IJE Advance Access originally published online on March 2, 2006
International Journal of Epidemiology 2006 35(3):779-786; doi:10.1093/ije/dyl022
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Association between reported exposure to road traffic and respiratory symptoms in children: evidence of bias
1 Swiss Paediatric Respiratory Research Group, Department of Social and Preventive Medicine, University of Berne, Switzerland
2 The Leicester Children's Asthma Centre, Division of Child Health, Department of Infection, Immunity & Inflammation, University of Leicester, Leicester, LE2 7LX, UK
* Corresponding author. Department of Social and Preventive Medicine, Finkenhubelweg 11, CH-3012 Bern, Switzerland. E-mail: kuehni{at}ispm.unibe.ch
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background Many studies showing effects of traffic-related air pollution on health rely on self-reported exposure, which may be inaccurate. We estimated the association between self-reported exposure to road traffic and respiratory symptoms in preschool children, and investigated whether the effect could have been caused by reporting bias.
Methods In a random sample of 8700 preschool children in Leicestershire, UK, exposure to road traffic and respiratory symptoms were assessed by a postal questionnaire (response rate 80%). The association between traffic exposure and respiratory outcomes was assessed using unconditional logistic regression and conditional regression models (matching by postcode).
Results Prevalence odds ratios (95% confidence intervals) for self-reported road traffic exposure, comparing the categories ‘moderate’ and ‘dense’, respectively, with ‘little or no’ were for current wheezing: 1.26 (1.13–1.42) and 1.30 (1.09–1.55); chronic rhinitis: 1.18 (1.05–1.31) and 1.31 (1.11–1.56); night cough: 1.17 (1.04–1.32) and 1.36 (1.14–1.62); and bronchodilator use: 1.20 (1.04–1.38) and 1.18 (0.95–1.46). Matched analysis only comparing symptomatic and asymptomatic children living at the same postcode (thus exposed to similar road traffic) showed similar ORs, suggesting that parents of children with respiratory symptoms reported more road traffic than parents of asymptomatic children.
Conclusions Our study suggests that reporting bias could explain some or even all the association between reported exposure to road traffic and disease. Over-reporting of exposure by only 10% of parents of symptomatic children would be sufficient to produce the effect sizes shown in this study. Future research should be based only on objective measurements of traffic exposure.
Keywords Child, preschool, asthma, cough, vehicle emissions, bias, epidemiological methods, questionnaires
Accepted 30 January 2006
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Although the individual health risks of air pollution are small its public-health consequences are substantial.1,2 The recent reduction in classical air pollutants such as SO2 or NO2 has masked an ongoing increase in exhaust emissions from road traffic, a complex mixture of pollutants that are not all individually measured. Several authors have, therefore, investigated whether exposure to road traffic is associated with respiratory illness. Their findings have been contradictory, some reporting considerable effects3–7 while others found small or no effects.8–12 One reason for these discrepancies could be varying misclassification of exposures, especially where self-reported exposure was used.3–6
Although population mean estimates of air pollution correlate with objective measures,13–15 individual estimates vary widely and are associated with a number of factors in addition to measured air pollution.14,16 Most important is the possible over-reporting of exposure by symptomatic participants, because of the publicity given to air pollution and respiratory health. Such differential reporting would tend to exaggerate any association between exposure and disease.17 Heinrich et al. have recently shown that self-reported and modelled assessment of exposure to air pollution are only weakly associated, but the possibility of reporting bias by symptom status has not been investigated in studies on respiratory symptoms in children, the most common subpopulation involved in air pollution studies, and its impact has not been quantified. Our aim was to investigate in a large population-based survey: (i) whether parent-reported road traffic density at home was associated with the prevalence of respiratory symptoms in pre-school children and (ii) whether any such association could be explained by biased reporting of traffic density.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Population and study design
In April 1998 we sent a respiratory questionnaire to a random sample of 8700 children aged 1.00–4.99 years, born and resident in Leicestershire, UK, using the Leicestershire Health Authority child health database as the sampling frame (Table 1). Parents were told that we were interested in coughs, colds, wheezes, and allergies in young children but not told that we had an interest in road traffic and air pollution. The methods of this survey have been reported in detail elsewhere.18,19 South Asians, the largest ethnic minority group in the UK were oversampled. The Leicestershire Health Authority Research Ethics Committee approved the study.
