In this study, we used EHR data to assess the association between residential distance to roadways on asthma exacerbation in children. There are a number of studies suggesting that proximity to high traffic areas, including roadways, have adverse effects on child respiratory health,40,41 hospital admissions for asthma,42 pulmonary function,43,44 and atopy,45 and there is some evidence that early life exposure to traffic-related pollution affects later asthma outcomes.46 In our assessment of exacerbation rates in children with diagnosed asthma, we found a negligible relationship between distance from 55 + MPH roadway and any asthma exacerbation. Distance from 35 + MPH roads did not affect exacerbations at all.
Though our results were negative, our study has a number of strengths that differentiates it from prior work. For one, our cohort was relatively large compared with previous studies, ranging from 2 times to 15 times larger.16,19,25,27,30 Moreover, the majority of previous studies used questionnaires to collect information.19,25,30 Such questionnaires are frequently complicated and time-consuming, and, because questionnaires rely on self-reports, they can cause some potential misclassification due to recall bias for outcomes and/or exposures.27 Instead of relying on self-report, we directly extracted patient information from the EHR system, which efficiently provides data on a large number of individuals and avoids biases found with self-reported data.
The lack of significant association between residential distance to roadway and asthma exacerbation rates could potentially be explained by factors related to air quality and the built environment of Durham. First, the air quality in Durham is generally good (Fig. 4). Other studies that reported a weak or no association between asthma and traffic-related air pollution evaluated populations living in locations with good air quality, such as Sweden, Norway, and Germany.27–30 Based on the worldwide air quality rankings for mid-2019, the air quality index for these countries were 17.45, 20.29 and 28.42 respectively.47 In contrast, studies that reported strong associations were conducted in places such as the United Kingdom, Poland, and South Korea, whose air pollution indexes were 40.63, 52.88, and 54.80, respectively.23–25, 48 Similar trends have been reported within the United States. A study conducted in Los Angeles reported that children and adults who suffer from asthma and live close to traffic are nearly three times more likely to visit the emergency department or be hospitalized than those who live near low traffic density.18 In fact, cities in California, including Los Angeles, have ranked to have the worst particle pollution in the US.49
A second reason our results may differ is that compared to other studies our cohort lived further from roadways. Multiple studies reported a threshold of 150 meters (~ 0.1 mile) for finding the highest concentrations of traffic-related pollutions.17,25 However, in our cohort, only 283 children live within this proximity to roadways. Two previous studies in California and South Korea utilizing radial density measures similar to ours found effects only at much closer distances of 500 feet and 200 meters respectively, than the 1-mile buffer we used.23,25
Only one other study, conducted in Minnesota, has utilized EHR data to assess the relationship between asthma exacerbation and traffic-related air pollution.50 The authors used vehicle kilometers traveled within 250 meters and 500 meters of each individual’s residence, as well as the traffic density as their exposures. Traffic exposures is similar to the radial density measures in our study. Assessing both pediatric and adult patients, the authors reported that traffic exposure at the residence increases the risk of asthma exacerbations. The authors reported lower odds ratio for vehicle kilometers traveled within 500 meters than the vehicle kilometers traveled within 250 meters and affirmed that 500 meters buffer distance does not effectively capture the traffic effect. Moreover, while air quality across Minnesota is generally within the healthy range, there are several days each year where fine particles levels can exceed to the unhealthy levels.51 Overall this paper helps to contextualize some of the present findings highlighting that it is both close exposure to roadways that matters and that there may be air quality threshold that an environment needs to drop below.
There are a few limitations in our study. While our results suggest no association between residential distance to roadway and asthma exacerbation rates, a null result is hard to prove, as statistical tests are not designed to detect them. Moreover, our results are limited to one geographic area. As we noted above, Durham County has several unique characteristics, including relatively good air quality, which may cause effect heterogeneity between studies. Finally, while we believe that the use of EHR data produces more objective data measurements, EHR data are prone to their own biases.52 In particular, type of health service utilization is often a reflection of both severity of disease as well as other social factors such as SES and health seeking behavior. Of note, though, our analyses did not indicate confounding due to nSES.