Few studies have investigated the spatial variation in road traffic indicators associated with lung cancer risk. The purpose of this study was to explore the relationship between road traffic-related indicators and lung cancer risk and to estimate its spatial variability. The population-based case control study was conducted, including all the newly diagnosed lung cancer patients (cases) and colorectal cancer patients (controls) in Jiading District, Shanghai from 2014 to 2016. Traffic intensity variable (traffic intensity in a 500 m buffer), residential distance to major road or highway, and greenness exposure at the residence were estimated for each individual. We conducted unconditional logistic regression with adjustment for age, sex, smoking status and NDVI values and geographically weighted logistic regression (GWLR). The clustering of lung cancer risk was analyzed by Bernoulli model of the SaTScan software. This study included 1461 lung cancer patients and 954 colorectal cancer patients. In multivariate logistic regression, smoking [OR 1.25 95% CI (1.15-1.35)], living <50 m from the major road [OR 1.43 95% CI (1.02-2.03)] were significantly associated with lung cancer risk. Residential Proximity to highway, residential greenness, and traffic intensity were not significantly associated with lung cancer risk. The GWLR model showed that the degree of correlation between residential proximity to major road and lung cancer risk varied geographically. The SaTScan results showed a lung cancer cluster in the southwest of Jiading District, Shanghai. Our study suggested that the distance from residence to the main road was significantly associated with lung cancer risk, which varied geographically. It is helpful to further study the traffic factors' spatial variation related to lung cancer risk and carry out reasonable regional planning.
Keywords: Air pollution; Geographically weighted logistic regression (GWLR); Lung cancer; Traffic.
Copyright © 2018 Elsevier B.V. All rights reserved.