Risk of childhood asthma prevalence attributable to residential proximity to major roads in Montreal, Canada.

Price, Karine ; Plante, Celine ; Goudreau, Sophie 等

Vehicle traffic emissions result in a complex mixture of carbon monoxide (CO), nitrogen oxides (NOx), particulate matter (mainly ultra-fine particles) and air toxics (1,3-butadiene, benzene, formaldehyde, polycyclic aromatic hydrocarbons-PAHs). For the province of Quebec, approximately 75% of NOx and 14% of [PM.sub.2.5] (particulate matter with median diameter of less than 2.5 yum) emissions are related to mobile sources. (1)

Epidemiological studies that evaluate the relation between long-term exposure to traffic emissions and health outcomes have been reviewed in a number of studies. (2-6) Exposure to traffic-related air pollutants has been linked to premature mortality. (7-10) Epidemiological studies have also linked traffic pollutant exposure to cardiovascular outcomes like myocardial infarction. (11-13) Birth cohort, case-control and cross-sectional studies have also been performed to assess childhood exposure to traffic emissions and incidence and prevalence of asthma and asthma-like symptoms. (3,14-16)

According to the Health Effects Institute (HEI) review of the literature, of the many outcomes of traffic-related exposures, asthma in children has been judged to have between sufficient and suggestive evidence to support a causal relationship with traffic exposures. (3) Furthermore, exacerbation of respiratory symptoms in children with asthma was judged sufficient to infer a causal association with traffic exposure. The increasing prevalence and incidence of asthma in Canadian cities, (17) especially in children, is an issue that requires further attention. Computation of asthma risks attributable to exposure to traffic-related pollutants can thus help orient policies and interventions related to transportation and public health.

The focus of this study was an attempt to assess the range of prevalent asthma cases in children attributable to residing near a major road or highway on the Island of Montreal.

METHODS

Initial scoping of health risk functions

The study selection on asthma prevalence and proximity to major roads or highways was based on the previous literature review performed by a panel of experts from the HEI report. (3) We selected this review because i) it is a comprehensive update of the latest studies published in relation to traffic pollution and health effects, and ii) a comprehensive approach to assess causality was applied. (18,19)

We selected studies from the HEI report if they related to asthma prevalence in children and if the exposure of interest was a dichotomous variable for proximity to a major road or highway (e.g., less than 50 or 75 m from a major road or a highway), since information on these parameters was available for further application to Montreal. Here, we did not assess incident cases attributable to this exposure as too few studies were available on asthma incidence and proximity to major roads or highways (less than 75 m). (3) We also did not assess exacerbation of respiratory symptoms in children with asthma since studies that related this outcome and proximity to major roads or highways were not available. Attributable fractions were computed using studies performed in locations other than Montreal, given the lack of published studies of traffic proximity and asthma prevalence in Montreal.

Based on the HEI report performed for studies from 1980 through October 2008, a total of six studies evaluated the association between proximity to major roads and prevalence of asthma in children (within 75 m of a roadway). (3,15,20-24) Only four studies from the Unites States and Europe were retained for this analysis (see Figure 1). (15,20-22) One study was not selected based on reservations put forward in the HEI report concerning study design issues and lack of control for confounders, making the study hard to interpret. (24) Another study was not retained for our assessment since it reported inconclusive results. (23) It should be mentioned that we also decided not to retain a positive but non-significant association between roadway proximity (<50 m from major roads or highways) and asthma prevalence in children aged 4 to 12 years from Montreal (unpublished results: OR: 1.09; 95% CI: 0.91-1.31, n=5,650).

Selected risk functions

The chosen four studies from which risk functions were obtained are listed in Table 1, accompanied by the study characteristics (epidemiological design, health outcome studied, reported risks, exposure categories and demographic characteristics of the study group). All retained studies had been adjusted for confounding and modifying factors and their ORs were used to calculate the number of cases attributable to residing in proximity to major roads. Point estimates between studies are similar and confidence intervals overlap, hence in these studies, the impact appears to be similar across study design and different age groups. Although a meta-analysis would provide more accurate results, too few studies are available for the same age groups and the same proximity to roadway variables to perform a meta-analysis.

