog and cat). Individuals predominantly sensitized to climate-sensitive aeroallergens will thus likely be most affected by climate change. As aeroallergens, like air pollutants, do not recognize national boundaries, further large multi-country studies that consider climatic effects on differentallergic rhinitis phenotypes are vital to our understanding of the global impact of climate change on allergic diseases. 1.2   Data sources and study population1.2.1   Traffic, Asthma and Genetics collaboration 1.2.1.1   Study population and outcome dataThe Traffic, Asthma and Genetics (TAG) study population, described in (MacIntyre et al. 2013), is composed of 15,299 children recruited in six birth cohorts followed for at least seven years (the Children, Allergy, Milieu, Stockholm, Epidemiological Survey (BAMSE) (Emenius et al. 2003; Wickman et al. 2002), the Canadian Asthma Primary Prevention Study (CAPPS) (Chan-Yeung et al. 2000), the German Infant study on the influence of Nutritional Intervention plus environmental and genetic influences on allergy development (GINIplus) (Filipiak et al. 2007; von Berg et al. 2003), the Lifestyle related factors, Immune System and the development of Allergies in East and West Germany plus the influence of traffic emissions and genetics study (LISAplus) (Heinrich et al. 2002a), the Prevention and Incidence of Asthma and Mite Allergy 17study (PIAMA) (Brunekreef et al. 2002) and the Study of Asthma, Genes and Environment (SAGE) (Kozyrskyj et al. 2009)). All cohorts were recruited during the mid-to-late 1990s primarily through hospitals, clinics and outpatient practices near the time of birth and prospectively followed. The SAGE cohort is the only exception; children born in 1995 were identified from a healthcare registry at the age of eight years and thus earlier-age information was collected retrospectively and later-age information prospectively. All cohorts were originallypopulation-based except CAPPS, for which only high-risk children were recruited (those with at least one first-degree relative with asthma or two first-degree relatives with other IgE-mediated allergic diseases (atopic dermatitis, seasonal or perennial allergic rhinitis or food allergy)). However, the analyses in this dissertation for BAMSE and SAGE are based on subsets from nested case-control studies within the original cohorts for wheeze and asthma, respectively. Data on several health outcomes, environmental exposures and important covariates were collected via either parent- or self-completed questionnaires according to each cohort's respectiveinformation collection strategy and harmonized into common variables. A detailed description of this harmonization process and the recruitment and follow-up of each cohort has been published (MacIntyre et al. 2013). Chapter 2 is based on only the German GINIplus and LISAplus birth cohorts for which longitudinal outcome data were available up to ten years of age. Chapters 3 and 4 include data from all six birth cohorts participating in TAG. Allergic rhinitis status was derived from parent-completed questionnaire data and was based on aphysician diagnosis of allergic rhinitis or hayfever for CAPPS, GINIplus, LISAplus and SAGE, areport of symptoms for PIAMA and either a doctor diagnosis or report of symptoms for BAMSE.Aeroallergen sensitization was assessed by skin prick testing for CAPPS and SAGE. A positive reaction was defined as having a wheal diameter > 3 mm. For BAMSE, GINIplus, LISAplus and PIAMA, aeroallergen sensitization was assessed by measuring allergen-specific IgE. A positive reaction was defined as any value > 0.35 kU/L. For Chapters 2 and 3, allergic rhinitis and sensitization were modelled as two separate outcome variables. For Chapter 4, a concomitant positive report of both allergic rhinitis and aeroallergen sensitization was used to define the primary outcome. The frequency of follow-up of each cohort is depicted in Figure 1. Further 18details on the cohorts' study designs and outcome assessments, including the specific aeroallergens tested, are provided in Table 2. 19Figure 1: Follow-up time points for each cohort participating in the Traffic, Asthma and Genetics collaboration* = aeroallergen sensitization data were also availableTable 2:  Characteristics of cohorts participating in the Traffic, Asthma and Genetics collaborationCohort Areas (country) Study type Recruitment Sample size a Allergic rhinitis definition Aeroallergens testedBAMSE Jarfalla, Solna, Sundbyberg, Stockholm (Sweden)Population based birthcohort with wheeze nested case-control 1994-6 982 Symptoms (sneezing, runny or blocked nose, itchy, red and watery eyes) after exposure to furred pets or pollen or a medical diagnosis of allergic rhinitis since previous questionnaireBirch, cat, dog, house dust mite (Dermatophagoides pteronyssinus), mould (Cladosporium herbarum), mugwort, timothy grassCAPPS Vancouver, Winnipeg (Canada)Randomized controlled study with asthma intervention1995 545 Medical diagnosis of allergic rhinitis assessed at follow-upAlternaria, cat, cockroaches, dog, feathers, grass, house dust mites, mould (Cladosporium herbarum), ragweed, trees, weedsGINIplus Munich, Wesel, Leipzig, Bad Honnef (Germany)Population based birthcohort. Subset selected for nutritionalintervention1995-8 5991 Medical diagnosis of allergic rhinitis or hayfever during the last12 months Birch, cat, dog, house dust mite (Dermatophagoides pteronyssinus), mould (Cladosporium herbarum), mugwort, rye, timothy grassLISAplus Munich, Wesel (Germany)Population based birthcohort1997-9 3095 Medical diagnosis of allergic rhinitis or hayfever during the last12 months Birch, cat, dog, house dust mite (Dermatophagoides pteronyssinus), mould (Cladosporium herbarum), mugwort, rye, timothy grassPIAMA Northern, central, and western communities (The Netherlands)Population based birthcohort. Subset selected for mattress cover intervention1996-7 3963 Sneezing, runny/blocked nose during the last 12 months without cold or fluAlternaria, birch, cat, Dactylis, dog, house dust mite (Dermatophagoidespteronyssinus)SAGE Winnipeg (Canada)Population-based cohort with asthma nested case-control1995 723 Medical diagnosis of allergic rhinitis assessed at follow-upCat, dog, feathers, grass, ragweed, trees, weeds a Number of children included in the Traffic, Asthma and Genetics database.201.2.1.2   Air pollution estimatesThroughout this dissertation, nearly all individual-level air pollution estimates were derived using LUR models. These models rely on two principles: that readily measurable predictor variables relating to land-use can be used to estimate environmental conditions and that the relationship between air pollutant concentrations and these land-use predictor variables can be determined using a small sample of air quality measurements (Briggs et al. 2000). Under these assumptions, regression models can be built to relate the small sample of air quality measurements with the land-use variables. These models can be subsequently used to assign a unique estimate of the annual average exposure to a pollutant to any participant's address. Individual NO2 estimates were available for participants from all birth cohorts in TAG. For Chapter 3, these estimates were derived using LUR modelling for all cohorts except BAMSE. The LUR models developed for GINIplus (Munich city), LISAplus (Munich city) and PIAMA were created as part of the Traffic Related Air Pollution and Childhood Asthma (TRAPCA) collaboration (Brauer et al. 2003; Hoek et al. 2002). Using a similar methodology, LUR models were developed for the two Canadian cohorts, CAPPS and SAGE (Allen et al. 2011; Henderson et al. 2007), and the cities of Wesel and Leipzig within the GINIplus and LISAplus cohorts (Hochadel et al. 2006). NO2 estimates for the BAMSE cohort were obtained from a dispersion model (Nordling et al. 2008). Individual PM2.5 mass and PM2.5 absorbance estimates, calculated using the same methodology as for NO2, were available for a subset of the cohorts. All of the aforementioned LUR models contained variables describing local traffic intensity, as well as other variables such as population and address density.For Chapters 2 and 4, NO2, PM2.5 mass and PM2.5 absorbance estimates derived as part of the ESCAPE project (Beelen et al. 2013; Cyrys et al. 2012; Eeftens et al. 2012a, 2012b) were used instead for all the European cohorts (BAMSE, GINIplus, LISAplus and PIAMA). All of the PM2.5 mass and PM2.5 absorbance models for these study areas contained traffic predictor variables describing local traffic intensity. The predictor variables in the NO2 models were more variable across study sites and generally included industry and/or population density variables. The ESCAPE-based estimates became available while this dissertation was being conducted and have the advantage of all being derived from the same standardized and harmonized protocol. 21Finally, individual-level ozone estimates were available for PIAMA, GINIplus and LISAplus participants, calculated as part of the Air Pollution Modelling for Support to Policy on Health and Environmental Risk in Europe (APMoSPHERE; www.apmosphere.org) project (Beelen et al. 2009), and for CAPPS participants, calculated using ambient monitoring network data (Marshall et al. 2008). Unlike the other pollutants, the ozone estimates were not derived using any specific traffic components and thus represent air pollution in general rather than traffic-related air pollution. 1.2.1.3   Genetic dataGenetic data on several candidate SNPs believed to be involved in inflammation and oxidative stress metabolism were available for all cohorts (details of genotyping provided in section 3.2.4).Imputed genome-wide data were also available for a subset of the birth cohorts (details of imputation provided in Appendix A). 1.2.2   International Study of Asthma and Allergies in Childhood 1.2.2.1   Study population and outcome dataISAAC is the largest worldwide collaborative research project. In ISAAC Phase One, worldwide prevalences of several allergic diseases were established (for example, (Strachan et al. 1997)). In ISAAC Phase Two, the relative importance of potential risk and protective factors that may contribute to international differences in prevalence were investigated using a smaller number of selected centres (Weiland et al. 2004a). ISAAC Phase Three, which was conducted between 2001and 2003, aimed to examine time trends in disease prevalence (as compared to Phase One), develop a more comprehensive map of disease prevalence and assess the potential impacts of environmental factors (Ellwood et al. 2005). The ecological global study described in Chapter 5 of this dissertation uses data from ISAAC centres that collected valid data on monthly rhinitis symptoms via standardized parent- (for 6-7 year-old children) or child- (for 13-14 year-old teenagers) completed questionnaires during ISAAC Phase Three (protocols available on the ISAAC website; isaac.auckland.ac.nz). These monthly rhinitis symptom data were used to definethe prevalences of intermittent (at least one symptom report but not for two consecutive months) and persistent (symptoms for at least two consecutive months) rhinitis symptoms per centre. In 22total, data from 222 centres in 94 countries (for 13-14 year-old teenagers) and 135 centres in 59 countries (for 6-7 year-old children) were included. 1.2.2.2   Environmental factors and covariatesData on monthly mean daily temperature (Celsius), total precipitation (millimeter) and vapour pressure (hectopascal), averaged over the period of 1991 to 2000 for 0.5o x 0.5o grids (approximately 3025 km2), were obtained from the Intergovernmental Panel on Climate Change Data Distribution centre (Mitchell 2004; Mitchell and Jones 2005). Normalized Difference Vegetation Index (NDVI) data, a biomass density indicator, were obtained for 2005 from the Global Land Cover Facility at a resolution of 0.07o on a 16-day basis and averaged per month (Pinzon et al. 2005; Tucker et al. 2005). Data on gross national income (GNI) per capita for 2003were obtained from the World Bank (Atlas Method 2003) (World Bank 2012). When missing, GNI data were imputed using information from the Central Intelligence Agency World Fact Book (2003; seven countries) (Central Intelligence Agency 2007). Population density data for 2005 were obtained from the Socioeconomic Data and Applications centre (Socioeconomic Data and Applications Center 2004). 1.3   Dissertation objectivesThe objective of this dissertation was to examine the influence of environmental (traffic-related air pollution and climate) and genetic risk factors, and their potential interactions, on the development of allergic rhinitis during childhood. As traffic-related air pollution concentrations are unlikely to decrease in the coming years and climate change will continue to occur, the issuesconsidered in this dissertation are of significant importance.The specific objectives of the dissertation were: 1. To examine associations between traffic-related air pollution and the development of asthma, allergic rhinitis and aeroallergen sensitization among children followed from birth to ten years of age in three areas in Germany (Chapter 2). 232. To examine whether traffic-related air pollution exposure is associated with childhood allergic rhinitis and aeroallergen sensitization at school-age in six birth cohorts, and to assess the influence of ten SNPs related to inflammation and oxidative stress metabolism in the GSTP1, TNF, TLR2 and TLR4 genes on any possible association (Chapter 3).3. To examine whether seven SNPs located at the asthma-risk 17q21 locus are associated with allergic rhinitis from early childhood to adolescence, and whether possible associations are modified in subjects also diagnosed with asthma (Chapter 4).4. To examine between- and within-country associations of climate measures and NDVI, a biomass density indicator, with the prevalence of two allergic rhinitis phenotypes in a global context (Chapter 5).1.4   Dissertation structureThe layout of this thesis conforms to the manuscript-based thesis guidelines of the University of British Columbia. Chapters 2 and 3 have been published in peer reviewed journals, Chapter 4 hasbeen submitted for publication and Chapter 5 has been accepted for publication in a peer reviewed journal. Chapter 2 is a longitudinal study in which the potential effects of traffic-relatedair pollution exposure on allergic health outcomes are explored in two German birth cohorts. This dataset is extended in Chapter 3 to include four more birth cohorts, and the influence of genetic risk factors and potential gene-environment interactions on allergic rhinitis and sensitization are explored. Chapter 4 focuses on the role of genetic variability at a known asthma-risk gene locus on allergic rhinitis, with and without comorbid asthma. In Chapter 5, climate and vegetation effects on the prevalences of two allergic rhinitis phenotypes are assessed in a global ecological study. Finally, the conclusions and implications of this dissertation, as well as its strengths and limitations, are discussed in Chapter 6. Recommendations for further research are provided. 242   A longitudinal analysis of associations between traffic-related air pollution with asthma, allergies and sensitization in the GINIplusand LISAplus birth cohorts 12.1   IntroductionThe rapid rise in asthma and allergic diseases in recent decades suggests a role for environmentalfactors. Whether traffic-related air pollution contributes to the development of childhood asthma,allergy and related symptoms has been the topic of several studies. Recent reviews (Guarnieri and Balmes 2014; Tager et al. 2010) and a meta-analysis (Anderson et al. 2013) concluded that the evidence for an association between TRAP and asthma exacerbation is sufficient/strong to support a causal relationship and support for this association continues to build, including from genetic and gene-environment studies (Carlsten and Melén 2012; Holloway et al. 2012). The evidence for other non-asthma allergic phenotypes, such as allergic rhinitis, eczema and aeroallergen sensitization, is generally weaker (Tager et al. 2010).Several epidemiological studies have examined the link between air pollution and allergic diseases in children in early life (Brauer et al. 2002; Clark et al. 2010; Gehring et al. 2002; Morgenstern et al. 2007) and childhood (Brauer et al. 2007; Carlsten et al. 2011; McConnell et al. 2006; Morgenstern et al. 2008; Nordling et al. 2008). All of these aforementioned studies report positive associations with at least one respiratory or allergic health outcome, despite the challenges associated with accurately assessing air pollution exposure levels and allergic health status in young children. A study on older children (ten to 11 years) found no association between traffic-related air pollution and general sensitization, although sensitization to house dust mites was associated with lifetime air pollution exposure (Oftedal et al. 2007). Two recent large cross-sectional multicentre European studies found no association between individually assigned traffic-related air pollution exposures and asthma or wheeze at four to five and eight to 1- A version of this manuscript has been published: Fuertes E, Standl M, Cyrys J, Berdel D, von Berg A, Bauer CP, Krämer U, Sugiri D, Lehmann I, Koletzko S, Carlsten C, Brauer M, and Heinrich J. 2013. A longitudinal analysis of associations between traffic-related air pollution with asthma, allergies and sensitization in the GINIplus and LISAplus birth cohorts. PeerJ. 1:e193. dx.doi.org/10.7717/peerj.19325ten years in approximately 10,000 children (Mölter et al. 2014) and sensitization at four to six and eight to ten years in approximately 7,000 children (Gruzieva et al. 2014).Only two longitudinal epidemiological analyses incorporating health data from birth up to eight years of age have been conducted. Gehring et al. (2010) reported positive associations between traffic-related air pollution and the incidence and prevalence of asthma and the prevalence of asthma symptoms in a study of 3,863 Dutch children (Gehring et al. 2010). Positive associations were also found for allergic rhinitis, but only among non-movers. No associations were found forsensitization at eight years, despite a previous positive finding at four years in this same cohort (Brauer et al. 2007). Gruzieva et al (2012) also found no association between traffic-related air pollution and allergic sensitization at eight years in a Swedish birth cohort (Gruzieva et al. 2012),again despite a previous documented association at four years of age (Nordling et al. 2008). These results may suggest that the timing of air pollution exposure relative to disease onset may be important. This hypothesis, as well as the long-term impact of air pollution on allergic diseaseprevalence in later childhood, is best explored in the context of longitudinal birth cohorts (Braback and Forsberg 2009).Recently, the ten year follow-ups of the GINIplus and LISAplus birth cohort studies were completed. These studies are unique in their long-term and frequent follow-up, large sample sizes, extensive health, demographic and lifestyle information, as well as the availability of air pollution estimates at different time points during life (birth, six and ten years). Previously published results from these cohorts suggest a possible adverse role of traffic-related air pollutionon symptoms during the first two years of life (Gehring et al. 2002; Morgenstern et al. 2007) and several allergic outcomes at six years (Morgenstern et al. 2008) for children living in the Munich (GINI/LISA South) area. In the northern part of these cohorts (Wesel area, GINI/LISA North), associations were found only with the prevalence of eczema (Krämer et al. 2009). The current study builds on these past efforts by examining whether NO2, PM2.5 mass, PM2.5 absorbance and ozone concentrations, the latter of which has not been previously considered in the context of theGINIplus and LISAplus studies, are associated with the prevalence of asthma, allergic rhinitis and aeroallergen sensitization among children followed for ten years in three areas in Germany. 262.2   Methods2.2.1   Study populationGINIplus is a prospective birth cohort of 5,991 children born at full-term and normal weight recruited in the areas of GINI/LISA South (a predominantly urban area) and GINI/LISA North (apredominantly rural area) between 1995 and 1998. Children with at least one atopic parent or sibling were allocated to an intervention study arm which investigated the effect of different hydrolyzed formulas consumed during the first year of life on allergy development (N=2,252) (von Berg et al. 2003). All children whose parents did not give consent for the randomized clinical trial or who did not have a family history of allergic diseases were allocated to the observation study arm (N=3,739). LISAplus is a population-based prospective birth cohort of 3,095 children born at full-term and normal weight recruited in GINI/LISA South, GINI/LISA North, Leipzig (LISA East; formally part of Eastern Germany) and Bad Honnef between 1997 and 1999 (original study size 3,097 but two participants withdrew their consent to participate). Detailed descriptions of the recruitment and follow-up strategy for both cohorts are available (Filipiak et al. 2007; Heinrich et al. 2002a). Both studies were approved by the local Ethics Committees (the Bavarian Board of Physicians (reference numbers: 01212 and 07098), University of Leipzig (reference number: 345/2007), and Board of Physicians of North-Rhine-Westphalia (reference numbers: 2003355 and 2008153)) and written consent was obtained from all parents of the participants.2.2.2   Questionnaire dataDemographic and health data were collected using parent-completed questionnaires administeredwhen the child was born and when the child was one, two, three, four, six and ten years-old for GINIplus participants and six, 12 and 18 months and two, four, six and ten years-old for LISAplus participants. Only information collected from three years onwards is included in this study as it is difficult to accurately diagnose allergic health outcomes at very young ages.A doctor diagnosis of asthma was defined as a positive response to “In the last 6/12 months, has your child been diagnosed with asthma?” A doctor diagnosis of allergic rhinitis was defined as a positive response to “In the last 6/12 months, has your child been diagnosed with hayfever or 27allergic rhinitis?” For GINIplus, this information was collected in one question at the three, four and six year follow-ups (diagnosis of hayfever or allergic rhinitis, as aforementioned). For the data collected at the ten year follow-up of GINIplus and for all LISAplus follow-ups, this information was collected in two separate questions which were subsequently combined. If the follow-up covered a period of greater than one year, the prevalence of the diagnosis was asked separately for each year. Ultimately, yearly diagnoses for both asthma and allergic rhinitis were available for age three to ten years. Eyes and nose symptoms were assessed at ages four, six and ten years only, and were defined based on two concomitant positive responses to “In the past 12 months, has your child had a clogged or itchy nose when he/she did not have a cold?” and “In thepast 12 months, has your child had a clogged or itchy nose accompanied by watery eyes?”, which is consistent with the rhinoconjunctivitis definition used in ISAAC (Aït Khaled et al. ‐2009).Specific IgE against common aeroallergens was assessed at ages six and ten years using the standardized CAP-RAST FEIA method (ThermoFischer, Freiburg, Germany). Sensitization to aeroallergens (SX1: birch, cats, dogs, house dust mites (Dermatophagoides pteronyssinus), mould (Cladosporium herbarum), mugwort, rye and timothy grass) was measured by a screeningtest, followed by single specific allergen tests if the overall screening test was positive. The detection limit of the CAP-RAST FEIA method is 0.35 kU/L IgE. A test was defined as positive if the specific IgE value was greater or equal to this limit. Although some aeroallergens can be present in both indoor and outdoor environments, in this analysis, birch, mugwort, rye and timothy grass were considered as outdoor aeroallergens and cats, dogs, house dust mites and moulds were classified as indoor aeroallergens.2.2.3   Air pollution estimatesAs part of the ESCAPE collaboration, ambient concentrations of NO2, PM2.5 mass and PM2.5 absorbance were estimated for each child's home address at birth, six and ten yearsusing LUR models for children living in GINI/LISA South and GINI/LISA North (Beelen et al. 2013; Cyrys et al. 2012; Eeftens et al. 2012a, 2012b). For children living in LISA East, NO2 and PM2.5 mass estimates at the home address at birth and six years were derived from a similar LUR 28model developed as part of the earlier TRAPCA collaboration which was conducted in the city ofMunich, Germany (Brauer et al. 2003; Cyrys et al. 2003; Hoek et al. 2002). PM2.5 absorbance data are not available for LISA East. Ozone estimates from the APMoSPHERE project were also available. For this latter pollutant, a 1 x 1 km resolution concentration map was developed across15 European Union States using several European-wide datasets on monitored air pollution, land cover, altitude, transport networks, meteorology and population (Beelen et al. 2009). Data from this map were used to assign ozone concentrations to the birth and six year home addresses of allparticipants. Table 3 summaries the key characteristics of the models used to estimate individual-level air pollution concentrations in the current study. As no air pollution models were developedfor the Bad Honnef area, children from this city were excluded from all analyses  (N=306).Table 3:  Characteristics of models used to estimate air pollution exposuresProject AreasAir pollution sampling description Pollutants R2 RMSEAssociated publicationsESCAPE GINI/LISA South20/40 particulate matter/NO2 sites in Munich, Augsburg and small nearby towns sampled for three two-week intervals between 10/2008 and 11/2009NO2 0.86 5.5 (Beelen et al. 2013; Cyrys et al.2012)PM2.5 mass 0.78 1.0 (Eeftens et al. 2012a, 2012b)PM2.5 absorbance 0.91 0.2GINI/LISA North20/40 particulate matter/NO2 sites in Dortmund, Duisburg, Essen and smaller towns sampled for three two-week intervals between 10/2008 and 10/2009NO2 0.89 4.3 (Beelen et al. 2013; Cyrys et al.2012)PM2.5 mass 0.88 0.9 (Eeftens et al. 2012a, 2012b)PM2.5 absorbance 0.97 0.1TRAPCA LISA East 30 sites sampled for four two-week intervals between 04/2004 and 03/2005 NO2 0.81 2.7PM2.5 mass 0.55 0.1APMoSPHERE GINI/LISA South and North, and LISA EastAir pollution data for 2001 obtained from Airbase (air quality database from routine air pollution monitoring covering 15 European member states; resolution 1 x 1 km)Ozone 0.70 7.7 (Beelen et al. 2009)R2 = Model explained variance; RMSE = Root-mean standard errors 292.2.4   Statistical analysisAll analyses were conducted using the statistical program R, version 2.13.1 (R Core Team 2012).Differences in population characteristics were assessed using the Chi-square test. Longitudinal associations between air pollutants estimated at the birth addresses with the prevalence of doctor diagnosed asthma and allergic rhinitis, nose and eye symptoms and aeroallergen sensitization were analyzed using generalized estimation equations with a logit link (geeglm function from thegeepack package (Halekoh et al. 2006)). An exchangeable correlation structure was used to account for repeated observations on the same individual. All models were adjusted for sex, age, presence of older siblings (yes/no), parental history of atopy, parental education (originally defined using three categories based on the highest number of years of education of either parent, but collapsed into two categories due to low numbers in the lowest category: less than or equal to ten years versus more than ten years), maternal smoking during pregnancy, secondhand smoke exposure in the home (ever between birth and four years), contact with furry pets during the first year of life, use of gas stove for cooking during the first year of life, dampness or indoor moulds in the home during the first year of life, intervention participation (GINIplus participants only), cohort and geographical area (for total models only). These confounders are the same as in previous analyses of these cohorts (Krämer et al. 2009; Morgenstern et al. 2008). Models for LISA East were not adjusted for intervention orcohort as only children for the LISAplus cohort were recruited from this area. Data on pneumonia infections in the first two years of life, which has been associated with air pollutants in a multicentre study (MacIntyre et al. 2014b), and percent of total green space and population density in a 500 m buffer around the home address (GINI/LISA South and North only) were also available for sensitivity analyses. Elevated risks of disease were analyzed per interquartile range increase of each air pollutant. ORs and 95% confidence intervals (CI) are presented for the total population and by geographical area.As 59% of the study population reported moving at least once during the first ten years of life, several sensitivity analyses were conducted to assess potential exposure misclassification. First, all models were rerun using air pollutants estimated at the six and ten year home addresses rather30than the birth address. Second, associations were assessed using the average of the birth, six and ten year air pollution concentrations for NO2, PM2.5 mass and PM2.5 absorbance, and the average of the birth and six year concentrations for ozone (information at ten years was not available). These averages should better reflect a true life-time exposure for participants who moved between birth and ten years. Lastly, associations were stratified by lifetime moving behaviour. To address the potential importance of exposure timing, mutually adjusted models including bothbirth and six or ten year address pollution concentrations were examined when the pollutants were not highly correlated (Pearson correlation coefficient < 0.70). Each such correlation (for example, between NO2 assessed at the birth address and NO2 assessed at the six year address) wasexamined separately in the total and area-specific datasets.2.3   ResultsIn total, 6,604 children had available information on at least one health outcome and air pollutant, 3,655 of which also had available sensitization data at one time point (Figure 2). Population characteristics and the prevalence of health outcomes at the age of ten years are provided for the total population and by area (Table 4), and for the subset of the total population who provided blood samples that were assessed for aeroallergen sensitization (Table 5).31Figure 2: Flow chart of study populationTable 4:  Characteristics of study participantsCharacteristicTotal population(N=6604)GINI/LISA South(N=3362)GINI/LISA North(N=2551)LISA East (N=691)n/N % n/N % n/N % n/N %Males 3386/6604 51.3 1747/3362 52.0 1304/2551 51.1 335/691 48.5Presence of older siblings 3021/6588 45.9 1398/3357 41.6 1378/2542 54.2 245/689 35.6Parental education < 10 years 2405/6573 36.6 759/3350 22.7 1342/2542 52.8 304/681 44.6> 10 years 4168/6573 63.4 2591/3350 77.3 1200/2542 47.2 377/681 55.4Smoking During pregnancy 943/6248 14.7 430/3268 13.2 403/2494 16.2 110/666 16.5Ever in home (0-4 years) 2212/5663 39.1 887/2987 29.7 1096/2099 52.2 229/577 39.7Parental history of atopy 3750/6534 57.4 2232/3340 66.8 1221/2535 48.2 297/659 45.1Owned furry pet during early life 1126/6343 17.8 504/3237 15.6 444/2430 18.3 178/676 26.3Gas used in home during early life 469/6447 7.3 257/3301 7.8 106/2474 4.3 106/672 15.8Mould/dampness in home during early life 1330/5312 25.0 797/2889 27.6 379/1751 21.6 154/672 22.9Moved between one and 10 years 3197/5407 59.1 1739/2935 59.3 1061/1930 55.0 397/542 73.2Cohort GINIplus 4386/6604 66.4 2107/3362 62.7 2279/2551 89.3 0/691 0.0LISAplus 2218/6604 33.6 1255/3362 37.3 272/2551 10.7 691/691 100.0Intervention participation a 1935/6604 29.3 1031/3362 30.7 904/2551 35.4 0/691 0.0Doctor diagnosed asthma (10 years) 164/4696 3.5 82/2589 3.2 68/1684 4.0 14/423 3.3Doctor diagnosed allergic rhinitis (10 years) 460/4623 10.0 275/2528 10.9 140/1676 8.4 45/419 10.7Eyes and nose symptoms (10 years) 628/4736 13.3 389/2585 15.0 173/1722 10.0 66/429 15.4Sensitized to aeroallergens (10 years) 1100/2735 40.2 678/1581 42.9 303/867 34.9 119/287 41.5Sensitized to indoor aeroallergens (10 years) 748/2732 27.4 449/1579 28.4 214/866 24.7 85/287 29.6Sensitized to outdoor aeroallergens (10 years) 809/2734 29.6 509/1581 32.2 222/866 25.6 78/287 27.2a Intervention only part of the GINIplus cohort. n = number of cases; N = number of children with available data32Table 5:  Characteristics of the total study participants with available serology data Characteristic n/N %Males 1891/3655 51.7Presence of older siblings 1679/3648 46.0Parental education < 10 years 1204/3640 33.1> 10 years 2436/3640 66.9Smoking During pregnancy 487/3581 13.6Ever in home (0-4 years) 1314/3473 37.8Parental history of atopy 2229/3621 61.6Owned furry pet during early life 590/3532 16.7Gas used in home during early life 270/3597 7.5Mould/dampness in home during early life 855/3341 25.6Moved between one and 10 years 1882/3296 57.1Area GINI/LISA South 2017/3655 55.2GINI/LISA North 1241/3655 34.0LISA East 397/3655 10.9Cohort GINIplus 2388/3655 65.3LISAplus 1267/3655 34.7Intervention participation a 1223/3655 33.5Doctor diagnosed asthma (10 years) 137/3208 4.3Doctor diagnosed allergic rhinitis (10 years) 357/3168 11.3Eyes and nose symptoms (10 years) 497/3242 15.3Sensitized to aeroallergens (10 years) 1100/2735 40.2Sensitized to indoor aeroallergens (10 years) 748/2732 27.4Sensitized to outdoor aeroallergens (10 years) 809/2734 29.6a Intervention only part of the GINIplus cohort. n = number of cases; N = number of children with available data33Compared to the original cohorts, children included in this study and the subset who provided blood samples were less likely to have been exposed to furry pets early in life or to have moved at least once between birth and ten years, but more likely to have participated in the nutritional intervention (GINIplus participants only) and to have a parent with more than ten years of education and a history of atopy. Furthermore, the children included in the main analyses were less likely to have been exposed to smoke in utero. The subset of children who provided blood samples were less likely to have been exposed to tobacco smoke in utero and during early life (birth to four years) or later childhood (six to ten years). 