Commentary: Traffic exposure and asthma: problems of interpretationBMJ 1996; 312 doi: http://dx.doi.org/10.1136/bmj.312.7032.676a (Published 16 March 1996) Cite this as: BMJ 1996;312:676
- David P Strachan, reader in epidemiologya
Neither air pollution nor asthma is a new problem, but in the past decade there has been increasing concern that the apparent increase in the prevalence of asthma may be due to increasing exposure to pollution, particularly from motor vehicles.1
Vehicles emit many pollutants, but in relation to asthma interest has focused on particulate matter and nitrogen oxides (mainly nitrogen monoxide). Nitrogen monoxide is rapidly oxidised to nitrogen dioxide, which in turn, and at a distance, catalyses the formation of ozone. Proximity to major roads or estimates of local traffic density have been used as indicators of exposure to vehicle exhaust fumes in several recent epidemiological studies, including that by Livingstone and colleagues.2 The relation of traffic exposure to pollutant exposure, however, is far from straightforward, and this complicates the interpretation of both positive3 4 and negative2 findings.
Although mean nitrogen oxide and particulate concentrations decrease with distance from the kerb,3 the decline beyond 20 m is small. No correlation between traffic density and ambient nitrogen dioxide concentrations was found in Munich,4 although carbon monoxide, benzene, and toluene (pollutants related to vehicles that are not implicated in asthma) were more concentrated in areas of higher traffic flow. Higher traffic density was inversely correlated with concentrations of ozone,4 which is formed at some distance from emission sources and scavenged in city centres by nitrogen monoxide from vehicles.
Hitherto, epidemiological studies have imputed traffic exposure from place of residence. Yet, for many people the environment close to home represents only a small part of their daily exposure to outdoor air. Furthermore, indoor sources (particularly cooking fuels and environmental tobacco smoke) are more important influences on personal exposure to both nitrogen oxides and airborne particulates. Only a small increase in personal exposure to these pollutants was discernible for subjects living close to major roads in Tokyo.5
When significant associations between recurrent wheezing and traffic exposure have been reported—for example, among Japanese women3 and German children4—the prevalence varies (in relative terms) by up to 50% across the range of traffic exposures. This is a modest difference but arguably too great to be explained by the subtle variations in measured exposures to pollutants related to local traffic density. This does not exclude the possibility that other vehicle related pollutants may have a hitherto unsuspected role in initiating or exacerbating asthma, but it argues for a cautious interpretation and careful consideration of possible confounding factors and reporting artefacts.
On the other hand, the failure of a statistically powerful study such as that by Livingstone and colleagues to show an association between local traffic density and prevalence of disease2 does not refute the possibility of a more general link between air pollution and asthma because place of residence is such a poor indicator of personal exposure to traffic related pollutants. Nevertheless, these findings do offer reassurance to city dwellers living close to busy roads that the location of their home does not place them or their children at substantially increased risk of asthma.