Diagnostics & AI
In human trials
Biological age clocks and AI biomarkers
Blood tests that estimate how fast you, or each organ, are aging.
How it works
Machine-learning models trained on DNA methylation, proteins or other data estimate biological age or pace of aging. They do not treat aging but aim to measure it and show whether an intervention is working.
Hallmarks of aging targeted
Epigenetic alterations · Altered intercellular communication
What the evidence says
Honest about what is proven, and what is not.
DunedinPACE, a DNA-methylation pace-of-aging test, is associated with morbidity, disability and mortality (Belsky, eLife 2022). Proteomic organ clocks in 44,498 UK Biobank participants predicted disease and death, with 8 or more aged organs tied to about 8-fold mortality risk (Oh, Nat Med 2025). Clocks are already used as trial endpoints (CALERIE, plasma exchange, semaglutide), but they are not yet accepted substitutes for clinical outcomes; ARPA-H’s PROSPR program is working to build validated biomarkers.
Key reference
Oh HS, Le Guen Y, Rappoport N, et al. Nat Med. 2025;31(8):2703-2711
Considering this for yourself?
Many longevity therapies are experimental, off-label or not FDA-approved for aging. A physician-guided evaluation weighs your labs, history and goals before anything is started.
Explore other emerging therapeutics
Foundations
Exercise and resistance training
The best-evidenced healthspan intervention we have: aerobic fitness plus strength.
Foundations
Sleep, nutrition and caloric restriction
Daily inputs that shape metabolic aging; the strongest data are for caloric restriction.
Metabolic
GLP-1 receptor agonists
Obesity and diabetes drugs with proven heart benefits and an open question on aging.