Social Capital and Self-rated Health: testing association with longitudinal and multilevel methodologies

Abstract: Since Durkheim’s seminal work over a century ago, research has repeatedly shown that individuals with higher levels of social integration, social networks and social support have better health status. However, the recent introduction of a contextual phenomenon known as social capital to the field of public health has sparked lively debate as to how it may also influence the health of individuals, if at all. Though critics have raised several points of contention regarding reported association between social capital and health over recent years, one outstanding issue remains: the lack of empirical research focusing on causal relationships, due to paucity of adequate longitudinal social capital data. The overall aim of this thesis is to test association between different social capital proxies and self-rated health (SRH), alongside other well-known health determinants, using multilevel and longitudinal data, whilst employing a variety of study designs and methods. All data used in this thesis come from the United Kingdom’s British Household Panel Survey (BHPS) from years 2000, 03, 05, 07 and 08. The underlying premise of this body of work is to investigate temporal (causal) relationships between social capital and health. All four papers of this thesis demonstrate that generalised trust is the most robust of all social capital proxies tested, it maintaining a positive association with SRH over time. Furthermore, results from paper III imply that prior trust levels can predict future SRH, lending weight to the hypothesis that trust is an independent determinant of health. However, debate remains as to whether generalised trust solely captures social capital or other, more tangible aspects of social cohesion.