With coauthor Paul Allison, PhD, professor and chair of the University of Pennsylvania Department of Sociology, Christakis examined the effect of illness in one spouse on the risk of illness in a partner--commonly called the "caregiver burden." Concurrently, they looked at the effect of the death of one spouse on the mortality of the other, known as the "widower effect." By considering these effects together for the first time, the authors were able to assess the implications of specific illnesses or diseases for a partner's risk of death.
The findings, says Christakis, are striking. "When a spouse is hospitalized, the partner's risk of death increases significantly and remains elevated for up to two years," he notes. The period of greatest risk is over the short run, within 30 days of a spouse's hospitalization or death. Over this time frame, hospitalization in a spouse can confer to a partner almost as much risk of dying as the actual death of a spouse.
"Spousal illness or death may impose stress on a partner or deprive a partner of social, emotional, economic, or other practical support," says Christakis. "When a spouse falls ill or dies, partners may increase harmful behavior, such as drinking. Stress and lack of social support may also adversely affect immunologic measures, so spousal hospitalization may have physiological effects on partners."
"This is a highly innovative study--in an enormous sample of older people--demonstrates yet another important connection between social networks and health," says Richard M. Suzman, PhD, associate director of the National Institute of Aging for Behavior and Social Research. "We don't yet know the full extent to which social networks affect health. We need to explore the mechanisms behind the stresses associated with these hospitalizations as we look for ways to protect people when their central relationships are disrupted."
The study found that certain illnesses in a spouse are more harmful to partners than others. The more a disease interferes with a spouse's physical or mental ability, and the more burdensome it is, the worse for the health of the partner.
The study also found that specific illnesses impacted the spouse differently. For instance, among men with hospitalized wives, if the wife was hospitalized for colon cancer, there was almost no effect on her husband's mortality. But if a wife was hospitalized with heart disease, her husband's risk of death was 12 percent higher than it would be if the wife was not sick at all. If a wife was hospitalized with dementia, her husband's risk was 22 percent higher. Similar effects were seen in women whose husbands were hospitalized.
Further analyses also confirmed that the death of a spouse within the past 30 days was harmful. The death of a wife in the previous 30 days increased her husband's risk of death 53 percent, and the death of a husband increased his wife's risk of death by 61 percent.
Beginning in 1993 and continuing for nine years, the authors studied 518,240 couples--more than 1 million people--between the ages of 65 and 98 who were enrolled in Medicare. Over the nine years, 383,480 husbands (74 percent) and 347,269 wives (67 percent) were hospitalized at least once, and 252,557 husbands (49 percent) and 156,004 wives (30 percent) died. The mean age of men in the study was 75 years and the mean age of women was 72 years. This is the largest study of its kind.
Additionally, the authors considered how age, race, and the level of poverty modified the mortality risk of a having a sick spouse. For women, the effect of a husband's hospitalization increased with age and the level of poverty. For men, the effect of a wife's hospitalization increased only with age.
Although illness and death both create stress and have a negative effect on social support for a partner, these results operate over various time frames. This study found that the stress effect may last for a few weeks or months, and the support effect may last for several years. Eventually, the healthy partner adapts to the stress effect, so that the health risks of being a caregiver decline. However, the lack of social support that is associated with the illness or death of a partner becomes a problem, and health risks in the partner increase again.
"Our work suggests that interventions might decrease the mortality of caregivers," says Christakis. "Interventions are likely to be useful in certain diseases, such as stroke and dementia, and the timing of such interventions might be matched to the riskiest times for caregivers, for example, just after the hospitalization of the spouse."
"It seems clear that a person's illness or death can have health consequences for others in his or her social network," Christakis says. "This means that efforts to reduce disease, disability, and death can be self-reinforcing, since a decrease in the burden of these events in one person can have a cascading benefit for others. The training and assistance of caregivers can lower costs and also improve the health of patients and caregivers alike."
This work was supported by the National Institute on Aging, a division of the National Institutes of Health.