|
Questionnaire
The questionnaire was designed in 1990 for use in pre-school children and was slightly adapted by adding core questions from the International Study of Asthma and Allergies in Childhood (ISAAC).20,21 It included questions on the 12 month period prevalence of wheeze, doctor-diagnosed asthma, bronchodilator use, night cough, and chronic rhinitis. ‘Possetting or vomiting’ in the first year of life was included as a symptom not expected to be related to air pollution in the minds of the children's parents. Exposure to road traffic was assessed by the question, ‘How would you describe the location of your house: (i) in a street with very dense traffic (main road); (ii) in a street with moderate traffic (residential road); (iii) in a quiet street with little or no traffic’. The questionnaire also included sections on socio-demographic conditions (parental education, single parents, overcrowding), family history of atopic disease, and a number of known or suspected environmental risk factors for respiratory disease.
Analysis
Statistical analysis was performed using STATA, version 8.0 for Windows (STATA Corporation, TX, USA). First, we investigated whether reported traffic was associated with increased prevalence of respiratory outcomes in all responders (n = 6811; Table 2). Relevant outcomes were wheeze, night cough, chronic rhinitis, and bronchodilator use in the past 12 months, doctor-diagnosed asthma ever, and possetting in the first year of life, with results expressed as proportions and odds ratios (ORs) with 95% confidence intervals (95% CIs), comparing symptom prevalence between exposure categories. Chi-square tests for trend and likelihood ratio tests were used to assess evidence of association. Then, the following putative confounders were entered one by one into the model and those changing the OR perceptibly were included in the final regression models: ethnicity, place of residence, age and sex of the child, paternal and maternal smoking, gas cooking, gas heating, household pets, number of siblings, enrolment in nursery school or day-care, breastfeeding, parental education, and single parenthood.
|
The association between reported traffic exposures and symptoms that we found might be due to (i) a true causal association between traffic-related air pollution and health, (ii) over-reporting of traffic exposure by families of symptomatic children, or (iii) a combination of the two. To investigate these possibilities, we matched the children by postcode, assuming that within this small area (a 7-digit postcode covers up to 15–16 dwellings) the true exposure to road traffic would be very similar and, therefore, the OR of the effect of traffic exposure, comparing symptomatic with asymptomatic children should be 1.0. Any remaining association between symptoms and reported traffic exposure in this matched analysis, using a conditional logistic regression model, would therefore suggest over-reporting of exposure by parents of symptomatic children.
For this analysis, we could use only data from postcodes where at least one symptomatic child and one asymptomatic child were living. As these subgroups differed from the total sample, including more families living in an inner city area and more south-Asian children (Table 1), we analysed the subgroup in two ways. First, we performed an analysis ignoring the matching (logistic regression) to assess whether the association between traffic and symptoms was similar to the one obtained in the full study population. Second, we performed a conditional logistic analysis that accounted for the matching on postcode to assess whether the association persisted once matching on postcode was appropriately accounted for. For that we compared the ORs from this matched analysis with the results of an unmatched analysis within the same subpopulation (Table 3). We tested these models for effect modification by including interaction terms.
|
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The response rate, after discounting 200 invalid addresses, was 80% (6811/8500) and there were <2% missing answers for most questions. Forty-one per cent of the children reported little or no traffic, 48% moderate, and 11% dense traffic at their home address (Table 1).
