Longevity & Aging
epigenetic clocks correlates with biological age
In plain terms: Can DNA "clocks" measure how fast you're really ageing?
Strong support Longevity & Aging 🔬 Includes disconfirming
Part of: • Aging
RefutedContestedStrong support
consensus score 0.72
Probably modestly yes for predicting age, but it's only a correlation, not proof they track true biological ageing.
📅 Last reviewed: 2026-07-14 ⓘ
Evidence ladder
How far up the ladder this claim has climbed. A high consensus on a low rung means "consistent so far," not "proven in people."
Top evidence so far: Population patterns (Observational)
MechanismIn-vitroAnimalObservationalRCTMeta-analysis
How the studies fall
13 support 1 contradict 0 tested null 7 mixed · 21 sources, 14 independent groups
What the evidence shows
DNA-methylation clocks (Horvath 2013) predict chronological age accurately, but they are CORRELATIVE biomarkers - clock movement does NOT equal proven rejuvenation or lifespan gain (Teschendorff 2025; clock signal is partly stochastic drift).
The evidence (21)
| Source | Grade | Stance | Quality | Finding |
|---|---|---|---|---|
| Li 2025 · EBioMedicine | observational | supports | moderate | CpG-site clocks limited in biological interpretability; pathway-level clock (PathwayAge) needed for disease-specific aging mechanisms |
| Teschendorff 2025 · Nat Rev Genet | observational | mixed | high | Review: clocks are robust age predictors but clock change does NOT equal proven rejuvenation or lifespan benefit |
| Ashapkin 2019 · Adv Exp Med Biol | mechanism | mixed | moderate | Review questions whether clock is convenient marker vs active driver of aging; much age-methylation is stochastic at repeats |
| Ibanez-Cabellos 2026 · Biogerontology | observational | supports | low | Review: clocks robust age estimators but cautions on causal interpretation and clinical readiness for personalized aging |
| Li 2021 · EBioMedicine | observational | supports | high | ESTHER cohort: GrimAge/PhenoAge acceleration independently predict all-cause mortality, validating clocks as biological-age biomarkers |
| Vetter VM 2026 · Commun Med (Lond) | observational | supports | high | In 1,083 Berlin Aging Study II participants (7.4-year follow-up), Allostatic Load Index and DunedinPACE showed the strongest, most consistent associations with age-related phenotypes (frailty, cognition, morbidity) among 16 biological age markers, improving prediction of incident frailty/metabolic s |
| Arpawong TE 2026 · Geroscience | observational | supports | high | PhysAge, a DNAm-based multi-system biomarker validated in HRS (n=3177) plus two independent cohorts (TILDA, NICOLA), was comparable to PhenoAge/GrimAge2/DunedinPACE in predicting mortality and functional health measures (grip strength, gait speed, frailty). |
| Bertucci-Richter 2024 · Aging (Albany) | mechanism | mixed | moderate | Clock signal driven by stochastic methylation disorder; clocks & true aging processes are decoupled - caution against treating clock as causal readout |
| Mitteldorf 2025 · Aging (Albany NY) | observational | mixed | low | Argues methylation clocks unreliable for evaluating anti-aging interventions; methylation under selection, clock change need not mean rejuvenation |
| Ying 2025 · Nature Aging | observational | contradicts | high | Curation framework: ability to predict chronological age does NOT correlate with predicting mortality/outcomes; decouples prediction from biological aging |
| Horvath 2013 · Genome Biol | observational | supports | high | Multi-tissue 353-CpG clock predicts chronological age accurately across human tissues - landmark biomarker, but correlation not causation ⚠️ correction-on-file (Crossref) - kept, corrigendum not retraction |
| Levine 2018 · Aging (Albany NY) | observational | supports | high | 2nd-gen DNAmPhenoAge built on clinical phenotype outperforms 1st-gen clocks at predicting mortality/healthspan across cohorts |
| Tan 2026 · Clinical Epigenetics | observational | mixed | moderate | Later-gen clocks (GrimAge/PhenoAge) predict breast-cancer survival better than Horvath/Hannum; first-gen age-predictors weakly prognostic |
| Jiang Y 2026 · Brief Bioinform | observational | mixed | moderate | StackAge, an ensemble multi-omics clock trained on 30,376 UK Biobank participants, predicted chronological age (Pearson r≈0.93) and achieved AUC>0.90 for predicting 12 chronic diseases including type 2 diabetes and Alzheimer's. |
| Yusupov N 2026 · Biol Psychiatry Glob Open Sci | observational | mixed | moderate | Systematic review of 57 studies (37,516 cross-sectional + 15,551 longitudinal) across 16 epigenetic clocks: epigenetic age acceleration consistently associated with worse cognitive performance/decline — broad validation that clocks track an aging-relevant biological signal. |
| Hannum 2013 · Molecular Cell | observational | supports | high | Landmark blood methylome model predicts chronological age (~96% correlation, ~3.9yr error); established methylation-age clock concept |
| An T 2025 · Geroscience | observational | supports | high | NHANES adults (n=2,532; 17-yr follow-up, 271 cancer deaths): GrimAge/GrimAge2 mortality-acceleration positively and independently associated with mortality — large prospective evidence clocks predict biological aging outcomes. |
| Chen Q 2026 · Nat Aging | observational | supports | high | OMICmAge (DNA-methylation + multi-omics + EMR) discovery n=3,451, validated in independent cohorts: accurately quantifies biological age and predicts health outcomes — a rigorously validated next-gen clock. |
| Meng X 2026 · Cancer Epidemiol Biomarkers Prev | observational | supports | high | Among 2,941 childhood cancer survivors, DNAm-based biological age acceleration (43 biomarkers) was elevated after high-dose cardiotoxic treatment and mediated up to 25.7-35.9% of treatment-cardiotoxicity associations; PCGrimAge was most strongly associated with myocardial infarction (OR=1.85). |
| Sosnowski DW 2026 · J Gerontol A Biol Sci Med Sci | observational | supports | high | In 396 ALIVE cohort participants (3,862 person-years), PhenoAge acceleration was associated with 3.28x (with HIV) and 2.12x (without HIV) higher all-cause mortality hazard; HIV and epigenetic age acceleration were independently associated with mortality. |
| Liu F 2026 · Clin Epigenetics | observational | supports | moderate | A targeted 74-CpG methylation clock for Chinese populations achieved R2=0.93-0.95 (MAE 3.4-4.0 years) predicting chronological age across training/test/external validation cohorts (n=610, n=188), and distinguished cancer-associated methylation drift from physiological aging. |
Disagree, or know a study we missed?
We grade by evidence, not opinions. The way to weigh in is to point us to a study we haven't cited (check the evidence table above first), or to flag a problem with one we have. Every submission is reviewed; if it holds up, the grade updates and shows in Science Changes Its Mind.
Opens a short form. You'll sign in with Google so submissions are tied to a real account — we don't display your identity, and we only accept a link we can verify (PubMed, DOI, ClinicalTrials.gov).
Educational only, not medical advice. Grades and scores reflect published evidence weighted by study design and quality; see the methodology.