Biological Age Testing: Which Epigenetic Clocks Are Worth Using
Your chronological age is the number of years you have been alive. Your biological age is the age your cells and tissues actually function at - and the two can diverge by decades. Epigenetic clocks estimate biological age from a blood sample. They are now sold directly to consumers, but they remain research tools rather than validated clinical tests.
- Epigenetic clocks measure biological age from patterns of DNA methylation at specific CpG sites. The deviation between estimated biological age and chronological age - epigenetic age acceleration - is the number these tests report; it is a research measure, not a clinical one.
- First generation clocks (Horvath, Hannum) predicted chronological age accurately. Second generation clocks (PhenoAge, GrimAge) were trained to predict mortality and age-related disease - making them better predictors of health outcomes in research cohorts.
- DunedinPACE measures the pace of aging rather than a static age estimate - which may make it better at detecting change over time. The CALERIE caloric restriction trial found DunedinPACE significantly slowed with CR while other clocks did not detect the effect.
- GrimAge strongly predicted time to death, coronary heart disease and cancer in large validation cohorts. A GrimAge well above your actual age signals higher risk on average; it is not a diagnosis.
- Consumer tests are sold by several companies and often report GrimAge and DunedinPACE. Results can differ between labs and between repeat samples, so single readings should be interpreted cautiously. In one analysis, technical noise alone produced differences of up to 9 years between repeat measurements of the same sample on six well-known clocks; versions rebuilt from principal components agreed within 1.5 years for most repeats.7
The idea that biological age might diverge significantly from chronological age has been a central intuition of aging science for decades. What changed in 2013 was the discovery by Steve Horvath at UCLA that the pattern of DNA methylation marks across the genome predicts chronological age with high accuracy - so accurately that a blood sample could be used to estimate a person's age to within a few years without any other information.1 Later clocks showed that people whose methylation looks older than their age tend to have higher mortality.4
How Epigenetic Clocks Work
DNA methylation is the addition of a methyl group to cytosine nucleotides at CpG sites. Methylation patterns change systematically with age across hundreds of CpG sites - some sites becoming more methylated, others less - in a pattern highly conserved across individuals and tissues. An epigenetic clock is a machine learning algorithm trained on methylation data from thousands of individuals of known ages, producing an estimated biological age from methylation levels at a specific set of CpG sites.2
The Clock Generations
First generation (Horvath and Hannum): trained to predict chronological age; Horvath's multi-tissue clock was accurate to a median of 3.6 years.1,3 Second generation (PhenoAge and GrimAge): trained on markers of health and mortality. GrimAge, built from DNA methylation estimates of plasma proteins and smoking history, predicted time to death, coronary heart disease and cancer in large validation cohorts.4 Third generation (DunedinPACE): estimates the pace of aging rather than a static age; a value of 1.0 means one year of biological aging per calendar year, and 0.8 means 20 percent slower.5 In the CALERIE trial, two years of calorie restriction slowed DunedinPACE while other clocks did not change significantly.6
What Accelerates and Slows Biological Age
Smoking raises GrimAge, which is built partly from a DNA methylation estimate of smoking pack-years.4 Two years of calorie restriction slowed DunedinPACE in a randomized trial.6 Links with exercise, diet and sleep come mainly from observational studies, and the size of the effect varies between clocks and cohorts.
Put this research into practice: Biological Age Calculator · What Should I Test Next?
References
- 1Horvath S. "DNA methylation age of human tissues and cell types." Genome Biol. 2013;14(10):R115. PubMed · DOI
- 2Jones PA. "Functions of DNA methylation: islands, start sites, gene bodies and beyond." Nat Rev Genet. 2012;13(7):484-92. PubMed · DOI
- 3Hannum G, et al. "Genome-wide methylation profiles reveal quantitative views of human aging rates." Mol Cell. 2013;49(2):359-367. PubMed · DOI
- 4Lu AT, et al. "DNA methylation GrimAge strongly predicts lifespan and healthspan." Aging (Albany NY). 2019;11(2):303-327. PubMed · DOI
- 5Belsky DW, et al. "DunedinPACE, a DNA methylation biomarker of the pace of aging." Elife. 2022;11. PubMed · DOI
- 6Waziry R, et al. "Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial." Nat Aging. 2023;3(3):248-257. PubMed · DOI
- 7Higgins-Chen AT, et al. "A computational solution for bolstering reliability of epigenetic clocks: Implications for clinical trials and longitudinal tracking." Nat Aging. 2022;2(7):644-661. PubMed · DOI
