Area Social Deprivation and Public Health: Analyzing the Spatial Non-stationary Associations Using Geographically Weighed Regression

The prevalence of non-communicable chronic diseases (NCDs) worldwide poses an alarming threat to public health. Recent literature has embraced the opinion that incorporating the social factors should advance the understanding of NCDs prevalence. In this context, examining the NCDs prevalence in association with area social deprivation should provide critical implications for coping with public health risks. However, few empirical studies have examined this specific issue, especially in the developing countries. Using the principal component analysis, an area social deprivation index (ASDI) is established for the Shenzhen city (China) by integrating ten indicators from four dimensions: education, housing, socially disadvantaged population, and economically disadvantaged population. The geographically weighted regression (GWR) is employed to analyze the associations between ASDI and the incidence rate of three prevalent NCDs at district scale. Spatial non-stationary relationships are identified for the three diseases. More specifically, prevalence of the three diseases is all positively correlated with the ASDI. Strength of the associations presents the geography that it generally decreases from the central city to the suburb. These findings suggest that greater possibility of NCDs prevalence would be expected in districts with higher social deprivation. Besides, the impact of social deprivation on NCDs prevalence is much stronger in the central city. The spatial stationarity can facilitate the formulation of location-specific preventive measurements. This paper is believed to provide an innovative insight for social indicators research.