Exposure: Proximity to major roads or highways

Exposure of Montreal children to traffic-related pollutants was estimated by using proximity of residences to major roads or highways. We estimated the location of residential addresses using the six-digit postal codes. In Montreal, a six-digit postal code usually corresponds to a city block side. Following the selection of studies, different traffic exposure scenarios (e.g., <50 m or <75 m from a major road or highway) were considered (see Table 1). The road traffic categories nearest to residential addresses were obtained through the DMTI database (DMTI Spatial Inc.). In this study, we used two categories: major roads and highways (which includes the DMTI classes of secondary and principal highway, expressway and major roads) as well as highways only (which includes the DMTI classes of secondary and principal highway and expressway). The major roads and highways category in Montreal had a traffic density that was superior to approximately 20,000 vehicles/day and was used when calculating attributable asthma cases based on the studies from McConnell et al., 2006, (15) Morgenstern et al., 2007 (20) and 2008.21 The highways only category corresponds, in Montreal, to a traffic density that is superior to approximately 150,000 vehicles/day; it was used to calculate attributable prevalent asthma cases based on the study by Kim et al., 2008. (22)

[FIGURE 1 OMITTED]

Population at risk and health data

In order to apply the risk functions presented in Table 1 to Montreal, the following information was also estimated:

i) population of children residing near major roads and highways in Montreal for different age groups;

ii) population of asthmatic children in different age groups of the chosen risk functions.

The population of children residing near major roads and highways was estimated using the population numbers assigned to the centroid location of six-digit postal codes of 1996. Population numbers for six-digit postal codes were not available for more recent years. In 1996, there were 1,797,912 people residing on the Island of Montreal. (25) In 2006, this population rose to 1,873,589 people. Given the limited increase in population numbers, we did not adjust the 1996 value.

In order to determine the proportion of children from each age group of the chosen risk functions (2, 4, 6, 5 to 7 and 8 to 10), we determined the percentage of children in 2006 in each age group (relative to the entire 2006 population) living on the Island of Montreal and applied the percentages to the population in each road traffic category. (25)

For the purpose of comparison, the prevalence of asthma was taken from two studies using different approaches: one performed on the Island of Montreal in 2006 (26) (questionnaire-based approach) and another performed in the province of Ontario, Canada from 1996 to 2005 (27) (based on diagnoses in health care service databases). For the asthma prevalence reported by questionnaire, results were cross-checked with the Medicare database (reported doctor diagnosis). Both sources of data reported similar prevalence rates for two or more asthma diagnoses in children's health files. For the Montreal prevalence study, the definition of asthma was the one used in the 1995-1996 Health Canada Student Lung Health Survey.

Estimation of health impacts

Using the information concerning population numbers for different age groups and proximity categories to major roads and highways for Montreal, corresponding odds ratios for the different categories were applied to Montreal. The attributable fraction (AF) estimation of asthma prevalence related to proximity to major roads or highways was calculated as follows: (28)

[AF.sub.i] = [p.sub.ei] ([RR.sub.i] - 1)/[RR.sub.i]

Equation 1

[p.sub.ei] = proportion of exposed cases at age "i" (living within 50 or 75 m of major roads or highways). The prevalence of asthma in the exposed population was not available to calculate this value. We assumed the prevalence of asthma in the exposed population to be equal to the prevalence of asthma in the whole population.

