A Blood Metabolite Clock Built on 18,476 People, Ages 1 to 96

Jeff's Daily Pick. A single blood draw can tell how old your metabolism looks. In 18,476 people, those whose blood looked older than their age had more diabetes, fatty liver and deaths.

In China, a team led from Shanghai Jiao Tong University reported in Science Translational Medicine in October 2026 one of the largest plasma metabolomic studies of human aging to date. Using nano assisted laser desorption and ionization mass spectrometry, the researchers profiled the blood of 18,476 participants aged 1 to 96 at six centers.

Building the clock

The analysis identified 56 metabolites whose levels shifted with age across the lifespan. They fall into recognizable groups: collagen turnover, amino acid and one carbon metabolism, redox metabolism of the gamma glutamyl cycle, kidney and muscle turnover, and compounds made jointly by the host and gut microbes. Machine learning combined them into a predicted metabolomic age. The key metric is metabolomic age acceleration, the difference between that predicted age and the calendar age.

What faster metabolic aging predicted

Participants in the top quarter of age acceleration had significantly higher risks of developing metabolic associated fatty liver disease, hypertension and type 2 diabetes, and higher all cause mortality, than those in the bottom quarter, according to Inside Precision Medicine's summary of the paper. High acceleration went with smoking, shift work and unhealthy diet; low acceleration with mixed grain intake and exercise. Within the same people over time, more physical activity and less fried food were associated with favorable changes, and new exposure to secondhand smoke with unfavorable ones. A second metric, age deviation, captured departures from the mortality risk expected for age and tracked disease burden and multimorbidity.

Metabolite clocks are cheaper and faster than epigenetic clocks and read physiology closer to the present moment. The open question is the one every aging clock faces: whether it moves in a trial where an intervention also changes hard outcomes. Our guide to biological age clocks explains why that link is still missing for most of them.