ABSTRACT
Mortality differentials by marital status have been examined in numerous studies, and show that married individuals have lower mortality rates than those who are widowed, divorced, or single (Hu and Goldman 1990; Goldman, Korenman, and Weinstein 1995; Sorlie, Backlund, and Keller 1995; Waldron, Hughes, and Brooks 1996). Among married people, loss of a spouse has been associated with excess mortality, even after controlling for potential confounding factors (Bowling 1987; Schaefer, Quesenberry, and Wi 1995; Martikainen and Valkonen 1996a). Given the stress associated with spousal loss, it is not surprising that bereavement greatly increases the risk of poorer physical and mental health (Parkes and Weiss 1983; Osterweis, Solomon, and Green 1984; Lund, Diamond, and Caserta 1985; Mor, McHorney, and Sherwood 1986; Jacobs et al. 1990; Bruce, Kim, Leaf, and Jacobs 1990; Parkes 1996). Other pathways leading from bereavement to poor health and mortality include role change (Lund, Caserta, and Dimond 1986; Bowling 1987; Lindstrom 1999) and loss of social and material support (Helsing, Szklo, and Comstock 1981; Zick and Smith 1991; Stylianos and Vachon 1993). With time, however, individuals gradually adapt to widowhood, and the risk of mortality decreases. Some studies suggest that the effect of death of a spouse is especially marked during the first weeks (Kaprio, Koskenvuo, and Rita 1987; Martikainen and Valkonen 1996a), months or years (Bowling 1987; Schaefer et al. 1995) after bereavement. Others, however, report both short-term and long-term excess mortality among widowed individuals (Mellstrom, Nilsson, Oden, Rundgren, and Svanborg 1982).
