Heavy-Traffic Roads, Benzene Exposure, and Childhood Leukemia
Heavy-Traffic Roads, Benzene Exposure, and Childhood Leukemia
Childhood leukemia may be associated with traffic-related environmental exposure to benzene, and additional data are needed. The Géolocalisation des Cancers Pédiatriques (GEOCAP) Study, a nationwide French case-control study, was designed to avoid selection bias due to differential participation and misclassification. The study compared the 2,760 childhood leukemia cases diagnosed in France between 2002 and 2007 (including 2,275 cases of acute lymphoblastic leukemia (ALL) and 418 cases of acute myeloblastic leukemia (AML)) with 30,000 contemporaneous child population controls. The residence addresses were precisely geocoded, and 3 indicators of residential proximity to traffic were considered. Estimates of benzene concentrations were also available for the Île-de-France region (including Paris). A 300-m increase in major road length within 150 m of the geocoded address was significantly associated with AML (odds ratio = 1.2, 95% confidence interval: 1.0, 1.4) but not with ALL (odds ratio = 1.0, 95% confidence interval: 0.9, 1.1), and the association was reinforced in the Île-de-France region when this indicator was combined with benzene estimates. These results, which were free from any participation bias and based on objectively determined indices of exposure, showed an increased incidence of AML associated with heavy-traffic road density near a child's home. The results support a role for traffic-related benzene exposure in the etiology of childhood AML.
A relationship between benzene exposure and leukemia has been demonstrated in adults highly exposed to benzene in the workplace. Children may be exposed to much lower concentrations of benzene, mainly due to anthropogenic sources, especially road traffic. Traffic-related air pollution is a complex mixture of many compounds. Only a few studies to date have specifically addressed the association between childhood leukemia (CL) and traffic-related benzene exposure using benzene concentration estimates obtained by means of dispersion models integrating parameters such as weather, wind speed, traffic density, distance to the nearest road, building heights, or street widths. Three of these studies showed a positive association with CL. Most of the studies were based on indicators of residential proximity to roads and traffic. In a recent meta-analysis based on 7 studies with diverse definitions of proximity and traffic density, Boothe et al. concluded that postnatal exposure to heavy-traffic roads might be related to CL. The information available did not allow meta-analysis by subtype of CL. A few studies used nitrogen dioxide concentration as a marker of traffic-related air pollution in the vicinity of roads, estimated most often by means of dispersion models, cokriging models, or land-use regression models. The results of those studies were not consistent.
We investigated the role of residential exposure to heavy-traffic roads, as a proxy for traffic-related pollution, in the occurrence of CL. All CL cases diagnosed in France from 2002 through 2007 were included, and controls were drawn from a population registry. In addition, the study focused on traffic-related benzene exposure in the subset of children living in the most urbanized region (Île-de-France), for which estimates of local benzene concentrations were available.
Abstract and Introduction
Abstract
Childhood leukemia may be associated with traffic-related environmental exposure to benzene, and additional data are needed. The Géolocalisation des Cancers Pédiatriques (GEOCAP) Study, a nationwide French case-control study, was designed to avoid selection bias due to differential participation and misclassification. The study compared the 2,760 childhood leukemia cases diagnosed in France between 2002 and 2007 (including 2,275 cases of acute lymphoblastic leukemia (ALL) and 418 cases of acute myeloblastic leukemia (AML)) with 30,000 contemporaneous child population controls. The residence addresses were precisely geocoded, and 3 indicators of residential proximity to traffic were considered. Estimates of benzene concentrations were also available for the Île-de-France region (including Paris). A 300-m increase in major road length within 150 m of the geocoded address was significantly associated with AML (odds ratio = 1.2, 95% confidence interval: 1.0, 1.4) but not with ALL (odds ratio = 1.0, 95% confidence interval: 0.9, 1.1), and the association was reinforced in the Île-de-France region when this indicator was combined with benzene estimates. These results, which were free from any participation bias and based on objectively determined indices of exposure, showed an increased incidence of AML associated with heavy-traffic road density near a child's home. The results support a role for traffic-related benzene exposure in the etiology of childhood AML.
Introduction
A relationship between benzene exposure and leukemia has been demonstrated in adults highly exposed to benzene in the workplace. Children may be exposed to much lower concentrations of benzene, mainly due to anthropogenic sources, especially road traffic. Traffic-related air pollution is a complex mixture of many compounds. Only a few studies to date have specifically addressed the association between childhood leukemia (CL) and traffic-related benzene exposure using benzene concentration estimates obtained by means of dispersion models integrating parameters such as weather, wind speed, traffic density, distance to the nearest road, building heights, or street widths. Three of these studies showed a positive association with CL. Most of the studies were based on indicators of residential proximity to roads and traffic. In a recent meta-analysis based on 7 studies with diverse definitions of proximity and traffic density, Boothe et al. concluded that postnatal exposure to heavy-traffic roads might be related to CL. The information available did not allow meta-analysis by subtype of CL. A few studies used nitrogen dioxide concentration as a marker of traffic-related air pollution in the vicinity of roads, estimated most often by means of dispersion models, cokriging models, or land-use regression models. The results of those studies were not consistent.
We investigated the role of residential exposure to heavy-traffic roads, as a proxy for traffic-related pollution, in the occurrence of CL. All CL cases diagnosed in France from 2002 through 2007 were included, and controls were drawn from a population registry. In addition, the study focused on traffic-related benzene exposure in the subset of children living in the most urbanized region (Île-de-France), for which estimates of local benzene concentrations were available.
Source...