The progress and future of biological aging clocks

A Nature Medicine review by Tony Wyss-Coray and Eric Topol maps the progress and future of biological aging clocks—from DNA methylation to organ-specific proteomic models—and why they matter for healthspan research. Clocks can predict mortality and disease risk, but they lack clinical standardization and do not yet prove causality.

Key Takeaways

• Aging clocks now span DNA methylation, plasma proteins, organ and cell-type models, and many other -omics and clinical inputs.
• Organ clocks tie organ-specific biological age to disease risk, and organs often age at different rates within one person.
• Clocks correlate with mortality and health outcomes but do not prove causality and are not clinically standardized.
• Exercise and related interventions may slow epigenetic aging; structured training can blunt parts of muscle aging.
• Refined, AI-assisted clocks could one day support prevention and early endpoints in longevity trials.

According to coverage of the review on Lifespan.io, biological aging clocks are computational models that track the pace of aging and estimate biological age for organisms, organs, or cells relative to a reference population. For readers following longevity and biohacking, that shift from philosophical debate to measurable molecular metrics is the core story.

What are biological aging clocks measuring today?

First-generation clocks were built on DNA methylation patterns. Later models added plasma proteins and other large-scale molecular arrays, plus functional measures such as hand strength, cognition, locomotion, sensory acuity, and psychological testing.

Organ- and cell-based clocks go further. Studies using organ clocks linked organ-specific biological age to organ-specific disease—for example, brain age with Alzheimer's disease. Among organ clocks, brain and immune clocks showed the strongest links to survival. Aging in one organ did not strongly track aging in others, suggesting organs age at different rates in the same person. Cell-type-specific clocks have also tied accelerated aging in particular cells to disease risk.

Beyond those designs, researchers build clocks from transcriptomics, glycomics, metabolomics, lipidomics, microbiomics, electronic health records, lifestyle factors, telomere length, standard lab tests, medical imaging, and physical measures such as grip strength and gait speed. Predictive power varies by system and data type; immune complexity, for example, makes that system especially hard to clock.

What still limits these tools in real-world care?

Training datasets can carry demographic bias when drawn from limited populations. Clocks have been used to predict all-cause and cause-specific mortality, healthspan, and other outcomes, yet they do not establish causality. Most evidence comes from population-level studies, so usefulness for individuals remains unclear.

Aging clocks are not established for clinical use. Consumer products exist, but the field lacks standardization and regulatory approval. That gap matters as much as technical progress for anyone hoping to use biological age as a personal health metric.

Where does the progress and future of aging clocks lead?

Combined with other tools, epigenetic and proteomic clocks have helped link biomarkers to disease mechanisms and potential drug targets. Studies also associate aging measures with medications, lifestyle, menopause, foods, and occupations. Epigenetic aging can be slowed by exercise and, to a lesser degree, by omega-3 and vitamin D, multivitamins, the shingles vaccine, and GLP-1 drugs.

A separate Nature Aging study found that planned training—not just daily activity—erased more than half of age-related gene-expression changes in older muscle, especially in mitochondrial and energy pathways, while roughly half of changes remained "exercise-proof."

Authors of the review argue that refined clocks, especially with AI, could enter mainstream care and help prevent age-related disease rather than only treat it. Using clocks as substitutes for long clinical follow-up still needs validation, but measuring early endpoints of geroprotective interventions is a near-term goal.