Biological Age Clocks:
What Epigenetic Methylation Reveals About How Fast You're Aging

The Mechanism: What Is DNA Methylation?

To understand biological age clocks, you first need to understand what they are actually measuring. Every cell in your body carries the same DNA sequence, yet a liver cell behaves completely differently from a neuron. This is because of epigenetics — a layer of chemical tags sitting on top of your DNA that determines which genes are active and which are silenced.

The most well-studied of these tags is DNA methylation: the addition of a methyl group (CH₃) to a cytosine nucleotide, almost always at a location called a CpG site (a cytosine nucleotide followed by a guanine, linked by a phosphate — hence cytosine-phosphate-guanine). Your genome contains roughly 28 million CpG sites, and the methylation pattern across these sites changes in highly predictable ways as you age.

This is the key insight that made epigenetic clocks possible: aging is not random at the methylation level. Certain CpG sites become more methylated with age; others lose methylation. The pattern is so consistent across individuals that, by reading a subset of these sites, scientists can infer your age to within a few years — and, more importantly, determine whether you are aging faster or slower than expected for someone born when you were.

The distinction between chronological age (how many years since birth) and biological age (how old your cells behave) is not merely semantic. Two people born on the same day can have methylation profiles suggesting their bodies differ by a decade or more. The person whose biological age exceeds their chronological age faces substantially elevated risk of cardiovascular disease, cancer, neurodegenerative disease, and all-cause mortality. The person running younger, by contrast, appears protected.

The Major Epigenetic Clocks: A Generation of Discovery

The Horvath Clock (2013) — The Foundation

In 2013, UCLA biostatistician Steve Horvath published what remains the foundational paper in the field. His clock uses 353 CpG sites drawn from methylation data across 51 tissue types and nearly 8,000 samples. The result was a single model capable of estimating chronological age from any tissue in the body with a correlation of R²=0.96 — extraordinary precision for a biological measurement.

The Horvath Clock's power lies in its universality. It works in blood, saliva, brain, liver, skin — making it what researchers call a "multi-tissue clock." Its limitation is that it was trained to predict chronological age, not health or mortality outcomes. A person who has smoked for 30 years might score deceptively young on Horvath while having profoundly accelerated disease risk captured by later clocks.

The Hannum Clock (2013) — Blood-Specific Precision

Published the same year, the Hannum Clock uses just 71 CpG sites but restricts itself to blood samples. Because it is tissue-specific, it achieves comparable predictive accuracy for blood-derived measurements and correlates with multiple hallmarks of aging including telomere length and gene expression patterns associated with immune aging. Hannum is still used in research settings, particularly for immune-focused longevity studies.

PhenoAge (Levine, 2018) — The Clinical Upgrade

Morgan Levine's PhenoAge represented a conceptual leap. Rather than training on chronological age, Levine first constructed a composite "phenotypic age" from 9 clinical biomarkers — albumin, creatinine, glucose, C-reactive protein, lymphocyte percentage, mean red cell volume, red cell distribution width, alkaline phosphatase, and white blood cell count — then combined this with chronological age. She then trained an epigenetic clock to predict this phenotypic age composite.

The result is a clock that is a significantly better predictor of mortality, cancer, and physical functioning than earlier clocks. When your PhenoAge biological age is higher than your chronological age, it signals that your physiological systems are deteriorating faster than the calendar suggests. PhenoAge is the algorithm underlying Elysium's Index test and is widely used in intervention studies.

GrimAge (Lu, 2019) — Trained on Death Itself

GrimAge is the most sobering of the clocks. Developed by Ake Lu and colleagues, it was trained not on age or clinical biomarkers but directly on time-to-death. This makes it the strongest predictor of all-cause mortality, lifespan, healthspan, coronary heart disease, and physical functioning across all epigenetic clocks developed to date.

One of GrimAge's components is an epigenetic surrogate for smoking pack-years — the number of packs smoked per day multiplied by years of smoking. This was included because smoking is the single largest lifestyle accelerant of epigenetic aging, adding an estimated 4-7 years of biological age per pack-year in GrimAge calculations. Even former smokers show residual GrimAge acceleration years after quitting, though this signal does attenuate over time.

