Two Proteins in Ordinary Blood Tests Track Who Loses Independence

The vial of blood drawn from an 85-year-old at a routine check already carries more than a cholesterol number. Two proteins that clinics can measure today, beta-2-microglobulin and cystatin C, turn out to track which of the very oldest adults will go on to lose their independence, according to a study of community-dwelling octogenarians in Japan and Italy. Higher levels of each were linked to a greater risk of developing disability years later, even after the researchers accounted for age, sex, kidney function and lifestyle.

Cystatin C, in particular, has already turned up in other large cohorts as a marker that tracks accelerating frailty and slipping physical function. The proteins do not cause the decline; they seem to register the slow biological weather, kidney strain and low-grade inflammation, that so often precedes it.

That distinction matters, because the very old are exactly the group where such signals have stayed frustratingly hidden. Routine blood work catches plenty of named diseases, yet reliable markers for the gradual slide into dependence have been elusive.

The stakes are hard to overstate in Japan, which has one of the world’s fastest-aging populations. Nearly 60 percent of Japanese adults aged 85 and older already draw on the national Long-Term Care Insurance system, so a test that flagged risk before disability set in could change who gets help, and when. The research team, led by Yusuke Osawa and Yasumichi Arai at Keio University with Luigi Ferrucci of the US National Institute on Aging, went looking for those markers the hard way: not by guessing which proteins mattered, but by screening dozens at once and letting the data point. They measured 29 plasma proteins in 230 disability-free adults aged 85 to 89 and followed them for about 4.5 years.

Two names kept surfacing. Beta-2-microglobulin (B2M) and cystatin C were each associated with a higher chance of later disability, with hazard ratios of 1.35 and 1.42 respectively. Reckon it as a persistent nudge upward in risk that survived adjustment for the usual confounders.

What Held Up Across Two Countries

A single cohort proves little on its own, so the team took the two proteins to a very different group: the InCHIANTI study, which has followed older Italians for up to 15 years. Once again, elevated B2M and cystatin C tracked with greater disability risk, most clearly among participants aged 80 and older. Other proteins that had looked promising for predicting death, including epidermal growth factor, did not replicate, which is itself a useful result. It tells you the disability signal was the sturdy one, and the mortality hunches were not.

Why these two? Both are tied to kidney function, and B2M also rises with the chronic, smouldering inflammation that gerontologists have taken to calling inflammaging, a state that itself tracks with frailty, disability and death across the aging literature. B2M is no bystander in that story: in 2015, researchers reporting in Nature Medicine identified it as a circulating pro-aging factor that climbs with age and drags on the aging body. Neither is exotic. That is rather the point: the appeal here is that clinicians would not need a new machine, only a new reason to read numbers they can already order.

Osawa, the Keio University epidemiologist who led the work, is careful about what that buys you. “Because B2M and cystatin C are already measurable using standard clinical assays, they have the potential to become practical tools for identifying older adults who may benefit from preventive support,” he says. The potential is the operative word.

His colleague Arai, who works on the biology of extreme old age at Keio, frames the shift in emphasis it points to. “Our findings suggest that preserving healthy aging requires attention not only to diseases but also to the biological processes that precede disability,” he explains.

A Signal, Not Yet a Screening Test

Here’s the caveat the headlines tend to skip. When the researchers asked their model not just to find the proteins but to actually predict who would become disabled, it managed only modest discrimination, a concordance index of 0.54, and did not reach statistical significance. The authors say their method was better at selecting variables than at forecasting individual fates. An association robust enough to replicate across two countries is a real finding; a working screening test is a different animal, and this is not yet that.

The study carries other limits. The Japanese participants were relatively healthy and disability-free at the start, which may not represent everyone in their eighties. Records did not always say why someone was certified for care, leaving room for misclassification. And the two cohorts differed in nearly everything except the question, from average age to how disability was measured, which makes the agreement between them more striking but the details harder to line up.

Still, the direction of travel is what interests the researchers. “Earlier identification of people at higher risk could create opportunities for timely interventions such as exercise, nutritional support, and rehabilitation before irreversible decline occurs,” Arai says. Whether reading the signal actually changes the ending is the next question, and it is not one a blood draw alone can answer.

Frequently Asked Questions