[RR.sub.i] = Relative risk (Prevalence ratio) for age group "i"

The attributable number (AN) of prevalent asthma cases due to residing near major roadways was determined as follows for each age group: (29)

[AN.sub.i] = [Pop.sub.i] * Po * [AF.sub.i]

Pop=population of children of age group "i"

Po=prevalence of asthma in age group "i"

Values of RR (or Prevalence ratios, PR) were not reported in the selected studies; only values of ORs were reported. In the case of diseases with very low incidence among the non-exposed, OR and RR (PR) are likely to be very similar. However, for outcomes that are not so rare, such as asthma, the ORs and RRs can differ. (30) In order to account for these differences, we estimated the RR (PR) as follows, using equation 2: (30)

RR = OR/(1 - [P.sub.o]) + ([P.sub.o] x OR) Equation 2

[P.sub.o] = prevalence of asthma in the initial studies

Equations 1 and 2 were then applied to the different exposure scenarios presented in Table 2, so as to obtain a range of attributable cases, based on the different risk functions that were selected. The confidence interval for the number of cases was calculated based on the 95% CI of the ORs, as reported in the selected studies (Table 2), assuming normal distribution of the error on the OR.

RESULTS

Based on the chosen risk functions, there were 5 age groups for which information was calculated (2, 4, 6, 5 to 7 and 8 to 10). Populations of children residing at different distances for each age category were estimated to range in Montreal from 650 at <75 m for the highway only category, and from 2,830 at <50 m to 11,809 at <75 m for the major road and highway category.

The numbers of prevalent cases that may be attributable to proximity of residences to major roads or highways were calculated using prevalence from the questionnaire-based Montreal study (26) and prevalence from the Ontario study. (27) Attributable cases using Montreal prevalence ranged from 44 for the 2-year age group at <50 m to 636 in the 5 to 7 year age group residing within 75 m of a major road or highway (Table 2). Attributable cases using the Ontario asthma prevalence ranged from 69 for the 2-year age group at <50 m to 778 in the 5 to 7 year age group residing within 75 m of a major road or highway (Table 2). Depending on the exposure scenario and prevalence rates used, we estimated that exposure to proximity to major roads and highways could have an impact of 0.7% to 6.4% of prevalent asthma cases in Montreal.

DISCUSSION

This study is an attempt to quantify the risk of asthma prevalence in children that is attributable to living near major roads or highways on the Island of Montreal, Canada. The results presented here should be seen as a potential range of attributable cases, as other factors not considered may also be responsible for the disease. Attributable fractions may be lower or higher than those calculated, as there are uncertainties in calculations and competing risk factors that we have not considered when applying risk functions to our specific population (such as environmental tobacco smoke). Overall health risks attributable to proximity to roads are, however, more important than the ones presented in this study where only asthma-prevalent cases are presented. Other studies have shown that exposure to traffic emissions is linked to several health effects not reported in this study, such as premature mortality and cardiovascular outcomes. (3)

Applying risk functions in populations other than the initial study population has also been performed by other authors to calculate prevalent asthma cases attributable to proximity to major roads. (31,32) In these studies, which were applied to Southern California, the percentage of prevalent asthma cases attributable to residing in proximity to major roads ranges between 6% and 9.3%. These fractions were estimated using the study by McConnell et al., 2006, (21) which was one of the studies retained for our assessment. Here, based on the ORs of the four studies that we retained, we report attributable fractions lower than the ones estimated for Southern California (range 0.7 to 6.4%), probably due to various factors like differences in ORs, prevalence rates and children's population within 50 or 75 m of major roads or highways. (15,20-22)

It should be stressed that attributable risk is only valid if causality between the exposure and disease exists. (28) We retained a health outcome (prevalence of asthma) for which a causal association due to exposure to traffic pollutants has been judged to be between sufficient and suggestive, but uncertainties regarding causality still remain since not all studies performed in different locations and population subgroups have shown consistent results (e.g., greater risks have been found in girls in some studies, but not in others (21,22)). The size of the differences in risk coefficients between studies in different locations can also be considerable. Thus, the study by Kim et al., 2008, has larger OR and confidence intervals since the study was performed with a very limited number of children residing within 75 m of a highway. (22)

In our assessment, the chosen studies had both strengths and weaknesses for the purpose of transfer of risk functions to the Montreal context. On one hand, risk functions used were adjusted for many confounding or modifying factors. On the other hand, the exposure response functions used for proximity to roadways were not developed specifically for Montreal children, and their applicability in other regions than the one under study (the Island of Montreal) is uncertain. Different locations may present differences in pollutant levels and population characteristics. Pollutant levels may differ greatly on major roadways between countries since characteristics of vehicle fleet are likely different (automobile, truck, and number of diesel engines) as well as the composition of the emission mixture (based on motor emission standards). Background levels of pollutants may also be different in other countries compared to those in Montreal.