2.3.1   Distribution of outcomesThe period prevalences of doctor diagnosed asthma and allergic rhinitis are presented in Figure 3. In the total population, the annual prevalence of doctor diagnosed asthma and allergic rhinitis ranged from 1.1% to 3.5% and from 1.6% to 10.0%, respectively. The distribution of doctor diagnosed asthma and allergic rhinitis prevalence across areas was similar. Only at age nine and ten years were the rates of doctor diagnosed allergic rhinitis significantly different across areas. The pooled prevalence of reported eye and nose symptoms rose steadily with age and differed byarea (3.7%, 6.8% and 13.3% for ages four, six and ten years, respectively), as did the prevalence of aeroallergen sensitization (29.0% and 40.2% for ages six and ten years, respectively). The prevalence of eye and nose symptoms and aeroallergen sensitization was highest in GINI/LISA South and lowest in GINI/LISA North at almost all ages. 342.3.2   Air pollution estimatesThe distributions of annual average air pollution estimates are presented in Table 6 for the total study population and by area. At the birth addresses, mean NO2 concentrations were highest in GINI/LISA North (23.8 μg/m3), mean PM2.5 mass concentrations were highest in LISA East (17.5μg/m3) and mean PM2.5 absorbance and ozone concentrations were highest in GINI/LISA South (1.7 10-5/m and 45.8 μg/m3, respectively). As the GINI/LISA North area is affected by the neighboring industrial Ruhr area in Germany, the elevated levels of PM2.5 mass but not PM2.5 absorbance in this area may suggest that there are PM2.5 mass sources other than traffic. The correlations between NO2 and the other pollutants were moderate in the pooled data (0.30, 0.32 and -0.50 for PM2.5 mass, PM2.5 absorbance and ozone, respectively).35Figure 3: Period prevalence of children with doctor diagnosed asthma (bars with diagonal lines) or allergic rhinitis (filled bars) at ages three to ten years in the total population (A) and stratified by area: GINI/LISA South (B), GINI/LISA North (C) and LISA East (D)Table 6:  Distribution of estimated annual average concentrations of NO2, PM2.5 mass, PM2.5 absorbance and ozone at the birth addresses in the total dataset and per areaAir pollutant N min 0.25 median mean 0.75 max IQRNO2 (μg/m3)Total population 6485 11.5 18.9 22.2 22.4 25.0 62.8 6.1GINI/LISA South 3306 11.5 17.3 20.7 21.7 25.4 61.1 8.1GINI/LISA North 2491 19.7 21.8 23.2 23.8 25.1 62.8 3.3LISA East 688 18.5 18.7 18.8 20.8 22.4 34.8 3.7PM2.5 mass (μg/m3)Total population 6485 0.4 13.3 15.4 15.3 17.3 21.5 4.0GINI/LISA South 3306 10.6 12.8 13.3 13.4 14.0 18.3 1.2GINI/LISA North 2491 15.8 16.9 17.3 17.4 17.8 21.5 0.9LISA East 688 0.4 17.2 17.8 17.5 18.34 20.1 1.2PM2.5 absorbance (10-5/m)Total population 5797 1.0 1.2 1.5 1.5 1.7 3.6 0.5GINI/LISA South 3306 1.3 1.6 1.7 1.7 1.8 3.6 0.2GINI/LISA North 2491 1.0 1.1 1.2 1.2 1.3 3.1 0.2LISA East NA NA NA NA NA NA NA NAOzone (μg/m3)Total population 6604 32.3 39.6 42.9 42.5 44.8 59.4 5.2GINI/LISA South 3362 34.1 44.1 44.7 45.8 45.5 59.4 1.4GINI/LISA North 2551 32.3 34.4 38.0 38.2 41.4 54.3 7.0LISA East 691 38.0 41.0 41.8 41.9 42.5 52.9 1.5IQR = interquartile range; max = maximum; min = minimum; N = number of children with air pollutant estimates and data on at least one health outcome; NA = not available2.3.3   Total and area-specific associationsCrude and adjusted associations between air pollution concentrations at the birth address and the prevalence of health outcomes were similar (adjusted associations provided in Table 7). The area-specific results were heterogeneous. In GINI/LISA North, the estimates for allergic rhinitis and eye and nose symptom prevalence were elevated for PM2.5 mass (1.22 [0.99,1.50] and 1.19 [0.99, 1.43], respectively). In LISA East, the estimates for ozone were elevated for all four outcomes, and that for allergic rhinitis and nose and eye symptom prevalence reached statistical significance (1.30 [1.02, 1.64] and 1.35 [1.16, 1.59], respectively). For GINI/LISA South, two 36associations with aeroallergen sensitization were significant (0.84 [0.73, 0.97] for NO2 and 0.87 [0.78, 0.97] for PM2.5 absorbance), as well as the association between allergic rhinitis and PM2.5 absorbance (0.83 [0.72, 0.96]). Given the heterogeneous area-specific effects, and the large influence of the GINI/LISA South cohort which represents 50.9% of the study population, risk estimates for the total population were generally null. 37Table 7:  Total and area-specific associations between air pollutants estimated to the birth address and health outcomes during the first ten years of life Total population GINI/LISA South GINI/LISA North LISA EastN OR [95% CI] N OR [95% CI] N OR [95% CI] N OR [95% CI]NO2 Asthma 4585 0.89 [0.73, 1.08] 2524 0.86 [0.62, 1.18] 1545 0.95 [0.77, 1.18] 516 1.02 [0.69, 1.50]Allergic rhinitis 4586 0.96 [0.85, 1.09] 2525 0.86 [ 0.71, 1.02] 1545 1.10 [0.96, 1.26] 516 1.18 [0.91, 1.54]Eyes and nose symptoms 4586 0.96 [0.87, 1.05] 2525 0.90 [0.78, 1.04] 1545 1.03 [0.92, 1.15] 516 0.96 [0.75, 1.24]Aeroallergen sensitization 3013 0.92 [0.84, 1.01] 1689 0.84 [0.73, 0.97] 975 1.06 [0.93, 1.21] 349 1.12 [0.86, 1.44]PM2.5 mass Asthma 4585 0.97 [0.59, 1.58] 2524 0.96 [ 0.74, 1.25] 1545 0.89 [ 0.64, 1.23] 516 1.06 [0.83, 1.35]Allergic rhinitis 4586 0.87 [0.60, 1.26] 2525 0.89 [0.76, 1.06] 1545 1.22 [0.99, 1.50] 516 0.95 [0.81, 1.11]Eyes and nose symptoms 4586 0.93 [ 0.67, 1.29] 2525 0.96 [0.84, 1.10] 1545 1.19 [0.99, 1.43] 516 0.93 [ 0.83, 1.05]Aeroallergen sensitization 3013 1.10 [0.83, 1.45] 1689 1.01 [0.89, 1.14] 975 1.14 [0.95, 1.37] 349 1.03 [ 0.89, 1.19]PM2.5 absorbanceAsthma 4069 0.82 [ 0.55, 1.21] 2524 0.94 [ 0.75, 1.18] 1545 0.88 [ 0.67, 1.14] NA NAAllergic rhinitis 4070 0.75 [0.58, 0.96] 2525 0.83 [0.72, 0.96] 1545 1.00 [0.84, 1.20] NA NAEyes and nose symptoms 4070 0.91 [0.75, 1.10] 2525 0.93 [ 0.84, 1.04] 1545 1.02 [0.88, 1.19] NA NAAeroallergen sensitization 2664 0.82 [0.68, 0.99] 1689 0.87 [0.78, 0.97] 975 1.02 [0.89, 1.18] NA NAOzone Asthma 4649 1.20 [0.98, 1.48] 2569 1.05 [0.98, 1.13] 1562 1.23 [ 0.78, 1.94] 518 1.10 [ 0.86, 1.42]Allergic rhinitis 4650 1.02 [0.90, 1.16] 2570 1.00 [0.96, 1.05] 1562 0.92 [0.68, 1.27] 518 1.30 [1.02, 1.64]Eyes and nose symptoms 4650 0.97 [0.87, 1.08] 2570 1.00 [0.96, 1.03] 1562 0.77 [0.57, 1.04] 518 1.35 [1.16, 1.59]Aeroallergen sensitization 3049 0.99 [0.89, 1.09] 1716 1.01 [0.98, 1.04] 982 0.79 [0.61, 1.02] 351 1.17 [0.96, 1.43]ORs are calculated per interquartile increase of each air pollutant. Models are adjusted for sex, age, parental history of atopy, parental education, older siblings, maternal smoking during pregnancy, secondhand smoke exposure in the home, contact with furry pets, use of gas stove for cooking, home dampness or indoor mould, intervention participation, cohort and area (total models only). Bold = p-value < 0.05; CI = confidence intervals; N = number of children included in the model; NA = not available; OR = odds ratio382.3.4   Sensitivity analysesAssociations with sensitization did not differ when aeroallergens were stratified into indoor and outdoor categories. Analyses which considered atopic asthma and more general allergic rhinitis (doctor diagnosis or nose and eye symptoms) as alternate outcomes yielded similar associations. The results remained robust when the models were further adjusted for percent total green space and population density in a 500 m buffer around the home address (not done for LISA East) and upon adjustment for pneumonia infections in the first two years of life. Additional analyses stratified by sex, parental history of atopic disease and secondhand smoke exposure in the home during early-life did not reveal a vulnerable subgroup.Air pollution concentrations estimated to the birth, six and ten year addresses were generally highly correlated with one another. Consequently, the risk estimates obtained using air pollution concentrations estimated at the six (Figure 4) and ten (Figure 5) year home addresses, and using the average of the birth, six and ten year air pollution estimates (Figure 6), were similar to those reported for air pollutants estimated at the birth address. When the pooled analyses were run stratified by moving status (never moved versus moved at least once), the risk estimates did not differ substantially between groups.As air pollution concentrations were generally highly correlated across time, we could only examine the potential relative importance of varying time periods of exposure, using models including both birth and six or ten year address pollution concentrations, for a few time-point combinations. The Pearson correlation coefficient was < 0.70 for two time-point combinations inthe total data (between NO2 assessed at the birth and six year addresses as well as between NO2 assessed at the birth and ten year addresses), and seven, three and three time-point combinations in the GINI/LISA South, GINI/LISA North and LISA East area-specific datasets. It was not possible to decipher a consistent trend as to which time period may be most important from the results of these models.3940Figure 4: Total and area-specific associations between NO2 (A), PM2.5 mass (B), PM2.5 absorbance (C) and ozone (D) estimated to the six-year address with doctor diagnosed asthma (red squares), doctor diagnosed allergic rhinitis (purple triangles), eye and nose symptoms (blue stars) and aeroallergen sensitization (black diamonds). Models are adjusted for sex, age, parental history of atopy, parental education, older siblings, maternal smoking during pregnancy, secondhand smoke exposure in the home, contact with furry pets, use of gas stove for cooking, home dampness or indoor mould, interventionparticipation, cohort and area (total models only). Odds ratios (OR) are calculated per interquartile (IQR) increase of each air pollutant. CI = confidence intervals; NA = not available41Figure 5: Total and area-specific associations between NO2 (A), PM2.5 mass (B) and PM2.5 absorbance (C) estimated to the ten year address with doctor diagnosed asthma (red squares), doctor diagnosed allergic rhinitis (purple triangles), eye and nose symptoms (blue stars) and aeroallergen sensitization (black diamonds). Models are adjusted for sex, age, parental history of atopy, parental education, older siblings, maternal smoking during pregnancy, secondhand smoke exposure in the home, contact with furry pets, use of gas stove for cooking, home dampness or indoor mould, interventionparticipation, cohort and area (total models only). Odds ratios (OR) are calculated per interquartile (IQR) increase of each air pollutant. CI = confidence intervals; NA = not available42Figure 6: Total and area-specific associations between NO2 (A), PM2.5 mass (B) and PM2.5 absorbance (C) averaged over the birth, six and ten year address estimates, and ozone (D) averaged over the birth and six year address estimates with doctor diagnosed asthma (red squares), doctor diagnosed allergic rhinitis (purple triangles), eye and nose symptoms (blue stars) and aeroallergen sensitization (black diamonds). Models are adjusted for sex, age, parental history of atopy, parental education, older siblings, maternal smoking during pregnancy, secondhand smoke exposure in the home, contact with furry pets, use of gas stove for cooking, home dampness or indoor mould, interventionparticipation, cohort and area (total models only). Odds ratios (OR) are calculated per interquartile (IQR) increase of each air pollutant. CI = confidence intervals; NA = not available2.4   DiscussionIn a longitudinal analysis of two German birth cohorts followed for ten years, we did not find consistent evidence that traffic-related air pollution exposure increases the risk of asthma, allergic rhinitis or aeroallergen sensitization in later childhood. The risk estimates for children living in two of the areas investigated (GINI/LISA North and LISA East) were null or elevated and those for the third (GINI/LISA South) tended to be below one. Given the heterogeneous area-specific effects, the risk estimates for the total population were inconclusive.The factors driving the differing associations observed across areas are unknown. The sources of pollutants which likely differ by area may be one explanation; air pollution in all three areas is predominantly attributable to traffic-sources but industry also contributes to air pollution levels in GINI/LISA North (Beelen et al. 2013; Eeftens et al. 2012a). It is also possible that residual confounding may be influencing the results. We attempted to adjust for individual-level socioeconomic status using parental education as a proxy in the final models, and marital status at the time of birth and household income per person (calculated according to (Sausenthaler et al.2011)) in sensitivity analyses, but these indicators may be imperfect markers of socioeconomic factors. Socioeconomic data at the neighborhood level is not available for a large proportion of participants, but has been shown not to strongly influence associations between air pollution and allergic outcomes in recent cross-sectional multicentre European studies (Mölter et al. 2014) and (Gruzieva et al. 2014)). Effect estimates also remained robust upon adjustment for surrounding green space and population density.The overall null findings reported here for the total population are in line with those of two large recent multicentre European studies on asthma (Mölter et al. 2014) and sensitization (Gruzieva etal. 2014) in which GINI/LISA South and GINI/LISA North were included. The current work differs from these previous studies in several respects. The current study uses a longitudinal analytical approach to optimize the use of the long-term prospectively collected health outcome data from three to ten years, whereas the previous two studies examined cross-sectional associations at two time points. Furthermore, the two multicentre studies did not include the LISA East study area nor were allergic rhinitis or nose and eye symptoms considered as 43outcomes or ozone as an air pollutant. With respect to previous studies with a long-term longitudinal design, our findings for the total study population are in contrast to the positive associations observed between traffic-related air pollution and asthma, as well as allergic rhinitis among non-movers, in a Dutch birth cohort followed for eight years (Gehring et al. 2010), but in line with the null associations for sensitization found in this Dutch cohort and in a similar Swedish birth cohort followed for eight years (Gruzieva et al. 2012). The area-specific findings reported for GINI/LISA North are generally in line with a previous study in this area which examined associations with outcomes up to six years of age (Krämer et al. 2009). For GINI/LISA South, the current findings are more challenging to reconcile with those reported for outcomes up to six years of age (Morgenstern et al. 2008). The sample sizes differ slightly between this previous and the current analyses (3,577 and 3,941 children included,respectively), the definitions were not identical for all outcomes and there were some differences in the methodologies of the exposure assessment. In the previous work, exposure estimates were derived using a LUR model developed as part of the TRAPCA project (Brauer et al. 2003). This initial model was subsequently applied to the GINI/LISA South metropolitan area (TRAPCA II), which includes the city of Munich and the surrounding districts (Morgenstern et al. 2007). Exposure estimates derived from this TRAPCA II model were significantly positively associated with several health outcomes at six years among GINIplus and LISAplus participants living in GINI/LISA South (Morgenstern et al. 2008). In contrast, the current analysis for this area, which yielded null or negative associations, is based on estimates derived from a different LUR model developed almost a decade later as part of the ESCAPE collaboration. The ESCAPE models explain more variation than the TRAPCA II models and the root-mean standard errors of the ESCAPE models are lower than for TRAPCA II. The distribution of PM2.5 mass and PM2.5 absorbance concentrations are similar between the two datasets, however the NO2 estimates are lower in the more recently derived ESCAPE dataset. The lower estimated NO2 concentrations may reflect true decreases as the air pollution measurements for the ESCAPE models were taken a decade after those for the TRAPCA models and actual air pollution levels have decreased in GINI/LISA South during this time.44To date, very few studies have reported on associations with ozone. In the current study, results were generally non-significant, with only two positive associations found for LISA East. A few factors may have hindered our ability to detect consistent associations. First, the resolution of thedatabase (1 x 1 km) used to estimate ozone concentrations to the home address was lower than for the other pollutants. Second, as the spatial distribution of ozone is more even than for other pollutants, with the exception of areas very close to traffic, detecting true associations may be more difficult. Third, ozone concentrations are likely higher in rural areas compared to areas with greater traffic densities. However, effect estimates remained similar when the analyses werestratified into the inner city of Munich and surrounding areas (for GINI/LISA South) or when those not living in the city area were excluded (for GINI/LISA North).Although the present work is among the very few studies which utilized such a rich and large longitudinal dataset, certain limitations should be acknowledged. Participation bias is always a concern for cohorts with a long follow-up. Children included in this study differed from those in the initial birth cohort with regard to several characteristics and this non-random retention of participants may have affected the effect estimates. Of the 9,086 children who were recruited in the GINIplus and LISAplus cohorts at birth, 5,078 participated in the ten year follow-up (55.9%).Outcome misclassification is also a concern when analyzing data collected by questionnaires, butobjective measures of aeroallergen sensitization were available for 55.3% (3,655/6,604) of the study population. No systematic differences between the results for the three parent-reported outcomes and the objective aeroallergen sensitization outcome are apparent. Furthermore, a positive response for asthma and allergic rhinitis was based on a parental-report of a doctor diagnosis and not only on a report of symptoms. The responses at the older ages are also likely more accurate as allergic disorders are easier to diagnose in later childhood. Finally, the data used to inform the LUR regression models for GINI/LISA South and North were collected approximately a decade after the commencement of the birth cohorts (approximately five years for LISA East) under the implicit assumption that the spatial variability in air pollution estimates would not have changed since the baseline periods of the cohorts. Three studies provide evidencesupporting this assumption for NO2 over a period of seven to 12 years (Cesaroni et al. 2012; Eeftens et al. 2011; Wang et al. 2013).45A greater importance of early-life exposures has been hypothesized as a possible explanation for why positive associations with traffic-related air pollution and sensitization have been found at four years of age but not at eight years in a Swedish birth cohort (Gruzieva et al. 2012), a findingthat was also observed in a Dutch birth cohort (Gehring et al. 2010). One previous study considered the relative importance of traffic-related air pollution exposure timing on the development of asthma and reported that NO and PM10 in utero exposures have an independent effect from post-birth exposures, although it was not possible for the authors to conclude which period may be most important (Clark et al. 2010). Similarly, the high correlation between exposures at birth and those at six and ten years in the current study rendered it challenging to disentangle the effects of these distinct exposure periods. Although we conducted several sensitivity analyses to reduce potential moving-related exposure misclassification and noted no changes in the results, it is possible that some exposure misclassification remains, especially at the older ages when children spend a larger proportion of their time at school. However, a Swedish and French study showed that exposures from traffic assessed at the home address are good approximations of those at schools, possibly because schools tend to be located in the close vicinity of homes (Gruzieva et al. 2012; Reungoat et al. 2005). Additionally, a study conducted in the United States found little differences between time-weighted averages of diesel exposures estimated at all addresses where a child spent more than eight hours per week and those estimated only at the home address (Ryan et al. 2008).In conclusion, we did not find consistent evidence that traffic-related air pollution increases the risk of childhood asthma or allergic diseases in later childhood using data from German birth cohort participants followed for ten years. Heterogeneous results were noted across the three geographical areas investigated.463   Childhood allergic rhinitis, traffic-related air pollution, and variability in the GSTP1, TNF, TLR2 and TLR4 genes 23.1   IntroductionRecent global estimates indicate that 8.5% of children aged six to seven suffer from allergic rhinitis and the prevalence is higher among 13-14 year-olds (14.6%) (Aït Khaled et al. 2009)‐ . The continued increase in prevalence in recent years in a majority of countries is especially concerning (Björkstén et al. 2008). Allergen exposure is strongly associated with allergic rhinitis onset. Early-life factors (young maternal age, multiple gestation and low birth weight), family history, ethnicity and environmental factors (secondhand smoke exposure, urban living, lifestyle,nutrition and air pollution) are also believed to be important (Bousquet et al. 2008; Kaiser 2004; Marshall 2004).Substantial experimental and toxicological evidence of the adverse effects of traffic-related air pollution on allergic disease exists and epidemiological evidence is building (Heinrich and Wichmann 2004), as summarized in a recent review (Tager et al. 2010). Given its likely association with asthma, traffic-related air pollution has been investigated as a potential cause of allergic rhinitis and several recent large studies support a positive association (Brunekreef and Sunyer 2003; Gehring et al. 2010). However, some studies have failed to find an association between the prevalence of allergic rhinitis symptoms and exposure to air pollution (Anderson et al. 2009; Nicolai et al. 2003; Wyler et al. 2000).Whether traffic-related air pollution increases the risk of allergic disease development and exacerbates symptoms in a genetically vulnerable subgroup remains largely unknown (Krzyzanowski et al.; Tager et al. 2010). Gene-environment interactions, which have been rarely considered in previous studies of allergic rhinitis, may provide some insight and have thus been 2 - A version of this manuscript has been published: Fuertes E, Brauer M, MacIntyre E, Bauer M, Bellander T, von Berg A, Berdel D, Brunekreef B, Chan-Yeung M, Cramer C, Gehring U, Herbarth O, Hoffmann B, Kerkhof M, Koletzko S, Kozyrskyj A, Kull I, Heinrich J, Melén E, Pershagen G, Postma D, Tiesler CM, Carlsten C. 2013. Childhood allergic rhinitis, traffic-related air pollution, and variability in the GSTP1, TNF, TLR2, and TLR4 genes: Results from the TAG Study. J Allergy Clin Immunol. 132(2):342–352.e2. dx.doi.org/10.1016/j.jaci.2013.03.00747recommended (Braback and Forsberg 2009). Many studies examining the interplay between genetic susceptibility and traffic-related air pollution on respiratory conditions have focused on genes in the oxidative stress and inflammation pathways (Holloway et al. 2012). Genetic variants of the GSTP1 gene have sparked considerable interest given the existence of common functional variants in the general population, GSTP1's role in cellular protection againstoxidative stress and the presence of the cytosolic glutathione S-transferase proteins in the human lung (Saxon and Diaz-Sanchez 2005). The evidence of a gene-environment interaction appears strongest for the Ile105Val (rs1695) SNP within the GSTP1 gene (Gerbase et al. 2011; Gilliland et al. 2004, 2006; Lee, Y-L et al. 2004; Melén et al. 2008; Romieu et al. 2006; Salam et al. 2007).Gene-environment interactions have also been observed for the G308A (rs1800629) SNP within the TNF gene with respect to passive smoke exposure and childhood asthma (Wu et al. 2007), and for ozone exposure with lung function and wheezing (Li et al. 2006; Yang et al. 2005). Furthermore, a gene-gene-environment interaction between the G-308A TNF variant, GSTP1 variants and NO2 exposure was documented for sensitization (Melén et al. 2008). Members of thetoll-like receptor family may also be important given their key roles in controlling innate and adaptive immune responses. Genetic polymorphisms in toll-like receptors have already been associated with allergic rhinitis (Gao et al. 2010) and may modify the link between particulate matter and childhood asthma (Kerkhof et al. 2010).Using a pooled analysis combining data from six birth cohorts with individual-level air pollution exposure assessment, we examined the associations between traffic-related air pollution with allergic rhinitis and aeroallergen sensitization in children, and the influence of ten SNPs related to inflammation and oxidative stress metabolism in the GSTP1, TNF, TLR2 and TLR4 genes. 3.2   Methods3.2.1   Study populationThe TAG study population is composed of 15,299 children recruited in six birth cohorts: BAMSE (Emenius et al. 2003; Wickman et al. 2002), CAPPS (Chan-Yeung et al. 2000), GINIplus (von Berg et al. 2003), LISAplus (Zutavern et al. 2006), PIAMA (Brunekreef et al. 482002) and SAGE (Kozyrskyj et al. 2009). Data on several health outcomes, environmental exposures and covariates were collected via either parent- or self-completed questionnaires at various time points according to each cohort's respective information collection strategy. Information across cohorts was harmonized into common variables, as previously described (MacIntyre et al. 2013). 3.2.2   Health outcomesThe assessment of allergic rhinitis differed slightly across cohorts; the two Canadian cohorts (CAPPS and SAGE) relied on a diagnosis during an assessment by a physician at a follow-up visit, the two German cohorts (GINIplus and LISAplus) relied on the report of a doctor's diagnosis during the last 12 months, the BAMSE cohort relied on the report of symptoms or a diagnosis of allergic rhinitis anytime in the last five years and the PIAMA cohort relied on the report of symptoms in the last 12 months (Table 2). The eight year follow-up was selected as the time point of interest as information on allergic rhinitis was available for all but one cohort at thisage. For CAPPS, the assessment was made at age seven.Sensitization was assessed by skin prick testing at age seven for CAPPS and at age eight for SAGE, with a positive reaction defined as having a wheal diameter of greater or equal than 3 mm. For GINIplus, LISAplus, BAMSE and PIAMA, sensitization was assessed by measuring specific IgE, with a positive reaction defined as any value > 0.35 kU/L (at age six years for the former two cohorts and eight years for the latter two). Although some aeroallergens can be present in both indoor and outdoor environments, in this analysis, birch, Dactylis, mugwort, ragweed, rye, timothy grass, trees and weeds were considered as outdoor aeroallergens and Alternaria, cats, cockroaches, dogs, feathers, house dust mites and moulds (Cladosporium herbarum) were considered as indoor aeroallergens. All available aeroallergens were included in the overall sensitization analysis. Not all cohorts had information on all aeroallergens (Table 2). 3.2.3   Air pollution estimatesUnique NO2 concentration estimates were available for 55.4% (8,470/15,299; 6/6 cohorts) of participants' home addresses at the time of birth. For all cohorts except BAMSE, the NO2 49estimates were derived using LUR modelling. The LUR models developed for the European cohorts (GINIplus and LISAplus (Munich city only), and PIAMA) were created as part of the TRAPCA collaboration (Brauer et al. 2003; Hoek et al. 2002). Using a similar methodology, LUR models were developed for the two Canadian cohorts (Allen et al. 2011; Henderson et al. 2007) and the cities of Wesel (Hochadel et al. 2006) and Leipzig within the GINIplus and LISAplus cohorts. NO2 estimates for the BAMSE cohort were obtained from a dispersion model,as previously described (Nordling et al. 2008). PM2.5 mass and PM2.5 absorbance concentrations, calculated using the same methodology as for NO2, were available for 38.5% (5,893/ 15,299; 3/6 cohorts) and 56.3% (8,615/ 15,299; 4/6 cohorts) of participants, respectively. Ozone estimates were available for 76.8% (11,757/15,299; 4/6 cohorts) of participants. These were calculated as part of the APMoSPHERE project for PIAMA, GINIplus and LISAplus (Beelen et al. 2009), andusing ambient monitoring network data for the CAPPS cohort (Marshall et al. 2008). Unlike for the other pollutants, the ozone estimates were not derived using any specific traffic components, and thus represent air pollution in general rather than traffic-related air pollution. The estimated exposures for NO2, PM2.5 mass and PM2.5 absorbance were positively correlated (r = 0.35, 0.81, 0.49 for NO2 and PM2.5 mass, NO2 and PM2.5 absorbance and PM2.5 mass and PM2.5 absorbance, respectively). Ozone was negatively correlated (r = -0.25, -0.18 and -0.15 for NO2, PM2.5 mass and PM2.5 absorbance, respectively). 3.2.4   Genetic dataIn total, 47.3% (7,229/15,299) of TAG participants were genotyped for at least one SNP of interest. For CAPPS and SAGE, genotyping was done using the Illumina BeadArray system (Illumina, San Diego, California, USA). Genotyping in PIAMA was performed by the Competitive Allele-Specific PCR using KASParTM genotyping chemistry (K-Biosciences, Herts, UK). For the two German cohorts (GINIplus and LISAplus) and the Swedish BAMSE cohort, SNPs were genotyped using the iPLEX (Sequenom, San Diego, CA, USA) method by means of the matrix assisted laser desorption ionization time of flight mass spectrometry method (MALDI-TOF MS, Mass Array; Sequenom, San Diego, CA, USA), with the exception of rs1695which was detected using the restriction fragment length polymorphism approach (GINIplus and LISAplus only) (Slama et al. 2010). All SNPs had a genotyping success rate greater than 93%. 503.2.5   Statistical analysisAssociations between a pollutant or a SNP and each outcome were assessed using logistic regression. The effect of each SNP on allergic rhinitis was examined in a dominant genetic model (carriers of at least one minor allele versus homozygous major allele carriers). Elevated risks of disease were analyzed per 10 μg/m3 increase for NO2 and ozone, per 5 μg/m3 increase forPM2.5 mass and per 10-5/m increase for PM2.5 absorbance (roughly the interquartile range of each pollutant in the pooled data). All models were adjusted for covariates selected a priori (city/centre, cohort, sex, birth weight, parental history of atopic disease, maternal smoking during pregnancy, secondhand smoke exposure at the time of follow-up, intervention status (when applicable) and maternal age at birth). The latter variable was used as a surrogate of socioeconomic status as women from a higher socioeconomic background tend to have children at older ages (Hemminki and Gissler 1996) and a positive association between maternal age at birth and socioeconomic status has been observed in previous studies of similar populations (Almqvist et al. 2005; Groen et al. 2012; Ruijsbroek et al. 2011). In order to assess whether the relationship between a pollutant and an outcome differed by genotype, models were run separately for homozygous major allele carriers and heterozygous/homozygous minor allele carriers. Gene-environment interactions were examined using interaction terms in the models.All results are presented by cohort, except for the GINIplus and LISAplus studies which are presented separately for Munich (previously referred to as GINI/LISA South in Chapter 2) and Wesel/Leipzig (previously referred to as GINI/LISA North and LISA East, respectively, in Chapter 2) as the measurement campaigns for the LUR modelling for these study areas were conducted at different time points. Pooled results, which take advantage of the full available statistical power, are also presented. To assess the influence of each cohort on our pooled findings, we examined the results after a step-wise exclusion of each cohort. All results are presented as ORs [95% CI]. All statistical analyses were conducted using R, version 2.13.1 (R Core Team 2012). 