If GrimAge exceeds your chronological age, your risk profile is elevated. A GrimAge 5 years older than your chronological age has been associated in prospective studies with significantly higher probability of dying in the next decade.

DunedinPACE (2022) — Measuring the Rate, Not the Position

All previous clocks measure a snapshot: how old do your cells look right now? DunedinPACE measures something different and arguably more actionable: how fast are you aging? Derived from the Dunedin longitudinal birth cohort study (which tracked individuals from birth through their 40s), DunedinPACE expresses aging as a rate — biological years per chronological year.

A DunedinPACE score of 1.0 means you are aging at the average rate. A score of 0.8 means you are aging 0.8 biological years per calendar year — meaningfully slower. A score of 1.2 means you are aging 20% faster than expected. Because DunedinPACE is a rate measure derived from a longitudinal cohort (not just a cross-sectional snapshot), it responds to interventions more sensitively and is increasingly used as the primary endpoint in lifestyle intervention trials.

What Accelerates Epigenetic Aging

The epigenetic clock is not fixed. It is continuously remodeled by the signals your cells receive — from diet, behavior, environment, and stress. Understanding what pushes the clock forward is the first step toward slowing it.

• Smoking — The single largest lifestyle accelerant. GrimAge analysis shows 4-7 additional years of biological age per pack-year of smoking history. Even current light smokers show measurable GrimAge acceleration.
• Obesity — Each unit increase in BMI is associated with approximately 0.3 additional years of PhenoAge. The mechanism involves chronic low-grade inflammation and altered metabolic signaling acting on methylation machinery.
• Physical inactivity — Sedentary individuals consistently show older epigenetic profiles compared to active counterparts, with some cross-sectional data suggesting 2-4 years of PhenoAge difference.
• Chronic psychological stress — Elevated cortisol and sympathetic nervous system activation alter DNA methylation at stress-response gene loci. Caregivers, individuals with PTSD, and those in chronic socioeconomic adversity show accelerated epigenetic clocks.
• Low socioeconomic status — A consistent independent predictor of epigenetic age acceleration, even after adjusting for smoking, diet, and BMI, suggesting that chronic adversity itself encodes biological aging.
• Ultra-processed food diet — Diets high in refined carbohydrates, industrial seed oils, and additives are associated with epigenetic age acceleration, likely through chronic inflammation and oxidative stress.
• Alcohol — Dose-dependent relationship. Heavy drinking (more than 14 drinks/week) consistently associates with older epigenetic clocks. The effect appears to operate primarily through liver epigenetics and systemic inflammation.
• Sleep deprivation — Chronic short sleep (<6 hours) is associated with accelerated epigenetic aging in multiple cohort studies. Total sleep deprivation rapidly alters methylation at circadian clock genes.

What Slows or Reverses Epigenetic Aging

Caloric Restriction — The CALERIE Trial

The most rigorous human evidence for reversing epigenetic aging comes from the CALERIE (Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy) trial. Participants who maintained 25% caloric restriction for 2 years showed a 2.2-year reduction in PhenoAge compared to controls in a methylation substudy. This is not a small effect — it represents a measurable shift in the epigenetic aging trajectory from a dietary intervention alone.

Mediterranean Diet

Adherence to a Mediterranean dietary pattern — rich in olive oil, legumes, vegetables, fish, and moderate red wine — consistently associates with slower epigenetic aging across observational and intervention studies. The combination of anti-inflammatory polyphenols, omega-3 fatty acids, and high fiber appears to act synergistically on methylation patterns.

Omega-3 Fatty Acids

A notable finding emerged from the VITAL (Vitamin D and Omega-3 Trial) ancillary methylation study. Supplementation with EPA and DHA (the active marine omega-3s) was associated with a 3.8-year younger PhenoAge in the African American participant subgroup — a remarkably large effect that warrants replication but points to meaningful anti-inflammatory epigenetic activity from omega-3 supplementation.

Exercise — Especially High-Intensity

Cross-sectional data consistently shows that physically active individuals have younger epigenetic profiles. HIIT (High-Intensity Interval Training) appears particularly potent, with some analyses suggesting 2-3 years of slower PhenoAge in regular HIIT practitioners compared to sedentary controls. The mechanism likely involves mitochondrial biogenesis, reduced systemic inflammation, and NAD+ metabolism — all of which feed back into epigenetic regulation.

