Social contexts and health in late life: Living arrangements, social networks, resources, and health

Most research on the association between social contexts and health inequality has focused on life events as major sources of stress and has viewed social connections as major moderating mechanisms (e.g., the "buffering") that reduce the impact of these unexpected events. The general underlying premise is that one additional close connection (or strong tie) affords one more access to resources, and thus, results in better health outcomes even after stressful and isolating life events. I challenge this assumption in two ways: first, that our understanding of source of stress is biased by only locating significance in major life events (as opposed to everyday life), and second, that the proportional assumption on additional close connections is problematic for our understanding of social connections and health outcomes.;For the first issue, I argue that one prominent source of everyday stress stems from the configurations (or composition) of members in social contexts. I draw on the stress process model, social exchange theory, and role strain theory to build a theory of the association between the configurations of members and their impact on health outcomes. Furthermore, I challenge the perspective that confines the role of social connections to mediating or moderating mechanisms and reconceptualize social connections as a micro-relational structure that exerts significant influence on older adults' health and well-being. For the second issue, which is related to our understanding of social roles and exchanges within contexts, I argue that an additional close connection may not only increase the size or the volume of interactions but at the same time increase the complexity of expectations and exchanges that may induce negative experiences in social interactions.;My analysis begins by identifying the types of member configurations in two domains of social context---households and social networks. For the configurations of household members, I obtained detailed information on persons living in an older adult's household and his or her relationships with those persons. For social networks, I developed a typology of configurations of network members using latent class analysis and reveal six configuration types: children, friends, kin-friend, spouse-diversified, spouse-children, and spouse-centered network. After revealing the configurations of members in each context, results confirmed that each configuration has different patterns of role expectations and exchange patterns, and thus, social and financial resources are not necessarily distributed evenly in proportion to increases in close connections---that is, an increase in the size and in the volume of interactions amplifies the complexity of expectations and exchanges among members such that we cannot assume monotonous outcomes based on these changes. Therefore, older adults in some types of living arrangements and network member configurations (i.e. composition) enjoy better health outcomes than others. Role strains and unbalanced social exchanges of tangible and emotional resources are often the underlying mechanisms that link social relation patterns and health outcomes. By accomplishing this work, I have expanded the concept of a source of stress to a micro-relational structure and develop moderating and mediating mechanisms based on the theory of social exchange and role strain. Therefore, I have reconceptualized interrelationships among source, pathways and health outcomes within sociological research.