Human lifespan has nearly doubled over the past century, but healthspan has not kept pace, leaving more people worldwide affected by age-related cardiovascular disease, cancer, dementia and frailty. Longevity and rejuvenation research, once regarded as speculative, has become a mainstream scientific pursuit, driven by converging advances in molecular biology, genomics, high-throughput proteomics, single-cell technologies and artificial intelligence (AI). Central to this progress is the recognition that aging is not a single uniform process but a collection of molecular and physiological changes unfolding at different rates within cells, tissues and organs of the same individual.
Despite these conceptual advances, translating the biology of aging into clinically useful tools has remained difficult. A key technical and scientific gap has been the lack of reliable metrics capable of quantifying the pace of aging at the level of the whole organism, individual organs and specific cell types. In this review, the authors set out to critically appraise the current landscape of so-called biological aging clocks, computational models built on epigenetic, proteomic, cellular, immune, imaging-based and multi-omic data, and to assess how these tools might inform the biology of aging, support disease prevention, and ultimately serve as endpoints in clinical trials of geroprotective interventions.
The review synthesizes evidence across several generations of aging clocks. Epigenetic clocks have progressed through successive generations, including PhenoAge, GrimAge2 and DunedinPACE, which are widely assessed for prediction of mortality, health outcomes and the pace of aging, to fourth-generation deep learning models such as AdaptAge and CpGPT. In parallel, proteomic clocks derived from plasma protein measurements have been used to construct organ-specific and cell-type-specific “clocks,” while other approaches draw on transcriptomic, metabolomic, immune and imaging data, including retinal, cardiac and brain imaging. The authors also review evidence for the nonlinearity of aging, describing coordinated molecular and structural “waves” identified across independent cohort studies, and examine interventions, ranging from lifestyle changes such as exercise and caloric restriction to pharmacological agents including rapamycin, metformin and GLP-1 receptor agonists, that have been tested for their capacity to modulate clock-based aging measures.
Several quantitative findings stand out. A consensus curve compiled across seven studies identifies coordinated aging transitions clustering around ages 33, 60, 69 and 78 years, pointing to three major phases of adult human aging. In a study of roughly 55,000 UK Biobank participants, advanced proteomic-based brain age was associated with a 3.1-fold increase in the hazard of developing Alzheimer’s disease (AD), independent of APOE genotype, while brain and immune system clocks showed the strongest links to survival among organ clocks examined. A related cell-type-resolved proteomic study of roughly 60,000 individuals found that amyotrophic lateral sclerosis showed the strongest association with accelerated skeletal myocyte aging, with a hazard ratio of 12.7 comparing extreme accelerated to youthful aging, and that individuals with 20 or more extremely aged cell types had markedly reduced 15-year survival relative to normal agers. Findings from the UK’s National Survey of Health and Development further showed that extreme organ aging, particularly when affecting three or more organs simultaneously, was associated with elevated mortality risk, and identified factors such as normal birthweight, higher childhood socioeconomic status, sustained physical activity and moderate alcohol intake as associated with healthier organ aging trajectories.
The authors indicate that these clocks hold potential significance for several applications, including identifying individuals at elevated risk of specific diseases before symptoms appear, serving as surrogate endpoints in trials of geroprotective drugs, and supporting a shift toward primary prevention of age-related conditions such as AD, cardiovascular disease and cancer through integration with genomic data, and multimodal AI. They note an ongoing clinical trial (ClinicalTrials.gov identifier NCT07646054) that will test whether an intensive lifestyle coaching intervention can reduce plasma phosphorylated tau217 and slow proteomic brain clock aging in individuals at high risk of AD, using these markers as surrogates for delaying or preventing AD. At the same time, the review emphasizes substantial unresolved limitations: the evidence base rests largely on cross-sectional cohorts with few longitudinal studies, the causal role of clock-associated markers remains poorly understood, training datasets such as the UK Biobank lack diversity, and commercially available aging-clock tests currently lack standardization and robust clinical validation despite being sold without regulatory approval.
Author: Ziya Burak Ekinci
Editor: Nehir Necem Ünlü
Reference: Wyss-Coray, T. & Topol, E. J (2026). Biological aging clocks in health and disease. Nature Medicine 32, 2383–2394). https://doi.org/10.1038/s41591-026-04495-3
-Bioinfocodes Scientific News Service-
News articles prepared by our team members, reviewing and compiling scientific research
published in journals with and impact factor greater than 20 (click here for the list)