An epigenetic pace-of-aging test was more strongly linked to risk of death than other aging biomarkers. Digest
Scientists have proposed many ways to measure biological aging, but it is still unclear which markers best predict hard outcomes such as death. A new study tested 14 aging biomarkers side by side to see which were most strongly associated with mortality.
You just missed this in your inbox
Every other week our Premium Members received this exact study plus Rhonda's practical commentary and 8+ other hand-picked papers.
The researchers analyzed data from 1,671 adults in the Berlin Aging Study II who were 60 to 80 years old at the start of the study. The biomarkers included inflammatory and hormone-related blood measures, physical function tests, cognitive performance, blood pressure, and DunedinPACE, an epigenetic score that uses chemical tags on DNA measured in blood to estimate the pace of biological aging. Researchers then examined which markers were most strongly associated with death across up to 16 years, during which 316 participants died. They first tested each biomarker on its own, then tested which smaller combination best preserved the predictive value of a larger 12-biomarker panel, since a marker that is weak alone could still add useful information when combined with others.
- DunedinPACE was the strongest individual predictor of all-cause mortality. Participants who scored one standard deviation higher on this test had a 64% higher rate of death during the study period.
- Several other markers also showed associations. Higher IL-6, an inflammatory marker, was linked to a higher rate of death, while stronger hand grip, unimpaired standing balance, unimpaired muscle strength, and better performance on a speed-based thinking test were linked to a lower rate of death.
- Other commonly discussed aging markers did not clearly predict death when analyzed on their own. These included CRP (another inflammation-related blood marker), IGF-1 (a hormone-related growth signal), blood pressure, walking speed, muscle mass, frailty status, the Timed-Up-and-Go mobility test, and an estimate of GDF15 (a protein linked to aging-related stress in the body).
- When the researchers combined DunedinPACE, muscle mass, and standing balance in one prediction model, it performed almost as well as the larger 12-biomarker model. This suggests that some aging markers may capture overlapping information rather than adding entirely new insight.
DunedinPACE may have stood out because it appears to capture both breadth and tempo. Instead of focusing on one blood protein, one body function, or one clinical test, it reads patterns of DNA methylation in blood that may reflect changes across multiple body systems. It is also designed to estimate the pace of biological aging, rather than simply estimating how old the body appears at a single moment. That combination may help DunedinPACE pick up a broader pattern of aging-related risk than narrower markers. However, a weaker link to mortality in this study does not mean those other measures are irrelevant. Some markers may just be better suited for tracking specific aspects of health and aging.
The findings are observational, so they show associations rather than proving cause and effect. A higher DunedinPACE score may identify people at greater mortality risk, but this study cannot tell whether changing that score would affect how long someone lives. Still, if these findings are confirmed in broader studies and intervention trials, aging tests such as DunedinPACE could become useful tools for tracking how lifestyle and other preventive strategies relate to long-term aging and disease risk. In episode #62, Dr. Steve Horvath discusses epigenetic aging clocks and their use in predicting biological aging and healthspan.