Networks and Neighborhoods

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): This Program Project involves an integrated and innovative set of investigations into the role of social networks and residential neighborhoods in the health behaviors, health care, and health outcomes of individuals. We have two overarching goals: (1) to ask new questions about the roles of social networks and local social environments in health; and (2) to develop methods for answering such questions using new statistical techniques and using new approaches to causal inference in the social sciences that exploit genetic information. We will focus on pressing problems in health and health care in the U.S., including cardiovascular disease, cancer, obesity, violence, and substance use. There are eight constituent Projects: (1) "The Spread of Health Behaviors in the Framingham Heart Study (FHS) Social Network" examines the extent of person-to-person spread in smoking, drinking, physical activity, healthy eating, and weight gain in a unique, densely inter-connected network of 12,630 people followed from 1971 to the present (the "FHS-Net" data); (2) "Genes as Instrumental Variables for Health Behavior Peer Effects in the FHS-Net" examines peer effects in obesity and smoking and exploits detailed genetic information as instrumental variables; (3) "Neighborhood Effects on Health in Framingham and the Role of Social Networks" uses the FHS-Net data, and its detailed address information, to evaluate the relative contribution of a person's social network and residential neighborhood to health; (4) "Neighborhood Effects on Cancer Course in the Aged" evaluates the influence of geography across the full course of cancer in 550,000 elderly patients from around the U.S.; (5) "Informal Physician Networks and Patterns of Care for Medicare Beneficiaries" examines the informal networks of up to 150,000 primary and specialist physicians that share patients or information and how these networks affect patient care; (6) "Neighborhood Networks and Social Mechanisms of Trajectories in Well Being" uses the PHDCN data set regarding over 6,000 children to examine how changing networks and neighborhoods affect health outcomes; (7) "The Genetic Basis of Civic Engagement and Social Networks and their Effect on Health" uses the AddHealth dataset, and 2,030 of its subjects, to investigate the genetic determinants of social networks and cooperation, and their role in obesity, smoking, and drinking; and (8) "Methods for the Analysis of Longitudinal Social Network Data" develops novel statistical methods to account for the complex correlational structure of network data. These projects are supported by three Cores: an Administrative Core, a Data and Statistics Core, and a Junior Faculty Development Core. The projects build on one another and shed light on important aspects of clinical care, basic social science, and health policy.