This study followed 518,240 elderly couples enrolled in Medicare in 1993 for nine years to determine whether there was an association between hospitalization of a spouse and a partner's risk of death. The "bereavement effect"—referring to the increased risk of death in an individual whose spouse has died—is well-documented, but effects of hospitalization on spouses is not. This effect, termed "caregiver burden," has been chiefly studied not in terms of spousal mortality, but in terms of worsening health of the caregiving spouse. The authors used hospital claims to establish the nature of spousal illness as well as length and dates of hospitalizations. They also attempted to determine whether some illnesses took a more severe toll on spouses. For comparison, the bereavement effect for women is 17 percent after a husband's death and for men it is 21 percent.

Key Findings

• Hospitalization of a woman was associated with average increased risk of a husband's death of 4.5 percent; hospitalization of a man increased his wife's death risk 2.7 percent.
• Psychiatric hospitalizations took the largest toll (8.6% increase for men after a wife's hospitalization and 5.0% increase for women after a husband's) and colon cancer took among the lowest. Other diseases that severely interfere with physical and mental functioning tended to have strong effects.
• For women, effects increased with age and poverty. For men, effects increased with age only. Race was not significant.
• The first 30 days post-hospitalization was the most crucial, with deaths increasing 35 percent for men and 44 percent for women—comparable to death risks seen in spousal mortality circumstances (53% for men and 61% for women).
• Graphs of spousal mortality over two years following hospitalizations were U-shaped, with a nadir at 90-180 days.

Hypothesized mechanisms to explain caregiver burden include: loss of social, emotional, economic or other support; an increase in harmful behaviors such as drinking, bad diets, etc.; and poorly understood feedback mechanisms between stress and loss of immune functions. These findings can help time the delivery of support services to spouses and help concentrate services on particularly severe mortality inducers in spouses, such as stroke and dementia.