513.3   ResultsTable 2 summarizes the study characteristics of the six participating cohorts, one of which (CAPPS) only recruited children with at least one first-degree relative with asthma or two first-degree relatives with other IgE-mediated allergic diseases and two of which are nested case-control studies (for asthma (SAGE) and wheeze (BAMSE)). After excluding children with no information on any of the air pollutants (N = 2,100) or health outcomes (N =4,416), 10,023 children remained in the study and are described in Table 8. However, not all children were included in all analyses due to missing covariate information. Overall, 1,298 (13.7%) children had allergic rhinitis at the time of follow-up. The two Canadian cohorts (CAPPS and SAGE) thatutilized an active physician assessment at the follow-up visit had the highest rates of allergic rhinitis. The cities in the German cohorts (Munich and Wesel/Leipzig) that relied on the report ofa doctor diagnosis of allergic rhinitis in the last 12 months had the lowest prevalences.52Table 8:  Characteristics of the study populationPooled BAMSE CAPPS Munich Wesel/Leipzig SAGE PIAMAN % N % N % N % N % N % N %Males 10023 51.6 919 52.8 372 53.8 2784 51.2 2355 51.3 235 56.2 3358 51.4Parental history of allergies 10017 50.2 919 58.1 372 92.5 2779 53.9 2354 34.4 235 67.7 3358 50.1Maternal smoking during pregnancy9328 14.7 919 12.9 369 7.9 2437 13.4 2060 16.4 227 11.9 3316 16.0Secondhand smoke exposure at home at age 7/8 yrs9423 18.8 911 17.8 372 18.0 2566 14.8 2107 26.8 229 20.5 3238 16.9Older siblings 9969 47.9 919 49.5 372 54.8 2781 42.7 2348 50.4 198 68.2 3351 48.0Intervention participation 10020 23.1 919 0 372 53.5 2784 31.4 2355 27.3 235 0 3358 17.8Birth weight (g) a 3489.3 (513.5) 3500.2 (577.3) 3495.5 (642.4) 3415.2 (437.5) 3527.0 (478.2) 3378.6 (636.3) 3507.2 (546.1)Maternal age at birth a 31 (4.1) 30.7 (4.5) 31.8 (5.0) 32.2 (4.1) 30.4 (3.9) 28.9 (5.3) 30.3 (3.9)Allergic rhinitis at age 7/8 yrs 9451 13.7 913 17.9 372 30.1 2606 7.7 2130 6.3 190 40.0 3240 18.9Sensitization to any aeroallergen 6212 31.7 766 24.8 359 45.1 1668 29.2 1319 26.2 234 37.2 1866 37.4Sensitization to indoor aeroallergen 6058 24.4 766 20.6 359 36.8 1668 17.8 1319 18.3 234 27.8 1712 34.1Sensitization to outdoor aeroallergen6206 19.3 762 18.1 358 21.8 1668 21.6 1319 19.0 234 20.9 1865 17.4Allergic rhinitis and sensitization 5640 11.3 760 14.5 359 20.6 1490 8.9 1094 7.2 189 16.9 1748 12.0Allergic rhinitis with available sensitization data925 na 143 na 106 na 148 na 86 na 75 na 367 naa mean (standard deviation) % = percentage of children with this covariate/outcome among those with available data; N = number of children with available data on this covariate/outcome (except for birth weight and maternal age [as noted] and allergic rhinitis with available sensitization data [where N represents children with a diagnosis of allergicrhinitis who also have available sensitization data]); na = not applicable53In total, 31.7% (1,968/6,212) of children were sensitized to at least one aeroallergen. Among subjects with data on allergic rhinitis and sensitization, 11.3% (637/5,640) had both conditions. Among children with a doctor diagnosis of allergic rhinitis who also had available sensitization data, 68.9% (637/925) were sensitized (range per cohort is 42.7% in SAGE and 91.9% in Wesel/Leipzig). Given that approximately 30% of our subjects with allergic rhinitis were not sensitized to any tested aeroallergen, it is likely that our disease definition includes both children with allergic and non-allergic rhinitis. The characteristics of each SNP are presented in Table 9 by cohort and in the pooled data. All SNPs were in Hardy-Weinberg equilibrium with the exception of rs4891 in the GSTP1 gene, which was excluded from the analysis. In this study, we focused on SNPs related to oxidative stress and inflammation that were available in at least three cohorts.54Table 9:  Single nucleotide polymorphism characteristics and genotype frequencies in the study populationGene SNP Location Alleles aPooled BAMSE CAPPS MunichWesel/Leipzig SAGE PIAMAN EAF N EAF N EAF N EAF N EAF N EAF N EAFGSTP1 rs1138272 Exon (Ala114Val) C/T 5470 9.1 861 9.0 345 7.0 903 9.4 1252 9.3 183 8.2 1926 9.3rs1695 Exon (Ile105Val) A/G 5996 34.7 897 32.9 345 31.2 1470 35.0 1194 34.7 181 31.5 1909 36.1rs4891 b Exon (synonymous) T/C 4246 39.1 873 30.5 NA NA 740 49.4 684 45.3 NA NA 1949 36.9TNF rs1800629 Promoter G/A 5296 16.9 854 15.5 346 14.2 823 14.7 1182 17.7 185 13.0 1906 18.9TLR2 rs4696480 Intron T/A 1685 48.8 NA NA NA NA 391 51.2 382 49.5 NA NA 912 46.9TLR4 rs10759930 Promoter C/T 2352 38.7 NA NA 347 40.2 823 40.3 1182 37.1 NA NA NA NArs10759931 Promoter G/A 2898 39.1 NA NA NA NA 824 40.3 1183 37.1 NA NA 891 40.7rs10759932 Promoter T/C 1679 12.6 NA NA NA NA 387 13.7 384 13.3 NA NA 908 11.9rs1927911 Intron C/T 3446 24.9 NA NA 347 26.5 824 25.5 1181 25.3 185 20.5 909 24.0rs2737190 5' untranslated region A/G 2926 32.1 NA NA NA NA 823 31.2 1184 32.6 NA NA 919 32.2rs2770150 Promoter T/C 3432 28.5 NA NA 347 25.6 822 28.5 1181 29.1 186 26.9 896 29.1a Reference allele / effect allele b SNP was not in Hardy-Weinberg equilibrium and subsequently eliminated from the analysis.EAF = effect allele frequency; N = number of children with data on genotype, at least one air pollutant and at least one health outcome; NA = not available; SNP = single nucleotide polymorphism553.3.1   Main environmental associationsThere was substantial overlap in the distribution of NO2 and PM2.5 absorbance between cohorts, but less so for ozone and PM2.5 mass (Figure 7). In the pooled analysis, the point estimates for theassociation between NO2, PM2.5 mass, and PM2.5 absorbance and allergic rhinitis at age seven/eight years were elevated, but only that for PM2.5 mass reached statistical significance aftercovariate adjustment (1.37 [1.01, 1.86] per 5 μg/m3 PM2.5 mass, 1.10 [0.95, 1.26] per 10 μg/m3 NO2, and 1.16 [0.96, 1.41] per 10-5/m PM2.5 absorbance; depicted in Figure 8 and estimates provided in Table 10). There were no significant associations between any of the pollutants and aeroallergen sensitization in the pooled analysis. Furthermore, all associations between air pollutants and “atopic allergic rhinitis” (allergic rhinitis and sensitization to any aeroallergen) were null.The elevated risk estimates found between allergic rhinitis and the air pollutants were heavily influenced by increased risks seen in the PIAMA cohort, as previously published (Gehring et al. 2010) (for example, 1.37 [1.01, 1.86] when all cohorts are included and 1.02 [0.62, 1.67] when PIAMA is excluded for the association between PM2.5 mass and allergic rhinitis). This observation is further supported by the relatively inconsistent trend in the results seen across cohorts. 5657Figure 7: Distribution of air pollutants pooled and by cohort. The interquartile range is indicated by each box height and median level by each dark line. The number ofchildren with health data that also had available air pollution data are given along the top of each graph. NA = not available58Figure 8: Associations between allergic rhinitis (stars) and aeroallergen sensitization (dark squares)with NO2, ozone, PM2.5 mass and PM2.5 absorbance. Models were adjusted for city/centre, cohort, sex, birth weight, parental history of atopic disease, maternal smoking during pregnancy, secondhand smoke exposure at the time of follow-up, maternal age at birth and intervention status (when applicable). CI = confidence intervals; NA = not available; OR = odds ratioTable 10: Pooled and cohort specific associations between allergic rhinitis and aeroallergen sensitization with air pollutantsPooled BAMSE CAPPS Munich Wesel/Leipzig SAGE PIAMAN OR[95%CI] NOR[95%CI] NOR[95%CI] NOR[95%CI] NOR[95%CI] NOR [95%CI] NOR [95%CI]Allergic rhinitisNO2 7364 1.10[0.95, 1.26] 8970.79[0.51, 1.23] 3681.04[0.67, 1.63] 10280.94[0.64, 1.40] 17401.26[0.67, 2.37] 1710.34[0.10, 1.12] 31601.24[1.04, 1.49]Ozone 7293 0.91[0.77, 1.08] NA NA 1860.96[0.32, 2.88] 21630.86[0.59, 1.26] 17930.94[0.57, 1.56] NA NA 31510.93[0.75, 1.13]PM2.5 mass 4373 1.37[1.01, 1.86] NA NA 1851.08[0.59, 1.96] 10280.89[0.34, 2.31] NA NA NA NA 31601.66[1.12, 2.46]PM2.5 absorbance 57121.16[0.96, 1.41] NA NA 1851.08[0.84, 1.40] 10280.83[0.38, 1.79] 13391.46[0.40, 5.27] NA NA 31601.51[1.07, 2.13]Aeroallergen sensitizationNO2 4603 0.94[0.82, 1.08] 7510.87[0.58, 1.32] 3550.94[0.61, 1.44] 6170.77[0.57, 1.06] 9401.00[0.62, 1.63] 2150.47[0.19, 1.17] 17251.10[0.89, 1.37]Ozone 4132 0.95[0.81, 1.12] NA NA 1771.99[0.66, 6.06] 12761.10[0.82, 1.46] 9610.97[0.66, 1.41] NA NA 17180.88[0.70, 1.11]PM2.5 Mass 2518 0.92[0.66, 1.27] NA NA 1760.78[0.43, 1.42] 6170.52[0.24, 1.11] NA NA NA NA 17251.22[0.77, 1.94]PM2.5 absorbance 32170.93[0.76, 1.13] NA NA 1760.94[0.72, 1.21] 6170.59[0.32, 1.08] 6990.76 [0.29, 1.99] NA NA 17251.17[0.78, 1.76]Models were adjusted for city/centre, cohort, sex, birth weight, parental history of atopic disease, maternal smoking during pregnancy, secondhand smoke exposure at the time of follow-up, maternal age at birth and intervention status (when applicable).Bold = p-value < 0.05; CI = confidence intervals; N = number of children included in the model; NA = not available; OR = odds ratio593.3.2   Main genetic associationsCarriers of at least one minor rs1800629 or rs1927911 allele were at increased risk of developingallergic rhinitis in the pooled analysis (1.19 [1.00, 1.41] and 1.24 [1.01, 1.53], respectively; depicted in Figure 9 and estimates provided in Table 11). When examining the cohort-specific analyses, the estimates for rs1800629 and rs1927911 were elevated in four of six and four of fivecohorts, respectively. Furthermore, during the step-wise exclusion of each cohort, the pooled point estimates remained similar, although loss of statistical significance was occasionally observed (for example: 1.19 [1.00, 1.41] and 1.21 [1.01, 1.46] for rs1800629 and 1.24 [1.01, 1.53] and 1.16 [0.94, 1.45] for rs1927911, with and without SAGE, respectively). No significant associations were documented between any of the SNPs investigated and aeroallergen sensitization in the single cohort and pooled analyses (depicted in Figure 9 and estimates provided in Table 12). These results remained unchanged when the analysis was stratified by indoor and outdoor aeroallergens. The ORs for atopic allergic rhinitis with rs1800629 and rs1927911 were elevated but not significant. This loss of significance may be due to a drop in sample size as sensitization data were only available for a subset of the population ormay reflect a true reduced effect on this outcome (allergic rhinitis: 1.19 [1.00, 1.41] and 1.24 [1.01, 1.53] vs. atopic allergic rhinitis: 1.13 [0.91, 1.40] and 1.13 [0.88, 1.46] for rs1800629 and rs1927911, respectively). 6061Figure 9: Associations between allergic rhinitis (stars) and aeroallergen sensitization (dark squares) with ten single nucleotide polymorphisms. Models were adjusted for city/centre, cohort, sex, birth weight, parental history of atopic disease, maternal smokingduring pregnancy, secondhand smoke exposure at the time of follow-up, maternal age at birth and intervention status (when applicable). The upper confidence limit in SAGE for rs1927911 and allergic rhinitis is 7.44. Po = Pooled; B = BAMSE; C = CAPPS; M = Munich;W/L = Wesel and Leipzig; S = SAGE; P = PIAMA; CI = confidence intervals; NA = not available; OR = odds ratioTable 11: Pooled and cohort specific associations between allergic rhinitis and ten single nucleotide polymorphismsPooled BAMSE CAPPS Munich Wesel/Leipzig SAGE PIAMASNP N OR [95%CI] NOR [95%CI] NOR [95%CI] NOR [95%CI] NOR [95%CI] NOR [95%CI] NOR [95%CI]rs1138272 4739 1.09[0.89, 1.35] 8410.89[0.54, 1.46] 3441.94[1.01, 3.72] 7161.10[0.55, 2.21] 9130.76[0.39, 1.49] 1340.83[0.30, 2.31] 17911.20[0.88, 1.63]rs1695 5077 1.02[0.87, 1.20] 8760.81[0.57, 1.16] 3441.51[0.93, 2.46] 11071.28[0.85, 1.93] 8430.85[0.51, 1.42] 1310.90[0.40, 2.05] 17761.00[0.79, 1.28]rs1800629 4601 1.19[1.00, 1.41] 8331.34[0.90, 1.98] 3450.99[0.58, 1.70] 6551.11[0.58, 2.12] 8600.94[0.55, 1.61] 1351.73[0.67, 4.47] 17731.21[0.95, 1.55]rs4696480 1505 0.91[0.66, 1.25] NA NA NA NA 3491.16[0.41, 3.28] 3130.83[0.30, 2.28] NA NA 8430.89[0.62, 1.27]rs10759930 1862 0.92[0.68, 1.25] NA NA 3460.82[0.49, 1.35] 6551.44[0.75, 2.76] 8610.79[0.48, 1.32] NA NA NA NArs10759931 2340 1.15[0.88, 1.49] NA NA NA NA 6551.44[0.75, 2.76] 8610.80[0.48, 1.33] NA NA 8241.29[0.91, 1.83]rs10759932 1501 1.13[0.81, 1.57] NA NA NA NA 3461.56[0.63, 3.89] 3140.91[0.34, 2.46] NA NA 8411.12[0.76, 1.65]rs1927911 2834 1.24[1.01, 1.53] NA NA 3451.07[0.66, 1.73] 6550.94[0.51, 1.71] 8581.63[0.99, 2.69] 1353.14[1.33, 7.44] 8411.13[0.81, 1.56]rs2737190 2366 1.00[0.78, 1.28] NA NA NA NA 6540.82[0.45, 1.48] 8611.30[0.78, 2.16] NA NA 8510.96[0.69, 1.32]rs2770150 2822 1.02[0.82, 1.25] NA NA 3450.89[0.55, 1.45] 6541.08[0.60, 1.96] 8590.99[0.60, 1.63] 1361.40[0.64, 3.06] 8281.05[0.76, 1.46]Models were adjusted for city/centre, cohort, sex, birth weight, parental history of atopic disease, maternal smoking during pregnancy, secondhand smoke exposure at the time of follow-up, maternal age at birth and intervention status (when applicable).Bold = p-value < 0.05; CI = confidence intervals; N = number of children included in the model; NA = not available; OR = odds ratio; SNP = single nucleotide polymorphism62Table 12: Pooled and cohort specific associations between aeroallergen sensitization and ten single nucleotide polymorphismsPooled BAMSE CAPPS Munich Wesel/Leipzig SAGE PIAMASNP N OR [95%CI] NOR [95%CI] NOR [95%CI] NOR [95%CI] NOR [95%CI] NOR [95%CI] NOR [95%CI]rs1138272 4091 0.95[0.79, 1.13] 7060.87[0.53, 1.40] 3321.30[0.68, 2.46] 5461.33[0.81, 2.16] 8420.68[0.45, 1.03] 1690.63[0.25, 1.56] 14961.01[0.76, 1.33]rs1695 4716 1.00[0.88, 1.13] 7351.02[0.72, 1.44] 3330.70[0.44, 1.10] 11291.11[0.86, 1.45] 8660.82[0.60, 1.11] 1671.93[0.99, 3.76] 14861.02[0.82, 1.27]rs1800629 3986 1.04[0.90, 1.20] 6990.88[0.59, 1.30] 3331.01[0.61, 1.66] 5071.33[0.86, 2.07] 7951.18[0.85, 1.65] 1711.09[0.51, 2.32] 14810.95[0.75, 1.19]rs4696480 1145 1.04[0.79, 1.39] NA NA NA NA 1930.95[0.44, 2.07] 2310.72[0.36, 1.45] NA NA 7211.14[0.81, 1.60]rs10759930 1637 0.98[0.79, 1.23] NA NA 3340.98[0.61, 1.57] 5070.93[0.61, 1.43] 7961.05[0.77, 1.45] NA NA NA NArs10759931 2010 1.00[0.82, 1.22] NA NA NA NA 5070.93[0.61, 1.43] 7961.05[0.76, 1.45] NA NA 7071.01[0.73, 1.37]rs10759932 1141 1.10[0.82, 1.47] NA NA NA NA 1901.46[0.69, 3.07] 2321.06[0.55, 2.05] NA NA 7191.07[0.74, 1.54]rs1927911 2525 0.89[0.75, 1.06] NA NA 3330.84[0.54, 1.32] 5070.92[0.61, 1.38] 7940.91[0.67, 1.25] 1711.00[0.52, 1.92] 7200.84[0.61, 1.14]rs2737190 2029 0.85[0.70, 1.03] NA NA NA NA 5060.86[0.57, 1.29] 7960.98[0.71, 1.33] NA NA 7270.76[0.56, 1.02]rs2770150 2515 1.05[0.89, 1.25] NA NA 3330.78[0.49, 1.23] 5071.23[0.82, 1.85] 7941.07[0.78, 1.46] 1720.49[0.26, 0.95] 7091.30[0.96, 1.77]Models were adjusted for city/centre, cohort, sex, birth weight, parental history of atopic disease, maternal smoking during pregnancy, secondhand smoke exposure at the time of follow-up, maternal age at birth and intervention status (when applicable).Bold = p-value < 0.05; CI = confidence intervals; N = number of children included in the model; NA= not available; OR = odds ratio; SNP = single nucleotide polymorphism633.3.3   Genotype stratification and interaction associationsStratified analyses did not reveal an increased risk of allergic rhinitis among heterozygous/homozygous minor allele carriers exposed to traffic-related air pollution (Table 13). Only the association between allergic rhinitis and PM2.5 mass among rs2737190 (TLR4) homozygous major allele carries was significant (2.77 [1.07, 7.15]), but this association was also driven by the PIAMA cohort. All interaction terms were non-significant (p-values ranged from 0.06 (rs10759931*NO2) to 0.99 (rs1800629*NO2 )). Table 13: Associations between air pollutants and allergic rhinitis among homozygous major and heterozygous/homozygous minor allele carriersAir pollutantHomozygous major Heterozygous/homozygous minorSNP N OR [95%CI] N OR [95%CI]NO2 rs1138272 3589 0.97 [0.78, 1.19] 756 1.20 [0.77, 1.87]rs1695 1902 1.08 [0.82, 1.43] 2581 0.96 [0.75, 1.21]rs1800629 2896 0.97 [0.77, 1.23] 1351 1.10 [0.78, 1.54]rs4696480 359 1.01 [0.52, 1.95] 960 1.24 [0.88, 1.73]rs10759930 561 0.86 [0.48, 1.55] 948 1.37 [0.90, 2.10]rs10759931 753 1.27 [0.82, 1.95] 1235 1.07 [0.74, 1.54]rs10759932 1004 1.22 [0.88, 1.71] 312 0.98 [0.49, 1.94]rs1927911 1400 1.35 [0.99, 1.85] 1081 0.77 [0.54, 1.11]rs2737190 936 1.30 [0.86, 1.95] 1078 1.05 [0.73, 1.53]rs2770150 1262 0.87 [0.62, 1.21] 1208 1.26 [0.90, 1.77]Ozone rs1138272 2945 0.92 [0.72, 1.18] 639 0.94 [0.51, 1.73]rs1695 1623 0.71 [0.50, 1.02] 2266 0.90 [0.68, 1.20]rs1800629 2335 0.81 [0.60, 1.09] 1116 0.99 [0.67, 1.46]rs4696480 391 1.54 [0.80, 2.94] 1110 0.94 [0.63, 1.41]rs10759930 630 0.95 [0.42, 2.15] 1060 0.92 [0.50, 1.66]rs10759931 873 1.21 [0.71, 2.04] 1463 1.00 [0.68, 1.45]rs10759932 1141 1.03 [0.69, 1.54] 357 1.15 [0.56, 2.34]rs1927911 1422 0.99 [0.67, 1.47] 1102 1.10 [0.70, 1.72]rs2737190 1100 1.03 [0.67, 1.59] 1262 1.14 [0.75, 1.74]rs2770150 1284 0.99 [0.65, 1.50] 1227 1.19 [0.79, 1.81]64Air pollutantHomozygous major Heterozygous/homozygous minorSNP N OR (95%CI) N OR (95%CI)PM2.5 mass rs1138272 1903 1.20 [0.76, 1.89] 402 1.29 [0.50, 3.35]rs1695 1010 1.72 [0.95, 3.13] 1467 0.97 [0.58, 1.65]rs1800629 1514 1.40 [0.87, 2.27] 749 0.91 [0.40, 2.08]rs4696480 288 1.52 [0.30, 7.58] 721 1.70 [0.81, 3.58]rs10759930 175 0.60 [0.21, 1.69] 317 1.41 [0.68, 2.95]rs10759931 419 2.07 [0.74, 5.76] 724 1.25 [0.53, 2.97]rs10759932 774 1.74 [0.82, 3.69] 231 1.28 [0.26, 6.20]rs1927911 752 1.42 [0.77, 2.65] 581 0.94 [0.48, 1.82]rs2737190 550 2.77 [1.07, 7.15] 619 1.07 [0.43, 2.66]rs2770150 676 1.05 [0.57, 1.93] 644 1.37 [0.69, 2.71]PM2.5 absorbancers1138272 2368 1.02 [0.80, 1.31] 505 1.24 [0.70, 2.17]rs1695 1271 1.09 [0.78, 1.53] 1828 1.03 [0.77, 1.37]rs1800629 1864 0.99 [0.76, 1.30] 932 1.31 [0.85, 2.00]rs4696480 288 1.59 [0.41, 6.23] 721 1.56 [0.81, 3.00]rs10759930 390 0.80 [0.52, 1.22] 636 1.12 [0.81, 1.55]rs10759931 635 1.62 [0.71, 3.72] 1042 1.14 [0.56, 2.32]rs10759932 774 1.66 [0.86, 3.18] 231 1.41 [0.35, 5.67]rs1927911 1049 1.14 [0.82, 1.60] 818 0.98 [0.71, 1.36]rs2737190 801 1.79 [0.83, 3.86] 902 1.11 [0.53, 2.33]rs2770150 948 1.09 [0.82, 1.45] 904 0.93 [0.62, 1.40]Models were adjusted for city/centre, cohort, sex, birth weight, parental history of atopic disease, maternal smoking during pregnancy, secondhand smoke exposure at the time of follow-up, maternal age at birth and intervention status(when applicable).Bold = p-value < 0.05; CI = confidence intervals; N = number of children included in the model; OR = odds ratio; SNP = single nucleotide polymorphismsAll stratified analyses for aeroallergen sensitization were null. Only one gene-environment interaction term for aeroallergen sensitization was significant in pooled analyses (p-values ranged from 0.03 (rs1695*ozone) to 0.99 (rs2737190*PM2.5 absorbance)). Stratification into indoor and outdoor aeroallergen categories yielded a significantly elevated risk between indoor aeroallergen sensitization and NO2 among minor rs1800629 allele carriers (1.52 [1.09, 2.12] per 10 μg/m3 increase in NO2) but no interaction was found (p-value=0.27); the result for outdoor aeroallergen sensitization for this SNP was null (1.01 [0.70, 1.45] per 10 μg/m3 increase in NO2). 653.4   DiscussionThe results of this large collaborative project do not suggest that traffic-related air pollution increases the risk of allergic rhinitis in general. Although the estimate for PM2.5 mass was significantly elevated, and those for both NO2 and PM2.5 absorbance were also elevated, these results were mainly driven by only one cohort (PIAMA) and were not replicated in the other five. No associations were observed for ozone, however, the spatial scale of the APMoSPHERE model from which the ozone estimates were estimated is relatively broad (1 × 1 km) and may incorporate more exposure misclassification than estimates for the other pollutants. In our study, we found suggestive evidence that children with at least one adenine at the 308 position in the TNF gene (rs1800629) may be at an elevated risk of allergic rhinitis at the age of seven/eight years. To our knowledge, only three other studies have investigated this association. Zhu et al. found no association between TNF and the development of atopy, asthma and rhinitis in a high risk population of 373 infants (Zhu et al. 2000). However, Gentile et al. found that amongst 124 infants, minor allele carriers of the TNF gene variant were at a higher risk of havinga parental history of allergic disease (Gentile et al. 2004). Moreover, a recent study found a strong association between the rs1800629 SNP and allergic rhinitis exacerbation in a population of 269 adult Pakistani patients (Minhas et al. 2010). Our study is the first to document this association in school-age children and our results are based on a substantially larger sample size than those used in previous studies. The association between the rs1800629 SNP and allergic rhinitis is biologically plausible. The rs1800629 SNP is located within the promoter region of the TNF gene and has been reported to affect the expression of the pleiotropic proinflammatory cytokine TNF-alpha by some studies (Louis et al. 1998; Wilson et al. 1997), but not others (Brinkman, et al. 1994). It is also difficult to distinguish the effect of TNF polymorphisms from those of other alleles at the human leukocyte antigen region because of linkage disequilibrium (LD) (Wilson et al. 1993). Nevertheless, elevated levels of TNF-α have been observed in human allergic rhinitis sufferers (Nonaka et al. 1996; Riccio et al. 2002) and studies in mice suggest that the lack of this cytokine inhibits allergic rhinitis development (Iwasaki et al. 2003). Functional and biological studies 66which elucidate the role of TNF-alpha in allergic rhinitis development are required and future epidemiological studies should aim to replicate our result.Our study results also suggest that carriers of the C allele in the rs1927911 SNP in the intron region of the TLR4 gene may be at an elevated risk of allergic rhinitis. No other studies have documented this association. However, eight other SNPs in the toll-like receptors have been linked to the prevalence of allergic rhinitis, including one in the TLR4 gene (rs4986790) (Senthilselvan et al. 2008). Unfortunately, we did not have data for this SNP in our study. Interestingly, we did not see an association between allergic rhinitis and the rs4696480 SNP in TLR2, as has been previously documented in European farmers (Eder et al. 2004). Both genetic findings of the current study were robust to step-wise exclusion of each cohort. We found no evidence to support the existence of gene-environment interactions between NO2, PM2.5 mass, PM2.5 absorbance or ozone and ten SNPs in the GSTP1, TNF, TLR2 and TLR4 genes. We did find a significant risk of sensitization to indoor aeroallergens among minor rs1800629 allele carriers exposed to NO2, however this result was not also observed for outdoor aeroallergen sensitization. The interaction term between ozone and the rs1695 SNP was also significant for overall aeroallergen sensitization. However, neither the main environmental nor genetic effect estimates were significant for this outcome and SNP.To date, we are the first to assess the existence of gene-air pollution interactions with respect to allergic rhinitis. However, gene-environment interactions have been reported for other environmental exposures (Bieli et al. 2007; Kim et al. 2012). With respect to sensitization, Melénet al. (2008) reported a significant interaction between NOx and the rs1695 SNP (GSTP1) using the BAMSE cohort. Although we included this cohort in our analysis and examined it individually, we were unable to replicate this finding. However, in the present analysis, sensitization was assessed at eight years of age and included only aeroallergens, whereas Melén et al. (2008) examined sensitization to food or aeroallergens at four years of age. A recent publication by the BAMSE cohort research group suggests that the adverse effects of air pollution on sensitization may be restricted to gestation and early childhood time points during 67which the immune system is rapidly developing (allergic rhinitis was not considered) (Gruzieva et al. 2012). This hypothesis, namely that the adverse effects of traffic-related air pollution may be limited to early life, may explain why a gene-environment interaction was observed when the BAMSE population was four years-old but not in the current study in which they are eight. However, we cannot rule out that interaction effects may exist between GSTP1, air pollutants andallergy-related outcomes. An even larger sample size, including a complete cover of variants in GSTP1, will likely give further insights into this complex interplay.Gilliland et al. also reported positive findings for gene-environment effects with respect to sensitization: nasal IgE levels were increased among genetically susceptible allergic individuals after diesel exhaust particle (Gilliland et al. 2004) and second-hand smoke (Gilliland et al. 2006) exposure. The discrepancy between these positive findings and our null results may reflect differences in study design, patient populations and phenotypes studied. Most notably, the studiesconducted by Gilliland et al. involved adult patients and a randomized, placebo-controlled crossover study design. Furthermore, epigenetic effects were not considered in our nor the other studies, but are likely to have important consequences for disease development, as described in a recent update on the current literature on air pollution, genetics (and epigenetics) and allergy (Carlsten and Melén 2012).One of the major issues of studies examining gene-environment interactions, in addition to manyother challenges, is that null findings may simply be due to lack of statistical power (Kauffmann and Demenais 2012). The TAG collaboration answers the numerous calls for the need to increasesample sizes by combining cohorts so that we are better poised to fully investigate the relationships and interactions between the genome, the environment and disease development. Nevertheless, we cannot exclude the possibility that our study may still be under-powered to detect real gene-environment interactions. For this reason, we also conducted stratified analyses, for which power may be less likely a concern but can still be limiting. For example, even by combining all available NO2, health and covariate data available among minor rs1800629 allele carriers (N=1360), we were only powered to detect associations with ORs greater than 1.36 (G*Power, version 3.1.3 (Faul et al. 2009), assuming alpha = 0.05 and power = 0.85). However, 68this limitation is unlikely to hinder the main environmental and genetic effect estimates, which have traditionally been estimated using smaller sample sizes. We nevertheless acknowledge that there remains the possibility that the positive results reported here may be due to chance. A few limitations should be noted. Common to all studies that combine data sources, the data were not collected using identical strategies across all cohorts. This is an especially relevant concern in this study as differing definitions of allergic rhinitis were used by each cohort, which may have affected the study-specific prevalence estimates. For example, the two German cohortswhich relied on the report of a doctor diagnosis in the last 12 months had the lowest prevalence rates of allergic rhinitis, although these rates were similar to that reported for Germany in a global study which relied on questionnaire-based report of symptoms (Asher et al. 2006). Any misclassification of the disease outcome would likely be non-differential and would drive the results towards the null. As such, non-differential misclassification cannot be ruled out as an explanation for our findings. Furthermore, not all participating cohorts were population based, which may influence the prevalence of disease, such as for the CAPPS cohort of children with hereditary allergies. However, our results remained stable when we adjusted for whether a child was a case in the nested case-control cohorts (BAMSE and SAGE), excluded these cases completely from the analysis or removed each cohort sequentially. Second, the panel of SNPs assessed was selective and may not include other genotypes which could influence the pathogenesis and expression of allergic rhinitis and aeroallergen sensitization. In fact, it is quite likely that a complex interaction of genes is required to determine susceptibility. Our selection was based on the literature which suggest that genes involved in inflammation or oxidative stressmetabolism may play a role and on the availability of the SNP in at least three cohorts. Third, although all exposure estimates were individually assigned to each participant, which is a major strength of this study, exposure misclassification remains possible. Furthermore, our approach only considered one air pollutant per analysis. This does not reflect an individual's true exposure,which is in reality, a complex combination of several components. Fourth, we did not have information on the moving patterns of the children from all cohorts. Thus, we were unable to assess the percentage of children for whom an estimation of traffic-related air pollution exposure at their home address at birth may not reflect exposures in later childhood. A previous 69examination of this issue found stronger associations between traffic-related air pollution and allergic diseases for children who had never moved (Gehring et al. 2010). As such, the impact of moving most likely biased our air pollution results towards the null. Population stratification is also likely of minimal concern as 95.1% of our study participants were Caucasian. Finally, selective drop-out is unlikely to have affected the main genetic results of this study as it is improbable that an individual's genotype influenced their decision to participate. In conclusion, a pooled analysis of six birth cohorts suggests that the generally null effect of traffic-related air pollution on allergic rhinitis and sensitization is not modified by ten SNPs in the GSTP1, TNF, TLR2 and TLR4 genes. Although traffic-related air pollution increased the risk of allergic rhinitis in the pooled analysis, this result was not robust to single cohort exclusions. Children with at least one minor rs1800629 allele in their TNF gene or one minor rs1927911 allele in their TLR4 gene may be at a higher risk of allergic rhinitis by school-age. Although the effect estimates observed are small, both SNPs are present in a large proportion of the population(31.2% and 43.5% in this study). The biological mechanisms behind these possible associations remain unknown.704   Associations between the 17q21 region and allergic rhinitis in five birth cohorts 34.1   IntroductionAllergic rhinitis prevalence is growing worldwide (Björkstén et al. 2008). Its causes are multifactorial, including environmental and genetic influences (Bousquet et al. 2012). The high heritability and comorbidity between allergic rhinitis, asthma and other allergic conditions suggests common etiologies and genetic susceptibility loci (Pinart et al. 2014), some of which have been identified in recent GWASs (Andiappan et al. 2011; Bønnelykke et al. 2013; Hinds et al. 2013; Ramasamy et al. 2011; Ferreira et al. 2014). The discovery of common genetic risk factors contributes to our understanding of allergic disease pathogenesis.Genetic studies on asthma have provided valuable insight regarding the possible mechanisms regulating this complex disease (Melén and Pershagen 2012). Among the most robust and replicated signals is the chromosome 17q21 locus, originally identified in the first GWAS on childhood asthma (Moffatt et al. 2007). Allele-specific differences in the expression of genes at this locus have been identified, such as ZPBP2, ORMDL3 and GSDMB. Expression levels of ORMDL3 and GSDMB are co-regulated and opposite to that of ZPBP2 (Verlaan et al. 2009a, 2009b). Evidence implicating other genes, such as GSDMA, also exists (Hao et al. 2012). Nevertheless, the above mentioned genes are unlikely to be the sole mechanism by which geneticvariability at the 17q21 locus affects disease susceptibility (Halapi et al. 2010), which remains largely unexplained. The few studies which have examined whether this strong asthma locus is also associated with allergic rhinitis are conflicting. Small cross-sectional studies have primarily reported null findings (Bisgaard et al. 2009; Balantic et al. 2013; Kavalar et al. 2012), however a recent cross-3 - A version of this manuscript has been published: Fuertes E, Söderhäll C, Acevedo N, Becker A, Brauer M, Chan-Yeung M, Dijk FN, Heinrich J, de Jongste J, Koppelman GH, Postma DS, Kere J, Kozyrskyj AL, Pershagen G, Sandford A, Standl M, Tiesler CMT, Waldenberger M, Westman M, Carlsten C, Melén E. Associations between the 17q21 region and allergic rhinitis in five birth cohorts. J Allergy Clin Immunol. doi:10.1016/j.jaci.2014.08.016.71sectional study on two distinct Japanese adult populations reported consistent significant associations between variants at the 17q21 locus and allergic rhinitis (Tomita et al. 2013). The first two allergic rhinitis GWASs reported null findings for this risk locus (Andiappan et al. 2011;Ramasamy et al. 2011) as did a GWAS on self-reported rhinitis symptoms (Hinds et al. 2013). However, a recent larger GWAS identified the 17q21 region as one of eleven regions significantly associated at the genome-wide significance level with asthma and hayfever modelled independently and as a combined phenotype (Ferreira et al. 2014). Further large, well-defined studies are needed to clarify these inconsistencies. In a pooled analysis of five birth cohorts, we examined whether seven SNPs located at the 17q21locus were associated with allergic rhinitis from early childhood to adolescence, and whether possible associations were modified in subjects also diagnosed with asthma. Given that  associations between the 17q21 region and asthma appear strongest for childhood asthma (Bisgaard et al. 2009; Bouzigon et al. 2008; Hirota et al. 2008; Leung et al. 2009; Moffatt et al. 2007; Wu et al. 2009), we also examined whether the strength of any association with allergic rhinitis depends on the age of onset using three cohorts with available longitudinal data. 4.2   Methods4.2.1   Study populationFive birth cohorts were included in this study: BAMSE (Wickman et al. 2002), PIAMA (Brunekreef et al. 2002), the two Canadian cohorts SAGE (Kozyrskyj et al. 2009) and CAPPS (Chan-Yeung et al. 2000) and the combined German cohorts GINIplus (Filipiak et al. 2007) and LISAplus (Heinrich et al. 2002a). Only data from the Munich centre from these latter two combined German cohorts were included in the current analysis, hereon referred to as GINI/LISA South, as genome-wide genetic data were only available for that area (Wesel, Leipzigand Bad Honnef were excluded). The health outcomes, environmental exposures and covariates of these five cohorts have been previously harmonized within the TAG collaboration (MacIntyre et al. 2013) and have been used to examine gene-environment interactions with air pollution for allergic rhinitis (Chapter 3 of this dissertation; (Fuertes et al. 2013)) and asthma (MacIntyre et al.2014a). Each cohort received ethical approval from their local Institutional Review Boards. 724.2.2   Health outcomesParent-completed questionnaires were used to collect health data. Allergic rhinitis definitions varied slightly by cohort (Table 2). A doctor diagnosis of allergic rhinitis was used in GINI/LISA South, CAPPS and SAGE. The assessment in PIAMA relied on a report of rhinitis symptoms (sneezing, runny blocked nose in the last 12 months when the child did not have a cold or flu). InBAMSE, either a doctor diagnosis or a report of symptoms was used. Asthma was defined using a doctor diagnosis for all cohorts except BAMSE, for which asthma was defined as at least four episodes of wheeze in the last 12 months or at least one episode of wheeze with use of inhaled corticosteroids in the last 12 months. In CAPPS and SAGE, children also underwent a medical examination to confirm allergic rhinitis and asthma diagnoses. Aeroallergen sensitization was assessed at four, eight and 16 years in BAMSE, six and ten years in GINI/LISA South and four, eight and 11/12 years in PIAMA. A positive reaction was defined as any specific IgE level > 0.35 kU/L. For CAPPS and SAGE, sensitization was assessed at seven and eight years, respectively, by skin prick testing. A positive reaction was defined as having a wheal diameter > 3 mm. All available aeroallergens were considered (Alternaria, birch, cats, cockroaches, Dactylis, dogs, feathers, grass, house dust mites, moulds (Cladosporium herbarum), mugwort, ragweed, rye, timothy grass, trees and weeds). Although not all cohorts had information on all aeroallergens, several aeroallergens were tested in each cohort (Table 2).The primary outcome in this analysis (allergic rhinitis) was defined as a positive report of allergic rhinitis (as defined above; cohort-specific definitions provided in Table 2) and a concomitant positive sensitization test to any aeroallergen. As this specific homogenous definition requires information on sensitization status, the primary outcome was only defined at ages at which sensitization data were available. Age-specific controls were defined as those without allergic rhinitis and who were not sensitized to any aeroallergen. Allergic rhinitis with and without concomitant asthma were defined as additional outcomes. 