We also recognize the limitations of our assessment concerning the evaluation of real exposure, and this could lead to an over- or underestimation of our findings. Our analysis was performed using proximity to major roads or highways as a surrogate of exposure to traffic-related pollutants, since measures of air pollutants were not available for the present study. Proximity to roads has commonly been used as a proxy for exposure to road traffic emissions in epidemiological studies as traffic pollutant levels (e.g., NOx) are higher than background levels in proximity to roadways (i.e., at a distance less than 200 m). (33) It should be stressed, however, that compared to Land Use Regression models that estimate exposure to pollutant levels, proximity to roadways is particularly prone to errors in estimating exposure and to classification bias. Nonetheless, proximity to roadways is a way to assess the risks associated with a mixture of air pollutants. In our assessment, we chose to be conservative and only report on risks attributable to distances of less than 75 m from roadways. (3)

Other uncertainties also arise from the estimation of attributable risk. In another study, attributable cases of asthma due to traffic-related pollutants and proximity to roadways were calculated using the standard equation (AF=[p.sub.p](RR-1)/([p.sub.p](RR-1)+1), where [p.sub.p]=proportion of exposed children). (32) However, the application of this equation requires the assumption that exposure to other factors not considered does not increase the disease risk, which we cannot ascertain in this study. Thus, we chose to apply an alternative formulation (34) (Equation 1 in this study, which is a function of the prevalence of exposure in diseased individuals). To adequately calculate this value, it is essential to have the prevalence of asthma in the exposed population. We did not have this information and we assumed this value to be equal to the prevalence of asthma in the whole population. This is a conservative approach since the prevalence of disease in the exposed population should be higher than in the whole population if exposure is associated with asthma prevalence.

Uncertainties are also associated with estimations of the number of exposed children. First, the population residing near major roads or highways was based on the location of postal codes of 1996, since more recent population numbers for postal codes were not available. This is an underestimation since population numbers between 1996 and 2006 increased by approximately 4%. Using the centroids of postal codes on a given street block is only an approximation of the location of the residences. Furthermore, the proportion of children of different age groups residing near these major roads or highways was deemed to be the same as that found across the Island of Montreal; this assumption requires validation. Better estimations would be provided with real population numbers.

Finally, we should recognize that retaining studies for which an association with the exposure to proximity to major roads was significant is an approach that would tend to lead to systematic bias. Meta-analysis including further studies and assessing publication bias are needed to provide more accurate results to estimate attributable risks.

From our assessment in Montreal, we identified a number of scientific gaps, including the development of risk functions specific to Montreal and possibly with estimates of exposure to air pollutants alongside major roadways. This input would be useful for either calculating the attributable fraction of prevalent asthma for the specific population, or to be included in a meta-analysis of relevant studies.

Even though uncertainties are present as cited in the preceding paragraphs, a range of childhood asthma cases that can be linked to residing near major roadways, as suggested by epidemiological risk functions, can be estimated, until more substantial information is made available. Our study is thus a first attempt to calculate crude risks of asthma prevalence attributable to proximity of residence to major roads or highways on the Island of Montreal. It points out the main limitations that we encountered and the directions that future studies must take to properly address the issue.

Acknowledgements: Elena Boldo was supported by a grant from the Ministry of Science and Innovation and the Carlos III Institute of Health, Spain (Bolsa de Ampliacion de Estudios de la Accion Estrategica de Salud. Plan Nacional I+D+i 20082011. BOE 58, 08-03-2010).

Conflict of Interest: None to declare.

Received: May 3, 2011

Accepted: November 5, 2011

REFERENCES

(1.) Environment Canada. National Pollutant Release Inventory Online Data, 2008. Available at: http://www.ec.gc.ca/pdb/websol/emissions/ap/ap_query_f.cfm (Accessed December 2, 2010).

