Abstract
PURPOSE: The aims of this study were to 1) characterize and visualize the association between social connection and physical functioning using network analysis, with attention to different types of measures of these constructs, and 2) identify key indicators that link social connection and physical functioning. METHODS: Data come from the 2014/2016 Health and Retirement Study (N = 7,270, mean age = 67.3). Network analysis was used to depict and explore the relationships between physical functioning and social connection. Both of these constructs were measured using a combination of "objective" (e.g., balance test, number of social connections) and "subjective" (e.g., self-evaluated activities of daily living (ADLs), perceived loneliness) indicators. The network was estimated using a regularized partial correlation network. Features of the network structure were characterized using 1) assortativity, 2) community detection, and 3) centrality indices (i.e., strength and betweenness). Betweenness centrality was used to identify the key nodes linking the two constructs. RESULTS: The network structure is determined by constructs (i.e., social connection vs. physical functioning, assortativity = 0.87), instead of types of measures (i.e., objective vs. subjective, assortativity = 0.45). There are five communities identified from the network, three of which include both subjective and objective indicators. The incomplete balance test is the key bridging indicator that links social connection and physical functioning. CONCLUSION: The linkages between social connection and physical functioning are multifaceted. Balance impairment is the key bridge between these constructs for older adults.