The associations between exposures to traffic and mortality risks in Hong Kong

Abstract

Background: Traffic is a main source of air pollution in Hong Kong. The associations between traffic related health effects have been 13 reported mostly for European and North America, but very few for densely populated urban environments in Asia. Objective: To assess traffic related health effects in the population of Hong Kong. Methods: Residential locations of persons who died during 2010 were documented in 209 Tertiary Planning Units (TPU) of Hong Kong. The 2010 Annual Average Daily Traffic (AADT) data were retrieved from traffic census reports. Three indicators of trafficrelated exposure by TPU were computed using ArcGIS (version 9.2): (i) Road Density (RD) = total length of roads divided by area; (ii) Traffic Density (TD) = RD times AADT; and (iii) Vehicle Density (VD) = AADT divided by area. They were further categorizing into three levels by tertile: low, middle and high. Poisson regression models of standardized mortality ratios were fitted with adjustment for ethnicity, marital status, education level, housing type and income. Results: The ranges of risk ratios (95% CI) of mortality for high or middle relative to low exposure level were: for all-nature causes 1.17 (1.13-1.21)-1.32(1.27-1.36), cardiovascular 1.16(1.09-1.23)- 1.27(1.19-1.37), respiratory 1.30(1.20-1.42)-1.44(1.33-1.56), and cancer 1.08(1.02-1.14)-1.28(1.20-1.36). The risk ratios were the highest for the middle level of TD or VD, but increased from middle to high level of RD. Conclusions: Traffic related exposures measured at the neighborhood levels are associated with mortality. Investigations at individual levels are necessary to account for confounding.