(2.) World Health Organization. Health effects of transport-related air pollution. Krzyzanowski M, Kuna-Dibbert B, Schneider J (Eds.). Copenhagen, Denmark: WHO, 2005;190.

(3.) Health Effects Institute. Traffic-Related Air Pollution: A Critical Review of the Literature on Emissions, Exposure, and Health Effects. HEI Panel on the Health Effects of Traffic-Related Air Pollution. HEI Special Report 17. Boston, MA: HEI, 2010.

(4.) Patel MM, Miller RL. Air pollution and childhood asthma: Recent advances and future directions. Curr Opin Pediatr 2009;21(2):235-42.

(5.) Braback L, Forsberg B. Does traffic exhaust contribute to the development of asthma and allergic sensitization in children: Findings from recent cohort studies. Environ Health 2009;8:17.

(6.) Delfino RJ. Epidemiologic evidence for asthma and exposure to air toxics: Linkages between occupational, indoor, and community air pollution research. Environ Health Perspect 2002;110(Supp. 4):573-89.

(7.) Beelen R, Hoek G, van den Brandt PA, Goldbohm RA, Fischer P, Schouten LJ, et al. Long-term effects of traffic-related air pollution on mortality in a Dutch cohort (NLCS-AIR study). Environ Health Perspect 2008;116:196-202.

(8.) Finkelstein MM, Jerrett M, Sears MR. Environmental inequality and circulatory disease mortality gradients. J Epidemiol Community Health 2005;59:481-87.

(9.) Gehring U, Heinrich J, Kramer U, Grote V, Hochadel M, Sugiri D, et al. Long-term exposure to ambient air pollution and cardiopulmonary mortality in women. Epidemiol 2006;17:545-51.

(10.) Jerrett M, Finkelstein MM, Brook JR, Arain MA, Kanaroglou P, Stieb DM, et al. A cohort study of traffic-related air pollution and mortality in Toronto, Ontario, Canada. Environ Health Perspect 2009;117:772-77.

(11.) Hoffman B. Residential exposure to traffic is associated with coronary atherosclerosis. Circulation 2007;116:489-96.

(12.) Rosenlund M, Berglind N, Pershagen G, Hallqvist J, Jonson T, Bellander T. Long-term exposure to urban air pollution and myocardial infarction. Epidemiol 2006;17:383-90.

(13.) Tonne C, Melly S, Mittleman M, Coull B, Goldberg R, Schwartz J. A case-control analysis of exposure to traffic and acute myocardial infarction. Environ Health Perspect 2007;115:53-57.

(14.) Brauer M, Hoek G, Van Vliet P, Meliefste K, Fischer PH, Wijga A, et al. Air pollution from traffic and the development of respiratory infections and asthmatic and allergic symptoms in children. Am J Respir Crit Care Med 2002;166:1092-98.

(15.) Morgenstern V. Respiratory health and individual estimated exposure to traffic-related air pollutants in a cohort of young children. Occup Environ Med 2007;64:8-16.

(16.) Oftedal B, Nystad W, Brunekreef B, Nafstad P. Long-term traffic-related exposures and asthma onset in schoolchildren in Oslo, Norway. Environ Health Perspect 2009;117:839-44.

(17.) Garner R, Kohen D. Changes in the prevalence of asthma among Canadian children. Health Rep 2008;19:45-50.

(18.) Hill AB. The environment and disease: Association or causation? Proc R Soc Med 1965;58:295-300.

(19.) Bates DV. Health indices of the adverse effects of air pollution: The question of coherence. Environ Res 1992;59(2):336-49.

(20.) Morgenstern V, Zutavern A, Cyrys J, Brockow I, Koletzko S, Kramer U, et al. Atopic diseases, allergic sensitization, and exposure to traffic-related air pollution in children. Am J Respir Crit Care Med 2008;177:1331-37.

(21.) McConnell R, Berhane K, Yao L, Jerrett M, Lurmann F, Gilliland F, et al. Traffic, susceptibility, and childhood asthma. Environ Health Perspect 2006;114:766-72.

