Can Lab Test Tarot Predict Death?
Certain Combinations of Lab Tests May Unlock Death's Door for Elderly
WebMD News Archive
Sept. 18, 2006 -- Certain combinations of 13 common lab test results predict death in healthy people over 70.
The finding comes from a detailed study of 1,189 remarkably healthy 70-something men and women, originally studied as examples of successful agingaging. Twelve years later, about half of these people were dead.
Was there some way these deaths could have been predicted? UCLA researcher Tara L. Gruenewald, PhD, and colleagues analyzed results from 13 lab tests performed when these men and women first entered the study.
These "biomarkers" covered risk factors for heart diseaseheart disease (such as blood pressure), tests of immune activation, tests for stress and steroidal hormones, and tests for metabolic factors (such as good and bad cholesterol).
Certain combinations of lab tests formed "trees" of death. For example, there were 28 deaths among 30 people in one particularly lethal "tree." This tree was a combination of high levels of adrenal hormones and high levels of C-reactive protein or CRP. CRP is linked to chronic immune activation, which is thought to underlie several different diseases.
People in two or three of these trees, it turned out, were living in a statistical forest of death.
"We found a variety of combinations that predict mortality, but certain combinations appeared frequently," Gruenewald tells WebMD. "Typically these were markers of immune activation, including CRP. And also stress hormones."
The findings appear in the Sept. 19 issue of the Proceedings of the National Academy of Sciences.
Gerontologist Robert McCann, MD, of the University of Rochester, N.Y., praises the Gruenewald team's work.
"We have known for years that a lot of these biomarkers can predict higher mortality," McCann tells WebMD. "What we don't know -- and what this work addresses -- is how these biomarkers can combine with each other to make better predictions."
Gruenewald says -- and McCann agrees -- that the study findings, while an important step in the right direction, aren't ready for prime time.
I don't think our data at this point suggests that doctors should start looking at all the biomarkers in this study," Gruenewald says. "But we should start considering the value of each one of these markers in the context of the value of another."