Exploring neighbourhood effects, genetic characteristics and individual behaviours' influence on health disparities within a geographical context...

Lead Research Organisation: University of Bristol
Department Name: Geographical Sciences


... based on social media data mining and individual level survey data.

Rapidly increasing prevalence of chronic diseases, such as obesity and diabetes has become a major worldwide challenge and threat requiring more attention and potential interventions. The health status of humans is inevitably affected by the living environment, including the built, social, economic and political environments, their individual characteristics and behaviours, and genetic variations. This research mainly focuses on geographical variations of health disparities and how geographical context plays the significant role in interactions between environments, genomic and individual behaviours, comprehensively understanding the underlying mechanisms of health inequality and investigating risk factors of chronic diseases. Individual level survey data will be utilized to explore, identify and classify the neighbourhood effect and genetic influence on health status based on different exploratory statistical and regression models. Additionally, widespread usage of the Internet provides an opportunity to share daily life, as well as an opportunity to observe health-related behaviours, such as healthy food consumption, physical activity, smoking, alcohol consumption, and poor sleep patterns (Pachucki, et al., 2011; Keating, et al., 2011; Rosenquist, et al., 2010; Mednick, et al., 2010). More health-behaviour patterns can be observed through social media data mining among the wider population, providing an opportunity to investigate the association between socio-cultural factors and health outcomes; also to identity the neighbourhood effect and risk factors of chronic diseases across space based on the combination of survey data and health-related behaviours, such as diet habits and physical activities derived from social media platforms. The analytical results will provide insight on how neighbourhood effect, genetic characteristics and individual behaviour have effects on health status, encouraging governments to raise the awareness of different types of risk factors and improve the living environment aiming at addressing high risks of chronic diseases.


10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
ES/P000630/1 01/10/2017 30/09/2027
2094832 Studentship ES/P000630/1 01/10/2018 01/10/2022 Yunqi Zhou
Description The causes of chronic diseases such as diabetes and obesity have been widely studied and related to various factors including individual behaviours and genetics, as well as to the neighbourhood environments in which people live. Less studied is the possibility that these causes and their effects are themselves modified according to a wider geographical context, creating geographical variations in health outcomes and their causes. This study uses UK biobank data with the recruitment of around 500,000 volunteer participators aged 40-69 between 2006 and 2010, throughout the UK to investigate how different types of risk factors contribute to the geographical variations of the prevalence of chronic diseases. This study has adopted methods of spatial analysis and geographical modelling approaches, such as multi-level models to explore geographical variations in the prevalence of diabetes and obesity, and whether their causes vary from places to places. The current work focuses on participators who attend assessment centres in London and aims to investigate how different types of risk factors influence individual BMI values. Firstly, this study examines the spatial patterns of health disparities and inequalities at the MSOA level through exploratory spatial analysis. There are apparent and noticeable spatial clusters of cold posts (low-low) for average BMI values in Southwestern London. Subsequently, multi-level regression models with consideration of hierarchical structure within data have been constructed to investigate how different types of risk factors have an influence on health outcomes. It is discovered that the weighted Multi-level model is feasible to better capture spatial clustering within individual BMI values without underestimations of interpretability of predictors on the between coordinate variances. Geographical context plays an important role in generating health disparities according to the null weighted two-level model and spatial clustering patterns in average BMI values at MSOA level. Added predictors are possible to interpret the majority between coordinate variances based on the dramatic reduction of VPC values. However, there is still apparent within coordinate variance which requires other datasets, such as genetic variances to interpret remaining individual variances of BMI values. According to evidence and finding derived from the finalized weighted two-level random intercept model, personal characteristics and lifestyle habits have more influence on individual BMI compared with neighbourhood effect. Usual walking pace has the most substantial influence on the risks of obesity, suggesting improvements in the neighbourhood environment and facilities for exercise. It is discovered besides physical activity and diet habits which have been discussed previously to have a substantial effect on being obese, sleep patterns have a considerable impact on individual BMI values, thereby requiring further analysis. Diet habits, including dietary pattern and diet composition, both play a significant role in obesity. However, in this study, household income has limited influence compared with lifestyle habits. Compared with white British participators, Asians have relative lower BMI values while Africans have higher BMI values.
Exploitation Route This study will adopt methods of spatial analysis and of geographical modelling to explore geographical variations in the prevalence of diabetes and obesity, and whether their causes vary from places to places. It aims to understand if and how geographical context plays an important role in generating health disparities and to examine the potentially geographically varying relationships between various types of risk factors (including the neighbourhood environment, personal factors and genetic variations). This study helps to understand the causes of geographical variation in the prevalence of obesity and of diabetes looking at various types of gene-environment interactions to construct improved disease prediction models and to better understand the causes of theses chronic diseases and if those causes appear to vary geographically. This study helps to investigate how the modelled relationships may themselves vary geographically. For example, does an unhealthy lifestyle matter more in one city compared with another?
Sectors Healthcare

Description In consequence, evidence discovered in this study suggests the consequences of health disparity and inequality is attributable to multifactorial influence and mainly influenced by individual behaviours, lifestyle habits and demographical characteristics compared with neighbourhood effect and socio-economic status. Governments are encouraged to increase public facilities in neighbours to encourage the higher frequency of physical activities and address the high risks of obesity.If there is geographical variation in what predicts the prevalence of these diseases, then knowledge of it is beneficial to avoid "one-size-fits-all" prevention and intervention strategies, instead of tailoring them to the individual and their context.
First Year Of Impact 2021
Sector Healthcare
Impact Types Policy & public services