Results from the whole study population (n = 6811)
Prevalence of reported wheeze, asthma diagnosis, bronchodilator use, night cough, and chronic rhinitis was higher in children reported as living on roads with moderate and dense traffic compared with those reporting little or no traffic (Table 2). Adjustment for a large number of confounders did not change these findings. Comparing the categories ‘moderate’ and ‘dense’ traffic exposure, respectively, with ‘little or no’, the ORs (95% CIs) for current wheezing were 1.26 (1.13–1.42) and 1.30 (1.09–1.55); for asthma diagnosis 1.29 (1.11–1.50) and 1.14 (0.90–1.45); for bronchodilator use 1.20 (1.04–1.38) and 1.18 (0.95–1.46); for night cough 1.17 (1.04–1.32) and 1.36 (1.14–1.62); and for chronic rhinitis 1.18 (1.05–1.31) and 1.31 (1.11–1.56). Possetting in the first year of life was not related to reported traffic density.
Results from the subgroup, where children could be matched by postcode (n = 1660)
Depending on the prevalence of the different symptoms, a varying number of postcodes including at least one symptomatic child and one asymptomatic child were used for the matched analysis. These were: for wheeze 627 areas with 1660 children, for asthma diagnosis 396 areas with 1047 children, for bronchodilator use 420 areas with 1147 children, for night cough 615 areas with 1662 children, for rhinitis 698 areas with 1832 children, and for possetting 379 areas with 989 children.
Using unconditional logistic regression analysis in these subgroups, the strength of the association between traffic exposure and outcomes was similar to that in the total study population, although CIs for the ORs included 1 in most cases, owing to the lower statistical power (Table 3, unmatched analysis).
Conducting a conditional logistic regression analysis after matching the children by postcode (Table 3, matched analysis) resulted in equal or even larger effects than in the unmatched analysis. The adjusted ORs (95% CIs), comparing children exposed to dense traffic with those exposed to little traffic, were 1.40 (0.88–2.23) for wheeze, increasing to 1.90 (1.06–3.42) for bronchodilator use and 2.26 (1.22–4.21) for asthma diagnosis, two features associated with more severe wheeze. ORs for night cough and rhinitis were smaller [1.33 (0.85–2.08) and 1.40 (0.91–2.17), respectively]. This implies that parents of children with more severe respiratory problems are particularly prone to overestimate traffic exposure.
We did not find consistent evidence of an effect modification, which would suggest more misclassification in subpopulations defined by paternal or maternal education, parental smoking, parental asthma, or living in an inner city, but statistical power for performing interaction tests was low.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In this population-based survey of pre-school children prevalence of respiratory symptoms, bronchodilator use, and asthma diagnosis were associated with reported exposure to road traffic, even after controlling for a large number of confounders. When we repeated the analysis after matching the children by postcode, an objective marker for comparing traffic exposure, the strength of associations remained similar or increased, especially for those with more severe symptoms. This suggests that the parents of children with respiratory symptoms over-reported their children's exposure to road traffic or that a third unmeasured factor, like ‘negative affectivity’,22 was present that led families to over-report both respiratory symptoms and traffic exposure.
Methodological considerations
The strengths of this study include its population-based sampling strategy, large sample size, good response rate, and inclusion of large numbers of South Asians, the largest group of ethnic minority population in the UK. Our results are, therefore, likely to be representative for the UK. The full postcode allowed us to allocate children's houses to small geographic areas, covering up to 15 dwellings. We assumed that true domiciliary exposure to traffic-related air pollution was uniform within a single postcode. Although there are certainly situations where traffic exposure might vary within a postcode, owing to increasing horizontal distance from a major road or differing vertical distance in multi-storey buildings the difference should be less within postcodes than between postcodes. Therefore, the ORs should be lower in the matched analysis compared with the unmatched analysis. This was not the case, the strength of association was similar in the matched and the unmatched analysis, implying that most or even all of the associations found in this study might be explained by biased over-reporting of traffic density by parents of symptomatic children.
