Ageing is a complex biological process, and accurately measuring how the body declines over time has long challenged scientists. People of the same chronological age can experience very different health trajectories, making it difficult to determine biological age and identify the true drivers of ageing. To address this, researchers from China’s Aging Biomarker Consortium (ABC) developed a computational framework called the Digital Aging Twin to predict biological age and track how different organs age over time.
The study, published in the journal Cell on May 8, involved researchers from the Chinese Academy of Sciences, Capital Medical University, and several other institutions. The team recruited 2,019 healthy participants aged 18 to 91 from multiple Chinese cities to establish a standardized multicentre cohort known as mCAS (multicentric Chinese Aging Standardized). Researchers gathered more than 240 physiological and molecular measurements from each participant, including clinical tests, brain and retinal imaging, cognitive and motor assessments, DNA methylation, proteins, metabolites, and gut microbiome data.
Using more than one billion high-quality data points, the researchers created a three-tier system of “ageing clocks.” The first tier, called the core capacity clock, integrates 240 physiological indicators to measure overall functional decline. The second tier, the multimodal clock, combines multiple layers of molecular data using deep-learning methods to estimate biological age with high accuracy, achieving a mean absolute error of only 3.87 years. The third tier consists of organ-specific clocks for organs such as the brain, liver, lungs, muscles, blood vessels, and skin.
One of the study’s most notable findings was that organs age at different rates. Researchers discovered that the liver reaches a major ageing turning point around age 40, while the brain’s ageing accelerates closer to age 50. The analysis also revealed two major waves of ageing-related change: one between ages 40 and 50 and another between ages 60 and 70. These findings suggest that ageing is not a uniform process but occurs in distinct phases across different organ systems.
To investigate the biological drivers of ageing, the researchers examined plasma proteins, human liver tissue, cultured human cells, and animal models. They identified liver-derived coagulation factors — particularly F13B, F9, and F10 — as direct contributors to vascular and systemic ageing. Experiments showed that exposing human vascular cells to these factors triggered signs of cellular senescence and inflammation, while injecting F13B into mice accelerated ageing across multiple tissues, including the liver, heart, kidneys, and blood vessels.
The researchers also developed simplified “proxy clocks” based on just 100 to 108 plasma proteins, raising the possibility that future blood tests could provide comprehensive ageing assessments. The study further identified lifestyle factors associated with slower biological ageing, including greater fruit intake, regular sleep schedules, and moderate walking, while smoking, poor sleep, and frequent eating were linked to accelerated ageing. Although the framework currently relies on cross-sectional data, researchers say the Digital Aging Twin represents a major shift in ageing science — from simply describing ageing to predicting it and identifying actionable biological drivers that could guide future interventions.
More information: Jiaming Li et al, Multimodal clocks of human aging, Cell. DOI: 10.1016/j.cell.2026.04.025
Journal information: Cell Provided by Chinese Academy of Sciences Headquarters
