BackgroundThere is increasing evidence that residential proximity to roadways is associated with an elevated risk of asthma exacerbation. However, there is no consensus on the distance at which these health effects diminishes to background levels. Therefore the optimal, clinically relevant measure of exposure remains uncertain. Using four spatially defined exposure metrics, we evaluated the association between residential proximity to roadways and emergency department (ED) presentation for asthma in Perth, Western Australia.MethodThe study population consisted of 1809 children aged between 0 and 19 years who had presented at an ED between 2002 and 2006 and were resident in a south-west metropolitan area of Perth traversed by major motorways. We used a 1:2 matched case-control study with gastroenteritis and upper limb injury as the control conditions. To estimate exposure to traffic emissions, we used 4 contrasting methods and 2 independently derived sources of traffic data (video-monitored traffic counts and those obtained from the state government road authority). The following estimates of traffic exposure were compared: (1) a point pattern method, (2) a distance-weighted traffic exposure method, (3) a simple distance method and (4) a road length method.ResultsRisk estimates were sensitive to socio-economic gradients and the type of exposure method that was applied. Unexpectedly, a range of apparent protective effects were observed for some exposure metrics. The kernel density measure demonstrated more than a 2-fold (OR 2.51, 95% CI 2.00 - 3.15) increased risk of asthma ED presentation for the high exposure group compared to the low exposure group.ConclusionWe assessed exposure using traffic data from 2 independent sources and compared the results of 4 different exposure metric types. The results indicate that traffic congestion may be one of the most important aspects of traffic-related exposures, despite being overlooked in many studies on the exacerbation of asthma.