Non-Linear Effects of the Built Environment and Social Environment on Bus Use among Older Adults in China: An Application of the XGBoost Model.

Lanjing Wang,Yi Zhang,Xumei Chen,Chaoyang Li,Chunli Zhao,Jiani Wu,Xiaofei Liu,Tao Wang

doi:10.3390/ijerph18189592

Abstract

Global aging has raised increasing concerns on the health and well-being of older adults. Public transport is a viable option to improve the mobility and quality of life among older adults. However, policies that promote the public transport use among older adults are rare. This study utilizes the eXtreme Gradient Boosting (XGBoost) decision tree to explore the non-linear associations of the built and social environment with bus use among older adults in China. The bus use of older adults was obtained from the Zhongshan Household Travel Survey (ZHTS) in 2012. Results show that non-linear relationships exist among all built environment and social environment characteristics. Within certain thresholds, the percentage of green space land use, land use mixture, bus-stop density, and dwelling unit density are positively related to bus use among older adults. Likewise, one social environment variable, the proportion of older adults in a neighborhood, is the key social environment variable. Furthermore, the dwelling unit density and proportion of older adults appear to have an inverse U-shaped relationship. Additionally, age, ownership of motorcycles, and distance from home to the nearest bus stop also show non-linearity. The findings presented in this paper facilitate effective planning interventions to promote bus use among older adults.

Highlights

Driven by the increase in life expectancy and the decline in fertility, the age structure of the world’s population continues to undergo sustained changes
This study introduces the XGBoost modeling approach, an improved version of gradient boosting decision tree (GBDT), to test the hypothesis and disentangle the complex non-linear relationships between five categories of independent variables and public transport use among older adults in China
This study introduced the XGBoost model to explore the non-linear associations of the built and social environment with bus use among older adults in Zhongshan, China

Summary

Introduction

Driven by the increase in life expectancy and the decline in fertility, the age structure of the world’s population continues to undergo sustained changes. The proportion and number of older adults in the total population are increasing dramatically [1]. One of the goals of the 2030 Agenda for Sustainable Development is to ensure healthy lives and promote well-being for all at all ages [3]. This pledge implies that all segments of society and all ages must be considered, with special attention to the most vulnerable groups, including older adults. In this context, the living arrangements for older adults are increasingly important to policymakers. Previous studies have highlighted that continuous mobility has a strong correlation with active aging and the health status, well-being, and quality of life

Methods

Results

Conclusion