Start a live chat
1-800-870-4800
Your epigenome
is not fixed — it is
a living record of your life.
The genome encodes your DNA sequence. The epigenome records everything that has happened since. Every year that passes, every environmental exposure, every metabolic pattern — these leave measurable molecular traces on the epigenome in the form of chemical marks on DNA and histones. Scientists can now read those marks and estimate biological age with striking accuracy. That system of marks is regulated, in part, by enzymes that require NAD+ to function.
I
The difference between
the genome and the epigenome.
The genome is static. Barring mutation, the DNA sequence you were born with is the DNA sequence you carry throughout your life — the same 3 billion base pairs in virtually every cell, unchanged by experience or time. The epigenome is not static. It is a dynamic layer of chemical modifications that sits on top of the DNA sequence and changes continuously in response to cellular signals, environmental exposures, metabolic conditions, and the passage of time itself.
These epigenetic modifications — primarily DNA methylation (the addition of methyl groups to cytosine bases at CpG sites) and histone modifications (acetylation, methylation, and other chemical changes to the protein spools around which DNA is wound) — do not change the underlying DNA sequence. They change how the cell reads it. A methylated gene promoter is silenced; the gene's information is still there, but the transcription machinery cannot access it. An acetylated histone loosens the chromatin in its region, making the nearby genes more accessible. The epigenome is, in this sense, a layer of annotation on top of the genome — one that determines which parts of the fixed instruction set are currently active.
What makes the epigenome particularly remarkable from a longevity biology perspective is that it changes with age in predictable, measurable ways. The pattern of DNA methylation across the genome shifts systematically as a cell ages — some sites gain methylation, others lose it — in patterns that are consistent enough across individuals that they can be used to estimate biological age. This insight forms the basis of epigenetic clocks: mathematical models trained on methylation data that predict age from the epigenetic state of a cell with an accuracy that often exceeds other biological markers. The epigenome is not just a regulatory system — it is also a clock.
The genome is the instruction set.
The epigenome is the record
of how those instructions
have been read, modified,
and responded to
across an entire lifetime.
Epigenetic Clocks
Three generations of biological age clocks
built from DNA methylation patterns.
Epigenetic clocks are among the most precise biological age estimators available. Each generation has refined the measurement — from calendar age prediction to mortality risk to what some researchers call the first-generation clocks specifically of biological aging.
First generation · 2013
The Horvath Clock — DNA methylation predicts chronological age
In 2013, researcher Steve Horvath published a mathematical model trained on DNA methylation data from multiple tissue types that could predict chronological age with a median error of approximately 3.6 years. The model used 353 CpG sites — specific locations in the genome where methylation patterns changed consistently with age across tissues as diverse as blood, brain, and skin. The Horvath Clock was the first demonstration that aging leaves a consistent, readable molecular signature in the epigenome — one that is distinct enough to function as a biological age estimator independent of other biomarkers. The clock is pan-tissue: the same 353 sites predict age regardless of which cell type is sampled.
Second generation · 2015–2018
The Hannum and PhenoAge Clocks — from age prediction to health prediction
Second-generation clocks moved beyond chronological age prediction toward mortality and health outcome prediction. The Hannum Clock (2013) was blood-specific and correlated methylation age acceleration with biological aging markers. The GrimAge clock and PhenoAge (both published 2018) were trained not on chronological age but on health outcomes and mortality risk — making them more directly relevant to the concept of biological age as distinct from calendar age. These clocks revealed that individuals of the same chronological age can have substantially different epigenetic ages, and that the gap between chronological and epigenetic age — called "age acceleration" — is associated with measurable differences in health trajectories.
Third generation · 2022–present
DunedinPACE and next-generation clocks — measuring the rate of aging itself
Third-generation clocks represent a conceptual shift: rather than estimating a static biological age, they estimate the pace of aging — how fast the epigenome is currently changing. DunedinPACE (Pace of Aging Computed from the Epigenome) was trained on longitudinal data from participants followed over decades, measuring the rate of change in multiple aging-related biomarkers simultaneously. The result is a measure of biological aging velocity — not where you are on the aging trajectory, but how fast you are moving along it. This distinction has significant implications for intervention research: a clock that measures pace is more sensitive to change than one that measures position.
II
What writes — and rewrites —
the epigenome across a lifetime.
