DESCRIPTION (Applicant's abstract): Social relationships have been reliably related to morbidity and mortality. However, the joint contribution of positive and negative aspects of social relationships, and its association to health, has not been adequately examined. In the proposed research, we present a general conceptual model suggesting that an individual's social network members may be heuristically categorized as sources of positivity (i.e., social support), sources of negativity (i.e., social aversion), or sources of both positivity and negativity (i.e., social ambivalence). The major aims of this research are to examine (a) the utility of this general conceptual framework for analyzing the health effects of social relationships and (b) potential mechanisms responsible for links between social relationships and health at differing levels of analysis (e.g.. health behaviors, psychological states, physiological function). We pursue these aims in 6 integrative studies. In our first two studies, we investigate the prevalence, mental health correlates, and ambulatory cardiovascular correlates of these different categories of social relationships. Studies 3 and 4 utilize laboratory paradigms to examine cardiovascular reactions to situations involving different categories of social relationships. In the first of these studies, we explore if the activation of a schemata associated with a supportive, aversive, or ambivalent network tie is associated with differences in cardiovascular reactivity. In our second laboratory study, we examine if supportive behaviors reduce cardiovascular reactivity to acute stress when the network member is a source of ambivalence. Study 5 investigates the developmental processes that may influence the formation of different social networks. Finally we examine an application of our model to a biologically significant outcome using an influenza vaccination paradigm (study 6). Importantly, our preliminary research provides evidence for the feasibility and potential of the proposed research.