(22.) Kim JJ, Huen K, Adams S, Smorodinsky S, Hoats A, Malig B, et al. Residential traffic and children's respiratory health. Environ Health Perspect 2008;116:1274-79.

(23.) 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.

(24.) Shima M, Nitta Y, Adachi M. Traffic-related air pollution and respiratory symptoms in children living along trunk roads in Chiba prefecture, Japan. J Epidemiol 2003;13:108-19.

(25.) Statistique Canada. Montreal, Quebec (Code2466023) (tableau). Profils des communautes de 2006, Recensement de 2006, produit no 92-591-XWF au catalogue de Statistique Canada. Ottawa, 2007. Available at: http://www12.statcan.ca/censusrecensement/2006/dp-pd/prof/92-591/ index.cfm?Lang=F (Accessed January 5, 2011).

(26.) Deger L, Plante C, Goudreau S, Smargiassi A, Perron S, Thivierge RL, et al. Home environmental factors associated with poor asthma control in Montreal children: A population-based study. J Asthma 2010;47:513-20.

(27.) Gershon AS, Guan J, Wang C, To T. Trends in asthma prevalence and incidence in Ontario, Canada, 1996-2005: A population study. Am J Epidemiol 2010;172:728-36.

(28.) Rockhill B, Newman B, Weinberg C. Use and misuse of population attributable fractions. Am J Public Health 1998;88(1):15-19.

(29.) World Health Organization. Comparative Quantification of Health Risks. Global and Regional Burden of Disease Attributable to Selected Major Risk Factors. Volume 2. Ezzati M, Lopez AD, Rodgers A, Murray CJL (Eds.). Geneva, Switzerland: WHO, 2004.

(30.) Zhang J, Yu KF. A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998;280:1690-91.

(31.) Perez L, Kunzli N, Avol E, Hricko AM, Lurmann F, Nicholas E, et al. Global goods movement and the local burden of childhood asthma in Southern California. Am J Public Health 2009;99:S622-S628.

(32.) Kunzli N, Perez L, Lurmann F, Hricko A, Penfold B, McConnell R. An attributable risk model for exposures assumed to cause both chronic disease and its exacerbations. Epidemiol 2008;19(2):179-85.

(33.) Gilbert NL, Woodhouse S, Stieb DM, Brook JR. Ambient nitrogen dioxide and distance from a major highway. Sci Total Environ 2003;312:43-46.

(34.) Benichou J. Biostatistics and epidemiology: Measuring the risk attributable to an environmental or genetic factor. CR Biologies 2007;330:281-98.

Karine Price, MSc, [1] Celine Plante, MSc, [1] Sophie Goudreau, [1] Elena Isabel Pascua Boldo, MSc, [2] Stephane Perron, MD, MSc, FRCPC, [1,3] Audrey Smargiassi, PhD [4]

Author Affiliations

[1.] Direction de sante publique de l'Agence de la sante et des services sociaux de Montreal, Montreal, QC

[2.] Cancer and Environmental Epidemiology Unit, National Center for Epidemiology, Carlos III Institute of Health, Madrid, Spain; Consortium for Biomedical Research in Epidemiology and Public Health (CIBER en Epidemiologia y Salud Publica CIBERESP), Spain

[3.] Departement de medecine sociale et preventive, Universite de Montreal, Montreal, QC

[4.] Chaire sur la pollution de l'air, les changements climatiques et la sante, Universite de Montreal; Institut national de sante publique du Quebec; Departement de sante environnementale et sante au travail, Universite de Montreal, Montreal, QC

Correspondence: Karine Price, Direction de sante publique de l'Agence de la sante et des services sociaux de Montreal, 1301, rue Sherbrooke Est, Montreal (Quebec) H2L 1M3, Tel: 514-528-2400, Fax: 514-528-2459, E-mail: kprice@santepub-mtl.qc.ca

Table 1. Description of Information Concerning the Retained Studies
for Application to the Montreal Context

