Death after Widow(er)hood: An Analysis of Mortality Rates up to 13 Years after Bereavement

Abstract

The mortality rates of a national random sample of 503 elderly people who were widowed in 1979 were analyzed. Over the thirteen and a quarter year period analyzed, 62 percent of the sample had died. Although earlier analyses had indicated excess mortality in the first six months of bereavement for men aged 75+, there were no further significant excess mortality rates over the thirteen and a quarter year period analyzed. The data were analyzed by fitting Cox's proportional hazard regression model to the widowed's survival time for factors predicting survival time. The best fitting model for males and females indicated that risk of mortality at each period was associated with older age, male sex, and poorer functional ability (measured in 1979). Frequency of telephone contacts as a risk factor for survival period was only significant for the younger widowed.