Show summary Hide summary
People of the same calendar age can show very different molecular signs of aging, a new longitudinal study reports — a finding that complicates attempts to reduce aging to a single score. The research suggests that what scientists call biological age may reflect multiple, shifting molecular paths rather than one fixed number.
Following the same people over time
Molecular aging differs between individuals and depends on context
What to Know About the Cosmopolitan Beauty Awards
Researchers tracked 335 women, aged 32 to 80, from the long-running TwinsUK cohort across roughly eight years. Each participant visited a clinic at least three times between 2009 and 2017, allowing repeated blood sampling and repeated measurements of gene activity and small molecules produced by metabolism.
The study design lets researchers compare how molecular features change inside the same person over time, rather than comparing different people at a single moment. As study co-author Julia El-Sayed Moustafa, a computational genomics researcher at King’s College London, put it: “Molecular aging is dynamic and unique to each person.”
Which genes and metabolites shifted
The team examined more than 16,000 genes and 915 metabolites. They report that 5,061 genes and 181 metabolites showed statistically significant change over the follow-up period. Most of the changing genes — 5,036 of them — moved consistently up or down across the group, and a smaller set of metabolites showed similar uniform trends.
Still, individual variation was prominent. As co-author Kerrin Small, a genomics professor at King’s College London, noted, “even when most people’s genes or metabolites moved in one direction, we still found groups moving the opposite way.” In practice, a metabolite that rose in one woman could fall in another, and each participant’s overall metabolite profile tended to diverge from her earlier profile.
Many affected genes mapped to pathways involved in immune function, energy metabolism and conditions linked to aging, such as cardiovascular and neurodegenerative disease.
Genes, clocks and environmental signals
Some patterns pointed to inherited influence: identical twins showed more similar gene-expression trajectories than fraternal twins, indicating a substantial genetic contribution. At the same time, the molecular signals depended on timing. About one-quarter of the genes and metabolites displayed seasonal variation, and up to 40% of metabolites shifted with the body’s 24-hour internal clock — demonstrating a clear circadian component.
The researchers also observed a drop in blood levels of certain per- and polyfluoroalkyl substances (PFAS), including PFOA and PFOS. The team suspects this decline reflects policy and market changes in the U.K. rather than individual-level biology. PFAS levels correlated with some gene and metabolite changes, though the study cannot establish that the chemicals caused those molecular shifts.
How this affects biological-age measurements
Many researchers use “aging clocks” — algorithms that infer biological age from molecular markers such as DNA chemical tags — to estimate whether tissues look older or younger than a person’s chronological age. This new work suggests those clocks may capture only part of a complex picture.
Raghav Sehgal, an associate research scientist at Yale who was not involved in the study, told Live Science that the results highlight limits of single-number summaries. “No single number can fully capture the many systems that age at different rates,” he said. He added that aging may be better characterized by a combination of broad biological-age scores and more targeted measures of immune, metabolic, brain and cardiovascular aging.
Limits and next steps
The authors acknowledge key limitations: the study is observational, included only women and relied exclusively on blood samples. Those factors mean the findings need testing in larger, more diverse cohorts and in other tissues.
The research team plans to continue monitoring participants for about 15 years in total, extending the current eight-year window. Over time, mapping these individualized molecular trajectories may help researchers separate benign, age-related changes from patterns that signal disease — though the authors and outside experts say that practical clinical use remains a longer-term goal.
Study source: Longitudinal Dynamics of gene expression and metabolomics in an aging population cohort. Science, 393(6815). DOI: 10.1126/science.aed6452.
