Socio-spatial inequalities in tobacco retail availability and smoking in megacities of China : a case study of Shanghai

Background: China is the world's biggest manufacturer and consumer of tobacco products. In 2019, the smoking prevalence was 26.6%, with a large gender disparity (50.5% of men and 2.1% of women). Over 1 million deaths and an estimate of USD 45.28 billion in health expenditure could be attributed to tobacco smoking (2015). Aiming to reduce the national-level smoking prevalence by 20% along with a reduction in health inequalities, Healthy China 2030 was proposed in 2016 as a national strategy, which makes the tobacco epidemic a public health priority in China. Increasing international evidence suggests tobacco retail availability is associated with smoking, with more tobacco retailers clustered within more vulnerable areas. Regulating tobacco retail availability provides policy opportunities for effective tobacco control. However, the evidence in this area is scarce in the Chinese context. Therefore, the aim of this thesis is to investigate the socio-spatial inequalities in tobacco retail availability and smoking behaviours in the context of Shanghai, the biggest megacity of China with the most developed economy and a well-established local tobacco retail market in a transition economy and rapid urbanisation. Methods: This thesis comprises of four related empirical studies using multiple datasets. Using web-based Points of Interest data (POIs) regarding tobacco retailers (2019) and the neighbourhood deprivation index constructed by the 2010 Chinese census, population-weighted kernel density estimation, the Kruskal-Wallis test and effect size estimation were utilised for Study 1 to examine the socio-spatial disparities of tobacco retail availability and its association with neighbourhood deprivation. Study 2 used registered public school data (2019) and POIs regarding tobacco retailers, along with the neighbourhood deprivation index, to investigate whether and to what extent tobacco retailers cluster around public schools, as well as the compliance of tobacco retailers with local spatial restrictions through spatial analysis techniques (i.e. network analysis and multi-distance spatial cluster analysis). Linking tobacco retail availability index and neighbourhood deprivation index with the International Tobacco Control China Survey (2009-2015), Study 3 used 2-level logistic regression and generalised estimating equations to determine the cross-sectional and longitudinal relationship between tobacco retail availability, neighbourhood deprivation, and smoking behaviours. Study 4 simulated a total of 12 key-stakeholder-informed scenarios. It used spatial analysis techniques to evaluate the impacts of plausible spatial restrictions on tobacco retail availability reduction. Results: Study 1 suggests the overall tobacco retail availability in Shanghai was high (9.50 retailers per km2 per neighbourhood), with a significant non-linear association between tobacco retail availability and neighbourhood deprivation, which further varied across levels of urbanity (p < 0.001), as well as the widespread availability of convenience stores. Study 2 revealed a significant spatial clustering of tobacco retailers around public schools, with 95.8% of public schools being exposed to tobacco retail and 60.1% of tobacco retailers in the vicinity of a public school within an 800 m distance. Additionally, about 5.9% of public schools were exposed to non-compliant tobacco retailers. Study 3 suggests that tobacco retail availability and neighbourhood deprivation were each significantly associated with individual-level smoking, and their impacts differed in cross-sectional and longitudinal models. Furthermore, the association between tobacco retail availability and individual-level smoking varied by urbanity and deprivation level. Study 4 indicates all spatial restrictions have the potential to reduce tobacco retail availability substantially, but the most effective restriction may greatly increase social inequality in tobacco retail availability. In addition, the impacts of spatial restrictions on reducing tobacco retail availability and inequality differed by urbanity level. Conclusion: Findings revealed the existence of socio-spatial inequalities in tobacco retail availability and smoking in the context of Shanghai, which calls for the development of a comprehensive tobacco control strategy. This thesis provides useful insights into regulating tobacco retail availability, which remains an opportunity for effective tobacco control. In addition to reducing overall availability, regulating availability at the neighbourhood level and focusing on retailer type in the process of tobacco retail licensing could be effective for the completion of tobacco control strategies. Furthermore, stricter restrictions on tobacco retail should be considered around school areas. Moreover, while spatial restrictions could reduce tobacco retail availability substantially, the most effective restrictions may actually increase inequality in tobacco retail availability. It is critical for policymakers to enact policies targeting different social groups and groups in certain geographical areas.