The epigenome is not written once at birth and then read passively for the rest of life. It is continuously written, erased, and rewritten by a molecular machinery that responds to signals from inside and outside the cell. DNA methyltransferases (DNMTs) add methyl groups to cytosine bases; TET enzymes remove them through active demethylation. Histone acetyltransferases (HATs) add acetyl groups to histone tails; histone deacetylases (HDACs) — including the sirtuin family — remove them. The balance of these opposing activities determines the epigenetic state of any given genomic region at any given moment.
This continuous rewriting is why the epigenome is sensitive to so many different biological inputs. Nutritional signals, hormonal changes, metabolic states, cellular stress responses, and the products of specific signaling pathways all feed into the enzymatic machinery that modifies epigenetic marks. The cell does not experience these inputs in isolation — they are integrated through the epigenome into changes in gene expression that reflect the cell's current biological context. A cell under sustained metabolic stress will have a different epigenetic state than a cell operating under optimal conditions — not necessarily because its DNA sequence has changed, but because the signals that regulate its epigenetic writers and erasers have changed.
Among the erasers of histone acetyl marks — the enzymes that tighten chromatin and regulate gene silencing — the sirtuin family is particularly noteworthy for a reason covered in the previous article: every deacetylation reaction consumes one molecule of NAD+. This means the activity of the sirtuin-family epigenetic erasers is directly linked to the cellular NAD+ pool. A cell with an abundant NAD+ pool can run its sirtuin-mediated epigenetic maintenance at full capacity. A cell with a depleted NAD+ pool — whether from aging, chronic demand, or inadequate precursor supply — cannot. The epigenome is, in this specific sense, sensitive to the NAD+ status of the cell that maintains it.
What Shapes the Epigenome
Five categories of input that the research
literature associates with epigenetic change.
These are documented associations in the epigenetic research literature — inputs whose relationship to epigenetic marks has been characterized in human and animal studies. They are biological observations, not outcome claims.
The most consistent driver of epigenetic change is simply the passage of time. The DNA methylation patterns that epigenetic clocks measure shift in predictable ways as cells age — some CpG sites gaining methylation, others losing it, across tissues and individuals. This age-related epigenetic drift is not fully understood mechanistically, but is thought to involve the gradual loss of fidelity in the epigenetic maintenance machinery — the enzymes that copy epigenetic marks when cells divide and that correct aberrant marks when they accumulate. SIRT1 and SIRT6, both NAD+-dependent, have documented roles in maintaining the fidelity of specific epigenetic patterns that drift with age.
The cell's metabolic state — its nutrient availability, energy balance, and the ratio of key metabolites — feeds directly into the epigenetic machinery. Many of the substrates and cofactors used by epigenetic writers and erasers are metabolic intermediates: acetyl-CoA (the acetyl group donor for histone acetyltransferases) comes from glucose and fatty acid metabolism; SAM (the methyl group donor for DNA and histone methyltransferases) comes from the methionine cycle; and NAD+ (the cofactor for sirtuin deacetylases) comes from the Salvage Pathway. This means the epigenome is metabolically sensitive — its state reflects, in part, the metabolic history of the cell that maintains it.
The circadian clock — the 24-hour oscillator that governs the timing of cellular processes — has direct connections to the epigenome. CLOCK and BMAL1, the core circadian transcription factors, regulate the acetylation of H3K9 and H3K14 at circadian gene promoters, with CLOCK itself having intrinsic histone acetyltransferase activity. SIRT1 provides the corresponding deacetylase activity in a NAD+-dependent manner — meaning the circadian epigenetic cycle has a built-in NAD+ dependency at its core. NAMPT — the rate-limiting enzyme of the Salvage Pathway — is itself circadian regulated, creating a feedback loop between the circadian clock, NAD+ production, SIRT1 activity, and the epigenetic state of circadian genes.
Inflammatory signaling — through NF-κB and other transcription factors — has documented effects on epigenetic marks at inflammatory gene promoters. SIRT6 suppresses NF-κB target gene expression through H3K9ac deacetylation; when SIRT6 activity is reduced (whether from NAD+ depletion or reduced SIRT6 expression), those target genes become more accessible, and inflammatory signaling is amplified. Chronic low-grade inflammation — which tends to rise with age through the process of inflammaging — therefore has an epigenetic dimension: sustained NF-κB activation can shift the epigenetic state of inflammatory gene promoters in ways that perpetuate the inflammatory response. This is one of the specific molecular mechanisms connecting inflammation, epigenetics, and the aging process.
