Over the past quarter century, much research has convincingly documented the beneficial effects of social networks and social support on morbidity, mortality, and a variety ofpositive chronic illness outcomes.11,12 In the late 1970s and early 1980s, simple measuresof social ties were found to predict mortality in Alameda County, California; Tecumseh,Michigan; Durham County, North Carolina; and Framingham, Massachusetts (seeBerkman and Glass11 and Kaplan and Toshima12 for reviews). In the early to mid-1980s,the focus shifted from the effects of network structure on health to the functions providedby that network structure,11 and a copious amount of research examined the differenttypes of support provided by social networks and the relationships between social supportand health outcomes. Although much of this research used widely different operationaland conceptual definitions of the support construct, and some was fraught with methodologicalproblems, by the end of the decade the body of literature across many social sciencedisciplines convincingly demonstrated that supportive social relationships have protectiveeffects for a variety of physical and mental health outcomes.11,12