The Effects of Social Relationships and Self-Efficacy On Mortality in Older Men and Women

Purpose: Research demonstrates that social relationships are substantially associated with mortality risks among older adults (e.g., House et al., 1988). Social support, as a coping resource, may influence mortality directly or indirectly by buffering harmful effects of life stress on mortality. There still remain many ambiguities, however, about the extent and the ways in which social relationships are associated with mortality in late life. For example, most studies have focused on structural aspects of social relationships (e.g., network size) with relatively little interest on formal and/or informal integration. Further, we have little knowledge of the effects of self-efficacy on mortality and of gender differences in the link between social relationships and mortality. This study examines whether and in what extent social relationships and self-efficacy are associated with mortality in older men and women.    

Methods: Data came from the nationally representative Americans' Changing Lives study, for the subgroup of respondents aged 60 to 96 (n=1,669). The dependent variable is all-cause mortality. Survival time was defined as the number of months from the month of first interview at ACL wave 1 (1986) to either the month of death or the end of ACL wave 3 survey in 1994. Aspects of social relationships were measured using such measures as marital status, formal integration (e.g., community activities), and informal integration (e.g., social contact). Self-efficacy was measured using six items that asked about respondents’ belief in their own competency (e.g., “I take a positive attitude toward myself” “I am no good at all”). This study included other important variables (e.g., gender, race, education, family income, and functional health) to control for spurious effects. A series of hierarchical Cox proportional hazards models (1972) were fit to estimate the relative mortality risk of social relationships and self-efficacy during the 7.5 follow-up period after adjustment for other risk factors.

Results: Findings showed substantial variations in the effects on mortality across different aspects of social relationships and gender. Formal integration and self-efficacy were predictive of mortality in older women. In contrast, informal integration was not associated with mortality in older women. Especially self-efficacy had more potent effects on mortality in older women. Social relationships and self-efficacy, however, had no substantial effects on mortality in older men. Age, family income, and functional health were associated with mortality risks in both older men and women. Overall findings demonstrated that social relationships and self-efficacy had significant effects on mortality in older women but not in older men.  

Implications: This study sheds lights on the importance of taking into account social relationships and self-efficacy in examining the risk and protective factors for mortality in older adults. Especially the findings showing the strong effects of formal integration and self-efficacy on mortality in older women indicate the critical role of social engagement in community activities and self-efficacy as coping resources for physical health in older women. Thus, social work interventions for older adults, especially older women, need to focus on promoting formal integration and personal resources as well as building support in informal networks.