Rapamycin — The mTOR Pathway

Rapamycin (sirolimus), an mTOR inhibitor originally developed as an immunosuppressant, has produced the most dramatic lifespan extension results of any compound tested in the Interventions Testing Program (ITP) in mice — up to 26% extension in median lifespan even when started late in life. The mechanism is complex, but mTOR suppression is known to influence epigenetic aging through DNMT (DNA methyltransferase) regulation and autophagy. Human data is limited, but rapamycin's epigenetic effects are an active area of investigation.

NAD+ Restoration: NMN and NR

NAD+ (nicotinamide adenine dinucleotide) declines approximately 50% by middle age, and this decline impairs sirtuins — epigenetic enzymes that regulate DNA methylation and histone modification. NMN (nicotinamide mononucleotide) and NR (nicotinamide riboside) are precursors that restore NAD+ levels. Animal data is consistently promising; human trials are emerging and generally show NAD+ restoration in blood but await definitive methylation endpoints.

Metformin — The TAME Trial

Metformin, a widely prescribed diabetes drug, activates AMPK and has shown epigenetic age-slowing properties in observational studies of diabetic patients who live longer than expected. The TAME (Targeting Aging with Metformin) trial is currently the largest prospective study of any anti-aging drug in humans and will provide definitive evidence within the next few years on whether metformin's epigenetic effects translate to meaningful healthspan extension.

The Frontier: Yamanaka Reprogramming and Reversing Epigenetic Age

Everything discussed so far — diet, exercise, supplements — nudges the epigenetic clock. The emerging field of cellular reprogramming aims to reset it.

In 2006, Shinya Yamanaka discovered that four transcription factors — Oct4, Sox2, Klf4, and c-Myc (collectively OSKM, or "Yamanaka factors") — could reprogram adult cells back into a pluripotent stem cell state. In doing so, these factors erased the epigenetic age of the cell entirely. The problem: full reprogramming destroys the cell's identity. A skin cell forced to fully reprogram becomes, briefly, a blank stem cell — losing its function.

The breakthrough came from partial reprogramming: cycling OSKM factors on and off before full dedifferentiation occurs. This appears to allow epigenetic rejuvenation without loss of cell identity. David Sinclair's lab at Harvard demonstrated in 2020 that partial reprogramming could restore vision in aged mice with optic nerve damage, reversing the epigenetic clock in retinal ganglion cells by 1.5-2.5 years (in mouse equivalent terms) and restoring functional connectivity.

The commercial race to apply this is now funded at a scale unprecedented in longevity science. Altos Labs, backed by Jeff Bezos and other investors with an estimated $3 billion in funding, is pursuing full reprogramming as a therapeutic platform. Alto Neuroscience and several other ventures are targeting tissue-specific partial reprogramming for age-related disease.

Full reprogramming therapeutics remain 10-20 years from clinical application. But the conceptual proof — that biological age is not a one-way ratchet — is now established. This changes the framing of everything else: lifestyle interventions and biomarker tracking are not just about slowing decline, but about bridging to a era when resetting may be possible.

Interpreting Your Results — What the Numbers Mean

When you receive an epigenetic test report, you will typically see several numbers. Here is how to read them:

• Biological age vs. chronological age: A biological age more than 5 years younger than your chronological age represents a meaningful longevity advantage. The effect on mortality risk is non-linear — being 8 years younger biologically is not twice as good as being 4 years younger; it likely compounds significantly.
• GrimAge acceleration: The most critical mortality signal. If your GrimAge is older than your chronological age by more than 3-5 years, prioritize intervention immediately — particularly if you smoke or have metabolic dysfunction.
• DunedinPACE: Your most actionable number. A score above 1.0 means you are aging faster than average. Below 0.8 means your pace of aging is meaningfully slower. Track this number over time in response to interventions.
• PhenoAge: The best predictor of current physiological function. High PhenoAge relative to chronological age often reflects inflammatory burden, metabolic dysfunction, or both — and tends to respond well to diet and exercise interventions.