The relationship of social environment, social networks, and health outcomes in the Seattle Longitudinal Study: two analytical approaches.

This study examined the relation of social networks and perceived social environment to health outcomes and cost utilization over a 1-year period in a community sample of 387 (173 males, 214 females). Two analytical strategies, a variable-oriented approach and a subject-oriented approach, were used to complement each other: structural equation modeling assessed the direct relationship between social relationships and health, while cluster analysis examined how social relationship patterns were related to health outcomes. Lower levels of perceived social environment were associated with an increased number of hospital visits. For unmarried individuals, increased age was associated with greater medical usage and estimated total health care cost, while low perceived social environment was related to a greater number of doctor visits. Among married individuals, women had more frequent doctor visits, medication usage, estimated outpatient costs, and primary care visits. Married individuals with low levels of social networks had increased total health care costs, outpatient costs, and primary care visits. Cluster analysis was used to group individuals by characteristics of perceived social environment and social networks. Members of the cluster groups with greater health problems were more likely to be isolated, had the least social contact, and had lower levels of education and income. Including subject-oriented approaches is useful in complementing more popular regression methods for data analysis.