A limitation of the study was the low statistical power in the matched analysis, due to the fact that the sampled children were dispersed over a large area. Only a quarter of the study families could thus be used for the matched analysis.
Comparison with other studies
Heinrich et al., using data from Dutch and German cohorts, have recently shown that self-reported and modelled assessments of exposure are only weakly associated. They did not, however, analyse their data by symptom status of the participating children.15
Other studies on road traffic and respiratory symptoms in children using self-reported traffic exposure (including our own unmatched data) tended to find larger effects than surveys relying on objective measurements. For instance, ORs for current wheeze in children, contrasting the categories ‘frequent’ and ‘constant’ truck traffic with ‘never’, were 1.53 and 2.15 in a survey of 12- to 15-year old children in Münster and 1.53 and 1.67 in 13- to 14-year olds in Bochum, Germany (Table 4).4,6 Hirsch et al., in 5421 children aged 5–11 years, found an OR of 2.09 for wheeze, comparing ‘constant’ with ‘no’ truck traffic, while they did not find an association between wheeze and objectively measured exposures to a number of traffic-related air pollutants.5 Studies using distance to the main road or traffic counts as exposures generally found smaller or no effects.8–10,12,23 For rhinitis, we found insufficient studies using measured exposure to draw any conclusions.
|
Implications for future research
Our findings suggest that systematic over-reporting of exposure to road traffic by families of symptomatic children might have led to biased effect estimates and could explain some or all of the exposure-disease association in our study. The size of this bias varied for the different health outcomes; while it was not noticed for the non-respiratory symptom of ‘possetting’, it was intermediate for night cough and rhinitis and largest for bronchodilator use and the diagnosis of asthma, which have received the broadest media coverage with regard to air pollution. Also, diagnosis or medication given by doctors might induce parents to regard the symptoms more seriously and thus to attribute (or misattribute) causes to these problems.
Although it has often been hypothesized that reporting bias might play a role in assessment of effects of air pollution, this has never been shown for studies on respiratory symptoms in children. The different results for different symptoms suggest that public concern about health effects of air pollution plays an important role. The extent of this bias is, therefore, likely to vary between regions and time periods, so that results from one study cannot be extrapolated to other situations. For instance, a population-based survey in Italy, where information about respiratory disorders and traffic near residences was collected by questionnaire, could evaluate the possibility of reporting bias by matching a subsample of cases and controls by address code. In this study, the raw association between case–control status and reported frequency of lorry traffic was 1.12, decreasing to 1.04 in the matched analysis, suggesting no systematic difference in traffic reporting between parents of symptomatic and asymptomatic children.3
With a prevalence of exposure to moderate or high traffic of 60% in our study, a relatively small proportion (10 and 20%, respectively) of families with symptomatic children falsely reporting high traffic exposure would be sufficient to bias the OR from 1.0 (no effect) to 1.2 and 1.4, respectively, effect sizes typically reported in epidemiological studies (Figure 1). With 38% of parents falsely reporting high exposure, the OR would be 2.0.
|
Our data illustrate that random errors (quantified with P-values or CIs) and confounding, the two main issues that are usually dealt with in epidemiology, are not the only threats to valid inference and in fact might be dwarfed by systematic errors such as biased reporting. Systematic errors unfortunately are not routinely considered in the interpretation of research results.17 Our findings parallel what in more traditional case–control studies, for example in childhood cancer and its association with antenatal risk factors, is termed ‘recall bias’ if exposures are assessed retrospectively or contemporaneously with the health outcome.
In conclusion, after matching for postcode our results provide evidence that most if not all of the association between reported road traffic and respiratory symptoms in this survey of pre-school children could be the result of a reporting bias. These findings point out that self-reported exposure to road traffic is unreliable and of limited use in aetiological research.
