Examining the nonlinear relationship between neighborhood environment and residents' health

Abstract

Although numerous studies have revealed linear associations between the urban environment and health, the nonlinear effects of subjective and objective neighborhood environment on health have been under-researched. Using Beijing as a case study, this study employs the random forest method to investigate the nonlinear relationship between subjective/objective neighborhood environment and physical/mental health, respectively. The findings indicate that the subjective environment is the most important for both physical and mental health, and subjective air and noise pollution are the key factors. Neighborhood environmental variables have nonlinear correlations with health. Improved perception of facility accessibility is positively associated with physical health, whereas extremely high levels of accessibility may inversely decrease the probability of reporting good health. Road connectivity and bus stop density exhibit inverted U-shaped patterns with physical health and mental health, separately. Furthermore, the study highlights different dimensions of green space affect residents' health in diverse ways.