4.2.3   Genetic dataSeven SNPs, which include top GWAS hits for asthma (rs7216389 and rs2305480 (Moffatt et al. 732007)), were genotyped in a subset of the BAMSE cohort (N=2,033 with DNA available (Acevedo et al. 2013a, 2013b)) by MALDI-TOF mass spectrometry (SEQUENOM® Inc). For the CAPPS, GINI/LISA South, PIAMA and SAGE cohorts, information on these seven SNPs was extracted from imputed genome-wide data (details provided in Appendix A). Data on only four SNPs (rs2305480, rs7216389, rs12603332 and rs3744246) were available for CAPPS and SAGE. All imputed dosage data were converted to hard coding. Genotypes were ultimately coded as 0, 1 and 2 according to the number of effect alleles. As a sensitivity analysis, the longitudinal model results for the GINI/LISA South cohort were compared using hard and dosage coding for the SNPs, and the results were very consistent. Pooled associations were also compared using a dominant genetic model rather than an additive genetic model and again the results were consistent. 4.2.4   Statistical analysisPooled and cohort-specific longitudinal associations were analyzed using generalized estimating equation models with a logit link assuming an exchangeable working correlation structure (geeglm function from the geepack package in R (Halekoh et al. 2006)), and were adjusted for age and cohort (pooled models only). Whether the strength of the association depended on age ofonset was explored by examining cross-sectional associations using logistic regression per available year per cohort. Risk estimates are presented as ORs with 95% CI. Statistical analyses were conducted using the statistical program R, version 2.13.1 (R Core Team 2012).4.2.5   Sensitivity analyses Given the known associations between genetic variation at the 17q21 locus and asthma, and between allergic rhinitis and asthma, sensitivity analyses were conducted to try and disentangle whether any observed association between the 17q21 region and allergic rhinitis was a consequence of a shared genetic risk factor for allergic rhinitis and asthma, a direct causal relationship between the two outcomes or a combination of both. First, allergic rhinitis with and without concomitant asthma were modelled as additional outcomes. Second, longitudinal associations with allergic rhinitis were adjusted separately for asthma during early-life (up to eight years) and asthma ever reported during follow-up and the cross-sectional associations were 74adjusted for concomitant asthma. Third, analyses with allergic rhinitis were stratified by asthma during early life (up to eight years) and asthma ever during follow-up. 4.3   Results4.3.1   Distribution of genotypes and outcomesIn total, 5,843 children had information on at least one SNP, 4,624 of which had available data on allergic rhinitis for at least one time point. Of these 4,624 children, which consist of the main study population in this analysis, 92.7% identified as Caucasian. The characteristics of the seven SNPs included in this analysis are presented by cohort in Table 14. Effect allele frequencies werevery similar across cohorts for all SNPs. The correlation between five of the SNPs (rs2305480, rs7216389, rs4065275, rs8076131 and rs12603332) was high (r2 > 0.7, D' > 0.9) and more moderate between these five SNPs and the other two (r2 < 0.3, D' > 0.9; rs17608925 and rs3744246). All SNPs were in Hardy-Weinberg equilibrium (p-value > 0.01). 75Table 14: Genetic information for seven single nucleotide polymorphisms at the 17q21 locusSNP GeneNon-effect/effect allelePosition(Build 37)BAMSE a CAPPS b GINI/LISA South b PIAMA b SAGE bN EAF N EAF N EAF N EAF N EAFrs2305480 GSDMB G/A 38062196 1934 0.47 80 0.47 1096 0.45 1386 0.44 109 0.40rs7216389 GSDMB C/T 38069949 1933 0.48 80 0.48 1096 0.49 1386 0.51 109 0.54rs4065275 ORMDL3 A/G 38080865 1939 0.50 NA NA 1096 0.50 1386 0.52 NA NArs8076131 ORMDL3 A/G 38080912 1935 0.47 NA NA 1096 0.45 1386 0.44 NA NArs12603332 ORMDL3 T/C 38082807 1948 0.49 79 0.50 1096 0.50 1386 0.52 109 0.56rs17608925 ORMDL3 T/C 38082831 1932 0.10 NA NA 1096 0.11 1386 0.11 NA NArs3744246 ORMDL3 C/T 38084350 1936 0.19 80 0.19 1096 0.21 1386 0.20 109 0.20a Genotypes derived from genotyped data.b Genotypes derived from imputed data.EAF = effect allele frequency; N = number of children with health outcome information for at least one time point and genotype data; NA = not available; SNP = single nucleotide polymorphism 76In the total population, 21.0% (969/4,624) of included children reported allergic rhinitis at least once. The pooled and cohort-specific proportion of allergic rhinitis cases increased with age, except for the high prevalence at age seven in the pooled population (Table 15). Only the high-risk (for asthma) CAPPS cohort contributed data at this age, which may explain the elevated prevalence at this time point. Direct comparison between cohorts is complicated by the limited overlap in ages at which outcome data were available. Table 15: Pooled and cohort-specific numbers of cases and controls [cases/controls]OutcomeAge (years) Pooled BAMSE CAPPSGINI/LISASouth PIAMA SAGEAllergic rhinitis a 4 152/1675 102/1319 NA NA 50/356 NA6 80/791 NA NA 80/791 NA NA7 22/58 NA 22/58 NA NA NA8 437/2398 253/1426 NA NA 159/888 25/8410 125/593 NA NA 125/593 NA NA11/12 154/505 NA NA NA 154/505 NA16 411/828 411/828 NA NA NA NAAllergic rhinitis with concomitant asthma b4 37/1594 28/1246 NA NA 9/348 NA6 12/780 NA NA 12/780 NA NA7 9/53 NA 9/53 NA NA NA8 115/2313 63/1367 NA NA 37/876 15/7010 25/585 NA NA 25/585 NA NA11/12 25/502 NA NA NA 25/502 NA16 69/797 69/797 NA NA NA NAAllergic rhinitis without concomitant asthma b4 114/1594 73/1246 NA NA 41/348 NA6 68/780 NA NA 68/780 NA NA7 13/53 NA 13/53 NA NA NA8 318/2313 186/1367 NA NA 122/876 10/7010 98/585 NA NA 98/585 NA NA11/12 129/502 NA NA NA 129/502 NA16 333/797 333/797 NA NA NA NAa Age-specific controls were defined as those without allergic rhinitis. b Age-specific controls were defined as those without allergic rhinitis or asthma. NA = not available774.3.2   Longitudinal associationsSignificant longitudinal associations between allergic rhinitis and six of the seven studied SNPs were observed in the pooled data after adjustment for age and cohort (OR: 0.85 [95% CI: 0.77, 0.94] for rs2305480:A, 1.15 [1.04, 1.27] for rs7216389:T, 1.14 [1.03, 1.26] for rs4065275:G, 0.86 [0.77, 0.95] for rs8076131:G, 1.15 [1.04, 1.27] for rs12603332:C and 0.86 [0.75, 0.97] for rs374426:T; Table 16). As the SNPs are in high LD, the observed associations are not independent. The directions of all observed associations for allergic rhinitis were consistent with those for asthma in the current study and in previous studies (Moffatt et al. 2007; Bouzigon et al. 2008; Galanter et al. 2008). Cohort-specific risk estimates were very consistent across all cohorts, with the exception of the smallest cohort CAPPS (Table 16). 78Table 16: Pooled and cohort specific associations between seven single nucleotide polymorphisms at the 17q21 locus and allergic rhinitisSNP (effect allele)Pooled a BAMSE(up to 16 yrs)CAPPS(at 7 years)GINI/LISA South(up to 10 yrs)PIAMA(up to 12 yrs)SAGE(at 8 years)NOR[95%CI] NOR[95%CI] NOR[95%CI] NOR[95%CI] NOR[95%CI] NOR[95%CI]rs2305480 (A) 4605 0.85[0.77, 0.94] 19340.87 [0.75, 1.00] 801.05 [0.52, 2.13] 10960.80 [0.64, 1.01] 13860.87 [0.72, 1.06] 1090.71 [0.36, 1.40]rs7216389 (T) 4604 1.15[1.04, 1.27] 19331.15 [1.00, 1.33] 800.98 [0.46, 2.05] 10961.16 [0.93, 1.44] 1386 1.14 [0.94, 1.38] 1091.25 [0.64, 2.42]rs4065275 (G) 4421 1.14[1.03, 1.26] 1939 1.15 [1.00, 1.33] NA NA 10961.09[0.87, 1.36] 1386 1.13 [0.93, 1.38] NA NArs8076131 (G) 4417 0.86[0.77, 0.95] 19350.87 [0.76, 1.01] NA NA 10960.81 [0.65, 1.02] 1386 0.87 [0.71, 1.06] NA NArs12603332 (C) 4618 1.15[1.04, 1.27] 19481.18 [1.03, 1.36] 791.00[0.47, 2.12] 10961.09 [0.87, 1.35] 1386 1.13 [0.93, 1.37] 1091.15 [0.60, 2.22]rs17608925 (C) 4414 0.88[0.74, 1.05] 19320.97 [0.77, 1.22] NA NA 10960.79 [0.52, 1.20] 1386 0.83 [0.60, 1.15] NA NArs3744246 (T) 4607 0.86[0.75, 0.97] 19360.91 [0.76, 1.08] 801.16 [0.49, 2.76] 10960.88 [0.65, 1.18] 1386 0.78 [0.61, 1.00] 1090.54 [0.21, 1.40]a p-values of significant pooled associations ranged from 0.002 (rs2305480) to 0.017 (rs3744246). Bonferroni corrected p-value corresponds to 0.05/7 = 0.007. Models were adjusted for age except for CAPPS and SAGE for which only one time point was available. Pooled models were additionally adjusted for cohort. The SNPs are in LD (r2 > 0.2, D' > 0.9). Bold = p-value < 0.05; CI = confidence intervals; N = number of children included in the model; NA = not available; OR = odds ratio; SNP = single nucleotide polymorphim79Longitudinal associations with asthma were significant for three of the seven SNPs (rs2305480, rs7216389 and rs8076131; Figure 10; previously reported in part for BAMSE (Acevedo et al. 2013a) and PIAMA (van der Valk et al. 2012)). Associations for allergic rhinitis with concomitant asthma appeared more pronounced (although not significantly) than when allergic rhinitis or asthma were modelled independently. Associations for allergic rhinitis without concomitant asthma were also significant or borderline significant for five of the SNPs. However, these risk estimates tended to be of slightly smaller magnitude than when allergic rhinitis or asthma were modelled independently, although the confidence intervals overlapped. For example, in the pooled data, associations for rs7216389:T were 1.15 [1.04, 1.27] for allergic rhinitis, 1.15 [1.01, 1.32] for asthma, 1.35 [1.12, 1.63] for allergic rhinitis with concomitant asthma and 1.10 [0.99, 1.22] for allergic rhinitis without concomitant asthma. 80Figure 10: Pooled longitudinal associations between seven single nucleotide polymorphisms at the 17q21 locus and allergic rhinitis (black squares), asthma (red stars), allergic rhinitis and concomitant asthma (blue triangles) and allergic rhinitis without concomitant asthma (purple circles). Models are adjusted for age and cohort. Effect alleles are indicated next to the rs number. The SNPs are in LD (r2 > 0.2, D' > 0.9). CI = confidence intervals; OR = odds ratioRisk estimates for allergic rhinitis were similar after adjustment for asthma during early-life (up to eight years) or asthma ever during follow-up (1.09 [0.98, 1.21] and 1.14 [1.01, 1.28] for rs7216389:T, respectively). However, analyses stratified by asthma in early-life and ever during follow-up indicated that associations between the SNPs and allergic rhinitis were only significantamong those with a history of asthma (Table 17). Associations between the SNPs and allergic rhinitis among those with no history of asthma were attenuated and not significant. 81Table 17: Pooled longitudinal associations between seven single nucleotide polymorphisms at the 17q21 locus and allergic rhinitis, stratified by history of asthma SNP (effect allele)Had early-life asthma a (3-8 years)No early-life asthma a (3-8 years)Had asthma ever b No asthma ever bN OR [95%CI] N OR [95%CI] N OR [95%CI] N OR [95%CI]rs2305480 (A) 1234 0.60 [0.48, 0.77] 1677 0.97 [0.84, 1.12] 1320 0.71 [0.58, 0.86] 1533 0.92 [0.79, 1.07]rs7216389 (T) 1231 1.49 [1.19, 1.87] 1680 1.03 [0.89, 1.19] 1318 1.31 [1.08, 1.59] 1535 1.08 [0.93, 1.26]rs4065275 (G) 1237 1.42 [1.14, 1.78] 1684 1.01 [0.88, 1.17] 1324 1.28 [1.06, 1.55] 1539 1.05 [0.91, 1.23]rs8076131 (G) 1234 0.66 [0.52, 0.83] 1677 0.97 [0.84, 1.12] 1320 0.75 [0.62, 0.91] 1533 0.93 [0.79, 1.08]rs12603332 (C) 1242 1.45 [1.16, 1.81] 1688 1.03 [0.89, 1.18] 1329 1.30 [1.08, 1.58] 1543 1.07 [0.92, 1.25]rs17608925 (C) 1232 0.99 [0.68, 1.44] 1675 0.92 [0.73, 1.17] 1318 1.02 [0.75, 1.39] 1531 0.91 [0.70, 1.18]rs3744246 (T) 1232 0.80 [0.59, 1.08] 1681 0.96 [0.80, 1.16] 1319 0.91 [0.71, 1.17] 1536 0.94 [0.77, 1.15]a The control group was defined as children with no allergic rhinitis or asthma at any age between three and eight years. b The control group was defined as children with no allergic rhinitis or asthma at any age during follow-up. Models were adjusted for age and cohort and only include cohorts for which data on early-life asthma were available (BAMSE, GINI/LISA South and PIAMA). The SNPs are in LD (r2 > 0.2, D' > 0.9).Bold = p-value < 0.05; CI = confidence intervals; N = number of children included in the model; OR = odds ratio; SNP = single nucleotide polymorphism824.3.3   Age-specific cross-sectional associations The results from the cross-sectional yearly analyses suggest that the associations between the five SNPs in high LD and allergic rhinitis may be most pronounced for allergic rhinitis diagnosed at later ages. This trend was most apparent in BAMSE (Figure 11A) but less clear in GINI/LISA South (Figure 11B) and PIAMA (Figure 11C). The observed associations did not change after adjustment for concomitant asthma. 83Figure 11: Cohort- and age-specific cross-sectional associations between seven single nucleotide polymorphisms at the 17q21 locus and allergic rhinitis for the three cohorts with available longitudinal data: BAMSE (A), GINI/LISA (B) and PIAMA (C). Effect alleles are indicated next to the rs number. The SNPs are in LD (r2 > 0.2, D' > 0.9).CI = confidence intervals; OR = odds ratio4.4   DiscussionGenetic variants at the 17q21 locus, including top GWAS hits for asthma (rs7216389 and rs2305480), were significantly associated with allergic rhinitis in a pooled longitudinal analysis of 4,624 children from five birth cohorts. Risk estimates appeared most pronounced for children with concomitant allergic rhinitis and asthma. Analyses stratified by asthma history suggest that the largest effect of the 17q21 locus on allergic rhinitis may involve an asthma-dependent mechanism. However, the existence of an asthma-independent mechanism cannot be dismissed by the study results. The finding of a consistent association between genetic variants at the 17q21 locus and allergic rhinitis is in line with the results of a recent GWAS study that reported genome-wide significant associations between the 17q21 region and asthma and hayfever modelled independently and as a combined phenotype (6,685 cases for the combined outcome) (Ferreira et al. 2014). Our resultsare also consistent with those from a recent candidate gene study of two distinct adult Japanese populations (Tomita et al. 2013). However, two smaller GWASs, one with 456 and 676 allergic rhinitis cases in the discovery and replication cohorts (mean age 21.4 years) (Andiappan et al. 2011) and one which included 3,933 self-reported allergic rhinitis cases from four European adult cohorts (Ramasamy et al. 2011), did not identify the 17q21 locus as a significant predictor of disease. This locus was also not associated at the genome-wide significance level with rhinitis symptoms in a GWAS on self-reported allergy (Hinds et al. 2013). A study conducted among 376six year-old Danish children found no association between rs7216389 and a cross-sectional diagnosis of allergic rhinitis at the age of six years or sensitization ever (Bisgaard et al. 2009). This latter study may have been limited by a small sample size and the difficulty of defining allergic rhinitis at the age of six years.In this study, risk estimates for concomitant allergic rhinitis and asthma appeared greater (although not significantly) than when either outcome was modelled independently. As significant associations were independently observed for both outcomes, the elevated risk estimates for the joint outcome are intuitive and were also observed in the GWAS that considereda joint asthma and hayfever outcome (Ferreira et al. 2014). These results may suggest that the 8417q21 locus is involved in the development of multiple clinical manifestations or a more severe type of disease, possibly via a causal pathway involving asthma. A potential asthma-dependent mechanism is further supported by the stratified analyses conducted (Table 17), in which associations between the SNPs and allergic rhinitis were consistently most pronounced for those with a history of asthma, especially during early-life. Furthermore, the 17q21 locus appears to bea stronger risk factor for asthma than for allergic rhinitis (Ferreira et al. 2014). However, this result is in contrast to a study on 154 asthmatic and 71 healthy Slovenian children in which ORMDL3 was associated with asthma without rhinitis but not with asthma with rhinitis (Kavalar et al. 2012), a result which was replicated among 493 Slovenian adults (Balantic et al. 2013).Weaker borderline significant associations for allergic rhinitis without concomitant asthma were also observed in the current study. Thus, we are unable to exclude the possibility that any effect of 17q21 variation on allergic rhinitis may also be mediated through an asthma-independent pathway. This hypothesis is consistent with the study by Tomita et al. (2013) in which associations with allergic rhinitis remained significant after excluding asthma cases, but needs to be confirmed in future studies.The age-specific cross-sectional analyses provide suggestive evidence that associations with the 17q21 locus were strongest for allergic rhinitis diagnosed during later childhood and adolescence. A possible explanation for this result may be that, as the 17q21 locus is associated with early-life asthma (Bisgaard et al. 2009; Bouzigon et al. 2008; Hirota et al. 2008; Leung et al. 2009; Moffatt et al. 2007; Wu et al. 2009), those with early-life asthma are at a higher risk for allergic rhinitis in adolescence (supporting the existence of an asthma-dependent mechanism). A second possible explanation may be that the 17q21 locus is a stronger risk factor for later childhood and adolescent allergic rhinitis. However, in addition to these possibilities, important methodological explanations need also be considered as alternative explanations. The prevalenceof allergic rhinitis increases with age and that of rhinitis caused by viral infections decreases withage. Thus, the cross-sectional analyses at older ages had less outcome misclassification and were better powered to detect associations. Further age-specific research is needed before firmer conclusions regarding any age-specific effect of the 17q21 locus on allergic rhinitis are possible. 85The biological mechanisms by which genetic variability at the 17q21 locus may affect allergic rhinitis are unknown. Previous efforts on this gene region have focused on ORMDL3, which mayaffect epithelial cell remodelling via calcium signalling, T-lymphocyte activation (Cantero-Recasens et al. 2010; Carreras-Sureda et al. 2013) and sphingolipid misregulation (Breslow et al.2010). A recent study also demonstrated a functional role for ORMDL3 in regulating eosinophil trafficking, recruitment and degranulation via regulation of integrins and CD48 (Ha et al. 2013). Eosinophils are involved in the immune and inflammatory responses to allergens and may represent a possible mechanism by which ORMDL3 gene expression affects allergic asthma and allergic rhinitis. As several SNPs at the 17q21 locus were associated with health outcomes in this and previous studies, detailed sequencing of the gene region, functional research and the use of bioinformatic tools are needed to identify which SNPs or other genetic variants causally affect the expression of ORMDL3 and nearby ZPBP2 and GSDMB. Of the SNPs included in this analysis, rs7216389 has been linked to altered levels of ORMDL3 mRNA (Moffatt et al. 2007) and the non-synonymous SNP rs2305480 (GSDMB/Ser311Pro) may affect protein structure and is in a LD block with SNPs with potential regulatory roles (Tulah et al. 2013). Other SNPs are undoubtedly also important. The intronic SNP rs12936231 in the ZPBP2 gene has been suggested to be a key regulator of the 17q21 region as it affects ZPBP2, GSDMB and ORMDL3 gene expression by modifying a CTCF-binding site and nucleosome occupancy (Verlaan et al. 2009a). A second SNP(rs479539) was found to influence the activity of the ZPBP2 promoter in vitro, a genetic effect that appears to be masked by variable methylation of exon 1 of ZPBP2 in lymphoblastoid cell lines (Berlivet et al. 2012). These studies are evidence of the complex genetic and epigenetic (and also possible environment and gene-environment) influences regulating 17q21 gene expression.The results of this study are based on a significantly larger sample size and longer follow-up thanany previous candidate gene study for allergic rhinitis on this region. Data from five birth cohortswere pooled and homogenous phenotypes defined to maximize statistical power. Previously reported criteria to pool and analyze heterogeneity between cohorts (assessment of genetic 86homogeneity, similar phenotype definitions and careful comparison of pooled versus stratified analysis) were considered in this analysis (Bottema et al. 2008). Particularly convincing are the highly consistent results observed across cohorts and in the pooled analyses. Only for the CAPPScohort were associations inconsistent, which may be the result of the recruitment strategy of this cohort (only those at high risk for asthma were recruited) or its small sample size. It is especially noteworthy that the cohort-specific analyses, although consistent in trend, did not reach statisticalsignificance on their own in almost all cases. This highlights the need to carefully pool or meta-analyze homogenous data in order to achieve sufficient statistical power. It nevertheless remains possible that our study may have been underpowered to detect associations that were found to be null or borderline significant. Three of the five birth cohorts included in this analysis had repeated follow-ups throughout childhood which increased our ability to capture cases and allowed a preliminary examination of a potential age-dependent effect of the 17q21 region on allergic rhinitis development. The trends observed in our analysis need to be replicated by future studies. Despite the strengths of this study, certain limitations need to be considered. First, our selection of SNPs was limited. We chose to study SNPs that were available for the largest subset of the cohort populations, which included top hits from previous GWASs and key studies on asthma (Bouzigon et al. 2008; Çalışkan et al. 2013; Moffatt et al. 2010, 2007). The fact that we observedsignificant associations with six of the seven selected markers in this study (which were in LD) isreassuring. Second, the questionnaire-derived allergic rhinitis question used to partially define this study's primary outcome differed slightly across cohorts. However, no differences in risk estimates were observed between cohorts using parental reports of a doctor diagnosis (CAPPS, GINI/LISA South and SAGE), rhinitis symptoms (PIAMA) or both (BAMSE). Third, as this study was hypothesis driven, we chose to present the primary results without adjustment for multiple testing. It is thus possible to argue that some associations may have occurred because of type-I statistical error. However, the consistency of associations across cohorts argues against this interpretation. Furthermore, four of the six pooled associations between allergic rhinitis and the 17q21 locus would have remained significant after a very conservative Bonferroni correction (Table 16). 87In conclusion, genetic variants at the 17q21 locus were significantly associated with allergic rhinitis in a pooled longitudinal analysis of five birth cohorts. Risk estimates appeared most pronounced when concomitant allergic rhinitis and asthma were modelled as a combined outcome. These results support the hypothesis of a shared genetic susceptibility between asthma and allergic rhinitis. Given the high allele frequency of the studied risk variants and the high individual and societal costs of allergic rhinitis, the potential population impact of this genetic locus on respiratory health is likely substantial.885   Childhood intermittent and persistent rhinitis prevalence and climate and vegetation: a global ecologic analysis 45.1   IntroductionThe prevalence of allergic rhinitis is increasing in the majority of countries (Björkstén et al. 2008). Environmental factors are suspected of being important contributing influences. There is now considerable evidence demonstrating that climate change is measurably altering the timing, distribution, quality and quantity of allergenic plants and aeroallergens (Beggs 2004; Shea et al. 2008), the primary risk factors for allergic rhinitis. Such changes are occurring via meteorological factors and through interactions with green-house gases. For example, ragweed inan urban site with higher temperature and carbon dioxide concentrations, similar to those associated with projected climate change, grew faster, flowered earlier and produced significantly greater above-ground biomass and ragweed pollen compared to ragweed grown in arural area (Ziska et al. 2003). Allergic responses may also be heightened by air pollutants acting directly upon the individual and/or through interactions with allergens (Ghiani et al. 2012).Previously, allergic rhinitis symptoms were classified as seasonal or perennial based on the timing of allergen exposure. However, this classification is not universally applicable and poorly reflects a patient's true experiences (Bousquet et al. 2008). An improved classification system, which categorizes rhinitis symptoms as either intermittent or persistent, is applicable worldwide, better suits a patient's needs and is now being recommended (Bousquet et al. 2001). Intermittent and persistent rhinitis are not synonymous with seasonal and perennial rhinitis. For example, a French study found that 43.7% of patients with seasonal rhinitis (as diagnosed by their doctor) had persistent rhinitis and 44.6% of patients with perennial rhinitis had intermittent rhinitis (Demoly et al. 2003). These differing types of rhinitis may also be differentially associated with risk factors (Bousquet et al. 2005). Studies that assess how climatic factors may influence these 4 - A version of this manuscript has been published: Fuertes E, Butland BK, Anderson HR, Carlsten C, Strachan DP, Brauer M and the ISAAC Phase Three study group. 2014. Childhood intermittent and persistent rhinitis prevalence and climate and vegetation: a global ecologic analysis. Ann. Allergy, Asthma, Immunol. 113(4):386-392e9; doi:10.1016/j.anai.2014.06.021.89two types of rhinitis and associated diseases are of interest (Lin and Zacharek 2012), and may provide indications as to the potential effects of future climate change on respiratory health.  One way to examine such associations is through the use of temporal data, as has been done in limited-area studies (Ariano et al. 2010; Breton et al. 2006; Kim et al. 2011; Newhouse and Levetin 2004). However, this approach is not currently feasible on a global scale. Alternatively, spatial associations can be examined. ISAAC is unique in its global scope and is well suited to assess spatial associations between disease and ecological measures of exposure, such as climate.A previous study by ISAAC Phase One reported a suggestive role for long-term climatic conditions on asthma and atopic eczema symptom prevalences in Western Europe (57 centres in 12 countries) (Weiland et al. 2004b). No significant associations were found with allergic rhinitissymptom prevalence. The current study extends this work by taking advantage of the substantially larger ISAAC Phase Three dataset and the newly adopted rhinitis classifications. Specifically, we aimed to assess spatial associations between the prevalence of intermittent and persistent rhinitis symptoms with climate and vegetation in a global context.5.2   Methods5.2.1   Study populationThe rationale and methods for ISAAC Phase Three have been published (Ellwood et al. 2005). The current analysis is limited to ISAAC centres that collected valid data on monthly rhinitis symptoms: 222 centres in 94 countries for 13-14 year-old teenagers and 135 centres in 59 countries for 6-7 year-old children. A diagram summarizing data availability is provided in Figure 12. All collaborating centres obtained ethical approval from their local ethics committee or board and the investigators for the Phase Three study groups included in this analysis are listed in Appendix B. Letters describing the survey were sent out to parents of all children. Parental completion of the questionnaire for the 6-7 year-olds implied consent. For the older age-group, passive consent for the teenager to complete their own questionnaire at school was used by the great majority of centres.90 5.2.2   Health outcomesMonthly data on rhinitis symptoms were collected via standardized parent- (for children 6-7 years-old) or child- (for teenagers 13-14 years-old) completed ISAAC questionnaires (protocols available on the ISAAC website; isaac.auckland.ac.nz/). Individuals were asked to indicate if in the last 12 months they (or their child) had experienced a problem with sneezing or a runny or blocked nose when they (or their child) DID NOT have a cold or flu. Following questions asked whether this nose problem was accompanied by itchy-eyes (at any time in the last 12 months) and in which of the past 12 months this nose problem occurred. Using the monthly data collectedfrom the last of these questions, the prevalences of intermittent (at least one symptom report but not for two consecutive months) and persistent (symptoms for at least two consecutive months) rhinitis symptoms per centre were calculated. Any apparent inconsistencies between stem and subsequent branch questions were accepted and not changed. 5.2.3   Environmental factors and covariatesData on monthly mean daily temperature (Celsius), total precipitation (millimeter), and vapour pressure (hectopascal), averaged over the period of 1991 to 2000 for 0.5o x 0.5o grids 91Figure 12: Data availability for the International Study of Asthma and Allergies in Childhood Phase Three study population(approximately 3025 km2), were obtained from the Intergovernmental Panel on Climate Change Data Distribution centre (Mitchell 2004; Mitchell and Jones 2005). NDVI data for 2005 were obtained from the Global Land Cover Facility at a resolution of 0.07o on a 16-day basis and averaged per month (Pinzon et al. 2005; Tucker et al. 2005). NDVI data ranges from -1 (water) to +1 (dense vegetation), with values close to zero indicating barren areas of rock, sand or snow. Using the monthly data on temperature, precipitation, vapour pressure and NDVI, the mean, maximum, minimum, standard deviation and maximal difference (difference between the monthly maximum and minimum) of monthly measurements for each factor were calculated. Data on GNI per capita were obtained from the World Bank (Atlas Method 2003) (World Bank 2012). When missing, GNI data were imputed using information from the Central Intelligence Agency World Fact Book (2003) (seven countries) (Central Intelligence Agency 2007). Population density data for 2005 were obtained from the Socioeconomic Data and Applications centre (Socioeconomic Data and Applications Center 2004). Centres were classified into five climate types according to the Köppen climate classification system (snow/polar, equatorial, arid,warm temperate with dry winter and warm temperate fully humid) (Kottek et al. 2006). The assignment of environmental variables to the centres has been previously described (Anderson et al. 2012). Briefly, coordinates for the study population were assigned to a 0.1o x 0.1o square and compared with the eight surrounding 0.1o x 0.1o squares (each square covers approximately 121 km2). Of these nine squares, the one with the highest population density was considered the centre grid and used for mapping. Climate data were mapped to this single coordinate. For population density and NDVI, the average values of the centre grid and eight surrounding grids were used (each sized 0.07o x 0.07o, approximately 59 km2; additional details provided in Appendix C). As the resolution of the NDVI data (0.07o x 0.07o) is not the same as used during the original geocoding of the centres (0.1o x 0.1o), the sizes of the grids used for mapping the environmental factors differ. For climate, population density and NDVI data, which were available at the centre level, country-level means were calculated (Begg and Parides 2003), which may not necessarily represent the true mean of a country. Data on GNI per capita were only available at the country-level. 925.2.4   Statistical analysisCorrelations between centre-level variables were assessed using Spearman correlation coefficients as not all variables were normally distributed. Linear regression mixed-models, which allow consideration of the hierarchical structure of the data, were used to assess associations between the prevalence of intermittent and persistent rhinitis symptoms with the mean, maximum, minimum, standard deviation and maximal difference of monthly measurements of temperature, precipitation, vapour pressure and NDVI (as implemented in the lme4 package (Bates et al. 2014)). Fully adjusted models included the exposure of interest, GNI per capita, population density and climate type. Furthermore, all models included country as a random intercept and fixed effects for both the centre- and country-level representation of each explanatory variable, except for GNI per capita which was available only at the country-level. GNI per capita was included in the models as it was associated with the prevalence of atopic symptoms in worldwide analyses (Stewart et al. 2001). Including a random-intercept for country,to allow the prevalence rates in countries to deviate from the estimated overall prevalence in order to avoid spurious results, has been shown to be important (Weiland et al. 2004b). In sensitivity analyses, models were further adjusted for air pollution (PM2.5 and NO2; data sources and assignment to centres previously described (Anderson et al. 2012)), and mutually adjusted for mean vegetation in the models estimating the effects of the climatic factors, and mean temperature, precipitation and vapour pressure in the models estimating vegetation effects.Between-country effects are presented as the average difference in country-level prevalence (per 100 children) associated with a one unit increase in country-level exposure, with corresponding 95% CI. For countries with greater than one participating centre, the average difference in centre-level prevalence (per 100 children) associated with a one unit increase in centre-level exposure were also estimated (within-country effects). All analyses were conducted using the statistical program R, version 2.13.1 (R Core Team 2012). 935.3   Results5.3.1   Distribution of intermittent and persistent rhinitis prevalences For the 13-14 age-group centres, the mean centre prevalence for intermittent rhinitis was highest in equatorial regions (20.0) and lowest in centres with snow/polar climates (8.9) and varied significantly by climate type (anova p-value < 0.001; Table 18). There was less variation by climate type for persistent symptoms (highest in warm temperate fully humid regions (16.1) and lowest in arid climates (11.9); anova p-value =0.04). Table 18: Intermittent and persistent rhinitis symptom prevalence by climate type for the 13-14 age-group centresClimate typeIntermittent rhinitis Persistent rhinitisN Mean ± SD Range Mean ± SD RangeSnow/polar 12 8.9 ± 8.0 0.4, 22.4 14.0 ± 9.3 3.4, 28.3Arid 23 13.3 ± 10.3 0.1, 45.4 11.9 ± 6.0 1.4, 25.6Equatorial 64 20.0 ± 11.0 0.0, 49.5 12.9 ± 8.5 0.0, 33.8Warm temperate with dry winter 21 9.5 ± 4.0 3.5, 18.6 14.6 ± 7.0 3.1, 26.8Warm temperate fully humid 102 15.4 ± 7.8 0.0, 44.9 16.1 ± 7.1 1.1, 32.1N = number of centres; SD = standard deviation Centre-specific sample sizes and intermittent and persistent rhinitis symptom prevalences are reported in Appendices D and E for the 13-14 and 6-7 age-group centres, respectively. Figures 13and 14 depict the global distribution of intermittent and persistent symptom prevalences, respectively, for the 13-14 age-group centres. 