KEY MESSAGES
|
|
Acknowledgments |
|---|
The authors would like to thank all the parents for filling in the questionnaire, Anthony Davis, Business Manager, Children's Directorate, Leicester City West Primary Care Trust, for his assistance and Nicola Low for very helpful comments on previous versions of this manuscript. The work presented in this paper was funded by the Swiss National Science Foundation (PROSPER grant 3233-069348 and 3200-069349, and SNF grant 823B-046481) and the Swiss Society of Paediatrics (Glaxo-Smithkline Scholarship for Paediatric Pulmonology 2001). Initial data collection was supported by a research grant from Trent NHS Executive (Trent Research Scheme, RBF # 98XX3).
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Kunzli N, Ackermann-Liebrich U, Brandli O, Tschopp JM, Schindler C, Leuenberger P. Clinically "small" effects of air pollution on FVC have a large public health impact. Swiss Study on Air Pollution and Lung Disease in Adults (SAPALDIA)-team. Eur Respir J 2000;15:131–36.[Abstract]
2 Kunzli N, Kaiser R, Medina S, Studnicka M, Chanel O, Filliger P et al. Public-health impact of outdoor and traffic-related air pollution: a European assessment. Lancet 2000;356:795–801.[CrossRef][Web of Science][Medline]
3 Ciccone G, Forastiere F, Agabiti N, Biggeri A, Bisanti L, Chellini E et al. Road traffic and adverse respiratory effects in children. SIDRIA Collaborative Group. Occup Environ Med 1998;55:771–78.
4 Duhme H, Weiland SK, Keil U, Kraemer B, Schmid M, Stender M et al. The association between self-reported symptoms of asthma and allergic rhinitis and self-reported traffic density on street of residence in adolescents. Epidemiology 1996;7:578–82.[Web of Science][Medline]
5 Hirsch T, Weiland SK, von Mutius E, Safeca AF, Grafe H, Csaplovics E et al. Inner city air pollution and respiratory health and atopy in children. Eur Respir J 1999;14:669–77.[Abstract]
6 Weiland SK, Mundt KA, Ruckmann A, Keil U. Self-reported wheezing and allergic rhinitis in children and traffic density on street of residence. Ann Epidemiol 1994;4:243–47.[Medline]
7 Brunekreef B, Janssen NAH, De Hartog J, Harssema H, Knape M, Van Vliet P. Air pollution from truck traffic and lung function in children living near motorways. Epidemiology 1997;8:298–303.[CrossRef][Web of Science][Medline]
8 Venn A, Lewis S, Cooper M, Hubbard R, Hill I, Boddy R et al. Local road traffic activity and the prevalence, severity, and persistence of wheeze in school children: combined cross sectional and longitudinal study. Occup Environ Med 2000;57:152–58.
9 Nicolai T, Carr D, Weiland SK, Duhme H, von Ehrenstein O, Wagner C et al. Urban traffic and pollutant exposure related to respiratory outcomes and atopy in a large sample of children. Eur Respir J 2003;21:956–63.
10 van Vliet P, Knape M, de Hartog J, Janssen N, Harssema H, Brunekreef B. Motor vehicle exhaust and chronic respiratory symptoms in children living near freeways. Environ Res 1997;74:122–32.[Medline]
11 Wilkinson P, Elliott P, Grundy C, Shaddick G, Thajrar B, Walls P et al. Case-control study of hospital admission with asthma in children aged 5–14 years: relation with road traffic in north west London. Thorax 1999;54:1070–74.
12 Lewis SA, Antoniak M, Venn AJ, Davies L, Goodwin A, Salfield N et al. Secondhand smoke, dietary fruit intake, road traffic exposures, and the prevalence of asthma: a cross-sectional study in young children. Am J Epidemiol 2005;161:406–11.
13 Forsberg B, Stjernberg N, Wall S. People can detect poor air quality well below guideline concentrations: A prevalence study of annoyance reactions and air pollution from traffic. Occup Environ Med 1997;54:44–48.
