Effects of social relationships on mortality among the elderly in a Japanese rural area: an 88-month follow-up study.

BACKGROUND The association between social relationships and lower mortality has been well documented in Western countries. This study aims to investigate that association among elderly Japanese in a rural area. METHODS An analysis was conducted with 1,994 subjects (58.1% women), 78.3% of the total elderly aged 65 and older in a town, who were independent in activities of daily living. A baseline survey was carried out in 1992, and subjects were followed until 1999. Cox proportional hazard models examined the association between social relationships (availability of casual friend/ support provider, group membership, job, living arrangement) and an 88-month mortality. RESULTS A significant association between social relationships and mortality was observed among the old-old (aged 75 and older). Among men, having a job and group membership were significantly associated with lower mortality with hazard ratios (95% confidence intervals) of 0.62 (0.41-0.94) and 0.60 (0.40-0.90), respectively, after adjustment for age, diagnosed illnesses, self-rated health, other social relationships, annual income, and home ownership. Among women, having a job and living alone were significantly associated with lower mortality with hazard ratios (95% confidence intervals) of 0.67 (0.45-0.99) and 0.35 (0.13-0.97), respectively. CONCLUSIONS Social relationships such as having a job and group membership were associated with lower mortality among the old-old. In addition, old-old women living alone were better off in terms of mortality after adjustment for possible confounders. This suggests the importance of considering family relationships in terms of quality in areas where multi-generation households prevail.