Detecting thresholds in the health impact of the urban built environment

Abstract

The urban built environment impacts human health through complex and nonlinear pathways. However, thresholds of urban built environment attributes associated with respiratory health are still relatively undetermined. Moreover, most existing studies assume the built environment to be static and only incorporate a cross-sectional approach to measuring it. In this case study examining the impact of the urban built environment on lung cancer risk in Shanghai, China, we perform a longitudinal analysis incorporating a range of environmental attributes and data spanning multiple years to tackle the synergistic effect of the urban built environment over an extended time period. We propose a novel approach combining nonlinear regressions and Geo-Detector to discern the general trend in the environment-disease association and pinpoint significant thresholds within this trend. Our findings showed that impervious land percentage, building coverage, green space coverage, and population density explained stratified heterogeneity of the respiratory health outcome by 8.7%, 21.2%, 8.1%, and 8.8%, respectively. Lung cancer incidence was significantly elevated in areas or during times with impervious surface percentage > 95%, building coverage > 25%, green space coverage < 15%, or population density > 10 thousand persons/km². These findings offer actionable insights for urban development regulation and policymaking.