Function Beter than Disease at Predicting Death Among Elderly

June 18, 2008
Contact: Steve Tokar (415) 221-4810

Among people 80 and older, limits on physical functioning predict the likelihood of dying within five years more accurately than do chronic diseases, according to a study by researchers at the San Francisco VA Medical Center.

"When we compare the effectiveness of hospitals and health plans, we look at mortality statistics, and the measure we use today to predict patient mortality is chronic disease diagnosis," lead author Dr. Sei Lee, a geriatrics researcher at SFVAMC and an assistant adjunct professor of medicine at the University of California, San Francisco, said. "But our findings suggest that for a really fair comparison of hospitals and health plans, we need to start looking at functional limitation measures, especially among patients who are 80 and older.

"This is significant news for risk adjustors, policy makers, and others whose job it is to assess the effectiveness of our health care system, since the over-80 group is the fastest-growing age segment in America,"Lee said.

The study appears in the July 2008 issue of the American Journal of Public Health.

In the study, researchers examined the health status reported in 1998 by 19,430 participants in the Health and Retirement Study, an ongoing national prospective study of health, income, and wealth sponsored by the National Institute on Aging. Participants were divided into age groups by decade. The authors then determined whether the participants were alive or dead in 2004.

For participants aged 50 to 59, chronic disease conditions such as high blood pressure and diabetes were significantly stronger predictors of five-year mortality than functional measures such as the ability to get out of bed unassisted, balance a checkbook, or walk unaided.

As participants aged, the predictive power of both measures began to decline somewhat, but that of chronic disease declined faster. Among participants aged 80 and older, functional limitation proved to be the stronger predictor of mortality by a statistically significant margin.

The study did not explore the reasons for the increasing predictive power of functional limitation with age, but Lee speculates on two possible causes.

The first is what he calls "a selection or survival effect." For a younger person, he explains, high blood pressure or diabetes "can be a very bad thing, with a likelihood of very bad consequences. But if you're 80, and you've lived with high blood pressure for 30 years, it's possible that it just doesn't mean the same thing to you as it does to the average person — that your body has adjusted to it in some way. And so it won't have nearly as much bearing on your chances of living another five years as your ability to walk down the block unaided or manage your own medications without help."

Another potential reason has to do with the "trajectory of decline among our oldest subjects," he said. Younger people are more likely to have a clear single cause of decline, he says, such as a heart attack or pneumonia. Older people, by contrast, are more likely to decline slowly, not as a result of one catastrophic event but "slowly, month by month, week by week, for no clear reason that you can put your finger on. Functional measures are much more likely to capture that trajectory of decline."

Lee noted that functional status is known to be an important component of quality of life for older people. "Most older patients will tell you that they are more afraid of losing their independence than they are of dying," he said.

As a follow up to the current study, Lee says that "even though self-reporting is well-known to be a valid indicator of health status, it would make sense to replicate this work with a larger sample and an objective data set such as formal medical records."

If his results are validated, he says, "I would strongly advocate that functional status become a standard component of medical records. It would be an important tool for policy makers, clinicians, and patients."

Co-authors of the study are Dr. Alan S. Go, of Northern California Kaiser Permanente, Oakland, Calif.; Karla Lindquist, MS, of UCSF; and Daniel Bertenthal, MPH, and Dr. Kenneth E. Covinsky, of SFVAMC and UCSF.

The research was supported by grants from the National Institute on Aging, some of which were administered by the Northern California Institute for Research and Education.

NCIRE — the Veterans Health Research Institute — is the largest research institute associated with a VA medical center. Its mission is to improve the health and well-being of veterans and the general public by supporting a world-class biomedical research program conducted by the UCSF faculty at SFVAMC.

SFVAMC has the largest medical research program in the national VA system, with more than 200 research scientists, all of whom are faculty members at UCSF.

This news release has been modified for the website