14 Oglesby L, Kunzli N, Monn C, Schindler C, Ackermann-Liebrich U, Leuenberger P. Validity of annoyance scores for estimation of long term air pollution exposure in epidemiologic studies: the Swiss Study on Air Pollution and Lung Diseases in Adults (SAPALDIA). Am J Epidemiol 2000;152:75–83.
15 Heinrich J, Gehring U, Cyrys J, Brauer M, Hoek G, Fischer P et al. Exposure to traffic related air pollutants: self reported traffic intensity versus GIS modelled exposure. Occup Environ Med 2005;62:517–23.
16 Hunter PR, Bickerstaff K, Davies MA. Potential sources of bias in the use of individual's recall of the frequency of exposure to air pollution for use in exposure assessment in epidemiological studies: a cross-sectional survey. Environ Health 2004;3:3.[CrossRef][Medline]
17 Phillips CV. Quantifying and reporting uncertainty from systematic errors. Epidemiology 2003;14:459–66.[Web of Science][Medline]
18 Kuehni CE, Davis A, Brooke AM, Silverman M. Are all wheezing disorders in very young (preschool) children increasing in prevalence? Lancet 2001;357:1821–25.[CrossRef][Web of Science][Medline]
19 Pierse N, Rushton L, Harris RS, Kuehni CE, Silverman M, Grigg J. Locally-generated particulate pollution and respiratory symptoms in young children. Thorax, doi:10.11136/Thx.2004.036418.
20 Luyt DK, Burton PR, Simpson H. Epidemiological study of wheeze, doctor diagnosed asthma, and cough in preschool children in Leicestershire. BMJ 1993;306:1386–90.
21 Asher MI, Keil U, Anderson HR, Beasley R, Crane J, Martinez F et al. International Study of Asthma and Allergies in Childhood (ISAAC): rationale and methods. Eur Respir J 1995;8:483–91.[Abstract]
22 Van Diest I, De Peuter S, Eertmans A, Bogaerts K, Victoir A, Van den Bergh O. Negative affectivity and enhanced symptom reports: differentiating between symptoms in men and women. Soc Sci Med 2005;61:1835–45.[Medline]
23 Venn AJ, Lewis SA, Cooper M, Hubbard R, Britton J. Living near a main road and the risk of wheezing illness in children. Am J Respir Crit Care Med 2001;164:2177–80.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. D. Spycher, M. Silverman, M. Egger, M. Zwahlen, and C. E. Kuehni Routine Vaccination Against Pertussis and the Risk of Childhood Asthma: A Population-Based Cohort Study Pediatrics, March 1, 2009; 123(3): 944 - 950. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G Cesaroni, C Badaloni, D Porta, F Forastiere, and C A Perucci Comparison between various indices of exposure to traffic-related air pollution and their impact on respiratory health in adults Occup. Environ. Med., October 1, 2008; 65(10): 683 - 690. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. N. Krivoshto, J. R. Richards, T. E. Albertson, and R. W. Derlet The Toxicity of Diesel Exhaust: Implications for Primary Care J Am Board Fam Med, January 1, 2008; 21(1): 55 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. E Kuehni, A. M Brooke, M.-P. F Strippoli, B. D Spycher, A. Davis, and M. Silverman Cohort Profile: The Leicester Respiratory Cohorts Int. J. Epidemiol., October 2, 2007; (2007) dym090v1. [Full Text] [PDF]
|
|
|
|
|
|
V Morgenstern, A Zutavern, J Cyrys, I Brockow, U Gehring, S Koletzko, C P Bauer, D Reinhardt, H-E Wichmann, and J Heinrich Respiratory health and individual estimated exposure to traffic-related air pollutants in a cohort of young children Occup. Environ. Med., January 1, 2007; 64(1): 8 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Michel, M. Silverman, M-P. F. Strippoli, M. Zwahlen, A. M. Brooke, J. Grigg, and C. E. Kuehni Parental understanding of wheeze and its impact on asthma prevalence estimates Eur. Respir. J., December 1, 2006; 28(6): 1124 - 1130. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||