Epigenetic clocks are statistical and machine learning models that estimate biological age from DNA methylation data. Since their introduction in 2013, they have become the most widely used molecular tools for measuring biological aging in both research and clinical settings.
Epigenetic clocks do not merely measure how old you are — they measure how fast you are aging, and what that rate predicts for your future health.
First-Generation Clocks: Tissue Age Estimation
The first generation of epigenetic clocks was developed to estimate chronological age from methylation data — essentially asking: "Given this methylation pattern, how old is this tissue?"
Horvath Clock (2013)
Developed by Steve Horvath at UCLA, the Horvath clock uses 353 CpG sites and was trained on 51 tissue and cell types. It remains one of the most broadly applicable clocks, capable of estimating age across diverse tissue types with a median absolute deviation of approximately 3.6 years. It was the first to demonstrate that epigenetic age acceleration is associated with cancer risk and all-cause mortality.
Hannum Clock (2013)
Developed by Gregory Hannum and colleagues, this clock uses 71 CpG sites and was trained specifically on blood samples. It shows strong correlation with chronological age in blood-based assays and was among the first to demonstrate that lifestyle factors such as BMI and smoking influence epigenetic age acceleration.
Second-Generation Clocks: Health and Mortality Prediction
Second-generation clocks were designed not just to estimate age, but to predict health outcomes — mortality risk, disease incidence, and functional decline. These clocks are more clinically relevant because they capture biological aging processes that directly affect patient outcomes.
PhenoAge
Developed by Morgan Levine and colleagues, PhenoAge was trained to predict "phenotypic age" — a composite of clinical biomarkers associated with mortality. It uses 513 CpG sites and has been shown to predict all-cause mortality, cancer incidence, and healthspan outcomes more accurately than first-generation clocks.
GrimAge
GrimAge, developed by Ake Lu and Steve Horvath, is currently considered the strongest predictor of lifespan among all epigenetic clocks. It uses a composite of plasma protein proxies derived from methylation data and has been validated in multiple large cohort studies as a predictor of time-to-death, cardiovascular disease, and cancer.
DunedinPACE
DunedinPACE (Pace of Aging Computed from the Epigenome) is a newer clock that measures the rate of aging rather than a static age estimate. Developed from the Dunedin longitudinal cohort, it tracks how quickly an individual is aging at the time of measurement — making it particularly useful for monitoring the effects of interventions over time.
How Clocks Are Used Clinically
- Baseline biological age assessment — establishing a patient's aging trajectory at enrollment
- Intervention monitoring — tracking epigenetic age changes in response to treatment
- Risk stratification — identifying patients with accelerated aging for prioritized intervention
- Research endpoints — using epigenetic age as a primary or secondary outcome in clinical trials
XELGEN incorporates epigenetic biomarker analysis derived from DNA methylation research to estimate biological age and evaluate molecular indicators of aging. The platform integrates multiple clock algorithms — including first and second-generation models — to provide clinicians with a comprehensive view of a patient's aging trajectory and enable longitudinal monitoring of intervention effects.
Explore how XELGEN integrates epigenetic clock science into methylation testingWhat is an epigenetic clock?
An epigenetic clock is a computational model that estimates biological age by analyzing DNA methylation patterns across selected CpG sites in the genome. Different clocks are optimized for different outcomes — from tissue age estimation to mortality prediction.