9495Figure 13: World map showing the centre prevalence of intermittent rhinitis symptoms for the 13-14 age-group centres.Figure 14: World map showing the centre prevalence of persistent rhinitis symptoms for the 13-14 age-group centres.5.3.2   Distribution and correlation of environmental factorsCharacteristics of modelled variables for the 13-14 age-group centres are provided in Table 19. All three measures of mean climatic variables were positively correlated with intermittent rhinitisprevalence. The prevalence of persistent rhinitis was negatively correlated with mean temperature. GNI per capita was negatively correlated with intermittent rhinitis prevalence, meantemperature and vapour pressure, and positively correlated with persistent rhinitis prevalence andmean NDVI. Population density was negatively correlated with mean NDVI and positively correlated with mean temperature and vapour pressure. Positive correlations were observed between all climatic factors. Mean precipitation was positively correlated with mean vegetation. Correlations were similar for the younger age-group centres although for persistent rhinitis, therewas a positive association with mean precipitation and no evidence of an association with mean temperature.96Table 19: Correlations between modelled variables for the 13-14 age-group centresVariable Time period Median (IQR)Spearman correlation withIntermittentrhinitisPersistentrhinitis Mean NDVIMeantemperatureMeanprecipitationMeanvapourpressureHealth outcomeIntermittent rhinitis ~2000-2003 14.1 (9.7, 20.0) - - 0.074 0.361 0.223 0.332Persistent rhinitis ~2000-2003 13.7 (8.6, 19.5) - - 0.105 -0.195 0.054 -0.084Economic/ populationGNI per capita a 2003 2950 (1483, 13070) -0.202 0.526 0.303 -0.441 -0.021 -0.318Population density 2005 809 (297, 2779) -0.025 0.048 -0.310 0.222 0.062 0.183Environmental factorsMean NDVI (NDVI unit) b 2005 0.4 (0.3, 0.5) - - - -0.058 0.422 0.084Mean temperature (oC) 1991-2000 17.9 (12.1, 24.4) - - - - 0.407 0.925Mean precipitation (mm) 1991-2000 81.7 (50.4, 124.5) - - - - - 0.568Mean vapour pressure (hPa) 1991-2000 14.1 (10.7, 21.7) - - - - - -a Correlations with GNI per capita are with country-level variables (94 countries).b Only 215 centres had information on NDVI.Bold = p-value < 0.05; GNI = gross national income; IQR = interquartile range; NDVI = Normalized Difference Vegetation Index97Correlations between different measures of the same climatic factor were also examined. For the older age-group centres, mean temperature was positively correlated with minimum and maximum temperature and negatively correlated with measures of variability. Mean vapour pressure was positively correlated with the minimum, maximum and standard deviation and negatively correlated with the maximal difference. Mean precipitation and vegetation were positively correlated with all other measures of precipitation and vegetation, respectively.5.3.3   Associations between rhinitis symptoms and environmental factors Between-country associations (average differences in country-level prevalence per 100 children associated with a one unit increase in country-level exposure) for the fully adjusted models for the 13-14 age-group centres are presented in Table 20. Intermittent symptom prevalence was positively associated with the mean and minimum monthly temperature and negatively associated with the maximal difference in monthly temperature. Intermittent symptom prevalence was also positively associated with the mean, maximum, standard deviation and maximal difference of monthly precipitation measurements, and the mean, maximum and minimum monthly vapour pressure measurements. Four of the ten significant associations observed for the older age-group centres were also significant for the younger age-group centres (Table 21). Persistent symptom prevalence was not associated with any climatic variable in the fully adjusted models. Confounder adjusted risk estimates for vegetation were elevated but non-significant for both outcomes. 98Table 20: Between-country associations between intermittent and persistent rhinitis prevalence and environmental factors for the 13-14 age-group centresCountry-level environmental factor (IQR)Average difference in country-level prevalence [95% CI] per100 children per one unit increase in country-level exposureIntermittent rhinitis Persistent rhinitisTemperature Mean (10.4) 0.59 [0.20, 0.98] -0.02 [-0.34, 0.29]Maximum (4.6) 0.33 [-0.15, 0.82] -0.10 [-0.47, 0.27]Minimum (16.1) 0.43 [0.15, 0.71] -0.02 [-0.25, 0.21]Standard deviation (2.1) -0.56 [-1.62, 0.49] 0.05 [-0.78, 0.88]Maximal difference (12.5) -0.36 [-0.67, -0.05] 0.07 [-0.18, 0.32]Precipitation Mean (67.0) 0.05 [0.01, 0.09] 0.01 [-0.02, 0.04]Maximum (166.1) 0.02 [0.01, 0.04] 0.00 [-0.01, 0.02]Minimum (42.7) 0.00 [-0.07, 0.07] 0.00 [-0.06, 0.05]Standard deviation (51.5) 0.09 [0.03, 0.14] 0.01 [-0.04, 0.05]Maximal difference (137.8) 0.03 [0.01, 0.05] 0.00 [-0.02, 0.02]Vapour pressure Mean (10.4) 0.60 [0.21, 0.99] 0.17 [-0.15, 0.48]Maximum (8.6) 0.44 [0.07, 0.81] 0.18 [-0.11, 0.47]Minimum (10.8) 0.55 [0.17, 0.93] 0.12 [-0.19, 0.44]Standard deviation (2.0) 0.25 [-1.24, 1.74] 0.75 [-0.38, 1.87]Maximal difference (4.9) -0.23 [-0.69, 0.23] 0.23 [-0.12, 0.58]Vegetation a Mean (0.1) 8.03 [-8.75, 24.81] 9.45 [-3.84, 22.75]Maximum (0.1) 6.33 [-8.54, 21.20] 7.98 [-3.91, 19.87]Minimum (0.1) 4.31 [-13.52, 22.14] 8.48 [-5.57, 22.53]Standard deviation (0.02) 11.57 [-41.84, 64.98] 19.32 [-23.92, 62.56]Maximal difference (0.1) 5.48 [-13.14, 24.11] 3.51 [-11.51, 18.53]a Vegetation data only available for 215 centres in 87 countries. Models included data from 222 centres in 94 countries and were adjusted for centre mean exposure of interest, centre and country mean population density, country gross national income per capita and climate type.Bold = p-value < 0.05; CI = confidence intervals; IQR = interquartile range 99Table 21: Between-country associations between intermittent and persistent rhinitis prevalence and environmental factors for the 6-7 age-group centresCountry-level environmental factor (IQR)Average difference in country-level prevalence [95% CI] per100 children per one unit increase in country-level exposureIntermittent rhinitis Persistent rhinitisTemperature Mean (10.6) 0.28 [-0.07, 0.64] 0.09 [-0.30, 0.47]Maximum (5.3) 0.06 [-0.33, 0.46] -0.17 [-0.65, 0.31]Minimum (16.8) 0.22 [-0.03, 0.47] 0.11 [-0.16, 0.39] Standard deviation (2.1) -0.24 [-1.04, 0.57] -0.52 [-1.46, 0.42]Maximal difference (13.4) -0.32 [-0.56, -0.07] -0.09 [-0.37, 0.18]Precipitation Mean (67.7) 0.03 [0.00, 0.07] 0.02 [-0.02, 0.06]Maximum (201.2) 0.02 [0.00, 0.04] 0.01 [-0.01, 0.03]Minimum (28.6) -0.04 [-0.10, 0.02] -0.01 [-0.08, 0.07]Standard deviation (64.6) 0.09 [0.04, 0.14] 0.04 [-0.02, 0.10]Maximal difference (182.99) 0.03 [0.01, 0.05] 0.02 [-0.01, 0.04]Vapour pressure Mean (11.1) 0.25 [-0.07, 0.57] 0.18 [-0.19, 0.54]Maximum (9.8) 0.12 [-0.16, 0.40] 0.04 [-0.30, 0.38]Minimum (10.6) 0.31 [0.00, 0.62] 0.19 [-0.16, 0.55]Standard deviation (1.9) 0.02 [-1.05, 1.09] -0.37 [-1.68, 0.94]Maximal difference (5.4) -0.30 [-0.61, 0.01] -0.11[-0.50, 0.28]Vegetation a Mean (0.1) 3.90 [-10.82, 18.62] 14.23 [-2.35, 30.81]Maximum (0.2) 4.84 [-8.09, 17.78] 10.06 [-4.81, 24.94]Minimum (0.1) 0.81 [-14.67, 16.29] 13.98 [-3.39, 31.35]Standard deviation (0.03) 3.62 [-39.50, 46.74] 11.00 [-37.10, 59.10]Maximal difference (0.1) 4.90 [-9.54, 19.33] 1.06 [-15.70, 17.82]a Vegetation data only available for 131 centres in 56 countries. Models included data from 132 centres in 57 countries and were adjusted for centre mean exposure of interest, centre and country mean population density, country gross national income per capita and climate type.Bold = p-value < 0.05; CI = confidence intervals; IQR = interquartile range 100For the 13-14 year-old age-group, 37 countries had greater than one participating centre (165 centres in total). These centres were used to assess the average difference in centre-level prevalence per 100 children associated with a one unit increase in centre-level exposure (Table 22). Within-country positive associations were observed between intermittent symptom prevalence and mean monthly temperature and the standard deviation of monthly vapour pressure measurements. Further positive associations were also observed between persistent symptom prevalence and mean and maximum monthly temperature and the mean, maximum, minimum and standard deviation of monthly vapour pressure measurements. When the within-country analyses were replicated in the 6-7 year-old age-group centres (98 centres in 23 countries), only the associations between persistent symptoms and mean, maximum and minimum monthly vapour pressure were significant (Table 23). A positive within-country association was also observed between the maximum monthly vegetation measurements and persistent symptom prevalence, but only among the 6-7 age-group centres. 101Table 22: Within-country associations between intermittent and persistent rhinitis prevalence and environmental factors among countries with two or more centres per country for the 13-14 age-group centresCentre-level environmental factor (IQR)Average difference in centre-level prevalence [95% CI] per100 children per one unit increase in centre-level exposureIntermittent rhinitis Persistent rhinitisTemperature Mean (10.7) 0.43 [0.02, 0.84] 0.39 [0.03, 0.76]Maximum (6.8) 0.32 [-0.05, 0.68] 0.37 [0.05, 0.70]Minimum (11.9) 0.26 [-0.10, 0.62] 0.16 [-0.16, 0.49]Standard deviation (2.6) 0.63 [-0.50, 1.75] 0.71 [-0.30, 1.72]Maximal difference (12.1) 0.12 [-0.28, 0.52] 0.22 [-0.13, 0.58]Precipitation Mean (68.7) -0.03 [-0.06, 0.01] -0.01 [-0.04, 0.03]Maximum (162.9) -0.01 [-0.02, 0.01] -0.01 [-0.03, 0.00]Minimum (35.1) -0.05 [-0.11, 0.01] 0.00 [-0.05 0.06]Standard deviation (58.5) -0.01 [-0.06, 0.04] -0.04 [-0.08, 0.00]Maximal difference (161.5) 0.00 [-0.02, 0.02] -0.01 [-0.03, 0.00]Vapour pressure Mean (8.4) 0.22 [-0.15, 0.59] 0.37 [0.04, 0.70]Maximum (10.1) 0.20 [-0.08, 0.49] 0.31 [0.06, 0.56]Minimum (6.2) 0.25 [-0.17, 0.67] 0.39 [0.02, 0.76]Standard deviation (2.4) 1.14 [0.06, 2.21] 1.32 [0.37, 2.27]Maximal difference (5.7) 0.30 [-0.10, 0.71] 0.28 [-0.09, 0.64]Vegetation Mean (0.2) -1.25 [-10.81, 8.31] -2.89 [-11.42, 5.64]Maximum (0.2) -0.40 [-8.82, 8.02] -0.25 [-7.79, 7.29]Minimum (0.2) -2.93 [-13.61, 7.76] -5.47 [-14.98, 4.04]Standard deviation (0.05) 0.75 [-36.50, 37.99] 7.53 [-25.95, 41.02]Maximal difference (0.2) 3.19 [-9.41, 15.78] 6.94 [-4.31, 18.20]Models included data from 165 centres in 37 countries and were adjusted for the country mean exposure of interest, centre and country mean population density, country gross national income per capita and climate type.Bold = p-value < 0.05; CI = confidence intervals; IQR = interquartile range 102Table 23: Within-country associations between intermittent and persistent rhinitis prevalence and environmental factors among countries with two or more centres per country for the 6-7 age-group centresCentre-level environmental factor (IQR)Average difference in centre-level prevalence [95% CI] per100 children per one unit increase in centre-level exposureIntermittent rhinitis Persistent rhinitisTemperature Mean (10.9) -0.37 [-0.85, 0.10] 0.42 [-0.01, 0.84]Maximum (7.2) -0.33 [-0.72, 0.07] 0.32 [-0.02, 0.67] Minimum (13.4) -0.32 [-0.77, 0.13] 0.32 [-0.09, 0.74]Standard deviation (2.4) -0.16 [-1.56, 1.24] 0.48 [-0.77, 1.72]Maximal difference (12.7) -0.02 [-0.48, 0.43] 0.13 [-0.28, 0.55]Precipitation Mean (57.2) 0.01 [-0.04, 0.05] 0.01 [-0.03, 0.05]Maximum (177.7) 0.00 [-0.02, 0.02] 0.00 [-0.02, 0.01]Minimum (30.2) -0.01 [-0.08, 0.06] 0.00 [-0.06, 0.06]Standard deviation (68.5) 0.00 [-0.04, 0.05] -0.01 [-0.05, 0.03]Maximal difference (172.05) 0.00 [-0.02, 0.02] 0.00 [-0.02, 0.01]Vapour pressure Mean (9.0) -0.38 [-0.79, 0.03] 0.42 [0.05, 0.79]Maximum (11.4) -0.30 [-0.62, 0.03] 0.32 [0.03, 0.61] Minimum (7.9) -0.25 [-0.71, 0.20] 0.45 [0.04, 0.86] Standard deviation (2.2) -0.09 [-1.38, 1.20] 0.63 [-0.50, 1.77]Maximal difference (5.7) -0.17 [-0.66, 0.31] 0.25 [-0.19, 0.69]Vegetation Mean (0.1) -9.05 [-19.66, 1.55] 8.37 [-1.04, 17.77]Maximum (0.2) -7.58 [-17.01, 1.85] 8.81 [0.53, 17.09]Minimum (0.1) -8.11 [-19.77, 3.55] 6.90 [-3.52, 17.31]Standard deviation (0.05) -30.47 [-74.18, 13.25] 33.08 [-5.55, 71.71]Maximal difference (0.2) -5.49 [-20.63, 9.66] 11.09 [-2.22, 24.40]Models included data from 98 centres in 23 countries and were adjusted for the country mean exposure of interest, centre and country mean population density, country gross national income per capita and climate type.Bold = p-value < 0.05; CI = confidence intervals; IQR = interquartile range When the analysis was restricted to children with intermittent or persistent rhinitis symptoms who also reported itchy-eye symptoms in the last 12 months, risk estimates remained in the samedirection but were generally smaller and not significant for one of the ten between-country associations and five of the eight within-country associations observed for the 13/14 age-group centres (results for the 13/14 age-group centres presented in Table 24). Between and within-country associations for the 13/14 age-group centres were generally robust to adjustments for air 103pollution (PM2.5 and NO2) and mutual adjustment for meteorological and vegetation factors, although a loss of statistical significance was occasionally observed (results for the 13/14 age-group centres presented in Table 25).104Table 24: Between- and within-country associations between environmental factors and the prevalence of intermittent rhinitis and itchy-eyes and persistent rhinitis and itchy-eyes for the 13-14 age-group centresEnvironmental factor Average difference in country-level prevalence [95% CI]per 100 children per one unit increase in country-level exposure aAverage difference in centre-level prevalence [95% CI]per 100 children per one unit increase in centre-level exposure bIntermittent and itchy eyes Persistent and itchy eyes Intermittent and itchy eyes Persistent and itchy eyesTemperature Mean 0.36 [0.16, 0.56] 0.03 [-0.14, 0.21] 0.23 [0.05, 0.42] 0.13 [-0.07, 0.32]Maximum 0.17 [-0.09, 0.43] -0.07 [-0.27, 0.14] 0.21 [0.04, 0.38] 0.15 [-0.02, 0.32]Minimum 0.27 [0.13, 0.41] 0.03 [-0.10, 0.16] 0.14 [-0.03, 0.30] 0.00 [-0.18, 0.17]Standard deviation -0.60 [-1.15, -0.06] -0.21 [-0.67, 0.25] 0.49 [-0.02, 1.01] 0.61 [0.08, 1.13]Maximal difference -0.25 [-0.40, -0.09] -0.04 [-0.18, 0.10] 0.10 [-0.08, 0.28] 0.18 [0.00, 0.37]Precipitation Mean 0.02 [0.00, 0.04] 0.00 [-0.01, 0.02] -0.02 [-0.04, 0.00] -0.01 [-0.03, 0.01]Maximum 0.01 [0.00, 0.02] 0.00 [-0.01, 0.01] 0.00 [-0.01, 0.00] -0.01 [-0.02, 0.00]Minimum 0.00 [-0.03, 0.04] 0.01 [-0.02, 0.04] -0.03 [-0.06, 0.00] -0.01 [-0.03, 0.02]Standard deviation 0.04 [0.01, 0.07] 0.00 [-0.03, 0.03] -0.01 [-0.03, 0.02] -0.02 [-0.04, 0.00]Maximal difference 0.02 [0.00, 0.03] 0.00 [-0.01, 0.01] 0.00 [-0.01, 0.01] -0.01 [-0.02, 0.00]Vapour pressureMean 0.33 [0.14, 0.53] 0.11 [-0.07, 0.28] 0.11 [-0.06, 0.28] 0.10 [-0.07, 0.28]Maximum 0.24 [0.04, 0.43] 0.09 [-0.07, 0.25] 0.11 [-0.02, 0.24] 0.10 [-0.03, 0.24]Minimum 0.32 [0.13, 0.51] 0.10 [-0.07, 0.28] 0.15 [-0.04, 0.35] 0.13 [-0.06, 0.33]Standard deviation 0.00 [-0.80, 0.80] 0.25 [-0.39, 0.88] 0.62 [0.13, 1.11] 0.60 [0.09, 1.10]Maximal difference -0.19 [-0.44, 0.05] 0.03 [-0.17, 0.23] 0.14 [-0.05, 0.32] 0.11 [-0.08, 0.30]Vegetation Mean 6.22 [-2.66, 15.10] c 6.00 [-1.36, 13.36] c -0.74 [-5.17, 3.70] -2.56 [-7.03, 1.91]Maximum 3.18 [-4.70, 11.05] c 4.22 [-2.37, 10.81] c -0.28 [-4.18, 3.62] -0.73 [-4.69, 3.24]Minimum 6.83 [-2.65, 16.32] c 6.39 [-1.39, 14.18] c -1.39 [-6.33, 3.56] -4.33 [-9.30, 0.63]Standard deviation -3.19 [-31.33, 24.95] c 6.49 [-17.42, 30.40] c 0.49 [-16.75, 17.73] 3.59 [-14.00, 21.18]Maximal difference -2.28 [-12.17, 7.60] c -0.19 [-8.52, 8.14] c 1.26 [-4.57, 7.09] 4.31 [-1.59, 10.22]105a Based on data from 222 centres in 94 countries. b Based on data from 165 centres in 37 countries. c Vegetation data was only available for 215 centres in 87 countries. Models were adjusted for centre mean exposure of interest (for between-country associations) or country mean exposure of interest (for within-country associa-tions), as well as the centre and country mean population density, country gross national income per capita and climate type. Bold = p-value < 0.05; CI = confidence intervals106Table 25: Between- and within-country associations between intermittent and persistent rhinitis prevalence and environmental factors for the 13-14 age-group centres, independently adjusted for PM2.5 mass and NO2, and mutually adjusted for meteorological and vegetation factors. Average difference in country-level prevalence [95%CI] per 100 children per one unit increase in country-level exposure Average difference in centre-level prevalence [95%CI] per 100 children per one unit increase in centre-level exposureIntermittent rhinitis Persistent rhinitis Intermittent rhinitis Persistent rhinitisAdjusted for PM2.5 mass aTemperature Mean 0.79 [0.30, 1.28] 0.12 [-0.28, 0.52] 0.46 [-0.02, 0.94] 0.51 [0.08, 0.95]Standard deviation 0.99 [-1.00, 2.97] -0.68 [-2.27, 0.90] 1.79 [-0.21, 3.79] 0.73 [-1.10, 2.56]Precipitation Mean 0.04 [-0.01, 0.10] 0.00 [-0.04, 0.04] -0.02 [-0.07, 0.03] -0.01 [-0.05, 0.04]Standard deviation 0.09 [0.02, 0.16] -0.01 [-0.07, 0.04] 0.00 [-0.06, 0.05] -0.03 [-0.08, 0.02]Vapour pressureMean 0.64 [0.15, 1.14] 0.28 [-0.13, 0.68] 0.18 [-0.25, 0.62] 0.48 [0.09, 0.86]Standard deviation 1.00 [-0.75, 2.74] 0.88 [-0.47, 2.22] 1.33 [0.03, 2.63] 1.57 [0.42, 2.72]Vegetation b Mean 6.31 [-15.91, 28.53] 7.88 [-9.63, 25.40] 0.72 [-10.96, 12.39] -3.71 [-14.19, 6.78]Standard deviation -8.89 [-80.35, 62.57] -7.01 [-64.62, 50.60] -2.16 [-47.11, 42.79] 8.95 [-31.78, 49.69]Adjusted for NO2 aTemperature Mean 0.67 [0.18, 1.15] 0.10 [-0.29, 0.48] 0.45 [-0.04, 0.93] 0.51 [0.08, 0.95]Standard deviation -0.34 [-1.60, 0.93] -0.33 [-1.29, 0.63] 0.74 [-0.55, 2.04] 0.88 [-0.29, 2.05]Precipitation Mean 0.03 [-0.02, 0.09] 0.01 [-0.03, 0.06] -0.03 [-0.09, 0.03] 0.00 [-0.05, 0.06]Standard deviation 0.03 [-0.27, 0.34] -0.14 [-0.40, 0.12] -0.18 [-0.52, 0.17] -0.31 [-0.62, -0.01]Vapour pressureMean 0.58 [0.09, 1.06] 0.29 [-0.10, 0.67] 0.17 [-0.27, 0.61] 0.47 [0.09, 0.86]Standard deviation 0.27 [-1.45, 1.99] 0.99 [-0.29, 2.27] 1.14 [-0.12, 2.40] 1.63 [0.52, 2.74]Vegetation b Mean 11.47 [-9.32, 32.27] 10.39 [-5.64, 26.41] -0.17 [-11.85, 11.52] -2.34 [-12.88, 8.21]Standard deviation 4.88 [-58.18, 67.94] 21.56 [-28.18, 71.30] -3.63 [-48.25, 40.98] 12.60 [-27.86, 53.06]107Average difference in country-level prevalence [95%CI] per 100 children per one unit increase in country-level exposure Average difference in centre-level prevalence [95%CI] per 100 children per one unit increase in centre-level exposureIntermittent rhinitis Persistent rhinitis Intermittent rhinitis Persistent rhinitisMutually adjusted for meteorological and vegetation factors cTemperature d Mean 0.63 [0.34, 0.92] -0.06 [-0.28, 0.17] 0.29 [-0.06, 0.65] 0.29 [-0.02, 0.61]Standard deviation -1.35 [-2.27, -0.43] -0.08 [-0.75, 0.59] 0.41 [-0.67, 1.49] 0.52 [-0.45, 1.49]Precipitation d Mean 0.08 [0.03, 0.12] 0.01 [-0.03, 0.04] -0.02 [-0.06, 0.02] 0.00 [-0.04, 0.03]Standard deviation 0.10 [0.05, 0.16] 0.01 [-0.03, 0.05] -0.01 [-0.05, 0.04] -0.03 [-0.07, 0.01]Vapour pressure dMean 0.67 [0.37, 0.97] 0.00 [-0.24, 0.23] 0.08 [-0.23, 0.38] 0.20 [-0.07, 0.48]Standard deviation 0.25 [-1.35, 1.85] 0.57 [-0.52, 1.66] 1.06 [-0.01, 2.13] 1.27 [0.31, 2.22]Vegetation e Mean 1.23 [-17.29, 19.74] 7.56 [-6.77, 21.90] 3.85 [-6.10, 13.80] -1.37 [-10.53, 7.78]Standard deviation 25.67 [-38.26, 89.61] -11.35 [-60.56, 37.87] 17.57 [-20.33, 55.47] 16.77 [-18.14, 51.67]a Between-country associations based on data from 174 centres in 80 countries and within-country associations based on data from 129 centres in 35 countries. b Between-country associations based on data from 169 centres in 75 countriesc Between-country associations based on data from 215 centres in 87 countries and within-country associations based on data from 165 centres in 37 countriesd Additionally adjusted for centre and country mean vegetatione Additionally adjusted for centre and country mean temperature, precipitation and vapour pressure.Models were adjusted for centre mean exposure of interest (for between-country associations) or country mean exposure of interest (for within-country associa-tions), as well as the centre and country mean population density, country gross national income per capita and climate type. Bold = p-value < 0.05; CI = confidence intervals1085.4   DiscussionSeveral spatial associations between climatic factors and the prevalence of intermittent and persistent rhinitis symptoms were observed in a non-randomly selected group of centres and countries which participated in ISAAC Phase Three. Our results suggest a general positive association of mean monthly temperature and vapour pressure (which were highly correlated, rs =0.925), and precipitation, with symptom prevalence. This cross-sectional global study is a first step in assessing how climate change may affect allergic rhinitis symptoms. However, the generalizability of the observed associations and the influence of other factors not accounted for in this analysis on the relationship between climate and rhinitis remain unknown. To account for the hierarchical nature of the data, we examined both between- and within-countyeffects (country- and centre-level associations, respectively). Associations with intermittent rhinitis prevalence were most consistent at the country-level and only limited evidence was observed at the centre-level. Persistent rhinitis prevalence was only associated with climatic factors at the centre-level. One explanation for these findings may be that intermittent symptoms are more strongly associated with temporal changes in certain aeroallergens affected by climate (for example, pollens and moulds), whereas lifestyle factors have a greater influence on persistent symptoms. Thus, we were able to detect the relatively strong between-country associations for climate with intermittent symptom prevalence but not with persistent symptom prevalence, as this latter association may be confounded by more important causal lifestyle factors that differ by country. Differences in sensitization patterns across areas are also likely to exist. As these factors are less likely to differ within countries, the potentially weaker association for persistent symptom prevalence with climatic factors may only be detectable in the within-country analyses which are less likely influenced by unmeasured confounding. Our finding of a positive association of temperature, vapour pressure and precipitation with rhinitis symptoms has been in part observed in previous single- or limited-area studies (Ariano etal. 2010; Kim et al. 2011; Newhouse and Levetin 2004), but others report null findings (Bhattacharyya 2009; Breton et al. 2006; De Marco et al. 2002; Zanolin et al. 2004). It is challenging to reconcile these previous studies with the current one as the effect of climate on 109aeroallergen distribution (and thus presumably rhinitis) likely varies with local vegetation and geographical area (D’Amato and Cecchi 2008). Furthermore, these studies are based on data collected from adults whereas the current study focuses on children and teenagers. There are onlytwo published multi-area studies on rhinitis. A study including 48 European centres conducted onadults concluded that climate can account for significant variability in (mostly asthma-related) respiratory symptom prevalence, although hayfever was only associated with temperature during the hottest month (Verlato et al. 2002). The second study, and the only one to include centres outside of Europe and to be based on children, utilized data from Phase One of ISAAC (Weiland et al. 2004b). Climatic factors and allergic rhinitis symptoms were not found to be associated, although suggestive evidence for asthma and atopic eczema was reported for Western European centres. World-wide associations were inconsistent. By contrast, the current study is based on a substantially larger number of centres (222 centres in 94 countries and 144 centres in 81 countries were used in the current and previous world-wide analyses for the 13-14 year-olds, respectively) and examines intermittent and persistent rhinitis prevalences as outcomes rather than any report of rhinitis in the last 12 months. As the prevalence of rhinitis symptoms has increased in the majority of centres between the timing of Phase One (1992-1997) and Phase Three (2000-2003) (Asher et al. 2006), associations with intermittent and persistent rhinitis symptoms may be easier to detect in the Phase Three data. The current study is the first to examine associations between vegetation and rhinitis prevalence in a global context. The mechanisms by which climatic factors may influence rhinitis are unclear, but may be indirect via changes in indoor factors, such as dampness, which are likely to affect the distribution of indoor aeroallergens, or via changes in pollen types, distributions and concentrations, a phenomenon well documented in previous studies covering smaller geographic areas (Shea et al. 2008). As an NDVI estimate assigned to one geographical coordinate is unlikely to accurately reflect the vegetation of a whole area, we used the average of nine NDVI estimates, each assigned to a 0.07o x 0.07o square (area coverage approximately 59 km2). The majority of the risk estimates obtained were elevated, although all but one were non-significant despite moderate variation in NDVI units between countries (range of mean country vegetation estimates was 0.44) and within-countries (for example, range of mean centre vegetation 110estimates was 0.68 in Chile with five participating centres and 0.49 in Spain with 12 participating centres). Including interaction terms between vegetation and temperature exposuresdid yield two significant associations for intermittent rhinitis (p-value = 0.04 for the interaction term between mean temperature and mean vegetation and p-value = 0.02 for the interaction term between minimum temperature and minimum vegetation). We also examined whether associations may be stronger among smaller countries where the NDVI estimate is more likely toreflect exposures throughout the country. Only when the analysis was restricted to countries smaller than 50,000 km2 (21 centres in 17 countries) was a positive between-country association found between intermittent symptom prevalence and the maximal difference and standard deviation of monthly vegetation estimates. As it is well established that pollen can disperse over large distances (D’Amato et al. 2007), it is possible that our ecological study design may not be ideal for assessing associations with vegetation which may be more heterogeneous over small areas. Using NDVI as a surrogate for vegetation also did not allow us to distinguish between plant species of differing allergenicity. An important strength of this study is the ability to examine associations between and within countries using a multilevel modelling approach. Examining between-country associations allows the use of the entire dataset, thereby taking full advantage of the large number and exposure contrasts of the countries participating in ISAAC Phase Three. However, it is likely thatsubstantial between-country differences were not captured by model adjustments for GNI per capita, climate type or air pollution, and thus we cannot exclude the possibility that residual confounding may be affecting the between-country associations. The differentiation between intermittent and persistent symptoms is, however, less likely to be affected by translation or awareness artifacts attributable to language and culture, which often hinder international comparisons of prevalence in ISAAC (and other questionnaire-based studies). Nevertheless, it remains possible that there may be some variation in the way the questionnaires were completed or administered. Confounding, although still possible, is less of a concern for the within-country associations as causal factors are less likely to differ to the same degree within countries as between countries. However, we did not account for individual risk factors, a limitation common to all ecological studies. Given that only countries with two or more participating centres per 111country can contribute to the within-country analysis, these results are hindered by a lack of statistical power, both in terms of fewer centres and smaller exposure variation. Finally, a possible limitation of this study is that associations were not adjusted for multiple testing as it is not clear that the different exposure metrics (i.e. the mean, maximum, minimum, standard deviation and maximal difference of temperature) represent the same exposure, especially not worldwide, and correcting for multiple testing can lead to conservative results when exposures are highly correlated. The ISAAC study, given its size and worldwide coverage which spans several different climate zones, is unique in its ability to assess the associations considered here. The collection of data from both 13-14 and 6-7 year-olds using the standardized and validated ISAAC protocol and questionnaire allowed for replication of analyses. The direction of estimates was the same for all significant between-country associations with intermittent symptoms and all but one within-country associations with persistent symptoms. Given that symptom data were collected for each month, and thus the time period in days and weeks of each rhinitis episode were not available, our persistent rhinitis definition differs slightly from that proposed by the ARIA workshop group,which requires symptoms to persist for greater than four days per week and for more than four consecutive weeks (Bousquet et al. 2001). The outcome definitions used here also deviate from the one commonly used by ISAAC for allergic rhinitis (rhinoconjunctivitis), which requires a positive report of both nose symptoms and itchy eyes (Björkstén et al. 2008). Data on itchy-eyes was only collected at one time point covering anytime in the last 12 months. A sensitivity analysis which considered intermittent or persistent symptoms and itchy eyes anytime in the last 12 months as alternate outcomes yielded similar results, especially for the between-country associations. Associations for both intermittent and persistent symptoms were stronger for children who reported that their nose problems interfered with daily activities “a moderate amount” and a “lot” compared to those who answered “not at all” or “no answer provided”. This may suggest that climate not only influences the prevalence of disease but also the severity of symptoms. However, we were unable to replicate this finding using the more stringent severe symptom definition normally used by ISAAC (which identifies severe symptoms as those that interfere with daily activities “a lot”) as the centre prevalences of severe symptoms were quite 112low. As with all questionnaire-based data, recall bias is possible. Symptom reports were highest in the months before and after the month the surveys were administered. We don't anticipate that this should affect the results as the outcomes considered were not month-specific. Caution shouldbe applied in interpretation of these data as representative of world prevalence as the participating centres and countries were not randomly selected. Finally, although during the design of this analysis we attempted to utilize data from the same time period, not all datasets overlap temporally.In conclusion, several between- and within-country spatial associations between climatic factors (temperature, vapour pressure and precipitation) and the prevalence of intermittent and persistentrhinitis symptoms were observed in this cross-sectional global analysis. These results provide suggestive evidence that climate influences the prevalence of rhinitis symptoms. Although not conclusive, our results represent a first step in investigating how future climate change may affect rhinitis symptom prevalence.1136   Conclusions of dissertation6.1   Summary of research and contributionsThis dissertation brings together several large, independent data sources and demonstrates the importance and value of international collaborations. Based on analyses of these combined data, three main conclusions can be drawn that can be used to guide future research on the etiology of allergic rhinitis. First, traffic-related air pollution did not consistently increase the risk of childhood allergic rhinitis at school age in the work presented in this dissertation. This conclusion is based on the longitudinal analyses conducted in three areas in Germany (Chapter 2) and a pooled analysis of six birth cohorts from Canada and Europe (Chapter 3). This conclusion is in line with the null findings found in a Dutch birth cohort study, although a positive association with allergic rhinitis at eight years was observed among a subset of non-movers (Gehring et al. 2010), as well as with the most recent studies on sensitization at older ages (Gehring et al. 2010; Gruzieva et al. 2012, 2014). Thus, the work in Chapters 2 and 3 adds to the growing number of studies which report null findings and decreases the strength of the already weak evidence for an adverse role of traffic-related air pollution on allergic rhinitis.Nevertheless, there remains much to be learned regarding for whom, when, how and in which contexts air pollution may impact disease. For example, in both of the aforementioned chapters, traffic-related air pollution significantly increased the risk of allergic rhinitis in subsets of the study populations (one of three areas in Chapter 2 and one of six cohorts in Chapter 3). The heterogeneity of associations was examined in great detail in Chapter 2, but the results remained variable across areas despite the use of standard outcome and exposure assessment strategies andregardless of the allergic outcome examined (allergic rhinitis, asthma and aeroallergen sensitization) or the age at which air pollution concentrations were assessed (birth, six and ten years). Associations with traffic-related air pollution also did not depend on genetic variability in the GSTP1, TNF, TLR2 and TLR4 genes (Chapter 3). Given that the work presented in Chapter 3 114is the first assessment of gene-air pollution interactions for allergic rhinitis, future studies are needed to either replicate or refute the null findings reported. Additional relevant genetic susceptibility loci should also be examined for gene-air pollution interactions given that the genes explored in Chapter 3 were limited. Second, carefully designed pooled analyses of existing cohort studies can be used to identify genetic risk factors for allergic rhinitis. Two SNPs (the rs1800629 SNP in the inflammation-related TNF gene and the rs1927911 SNP in the immunity-related TLR4 gene) were identified as genetic risk factors for developing allergic rhinitis by school age (Chapter 3). Evidence for a role of the TNF rs1800629 SNP on allergic rhinitis had been previously identified in two other studies(Gentile et al. 2004; Minhas et al. 2010) but a third reported null findings (Zhu et al. 2000). Our study was the first to document this association in school-age children and our results were basedon a substantially larger sample size than used in previous studies. The association reported with the rs1927911 SNP in the TRL4 gene in this dissertation was novel, although eight other SNPs in the TLRs had been linked to the prevalence of allergic rhinitis, including one in the TLR4 gene (Senthilselvan et al. 2008). Genetic variation at the 17q21 gene locus, which has been repeatedly implicated with asthma, was also found to increase the risk of allergic rhinitis from early childhood to adolescence (Chapter 4). Associations were strongest for concomitant allergic rhinitis and asthma modeled as a combined outcome, as has been observed in a recent GWAS (Ferreira et al. 2014). The identification of SNPs as potential independent risk factors for allergicrhinitis may have important public health relevance as the studied SNPs are present in a large proportion of the population, although the associated effect estimates are small. Furthermore, as genetic variability in these regions has also been linked to asthma, these results provide evidence to support the hypothesis of shared genetic susceptibility between asthma and allergic rhinitis.Third, future changes in climate may alter allergic rhinitis prevalences. This conclusion is based on the work in Chapter 5 in which several spatial associations between climatic factors and the prevalence of both intermittent and persistent rhinitis symptoms were identified using cross-sectional data from the ISAAC study. No associations were found between intermittent and persistent rhinitis prevalences and NDVI, a measure of biomass density. Prior to this work, only 115two multi-area studies had investigated associations between climatic factors and rhinitis symptoms (Verlato et al. 2002), only one of which included children (Weiland et al. 2004b). No consistent associations between climatic factors and rhinitis symptoms were reported in either of these previous studies. The work presented in Chapter 5 is based on a large sample size with worldwide coverage and a standardized exposure and outcome assessment, and allowed for examination of both between and within country-level associations. Furthermore, it is the only multi-area study which utilized the intermittent and persistent rhinitis classifications as outcomesand which investigated potential effects of vegetation on rhinitis. Although not conclusive, the novel results reported represent a first step in investigating how climate change may affect rhinitis symptom prevalence.6.2   Methodological considerationsThe first three research chapters of this dissertation are based on individual-level analyses using birth cohort data while the last chapter is an ecological cross-sectional analysis of global rhinitis prevalence data. Although complementary, these two sections are distinct and different and their strengths and limitations are best discussed separately. 