When DNA is damaged — by oxidative stress, UV radiation, or metabolic errors — the DNA damage response recruits epigenetic regulators to the break site. SIRT1 relocates from its normal genomic positions to DNA break sites, where it participates in repair coordination. This relocalization is thought to cause epigenetic drift at SIRT1's normal positions — as SIRT1 leaves those sites to respond to damage elsewhere, the histone acetylation patterns it normally maintains begin to spread. This "epigenetic relocalization" model — where the diversion of epigenetic regulatory enzymes to damage response causes drift at their normal regulatory positions — is a proposed mechanism for how accumulated DNA damage contributes to age-related epigenetic change.
The Epigenome in Numbers
What the epigenetic aging story
looks like as measurable facts.
353
CpG methylation sites used by the original Horvath Clock to predict biological age — from a genome of 28 million total CpG sites
The Horvath Clock uses 353 of the approximately 28 million CpG sites in the human genome to predict biological age across tissue types with a median error of roughly 3.6 years. The fact that 353 sites — out of 28 million — carry enough information to estimate age with that precision is a remarkable demonstration of how systematically aging writes itself into the epigenome. The sites were identified by training machine learning models on methylation data from thousands of samples across dozens of tissue types, then selecting the sites that predicted age most consistently across all of them.
~3.6
Years — median error of the Horvath Clock's biological age estimate from DNA methylation data alone
A median error of approximately 3.6 years means that the epigenetic clock's estimate of biological age falls within 3.6 years of the actual chronological age in half of all cases — and this is using only methylation data, with no other biological information. By comparison, most other biological age estimators perform substantially less precisely. The clock's accuracy reflects how reliably the epigenome records the passage of time — not as a perfect timer, but as a consistent pattern of change that machine learning models can read and translate into an age estimate.
1
NAD+ molecule consumed per sirtuin deacetylation event — the biochemical cost of each epigenetic mark the sirtuin family removes from the genome
The connection between NAD+ and the epigenome runs through this stoichiometry: one NAD+ consumed per acetyl group removed from a histone by a sirtuin. SIRT1 and SIRT6 — the two nuclear sirtuins most directly involved in epigenetic regulation and the maintenance of epigenetic age patterns — both pay this cost for every regulatory event they perform. The adequacy of the nuclear NAD+ pool, maintained by NMNAT1, therefore directly governs how actively these sirtuins can maintain the epigenetic patterns whose erosion is a documented feature of biological aging. Studies were conducted independently and did not involve any specific Codeage product.
III
The epigenome, NAD+,
and what the connection means.
The relationship between epigenetics and longevity biology is one of the most active areas of current research — and one of the most intellectually rich. The epigenome is simultaneously a regulatory system (controlling which genes are expressed in which cells), a record (encoding the history of a cell's exposures and experiences as molecular marks), and a clock (whose age-related drift can be measured and used to estimate biological age). These three functions are not separate — they are aspects of the same molecular system.
NAD+ enters this picture through the sirtuin family: the histone deacetylases that are among the primary epigenetic erasers in the nucleus. Their NAD+ dependency is not incidental — it is the mechanism by which the cell's metabolic state is coupled to its epigenetic regulatory capacity. When the NAD+ pool is adequate, sirtuins can maintain the epigenetic patterns that define cellular identity and function. When the NAD+ pool is depleted — whether from the NAMPT decline of aging, from CD38-driven consumption by inflammation, or from other sources of NAD+ demand — the sirtuin-mediated epigenetic maintenance runs at reduced capacity. Whether this reduced capacity contributes to the epigenetic drift documented in aging tissue is an active area of research, and the connections described here reflect current understanding in a field that continues to develop rapidly.
For the specific sirtuin biology and which enzymes govern which aspects of the epigenome, the genome article covers the four nuclear sirtuins in depth. For the NAD+ system that supplies their cofactor, the NAMPT article covers the Salvage Pathway's rate-limiting step. Both connect to Cellular Longevity — Pillar 03 of The Longevity Code.
The epigenome is simultaneously
a regulatory system,
a record, and a clock.
These are not separate functions —
they are aspects of
the same molecular system.
Codeage · Pillar 03 · Cellular Longevity
Built for the
cellular long game.
Cellular Longevity is Pillar 03 of The Longevity Code — the dimension of the system built around NAD+ biology, mitochondrial health, and the science of cellular aging.
Explore Cellular Longevity →
