PURPOSE: To assess the impact of changes in marital status on the mortality of elderly Japanese men and women. METHODS: In a baseline survey conducted in 1992, 2039 male and 1466 female residents in Takayama City, Gifu, who were married and aged 65 years or over responded to a detailed health and lifestyle questionnaire. Information regarding deaths of subjects and their spouses, the causes of death, and whether the subjects and spouses moved away from the city between 1992 and 1999 was obtained from the National Vital Statistics and the residential registers of the city. A proportional hazard model was used including marital status as time- varying independent variable. RESULTS: During the study period, six men and two women became separated/divorced and 151 men and 448 women became widowed. Widowhood was not significantly associated with mortality in men. Duration of widowhood was significantly inversely associated with mortality in women (p = 0.04). A significant decreased hazard ratio (=0.40, p = 0.04) was observed for women widowed for 3 years or more. CONCLUSIONS: We found no evidence indicating that widowed men and women have an increased mortality rate. Instead, the data suggested a decreased mortality rate among long-term widowed women.
PURPOSE: To assess the impact of changes in marital status on the mortality of elderly Japanese men and women. METHODS: In a baseline survey conducted in 1992, 2039 male and 1466 female residents in Takayama City, Gifu, who were married and aged 65 years or over responded to a detailed health and lifestyle questionnaire. Information regarding deaths of subjects and their spouses, the causes of death, and whether the subjects and spouses moved away from the city between 1992 and 1999 was obtained from the National Vital Statistics and the residential registers of the city. A proportional hazard model was used including marital status as time- varying independent variable. RESULTS: During the study period, six men and two women became separated/divorced and 151 men and 448 women became widowed. Widowhood was not significantly associated with mortality in men. Duration of widowhood was significantly inversely associated with mortality in women (p = 0.04). A significant decreased hazard ratio (=0.40, p = 0.04) was observed for women widowed for 3 years or more. CONCLUSIONS: We found no evidence indicating that widowed men and women have an increased mortality rate. Instead, the data suggested a decreased mortality rate among long-term widowed women.