6.2.1   Traffic, Asthma and Genetics collaboration 6.2.1.1   Study designA large component of this dissertation (Chapters 2, 3 and 4) utilized data collected within six birth cohorts (or a subset of these cohorts) participating in the TAG collaboration. For all cohorts except SAGE, children were recruited near the time of birth and prospectively followed, allowing for a clear examination of the temporal sequence of events (direction of causality). Although recall bias is always possible when data are collected using questionnaires, it is less of a concern when these data are collected prospectively. Throughout this dissertation, results were often consistent between questionnaire-derived outcomes (parental reports of doctor diagnosed allergic rhinitis) and objectively assessed outcomes (aeroallergen sensitization status). An inherent weakness of a birth cohort study design is the selection bias attributable to loss-of-follow up. Although all cohorts in TAG were designed to be population-based, except for CAPPSin which only children at high risk for allergic diseases were recruited, children with parents of 116high socioeconomic status and education level were over-represented. It is thus unknown if the results reported are generalizable to individuals of low socioeconomic status. In Chapters 3 and 4, data from the six birth cohorts were pooled to maximize available statistical power. Although this effort represents one of the largest combined sample sizes of its kind for allergic rhinitis, it remains possible that some of the analyses were underpowered, especially relating to the gene-environment interactions examined in Chapter 3. Furthermore, although a clear combined allergic rhinitis definition was established for these pooled analyses, the original data used to inform this combined variable differed slightly across cohorts as did the frequency of follow-up (Table 2 and Figure 1). These deviations may have introduced systematic differences between the cohorts with respect to disease frequency. Any misclassification of the disease outcome would likely be non-differential and would have driven the results toward the null. To address this limitation, all combined and cohort-specific associations were carefully compared to assess the consistency of effects across populations. Additional phenotypes, environmental exposures and confounders were also defined as similarly as possible across cohorts (MacIntyre et al. 2013). However, because of limited data or differences in the wording and language used in each cohort's specific questionnaire, a small number of potential confounders could not be derived and were consequently not included in the models (mode of birth delivery, duration of breastfeeding, parity, use of gas stove, visible mould and pets in the home). Area-level (or neighbourhood-level) variables were also unavailable. Genetic homogeneity (further discussed in section 6.2.1.3) was not a main concern as 95.5% of the TAG population were identified as Caucasian by their parents. Although this genetic homogeneity is a strength for the genetic analyses conducted, is was thus not possible to assess whether the associations observed in the TAG population are generalizable to other ethnicities. 6.2.1.2   Air pollution exposure assessmentLong-term exposures to traffic-related air pollutants were estimated to the home address at birth of all cohort participants using previously developed and validated LUR models. These types of models are commonly used in epidemiological studies and have been shown to be temporally stable for NO2 for periods of seven to 12 years (Cesaroni et al. 2012; Eeftens et al. 2011; Wang et117al. 2013). Exposure levels at the birth address were of primary interest in this dissertation as early-life has been hypothesized to be a period of increased susceptibility (Gruzieva et al. 2012). Nevertheless, despite the fact that several factors differ between movers and non-movers, stronger air pollution effects documented for non-movers in some studies provide suggestive evidence that current exposures may also be relevant (Gehring et al. 2010). Information on moving behaviour was only available for a subset of the TAG cohorts, thus it is unknown for how much of the study population an estimate at the birth address may not represent cumulative life-time exposure. Misclassification attributable to moving was examined in detail in Chapter 2 for the two German birth cohorts. Risk estimates were very similar for air pollutants assessed to the birth, six and ten year addresses and between movers and non-movers. The relative importance of birth address versus current home address exposure effects was also examined. However, we were unable to disentangle the effects of distinct exposure periods because the air pollutant estimates were highly correlated, an obstacle also reported by others (Clark et al. 2010).The effect of exposures which occur away from the home were not considered in this dissertation(for example, on the way to or while at school). Although previous examinations of this issue suggest that air pollution exposures at schools may not be that different from those at home, and thus exposure misclassification is likely to be low (Gruzieva et al. 2012; Reungoat et al. 2005; Ryan et al. 2008), the lack of these exposure data remains a limitation of this dissertation. Additionally, exposure levels during the pregnancy period or indoors were not available nor weredata on time spent outdoors near the home. 6.2.1.3   Genetic dataWith the exception of the SAGE cohort, none of the other cohorts included in the genetic association analyses of this dissertation were originally designed with the primary aim of studying genetic effects on childhood allergic diseases. Hence, certain data which might have been informative, such as parental DNA, were unavailable in some cohorts. Additionally, although the vast majority of participants were identified as Caucasian (95.5%), the information on which this percentage is based is vague. For example, only children whose parents could speak German were recruited into the two German cohorts, and consequently, all participants 118were assumed to be Caucasian. We are thus unable to completely exclude the possibility that any observed genetic associations may be due to population stratification (that is, cases and controls have systematically different allele frequencies which are irrelevant for disease development). Which SNPs to examine is a vital question in all candidate gene studies. Throughout this dissertation, the selection of SNPs for study was based on prior knowledge and data availability. In Chapter 3, ten SNPs in genes involved in inflammation, oxidative stress metabolism or immunity development, processes which may mediate gene-air pollution interactions, that were available in at least three of the six participating cohorts were selected. As only a limited number of SNPs were investigated, the work in Chapter 3 does not rule out the existence of individuals genetically susceptible to potential adverse effects of air pollution on allergic rhinitis. In Chapter 4, seven SNPs at the 17q21 asthma-risk locus which were available for the largest subset of the TAG population were selected. Although these SNPs included top hits from previous GWASs and key studies on asthma, it remains unknown which or if any of the genetic variants investigated are causal variants. As we chose to follow a hypothesis-driven approach throughout this dissertation and thus selected a limited number of SNPs to study, a stringent conservative adjustment for multiple testing was not as crucial as it is for hypothesis-free GWASs.6.2.2   International Study of Asthma and Allergies in Childhood 6.2.2.1   Study designThe last research chapter of this dissertation (Chapter 5) utilized cross-sectional data collected bystandardized and validated questionnaires completed by parents (for the 6-7 year-old children) and children (for the 13-14 year-old teenagers) in ISAAC Phase Three. Evidence of recall bias was suggested as monthly symptom reports were highest in the months before and after the month the surveys were administered. However, this is not expected to have affected the results as the outcomes analyzed were defined based on symptom duration and were thus not month-specific. As data were collected for two separate groups of children using the same questionnaire,all analyses could be replicated and the direction of effect estimates were consistent for many of the associations considered. Additionally, both centre- and country-level associations could be reported using a multilevel modelling approach. The global scope of the ISAAC study and its useof consistent methodology across all centres lends credit to its generalizability. However, the 119world prevalences reported should be interpreted with caution as the participating centres and countries were not randomly selected. It was not possible to comment on the direction of causality of the observed associations as the data collected were cross-sectional. Given the ecological design of the study, it is possible the associations observed at the group level do not necessarily hold true for a single individual chosen from that group (the ecological fallacy (Robinson 1950)). Indeed, residual confounding is an important concern. It is likely that substantial between-country differences were not captured by model adjustments for GNI per capita, climate type or air pollution, and thus the between-country associations reported may have been affected by residual confounding. Although still possible, confounding is likely less ofa concern for the within-country associations as causal factors are less likely to differ to the samedegree within countries as between countries. However, individual-level risk factors could not beaccounted for, as is common to all ecological studies. Nevertheless, despite these limitations, ecological analyses are vital in generating hypotheses that can subsequently be tested using observational or intervention studies. 6.2.2.2   Environmental assessmentDate on temperature, precipitation, vapour pressure and NDVI were mapped to geographical coordinates and used as exposure estimates. Climate factors were mapped to the 0.1o x 0.1o square with the highest population density surrounding a given centre. NDVI estimates were calculated as the average of this square and the eight surrounding squares, as vegetation may be more heterogeneous over small areas than climate. It is unknown whether these average exposureestimates may have masked more complicated relationships with disease at an individual level. Furthermore, although this study is the first to consider global associations between vegetation and rhinitis prevalence, the use of NDVI as a surrogate for vegetation might not have been optimal as NDVI estimates do not allow plant species of differing allergenicity to be distinguished. 6.3   Recommendations for future researchOne of the main challenges for allergic diseases is understanding their complexity. Allergic rhinitis with and without comorbidities represents a variety of conditions with unique and 120common risk factors. Efforts to establish standard and comparable phenotypes are well warrantedand underway (for example, by the Allergic Rhinitis and its Impact on Asthma (Bousquet et al. 2012) and European Mechanisms of the Development of Allergy (Bousquet et al. 2011) collaborations). Future studies should thus not only aim to replicate the findings of this dissertation using standard definitions of allergic rhinitis, but also should consider phenotypes that incorporate age of development, comorbidities and severity in order to further inform our understanding of this dynamic disease. Further recommendations are given below relating to the three primary risk factor themes discussed in this dissertation: traffic-related air pollution, genetic variability and climate. During the development of this dissertation, several large observational studies (including those in Chapters 2 and 3) have examined the adverse effects of traffic-related air pollution on respiratory health. With regard to allergic rhinitis diagnoses and symptoms, the evidence supporting a true association has decreased, despite the use of standardized, homogenous phenotypes and exposure assessment methods. Although the limitations discussed in section 6.2.1 may in part account for some of the null findings, more detailed and comprehensive assessments may be required. For example, the Risk of Airborne Particles: a Toxicological-Epidemiological Hybrid Study and Transport-Related Air Pollution and Health Impacts - Integrated Methodologies for Assessing Particulate Matter project (www.transphorm.eu) have begun to examine the relative toxicity of particulate matter components, which are known to vary in composition, size and oxidative potential, in relation to health outcomes (Eeftens et al. 2014; Fuertes et al. 2014; Strak et al. 2012; Wang et al. 2014). Such efforts may also be worthwhile for allergic rhinitis. Additionally, as the birth cohort participants continue to be followed-up into adolescence, the influence of exposures that occur away from the home may increase in importance, as will personal behaviours (for example, smoking). Future studies on airpollution should consider these additional exposures and should also aim to incorporate area-level (or neighbourhood-level) effects on health, which until recently (for example, (Gehring et al. 2013; Gruzieva et al. 2014)), were rarely considered in studies on air pollution and allergies. 121Genetic association studies are unlikely to have reached the expectations of the scientific community. Much missing heritability remains unexplained, possibly because of the under-investigated influences of genetic variations other than SNPs, such as epigenetics, copy number variations, gene-gene interactions, gene-environment interactions (Manolio et al. 2009) and epigenetic-environment interactions (Breton and Marutani 2014). Additional studies that examine this complex network of genetic and environmental influences are needed. For example,at the time of publication of this thesis, the work conducted in Chapter 2 remained the only published epidemiological (observational) gene-air pollution interaction study on allergic rhinitis, although smaller epidemiological (Melén et al. 2008) and experimental studies (Gillilandet al. 2004, 2006) on allergic sensitization exist. Gene-environment interaction studies are important as they can help elucidate possible mechanisms and can also help clarify the heterogeneity in effect estimates observed for environmental risk factors. For example, despite the fact that no significant main effect was observed for traffic-related air pollution in the pooled analyses of the six birth cohorts in Chapter 3 of this dissertation, it was necessary to examine potential gene-environment interactions to determine whether traffic-related air pollution may only be causing an adverse effect among a genetically susceptible subset of the population. As the field of genetic epidemiology continues to evolve and new studies are reported, it is imperative that their results be replicated in different populations and environments. More than 100 SNPs have been associated with allergic rhinitis, but very few have been replicated or are understood. Systematic evaluations, as was conducted for allergic rhinitis by Nilsson et al. (2013), combined with the use of bioinformatic tools will direct researchers to the most important SNPs or genetic regions and should be encouraged. Functional studies can then assess the biological plausibility of these “hits” and their capacity to cause disease or increase an individual's susceptibility to other risk factors. Linking results from epidemiological studies with those from functional and experimental studies is of vital importance in order to fully understand, in terms of the biology and pathophysiology, what stands behind any observed statistical association. Although not a requirement for taking action, an enhanced mechanistic understanding of how traffic-related air pollution may lead to or exacerbate disease is key to identifying safe exposures for populations (Carlsten et al. 2014) and to developing effective 122preventative strategies. For example, identifying which aspect(s) of air pollution are most detrimental or which populations are most vulnerable would help prioritize control strategies. The Earth's climate is changing, but to what extent and how this will affect allergic disease prevalences is poorly understood. Although predictions are based on models and thus may be imprecise, the quantification of potential effect estimates is crucial for health impact assessmentswhich support the health care community and decision making processes. The work presented in Chapter 5, although not conclusive, represents a first step at investigating how climate change may affect rhinitis symptom prevalence. Future studies should continue this investigation, but should also aim to identify and consider potential vulnerable subgroups, such as children or the elderly, those with pre-existing respiratory diseases or comorbidities and those of lower socioeconomic status who may be less able to adapt to climate change. Additionally, climate effects on respiratory health that may develop throughout pathways other than via changes in aeroallergen growth patterns and characteristics should be investigated. For example, wildfires and dust storms can dramatically adversely affect air quality and severe thunderstorms, which areprojected to increase in frequency, can cause severe bronchoconstriction among individuals sensitized to pollen, even among those with no history of asthma (D’Amato et al. 2010). There is an increasing need to consider the interacting effects of climate, air pollution and aeroallergens on allergic rhinitis. The mechanisms regulating these multidimensional interactions, and their health consequences, are largely unknown. Micro-environments resembling future climate change scenarios already exist in urban centres and should continue to be investigated. As climate-induced changes are expected to vary by region, studies that vary in vegetation, geography, climate and air pollution levels and sources will be required to unravel observed associations. Although current risk estimates associated with the environmental factors considered in this dissertation appear uncertain and small, these factors are significant public health concerns as a very large number of people are potentially at risk, air pollution concentrations are unlikely to decrease soon and climate change will continue in the coming decades. Indeed, although the 123evidence for an association between traffic-related air pollution and allergic rhinitis is weak, it is much stronger for several other health outcomes, such as cardiovascular disease and mortality, asthma, adverse pregnancy outcomes and lung cancer. Consequently, mitigation strategies that reduce population exposure to traffic-related air pollution, such as land-use planning and transportation management, reduction of vehicle emissions, modification of existing structures and encouraging behaviour change, as previously reviewed (Brauer et al. 2012), are well warranted and likely to have co-benefits. For example, strategies that promote active transportation or the use of public transportation are likely to increase physical fitness and reducetraffic-related air pollution, the latter of which has both health and climate-change associated benefits (Younger et al. 2008). Similarly, climate change policies are likely to positively impact air quality. An evaluation of climate policies on air pollution in Europe concluded that co-benefits exist for all investigated emission scenarios (Colette et al. 2012). Regardless of the future studies undertaken, the way forward will require interdisciplinary and international collaboration. Epidemiologists, climatologists, respiratory and allergy specialists, geneticists, policy makers and public health professionals need to be involved in 1) designing future studies that adequately address unanswered questions and 2) translating study results into action via the development and implementation of effective public health policies. 6.4   SummaryThe purpose of this dissertation was to examine the influence of environmental (traffic-related airpollution and climate) and genetic risk factors on the development of allergic rhinitis during childhood. Epidemiological analyses of data from several birth cohorts did not identify traffic-related air pollution as a consistent risk factor for allergic rhinitis; this null effect was not modified by genetic variability in the GSTP1, TNF, TLR2 and TLR4 genes. However, genetic variability in the TNF and TLR4 genes and at the 17q21 gene locus was independently associatedwith childhood allergic rhinitis. Finally, using global cross-sectional data, several spatial associations between climatic factors and the prevalence of both intermittent and persistent rhinitis symptoms were identified, providing a first step in the investigation of the potential effects of future climate change on allergic rhinitis. 124ReferencesAcevedo N, Reinius L, Gref A, Greco D, Orsmark Pietras C, Pershagen G, et al. 2013a. Decreased DNA methylation in the ORMDL3 promoter of children with asthma and severe asthma [abstract]. Allergy 68(Suppl. 97): 637–707.Acevedo N, Sääf A, Söderhäll C, Melén E, Mandelin J, Pietras CO, et al. 2013b. Interaction between Retinoid Acid Receptor-Related Orphan Receptor Alpha (RORA) and Neuropeptide S Receptor 1 (NPSR1) in asthma. PLoS ONE 8(4):e60111; doi:10.1371/journal.pone.0060111.Aït Khale‐ d N, Pearce N, Anderson H, Ellwood P, Montefort S, Shah J, et al. 2009. Global map ofthe prevalence of symptoms of rhinoconjunctivitis in children: The International Study of Asthma and Allergies in Childhood (ISAAC) Phase Three. Allergy 64(1):123–148; doi:10.1111/j.1398-9995.2008.01884.x.Allen RW, Amram O, Wheeler AJ, Brauer M. 2011. The transferability of NO and NO2 land use regression models between cities and pollutants. Atmos. Environ. 45(2):369–378; doi:10.1016/j.atmosenv.2010.10.002.Almqvist C, Pershagen G, Wickman M. 2005. Low socioeconomic status as a risk factor for asthma, rhinitis and sensitization at 4 years in a birth cohort. Clin. Exp. Allergy 35(5):612–618; doi:10.1111/j.1365-2222.2005.02243.x.Anderson H, Butland B, van Donkelaar A, Brauer M, Strachan D, Clayton T, et al. 2012. Satellite-based estimates of ambient air pollution and global variations in childhood asthma prevalence. Environ. Health Perspect. 120(9):1333–1339; doi:10.1289/ehp.1104724.Anderson H, Ruggles R, Pandey K, Kapetenakis V, Brunekreef B, Lai CKW, et al. 2009. Ambient particulate pollution and the world-wide prevalence of asthma, rhinoconjunctivitis and eczema in children: Phase One of the International Study of Asthma. Occup. Environ. Med. 67:293–300; doi:10.1136/oem.2009.048785.Anderson HR, Favarato G, Atkinson RW. 2013. Long-term exposure to air pollution and the incidence of asthma: meta-analysis of cohort studies. Air Qual. Atmosphere Health 6(1):47–56; doi:10.1007/s11869-011-0144-5.Andiappan AK, Wang DY, Anantharaman R, Parate PN, Suri BK, Low HQ, et al. 2011. Genome-wide association study for atopy and allergic rhinitis in a Singapore Chinese population. PLoS ONE 6(5):e19719; doi:10.1371/journal.pone.0019719.Ariano R, Canonica GW, Passalacqua G. 2010. Possible role of climate changes in variations in pollen seasons and allergic sensitizations during 27 years. Ann. Allergy. Asthma. Immunol. 104(3):215–222; doi:10.1016/j.anai.2009.12.005.125Asher M, Montefort S, Björkstén B, Lai C, Strachan DP, Weiland SK, et al. 2006. 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. The Lancet 368(9537):733–743; doi:10.1016/S0140-6736(06)69283-0.Balantic M, Rijavec M, Fležar M, Camlek T, Hudoklin I, Košnik M, et al. 2013. A polymorphismin ORMDL3 is associated not only with asthma without rhinitis but also with chronic obstructive pulmonary disease. J. Investig. Allergol. Clin. Immunol. 23(4): 256–261.Ballardini N, Kull I, Lind T, Hallner E, Almqvist C, Östblom E, et al. 2012. Development and comorbidity of eczema, asthma and rhinitis to age 12 – data from the BAMSE birth cohort. Allergy 67(4):537–544; doi:10.1111/j.1398-9995.2012.02786.x.Barnes CS, Alexis NE, Bernstein JA, Cohn JR, Demain JG, Horner E, et al. 2013. Climate change and our environment: the effect on respiratory and allergic disease. J. Allergy Clin. Immunol. Pract. 1(2):137–141; doi:10.1016/j.jaip.2012.07.002.Bates D, Maechler M, Bolker B, Walker S. 2014. Lme4: Linear mixed-effects models using Eigen and S4. R package version 1.1-6. www.CRAN.R-project.org/package=lme4.Beck I, Jochner S, Gilles S, McIntyre M, Buters JTM, Schmidt-Weber C, et al. 2013. High environmental ozone levels lead to enhanced allergenicity of birch pollen. PLoS ONE 8(11):e80147; doi:10.1371/journal.pone.0080147.Beelen R, Hoek G, Pebesma E, Vienneau D, de Hoogh K, Briggs DJ. 2009. Mapping of background air pollution at a fine spatial scale across the European Union. Sci. Total Environ. 407(6):1852–1867; doi:10.1016/j.scitotenv.2008.11.048.Beelen R, Hoek G, Vienneau D, Eeftens M, Dimakopoulou K, Pedeli X, et al. 2013. 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. 72:10–23; doi:10.1016/j.atmosenv.2013.02.037.Begg MD, Parides MK. 2003. Separation of individual-level and cluster-level covariate effects inregression analysis of correlated data. Stat. Med. 22(16):2591–2602; doi:10.1002/sim.1524.Beggs PJ. 2004. Impacts of climate change on aeroallergens: past and future. Clin. Exp. Allergy 34(10):1507–1513; doi:10.1111/j.1365-2222.2004.02061.x.Beijsterveldt CEM van, Boomsma DI. 2007. Genetics of parentally reported asthma, eczema andrhinitis in 5-yr-old twins. Eur. Respir. J. 29(3):516–521; doi:10.1183/09031936.00065706.Benjamini Y, Hochberg Y. 1995. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B Methodol. 57(1): 289–300.126Berlivet S, Moussette S, Ouimet M, Verlaan DJ, Koka V, Tuwaijri AA, et al. 2012. Interaction between genetic and epigenetic variation defines gene expression patterns at the asthma-associated locus 17q12-q21 in lymphoblastoid cell lines. Hum. Genet. 131(7):1161–1171;doi:10.1007/s00439-012-1142-x.Bhattacharyya N. 2009. Does annual temperature influence the prevalence of otolaryngologic respiratory diseases? The Laryngoscope 119(10):1882–1886; doi:10.1002/lary.20613.Bieli C, Eder W, Frei R, Braun-Fahrländer C, Klimecki W, Waser M, et al. 2007. A polymorphism in CD14 modifies the effect of farm milk consumption on allergic diseasesand CD14 gene expression. J. Allergy Clin. Immunol. 120(6):1308–1315; doi:10.1016/j.jaci.2007.07.034.Bisgaard H, Bønnelykke K, Sleiman PMA, Brasholt M, Chawes B, Kreiner-Møller E, et al. 2009. Chromosome 17q21 gene variants are associated with asthma and exacerbations but not atopy in early childhood. Am. J. Respir. Crit. Care Med. 179(3):179–185; doi:10.1164/rccm.200809-1436OC.Björkstén B, Clayton T, Ellwood P, Strachan D, ISAAC Phase Three Study Group. 2008. Worldwide time trends for symptoms of rhinitis and conjunctivitis: Phase III of the International Study of Asthma and Allergies in Childhood. Pediatr. Allergy Immunol. 19(2):110–124; doi:10.1111/j.1399-3038.2007.00601.x.Bønnelykke K, Matheson MC, Pers TH, Granell R, Strachan DP, Alves AC, et al. 2013. Meta-analysis of genome-wide association studies identifies ten loci influencing allergic sensitization. Nat. Genet. 45(8):902–906; doi:10.1038/ng.2694.Bottema RWB, Reijmerink NE, Kerkhof M, Koppelman GH, Stelma FF, Gerritsen J, et al. 2008. Interleukin 13, CD14, pet and tobacco smoke influence atopy in three Dutch cohorts: the allergenic study. Eur. Respir. J. 32(3):593–602; doi:10.1183/09031936.00162407.Bousquet J, Annesi-Maesano I, Carat F, Léger D, Rugina M, Pribil C, et al. 2005. Characteristics of intermittent and persistent allergic rhinitis: DREAMS study group. Clin. Exp. Allergy 35(6):728–732; doi:10.1111/j.1365-2222.2005.02274.x.Bousquet J, Anto J, Auffray C, Akdis M, Cambon Thomsen A, Keil T, et al. 2011. MeDALL ‐(Mechanisms of the Development of ALLergy): an integrated approach from phenotypes to systems medicine. Allergy 66(5):596–604; doi:10.1111/j.1398-9995.2010.02534.x.Bousquet J, Cauwenberge P, Khaltaev N. 2001. Allergic rhinitis and its impact on asthma. J. Allergy Clin. Immunol. 108(Suppl. 5):S147–S334; doi:10.1067/mai.2001.118891.Bousquet J, Khaltaev N, Cruz AA, Denburg J, Fokkens WJ, Togias A, et al. 2008. Allergic rhinitis and its impact on asthma (ARIA). Allergy 63(Suppl. 86):8–160; doi:10.1111/j.1398-9995.2007.01620.x.127Bousquet J, Schünemann HJ, Samolinski B, Demoly P, Baena-Cagnani CE, Bachert C, et al. 2012. Allergic rhinitis and its impact on asthma (ARIA): achievements in 10 years and future needs. J. Allergy Clin. Immunol. 130(5):1049–1062; doi:10.1016/j.jaci.2012.07.053.Bouzigon E, Corda E, Aschard H, Dizier M-H, Boland A, Bousquet J, et al. 2008. Effect of 17q21 variants and smoking exposure in early-onset asthma. N. Engl. J. Med. 359(19):1985–1994; doi:10.1056/NEJMoa0806604.Braback L, Forsberg B. 2009. Does traffic exhaust contribute to the development of asthma and allergic sensitization in children: findings from recent cohort studies. Environ. Health 8(1):17; doi:10.1186/1476-069X-8-17.Brauer M, Hoek G, Smit HA, de Jongste JC, Gerritsen J, Postma DS, et al. 2007. Air pollution and development of asthma, allergy and infections in a birth cohort. Eur. Respir. J. 29(5):879 –888; doi:10.1183/09031936.00083406.Brauer M, Hoek G, Van Vliet P, Meliefste K, Fischer PH, Wijga A, et al. 2002. Air pollution from traffic and the development of respiratory infections and asthmatic and allergic symptoms in children. Am. J. Respir. Crit. Care Med. 166(8):1092–1098; doi:10.1164/rccm.200108-007OC.Brauer M, Hoek G, Vliet PV, Meliefste K, Fischer P, Gehring U, et al. 2003. Estimating long-term average particulate air pollution concentrations: application of traffic indicators and geographic information systems. Epidemiology 14(2):228–239; doi:10.1097/01.EDE.0000041910.49046.9B.Brauer M, Reynolds C, Hystad P. 2012. Traffic-related air pollution and health : a Canadian perspective on scientific evidence and potential exposure-mitigation strategies.Breslow DK, Collins SR, Bodenmiller B, Aebersold R, Simons K, Shevchenko A, et al. 2010. Orm family proteins mediate sphingolipid homeostasis. Nature 463(7284):1048–1053; doi:10.1038/nature08787.Breton CV, Marutani AN. 2014. Air pollution and epigenetics: recent findings. Curr. Environ. Health Rep. 1(1):35–45; doi:10.1007/s40572-013-0001-9.Breton M-C, Garneau M, Fortier I, Guay F, Louis J. 2006. Relationship between climate, pollen concentrations of Ambrosia and medical consultations for allergic rhinitis in Montreal, 1994–2002. Sci. Total Environ. 370(1):39–50; doi:10.1016/j.scitotenv.2006.05.022.Briggs DJ, de Hoogh C, Gulliver J, Wills J, Elliott P, Kingham S, et al. 2000. A regression-based method for mapping traffic-related air pollution: application and testing in four contrasting urban environments. Sci. Total Environ. 253(1):151–167; doi:10.1016/S0048-9697(00)00429-0.128Brinkman, B, Zuijdeest D, Kaijzel E, Breedveld F, Verweij C. 1994. Relevance of the tumor necrosis factor alpha (TNF alpha) -308 promoter polymorphism in TNF alpha gene regulation. J. Inflamm. 46(1): 32–41.Brunekreef B, Smit J, de Jongste J, Neijens H, Gerritsen J, Postma D, et al. 2002. The preventionand incidence of asthma and mite allergy (PIAMA) birth cohort study: design and first results. Pediatr. Allergy Immunol. 13(15):55–60; doi:10.1034/j.1399-3038.13.s.15.1.x.Brunekreef B, Sunyer J. 2003. Asthma, rhinitis and air pollution: is traffic to blame? Eur. Respir. J. 21(6):913–915; doi:10.1083/09031936.03.00014903.