Study    Health Outcome       Type

[15]     Asthma prevalence    * Prospective birth
         (cross-sectional       cohort (Munich,
         analysis from          Germany)
         birth cohort)
                              * logistic
                                regressions
                                adjusted for sex,
                                parental atopy,
                                maternal education,
                                siblings,
                                environmental
                                tobacco smoke (ETS),
                                gas cooking, moulds,
                                dampness, pets

[20]     Asthma prevalence    * Prospective birth
         (cross-sectional       cohort (continuing
         analysis from          results from [15])
         birth cohort)
                              * logistic regressions
                                adjusted for sex,
                                age, parental atopy,
                                maternal education,
                                siblings, ETS, gas
                                cooking, dampness,
                                moulds, pets

[21]     Asthma prevalence    * Cohort study
         (cross-sectional       (Southern California)
         analysis from
         cohort)              * logistic regressions
                                adjusted for age, sex,
                                race, community,
                                language, housing
                                characteristics

[22]     Current asthma       * Cross-sectional study
                                (San Francisco Bay)

                              * logistic regressions
                                adjusted for crowding,
                                pests, moulds, chest
                                illness before age of
                                2, maternal asthma,
                                demographic variables,
                                housing characteristics

Study       Surrogate for Exposure: Proximity         Age (yrs)

                  Measure             OR (95% CI)

[15]     <50 m main road             1.23             2
         (motorway, federal or       (1.00, 1.51)
         state roads)

[20]     <50 m main road             1.66             4 and 6
         (motorway, federal or       (1.01, 2.59)
         state roads)

[21]     <75 m major road            1.50             5 to 7
         (freeway, highway,          (1.16, 1.95)
         arterial roads;
         approximately
         50,000-270,000
         v/day (24))

[22]     <75 m from freeway/         3.80 *           8 to 10
         highway; approximately      (1.20, 11.71)    (3rd to
         92,070-184,521 v/day)                        5th grade)

* The OR for this study is between 2 and 3 times higher
than for the other studies. v = vehicles.

Table 2. Number of Cases of Asthma in Children Attributable to
Proximity to Major Roads Corresponding to Different Exposure Scenarios

Ref.    Pop.       Pop. Child     Pop. of    Proportion     Prevalence
        at Risk    Residents      Children   of Exposed     of Asthma
        (yrs)      in Proximity   ([Pop.     Cases          in Montreal
                   to Major       sub.I])    ([P.sub.ei])   Children
                   Roads or                                 ([P.sub.o])
                   Highways in
                   Montreal

[15]      2          2973           18,850     0.16           0.10
[20]      4          2830           17,945     0.16           0.13
[20]      6          2887           18,310     0.16           0.19
[21]    5 to 7     11,809           54,450     0.22           0.18
[22]    8 to 10     650 *           57,470     0.01           0.19

Ref.         Number of Asthma Cases
           on the Island of Montreal
              at Specified Age

          Based on        Based on
         Prevalence      Prevalence
            From            From
        Ontario (17)    Montreal (26)

[15]         2903            1855
[20]         2764            2388
[20]         4120            3506
[21]       12,251          10,013
[22]       14,655          10,735

Ref.      Prevalent Cases Attributable to Proximity

          Based on           Based on Prevalence
         Prevalence       From Montreal (26) (95% CI)
            From
        Ontario (17)
          (95% CI)

                         Attributable    Attributable
                           Number of       Fraction
                          Cases (AN)        (AF,%)

[15]     69 (0-125)       44 (0-80)      2.4 (0-4.3)
[20]    154 (4-238)      133 (3-206)     5.6 (0.1-8.6)
[20]    241 (6-372)      205 (5-317)     5.9 (0.1-9.0)
[21]    778 (322-1138)   636 (263-930)   6.4 (2.6-9.3)
[22]    108 (24-134)      79 (18-98)     0.7 (0.2-0.9)

* In proximity to highways only.

Note: Values of P were obtained as follows:

[P.sub.ei] = Pop.child residents in proximity to major roads or
highways in Montreal/[Pop.sub.i]