Risk of Death of a Partner after the Hospitalization or Death of a Spouse, According to the Spouse's Diagnosis
From Table 2, NEJM paper by Christakis and Allison, Embargoed until Feb. 15, 5pm US EST
| Spousal diagnosis on hospitalization | Male Partner | Female Partner |
| Dementia | 22 percent increase | 28 percent increase |
| Psychiatric disease | 19 percent increase | 32 percent increase |
| Hip or other serious fracture | 15 percent increase | 11 percent increase |
| Chronic obstructive pulminary disease | 12 percent increase | 13 percent increase |
| Congestive heart failure | 12 percent increase | 15 percent increase |
| Stroke | 6 percent increase | 5 percent increase |
| Ischemic heart disease | 5 percent increase | no effect |
| Sepsis | 9 percent increase | no effect |
| Pneumonia | 6 percent increase | 6 percent increase |
| Abdominal surgical disease | 4 percent increase | no effect |
| Leukemia or lymphoma | no effect | no effect |
| Pancreatic Cancer | no effect | no effect |
| Colon Cancer | no effect | no effect |
| Lung Cancer | no effect | no effect |
| All other forms of cancer | no effect | no effect |
| All other diagnoses | 2 percent increase | no effect |
Spousal Death: Overall widower effect 21 percent increase 17 percent increase
Bio of Nicholas Christakis, MD, PhD
Nicholas Christakis, MD, PhD, is an internist and social scientist who investigates social factors that affect the delivery and outcomes of medical care. He is a professor in the Department of Health Care Policy at Harvard Medical School and in the Department of Sociology at the Harvard Faculty of Arts and Sciences; he is also an attending physician in the Palliative Medicine Program at Massachusetts General Hospital, caring for people who are terminally ill.
Christakis' past work has examined the role of prognosis in medicine, ways of improving end-of-life care, and the impact of a good versus bad deaths on the health of bereaved spouses. Currently, he is principally concerned with how illness, health risks, and death in one person can spread and have consequences for others in a person's social network. Some current work, including the present study, is focused on the health benefits of marriage and on how ill health in one spouse can have cascading effects on the other spouse. It seems likely that improving the health of one partner in a couple can have meaningful effects on the health of the other, and that both of the parties and society-at-large would benefit from this.
Bio of Paul Allison, PhD
Paul D. Allison, PhD, is professor and chair of the Department of Sociology at the University of Pennsylvania. Since 1976, he has published seven books and 40 articles on statistical methods in the social sciences. These have dealt with a variety of methods including linear regression, log-linear analysis, logit analysis, probit analysis, measurement error, inequality measures, missing data, Markov processes, and event history analysis. At present, his principal methodological research is focused on the analysis of longitudinal data, especially with determining the causes and consequences of events, and on methods for handling missing data.
Bio of Suzanne Salamon, MD
Suzanne Salamon is a geriatrician at Beth Israel Deaconess Medical Center, a teaching affiliate of Harvard Medical School. She is the associate chief for clinical geriatrics there and is an instructor at Harvard Medical School. She is also a volunteer in the Medical Reserve Corps in Brookline, Massachusetts.
HARVARD MEDICAL SCHOOL
http://hms.harvard.edu/
Harvard Medical School has more than 7,000 full-time faculty working in eight academic departments based at the School's Boston quadrangle or in one of 47 academic departments at 18 Harvard teaching hospitals and research institutes. Those Harvard hospitals and research institutions include Beth Israel Deaconess Medical Center, Brigham and Women's Hospital, Cambridge Health Alliance, the CBR Institute for Biomedical Research, Children's Hospital Boston, Dana-Farber Cancer Institute, Forsyth Institute, Harvard Pilgrim Health Care, Joslin Diabetes Center, Judge Baker Children's Center, Massachusetts Eye and Ear Infirmary, Massachusetts General Hospital, Massachusetts Mental Health Center, McLean Hospital, Mount Auburn Hospital, Schepens Eye Research Institute, Spaulding Rehabilitation Hospital, and VA Boston Healthcare System.
Journal
New England Journal of Medicine