Seniors who care for their seriously ill spouses may pay the ultimate price.

A groundbreaking new study shows that caring for a sick spouse can raise the caretaker's risk of death. For some particularly disabling illnesses -- dementia, in particular -- the toll on the caretaker is worse than the toll of a spouse's death.

It's been known for more than 150 years that the death of a spouse ups the surviving partner's risk of death. Now it's shown that illness, too, can break your heart.

"We showed you can die of a broken heart not just when your partner dies, but when your partner falls ill," researcher Nicholas Christakis, MD, PhD, said in a news conference. "We showed it is not just death that can give you a broken heart, but illness -- even when the spouses don't die."

Christakis, professor health care policy at Harvard Medical School in Boston, and colleagues report the findings in the Feb. 16 issue of The New England Journal of Medicine.

Christakis and colleagues analyzed Medicare data on more than half a million couples aged 65 to 98. The average age of men in the study was 75; the average age for women was 73.

The death of a spouse increased a man's risk of death by 21 percent and a woman's risk of death by 17 percent. Overall, the illness of a spouse was only one-fifth as deadly to caretakers as the death of a spouse.

But some spousal illnesses took at least as great a toll as death:

--A spouse's psychiatric illness raised the risk of death by 19 percent for men and by 32 percent for women.

--A spouse's dementia raised the risk of death by 22 percent for men and by 28 percent for women.

--Other diseases that take a heavy toll on caretakers included heart failure, hip fracture or other serious fracture, and chronic lung disease.

--A spouse's cancer did not increase caretaker death risk.

"It is the disablement and not the lethality of a spouse's illness that can be harmful to you and contribute to your risk of dying," Christakis says.

There's another factor that makes it hard on spouses: poverty.

"If you are living at the margin, economically or in terms of age or being sicker, you are more vulnerable to your spouse being sick," Christakis says. "If I am richer or younger, it is not as big a shock."

Poorer elderly people in the U.S. have limited access to health care, notes Suzanne Salamon, MD, associate chief for clinical geriatrics at Beth Israel Deaconess Medical Center in Boston. Salamon was not involved in the Christakis study but participated in the Harvard news conference.

"What I have noted with poor people is there are more of the things that affect your health negatively," Salamon says. "You often have more health problems in poor people because of obesity-related issues -- high blood pressure, diabetes, small strokes -- and because of less access to health care."

Spouse Illness: Early Risk and Later Risk

Why does a spouse's illness boost a partner's risk of death? There are two danger periods, Christakis says.

In the first weeks or months there's a stress effect. This may lead to an increase in harmful behaviors -- drinking, for example, or eating an unhealthy diet. And the stress may also lead to illnesses and accidents.

"Initially the partner is at increased risk from heart attack, suicide, and accidents," Christakis says. "I am shocked my wife is sick and I stop paying attention when driving. And there is an increase in infections."

Eventually, these stress effects diminish. But then the partner faces a new set of health risks from waning social support.

"At first there are a lot of people around and there is a lot of support," Salamon says. "But as time goes by, people don't come by, people don't bring food. And over time the loneliness sets in."

Your Health, My Health Connected

Christakis hopes that one result of the study will be that health care providers and insurers will realize that they can save future costs by giving more attention to the health of caretakers.

"What our work does is shed light on particular vulnerabilities to elderly people, and show there are time windows to target interventions," he says. "Seeing people as interconnected might change the way we see the costs of health care. Taking care of both spouses while one is dying increases the health benefits for the surviving partner."

What about other relationships?

"Because people are interconnected, we think this phenomenon we studied in elderly married couples applies more generally," Christakis says. "We are looking at broader connections -- between parent and child, brother and sister, neighbors, and friends."

By Daniel J. DeNoon, reviewed by Louise Chang, MD

SOURCES: Christakis, N.A. The New England Journal of Medicine, Feb. 16, 2006; vol 354: pp 719-730. Nicholas Christakis, MD, PhD, professor, department of health care policy, Harvard Medical School, Boston. Suzanne Salamon, MD, associate chief, clinical geriatrics, Beth Israel Deaconess Medical Center, Boston.