A new study suggests patterns in common blood tests could reveal our fates as we get older.
By Nathan Collins
(Photo: Angelo Su/Flickr)
To many of us, aging means gray hair, wrinkles, and a memory that starts to go—things that, while unfortunate, seem more or less inevitable. But while two elderly people are likely to have a lot in common, not everyone grows old in quite the same way, and, now, researchers think they’ve figured out a way to predict such differences, based only on a set of relatively common blood tests.
“Because people age differently, age is not a sufficient marker of susceptibility to disabilities, morbidities, and mortality,” a team of researchers led by Paola Sebastiani writes in Aging Cell. Indeed, no one variable or even set of variables is going to tell you what diseases you’re likely to develop in your old age, or what’s ultimately going to kill you.
Not that this has stopped people from trying, of course. One broad class of approaches is to to try to calculate your “biological age,” as opposed to your chronological age, using chemical signatures such as DNA methylation. The trouble is, those measures are one-dimensional, and as such can’t capture the diversity of aging.
For a better approach, Sebastiani and her colleagues turned to data compiled as part of the Long Life Family Study, which collected detailed medical histories along with blood samples from 4,704 people between the ages of 30 and 110. From those samples, the researchers identified a total of 19 blood tests that changed with age, among them indicators of inflammation, blood cell function, diabetes, and kidney disease.
But rather than focus on one test at a time or create a measure of biological age, the researchers instead looked for patterns across the 19 tests. Using a standard method called agglomerative clustering, they identified 26 subgroups of participants who shared similar blood test results—one cluster, for example, included 1,128 people whose test results indicated slightly better than average kidney function and slightly lower than average levels of inflammation.
Next, the team looked to see if any of their subgroups were more prone to disease or other medical problems. The team used the largest subgroup (2,262 people with normal blood test results) as a reference. Compared to the reference, the team found that a cluster of 387 people with signs of diabetes and slightly better-than-average kidneys also had relatively poor grip strength, slower gaits, and reduced lung function. Another group of 140 with signs of inflammation had slower gaits, worse lungs, cognitive deficits, and higher pulse pressure (the latter is a measure of heart function). The team replicated some of the results using data from the Framingham Heart Study.
This doesn’t mean doctors can look in a crystal ball to see how you’ll age, but they may have real value for researchers and clinicians alike. “If these analyses replicate in larger cohorts, biomarker signatures could be used … in the design and analysis of clinical trials and go beyond studies to become tools for early preclinical diagnoses and more efficacious patient treatment,” the team writes.