Çalışkan M, Bochkov YA, Kreiner-Møller E, Bønnelykke K, Stein MM, Du G, et al. 2013. Rhinovirus wheezing illness and genetic risk of childhood-onset asthma. N. Engl. J. Med.368(15):1398–1407; doi:10.1056/NEJMoa1211592.Cantero-Recasens G, Fandos C, Rubio-Moscardo F, Valverde MA, Vicente R. 2010. The asthma-associated ORMDL3 gene product regulates endoplasmic reticulum-mediated calcium signaling and cellular stress. Hum. Mol. Genet. 19(1):111–121; doi:10.1093/hmg/ddp471.Carlsten C, Brauer M, Brinkman F, Brook J, Daley D, McNagny K, et al. 2014. Genes, the environment and personalized medicine. EMBO Rep. 15(7);736-739; doi:10.15252/embr.201438480.Carlsten C, Dybuncio A, Becker A, Chan-Yeung M, Brauer M. 2011. Traffic-related air pollutionand incident asthma in a high-risk birth cohort. Occup. Environ. Med. 68(4):291–295; doi:10.1136/oem.2010.055152.Carlsten C, Melén E. 2012. Air pollution, genetics, and allergy. Curr. Opin. Allergy Clin. Immunol. 12(5):455–461; doi:10.1097/ACI.0b013e328357cc55.Carreras-Sureda A, Cantero-Recasens G, Rubio-Moscardo F, Kiefer K, Peinelt C, Niemeyer BA, et al. 2013. ORMDL3 modulates store-operated calcium entry and lymphocyte activation.Hum. Mol. Genet. 22(3):519–530; doi:10.1093/hmg/dds450.Central Intelligence Agency. 2007. The World Factbook 2003. Washington, DC: Central Intelligence Agency. Available: https://www.cia.gov/library/publications/download/download-2003/index.html [accessed:March 13, 2012].Cesaroni G, Porta D, Badaloni C, Stafoggia M, Eeftens M, Meliefste K, et al. 2012. Nitrogen dioxide levels estimated from land use regression models several years apart and association with mortality in a large cohort study. Environ. Health 11:48–57; doi:10.1186/1476-069X-11-48.129Chan-Yeung M, Manfreda J, Dimich-Ward H, Ferguson A, Watson W, Becker A. 2000. A randomized controlled study on the effectiveness of a multifaceted intervention program in the primary prevention of asthma in high-risk infants. Arch. Pediatr. Adolesc. Med. 154(7):657–663; doi:10.1001/archpedi.154.7.657.Chen LL, Tager IB, Peden DB, Christian DL, Ferrando RE, Welch BS, et al. 2004. EFfect of ozone exposure on airway responses to inhaled allergen in asthmatic subjects*. CHEST 125(6):2328–2335; doi:10.1378/chest.125.6.2328.Clark NA, Demers PA, Karr CJ, Koehoorn M, Lencar C, Tamburic L, et al. 2010. Effect of early life exposure to air pollution on development of childhood asthma. Environ. Health Perspect. 118(2):284–290; doi:10.1289/ehp.0900916.Colette A, Koelemeijer R, Mellios G, Schucht S, Péré J-C, Kouridis C, et al. 2012. Cobenefits of climate and air pollution regulations. The context of the European Commission Roadmap for moving to a low carbon economy in 2050. Technical Paper of the European Topic Centre on Air Pollution and Climate Change Mitigation (ETC/ACM).Connell JT. 1969. Quantitative intranasal pollen challenges: III. The priming effect in allergic rhinitis. J. Allergy 43(1):33–44; doi:10.1016/0021-8707(69)90018-5.Cordell H., Clayton D. 2005. Genetic epidemiology 3. Genetic association studies. The Lancet 366(9491):1121–1131; doi:10.1016/S0140-6736(05)67424-7.Custovic A, Simpson BM, Murray CS, Lowe L, Woodcock A, on behalf of the NAC Manchester Asthma and Allergy Study Group. 2002. The national asthma campaign Manchester Asthma and Allergy Study. Pediatr. Allergy Immunol. 13(s15):32–37; doi:10.1034/j.1399-3038.13.s.15.3.x.Cyrys J, Eeftens M, Heinrich J, Ampe C, Armengaud A, Beelen R, et al. 2012. Variation of NO2 and NOx concentrations between and within 36 European study areas: results from the ESCAPE study. Atmos. Environ. 62:374–390; doi:10.1016/j.atmosenv.2012.07.080.Cyrys J, Heinrich J, Hoek G, Meliefste K, Lewné M, Gehring U, et al. 2003. Comparison between different traffic-related particle indicators: elemental carbon (EC), PM2.5 mass, and absorbance. J. Expo. Sci. Environ. Epidemiol. 13(2):134–143; doi:10.1038/sj.jea.7500262.D’Amato G, Cecchi L. 2008. Effects of climate change on environmental factors in respiratory allergic diseases. Clin. Exp. Allergy 38(8):1264–1274; doi:10.1111/j.1365-2222.2008.03033.x.D’Amato G, Cecchi L, Bonini S, Nunes C, Annesi-Maesano I, Behrendt H, et al. 2007. Allergenic pollen and pollen allergy in Europe. Allergy 62(9):976–990; doi:10.1111/j.1398-9995.2007.01393.x.130D’Amato G, Liccardi G, Cecchi L, D’Amato M. 2010. Urban air pollution and climate change asenvironmental risk factors of respiratory allergy: An update. J. Investig. Allergol. Clin. Immunol. 20(2): 95–102.D’Amato PG, Bergmann KC, Cecchi L, Annesi-Maesano I, Sanduzzi A, Liccardi G, et al. 2014. Climate change and air pollution. Allergo J. Int. 23(1):17–23; doi:10.1007/s40629-014-0003-7.De Marco R, Poli A, Ferrari M, Accordini S, Giammanco G, Bugiani M, et al. 2002. The impact of climate and traffic-related NO2 on the prevalence of asthma and allergic rhinitis in Italy. Clin. Exp. Allergy 32(10):1405–1412; doi:10.1046/j.1365-2745.2002.01466.x.Delaneau O, Marchini J, Zagury J-F. 2012. A linear complexity phasing method for thousands of genomes. Nat. Methods 9(2):179–181; doi:10.1038/nmeth.1785.Delaneau O, Zagury J-F, Marchini J. 2013. Improved whole-chromosome phasing for disease and population genetic studies. Nat Meth 10(1):5–6; doi:10.1038/nmeth.2307.Demoly P, Allaert F-A, Lecasble M, Bousquet J, PRAGMA. 2003. Validation of the classification of ARIA (allergic rhinitis and its impact on asthma). Allergy. 58(7):672-675; doi:10.1034/j.1398-9995.2003.t01-1-00202.x. Eder W, Klimecki W, Yu L, von Mutius E, Riedler J, Braun-Fahrländer C, et al. 2004. Toll-like receptor 2 as a major gene for asthma in children of European farmers. J. Allergy Clin. Immunol. 113(3):482–488; doi:10.1016/j.jaci.2003.12.374.Eeftens M, Beelen R, de Hoogh K, Bellander T, Cesaroni G, Cirach M, et al. 2012a. Development of land use regression models for PM2.5, PM2.5 absorbance, PM10 and PM coarse in 20 European study areas; results of the ESCAPE project. Environ. Sci. Technol. 46(20):11195–11205; doi:10.1021/es301948k.Eeftens M, Beelen R, Fischer P, Brunekreef B, Meliefste K, Hoek G. 2011. Stability of measuredand modelled spatial contrasts in NO2 over time. Occup. Environ. Med. 68(10):765–770; doi:10.1136/oem.2010.061135.Eeftens M, Hoek G, Gruzieva O, Mölter A, Agius R, Beelen R, et al. 2014. Elemental composition of particulate matter and the association with lung function in 5 European birth cohorts – results of the ESCAPE & TRANSPHORM projects. Epidemiology. 25(5):648-657; doi:10.1097/EDE.0000000000000136. Eeftens M, Tsai M-Y, Ampe C, Anwander B, Beelen R, Bellander T, et al. 2012b. Spatial variation of PM2.5, PM10, PM2.5 absorbance and PM coarse concentrations between and within 20 European study areas and the relationship with NO2 – results of the ESCAPE project. Atmos. Environ. 62:303–317; doi:10.1016/j.atmosenv.2012.08.038.131Ellwood P, Asher M, Beasley R, Clayton TO, Stewart AW, ISAAC Steering Committee. 2005. The International Study of Asthma and Allergies in Childhood (ISAAC): Phase Three rationale and methods. Int. J. Tuberc. Lung Dis. 9(1): 10–16.Emenius G, Pershagen G, Berglind N, Kwon H-J, Lewne M, Nordvall SL, et al. 2003. NO2, as a marker of air pollution, and recurrent wheezing in children: a nested case-control study within the BAMSE birth cohort. Occup. Environ. Med. 60(11):876–881; doi:10.1136/oem.60.11.876.Esparza-Gordillo J, Marenholz I, Lee Y. 2010. Genome-wide approaches to the etiology of eczema. Curr. Opin. Allergy Clin. Immunol. 10(5):418–426; doi:10.1097/ACI.0b013e32833dc75e.Faul F, Erdfelder E, Buchner A, Lang A-G. 2009. Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behav. Res. Methods 41(4):1149–1160; doi:10.3758/BRM.41.4.1149.Ferreira MAR, Matheson MC, Tang CS, Granell R, Ang W, Hui J, et al. 2014. Genome-wide association analysis identifies 11 risk variants associated with the asthma with hay fever phenotype. J. Allergy Clin. Immunol. 133(6):1564–1571; doi:10.1016/j.jaci.2013.10.030.Filipiak B, Zutavern A, Koletzko S, von Berg A, Brockow I, Grübl A, et al. 2007. Solid food introduction in relation to eczema: results from a four-year prospective birth cohort study.J. Pediatr. 151(4):352–358; doi:10.1016/j.jpeds.2007.05.018.Fokkens WJ. 2002. Thoughts on the pathophysiology of nonallergic rhinitis. Curr. Allergy Asthma Rep. 2(3):203–209; doi:10.1007/s11882-002-0020-3.Fuertes E, Brauer M, MacIntyre E, Bauer M, Bellander T, von Berg A, et al. 2013. Childhood allergic rhinitis, traffic-related air pollution, and variability in the GSTP1, TNF, TLR2, and TLR4 genes: Results from the TAG Study. J. Allergy Clin. Immunol. 132(2):342–352.e2; doi:10.1016/j.jaci.2013.03.007.Fuertes E, MacIntyre E, Agius R, Beelen R, Brunekreef B, Bucci S, et al. 2014. Associations between particulate matter elements and early-life pneumonia in seven birth cohorts: Results from the ESCAPE and TRANSPHORM projects. Int. J. Hyg. Environ. Health. 217(8):819-829; doi:10.1016/j.ijheh.2014.05.004.Galanter J, Choudhry S, Eng C, Nazario S, Rodriguez-Santana JR, Casal J, et al. 2008. ORMDL3 gene is associated with asthma in three ethnically diverse populations. Am J Respir Crit Care Med 177(11):1194–1200; doi:10.1164/rccm.200711-1644OC.Gao Z, Rennie D, Senthilselvan A. 2010. Allergic rhinitis and genetic components: focus on toll-like receptors (TLRs) gene polymorphism. Appl. Clin. Genet. 3:109–120; doi:10.2147/TACG.S8380.132Gehring U, Cyrys J, Sedlmeir G, Brunekreef B, Bellander T, Fischer P, et al. 2002. Traffic-related air pollution and respiratory health during the first 2 yrs of life. Eur. Respir. J. 19(4):690 –698; doi:10.1183/09031936.02.01182001.Gehring U, Gruzieva O, Agius RM, Beelen R, Custovic A, Cyrys J, et al. 2013. Air pollution exposure and lung function in children: the ESCAPE project. Environ. Health Perspect. 121(11-12):1357–1364; doi:10.1289/ehp.1306770.Gehring U, Wijga A, Brauer M, Fischer P, de Jongste JC, Kerkhof M, et al. 2010. Traffic-related air pollution and the development of asthma and allergies during the first 8 years of life. Am. J. Respir. Crit. Care Med. 181(6):596–603; doi:10.1164/rccm.200906-0858OC.Gentile DA, Doyle WJ, Zeevi A, Howe-Adams J, Trecki J, Skoner DP. 2004. Association between TNF-α and TGF-β genotypes in infants and parental history of allergic rhinitis and asthma. Hum. Immunol. 65(4):347–351; doi:10.1016/j.humimm.2004.01.014.Gerbase MW, Keidel D, Imboden M, Gemperli A, Bircher A, Schmid Grendelmeier P, et al. ‐2011. Effect modification of immunoglobulin E mediated atopy and rhinitis by ‐glutathione S transferase genotypes in passive smokers. Clin. Exp. Allergy 41(11):1579–‐1586; doi:10.1111/j.1365-2222.2011.03807.x.Ghiani A, Aina R, Asero R, Bellotto E, Citterio S. 2012. Ragweed pollen collected along high-traffic roads shows a higher allergenicity than pollen sampled in vegetated areas. Allergy 67(7):887–894; doi:10.1111/j.1398-9995.2012.02846.x.Gibbs RA, Belmont JW, Hardenbol P, Willis TD, Yu F. 2003. The International HapMap Project. Nature 426(6968):789–796; doi:10.1038/nature02168.Gilliland F, Li Y-F, Gong H, Diaz-Sanchez D. 2006. Glutathione S-transferases M1 and P1 prevent aggravation of allergic responses by secondhand smoke. Am. J. Respir. Crit. CareMed. 174(12):1335 –1341; doi:10.1164/rccm.200509-1424OC.Gilliland F, Li Y-F, Saxon A, Diaz-Sanchez D. 2004. Effect of glutathione-S-transferase M1 and P1 genotypes on xenobiotic enhancement of allergic responses: randomised, placebo-controlled crossover study. Lancet 363(9403):119–125; doi:10.1016/S0140-6736(03)15262-2.Groen EH, Rossem L van, Jongste JC de, Mohangoo AD, Moll HA, Jaddoe VWV, et al. 2012. The role of prenatal, perinatal and postnatal factors in the explanation of socioeconomic inequalities in preschool asthma symptoms: the Generation R Study. J. Epidemiol. Community Health 66(11):1017–1024; doi:10.1136/jech-2011-200333.Gruzieva O, Bellander T, Eneroth K, Kull I, Melén E, Nordling E, et al. 2012. Traffic-related air pollution and development of allergic sensitization in children during the first 8 years of life. J. Allergy Clin. Immunol. 129(1):240–246; doi:10.1016/j.jaci.2011.11.001.133Gruzieva O, Gehring U, Aalberse R, Agius R, Beelen R, Behrendt H, et al. 2014. Meta-analysis of air pollution exposure association with allergic sensitization in European birth cohorts. J. Allergy Clin. Immunol. 133(3):767–776.e7; doi:10.1016/j.jaci.2013.07.048.Guarnieri M, Balmes JR. 2014. Outdoor air pollution and asthma. The Lancet 383(9928):1581–1592; doi:10.1016/S0140-6736(14)60617-6.Ha SG, Ge XN, Bahaie NS, Kang BN, Rao A, Rao SP, et al. 2013. ORMDL3 promotes eosinophil trafficking and activation via regulation of integrins and CD48. Nat. Commun.4:2479; doi:10.1038/ncomms3479.Halapi E, Gudbjartsson DF, Jonsdottir GM, Bjornsdottir US, Thorleifsson G, Helgadottir H, et al.2010. A sequence variant on 17q21 is associated with age at onset and severity of asthma.Eur. J. Hum. Genet. 18(8):902–908; doi:10.1038/ejhg.2010.38.Halekoh U, Højsgaard S, Yan J. 2006. The R package geepack for generalized estimating equations. J. Stat. Softw. 15(2):1–11.Hao K, Bossé Y, Nickle DC, Paré PD, Postma DS, Laviolette M, et al. 2012. Lung eQTLs to helpreveal the molecular underpinnings of asthma. PLoS Genet 8(11):e1003029; doi:10.1371/journal.pgen.1003029.Heinrich J. 2010. Air pollution, asthma and allergies in children. Occup. Environ. Med. 67(5):290–291; doi:10.1136/oem.2009.051193.Heinrich J, Bolte G, Hölscher B, Douwes J, Lehmann I, Fahlbusch B, et al. 2002a. Allergens andendotoxin on mothers’ mattresses and total immunoglobulin E in cord blood of neonates. Eur. Respir. J. 20(3):617–623; doi:10.1183/09031936.02.02322001.Heinrich J, Hoelscher B, Frye C, Meyer I, Wjst M, Wichmann H-E. 2002b. Trends in prevalence of atopic diseases and allergic sensitization in children in Eastern Germany. Eur. Respir. J. 19(6):1040–1046; doi:10.1183/09031936.02.00261802.Heinrich J, Wichmann H. 2004. Traffic related pollutants in Europe and their effect on allergic disease. Curr. Opin. Allergy Clin. Immunol. 4(5): 341–348.Hemminki E, Gissler M. 1996. Births by younger and older mothers in a population with late andregulated childbearing: Finland 1991. Acta Obstet. Gynecol. Scand. 75(1):19–27.Henderson SB, Beckerman B, Jerrett M, Brauer M. 2007. Application of land use regression to estimate long-term concentrations of traffic-related nitrogen oxides and fine particulate matter. Environ. Sci. Technol. 41(7):2422–2428; doi:10.1021/es0606780.Hinds DA, McMahon G, Kiefer AK, Do CB, Eriksson N, Evans DM, et al. 2013. A genome-wideassociation meta-analysis of self-reported allergy identifies shared and allergy-specific susceptibility loci. Nat. Genet. 45(8):907–911; doi:10.1038/ng.2686.134Hirota T, Harada M, Sakashita M, Doi S, Miyatake A, Fujita K, et al. 2008. Genetic polymorphism regulating ORM1-like 3 (Saccharomyces cerevisiae) expression is associated with childhood atopic asthma in a Japanese population. J. Allergy Clin. Immunol. 121(3):769–770; doi:10.1016/j.jaci.2007.09.038.Hirschhorn JN, Daly MJ. 2005. Genome-wide association studies for common diseases and complex traits. Nat. Rev. Genet. 6(2):95–108; doi:10.1038/nrg1521.Hochadel M, Heinrich J, Gehring U, Morgenstern V, Kuhlbusch T, Link E, et al. 2006. Predictinglong-term average concentrations of traffic-related air pollutants using GIS-based information. Atmos. Environ. 40(3):542–553; doi:10.1016/j.atmosenv.2005.09.067.Hoek G, Meliefste K, Cyrys J, Lewné M, Bellander T, Brauer M, et al. 2002. Spatial variability of fine particle concentrations in three European areas. Atmos. Environ. 36(25):4077–4088; doi:10.1016/S1352-2310(02)00297-2.Holloway JW, Savarimuthu Francis S, Fong KM, Yang IA. 2012. Genomics and the respiratory effects of air pollution exposure. Respirology 17(4):590–600; doi:10.1111/j.1440-1843.2012.02164.x.Howie BN, Donnelly P, Marchini J. 2009. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet 5(6):e1000529; doi:10.1371/journal.pgen.1000529.Hwang B, Jaakkola J, Lee Y, Lin Y, Guo Y-L. 2006. Relation between air pollution and allergic rhinitis in Taiwanese schoolchildren. Respir. Res. 7(1):23–29; doi:10.1186/1465-9921-7-23.IPCC, Stocker TF, Qin D, Plattner G-K, Tignor MT, Allen SK, et al. 2013. Summary for policymakers in: Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.Iwasaki M, Saito K, Takemura M, Sekikawa K, Fujii H, Yamada Y, et al. 2003. TNF-α contributes to the development of allergic rhinitis in mice. J. Allergy Clin. Immunol. 112(1):134–140; doi:10.1067/mai.2003.1554.Kaiser H. 2004. Risk factors in allergy/asthma. Allergy Asthma Proc. 25(1): 7–10.Kauffmann F, Demenais F. 2012. Gene-environment interactions in asthma and allergic diseases: Challenges and perspectives. J. Allergy Clin. Immunol. 130(6):1229–1240; doi:10.1016/j.jaci.2012.10.038.135Kavalar MS, Balantic M, Silar M, Kosnik M, Korosec P, Rijavec M. 2012. Association of ORMDL3, STAT6 and TBXA2R gene polymorphisms with asthma. Int. J. Immunogenet.39(1):20–25; doi:10.1111/j.1744-313X.2011.01051.x.Kay AB. 2001. Allergy and allergic diseases. N. Engl. J. Med. 344(1):30–37; doi:10.1056/NEJM200101043440106.Kerkhof M, Postma D, Brunekreef B. 2010. Toll-like receptor 2 and 4 genes influence susceptibility to adverse effects of traffic-related air pollution on childhood asthma. Thorax 65(8):690–697; doi:10.1136/thx.2009.119636.Kim S-H, Park H-S, Jang J-Y. 2011. Impact of meteorological variation on hospital visits of patients with tree pollen allergy. BMC Public Health 11(1):890; doi:10.1186/1471-2458-11-890.Kim WK, Kwon J-W, Seo J-H, Kim HY, Yu J, Kim B-J, et al. 2012. Interaction between IL13 genotype and environmental factors in the risk for allergic rhinitis in Korean children. J. Allergy Clin. Immunol. 130(2):421–426.e5; doi:10.1016/j.jaci.2012.04.052.Kottek M, Grieser J, Beck C, Rudolf B, Rubel F. 2006. World Map of the Koeppen-Geiger climate classification updated. Meteorol. Z. 15(3):259–263; doi:10.1127/0941-2948/2006/0130.Kozyrskyj AL, HayGlass KT, Sandford A, Pare PD, Chan-Yeung M, Becker AB. 2009. A novel study design to investigate the early-life origins of asthma in children (SAGE study). Allergy 64(8):1185–1193; doi:10.1111/j.1398-9995.2009.02033.x.Krämer U, Koch T, Ranft U, Ring J, Behrendt H. 2000. Traffic-related air pollution is associated with atopy in children living in urban areas. Epidemiology 11(1): 64–70.Krämer U, Sugiri D, Ranft U, Krutmann J, von Berg A, Berdel D, et al. 2009. Eczema, respiratory allergies, and traffic-related air pollution in birth cohorts from small-town areas. J. Dermatol. Sci. 56(2):99–105; doi:10.1016/j.jdermsci.2009.07.014.Krzyzanowski M, Kuna-Dibbert B, Schneider S. 2005. Health effects of transport-related air pollution. WHO Regional Office Europe.Lee Y-L, Shaw C-L, Su H-J, Lai J-S, Ko Y-C, Huang S-L, et al. 2003. Climate, traffic-related air pollutants and allergic rhinitis prevalence in middle-school children in Taiwan. Eur. Respir. J. 21(6):964–970; doi:10.1183/09031936.03.00094602.Lee, Y-L, Lin, Y-C, Lee, J-Y, Wang, T-R, Guo, Y.L. 2004. Glutathione S-transferase P1 gene polymorphism and air pollution as interactive risk factors for childhood asthma. Clin. Exp. Allergy 34(11):1707–1713; doi:10.1111/j.1365-2222.2004.02099.x.136Leung TF, Sy HY, Ng MCY, Chan IHS, Wong GWK, Tang NLS, et al. 2009. Asthma and atopy are associated with chromosome 17q21 markers in Chinese children. Allergy 64(4):621–628; doi:10.1111/j.1398-9995.2008.01873.x.Li Y, Willer CJ, Ding J, Scheet P, Abecasis GR. 2010. MaCH: using sequence and genotype data to estimate haplotypes and unobserved genotypes. Genet. Epidemiol. 34(8):816–834; doi:10.1002/gepi.20533.Li Y-F, Gauderman WJ, Avol E, Dubeau L, Gilliland FD. 2006. Associations of tumor necrosis factor G-308A with childhood asthma and wheezing. Am. J. Respir. Crit. Care Med. 173(9):970–976; doi:10.1164/rccm.200508-1256OC.Lin GC, Zacharek MA. 2012. Climate change and its impact on allergic rhinitis and other allergic respiratory diseases. Curr. Opin. Otolaryngol. Head Neck Surg. 20(3):188–193; doi:10.1097/MOO.0b013e3283524b14.London S. 2007. Gene-air pollution interactions in asthma. Proc. Am. Thoraic Soc. 4(3):217–220; doi:10.1513/pats.200701-031AW.Louis, Franchimont, Piron, Gevaert, Schaaf Lafontaine, Roland, et al. 1998. Tumour necrosis ‐factor (TNF) gene polymorphism influences TNF α production in lipopolysaccharide ‐(LPS) stimulated whole blood cell culture in healthy humans. Clin. Exp. Immunol. ‐113(3):401–406; doi:10.1046/j.1365-2249.1998.00662.x.MacIntyre E, Brauer M, Melen E, Bauer C-P, Bauer M, Berdel D, et al. 2014a. GSTP1 and TNF gene variants and associations between air pollution and incident childhood asthma: The Traffic, Asthma and Genetics (TAG) Study. Environ. Health Perspect. 122(4):418–424; doi:10.1289/ehp.1307459.MacIntyre EA, Carlsten C, MacNutt M, Fuertes E, Melén E, Tiesler CMT, et al. 2013. Traffic, asthma and genetics: combining international birth cohort data to examine genetics as a mediator of traffic-related air pollution’s impact on childhood asthma. Eur. J. Epidemiol. 28(7):597–606; doi:10.1007/s10654-013-9828-5.MacIntyre EA, Gehring U, Mölter A, Fuertes E, Klümper C, Krämer U, et al. 2014b. Air pollution and respiratory infections during early childhood: an analysis of 10 European birth cohorts within the ESCAPE project. Environ. Health Perspect. 122(1):107–113; doi:10.1289/ehp.1306755.Manolio TA, Collins FS, Cox NJ, Goldstein DB, Hindorff LA, Hunter DJ, et al. 2009. Finding the missing heritability of complex diseases. Nature 461(7265):747–753; doi:10.1038/nature08494.Marchini J, Cardon LR, Phillips MS, Donnelly P. 2004. The effects of human population structure on large genetic association studies. Nat. Genet. 36(5):512–517; doi:10.1038/ng1337.137Marenholz I, Bauerfeind A, Esparza-Gordillo J, Kerscher T, Granell R, Nickel R, et al. 2011. Theeczema risk variant on chromosome 11q13 (rs7927894) in the population-based ALSPACcohort: a novel susceptibility factor for asthma and hay fever. Hum. Mol. Genet. 20(12):2443–2449; doi:10.1093/hmg/ddr117.Marshall G. 2004. Internal and external environmental influences in allergic diseases. J. Am. Osteopath. Assoc. 105(Suppl. 5): S1–S6.Marshall JD, Nethery E, Brauer M. 2008. Within-urban variability in ambient air pollution: Comparison of estimation methods. Atmos. Environ. 42(6):1359–1369; doi:10.1016/j.atmosenv.2007.08.012.McConnell R, Berhane K, Yao L, Jerrett M, Lurmann F, Gilliland F, et al. 2006. Traffic, susceptibility, and childhood asthma. Environ. Health Perspect. 114:766–772; doi:10.1289/ehp.8594.Melén E, Nyberg F, Lindgren CM, Berglind N, Zucchelli M, Nordling E, et al. 2008. Interactionsbetween glutathione S-transferase P1, tumor necrosis factor, and traffic-related air pollution for development of childhood allergic disease. Environ. Health Perspect. 116(8):1077–1084; doi:10.1289/ehp.11117.Melén E, Pershagen G. 2012. Pathophysiology of asthma: lessons from genetic research with particular focus on severe asthma. J. Intern. Med. 272(2):108–120; doi:10.1111/j.1365-2796.2012.02555.x.Minhas K, Micheal S, Ahmed F, Ahmed A. 2010. Strong association between the -308 TNF promoter polymorphism and allergic rhinitis in pakistani patients. J. Investig. Allergol. Clin. Immunol. 20(7): 563–566.Mitchell T. 2004. High resolution observational climatologies version 2.1. University of East Anglia: Climate Research Unit. Available: www.ipcc-data.org/obs/cru_ts2_1.html [accessed: April 4, 2012].Mitchell TD, Jones PD. 2005. An improved method of constructing a database of monthly climate observations and associated high-resolution grids. Int. J. Climatol. 25(6):693–712; doi:10.1002/joc.1181.Moffatt M, Gut I, Demenais F, Strachan D, Bouzigon E, Heath S, et al. 2010. A large-scale, consortium-based genomewide association study of asthma. N. Engl. J. Med. 363(13):1211–1221; doi:10.1056/NEJMoa0906312.Moffatt MF, Kabesch M, Liang L, Dixon AL, Strachan D, Heath S, et al. 2007. Genetic variants regulating ORMDL3 expression contribute to the risk of childhood asthma. Nature 448(7152):470–473; doi:10.1038/nature06014.138Mölter A, Simpson A, Berdel D, Brunekreef B, Custovic A, Cyrys J, et al. 2014. A multi-centre study of air pollution exposure and childhood asthma – the ESCAPE project. Eur Respir J. [accepted]. Morgenstern V, Zutavern A, Cyrys J, Brockow I, Koletzko S, Kramer U, et al. 2008. Atopic diseases, allergic sensitization, and exposure to traffic-related air pollution in children. Am. J. Respir. Crit. Care Med. 177(12):1331–1337; doi:10.1164/rccm.200701-036OC.Morgenstern, Zutavern A, Cyrys J, Brockow I, Gehring U, Koletzko S, et al. 2007. Respiratory health and individual estimated exposure to traffic-related air pollutants in a cohort of young children. Occup. Environ. Med. 64(1):8 –16; doi:10.1136/oem.2006.028241.Motta AC, Marliere M, Peltre G, Sterenberg PA, Lacroix G. 2006. Traffic-related air pollutants induce the release of allergen-containing cytoplasmic granules from grass pollen. Int. Arch. Allergy Immunol. 139(4):294–298; doi:10.1159/000091600.Namork E, Johansen BV, Løvik M. 2006. Detection of allergens adsorbed to ambient air particlescollected in four European cities. Toxicol. Lett. 165(1):71–78; doi:10.1016/j.toxlet.2006.01.016.Newhouse CP, Levetin E. 2004. Correlation of environmental factors with asthma and rhinitis symptoms in Tulsa, OK. Ann. Allergy. Asthma. Immunol. 92(3):356–366; doi:10.1016/S1081-1206(10)61575-X.Nicolai T, Carr D, Weiland S, Duhme H, von Ehrenstein O, Wagner C, et al. 2003. Urban traffic and pollutant exposure related to respiratory outcomes and atopy in a large sample of children. Eur. Respir. J. 21(6):956–963; doi:10.1183/09031936.03.00041103a.Nilsson D, Andiappan AK, Halldén C, Tim CF, Säll T, Wang DY, et al. 2013. Poor reproducibility of allergic rhinitis SNP associations. PLoS ONE 8(1):e53975; doi:10.1371/journal.pone.0053975.Nonaka M, Nonaka R, Jordana M, Dolovich J. 1996. GM-CSF, IL-8, IL-1R, TNF-alpha R, and HLA-DR in nasal epithelial cells in allergic rhinitis. Am. J. Respir. Crit. Care Med. 153(5):1675–1681; doi:10.1164/ajrccm.153.5.8630619.Nordling E, Berglind N, Melén E, Emenius G, Hallberg J, Nyberg F, et al. 2008. Traffic-related air pollution and childhood respiratory symptoms, function and allergies. Epidemiology 19(3):401–408; doi:10.1097/EDE.0b013e31816a1ce3.Nowak, Wichmann, Magnussen. 1998. Asthma and atopy in Western and Eastern communities - current status and open questions. Clin. Exp. Allergy 28(9):1043–1046; doi:10.1046/j.1365-2222.1998.00373.x.139O’Regan GM, Campbell LE, Cordell HJ, Irvine AD, McLean WHI, Brown SJ. 2010. Chromosome 11q13.5 variant associated with childhood eczema: An effect supplementary to filaggrin mutations. J. Allergy Clin. Immunol. 125(1):170–174.e2; doi:10.1016/j.jaci.2009.10.046.Oftedal B, Brunekreef B, Nystad W, Nafstad P. 2007. Residential outdoor air pollution and allergen sensitization in schoolchildren in Oslo, Norway. Clin. Exp. Allergy 37(11):1632–1640; doi:10.1111/j.1365-2222.2007.02823.x.Passalacqua G. 2000. United airways disease: therapeutic aspects. Thorax 55(Suppl. 2):26S–27; doi:10.1136/thorax.55.suppl_2.S26.Pinart M, Benet M, Annesi-Maesano I, von Berg A, Berdel D, Carlsen KCL, et al. 2014. Comorbidity of eczema, rhinitis, and asthma in IgE-sensitised and non-IgE-sensitised children in MeDALL: a population-based cohort study. Lancet Respir. Med. 2(2):131–140; doi:10.1016/S2213-2600(13)70277-7.Pinzon J, Brown M, Tucker C. 2005. Satellite time series correction of orbital drift artifacts usingempirical mode decomposition. Hilbert-Huang transform: introduction and applications (Part II). 167-186.Pitz M, Kreyling WG, Hölscher B, Cyrys J, Wichmann HE, Heinrich J. 2001. Change of the ambient particle size distribution in East Germany between 1993 and 1999. Atmos. Environ. 35(25):4357–4366; doi:10.1016/S1352-2310(01)00229-1.Querol X, Alastuey A, Ruiz CR, Artiñano B, Hansson HC, Harrison RM, et al. 2004. Speciation and origin of PM10 and PM2.5 in selected European cities. Atmos. Environ. 38(38):6547–6555; doi:10.1016/j.atmosenv.2004.08.037.R Core Team. 2012. R: A language and environment for statistical computing. R   Foundation forStatistical Computing, Vienna, Austria. ISBN 3-900051-07-0, www.R-project.org/.Ramasamy A, Curjuric I, Coin LJ, Kumar A, McArdle WL, Imboden M, et al. 2011. A genome-wide meta-analysis of genetic variants associated with allergic rhinitis and grass sensitization and their interaction with birth order. J. Allergy Clin. Immunol. 128(5):996–1005; doi:10.1016/j.jaci.2011.08.030.Räsäsnen M, Laitinen T, Kaprio J, Koskenvuo M, Laitinen LA. 1998. Hay fever – a Finnish nationwide study of adolescent twins and their parents. Allergy 53(9):885–890; doi:10.1111/j.1398-9995.1998.tb03996.x.Reungoat P, Chiron M, Gauvin S, Zmirou-Navier D, Momas I. 2005. Retrospective assessment of exposure to traffic air pollution using the ExTra index in the VESTA French epidemiological study. J. Expo. Sci. Environ. Epidemiol. 15(6):524–533; doi:10.1038/sj.jea.7500430.140Riccio AM, Tosca MA, Cosentino C, Pallestrini E, Ameli F, Canonica GW, et al. 2002. Cytokine pattern in allergic and non allergic chronic rhinosinusitis in asthmatic children. Clin. Exp.‐Allergy 32(3):422–426; doi:10.1046/j.1365-2222.2002.01315.x.Robinson W. 1950. Ecological correlations and the behavior of individuals. Am. Sociol. Rev. 15(3): 351–357.Romieu I, Castro-Giner F, Kunzli N, Sunyer J. 2008. Air pollution, oxidative stress and dietary supplementation: a review. Eur. Respir. J. 31(1):179 –197; doi:10.1183/09031936.00128106.Romieu I, Ramirez-Aguilar M, Sienra-Monge JJ, Moreno-Macías H, del Rio-Navarro BE, DavidG, et al. 2006. GSTM1 and GSTP1 and respiratory health in asthmatic children exposed to ozone. Eur. Respir. J. 28(5):953 –959; doi:10.1183/09031936.06.00114905.Ruijsbroek A, Wijga A, Kerkhof M, Koppelman G, Smit H, Droomers M. 2011. The development of socio-economic health differences in childhood: results of the Dutch longitudinal PIAMA birth cohort. BMC Public Health 11(1):225; doi:10.1186/1471-2458-11-225.Ryan PH, LeMasters GK, Levin L, Burkle J, Biswas P, Hu S, et al. 2008. A land-use regression model for estimating microenvironmental diesel exposure given multiple addresses from birth through childhood. Sci. Total Environ. 404(1):139–147; doi:10.1016/j.scitotenv.2008.05.051.Salam MT, Lin P-C, Avol EL, Gauderman WJ, Gilliland FD. 2007. Microsomal epoxide hydrolase, glutathione S-transferase P1, traffic and childhood asthma. Thorax 62(12):1050–1057; doi:10.1136/thx.2007.080127.Sario MD, Katsouyanni K, Michelozzi P. 2013. Climate change, extreme weather events, air pollution and respiratory health in Europe. Eur. Respir. J. 42(3):826–843; doi:10.1183/09031936.00074712.Sausenthaler S, Standl M, Buyken A, Rzehak P, Koletzko S, Bauer CP, et al. 2011. Regional and socio-economic differences in food, nutrient and supplement intake in school-age children in Germany: results from the GINIplus and the LISAplus studies. Public Health Nutr. 14(10):1724–1735; doi:10.1017/S1368980010003575.Saxon A, Diaz-Sanchez D. 2005. Air pollution and allergy: you are what you breathe. Nat. Immunol. 6(3): 223–226.Senthilselvan A, Rennie D, Chénard L, Burch LH, Babiuk L, Schwartz DA, et al. 2008. Association of polymorphisms of toll-like receptor 4 with a reduced prevalence of hay fever and atopy. Ann. Allergy. Asthma. Immunol. 100(5):463–468; doi:10.1016/S1081-1206(10)60472-3.141Shea KM, Truckner RT, Weber RW, Peden DB. 2008. Climate change and allergic disease. J. Allergy Clin. Immunol. 122(3):443–453; doi:10.1016/j.jaci.2008.06.032.Sheikh A, Hurwitz B, Nurmatov U, van Schayck CP. 2010. House dust mite avoidance measures for perennial allergic rhinitis. Cochrane Database Syst. Rev. 7; doi:10.1002/14651858.CD001563.pub3.Skoner DP. 2001. Allergic rhinitis: Definition, epidemiology, pathophysiology, detection, and diagnosis. J. Allergy Clin. Immunol. 108(Suppl. 1):S2–S8; doi:10.1067/mai.2001.115569.Slama R, Gräbsch C, Lepeule J, Siroux V, Cyrys J, Sausenthaler S, et al. 2010. Maternal fine particulate matter exposure, polymorphism in xenobiotic-metabolizing genes and offspring birth weight. Reprod. Toxicol. 30(4):600–612; doi:10.1016/j.reprotox.2010.07.001.Socioeconomic Data and Applications Center. 2004. Gridded Population of the World (GPW), v3. Available: http://sedac.ciesin.columbia.edu/gpw/ [accessed: August 20, 2012].Spergel J. 2005. Atopic march: link to upper airways. Curr. Opin. Allergy Clin. Immunol. 5(1): 17–21.Stewart A, Mitchell E, Pearce N, Strachan DP, Weiland SK, ISAAC Steering Committee. 2001. The relationship of per capita gross national product to the prevalence of symptoms of asthma and other atopic diseases in children (ISAAC). Int. J. Epidemiol. 30(1):173–179; doi:10.1093/ije/30.1.173.Strachan D, Sibbald B, Weiland S, Ait-Khaled N, Anabwani G, Anderson HR, et al. 1997. Worldwide variations in prevalence of symptoms of allergic rhinoconjunctivitis in children: the International Study of Asthma and Allergies in Childhood (ISAAC). Pediatr.Allergy Immunol. 8(4):161–168; doi:10.1111/j.1399-3038.1997.tb00156.x.Strak M, Janssen NA., Godri KJ, Gosens I, Mudway IS, Cassee FR, et al. 2012. Respiratory health effects of airborne particulate matter: the role of particle size, composition, and oxidative potential--The RAPTES Project. Environ. Health Perspect. 120(8):1183–1189; doi:10.1289/ehp.1104389.Strand V, Rak S, Svartengren M, Bylin G. 1997. Nitrogen dioxide exposure enhances asthmatic reaction to inhaled allergen in subjects with asthma. Am. J. Respir. Crit. Care Med. 155(3):881–887; doi:10.1164/ajrccm.155.3.9117021.Tager I, Demerjian K, Frampton M, et al. 2010. Traffic-related air pollution: a critical review of the literature on emissions, exposure, and health effects. No. 17. Health Effects Institute.142Takai T, Shigaku I. 2011. Barrier dysfunction caused by environmental proteases in the pathogenesis of allergic diseases. Allergol. Int. 60(1):25–35; doi:10.2332 allergolint.10-RAI-0273.Thomas D. 2010. Gene–environment-wide association studies: emerging approaches. Nat. Rev. Genet. 11(4):259–272; doi:10.1038/nrg2764.Thomsen SF, Ulrik CS, Kyvik KO, Skadhauge LR, Steffensen I, Backer V. 2006. Findings on theatopic triad from a Danish twin registry. Int. J. Tuberc. Lung Dis. 10(11): 1268–1272.Tomita K, Sakashita M, Hirota T, Tanaka S, Masuyama K, Yamada T, et al. 2013. Variants in the 17q21 asthma susceptibility locus are associated with allergic rhinitis in the Japanese population. Allergy 68(1):92–100; doi:10.1111/all.12066.Traidl-Hoffmann C, Kasche A, Menzel A, Jakob T, Thiel M, Ring J, et al. 2003. Impact of pollen on human health: more than allergen carriers? Int. Arch. Allergy Immunol. 131(1):1–13; doi:10.1159/000070428.Tucker CJ, Pinzon JE, Brown ME, Slayback DA, Pak EW, Mahoney R, et al. 2005. An extended AVHRR 8 km NDVI dataset compatible with MODIS and SPOT vegetation NDVI data. ‐Int. J. Remote Sens. 26(20):4485–4498; doi:10.1080/01431160500168686.Tulah AS, Holloway JW, Sayers I. 2013. Defining the contribution of SNPs identified in asthma GWAS to clinical variables in asthmatic children. BMC Med. Genet. 14(1):100; doi:10.1186/1471-2350-14-100.United States Environmental Protection Agency. 2008. A review of the impact of climate variability and change on aeroallergens and their associated effects. EPA/600/R-06/164F.Van der Valk RJP, Duijts L, Kerkhof M, Willemsen SP, Hofman A, Moll HA, et al. 2012. Interaction of a 17q12 variant with both fetal and infant smoke exposure in the development of childhood asthma-like symptoms. Allergy 67(6):767–774; doi:10.1111/j.1398-9995.2012.02819.x.Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, et al. 2001. The sequence of the human genome. Science 291(5507):1304 –1351; doi:10.1126/science.1058040.Vercelli D. 2008. Discovering susceptibility genes for asthma and allergy. Nat. Rev. Immunol. 8(3):169–182; doi:10.1038/nri2257.Verlaan DJ, Berlivet S, Hunninghake GM, Madore A-M, Larivière M, Moussette S, et al. 2009a. Allele-specific chromatin remodeling in the ZPBP2/GSDMB/ORMDL3 locus associated with the risk of asthma and autoimmune disease. Am. J. Hum. Genet. 85(3):377–393; doi:10.1016/j.ajhg.2009.08.007.143Verlaan DJ, Ge B, Grundberg E, Hoberman R, Lam KCL, Koka V, et al. 2009b. Targeted screening of cis-regulatory variation in human haplotypes. Genome Res. 19(1):118–127; doi:10.1101/gr.084798.108.Verlato G, Calabrese R, De Marco R. 2002. Correlation between asthma and climate in the European Community Respiratory Health Survey. Arch. Environ. Health 57(1):48–52; doi:10.1080/00039890209602916.Von Berg A, Koletzko S, Grübl A, Filipiak-Pittroff B, Wichmann H-E, Bauer CP, et al. 2003. Theeffect of hydrolyzed cow’s milk formula for allergy prevention in the first year of life: The German Infant Nutritional Intervention Study, a randomized double-blind trial. J. Allergy Clin. Immunol. 111(3):533–534; doi:10.1067/mai.2003.101.Wang M, Beelen R, Stafoggia M, Raaschou-Nielsen O, Andersen ZJ, Hoffmann B, et al. 2014. Long-term exposure to elemental constituents of particulate matter and cardiovascular mortality in 19 European cohorts: Results from the ESCAPE and TRANSPHORM projects. Environ. Int. 66:97–106; doi:10.1016/j.envint.2014.01.026.Wang R, Henderson SB, Sbihi H, Allen RW, Brauer M. 2013. Temporal stability of land use regression models for traffic-related air pollution. Atmos. Environ. 64:312–319; doi:10.1016/j.atmosenv.2012.09.056.Weiland SK, Björkstén B, Brunekreef B, Cookson WOC, Mutius E von, Strachan DP. 2004a. Phase II of the International Study of Asthma and Allergies in Childhood (ISAAC II): rationale and methods. Eur. Respir. J. 24(3):406–412; doi:10.1183/09031936.04.00090303.Weiland SK, Hüsing A, Strachan DP, Rzehak P, Pearce N. 2004b. Climate and the prevalence of symptoms of asthma, allergic rhinitis, and atopic eczema in children. Occup. Environ. Med. 61(7):609–615; doi:10.1136/oem.2002.006809.Wickman M, Pershagen G, Nordvall SL. 2002. The BAMSE Project: presentation of a prospective longitudinal birth cohort study. Pediatr. Allergy Immunol. 13(suppl. 15):11–13; doi:10.1034/j.1399-3038.13.s.15.10.x.Wilson AG, Symons JA, McDowell TL, McDevitt HO, Duff GW. 1997. Effects of a polymorphism in the human tumor necrosis factor α promoter on transcriptional activation. Proc. Natl. Acad. Sci. 94(7):3195–3199.Wilson AG, Vries N de, Pociot F, Giovine FS di, Putte LB van der, Duff GW. 1993. An allelic polymorphism within the human tumor necrosis factor alpha promoter region is strongly associated with HLA A1, B8, and DR3 alleles. J. Exp. Med. 177(2):557–560; doi:10.1084/jem.177.2.557.144World Bank. 2012. GNI per Capita, Atlas Method (current US$). Available: http://data.worldbank.org/indicator/NY.GNP.PCAP.CD?order=wbapi_data_value_2001+wbapi_data_value+wbapi_data_value-last&sort=asc&page=2 [accessed: March 13, 2012].Wu H, Romieu I, Sienra-Monge J-J, del Rio-Navarro BE, Anderson DM, Dunn EW, et al. 2007. Parental smoking modifies the relation between genetic variation in tumor necrosis factor-α (TNF) and childhood asthma. Environ. Health Perspect. 115(4):616–622; doi:10.1289/ehp.9740.Wu H, Romieu I, Sienra-Monge J-J, Li H, Del Rio-Navarro BE, London SJ. 2009. Genetic variation in ORM1-like 3 (ORMDL3) and gasdermin-like (GSDML) and childhood asthma. Allergy 64(4):629–635; doi:10.1111/j.1398-9995.2008.01912.x.Wyler C, Braun-Fahrländer C, Künzli N, Schindler C, Ackermann-Liebrich U, Perruchoud A, et al. 2000. Exposure to motor vehicle traffic and allergic sensitization. Epidemiology 11(4):450–456.Yang IA, Holz O, Jorres RA, Magnussen H, Barton SJ, Rodriguez S, et al. 2005. Association of tumor necrosis factor-α polymorphisms and ozone-induced change in lung function. Am. J. Respir. Crit. Care Med. 171(2):171–176; doi:10.1164/rccm.200402-194OC.Younger M, Morrow-Almeida HR, Vindigni SM, Dannenberg AL. 2008. The built environment, climate change, and health: opportunities for co-benefits. Am. J. Prev. Med. 35(5):517–526; doi:10.1016/j.amepre.2008.08.017.Zanolin ME, Pattaro C, Corsico A, Bugiani M, Carrozzi L, Casali L, et al. 2004. The role of climate on the geographic variability of asthma, allergic rhinitis and respiratory symptoms: results from the Italian study of asthma in young adults. Allergy 59(3):306–314; doi:10.1046/j.1398-9995.2003.00391.x.Zheng T, Yu J, Oh MH, Zhu Z. 2011. The atopic march: progression from atopic dermatitis to allergic rhinitis and asthma. Allergy Asthma Immunol Res 2(3):67–73; doi:10.4168/aair.2011.3.2.67.Zhu S, Chan-Yeung M, Becker AB, Dimich-Ward H, Ferguson AC, Manfreda J, et al. 2000. Polymorphisms of the IL-4, TNF- α , and Fc α RI β genes and the risk of allergic disorders in at-risk infants. Am. J. Respir. Crit. Care Med. 161(5):1655–1659; doi:10.1164/ajrccm.161.5.9906086.Ziska L, Knowlton K, Rogers C, Dalan D, Tierney N, Elder MA, et al. 2011. Recent warming by latitude associated with increased length of ragweed pollen season in central North America. Proc. Natl. Acad. Sci. 108(10):4248–4251; doi:10.1073/pnas.1014107108.145Ziska LH. 2003. Evaluation of the growth response of six invasive species to past, present and future atmospheric carbon dioxide. J. Exp. Bot. 54(381):395 –404; doi:10.1093/jxb/erg027.Ziska LH, Gebhard DE, Frenz DA, Faulkner S, Singer BD, Straka JG. 2003. Cities as harbingers of climate change: Common ragweed, urbanization, and public health. J. Allergy Clin. Immunol. 111(2):290–295; doi:10.1067/mai.2003.53.Zutavern A, Brockow I, Schaaf B, Bolte G, von Berg A, Diez U, et al. 2006. Timing of solid foodintroduction in relation to atopic dermatitis and atopic sensitization: results from a prospective birth cohort study. Pediatrics 117(2):401–411; doi:10.1542/peds.2004-2521.146AppendicesAppendix A: Genotype imputation For GINI/LISA, 1,027 and 69 DNA samples were analyzed using the Affymetrix Human SNP Arrays 5.0 and 6.0, respectively. Genotypes were called using the BRLMM-P (Affymetrix 5.0) orBIRDSEED V2 algorithm (Affymetrix 6.0). The genotype data were subjected to quality control filters on the variant and individual levels. Variants were excluded if the call rate was below 95%, the minor allele frequency was below 1% or the Hardy-Weinberg equilibrium p-value was below 1E-5 (~18% excluded). Individuals who had a call rate above 95% and a heterozygosity value within +/- 4 standard deviations of the mean, and who passed a sex check and the similarity quality control step based on multidimensional scaling plots were retained for subsequent imputation. The genotype data were prephased using SHAPEIT v2 (Delaneau et al. 2012, 2013) and imputed using IMPUTE v2 (Howie et al. 2009) against reference haplotypes from the 1000 Genomes Project (Phase I integrated variant set (v3); March 2012, updated August26, 2012; limited to variants with more than one minor allele copy). All seven SNPs used in the analysis in Chapter 4 were imputed with good imputation quality (IMPUTE v2 INFO > 0.94). For PIAMA, DNA was extracted from blood or buccal swabs. DNA of 1,377 children was genotyped on the Illumina Omni Express Exome Chip and DNA of 288 children was genotyped with the Omni Express chip, both at the Genomics Facility of the University Medical Center Groningen. DNA of 404 children was genotyped at the Centre National de Genotypage (CNG, Evry, France) as part of the GABRIEL consortium (Moffatt et al. 2010). SNPs were harmonized by base pair position annotated to genome build 37, name and annotation of strand for each platform. Discordant or duplicate SNPs or SNPs that showed large differences in allele frequencies (> 15 %) were removed. After quality control, a total of 1,968 individuals remained and imputation was performed per platform using IMPUTE 2.0 against the reference data set of the CEU panel of the 1000 Genomes project (version March 2012). SNPs of high quality (info-score IMPUTE ≥ 0.7) were merged into one dataset using GTOOL and used for further analysis. 147For CAPPS and SAGE, data on only four SNPs (rs2305480, rs7216389, rs12603332 and rs3744246) were available. In total, 956 samples from parents and children were genotyped using the Illumina HumanHap550 SNP Array. Genotypes were subjected to the following qualitycontrol filters for variants and individual samples. Variants were excluded on the basis of the following criteria: a call rate below 95%, a minor allele frequency below 1%, failure of Hardy-Weinberg equilibrium (p-value below 1E-4) and SNPs with > 2 Mendelian errors. Individual samples were retained if they had a call rate > 97% and a heterozygosity value within +/- 3 standard deviations of the mean. Samples additionally had to pass gender and Mendelian transmission error checks. Multi-Dimensional Scaling was used to check for population stratification and to identify monozygotic twins (n = 2) and duplicate samples. Imputation was performed using MaCH (Li et al. 2010) and HapMap2 r22 reference population. SNPs needed to be imputed with an r2≥0.99 in order to be used in the analysis in Chapter 5. 148Appendix B: International Study of Asthma and Allergies in Childhood Phase Three study groupsISAAC Steering CommitteeN Aït-Khaled* (International Union Against Tuberculosis and Lung Diseases, Paris, France); HR Anderson (Division of Population Health Sciences and Education, St Georges, University of London, London, UK); MI Asher (Department of Paediatrics: Child and Youth Health, Faculty ofMedical and Health Sciences, The University of Auckland, New Zealand); R Beasley* (Medical Research Institute of New Zealand, Wellington, New Zealand); B Björkstén* (Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden); B Brunekreef (Institute of Risk Assessment Science, Universiteit Utrecht, Netherlands); J Crane (Wellington Asthma Research Group, Wellington School of Medicine, New Zealand); P Ellwood (Department of Paediatrics: Child and Youth Health, Faculty of Medical and Health Sciences, The University of Auckland, New Zealand); C Flohr (Department of Paediatric Allergy & Dermatology, St John’s Institute of Dermatology, London, UK); S Foliaki* (Centre for Public Health Research, Massey University, Wellington, New Zealand); F Forastiere (Department of Epidemiology, Rome E Health Authority, Rome, Italy); L García-Marcos (Respiratory Medicine and Allergy Units, 'Virgen de la Arrixaca' University Children's Hospital, University of Murcia, Spain); U Keil* (Institut für Epidemiologie und Sozialmedizin, Universität Münster, Germany); CKW Lai* (Department of Medicine and Therapeutics, The Chinese University of Hong Kong SAR, China);J Mallol* (Department of Paediatric Respiratory Medicine, University of Santiago de Chile, Chile); EA Mitchell (Department of Paediatrics: Child and Youth Health, Faculty of Medical and Health Sciences, The University of Auckland, New Zealand); S Montefort* (Department of Medicine, University of Malta, Malta), J Odhiambo* (Centre Respiratory Diseases Research Unit, Kenya Medical Research Institute, Nairobi, Kenya); N Pearce (Department of Medical Statistics, Faculty of Epidemiology and Public Health, London School of Hygiene and Tropical Medicine, London, UK); CF Robertson (Murdoch Children's Research Institute, Melbourne, Australia); AW Stewart (Population Health, Faculty of Medical and Health Sciences, The University of Auckland, New Zealand); D Strachan (Division of Population Health Sciences and Education, St Georges, University of London, London, UK); E von Mutius (Dr von Haunerschen149Kinderklinik de Universität München, Germany); SK Weiland† (Department of Epidemiology, University of Ulm, Germany); G Weinmayr (Institute of Epidemiology, University of Ulm, Germany); H Williams (Centre for Evidence Based Dermatology, Queen’s Medical Centre, University Hospital, Nottingham, UK); G Wong (Department of Paediatrics, Prince of Wales Hospital, Hong Kong SAR, China).* Regional Coordinator; † DeceasedISAAC International Data CentreMI Asher, TO Clayton, E Ellwood, P Ellwood, EA Mitchell, Department of Paediatrics: Child and Youth Health, and AW Stewart, School of Population Health, Faculty of Medical and Health Sciences, The University of Auckland, New Zealand.ISAAC Phase Three Principal InvestigatorsAlbania: Prof A Priftanji* (Tiranë); Algeria: Prof B Benhabylès (Wilaya of Algiers); Argentina: Dr CE Baena-Cagnani* (Córdoba), Prof Dr CD Crisci (Rosario City), Dr M Gómez (Salta), Prof GE Zabert (Neuquén); Australia: Prof CF Robertson* (Melbourne); Austria: Assoc Prof G Haidinger* (Kärnten, Urfahr-Umgebung); Barbados: Dr ME Howitt* (Barbados); Belgium: Prof J Weyler (Antwerp); Bolivia: Dr R Pinto-Vargas* (Santa Cruz); Brazil: Dr CdSD Bernhardt (Itajaí), Dr WG Borges (Brasília), Prof PAM Camargos (Belo Horizonte), Dra MdS Cardoso (Manaus Amazonas), Prof AJLA da Cunha (Nova Iguaçu), Dr GB Fischer (Porto Alegre), Dr JM Motta (Aracaju), Prof FJ Passos (Maceió), Dr AC Pastorino (São Paulo West), Dr AC Porto Neto(Passo Fundo), Prof N Rosário (Curitiba), Assis Prof A Silva (Caruaru), Prof D Solé* (Rural Santa Maria, Santa Maria, São Paulo), Assoc Prof N Wandalsen (Santo Andre), Dr M de Britto (Recife), Assoc Prof L de Freitas Souza (Feira de Santana, Salvador, Vitória da Conquista); Bulgaria: Dr T Popov* (Sofia); Cameroon: Prof C Kuaban* (Yaounde); Canada: Prof A Ferguson(Vancouver), Prof D Rennie (Saskatoon); Channel Islands: Ms R Goulding (Jersey), Dr P Standring (Guernsey); Chile: Dr P Aguilar (South Santiago), Dr L Amarales (Punta Arenas), Dr LAV Benavides (Calama), Dr MA Calvo (Valdivia), Dra A Contreras (Chiloe); China: Prof Y-Z Chen* (Beijing, Tong Zhou), Assis Prof O Kunii (Tibet), Dr Q Li Pan (Wulumuqi), Prof N-S Zhong (Guangzhou); Colombia: Dr G Aristizábal (Bogotá), Dr AM Cepeda (Barranquilla), Dr 150GA Ordoñez (Cali); Congo: Prof J M'Boussa (Brazzaville); Cook Islands: Dr R Daniel* (Rarotonga); Costa Rica: Dr ME Soto-Quirós* (Costa Rica); Croatia: Dr K Lah Tomulic (Rijeka); Cuba: Dra P Varona Peréz* (La Habana); Ecuador: Dr S Barba* (Quito), Dr C Bustos (Guayaquil); Egypt: Dr ML Naguib (Cairo); El Salvador: Dr M Figueroa Colorado* (San Salvador); Estonia: Dr M-A Riikjärv* (Tallinn); Ethiopia: Assoc Prof K Melaku (Addis Ababa); Fiji: Dr R Sa'aga-Banuve (Suva); Finland: Dr J Pekkanen* (Kuopio County); Former Yugoslav Republic of Macedonia (FYROM): Assoc Prof E Vlaski* (Skopje); Gabon: Dr IE Hypolite* (Port-Gentil); Georgia: Dr M Gotua* (Kutaisi); Germany: Prof Dr U Keil* (Münster); Greece: Assoc Prof J Tsanakas (Thessaloniki); Honduras: Dr A Bueso-Engelhardt* (San Pedro Sula); Hong Kong: Prof YL Lau (Hong Kong), Prof G Wong (Hong Kong); Hungary: Dr Z Novák (Szeged), Dr G Zsigmond* (Svábhegy); India: Prof S Awasthi (Lucknow), Prof J Chhatwal (Ludhiana), Prof L Kumar (Chandigarh), Dr SN Mantri (Mumbai (29)), Prof S Rego (Bangalore), Prof M Sabir (Bikaner), Dr S Salvi (Nagpur, Pimpri), Dr G Setty (Chennai (3)), Prof SK Sharma (New Delhi (7)), Prof V Singh (Jaipur), Dr PS Suresh Babu (Davangere); Indonesia: Prof Dr CB Kartasasmita (Bandung), Prof P Konthen (Bali), Dr W Suprihati (Semarang); Iran: Dr M-R Masjedi* (Birjand, Rasht, Tehran, Zanjan); Isle of Man: Dr A Steriu (Isle of Man); Italy: Dr L Armenio (Bari), Dr L Bisanti (Milano), Dr E Bonci (Cosenza), Dr E Chellini (Firenze), Dr G Ciccone (Torino), Dr V Dell'Orco (Colleferro-Tivoli), Dr F Forastiere* (Roma), Dr C Galassi (Emilia-Romagna), Dr G Giannella (Mantova), Dr S La Grutta (Palermo), Dr MG Petronio (Empoli), Dr P Sestini (Siena), Dr S Piffer (Trento); Japan: Dr H Odajima (Fukuoka), Prof M Sohei (Tochigi); Jordan: Dr F Abu-Ekteish (Amman); Kenya: Dr FO Esamai (Eldoret), Dr L Ng’ang’a* (Nairobi); Kingdom of Tonga: Dr S Foliaki (Nuku'alofa); Kuwait: Dr JA al-Momen (Kuwait); Kyrgyzstan: Dr C Imanalieva* (Balykchi, Bishkek), Prof S Sulaimanov (Jalalabat); Latvia: Dr V Svabe (Riga); Lithuania: Prof J Bojarskas (Panevezys, Siauliai), Assoc Prof J Kudzyte* (Kaunas); Malaysia: Assoc Prof J de Bruyne* (Klang Valley), Prof BS Quah (Kota Bharu), Dr KH Teh (Alor Setar); Malta: Prof S Montefort* (Malta); Mexico: Dr M Baeza-Bacab* (Mérida), Dra M Barragán-Meijueiro (Ciudad de México (3)), Dra BE Del-Río-Navarro (Ciudad de México (1)), Dr R García-Almaráz (Ciudad Victoria), Dr SN González-Díaz (Monterrey), Dr FJ Linares-Zapién (Toluca), Dr JV Merida-Palacio (Mexicali Valley), Dra N Ramírez-Chanona (Ciudad de México (4)), Dr S Romero-Tapia (Villahermosa), Prof I Romieu 151(Cuernavaca); Morocco: Prof Z Bouayad* (Benslimane, Boulmene, Casablanca, Marrakech); Netherlands: Prof R Engels (Netherlands); New Zealand: Prof MI Asher* (Auckland), Dr C Moyes (Bay of Plenty), Dr R MacKay (Nelson), Assoc Prof P Pattemore (Christchurch), Prof N Pearce (Wellington); Nicaragua: Dr JF Sánchez* (Managua); Nigeria: Prof BO Onadeko (Ibadan); Niue: Ms M Magatogia (Niue Island); Nouvelle Caledonie: Dr I Annesi-Maesano (Nouvelle Caledonie); Pakistan: Dr N Mahmood* (Karachi), Dr MO Yusuf (Islamabad); Palestine: Dr N El Sharif* (Ramallah), Mr S Mortaja (North Gaza); Panama: Dr G Cukier* (David-Panamá); Paraguay: Dr JA Guggiari-Chase* (Asunción); Peru: Dr P Chiarella* (Lima); Philippines: Prof F Cua-Lim* (Metro Manila); Poland: Assoc Prof A Brêborowicz (Poznan), Assoc Prof G Lis* (Kraków); Polynesie Francaise: Dr I Annesi-Maesano (Polynesie Francaise); Portugal: Dr ML Chiera (Coimbra), Dra R Câmara (Funchal), Dr JM Lopes dos Santos (Porto), Dr C Nunes (Portimao), Dr J Rosado Pinto* (Lisbon); Republic of Ireland: Prof L Clancy (Republic of Ireland); Republique Democratique du Congo: Prof Dr J-M Kayembe (Kinshasa); Reunion Island: Dr I Annesi-Maesano (Reunion Island); Romania: Prof D Deleanu* (Cluj); Russia: Prof Dr EG Kondiourina (Novosibirsk); Samoa: Ms P Fuimaono V Pisi (Apia); Serbia and Montenegro: Dr O Adzovic (Podgorica), Dr M Hadnadjev (Novi Sad), Dr E Panic (Sombor),Dr S Zivanovic (Nis), Dr Z Zivkovic* (Belgrade); South Africa: Prof K Voyi (Polokwane), Prof HJ Zar* (Cape Town); South Korea: Prof H-B Lee* (Provincial Korea, Seoul); Spain: Dr A Arnedo-Pena (Castellón), Dr J Batlles-Garrido (Almeria), Prof A Blanco-Quirós (Valladolid), Dr RM Busquets (Barcelona), Dr I Carvajal-Urueña (Asturias), Dr G García-Hernández (Madrid), Prof L García-Marcos* (Cartagena), Dr C González Díaz (Bilbao), Prof F Guillén-Grima (Pamplona), Dr A López-Silvarrey Varela (A Coruña), Prof MM Morales Suárez-Varela (Valencia), Prof EG Pérez-Yarza (San Sebastián); Sri Lanka: Dr KD Gunasekera* (Sri Lanka); Sudan: Prof OAA Musa (Khartoum); Sultanate of Oman: Prof O Al-Rawas* (Al-Khod); Sweden:Dr H Vogt (Linköping); Syrian Arab Republic: Dr S Mohammad* (Tartous), Prof Y Mohammad (Lattakia), Dr K Tabbah (Aleppo); Taiwan: Dr J-L Huang* (Taipei), Dr C-C( Kao (Taoyuan); Thailand: Dr A Kongpanichkul (Nakorn Pathom), Dr R Nettagul (Chiangrai), Dr T Prasarnphanich (Chantaburi), Assoc Prof J Teeratakulpisarn (Khon Kaen), Assoc Prof M Trakultivakorn (Chiang Mai), Dr P Vichyanond* (Bangkok); Togo: Prof O Tidjani (Lome); Tokelau: Dr T Iosefa* (Tokelau); Trinidad and Tobago: Dr MA Monteil (St Augustine, Tobago); 152Tunisia: Prof M Jerray (Sousse), Prof F Khaldi (Grand Tunis); USA: Prof GJ Redding (Seattle), Dr HH Windom (Sarasota); Ukraine: Assoc Prof V Ognev* (Kharkiv, Rural Kharkiv); United Kingdom: Prof HR Anderson* (North Thames, South Thames), Dr JB Austin (Scotland), Dr M Burr (Wales), Prof D Strachan (Surry-Sussex); Uruguay: Dra D Holgado* (Montevideo), Dra MC Lapides (Paysandú); Venezuela: Dr O Aldrey* (Caracas); Vietnam: Dr B Vaên Cam (Ho Chi Minh City).* National CoordinatorISAAC National Coordinators not identified aboveCanada: M Sears; Channel Islands: HR Anderson; Chile: V Aguirre; Colombia: J Mallol interim); Croatia: V Ahel; Fiji: L Waqatakirewa; India: J Shah; Isle of Man: HR Anderson; Kingdom of Tonga: T Fakakovi; Malaysia: J de Bruyne; Netherlands: R Otten; Nouvelle Caledonie: S Barny; Polynesie Francaise: R Chansin; Republic of Ireland: P Manning; Republique Democratique du Congo: E Bahati; Reunion Island: C Catteau; Russia: RM Khaitov;Samoa: N Tuuau-Potai; Sudan: A El Sony; Sweden: L Nilsson.153Appendix C: Methods for assigning the Normalized Difference Vegetation Index and population density valuesNormalized Difference Vegetation Index valuesNDVI data for 2005 were downloaded from the Global Land Cover Facility via file transfer protocol (www.glcf.umiacs.umd.edu/data/gimms; (Pinzon et al. 2005; Tucker et al. 2005)). The global NDVI data are available at a spatial resolution of 0.07272 degrees (approximately 4.36 arcminutes or 8km at the equator). Rather than taking the single value underlying the study coordinate locations, the average of the nine surrounding values was taken. This was achieved in ArcGIS according to the following: • Converted one NDVI raster file to points to get central coordinates of all NDVI cells.• Ran the Analysis Tools -> Proximity -> Nearest Feature tool to get the NDVI cell ID closest to each coordinate ID. • Selected the specific NDVI cells IDs from all possible cells and created a new file. Note that 295 unique NDVI cell IDs were matched to 317 unique coordinate IDs, with some cells containing up to four sets of coordinates. • Used Analysis Tools -> Proximity -> Buffer to create a 0.07272*1.5 circular buffer around each NDVI cell centroid. • Used Data Management -> Features -> Feature Envelope to Polygon to generate square buffers (0.07272*3 each side).The raw NDVI GeoTIFF files were extracted from the compressed .gz format using the gunzip() function of the R.utils package in R. This process overwrites the compressed folder with an extracted folder with all of the contained information. The raw NDVI values were imported into R using the raster() function of the raster package, which preserves the geotagging. The ocean and missing values (-10000 and -5000 in the raw data, respectively) were converted to NA (the Rindicator for missing) and the remaining raw values were converted to NDVI values (ranging from -1 to 1) via the following: NDVI = raw/1000.154The extract() function of the raster package was used to get a list of the nine NDVI values in each of the NDVI square buffers. The values were averaged for each buffer and the results were merged with the unique identifiers for the study coordinates. The resulting file includes the unique study ID, the final latitude and longitude coordinates and an NDVI value for each period, where “feb05a” indicates the first half of February, 2005 and “jul05b” indicates the second half of July, 2005, etc. Note that values are null for all study sites on the small islands of Cook Islands, Fiji, French Polynesia, Niue, Samoa and Tokelau (10 of 317 sites). Population density values The gridded population of the world version 3 (GPWv3) population density data were downloaded from the Socioeconomic Data and Applications Center for 2005 (Socioeconomic Data and Applications Center 2004). The global GPWv3 data are available at a spatial resolution of 2.5 arcmintues (approximately 0.04167 degrees or 4.6km at the equator). The same process as described for the NDVI values above was applied to the population denisty values. The resulting file includes the unique study ID, the final latitude and longitude coordinates and a population density value (in persons per square kilometer), where “pop2005ndvisq” indicates the result fromthe NDVI square buffers in 2005, etc.155Appendix D: Descriptive statistics for the 13-14 age-group International Studyof Asthma and Allergies in Childhood centresCountry CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceAfricaAlgeria West Algiers 4203 0.07 27.96Cameroon Yaounde 2983 0.07 25.54Congo Brazzaville 1012 49.51 6.42Ethiopia Addis Ababa 3195 19.69 5.67Gabon Port-Gentil 3166 36.13 0.00Kenya Eldoret 3289 27.79 5.66Nairobi 3023 21.63 12.44Morocco Casablanca 1777 21.72 28.59Marrakech 1689 20.07 11.84Benslimane 1254 14.51 7.89Boulmene 1008 32.04 5.85Nigeria Ibadan 3142 35.65 0.00République Democratique du Congo Kinshasa 2930 14.71 3.72Reunion Island Reunion Island 2362 19.26 20.58South Africa Cape Town 5037 25.99 18.60Polokwane 4660 39.94 8.71Sudan Khartoum 2896 8.91 4.39Togo Lome 3090 21.17 11.62Tunisia Grand Tunis 6119 40.42 18.35Sousse 3042 45.40 18.28Asia-PacificChina Beijing 3530 0.48 28.33Guangzhou 3514 14.34 29.00Tibet 2878 0.87 3.41Tong Zhou 3542 4.15 6.72Wulumuqi 3884 8.32 18.95Hong Kong Hong Kong 3321 13.91 26.80Indonesia Bali 2569 26.00 5.10Bandung 2826 11.92 5.87Semarang 2435 10.27 4.31156Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceJapan Fukuoka 2520 15.40 27.98Tochigi 4466 3.56 26.76Malaysia Alor Setar 2941 24.18 17.68Klang Valley 3025 15.27 21.79Kota Bharu 2989 27.57 6.96Philippines Metro Manila 3658 6.83 3.66South Korea Provincial Korea 7375 9.26 19.43Seoul 2888 9.38 17.73Taiwan Taipei 6378 8.97 32.08Taoyuan 3190 14.83 27.18Thailand Bangkok 4669 24.54 33.78Chantaburi 2901 28.82 18.58Chiang Mai 3538 15.72 31.85Chiangrai 1809 12.38 15.15Khon Kaen 3410 17.30 22.84Nakorn Pathom 6975 16.72 12.40Vietnam Ho Chi Minh City 4240 39.15 25.54Eastern MediterraneanEgypt Cairo 3047 3.48 1.35Iran Birjand 2829 12.12 8.62Rasht 3004 0.37 25.60Tehran 3119 10.52 12.95Zanjan 2805 14.15 6.95Jordan Amman 2447 15.53 4.54Kuwait Kuwait 2882 19.99 6.14Malta Malta 4136 16.20 26.04Pakistan Islamabad 4069 13.03 21.82Karachi 2999 11.64 10.87Palestine North Gaza 3627 14.09 1.08Ramallah 3929 10.79 1.99Sultanate of Oman Al-Khod 3747 20.12 5.47157Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceSyrian Arab Republic Aleppo 3063 4.70 6.33Lattakia 3010 5.22 3.92Tartous 2995 1.64 1.50Indian subcontinentIndia Bangalore 3440 16.57 8.95Bikaner 3059 22.20 19.75Chandigarh 3122 14.64 6.47Davangere 2945 0.58 9.95Jaipur 3607 29.55 8.34Lucknow 3000 28.77 11.80Ludhiana 3108 24.00 18.89Chennai 2181 7.79 1.65Mumbai 1829 4.70 3.34Nagpur 4150 5.20 4.46New Delhi 3469 12.19 13.69Pimpri 3128 1.79 0.74Sri Lanka Sri Lanka 3717 17.73 6.91Latin AmericaArgentina Córdoba 3445 14.72 23.08Neuquén 3172 18.13 19.83Rosario City 3099 11.52 11.55Salta 3000 15.90 21.73Bolivia Santa Cruz 3257 31.29 11.30Brazil Aracaju 3043 24.71 9.53Belo Horizonte 3088 20.95 7.71Brasília 3009 23.63 11.43Caruaru 3026 18.21 9.25Curitiba 3628 8.68 14.72Feira de Santana 1732 28.29 13.28Itajaí 2737 15.93 8.62Maceió 2745 12.68 12.09Manaus Amazonas 3009 15.29 6.68Nova Iguaçu 3185 11.65 4.27158Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceBrazil cont'd Passo Fundo 2949 18.51 12.55Porto Alegre 3007 18.52 14.57Recife 2865 0.28 31.13Rural Santa Maria 3057 18.29 6.48Salvador 3020 19.50 27.28Santa Maria 3065 17.32 8.42Santo Andre 3232 0.03 25.77São Paulo 3161 13.67 11.80São Paulo West 3181 17.73 14.05Vitória da Conquista 1679 30.49 19.54Chile Calama 1618 18.73 17.06Chiloe 3000 18.40 16.40Punta Arenas 3044 0.43 26.87South Santiago 3026 18.31 18.54Valdivia 3105 16.55 25.44Colombia Barranquilla 3204 36.11 15.79Bogotá 3830 23.94 15.56Cali 3100 30.45 14.06Costa Rica Costa Rica 2436 28.94 29.72Cuba La Habana 3026 16.23 17.32Ecuador Guayaquil 3082 20.12 13.95Quito 3014 23.56 12.14El Salvador San Salvador 3260 40.95 10.34Honduras San Pedro Sula 2675 33.16 7.07Mexico Ciudad de México (1) 3891 26.21 13.03Ciudad de México (3) 3474 26.11 4.35Ciudad de México (4) 2662 29.56 7.89Ciudad Victoria 3122 19.25 13.45Cuernavaca 1431 8.67 7.69Mérida 3019 25.44 10.40Mexicali Valley 2988 30.86 9.40159Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceMexico cont'd Monterrey 3006 18.76 10.91Toluca 3021 5.96 3.08Villahermosa 3109 25.96 13.70Nicaragua Managua 3263 33.56 9.99Panama David-Panamá 3183 21.05 14.58Paraguay Asunción 3000 44.90 30.87Peru Lima 3022 0.10 17.97Uruguay Montevideo 3177 7.15 16.18Paysandú 1738 7.71 10.87Venezuela Caracas 3000 14.07 25.30North AmericaBarbados Barbados 2498 9.53 7.69Canada Vancouver 2853 14.72 22.99Trinidad and Tobago St Augustine 3512 11.10 22.61Tobago 1464 11.89 19.54USA Sarasota 1245 9.96 18.15Seattle 2422 15.32 17.22Northern and Eastern EuropeAlbania Tiranë 2983 12.67 3.96Bulgaria Sofia 1926 12.51 8.88Croatia Rijeka 2194 8.34 8.43Estonia Tallinn 3603 12.13 12.96Finland Kuopio County 3051 13.01 27.56Former Yugoslav Republic of Macedonia Skopje 3026 8.66 14.71Georgia Kutaisi 2650 7.17 6.94Hungary Svábhegy 4219 3.48 12.97Szeged 2889 3.98 9.69Kyrgyzstan Balykchi 1382 2.10 13.39Bishkek 5048 3.68 11.51Jalalabat 2404 22.42 6.03Latvia Riga 1283 8.81 9.74160Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceLithuania Kaunas 2723 8.01 9.33Panevezys 1187 14.15 19.63Siauliai 3516 12.71 12.37Poland Krakow 2545 9.59 25.85Poznan 1875 10.72 25.97Romania Cluj 3019 15.07 21.00Russia Novosibirsk 3769 16.95 15.47Serbia and Montenegro Belgrade 3228 11.59 9.32Nis 1207 18.64 13.67Novi Sad 1171 7.94 6.58Podgorica 1014 10.85 7.89Sombor 1105 11.76 3.08Sweden Linköping 2679 3.84 15.45Ukraine Kharkiv 2428 2.59 6.26Rural Kharkiv 3968 20.29 4.36OceaniaAustralia Melbourne 2192 16.42 20.39Cook Islands Rarotonga 445 24.27 2.47Fiji Suva 3093 31.52 10.67Kingdom of Tonga Nuku'alofa 2671 0.04 17.48New Zealand Auckland 2870 20.63 19.79Bay of Plenty 1976 13.56 17.26Christchurch 3116 11.07 17.36Nelson 2305 10.07 19.70Wellington 3050 16.95 28.72Niue Niue Island 79 22.78 18.99Nouvelle Caledonie Nouvelle Caledonie 7247 24.49 11.77Polynesie Francaise Polynesie Francaise 4289 20.73 10.89Samoa Apia 2986 11.82 2.18Tokelau Tokelau 66 34.85 13.64161Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceWestern EuropeAustria Urfahr-Umgebung 1439 7.44 11.88Belgium Antwerp 3250 9.72 25.69Channel Islands Guernsey 1248 12.74 18.91Jersey 773 11.64 16.30Germany Münster 4132 10.91 19.46Isle of Man Isle of Man 1716 16.84 21.45Italy Bari 1287 18.34 18.57Colleferro-Tivoli 1361 9.63 10.14Cosenza 925 6.49 8.76Emilia-Romagna 1347 13.29 19.90Empoli 1229 15.79 15.30Firenze 1383 12.65 18.51Mantova 1114 11.13 23.07Milano 1410 10.85 22.34Palermo 1221 11.96 18.18Roma 1325 8.83 26.94Siena 1082 11.74 24.77Torino 1180 12.37 21.36Trento 1311 4.35 11.82The Netherlands The Netherlands 6896 8.45 18.16Portugal Coimbra 1177 10.79 11.21Funchal 3161 10.16 10.09Lisbon 3024 12.10 15.31Portimao 1109 0.09 20.02Porto 3336 14.81 15.08Republic of Ireland Republic of Ireland 3089 13.05 20.78Spain A Coruña 2979 14.84 19.54Almeria 4051 8.49 14.76Asturias 4184 16.32 18.57Barcelona 3066 9.65 12.26Bilbao 3401 8.17 14.61Cartagena 3998 14.76 15.68162Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceSpain cont'd Castellón 4024 11.16 19.38Madrid 2652 12.14 14.86Pamplona 2932 9.89 13.27San Sebastián 1195 10.88 13.39Valencia 3132 12.77 11.11Valladolid 2944 15.18 15.29United Kingdom North Thames 2356 16.38 20.12Scotland 4662 12.91 22.31South Thames 2432 15.79 17.56Surrey-Sussex 5082 12.99 18.91Wales 2501 13.39 21.63163Appendix E: Descriptive statistics for the 6-7 age-group International Study ofAsthma and Allergies in Childhood centresCountry CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceAfricaNigeria Ibadan 2396 16.65 0.00South Africa Polokwane 3480 26.15 4.57Asia-PacificHong Kong Hong Kong 4448 13.02 23.40Indonesia Bandung 2503 12.74 2.36Japan Fukuoka 2958 7.98 18.09Malaysia Alor Setar 3786 10.17 3.86Klang Valley 3044 9.59 4.86Kota Bharu 3110 11.32 3.83South Korea Provincial Korea 4258 6.95 16.67Seoul 1760 7.05 18.58Taiwan Taipei 4832 6.21 31.42Taoyuan 3293 8.78 30.31Thailand Bangkok 4209 19.51 22.62Chantaburi 3321 18.85 14.39Chiang Mai 3106 15.04 13.17Chiangrai 1677 10.73 12.10Khon Kaen 2658 11.96 18.81Nakorn Pathom 1821 12.85 12.03Vietnam Ho Chi Minh City 3879 19.49 15.49Eastern MediterraneanIran Birjand 2693 7.98 4.68Rasht 3057 0.10 30.75Tehran 3008 3.36 4.89Zanjan 2777 17.21 8.64Jordan Amman 2598 11.97 4.35Malta Malta 3795 9.01 16.02Pakistan Islamabad 3966 5.77 12.36Karachi 2113 6.39 5.92164Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalencePalestine North Gaza 3575 11.78 2.66Ramallah 3754 8.98 2.66Sultanate of Oman Al-Khod 4130 10.94 3.95Syrian Arab Republic Lattakia 2373 3.75 5.44Tartous 2734 7.68 10.02Indian subcontinentIndia Davangere 3043 0.82 7.59Jaipur 2545 17.25 9.12Lucknow 3000 15.07 7.57Ludhiana 3225 9.64 7.26Mumbai 1833 3.66 3.76Nagpur 4294 3.89 3.31New Delhi 3706 7.37 9.15Pimpri 3838 4.22 3.31Sri Lanka Sri Lanka 3345 10.40 7.41Latin AmericaArgentina Neuquén 1930 11.35 16.48Rosario City 2952 6.30 9.79Brazil Aracaju 2443 13.10 8.64Itajaí 1511 11.05 9.79Maceió 1990 11.96 10.10Manaus Amazonas 3011 11.19 6.14Nova Iguaçu 3249 14.40 10.19Salvador 1069 11.51 26.47Santo Andre 2167 0.14 29.72São Paulo 3047 11.68 14.83São Paulo West 3312 16.06 13.35Chile Punta Arenas 3052 0.16 24.02South Santiago 3075 7.97 18.05Valdivia 3183 5.94 21.77Colombia Barranquilla 3209 17.23 13.24Bogotá 3256 15.14 14.28Cali 3005 20.30 14.08165Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceCosta Rica Costa Rica 3234 29.00 33.21Cuba La Habana 1803 13.59 20.91Ecuador Quito 3055 8.84 5.70El Salvador San Salvador 1365 27.69 10.40Honduras San Pedro Sula 1907 20.35 7.24Mexico Ciudad de México (1) 3205 33.35 9.64Ciudad de México (3) 3493 41.80 6.44Ciudad Victoria 2603 11.64 9.80Cuernavaca 2579 9.00 10.74Mérida 2896 16.23 19.68Mexicali Valley 2568 14.33 12.77Monterrey 3030 10.40 10.59Toluca 3235 10.39 7.54Villahermosa 2678 20.43 13.18Nicaragua Managua 3286 18.62 15.61Panama David-Panamá 2942 20.90 14.41Uruguay Paysandú 1512 6.75 11.11Venezuela Caracas 2999 8.64 21.94North AmericaBarbados Barbados 2759 5.15 4.49Canada Saskatoon 1255 3.43 21.35Northern and Eastern EuropeAlbania Tiranë 2896 8.87 5.18Bulgaria Sofia 1181 6.01 5.50Croatia Rijeka 1633 3.74 12.80Estonia Tallinn 2385 4.99 7.59Georgia Kutaisi 2666 3.19 5.14Hungary Svábhegy 2451 1.75 10.53Kyrgyzstan Bishkek 3146 2.29 7.95Jalalabat 1664 14.72 7.39Lithuania Kaunas 2772 8.30 11.22Panevezys 1176 7.65 13.69Siauliai 1341 8.65 11.93166Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalencePoland Krakow 2497 4.45 25.11Poznan 1999 3.75 22.41Russia Novosibirsk 2730 9.78 11.10Serbia and Montenegro Belgrade 1932 8.07 10.97Nis 1002 3.29 9.58Novi Sad 1044 6.51 6.99Sombor 1029 3.98 5.93Sweden Linköping 2089 1.58 10.29Ukraine Kharkiv 1950 2.31 6.21Rural Kharkiv 3000 15.57 4.87OceaniaAustralia Melbourne 2968 5.63 18.09New Zealand Auckland 3541 6.07 16.77Bay of Plenty 2150 4.65 18.00Christchurch 3315 3.59 19.16Nelson 1867 3.00 15.05Niue Niue Island 47 12.77 6.38Western EuropeAustria Kärnten 4847 2.58 7.49Urfahr-Umgebung 2029 2.61 9.46Belgium Antwerp 5645 3.21 12.31Germany Münster 3830 2.61 12.72Greece Thessaloniki 1228 10.26 4.97Isle of Man Isle of Man 1096 2.65 13.14Italy Bari 1943 7.05 12.09Colleferro-Tivoli 1143 2.97 12.51Emilia-Romagna 2265 4.06 12.54Empoli 1152 4.17 12.15Firenze 1036 4.15 13.32Mantova 1288 3.65 12.27Milano 2249 4.36 13.34Roma 2224 4.27 14.61Torino 2361 4.62 12.88167Country CentreNumber ofchildrenIntermittentrhinitis prevalencePersistent rhinitis prevalenceItaly cont'd Trento 2359 2.67 9.71Portugal Funchal 1819 8.80 11.16Lisbon 2477 8.48 15.46Portimao 1069 0.00 21.33Porto 2464 8.16 11.73Spain A Coruña 3016 6.60 13.66Almeria 3349 5.32 11.38Asturias 3193 7.23 12.18Barcelona 3002 2.73 5.26Bilbao 3157 5.99 10.96Cartagena 2948 5.90 9.77Castellón 3915 3.32 10.19Madrid 2347 10.01 11.55Pamplona 3176 3.21 6.77Valencia 3398 6.00 7.56168

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0135589/manifest

Comment

Related Items